Identification of canine glial cells by nonradioactive in situ hybridization

Studies on the development of the canine central nervous system and on demyelinating diseases demand unequivocal identification of the glial cells. For that reason, nonradioactive in situ hybridization (ISH) was performed in primary dog brain cell cultures (DBCC) and in brain sections of neonatal dogs. Specific RNA probes were used to detect messenger RNA (mRNA) coding for proteolipid protein (PLP), myelin basic protein (MBP) and glial fibrillary acidic protein (GFAP). PLP and MBP are markers for oligodendrocytes, GFAP for astrocytes. Oligodendrocytes positive for PLP and MBP mRNA were found in both DBCC and brain sections of neonatal dogs. Astrocytes expressing GFAP specific mRNA were detected in DBCC and in brain sections. These cells were evenly distributed in the white matter with additional accumulation in the membrana limitans gliae superficialis, around the ventricles and blood vessels. ISH clearly improves the study of oligodendrocytes in brain sections as, in contrast to the immunohistochemical methods, this technique allows to identify individual cells.

In situ hybridization of virulent canine distemper virus in brain tissue, using digoxigenin-labeled probes

Only a few hybridization experiments have been performed for detection of canine distemper virus (CDV) nucleic acid sequences in tissue cultures and in various tissues. Those published studies used probes derived from tissue culture-adapted CDV, and hybridization signals were not obtained in the CNS tissue, although infective CDV and viral antigen were detectable in this tissue. We developed probes complementary to virulent CDV and were able to detect viral RNA not only in primary brain cell cultures, but also in brain tissues, by use of in situ hybridization. Sensitivity of the test at least equaled that of immunohistochemistry. We applied digoxigenin-labeled, strand-specific RNA probes complementary to the nucleoprotein-coding viral nucleic acid sequence. Our results indicate that to detect CDV nucleic acid sequences in brain tissues, it is essential to use probes derived from the virulent virus.

Restricted canine distemper virus infection of oligodendrocytes

BACKGROUND

Canine distemper virus, a morbillivirus induces multifocal demyelination in the central nervous system. The acute demyelination correlates with virus replication in brain cells, especially astrocytes. Observations in vivo and in vitro demonstrated degeneration of oligodendrocytes, the myelin producing cells. However, the mechanism of oligodendroglial degeneration in distemper remained unexplained. Infection of the myelin producing cells, the most obvious explanation for the phenomenon of demyelination, could not be supported by extensive searches for viral particles or antigens in these cells neither in vivo nor in vitro.

EXPERIMENTAL DESIGN

In the present study, we combined in situ hybridization to visualize viral nucleic acid sequences with immunofluorescence for oligodendroglial antigens.

RESULTS

The nonradioactive in situ hybridization technique in combination with contrast enhanced video microscopy allowed us to unequivocally demonstrate the presence of canine distemper virus nucleic acid sequences in cultured oligodendrocytes. Many oligodendrocytes close to infected foci in the brain cell cultures were found to contain viral nucleic acid sequences. Only 1% of the viral nucleic acid sequences containing oligodendrocytes also contained viral antigen. Canine distemper virus replication in these cells is clearly restricted.

CONCLUSIONS

Different possibilities why oligodendrocytes do not support a productive virus infection and mechanisms by which such a restricted infection leads to oligodendroglial degeneration and ensuing demyelination are discussed. While our results have advanced our understanding of the pathogenesis of acute demyelination in distemper, they may also offer a possible explanation for the chronic progressive or even relapsing course of the disease. A restricted infection of the oligodendrocytes may be the mechanism by which canine distemper virus persists in the central nervous system. Virus persistence is probably a key event in many chronic viral induced inflammatory demyelinating diseases.

Saponin treatment for in situ hybridization maintains good morphological preservation

The commonly used procedures for in situ hybridization require treatment of the tissue with non-ionic detergents and proteolytic enzymes, resulting in considerable loss of morphological detail. In this study proteinase pre-treatment of the tissue was replaced by saponin, a highly surface-active plant glycoside. This saponin treatment allowed good preservation of tissue morphology, as determined by differential interference and contrast enhanced video microscopy. Saponin pre-treatment resulted in an equal or even better hybridization sensitivity with probes recognizing viral (canine distemper virus) and cellular (myelin) nucleic acid sequences in tissue cultures as well as in paraffin sections. Probable mechanisms of how saponin allows probe penetration while maintaining the morphological details are discussed.

Canine distemper virus increases procoagulant activity of macrophages

Inflammatory demyelination in canine distemper has been proposed to be due to a "bystander" mechanism, in which macrophages play an important role. In the present work we studied whether infection of macrophages by canine distemper virus (CDV) results in changes of macrophage functions, including Fc receptor-dependent and -independent phagocytosis, release of reactive oxygen species (ROS), and procoagulant activity (PCA). As a source of macrophages, dog bone marrow cells were seeded in teflon bags and grown for 1-2 weeks, at which time a marked enrichment of macrophages was noted. These cells were infected with the A75/17 strain of CDV. We could not detect any significant difference between uninfected and CDV-infected macrophages with respect to Fc receptor-dependent or -independent phagocytosis or with respect to the release of ROS. However, from Day 4 p.i. to the end of our observation period (10 days p.i.), PCA was up to 10-fold higher in CDV-infected unstimulated macrophage cultures than in uninfected unstimulated cultures of the same age. Increase in PCA was not due to the inoculation procedure by itself nor to components of the inoculum other than CDV; in particular, PCA was not due to contaminating endotoxin. Thus, several important macrophage functions do not appear to be impaired by CDV infection. The marked increase of macrophage PCA expression suggests that certain macrophage functions may even be enhanced as a result of infection. Such macrophage activation might contribute to the pathogenesis of the disease.

[Spongiform encephalopathies with special reference to bovine spongiform encephalopathy]

In Switzerland bovine spongiform encephalopathy (BSE) was detected for the first time in November 1990. It is a transmissible disease of the central nervous system similar to Creutzfeldt-Jakob disease (CJD), Gerstmann-Sträussler-Scheinker syndrome (GSS) and kuru in man, and, in animals, scrapie in sheep and goats, chronic wasting disease (CWD) in captive mule deer and elk of North America and transmissible mink encephalopathy (TME) of farm reared mink. The infectious agent of the spongiform encephalopathies (SEs) is extremely small (greater than 100 microns) and very resistant against physical and chemical disinfectants. The infectiousness is closely linked to the so-called prion proteins, the biological and genetic properties of which are not yet fully understood. The first outbreak of BSE occurred in England, and the disease spread rapidly all over Great Britain. The source of infection proved to be bone and meat meal obtained from scrapie-infected slaughter wastes. So far we have diagnosed 17 cases of BSE in our country. The risk of humans contracting CJD by ingestion of meat, milk and their products contaminated with BSE is probably minimal.

Canine distemper virus-immune complexes induce bystander degeneration of oligodendrocytes

Demyelination in chronic canine distemper encephalitis may be the result of a bystander effect in which the antiviral immune response is involved. In the present report we demonstrate that canine distemper virus-antiviral antibody immune complexes induce oligodendroglial degeneration in mixed brain cell cultures, particularly at the level of the cell processes. The involvement of macrophages as effector cells in this process was confirmed by depletion of these cells from the cultures which prevented the immune complex-mediated oligodendroglial degeneration. Canine distemper virus-immune complex-induced oligodendroglial pathology is thought to be mediated by toxic factors released from stimulated macrophages, this bystander effect demonstrated here in vitro may be relevant to the mechanisms of demyelination in vivo, in which virus persistence plays an important role.

Loss of virulence of canine distemper virus is associated with a structural change recognized by a monoclonal antibody

The monoclonal antibody (mAB) L1, which binds to the nucleocapsid protein of canine distemper virus (CDV), was shown to bind to avirulent CDV obtained after serial passages in Vero cells, but not to two different virulent demyelinating CDV-strains propagated in dog glial cell cultures. However, when both virulent CDV-strains were passaged through Vero cells they expressed, after a number of passages, an epitope recognized by mAB L1. The occurrence of the L1 epitope appeared to coincide with loss of virulence in animal inoculation experiments.

Direct evidence of a role for amino acid 101 of VP-1 in central nervous system disease in Theiler's murine encephalomyelitis virus infection

The DA virus, a member of the TO subgroup of Theiler's virus, invokes a chronic demyelinating disease in its natural host, the mouse, RNA transcripts from a cDNA clone, pDAFL3, are infectious, and the resulting virus, DAFL3, produces in mice a disease indistinguishable from that caused by the DA virus. Using oligonucleotide-directed site-specific mutagenesis, a single nucleotide, cytosine at position 3305 (viral genome), was changed in this infectious cDNA to a thymine. The mutated nucleotide is located in an area coding for a neutralizing epitope on loop II of VP-1. Virus OSM101, produced from the mutagenized plasmid pDA101, had the same growth characteristics and plaque phenotype in vitro as the virus DAFL3 produced from clone pDAFL3. However, in vivo in the mouse, virus OSM101 was markedly less neurovirulent than DAFL3. Central nervous system tissues from mice infected 4 to 6 weeks previously with the OSM101 virus contained less infectious virus and fewer infected cells than central nervous system tissues from animals infected with the control virus, DAFL3. Thus, we demonstrated that the single nucleotide change resulting in an amino acid substitution at position 101 (threonine to isoleucine) of VP-1 determines one aspect of Theiler's virus persistence and disease in mice.

Common immunologic determinant between human immunodeficiency virus type 1 gp41 and astrocytes

Monoclonal antibodies against a synthetic 12-amino-acid peptide that comprises the immunodominant domain of human immunodeficiency virus type 1 gp41 (amino acids 598 through 609) reacted with astrocytes found in human and rodent central nervous system tissue. The monoclonal antibodies bound to a 43-kDa protein found in central nervous system tissue preparations. These results indicate that human immunodeficiency virus type 1 gp41 contains a common epitope with astrocytes and that an immune response to human immunodeficiency virus type 1 gp41 could generate antibodies that are cross-reactive to astrocytes. Furthermore, anti-astrocyte antibodies, which were directed at a common epitope with the gp41 sequence, were found to be present in cerebrospinal fluid from some AIDS patients with central nervous system complications. Astrocytes regulate the environment for appropriate neuronal function, and astrocyte hyperactivity (astrocytosis) is known to be the common and early pathologic event in brains from patients with central nervous system AIDS. We suggest that antibody-induced effect(s) on astrocytes could lead to the physiologic neuronal dysfunctions observed in AIDS patients.

Pathogenesis of Theiler's murine encephalomyelitis virus

Theiler's murine encephalomyelitis virus belongs to the family of picornaviridae. Picornaviruses are small ( “pico”), phylogenetically related RNA viruses. Based on different biochemical and biophysical characteristics picornaviruses are subdivided into four groups: enteroaphthovirus (foot-and-mouth disease virus), cardiovirus [encephalomyocarditis virus (EMCV), Mengo virus], and rhinovirus (human rhinovirus). Theiler's murine encephalomyelitis virus was originally classified among the picornaviridae as an enterovirus because of its biological similarities with poliovirus. Further comparison of the complete genome of TMEV BeAn 8386 strain identifies remarkable similarities at the level of nucleotides and predicted amino acids between BeAn and the cardioviruses EMCV and Mengo virus. Theiler's murine encephalomyelitis virus is a single-stranded nonenveloped RNA virus. The viral RNA is of positive sense, having the same polarity as mRNA. Viral mRNA lacks the cap structure found at the 5’ end of almost all eukaryotic mRNAs.

The relationship between viral RNA, myelin-specific mRNAs, and demyelination in central nervous system disease during Theiler's virus infection

The DA strain of Theiler's murine encephalomyelitis virus (DAV) causes a chronic demyelinating disease in susceptible mouse strains. To elucidate the pathogenesis of DAV-induced demyelination, the authors investigated the spatial and chronologic relationship between virus (antigen and RNA), myelin-specific mRNAs, and demyelination in DAV-infected mice using immunohistochemistry, in situ hybridization, and slot blot hybridization analyses. In spinal cord white matter, viral RNA was detected easily in ventral root entry zones 1 to 2 weeks after infection. Viral RNA increased to maximum levels by 4 weeks after infection, which was associated with inflammation and mild demyelination. At 8 to 12 weeks after infection, when demyelination became most extensive, viral RNA was significantly decreased. Demyelination did not chronologically or spatially parallel the presence of viral RNA within the spinal cord. Decrease of myelin-specific mRNAs, including myelin-basic protein and proteolipid protein mRNAs, was observed within the demyelinating lesions with or without detectable viral RNA. These results indicate that a viral infection of white matter in the early phase of the infection initiates spinal cord disease leading to demyelination, but later an ongoing immunopathologic process contributes to the presence of extensive demyelination.

Monoclonal antibody to Theiler's murine encephalomyelitis virus defines a determinant on myelin and oligodendrocytes, and augments demyelination in experimental allergic encephalomyelitis

Theiler's murine encephalomyelitis virus (TMEV) causes a chronic demyelinating disease in mice. The mechanisms underlying the demyelination have not been fully elucidated. We have raised a mAb to TMEV (DA strain), H8, that reacts both with TMEV VP-1 and galactocerebroside (GC). In mouse brain cultures, cells positive for the mAb H8 epitope were double labeled with antibody to myelin basic protein, indicating that those cells were oligodendrocytes. Further, mAb H8 could immunostain myelin structures in frozen sections from mouse brains. When injected intravenously into mice with acute allergic encephalomyelitis, mAb H8 increased by 10-fold the size of demyelinated areas within the spinal cords. This is the first report demonstrating that an antibody to virus can enhance demyelination of a central nervous system disease. Ig fractions from the sera of mice with chronic TMEV infection had antibody(s) to GC, as well as to TMEV, as determined by ELISA. Furthermore, a competition ELISA for TMEV or GC antigen revealed that sera from these infected mice contained antibody(s) with the same specificity as mAb H8. Our results indicate that antibodies generated by immune response to TMEV can react with myelin and oligodendrocytes, and contribute to demyelination through an immune process.

Alteration of amino acid 101 within capsid protein VP-1 changes the pathogenicity of Theiler's murine encephalomyelitis virus

Chronic Theiler's murine encephalomyelitis virus infection of susceptible mice is an animal model for human demyelinating diseases. Previously we described an altered and diminished pattern of central nervous system disease in immunocompetent SJL/J mice infected with a variant virus. This variant virus H7A6-2 was selected with a neutralizing mAb recognizing the capsid protein VP-1 of Theiler's virus. Here we characterize the variant virus by ELISA and neutralization assays and by sequencing selected regions of the viral RNA genome and relate the alteration to disease. The variant virus contains one single point mutation within a neutralizing epitope of VP-1. This nucleotide change lead to an amino acid replacement at amino acid 101 of VP-1, a threonine (wild type) to an isoleucine (variant). Model building based on sequence alignments and the known structure of the related Mengo virus indicates that the altered amino acid is located in an exposed loop on the surface of the virus at the periphery of a site that has been proposed to be the receptor binding site. The results of ELISA, neutralization assay, and direct RNA sequencing provide for the first time an opportunity to precisely map an important structural determinant of neurovirulence.

Survival of athymic (nu/nu) mice after Theiler's murine encephalomyelitis virus infection by passive administration of neutralizing monoclonal antibody

Little or no antiviral immune response is mounted in athymic nude mice infected with the Daniels strain of Theiler's murine encephalomyelitis virus. In these athymic mice, increasing levels of infectious virus could be detected in the central nervous system. Seventy-five percent (9 of 12) of the nude mice were moribund or dead by 4 weeks postinfection. In contrast, treatment of Theiler's virus-infected nude mice with a neutralizing monoclonal antibody (H7-2) against the viral protein VP-1 resulted in a dramatic reduction of infectious virus within the central nervous system. All antibody-treated nude animals survived beyond 4 weeks postinfection. Monoclonal antibody titers could be maintained by passive transfer in treated nude mice at levels comparable to those of polyclonal antibody titers found in heterozygous infected nu/+ littermates. Areas of demyelination were detected in the untreated animals as early as 7 days after infection with little or no remyelination present. In approximately one-half of the antibody-treated nude animals, no demyelinating lesions were found. However, the rest of these treated mice were found to have areas of both demyelination and remyelination. Thus, anti-Theiler's murine encephalomyelitis virus antibody against VP-1 can play a dramatic role in the survival of mice, clearance of virus, limiting viral spread, and altering the pattern of disease in the absence of a functional T-cell response.

A neutralization-resistant Theiler's virus variant produces an altered disease pattern in the mouse central nervous system

Theiler's murine encephalomyelitis virus infection of mice is an animal model for human demyelinating diseases. To further define the role of this virus in the disease process, we selected a virus variant resistant to neutralization by a monoclonal antibody to VP-1. This virus variant was then injected into SJL/J mice. Central nervous system tissue was compared between variant virus- and wild-type virus-infected mice. Within the brain, no large differences were observed between the two groups as to the distribution of inflammatory infiltrates around the injection site and the number of viral antigen-positive cells during the first weeks of the observation period. In contrast, in the spinal cord major differences were found between variant virus- and wild-type virus-infected mice regarding the number of inflammatory lesions, infected cells, and the size of the areas involved with time. By immunohistochemistry, equivalent numbers of infected cells could be found in the spinal cord 1 week postinfection (p.i.): however, after that time, the number of infected cells in the wild-type virus-infected mice continued to increase, whereas the virus-positive cells from the variant virus-infected mice gradually decreased. Thus, the number of viral antigen-containing cells peaked by 1 week p.i. in the variant virus-infected animals. Conversely, the number of infected cells in the spinal cords from mice inoculated with wild-type virus steadily increased until 8 weeks p.i. At this time (8 weeks p.i.), no more variant virus antigen-positive cells could be observed within the spinal cord. Plaque assay of central nervous system tissue confirmed these differences between the two groups observed by immunohistochemistry. No infectious variant virus could be isolated after 2 weeks p.i. from the brain and 4 weeks p.i. from the spinal cord, whereas infectious wild-type virus could be detected up to the end of the observation period (12 weeks p.i.). Virus which was isolated from variant virus-infected mice still retained the neutralization-resistant phenotype. These studies emphasize the important biological in vivo activity of Theiler's virus VP-1 in determining neurovirulence.

Is Theiler's murine encephalomyelitis virus infection of mice an autoimmune disease?

Viruses can initiate disease by many different means. Direct viral, immune mediated and host factors all play important parts. Molecular mimicry or having cross-reacting determinants that result in immune responses which have the potential to cause damage can be incorporated into this framework. Here, autoimmune responses generated by virus infection have been presented in relation to these other parameters. The cross-reacting immune response originally generated by virus would have to be directed toward or involve a disease inducing site such as an EAE (encephalitogenic), thyroiditis, or diabetogenic site. If the cross-reaction took place at a nondisease inducing site, the ensuring immune response may result in the production of autoantibodies, however no disease would occur. In other systems autoantibodies can potentiate an ongoing inflammatory response. This may be the case that is described here with Theiler's murine encephalomyelitis virus infection. Lastly, viruses having common determinants with MHC determinants may modify immune responses leading to immunosuppression and allowing virus to persist. In addition, similar determinants may lead to disease by an alternative route. For example, we have described a region of human cytomegalovirus that has a common determinant with HLA DR beta chain. This region is associated with diabetes in humans (Todd et al. 1988). Thus, many factors are involved in the outcome of disease induction by viruses of which autoimmunity is one.

Monoclonal antibody defines determinant between Theiler's virus and lipid-like structures

Theiler's murine encephalomyelitis virus is known to cause a chronic demyelinating disease in mice. The contributions of immunologic factors, i.e. humoral and cellular responses to virus and/or myelin components, and direct virus-cell interactions leading to demyelination are still unclear. One important factor could be antibody initiation of myelin destruction. Here we describe four monoclonal antibodies that react with Theiler's murine encephalomyelitis virus. Three of these neutralize the virus and one of these three could also bind to various lipid-like structures including galactocerebroside, a myelin component. Further, this monoclonal antibody reacted with oligodendrocyte-like cells in vitro. All four monoclonal antibodies reacted with VP-1 by Western blot analysis. Thus, an immune response generated by virus that cross-reacts with a myelin element such as galactocerebroside could play a role in directing autoimmune processes toward myelin destruction.

Theiler's virus infection in nude mice: viral RNA in vascular endothelial cells

Infection of athymic (nu/nu) mice with Theiler's murine encephalomyelitis virus results in an acute encephalitis which resembles poliomyelitis. Immunohistochemistry and in situ hybridization were used to delineate the presence of viral proteins and RNA in the nervous systems of nude mice infected with the Daniels strain of Theiler's virus. This system permits the analysis of a viral infection in the absence of an effective immune response. By immunohistochemistry, viral antigen was found in the processes and cell bodies of neurons and glial cells. Besides the presence of viral antigen in these cell types, by in situ hybridization, Theiler's virus RNA was also found in cells associated with vascular endothelium in the brains and spinal cords of these infected mice. Theiler's virus RNA-positive endothelial cells were observed not only near the primary lesions but also away from demonstrable lesions in normal-appearing regions in the central nervous system. Earlier work had suggested an intra-axonal dissemination for this virus (M. C. Dal Canto and H. L. Lipton, Am. J. Pathol. 106:20-29, 1982). Our findings are consistent with this model but also suggest an additional mechanism for virus spread within the central nervous system, i.e., by infecting vascular cells and crossing the blood-brain barrier. Lastly, after Theiler's murine encephalomyelitis virus infection, not only glial cells but also endothelial cells express major histocompatibility complex class II (la) antigen on their surface (M. Rodriguez, M. L. Pierce, and E. A. Howie, J. Immunol. 138:3438-3442, 1987). Our demonstration of Theiler's virus-infected endotheliumlike cells may explain interactions of virus products in stimulating antigen presentation.

Demyelinating, non-demyelinating and attenuated canine distemper virus strains induce oligodendroglial cytolysis in vitro

A virulent canine distemper virus (CDV) strain that causes demyelination in vivo has been shown to induce oligodendroglial degeneration in vitro. In order to investigate if this effect on oligodendrocytes is specific for demyelinating strains only, primary brain cell cultures were infected with either virulent demyelinating strains (A75/17 and CH84-CDV), a virulent non-demyelinating strain (SH-CDV) or a non-virulent strain (OP-CDV). All virulent viruses caused a persistent type infection with moderate cytolysis whereas the non-virulent strain was highly cytolytic. All strains induced a similar pattern of oligodendroglial degeneration. It was concluded that the ability to induce oligodendroglial degeneration, which is thought to be the in vitro correlate of demyelination in vivo, is inherent to CDV irrespective of the strain. The discrepancy between biological behaviour of CDV strains in brain cell cultures and in vivo can be explained by the more complex virus-cell interactions in vivo than in vitro.

Oligodendroglial pathology in canine distemper virus infection in vitro

Dog brain cell cultures were infected with different canine distemper virus (CDV) strains to study the oligodendrocytes, which were characterized with eight different antibodies to cover the whole oligodendroglial population in the culture. A few weeks after infection all oligodendroglial cell types started to degenerate and disappeared from the culture. However, since no CDV protein could be demonstrated in the degenerating oligodendrocytes with extensive double-labelling studies, this lesion can not be explained as being a result of cytolytic infection. This conclusion was further supported in experiments with plaque-forming CDV, in which viral replication is restricted to the cytolytic areas only; oligodendrocytes also degenerated in virus-free areas between the plaques. The hypothesis of toxic factors released by other infected cell types in the culture leading to secondary damage of the oligodendrocyte could not be confirmed by transferring supernatants from infected to normal cultures. Whereas the presence of toxic factors can not be completely excluded, the possibility of an abortive infection of the oligodendrocytes with no or very limited viral protein synthesis is discussed.

Secondary degeneration of oligodendrocytes in canine distemper virus infection in vitro

To study the pathogenesis of demyelination in canine distemper virus (CDV) infection, primary canine brain cell cultures were infected with CDV to examine specific virus-induced glial cell changes. Cultures were harvested at regular intervals after inoculation and were immunostained for the specific demonstration of astrocytes, oligodendrocytes, and CDV antigen. The infection spread slowly with moderate cytolysis and cell fusion. Soon after inoculation, infection of astrocytes was found by means of double immunofluorescent labeling. One week after inoculation, CDV-induced astrocytic fusion and rearrangement of astroglial fibrils became apparent. The astrocytic changes progressed during the observation period. Double immunofluorescent labeling failed to show oligodendroglial infection. Despite clear absence of viral replication within oligodendrocytes at all stages of the experiment, these cells exhibited marked pathologic changes starting at about 20 days after inoculation and progressed to complete cytolysis within 10 days. The degeneration of the oligodendrocytes was thought to be secondary to CDV-induced changes in other cell types of the culture probably through the release of toxic factors in the tissue culture medium. Since little evidence has been found for oligodendroglial infection in demyelinating lesions in canine distemper in vivo, the present tissue culture findings suggest that demyelination in vivo could be the result of indirect oligodendroglial damage caused by CDV-induced changes in other cell types such as astrocytes.