key: cord-1024827-a471a1xo authors: Bieber, A. J.; Rodriguez, M. title: Chapter 44 Experimental Models of Virus-Induced Demyelination date: 2004-12-31 journal: Myelin Biology and Disorders DOI: 10.1016/b978-012439510-7/50097-8 sha: b6dcd0d0b3ace5c799ca92533a57047956ea82ac doc_id: 1024827 cord_uid: a471a1xo Publisher Summary This chapter reviews two of the most widely studied animal models of virus-induced demyelinating disease. These are Theiler's murine encephalomyelitis virus and murine hepatitis virus. Both viruses produce acute inflammatory encephalitis that is followed by chronic central-nervous-system (CNS) demyelinating disease. The clinical and pathologic correlates of virus-induced demyelination are largely immune mediated. Furthermore, several pathologic mechanisms have been proposed to explain the development of myelin damage and neurologic deficits, and each of the proposed mechanisms may play a role in disease progression depending on the genetic constitution of the infected animal. The induction of demyelinating disease by virus may be directly relevant to human MS. Several viruses are known to cause demyelination in humans and viral infection is an epidemiologic factor that is consistently associated with clinical exacerbation of MS. It is suggested that viral infection may be a cause of MS, although no specific virus has been identified as a causative agent. The clinical and pathologic presentation of virus-induced demyelination, the immunemediated nature of the disease mechanisms, and the variability of the disease presentation depending on genetic background are all very similar to what is observed in MS. The many similarities between human MS and the demyelinating diseases that are induced by infection with TMEV and MHV make these animal models very attractive for the study of multiple sclerosis. Theiler's murine encephalomyelitis Virus (TMEV) belongs to the cardiovirus genus of the picornavirus family (Nitayaphan et al., 1986; Pevear et al., 1987) . The original isolation and characterization of the virus was reported by Theiler in 1937 when the eVects of TMEV infection of the CNS were observed in a mouse that spontaneously developed Xaccid paralysis. The virus was subsequently transmitted to other mice by intracerebral inoculation with a suspension made from the brain and spinal cord of the infected mouse. TMEV is a ubiquitous enteric pathogen that usually causes asymptomatic intestinal infections in mice. Occasionally, the virus will spread beyond the intestinal tract and enter the CNS, resulting in both acute and chronic CNS infections. Chronic CNS infection results in extensive demyelination with accumulating neurologic deWcits and has provided a valuable experimental model for human demyelinating disease. Since its initial discovery, isolates of TMEV have been recovered in several laboratories. The diVerent TMEV strains have been divided into two subgroups depending on the disease that they induce after infection of the CNS (Lorch et al., 1981) . The GDVII and FA strains are highly virulent with an LD 50 as low as 1 to 10 plaque forming units (PFU). CNS infection causes acute encephalitis that is characterized by the destruction of a large number of CNS neurons, especially in the cortex, hippocampus, thalamus, brain stem, and in the anterior horns of the spinal cord. The encephalitis is usually fatal, and the highly virulent strains do not result in persistent infection of the CNS. A second TMEV subgroup is far less virulent (LD 50 > 10 6 PFU) and results in a distinctly diVerent disease presentation. This subgroup, which includes the TO4, DA, BeAn 3886, Yale, and WW strains, induces a biphasic CNS disease (Daniels et al., 1952; Lipton, 1978; Wroblewska et al., 1977) . Following intracerebral infection, these strains replicate in neurons in the brain resulting in encephalitis similar to that observed with the GDVII strain but nonlethal. The virus is cleared from the brain by the host immune response but persists in the spinal cord, eventually resulting in the development of chronic demyelinating disease (Lehrich et al., 1976; Lipton, 1975; Njenga et al., 1997; Rodriguez et al., 1987) . The disease is characterized by viral persistence in oligodendrocytes (Rodriguez et al., 1983) and macrophages , with chronic demyelination and progressive loss of motor function . The pathology is largely immune mediated with animals demonstrating a range of disease phenotypes depending on their genetic background. In the SJL strain, demyelination is evident within 30 days after infection. By 90 days, infected animals begin to develop spasticity and gait abnormalities, and weakness of the lower extremities, with paralysis eventually occurring by 6 to 9 months (Lipton and Dal Canto, 1976a) . The X-ray crystallographic structures of DA, BeAn and GDVII viruses have been determined to about 3 angstrom resolution (Grant et al., 1992; Luo et al., 1992 Luo et al., , 1996 Toth et al., 1993) . A schematic representation of the TMEV capsid, based on the crystallograghic data, is presented in Figure 44 .1. Each virus consists of a protein shell that is composed of 60 au2 FIGURE 44.1 TMEV capsid and genome structure. The upper panel shows a schematic representation of the TMEV capsid, based on the X-ray crystallograghic data. Lower panel shows a map of the TMEV DA strain genome. The 5' and 3' untranslated (UTR) regions are indicated. The 5' end of the RNA is bound by the viral VPg protein (ball) and the 3' end is polyadenylated (AAAAAn) . The open bar indicates the open reading frame from nucleotides 1066 to 7968. The heavy lines dividing the bar indicate the positions of the initial proteolytic cleavages after the leader peptide (L) and between the P1, P2, and P3 precursor peptides. The lighter lines dividing the bar indicate the arrangement of the Wnal gene products. The identities of the Wnal gene products are discussed in the text. The TMEV capsid structure is reproduced with permission from Webster and GranoV (1995). protomers, arranged as 12 pentamers with icosahedral symmetry. Each protomer contains one of each of the four capsid peptides (VP1, VP2, VP3, VP4). The VP1, VP2, and VP3 capsid proteins are each composed of an eight-stranded antiparallel b-barrel. The loops that connect the strands of the b-barrel form the surface structure of the viral capsid and may play important roles as sites of binding for neutralizing antibody and in conformational determinants of virus persistence. A second surface feature, the ''pit'' formed at the contact region between VP1 and VP3, is the likely binding site for the cellular receptor (Zhou, et al., 2000; Jnaoui, et al., 2002) . The TMEV genome consists of a positive sense, single stranded RNA molecule. Genomes for both the GDVII and TO subgroups have been sequenced and found to be very similar, with about 90% identity at the nucleotide level and 96% at the amino acid level . The genome of the DA strain is 8093 nucleotides long (Ohara et al., 1988) and contains a single 6903 nucleotide open reading frame that starts at nucleotide 1066 and encodes a 2301 amino acid polyprotein. The genome terminates with a poly(A) tail. The 12 mature TMEV gene products are generated from the polyprotein by post-translational proteolytic cleavage. The gene products include a leader peptide, four capsid polypeptides (VP1-4), two viral proteases, a polymerase/helicase, the RNA-dependent RNA polymerase that is involved in genome replication, and a small basic protein which becomes covalently linked to the 5' end of viral RNAs. Two other gene products have unknown functions. The TMEV infection cycle is typical of the picornavirus family (Rueckert, 1996) . Attachment of the virion to speciWc cell surface receptors serves to position the virus close to the cell membrane. Attachment also induces a conformational change in the virion that results in the loss of the VP4 capsid protein and the translocation of the RNA genome across the cell membrane and into the cytoplasm. Co-opting the translational machinery of the cell, the RNA genome directs the synthesis of a single large polyprotein. While nascent on the ribosome, the polyprotein is cleaved by virus encoded protease activity, into three large precursor proteins: P1, P2, and P3. The precursor proteins are further cleaved to produce the viral structural proteins and proteins necessary for completion of the infection cycle ( Fig. 44.1) . Precursor P3 is processed to produce a viral proteinase (designated 3C), a protein involved in initiating replication of the viral genome (3B or VPg) and the viral RNAdependent RNA polymerase (3D). The viral polymerase copies the incoming viral genome to produce a negative-strand RNA that then serves as the template for replication of the positive-strand viral genome. The newly formed positive-strand RNA molecules may be copied to form more negative-strand templates or be packaged into virions as virus assembly proceeds. The P1 precursor protein is cleaved to produce the viral capsid proteins VP0 (1AþB), VP1 (1D), and VP3 (1C). As the concentration of capsid proteins increases in the cell, they begin to aggregate into protomers composed of one copy of each of the capsid proteins. These protomers then assemble into pentamers, and 12 pentamers assemble with a positivestranded, VPg bound viral RNA, to produce a noninfectious provirion. The Wnal step in maturation to infectious virus requires a maturation cleavage in which the VP0 protein is cleaved to give the Wnal VP2 (1B) and VP4 (1A) capsid proteins that are found in the mature virus. In neonates and in some cells in culture, mature virus may accumulate to very high levels in infected cells and can often be observed as paracrystalline arrays in electron micrograghs ( Fig. 44 .2). Virus is generally released from the cell by infection-mediated cytolysis. The time that is required to complete the infection cycle is generally 7 to 12 hours. In culture, TMEV is able to infect a wide variety of cell types from several diVerent species. Although a unique protein receptor for the virus has not been identiWed, several lines of evidence suggest that cell surface carbohydrate residues may play a central role in the binding and entry of virus. Removal of sialic acid from the cell surface by treatment with sialidase signiWcantly reduces the infectivity of BeAn virus in BHK cells (Fotiadis et al., 1991) . Similar decreases are observed when binding to sialic acid is blocked with wheat germ agglutinin, which binds to sialic acid on the cell surface (Fotiadis et al., 1991) . Crystalline virus in neonatal brain. A paracrystalline viral array in the cytoplasm of a spinal cord neuron from a neonatal SJL mouse infected with the DA strain of TMEV. An intact mitochondrion is visible (m), but the numerous cytoplasmic vacuoles (v) indicate the cytopathic eVects of viral infection. Sialyllactose comprises the terminal three sugar residues of cell surface oligosaccarides, and inclusion of sialyllactose in cell culture medium inhibits viral binding and infectivity by blocking receptor binding sites on the virus (Zhou et al., 1997) . Together these data demonstrate an important role for sialic acid in the recognition of the cell surface by the TO subgroup TMEV strains. Sialyloligosaccaride addition is a common post-translation modiWcation of cell surface proteins, which is consistent with the observation that TMEV can infect a variety of cell types. Which speciWc sialic acid containing proteins might be required for virus binding and entry has not yet been established. The neurovirulent strains of TMEV (GDVII and FA) do not require sialic acid for cell surface binding (Fotiadis et al., 1991) but instead use cell surface heparin sulfate proteoglycan. Pretreatment of virus with heparin sulfate, treatment of host cells with heparin sulfate cleaving enzymes, or treatment with heparin sulfate speciWc antibodies all inhibit GDVII binding and infection (Reddi and Lipton, 2002) . Proteoglycan deWcient CHO cells are resistant to infection, further supporting a role for proteoglycans in GDVII binding. The interaction of proteins with heparin sulfate is usually mediated by heparin-binding domains (HBD) and a heparin-binding domain consensus sequence has been described (Cardin and Weintraub, 1989) . A putative HBD is present in the amino acid sequence of the VP1 capsid protein of both the GDVII and TO subgroups. This sequence may play a role in the binding of GDVII to heparin sulfate but if so, conformational diVerences must exist between VP1 peptides of the TMEV subgroups because the TO subgroup viruses do not bind heparin sulfate. Proteoglycans are known to serve as co-receptors or attachment factors for a variety of diVerent viruses. For several of these viruses, heparin sulfate acts primarily as an attachment factor and is not fully suYcient to mediate virus entry. The extent to which additional protein entry receptors are needed for GDVII infection is not yet clear (Reddi and Lipton, 2002) . The encephalitic phase of TMEV-induced disease develops rapidly after intracerebral infection. Virus infects and replicates in neurons of the brain and spinal cord, with titers of infectious virus reaching their peak by 5 to 7 days. Immunohistochemical staining demonstrates the highest viral antigen loads in the thalamus, hypothalamus, midbrain, brain stem, and spinal cord gray matter. Neuronal infection and lysis is particularly pronounced with the GDVII subgroup viruses, and infection usually results in death soon after inoculation . Acute disease following infection with the TO subgroup viruses is also characterized by intense mononuclear cell inWltration in the brain and spinal cord. The combined eVects of direct viral lysis of neurons and damage from inWltrating cells may result in transient Xaccid paralysis within 2 weeks after infection, but in many cases the early phase of the disease is largely asymptomatic. The acute phase of the disease is followed by clearance of virus from the gray matter by the host immune response (Njenga et al., 1997) . Resistant strains of mice achieve virtually complete viral clearance. In susceptible strains, virus is cleared from the gray matter but persists in the white matter of the spinal cord for the lifetime of the animal. During persistent infection, the amount of infectious virus, and the number of cells with detectable virus antigen in the spinal cord is usually quite low. This is in contrast to the high level of viral genome that is present throughout the disease (Trottier et al., 2001) . Infected cells most often appear in the lateral or anterior columns of the thoracic spinal cord and the presence of infected cells is often accompanied by an inXammatory parenchymal inWltrate and perivascular cuYng. The diVuse inXammatory inWltrate consists of large numbers of CD4 þ and CD8 þ T lymphocytes, activated macrophages and microglia, B cells, and reactive astrocytes (Fig. 44.3; Lindsley et al., 1989) . The presence of inXammatory cells correlates closely with the development of myelin damage, and electron microscopy reveals inWltrating mononuclear cells stripping myelin from axons. Macrophages containing engulfed myelin debris are numerous in demyelinating lesions. The speciWc roles that these diVerent cell types play in the development of demyelination and neurologic deWcits will be discussed in subsequent sections. Demyelination is apparent by as early as 22 days after infection and in many mouse strains, as demyelination progresses, neurologic deWcits begin to appear. Demyelination pathology is illustrated in Figure 44 infected mice involves not just the disruption of neuronal function but neuronal and axonal loss as well (McGavern et al., 2000) . Quantitative assessment of CNS injury during TMEV-induced disease shows that signiWcant atrophy of the lateral and anterior white matter regions of the spinal cord can be detected by 45 days post-infection. Atrophy is most pronounced in the posterior cervical and anterior thoracic regions of the spinal cord. While levels of demyelination reach a peak by 100 days post-infection, spinal cord atrophy continues to increase between 100 and 220 days. A 25% reduction was observed in the area of the anterior and lateral white matter at the C8-T11 level of the spinal cord by 195 to 229 days after infection. In this study, atrophy was speciWcally assessed in areas of normal appearing white matter and not in lesion areas, to avoid complications due to area changes resulting from inXammation and edema. This marked spinal cord atrophy is the result of a 30% loss of medium to large axonal Wbers by 195 to 229 days post-infection. The progressive loss of axons correlates well with electrophysiologic changes in the spinal cord and with changes in motor ability. These observations demonstrate that the neurologic deWcits that develop during chronic TMEV-induced demyelinating disease are primarily due to axonal damage and loss rather than to demyelination alone. Intracerebral inoculation with the highly virulent GDVII subgroup viruses results in fatal encephalitis, while infection with the less virulent TO viruses results in milder encephalitis, followed by clearance of virus from the brain and virus persistence in the spinal cord. There is still some debate concerning whether the diVerence in disease that is induced by these viruses is due simply to diVerences in virulence. Would the GDVII strains persist and cause demyelination if animals survived the initial encephalitis or are determinants for persistence and demyelination a speciWc genetic property of the TO strains? Recombinant chimeric viruses, resulting from the exchange of sequences between the GDVII and BeAn (or DA) virus genomes, have been useful in beginning to map the determinants of virus neurovirulence and persistence. Initial studies localized the majority of neurovirulence to a segment of the GDVII genome from the middle of 1B to the middle of 2C, although additional upstream sequences had a more minor eVect (Fig. 44.1; Fu et al., 1990a) . Infection with recombinant DA virus that contains these GDVII regions results in death at a much reduced LD 50 compared to wildtype DA virus. Subsequent studies narrowed the critical segment to a region of P1 from 1B to 1D, the capsid coding region . Several studies have since conWrmed that major determinants of viral persistence lie in the region of the viral genome that encodes the capsid proteins and have identiWed regions in two of these proteins, VP1 and VP2, that appear to be necessary for persistence (Adami et al., 1998; Lin et al., 1998) . Molecular modeling suggests that these sequences may be positioned on the surface of the virus in a manner consistent with a role in the binding of the virus to its cellular receptor (Zhou et al., 1997) . DiVerences in receptor binding between the TMEV subgroups may therefore play an important role in determining the balance between neurovirulence and the ability to produce persistent infection. Several mutated strains of the GDVII virus have been characterized in which neurovirulence is greatly attenuated, and while these strains are viable, replicate in the brain, and may even induce encephalitis, they do not persist or induce chronic demyelinating disease . These experiments suggest that ability of a TMEV strain to produce persistent infection and demyelination is not solely a function of neurovirulence but may be a speciWc genetic property of the strain. Other investigators have reported that insertion into the GDVII backbone, of DA strain segments from across the length of the DA genome, attenuates virulence and results in virus persistence and demyelination (Fu et al., 1990b; Rodriguez and Roos, 1992 ; see also Jakob and Roos, 1996) . Further studies will be needed to determine whether there are speciWc genetic determinants in the TO subtype that control persistence and to map these determinants to speciWc viral proteins. An additional interesting diVerence between the GDVII and TO subgroups involves the expression of the leader peptide region (L) of the TMEV genome at the 5' end of the open reading frame (see Fig. 44 .1). The TO strains use an alternative initiation codon to produce L*, a protein whose coding sequence is out of frame with the polyprotein reading frame. The GDVII strains do not encode the L* protein. Mutations in L* dramatically decrease virus-induced demyelination indicating that it plays a role in the development of disease (Chen et al., 1995) . The L* protein is essential for replication of the DA strain in a macrophage-like cell line in culture, while the GDVII strain does not replicate in these cells (Takata et al., 1998) . L* appears to function by preventing apoptosis of the DAinfected cells (Ghadge, et al., 1998) . The production of the L* protein therefore plays an essential role in the ability of TO subtype viruses to persist in speciWc cell types in the CNS. Analysis of the cytotoxic T lymphocyte (CTL) response in the CNS of mice infected with L* mutant virus, suggests that the inhibition of apoptosis by L* may prevent development of virus-speciWc cytotoxicity, permitting persistent virus infection and demyelinating disease . TMEV persistence in susceptible mice is associated with continuous virus replication and infectious virus and viral RNA can be isolated from the CNS for many months after initial infection. There is no single obvious cellular site of virus persistence in the spinal cord. Both viral antigen and RNA have been reported in oligodendrocytes, macrophages, and astrocytes (Aubert et al., 1987; Brahic et al., 1981; Lipton et al., 1995; Rodriguez et al., 1983) . The typical observation that only very small numbers of infectious virus can be recovered from chronically infected spinal cord initially led to the notion that the virus persists at low levels in the tissue. However, recent quantitative studies on the level of viral RNA in the CNS have demonstrated the presence of large amounts of full-length viral RNA, about 10 9 copies/spinal cord, throughout the course of the chronic disease (Trottier et al., 2001) . In contrast, infectious virus reaches a peak of >10 6 PFU/spinal cord by 15 days after intracerebral inoculation but then falls signiWcantly to 10 2 to 10 4 PFU/spinal cord after 28 days. This diVerence between viral genome load and infectious virus may reXect the rapid neutralization of virus by virus-speciWc antibodies. These data indicate that active viral replication continues throughout persistent viral infection. The identiWcation of a primary cellular site and speciWc mechanism of TMEV persistence has been diYcult. Based on immunostaining experiments, the predominant virus antigen burden in infected spinal cord has been reported to reside in either oligodendrocytes or macrophages Rodriguez et al., 1983) . More recently, fractionation of macrophages and oligodendrocytes from the chronically infected CNS revealed large amounts of viral genome in both cell types (Trottier et al., 2001) . Restriction of virus replication has been suggested as a mechanism for virus persistence and such restriction has been reported in macrophages isolated from TMEV infected CNS (Clatch et al., 1990; Levy et al., 1992) . However, the ratio of plus-to minus-strand TMEV RNA appears to be similar in macrophages and oligodendrocytes from infected spinal cord, suggesting that neither cell type is generally restricted for virus replication (Trottier et al., 2001) . As a potential site for TMEV persistence, macrophages have been the most widely studied. As mentioned earlier, restriction of virus replication has been reported in macrophages isolated from the TMEV infected CNS, and such restriction has been suggested as a possible mechanism for persistence. Depletion of blood borne macrophages early during the course of disease (7 to 21 days post-infection) decreases the amount of viral RNA in the CNS, supporting a role for macrophages in viral persistence . Studies in macrophage cell lines have demonstrated a block at the point of virus assembly rather than replication and it has been demonstrated that the diVerentiation state of the macrophage may be critical for the ability of the cell to support persistence Shaw-Jackson and Michiels, 1997) . Similar experiments, to look at potential viral persistence in the precursors of other glial types, have not yet been reported. Other investigators have suggested a role in virus persistence for other CNS cells, such as astrocytes and vascular endothelial cells (Sapatino et al., 1995; Zheng et al., 2001) . Whether there is a primary cellular site and a unique mechanism for TMEV persistence in the chronically infected CNS, is still an unresolved matter. au5 THEILER'S MURINE ENCEPHALOMYELITIS VIRUS (TMEV) Inbred strains of mice can be either resistant or susceptible to persistent TMEV infection and demyelinating disease. Upon intracerebral infection with virus, resistant strains mount an anti-viral immune response that controls virus replication and eventually clears virus from the animal. Upon clearance, infectious virus, virus antigen, and viral RNA can no longer be detected. Susceptible strains also mount a strong immune response against virus but virus is not cleared. A persistent viral infection is established along with the development of demyelinating disease, as described earlier. Since virus clearance is mediated by the immune response, mutations that disrupt major functional elements of the immune system generally result in some level of susceptibility, even in an animal with an otherwise resistant genetic background. The eVects of mutations of this type will be discussed in subsequent sections. Here we review naturally occurring genetic loci that act as modiWers of the response to TMEV infection and that have been identiWed using standard techniques of genetic analysis in mice. Early studies with diVerent strains of mice demonstrated a wide range of susceptibility to TMEV-induced disease (Lipton and Del Canto, 1979) . F1 and F2 backcrossing experiments between resistant and susceptible strains revealed that several loci were important for virus clearance and that one of these loci was linked to the H-2 major histocompatibility complex on chromosome 17 (Lipton and Melvold, 1984) . Further genetic analysis of this region showed that mice with certain H-2 haplotypes (H-2 f, p, q, r, s, v ) are susceptible to persistent virus infection, while others (H-2 d,b,k ) conferred resistance (Rodriguez and David, 1985) . Resistance was a dominant eVect and mapped to the D region of the H-2 complex, suggesting that a class-I mediated element of the immune response was essential for virus resistance (Clatch et al., 1985; Rodriguez et al., 1986b) . Direct demonstration that an H-2 linked class-I gene was responsible for virus resistance was provided when the H-2D d class-I gene was introduced into a susceptible mouse strain and these transgenic animals were shown to be resistant (Azoulay et al., 1994) . A key role for class-I restricted CD8 þ T cells as eVectors of resistance has since been conWrmed with targeted mutations in CD8 and b2-microglobulin (which lacks MHC class-I function). Both of these mutations result in TMEV susceptibility in animals with otherwise resistant genetic backgrounds (Fiette et al., 1993; Murray et al., 1998; Pullen et al., 1993; Rodriguez et al., 1993) . The H-2D b gene exerts its eVect on the stimulation of virus-speciWc CTL through the presentation of an immunodominant peptide from the VP2 capsid protein: VP2 121-130 (Dethlefs et al., 1997; Johnson et al., 1999) . There are other class-I molecules whose peptide binding sites are identical to those of the D locus and therefore have the potential to present TMEV peptide. These molecules, however, have not been shown to play a role in establishing resistance, possibly due to diVerential patterns of expression or to structural diVerences outside of the peptide binding domain that are important for antigen presentation (Altintas et al., 1993; Lin et al., 1997a) . Non-MHC loci also have an inXuence on virus resistance/susceptibility. The T-cell receptor genes lie on chromosome 6 (the Tcrb locus) in the mouse, and genetic mapping studies have indicated that a gene aVecting resistance maps close to the Tcrb locus. Several commonly used strains of susceptible mice have large deletions of the Vb T-cell receptor genes, further suggesting a link between T-cell function and virus resistance . Genetic analysis of congenic mice with speciWc Vb deletions indicates that in susceptible genetic backgrounds, Vb deletions enhance susceptibility but that have little eVect on resistance. These data generally support a role for T-cell receptor genes in determining virus resistance (Rodriguez et al., 1994) . Mapping experiments using the susceptible SJL strain and the resistant B10.S strain, two strains that both have the same H-2 S haplotype, have identiWed loci on chromosomes 10 and 18 (Bihl et al., 1999; Bureau et al., 1993) . There are two loci close to the gene for interferon-g (Ifn-g) on chromosome 10. While the possibility of a resistance locus at the site of the Ifn-g gene was of obvious interest, Wne mapping experiments have excluded Ifn-g as a candidate (Bihl et al., 1999) . A second locus is near the gene for myelin basic protein (MBP) on chromosome 18. The exact identity of the gene involved has not yet been au6 au7 au8 established but the possibility that TMEV resistance might be inXuenced by one of the major structural genes for myelin is intriguing. Studies on TMEV infection of MBP deWcient shiverer mice demonstrate enhanced susceptibility in the absence of MBP, supporting the idea that MBP might play a role in determining resistance (Bihl et al., 1997) . Loci that aVect TMEV resistance and persistence have also been identiWed on chromosome 3 near the carbonic anhydrase-2 (Car-2) locus , on chromosome 11 (Aubagnac et al., 1999) , and on chromosome 14 (Bureau et al., 1998) . The identity of the genes that are involved at these loci has not yet been established. Gender has also been reported to inXuence the susceptibility of certain mouse strains to TMEV-induced demyelinating disease (Kappel et al., 1990) . When considering TMEV-induced disease as a model for human demyelinating disease, questions regarding the mechanisms of demyelination and axonal damage are of central importance. Several diVerent mechanisms have been proposed to explain the pathology and neurologic deWcits that usually accompany persistent TMEV infection, but there is still disagreement in the Weld about the relative importance of the various potential mechanisms for the development of demyelination and axonal damage. Whatever the exact mechanisms of disease resistance and susceptibility, it seems certain that a balance between persistent virus infection and immune cell activation determines whether demyelinating disease will develop in immunocompetent mice. Mice with severe combined immunodeWciency (SCID) normally die from acute encephalitis when infected with TMEV. Adoptive transfer of normal spleen cells can result in survival and the development of demyelinating disease, but the results of adoptive transfer depend critically on the number of cells transferred . Transfer of too few cells fails to ameliorate the SCID phenotype and most mice die. Transfer of too many cells results in complete virus clearance and the mice survive but do not develop disease. Only when an intermediate number of cells is transferred do the surviving mice develop persistent virus infection and demyelination. These results demonstrate the balance between immune activation and virus persistence that determines the quality and extent of demyelinating disease. Proposed mechanisms of TMEV-induced demyelination include (1) demyelination resulting from direct virus lysis of infected oligodendrocytes or as a ''dying back'' response of oligodendrocytes following infection; (2) TMEV-speciWc, immune-mediated destruction of infected oligodendrocytes; (3) ''bystander'' damage to glia by toxic mediators from activated macrophages; (4) ''epitope spreading,'' an autoimmune attack against myelin antigens resulting from the de novo processing and immune recognition of myelin antigens that follows the initial events of virus-induced myelin damage; and (5) ''molecular mimicry,'' an autoimmune attack against myelin antigens based on the activity of virusactivated T cells that are cross-reactive with myelin antigens. While several distinct mechanisms of demyelination have been proposed and supported by experimental evidence, this does not preclude the likelihood that diVerent combinations of these mechanisms may be active at various times during the course of TMEV-induced disease and in animals with varying genetic composition. Direct virus effects on oligodendrocytes While it is generally agreed that immune activation plays a major role in demyelination, several lines of evidence also support a role for direct eVects of viral infection on oligodendrocytes. In cultures of mixed neonatal brain cells, TMEV infection of oligodendrocytes is primarily lytic (Graves et al., 1986; Rodriguez et al., 1988) . Two days after infection, many cultured oligodendrocytes display cytopathic ultrastructural changes characteristic of picornavirus infection, such as numer-au9 THEILER'S MURINE ENCEPHALOMYELITIS VIRUS (TMEV) ous cytoplasmic vacuoles and disruption of the endoplasmic reticulum, with preservation of mitochondria (see Fig. 44 .2). The number of oligodendrocytes in these cultures begins to diminish within 48 hours after infection and by 7 days has decreased to 50% of the value in uninfected control cultures (Rodriguez et al., 1988) . Since TMEV infects oligodendrocytes in vivo (Graves et al., 1986; Ohara et al., 1990; Rodriguez et al., 1988) and kills these cells in vitro, demyelination as a direct cellular consequence of the infection of oligodendrocytes seems plausible. A direct demonstration that infection can produce demyelination without the need for an accompanying activation of the immune response comes from studies of demyelination in nude mice which lack circulating T cells and cell mediated immunity (Roos and Wollmann, 1984; Rosenthal et al., 1986) . In these studies, infection of BALB/c nu/nu mice with the DA strain of TMEV resulted in the development of signiWcant neurologic deWcits by 2 to 3 weeks and death within 4 to 8 weeks. Demyelinated foci were observed in the spinal cord by 14 days post-infection even though inXammatory inWltrates were minimal. Electron microscopy revealed numerous infected and degenerating glial cells, some that were identiWed as oligodendrocytes. Damaged myelin was observed around intact axons as well as completely demyelinated axons. In general, the demyelinated lesions in these mice were small and inXammation is minimal compared to what might be expected in most susceptible strains of mice, suggesting that other mechanisms of demyelination may normally play a more prominent role. However, these observations do indicate that TMEV-induced glial damage can occur in the absence of a T-cell mediated immune response. In immunocompetent mice, the cytoplasm of most oligodendrocytes from infected animals appears normal on electron microscopic examination (Dal Canto and Lipton, 1975; Rodriguez et al., 1983) , even though viral RNA and antigen have been detected in oligodendrocytes throughout the course of persistent infection (Brahic et al., 1981; Rodriguez et al., 1983; Trottier et al., 2001) . Rodriguez (Rodriguez et al., 1983; Rodriguez, 1985) reported ultrastructural abnormalities of the innermost extent of the myelin sheath and has proposed a ''dying back'' mechanism for demyelination. In persistently infected SJL mice, the inner tongue of the myelin sheath, which is the most distal extent of the myelin lamellae, often exhibits swelling and abnormalities in the normal cytoachitecture. Abnormalities included vacuolization, degeneration of mitochondria, increased numbers of microWlaments, and the appearance of electron-dense bodies and debris. These indications of damage to the innermost layer of the myelin sheath were observed in the absence of any signs of damage to the external surface of the sheath, indicating that damage to the oligodendrocyte may occur before immune-mediated damage to the myelin surface. Immunoperoxidase studies revealed that viral antigen was present on both the inner and outer extents of the myelin lamellae, even though the ultrastructural abnormalities were restricted to the inner loops. The observed degeneration appears to begin in the inner loop of the oligodendrocyte lamellae and then proceeds proximally toward the cytoplasm with the ultimate destruction of the myelin sheath. The ''dying back'' concept was initially proposed as a mechanism underlying peripheral neuropathies in which the cell body is unable to support the structure and metabolism of a distant cellular process (Spencer and Schaumburg, 1976) . It seems particularly relevant that Ludwin and Johnson (1981) initially reported the ''dying back'' of oligodendrocytes in a very diVerent model for CNS demyelination in which demyelination is induced by ingestion of the toxic, copper-chelating agent, cuprizone. Dying back oligodendrogliopathy may therefore represent a general mechanism for demyelination in the CNS. TMEV-specific immune damage Dying back oligodendrogliopathy could be the direct result of viral infection on the cellular physiology of oligodendrocytes. However, it has been demonstrated that the demyelination observed following persistent TMEV infection in susceptible mouse strains is, at least in part, immune mediated (Lipton and Dal Canto, 1976b) . Dying back might also be the result of damage due to TMEV-speciWc immune reactions directed against infected oligodendrocytes. In cultures of mixed neonatal brain cells, immunoperoxidase staining visualized with electron microscopy demonstrated large amounts of viral antigen on the oligodendrocyte cell membrane (Rodriguez et al., 1988) . Similarly, staining in vivo can detect virus antigen in persistently infected oligodendrocytes by 28 days post-infection (Rodriguez et al., 1983) . Antigen is present in the cell cytoplasm as well as in the myelin lamellae. Infected animals typically produce high titers of virus-speciWc antibody and the recognition of virus antigen on the cell surface by these immunoglobulins may result in oligodendrocyte injury through several antibody dependent mechanisms including complement activation, antibody-dependent cell-mediated cytotoxicity (ADCC), or receptor mediated phagocytosis by macrophages. Alternatively, viral antigen might also be presented on the cell surface in the context of the MHC class I or class II proteins. MHC class I, in particular, is known to be upregulated in oligodendrocytes following TMEV infection of the CNS (Altintas et al., 1993; Lindsley et al., 1992) . In susceptible mouse strains, class-I MHC is up-regulated in the CNS by one day after infection and remains high in the spinal cord white matter throughout chronic infection and demyelination. A novel antigen such as viral peptide, presented on the cell surface by MHC class-I, might generate a class-I restricted cytotoxic T lymphocyte (CTL) response, which might cause primary demyelination as the result of chronic nonlethal injury to TMEV-infected oligodendrocytes. However, mice with a disrupted b 2microglobulin (b 2 -m) gene do not express signiWcant levels of MHC class-I, do not have functional CD8 þ T cells, and therefore cannot mount a CTL response. When b 2 -m knockout mice from a resistant haplotype are infected with TMEV, resistance is abrogated and the virus establishes persistent CNS infection. However, signiWcant demyelination develops in these mice (Fiette et al., 1993; Pullen et al., 1993; Rodriguez et al., 1993) , demonstrating that CD8 þ class I-restricted cytotoxic T cells are probably not the primary mediators of CNS demyelination. In general agreement with these observations, b 2 -m knockout mice from a susceptible haplotype display somewhat enhanced levels of demyelination (Begolka et al., 2001) . Similar results have been obtained with mice carrying a genetic deletion of the gene encoding CD8, which display little alteration in the disease phenotype in either resistant or susceptible strains (Murray et al., 1998) . Bystander damage In contrast to observations with MHC class-I and CD8, treatment of susceptible mice with neutralizing antibodies against MHC class-II (anti-Ia) and anti-CD4 has been shown to reduce demyelination (Gerety et al., 1994; Rodriguez et al., 1986a; Welsh et al., 1987) . These observations suggest that class-II restricted CD4 þ T lymphocytes might play a role in disease pathogenesis. Clatch and coworkers (1986) attempted to correlate the development of clinical disease with several pathophysiologic parameters including virus titer, titer of anti-TMEV speciWc antibody, delayedtype hypersensitivity (DTH), and T-cell proliferative responses. The extent of TMEVinduced demyelinating disease correlated most closely with the presence of high levels of TMEV-speciWc DTH. The development of TMEV speciWc delayed-type hypersensitivity would require an initial presentation of virus antigen to T H cells by MHC class-II, resulting in activation and clonal expansion. Generally, the activated T cells are of the CD4 þ T H 1 subtype and are often referred to as T DTH cells. Upon secondary contact with antigen, T DTH cells secrete cytokines that recruit and activate macrophages and other nonspeciWc inXammatory cells. Development of the DTH response may take 48 to 72 hours and by the time the reaction is fully developed, the vast majority of the activated cells at the site of inXammation are nonspeciWc immune cells and macrophages which act as the primary eVector cells of the DTH response. These cells release a variety of cytotoxic mediators which results in nonspeciWc ''bystander'' damage to cells in the area of the response. In the CNS of an animal with persistent TMEV infection, any cell that presents virus antigen with MHC class-II might therefore serve as a nidus for the development of a DTH response. The concept of DTH as a mechanism of demyelination is supported by adoptive transfer experiments using the sTV1 T-cell line or using TMEV sensitized lymph node T cells. sTV1 is a DTH-mediating CD4 þ T H 1 line that is speciWc for the immunodominant au11 THEILER'S MURINE ENCEPHALOMYELITIS VIRUS (TMEV) VP2 74-86 peptide of TMEV. Transfer of 5Â10 6 peptide-stimulated sTV1 cells into SJL mice that had been infected with a suboptimal dose of virus, signiWcantly potentiated clinical disease in the recipient mice. Animals that received cells but did not receive virus to provide secondary cellular stimulation did not develop disease. An experimental scheme using lymph node T cells from mice primed with UV-inactivated virus instead of sTV1 cells resulted in similar potentiation of disease but only when virus was present for secondary stimulation (Gerety et al., 1994) . Further supporting the notion of DTH mediated demyelination, induction of peripheral immune tolerance to TMEV signiWcantly reduces the development of DTH, general inXammation, virus-speciWc immune responses, and demyelination, following infection with the virus (Karpus et al., 1995) . Epitope spreading While the precise mechanisms may not yet be clear, it is well accepted that TMEV persistence in the CNS stimulates a virus-speciWc immune response that initiates damage to CNS cells and myelin. Recent studies have also detected myelinreactive T cells in chronically infected mice. The appearance of T cells that are reactive to myelin autoepitopes is presumably due to the development of an immune response to myelin antigens that are released during the initial stages of TMEV-induced demyelination. It has been proposed that these autoreactive T cells are pathogenic and that they may be a major source of pathology during the chronic stages of demyelinating disease. The generation of a myelin-speciWc pathologic immune response, following an initial event of virus-induced demyelination, is called ''Epitope spreading'' (Lehmann et al., 1993; McRae et al., 1995) . Earlier attempts to detect myelin reactive T cells from TMEV infected mice had been unsuccessful . However, Miller and coworkers (1997) have now reported that 3 to 4 weeks after initiation of demyelinating disease, T-cell responses to a series of myelin autoepitopes begin to appear in an ordered progression. T-cell proliferation and DTH responses to myelin peptides were Wrst demonstrated at about 50 days post-infection. By 164 days post-infection, reactivity was shown for multiple peptide epitopes from proteolipid protein (PLP), myelin basic protein (MBP), myelin oligodendrocyte glycoprotein (MOG), and mouse spinal cord homogenate. T-cell hybridoma clones speciWc for the immunodominant peptides of TMEV, and lymph node T cells from TMEV primed mice, did not show reactivity to myelin peptides. Conversely, lymph node T cells from myelin primed mice did not show reactivity to TMEV peptides. These results suggest that the reactivity to myelin epitopes was speciWc and not due to cross-reactivity with TMEV. Additionally, sequence comparison of the reactive myelin peptides to TMEV showed no strong homologies that might indicate a likely source of cross-reactivity. An epitope spreading mechanism for TMEV-induced demyelinating disease would be very similar to the pathologic mechanism of EAE. Adoptive transfer of myelin reactive immune cells from animals with EAE eVectively transfers the disease to naïve animals. However, attempts to transfer TMEV-induced demyelinating disease by the transfer of T cells from infected animals have been unsuccessful (Barbano et al., 1984) . Initial attempts to suppress TMEV-induced disease by inducing neuroantigen-speciWc tolerance to a heterogeneous mixture of CNS antigens were unsuccessful . More recently, induction of myelin-speciWc tolerance with a fusion protein containing MBP and PLP sequences has been reported to reduce demyelination and CNS inXammation in mice with TMEV-induced demyelinating disease (Neville et al., 2002) . Studies of TMEV-induced disease in CD4 knockout mice are in potential disagreement with the idea of CD4-dependent mechanisms of demyelination such as bystander damage (DTH) or epitope spreading. Murray and coworkers (1998) examined the eVect of CD4 deletion in both resistant (C57BL/6) and susceptible (SJL and PLJ) genetic backgrounds. In a resistant background, the absence of CD4 resulted in the establishment of persistent viral infection in the CNS and with the development of demyelination by 90 days postinfection. In both susceptible strains, demyelination was dramatically increased. It is not immediately obvious how best to address and reconcile the various observations described here. It might be noted, however, that CD4 plays an important role in immune control of viral infection and that TMEV titers in CD4 knockout mice were generally au12 elevated during the disease. The development of TMEV-induced pathology likely reXects a delicate balance between viral titer and activity and a dynamic immune response to the virus. As experimental protocols alter various aspects of this balance, such as virus titer and the quality and quantity of the participating immune cells, the mechanisms of pathology may also change in ways that make the interpretation of the data less straightforward than expected. In addition to the complexity of the system, diVerent strains of TMEV are currently being used by the most active research groups. The BeAn 8386 and Daniel's (DA) strains are both widely used and although both establish robust, persistent infections of the spinal cord, the BeAn strain is generally less virulent that the DA strain. The precise signiWcance of the use of these two strains to the interpretation and comparison of data is not yet clear. Like the epitope spreading model, molecular mimicry involves pathogenesis due to myelin reactive autoimmunity. However, in molecular mimicry, the autoimmune response results from structural homologies between TMEV and myelin proteins. Immune recognition of the virus therefore also gives rise to an anti-myelin response, and this autoimmunity is pathogenic. While molecular mimicry is probably an important mechanism for other CNS diseases (Rouse and Deshpande, 2002; Yuki, 2001) , several observations suggest that it may not play a prominent role in TMEV-induced disease. TMEV infection is followed by a rapid and robust immune response that includes the development of both TMEV-speciWc antibodies and T cells by 7 days post-infection. Demyelinating disease is Wrst observed by about 21 days post-infection and does not reach a peak until 90 to 100 days. The temporal diVerence between the development of the TMEV-speciWc immune response and the appearance of demyelinating disease argues against molecular mimicry as an important component of the demyelination mechanism. Furthermore, the initial immune response to TMEV does not cross-react with several myelin antigens including MBP, PLP, and MOG (Miller et al., 1997) . No strong homologies have been reported between TMEV protein sequences and those of the major myelin proteins. In the past, demyelination alone has often been considered to be suYcient for the development of axonal damage and neurologic deWcits. However, both the timing and mechanisms of demyelination and axonal damage can be distinct and diVerent in the TMEV model. Quantitative assessment of demyelination and neurologic ability in two susceptible mouse strains shows that small demyelinated lesions, accompanied by statistically signiWcant decreases in motor ability, can occur as early as 24 days post-infection . However, axonal damage and neurologic deWcits do not necessarily follow demyelination. Several reports have documented that deWciency of MHC class-I (b 2 -microglobulin mutants) or MHC class-II (Ab o mutants) function, will abrogate resistance to TMEV-induced disease in otherwise resistant strains of mice (Fiette et al., 1993; Njenga et al., 1996; Pullen et al., 1993; Rodriguez et al., 1993) . Resistant mice that are homozygous for either of these mutations develop extensive demyelination. However, class-II deWcient mice develop severe neurologic deWcits as indicated by signiWcant reductions in spontaneous activity, while class-I deWcient mice show no clinical signs. Electrophysiologic measurement of the conduction velocity of motor-evoked potentials shows that decreased conduction velocities accompany deWcits in class-II deWcient mice with chronic demyelination. Conduction velocities in class-I deWcient mice are no diVerent that those in uninfected controls . These Wndings demonstrate that demyelination and neurologic deWcits are genetically and functionally separable, and they implicate a role for MHC class-I in the development of neurologic deWcits following demyelination. A role for MHC class-I in the development of neurologic deWcits implicates CD8 þ T cells as the pathologic eVector cell for the induction of neurologic disease. Granule exocytosis (perforin release) and Fas ligand expression are two of the common cytopathic eVector functions of CD8 cells, and mice deWcient for perforin, Fas, and Fas ligand were tested to determine whether these molecules play a role in the development of demyelination and clinical disease (Murray et al., 1998) . Lpr (Fas mutation) and Gld (Fas ligand) mutant mice, on a TMEV-resistant genetic background, maintained resistance to viral persistence and demyelinating disease. Perforin-deWcient mice developed persistent viral infection and chronic demyelination of spinal cord white matter. Despite demyelination and virus persistence, these mice showed only minimal neurologic deWcits. These studies indicate that perforin release is the pathologic eVector for the induction of neurologic disease by CD8 þ T cells. The experiments described earlier for the determination of the roles played by class-I and class-II related mechanisms for the development of axonal damage were all performed in mice with the C57BL/6 or C57BL/6Â129 genetic background. C57BL/6 mice mount a vigorous CTL response to TMEV. This response is predominantly directed against a single immunodominant TMEV peptide that represents amino acids 121 to 130 of the VP2 capsid protein, which is presented in the context of the H-2D b MHC class-I molecule (Borson et al., 1997) . At 7 days post-infection, staining of CD8 þ brain inWltrating lymphocytes with D b :VP2 121-130 tetramers shows that 50 to 63% of the inWltrating CD8 þ cells are virus speciWc (Johnson et al., 1999) . Mice with a targeted disruption of the Ifn-g receptor gene respond to virus infection with rapidly progressing disease and severe clinical deWcits. These mice also develop a class-I restricted anti-viral response that is dominanted by the VP2 121-130 epitope. In these mice, intravenous injection of the VP2 peptide one day before virus infection completely eliminates the immunodominant T-cell response (Johnson et al., 2001) . Elimination of the response does not cause a signiWcant increase in viral titer. There is no change in the overall pattern of infection: by 45 days virus is primarily localized to the white matter of the spinal cord, there is no signiWcant increase in inXammation, and demyelination is similar in extent to control animals. Elimination of the response did, however, signiWcantly preserve motor function in infected animals. At 45 days after infection, VP2 peptide treated animals performed signiWcantly better than control animals in tests of motor ability, although not as well as uninfected animals. These results suggest that the CD8 þ T cells that mediate neurologic dysfunction are most likely the TMEV-speciWc cells that inWltrate the CNS following infection. In contrast to the work done in mice with the TMEV-resistant C57BL/6 genetic background, susceptible SJL mice that lack CD8 þ eVector function (b 2 -microglobulin knockouts) are not rescued from TMEV-induced neurologic deWcits (Begolka et al., 2001) . Instead, these mice show an earlier onset of neurologic disease, with enhanced CNS demyelination and macrophage inWltration at 50 days post-infection. These results indicate that in these mice, CD8 þ cells are not required for the initiation or progression of demyelinating disease and may illustrate the strong inXuence that genetic background can have on the development of this disease. Myelin repair, or remyelination, is a normal physiologic response to myelin damage. However, the ability of mice to repair the damage caused by TMEV-induced demyelinating disease is strongly dependent on the genetic constitution of the animal. b 2 -microglobulin deWcient mice lack MHC class-I function. On a C57BL/6Â129 genetic background, these mice develop extensive demyelination following TMEV infection . Despite the presence of persistent viral infection, these animals show extensive remyelination by 6 months post-infection . This spontaneous remyelination of CNS axons involves both oligodendrocytes and Schwann cells. In contrast, MHC class-II deWcient mice on the same genetic background also develop demyelinating disease but subsequent remyelination is largely absent . Class-II deWcient mice fail to produce TMEV-speciWc IgG and eventually developed high viral titers. However, the animals do not die during the acute phase of the disease but survive well into the chronic demyelinating phase, even though most die by 120 days postinfection. Failure to observe remyelination may simply reXect insuYcient time for myelin repair to occur before death. SigniWcant axonal damage is also apparent in these mice and this damage may also preclude remyelination. Like MHC class-II deWcient mice, the SJL strain is susceptible to TMEV infection, develops extensive demyelinating disease, and does not repair myelin damage. In contrast to class-II deWcient mice, SJL mice control viral infection eVectively and can survive for over a year with persistent infection and chronic demyelinating disease. The disease in SJL mice progresses from demyelination, through mild neurologic deWcits, to profound deWcits and paralysis. Two general hypotheses have been proposed to explain the absence of remyelination in these mice (Miller et al., 1996) . First, an inhibitory local environment may be present, which prevents spontaneous repair. Local inXammation, reactive astrocytes, microglia, or damaged axons might play a role. Second, an absence of growth factors or oligodendrocyte precursor cells may prevent eYcient remyelination. Which of these mechanisms is the most important remains unclear. The absence of spontaneous remyelination in SJL mice, however, has provided an excellent experimental system for the study of ways to enhance the remyelination process, and several potential therapeutic approaches have been studied and developed using this system (Asakura et al., 1998; Drescher et al., 1998; Njenga et al., 2000; Ure and Rodriguez, 2002; Warrington et al., 2000) . Whatever the mechanisms of remyelination, it seems clear that remyelination is associated with improved neurologic function. Like the SJL strain, PL/J mice develop chronic demyelinating disease but do not remyelinate. PL/J mice that are deWcient for either CD8 þ or CD4 þ T cells show a signiWcant increase in the severity of disease pathogenesis (Murray et al., 1998) . The disease is particularly severe in PL/J CD4 mutant mice. While early mortality is common in these animals, many survive up to 6 months post-infection, and many of the mice that survive to the chronic phase of the disease have nearly complete spontaneous remyelination. Importantly, all of the surviving mice show partial recovery of motor function that coincides with myelin repair, and the extent of recovery correlates strongly with the percentage of lesioned white matter area that has remyelinated (Murray et al., 2001) . These observations clearly demonstrate that functional recovery is possible despite previous demyelination and ongoing persistent virus infection, and that remyelination correlates with this recovery. Murine hepatitis virus (MHV) is a member of the Coronaviridae, a family of large, enveloped, plus-strand RNA viruses. MHV is a common enteric pathogen of mice and rats that occasionally disseminates to the CNS. Like TMEV, the original isolation and characterization of one of the most commonly studied strains, MHV-JHM, occurred with the spontaneous appearance of two paralyzed mice, and the subsequent passage of the virus in mouse brain were it continued to produce paralytic disease (Cheever et al., 1948; Pappenheimer, 1958) . Later passages of the virus resulted primarily in encephalitic disease with accompanying demyelination. The demyelinating phase of MHV infection has been used as a model for human demyelinating disease The MHV genome is a 31 kb message-strand RNA that is capped and polyadenylated. The genomic RNA is complexed with the viral nucleocapsid phosphoprotein (N) to form a long, helical nucleocapsid. The nucleocapsid lies within a lipoprotein envelop that consists largely of intracellular host cell membrane and three viral capsid proteins: membrane protein (M), spike protein (S), and hemagglutinin-esterase protein (HE). MHV virions bind to the plasma membrane of the host cell by the interaction of the spike proteins with the host receptor. Biliary glycoprotein 1a (Bgp1a), a member of the immunoglobulin superfamily, is a known MHV receptor (Williams et al., 1991) . The main determinant of cellular tropism is the ability of the spike protein to react with a suitable receptor on the cell surface and infected cell populations in the animal closely match those with demonstrated expression of the Bgp1a receptor. However, very little Bgp1a is expressed on CNS glia and neurons, and the mechanism of virus entry into these cells is still uncertain. The MHV virion also contains the HE protein. Binding of HE to 9-O-acetylated neuraminic acid residues on the plasma membrane may serve as an important pre-receptor function that facilitates viral entry. After receptor binding, the viral envelope fuses with the plasma membrane in an Sprotein mediated event. Once the RNA genome has entered the cytoplasm of the cell, it is translated into a polyprotein that is co-and post-translationally processed into multiple proteins. These proteins, which include the capsid proteins (N, M, S, and HE), the virus-speciWc RNA-dependent RNA polymerase, and virus encoded proteases, play various roles in the replication and development of infectious virus. The RNA-dependent RNA polymerase uses the positive-strand genomic RNA as a template to produce full-length negative-strand RNA, which in turn is used to make more positive-strand genomic RNA. Overlapping sets of 3' co-terminal subgenomic positive-strand RNAs are also produced and translated into viral gene products. The newly synthesized genomic RNA assembles with nucleocapsid (N) protein to form the helical nucleocapsid. The S, HE, and M glycoproteins are all translated on membrane bound polysomes and are co-translationally inserted into intracellular membranes. Virions begin to form when nucleocapsids bind to intracellular membranes that contain the viral M protein and virus then buds from these specialized membranes. S and HE proteins are incorporated into the virions during budding while host proteins are excluded. After budding, mature virions accumulate in large, smooth-walled vesicles that release virus when they fuse with the plasma membrane. The outcome of MHV infection depends on several variables such as virus strain, virus dose and route of infection, age, and genetic composition of the host. Two diVerent MHV strains, MHV-A59 and MHV-JHM, have been used in most studies of MHV-induced demyelination. As the name suggests, hepatitis is a common outcome of MHV infection but while infection with the MHV-A59 strain results in prominent hepatitis, MHV-JHM infected animals show little or no evidence of hepatitis. However, intracerebral or intranasal infection with either strain results in acute, sometimes fatal, encephalitis, followed by the development of chronic demyelination in animals that survive. To decrease mortality and thereby facilitate the study of MHV pathogenesis, several MHV-JHM variants have been developed that have limited neurotropism related lethality but that continue to produce an early encephalitis followed by acute primary demyelination (Erlich et al., 1987; Fleming et al., 1986; Haspel et al., 1978) . MHV infects a variety of CNS cell types including astrocytes, oligodendrocytes, microglia and neurons. As virus replicates to high titers in the brain and spinal cord, it is controlled by the development of an anti-viral immune response that is largely restricted to the CNS. Control of infection correlates well with the appearance of an inXammatory inWltrate by 4 to 7 days post-infection. This acute inXammatory response is characterized by the inXux of many types of immune eVector cells including CD4 þ and CD8 þ T cells, B cells, macrophages, and natural killer cells (Williamson et al., 1991) . Virus neutralizing antibodies appear in the serum by 7 to 9 days post-infection but probably do not play a critical role in initial virus clearance. Infectious virus is cleared from the CNS by 10 to 14 days but complete clearance is not achieved and virus persists in the CNS. Infectious virus is rarely recovered from spinal cord tissue after the acute phase of infection. Virus antigen can be detected for several months after initial clearance and appears to persist in astrocytes and oligodendrocytes (Sun et al., 1995) . Viral RNA can be detected in the CNS for at least a year after infection (Lavi et al., 1984) . Demyelination with axonal sparing can be detected as early as 5 days post-infection. MHV-induced demyelination is illustrated in Figure 44 .5. Relatively few inXammatory cells but large numbers of debris containing macrophages characterize demyelinated lesions, and their appearance correlates well with the development of lesions (Lane et al., 2000) . Electron microscopy demonstrates macrophage processes extending between the au13 FIGURE 44.5 Pathology of the MHV-infected spinal cord. (A) A section of spinal cord tissue from an SJL mouse infected with MHV-A59 and then stained for myelin with para-phenylene-diamine shows multiple white matter lesions (arrows). (B) Normal white matter lies adjacent to a demyelinated lesion in the ventral spinal cord tracts. Myelin destruction in the lesioned area is apparent. layers of myelin sheaths, strongly suggesting an active role for macrophages in the demyelination process (Powell and Lampert, 1975) . Demyelination continues to develop until about 21 days after infection and then begins to slow. Spontaneous remyelination is common following the initial phase encephalitis and demyelination. After the initial phase of encephalitis, demyelination, and viral clearance, virus persistence results in a second phase that probably continues for the life of the animal and that is associated with the recurrent episodes of demyelination. Infected adult mice generate a potent immune response that clears the virus in 10 to 14 days. CD4 þ and CD8 þ T cells are the most critical elements of this response. Viral infection results in the rapid recruitment of large numbers of CD8 þ T cells to the CNS parenchyma, while CD4 þ cells are primarily localized to the perivasculature and meninges. CD8 þ T cells are the primary immune eVectors of virus clearance. MHC class-I tetramer staining shows that over 50% of the CD8 þ T cells in the CNS after infection are virus speciWc . As virus is cleared, the total number of CD8 þ cells in the CNS decreases but the percentage of cells that are virus-speciWc remains high. Initially the inWltrating CD8 þ cells display virus-speciWc cytolytic activity. As clearance progresses and persistence is established, these cells lose their cytolytic function while retaining their ability to secrete Ifn-g, indicating diVerential regulation of CD8 þ eVector functions as the course of infection progresses. Consistent with a role for CD8 þ cells in virus clearance, CD8 deWcient mice show increases in viral titer, increased mortality, and an inability to eYciently clear virus after infection (Lane et al., 2000; Williamson et al., 1990) . CD8 þ cells probably use two diVerent eVector mechanisms to clear virus from diVerent subsets of infected cells. Perforin-mediated cytolysis may play an essential role in the control of virus infection of astrocytes and microglia. Perforin deWcient mice are able to clear MHV from the CNS but the clearance is delayed by several days (Lin et al., 1997b) , indicating that perforin-dependent CTL plays a role in viral clearance from the CNS. Histologic analysis has shown that virus-speciWc CTL are more eYcient at the clearance of virus from astrocytes and macrophages than from oligodendrocytes (Stohlman et al., 1995) . Ifn-g is the antiviral mechanism primarily responsible for controlling the infection of oligodendrocytes (Parra et al., 1999) . In Ifn-g deWcient mice, MHV infection results in increased neurologic disability and mortality that is associated with increased viral titers and delayed viral clearance. The anti-viral antibody response and the CTL response are not aVected. Increases of viral antigen in Ifn-g deWcient mice are located almost exclusively in oligodendrocytes and are accompanied by increased numbers of CD8 þ cells in CNS white matter. CD8 þ cells are therefore major participants in MHV clearance and employ at least two diVerent eVectors mechanisms to clear virus from diVerent types of infected cells. Some studies have also suggested that the Fas-Fas ligand system may play a role (Parra et al., 2000) . Although the ability to control MHV replication is predominantly controlled by CD8 þ cells, this response is highly dependent on the CD4 þ T-cell response (Stohlman et al., 1998) . Adoptive transfer of activated CTL into recipients that have been depleted of CD4 þ cells shows that although the CD8 þ cells can eVectively traYc to sites of CNS infection in the absence of CD4 þ cells, once they arrive they rapidly undergo apoptosis. These observations indicate that the survival of the CD8 þ cells that are primarily responsible for virus clearance may depend critically on support from CD4 þ cells. Consistent with these observations, studies on CD4 deWcient mice show increases in viral titer, increased mortality, and an inability to eYciently clear virus, a pathologic phenotype very similar to that observed for CD8 deWcient mice (Lane et al., 2000) . CD4 deWcient mice also exhibit a signiWcant reduction in the number of activated macrophages and microglia in the CNS of infected animals (Lane et al., 2000) . This may result from a decrease in the expression of the chemokine RANTES that is secreted by CD4 þ cells and acts to recruit inWltrating mononuclear cells to the CNS. In strong support of this idea, in vivo depletion of RANTES with RANTES-speciWc antiserum mimics the phenotype seen in CD4 deWcient mice. Despite the increases in viral titer, increased mortality, and inability to clear virus, CD4 mutant mice that survive infection show a general decrease in inXammation and demyelination. It has been proposed that these decreases may result directly from a reduction in the inWltration of macrophages that are involved in myelin destruction. However, depletion of blood-borne macrophages with toxic liposomes prior to infection has little eVect on demyelination (Xue et al., 1999) . Large numbers of activated macrophages/microglia are still present in the CNS after liposome treatment suggesting that the activity of resident microglia may play a central role in demyelination. Macrophage depletion prior to infection does result in increased mortality, suggesting a role for macrophages in the control of viral titers prior to the development of the T-cell responses (Wijburg et al., 1997; Xue et al., 1999) . The role of antibody in viral clearance and the establishment of persistence has been studied in mice that lack a normal humoral immune response . Mice homozygous for a disruption in the immunoglobulin mu gene lack B cells. In these mice, virus is cleared normally during acute infection but then virus titers increase dramatically after about 11 days post-infection, when the titer of infectious virus in wild-type animals has fallen to undetectable levels. Passive transfer of MHV immune serum following initial virus clearance prevents this secondary rise in virus titers. These data demonstrate that initial clearance of virus is largely antibody independent but that antibody may play a key role in controlling the reemergence of virus in the CNS during persistent infection. The mechanisms of demyelination during MHV infection are not yet entirely clear. A priori, the potential mechanisms involved in MHV-induced demyelination might be similar to those involved in TMEV-induced demyelinating disease-that is, direct virus lysis of infected oligodendrocytes, virus-speciWc immune-mediated destruction of oligodendrocytes and myelin, macrophage-mediated ''bystander'' damage to oligodendrocytes and myelin, and myelin damage by ''epitope spreading'' or ''molecular mimicry.'' Several studies have shown that virus actively replicates in oligodendrocytes and that morphologic and biochemical changes accompany oligodendrocyte infection in vivo (Barac-Latas et al., 1997; Powell et al., 1975) . Down-regulation of oligodendrocyte gene expression along with oligodendrocyte destruction by necrosis and apoptosis have been reported for both the early and late stages of demyelinating disease (Barac-Latas et al., 1997) . These types of observations have supported the suggestion that direct viral cytolysis might play a role in the development of demyelination. However, it seems clear that an immune component is essential for the full development of demyelinating disease. JHM-MHV infection of severe combined immunodeWciency (SCID) mice, which lack T lymphocytes, does not result in either demyelination or paralysis (Houtman and Fleming, 1996) . Similarly, mice deWcient in recombinase-activating gene (RAG) activity lack both Band T-cell function and do not develop demyelination after MHV infection (Wu and Perlman, 1999) . Infected RAG mice, whose immune system has been reconstituted by the adoptive transfer of spleen cells from genetically identical immunocompetent mice that have been immunized for MHV, developed demyelination within 7 to 9 days after transfer. Demyelination rarely occurs if donor mice have not been preimmunized against MHV, demonstrating that MHV-speciWc T cells are critical for the development of demyelination. The demyelination that follows reconstitution of immune function is characterized by the extensive recruitment of activated macrophages and microglia to sites of demyelinating lesions, consistent with a potential role for these cells in the development of myelin damage. These observations are also consistent with the idea that RANTES, secreted by CD4 þ cells, plays a key role in the recruitment of macrophages to lesions and that activated macrophages and microglia may be primary mediators of myelin destruction (Lane et al., 2000) . A mechanism of CD4 þ -dependent, nonspeciWc damage mediated by inWltrating macrophages bears strong resemblance to the ''bystander'' damage induced by the development of a DTH response in the TMEV model. Although these data suggest a signiWcant role for the immune system in the development of demyelination, a recent report on the infection of RAG mice with MHV-A59 demonstrated typical demyelinated lesions (Matthews et al., 2002) . T cells may therefore play a central role in the development of demyelination but other mechanisms may also be important, and the relative contributions of speciWc mechanisms may depend on variables such as virus strain and the host genetic background. The question of axonal damage after MHV infection is separate from the question of demyelination. MHV infection is generally characterized by demyelination with axonal sparing, but axonal injury at sites of demyelination has been described (Stohlman and Weiner, 1981) . In contrast to the TMEV model, where neurologic deWcits tend to occur well after primary demyelination is established, most studies have shown that MHV induced paralysis and demyelination occur almost simultaneously, indicating that axonal damage occurs early in disease. In the experiments described earlier, animals that lack T cells, such as RAG and SCID mice, do not develop demyelination, but reconstitution of these animals by the adoptive transfer of spleen cells results in demyelination by 7 days post-transfer. Axonal damage appears with exactly the same pattern following MHV infection in these mice (Dandekar et al., 2001) . Using the appearance of nonphosphorylated neuroWlament as a marker for axonal damage, the onset of axonal damage was shown to be coincident with the appearance of demyelination. Damaged axons were detected not only at sites of active demyelination but also in areas with heavy macrophage inWltration but no obvious demyelination. Axonal damage therefore occurs very early in the disease process and nonspeciWc inXammation may play a key role. In the TMEV model, the secondary development of a pathogenic immune response directed against myelin antigens or ''epitope spreading'' has been proposed as an important mechanism for the exacerbation and maintenance of demyelination. JHM-MHV infection of either Brown Norway or Lewis rats results in a subacute demyelinating disease. Infection of Lewis rats also results in lymphocytes that proliferate in response to myelin basic protein and that produce mild encephalitis upon adoptive transfer into naïve recipients (Watanabe et al., 1983) . Lymphocytes from infected Brown Norway rats did not develop MBP sensitivity, demonstrating a strong host strain dependence for this response (Watanabe et al., 1987) . The production of pathogenic anti-myelin antibodies demonstrates potential for an epitope spreading mechanism of MHV-induced disease, but the extent to which this mechanism contributes to the development of the chronic demyelinating disease is an unresolved issue. Antibodies, either autoreactive or anti-viral, might directly mediate demyelination by mechanisms such as antibody-dependent or complement-mediated cytotoxicity. In general, little correlation has been found between anti-MHV antibody titers and the severity of demyelination. The role of B cells and antibody has been studied directly in mice with mutations in the immunoglobulin mu gene; these mice lack B cells and make no antibody. In B-cell deWcient mice, demyelination following MHV infection is sign-iWcantly more severe than in controls at 30 to 60 days post-infection (Matthews et al., 2002) . This increase may be due to the inability of the animals to control virus after the acute phase of the disease. Animals that can produce antibody but lack antibody receptors (Fc) or components of the complement pathway are able to clear virus but still develop demyelination (Matthews et al., 2002) . These data indicate that antibody and antibody-dependent eVector mechanisms are probably not central to the demyelination process. Demyelination after MHV infection is clearly a complex event and probably results from some combination of direct viral destruction and immune-mediated damage. There does not, however, appear to be any single lymphocyte or monocyte function that is primarily responsible for the immune component of the damage. Both MHV-speciWc and nonspeciWc immune responses are probably involved. A determinant for central nervous system persistence localized in the capsid of Theiler's murine encephalomyelitis virus by using recombinant viruses DiVerential expression of H-2K and H-2D in the central nervous system of mice infected with Theiler's virus Targeting of IgM kappa antibodies to oligodendrocytes promotes CNS remyelination Viral load and a locus on chromosome 11 aVect the late clinical disease caused by Theiler's virus IdentiWcation of Theiler's virus infected cells in the central nervous system of the mouse during demyelinating disease FVB mice transgenic for the H-2Db gene become resistant to persistent infection by Theiler's virus Patterns of oligodendrocyte pathology in coronavirus-induced subacute demyelinating encephalomyelitis in the Lewis rat Serum and cells from Theiler's virus-infected mice fail to injure myelinating cultures or to produce in vivo transfer of disease. The pathogenesis of Theiler's virus-induced demyelination appears to diVer from that of EAE CD8-deWcient SJL mice display enhanced susceptibility to Theiler's virus infection and increased demyelinating pathology Inverted immunodominance and impaired cytolytic function of CD8 þ T cells during viral persistence in the central nervous system Two loci, Tmevp2 and Tmevp3, located on the telomeric region of chromosome 10, control the persistence of Theiler's virus in the central nervous system of mice The shiverer mutation aVects the persistence of Theiler's virus in the central nervous system Brain-inWltrating cytolytic T lymphocytes speciWc for Theiler's virus recognize H-2D b molecules complexed with a viral VP2 peptide lacking a consensus anchor residue Theiler's virus persists in glial cells during demyelinating disease Chromosome 14 contains determinants that regulate susceptibility to Theiler's virus-induced demyelination in the mouse Mapping loci inXuencing the persistence of Theiler's virus in the murine central nervous system Molecular modelling of protein-glycosaminoglycan interactions A murine virus (JHM) causing disseminated encephalomyelitis with extensive destruction of myelin A picornaviral protein synthesized out of frame with the polyprotein plays a key role in a virus-induced immune-mediated demyelinating disease Characterization of Theiler's murine encephalomyelitis virus (TMEV)-speciWc delayed-type hypersensitivity responses in TMEV-induced demyelinating disease: Correlation with clinical signs Theiler's murine encephalomyelitis virus (TMEV)-induced demyelinating disease in mice is inXuenced by the H-2D region: Correlation with TEMV-speciWc delayed-type hypersensitivity Monocytes/macrophages isolated from the mouse central nervous system contain infectious Theiler's murine encephalomyelitis virus (TMEV) Primary demyelination in Theiler's virus infection. An ultrastructural study Axonal damage is T cell mediated and occurs concomitantly with demyelination in mice infected with a neurotropic coronavirus Observations on Encephalitis of Mice Theiler's virus and Mengo virus induce cross-reactive cytotoxic T lymphocytes restricted to the same immunodominant VP2 epitope in C57BL/6 mice Failure of treatment with Linomide or oral myelin tolerization to ameliorate demyelination in a viral model of multiple sclerosis Experimental neuropathology of chronic demyelination induced by a JHM virus variant (DS) Theiler's virus infection of beta 2-microglobulin-deWcient mice Pathogenicity of antigenic variants of murine coronavirus JHM selected with monoclonal antibodies Comparison of the binding characteristics to BHK-21 cells of viruses representing the two Theiler's virus neurovirulence groups Strains from both Theiler's virus subgroups encode a determinant for demyelination Neurovirulence determinants of genetically engineered Theiler viruses A protein critical for a Theiler's virus-induced immune system-mediated demyelinating disease has a cell type-speciWc antiapoptotic eVect and a key role in virus persistence Class II-restricted T cell responses in Theiler's murine encephalomyelitis virus-induced demyelinating disease. VI. Potentiation of demyelination with and characterization of an immunopathologic CD4 þ T cell line speciWc for an immunodominant VP2 epitope Three-dimensional structure of Theiler virus Theiler's virus in brain cell cultures: Lysis of neurons and oligodendrocytes and persistence in astrocytes and macrophages Temperature-sensitive mutants of mouse hepatitis virus produce a high incidence of demyelination Mutation of predicted virion pit residues alters binding of Theiler's murine encephalomyelitis virus to BHK-21 cells Dissociation of demyelination and viral clearance in congenitally immunodeWcient mice infected with murine coronavirus JHM Molecular determinants of Theiler's murine encephalomyelitis-induced disease Theiler's virus growth in murine macrophage cell lines depends on the state of diVerentiation Mutations that aVect the tropism of DA and GDVII strains of Theiler's virus in vitro inXuence sialic acid binding and pathogenicity Prevalent class I-restricted T-cell response to the Theiler's virus epitope D b :VP2 121-130 in the absence of endogenous CD4 help, tumor necrosis factor alpha, gamma interferon, perforin, or costimulation through CD28 Preservation of motor function by inhibition of CD8 þ virus peptide-speciWc T cells in Theiler's virus infection InXuence of sex on susceptibility in the Theiler's murine encephalomyelitis virus model for multiple sclerosis Inhibition of Theiler's virus-mediated demyelination by peripheral immune tolerance induction A central role for CD4(þ) T cells and RANTES in virus-induced central nervous system inXammation and demyelination Persistence of mouse hepatitis virus A59 RNA in a slow virus demyelinating infection in mice as detected by in situ hybridization Determinant spreading and the dynamics of the autoimmune T-cell repertoire Demyelinative myelopathy in mice induced by the DA virus Theiler's virus replication in brain macrophages cultured in vitro Antibody prevents virus reactivation within the central nervous system Mouse hepatitis virus is cleared from the central nervous systems of mice lacking perforin-mediated cytolysis DiVerential generation of class I H-2D-versus H-2K-restricted cytotoxicity against a demyelinating virus following central nervous system infection Molecular characterization of a nondemyelinating variant of Daniel's strain of Theiler's virus isolated from a persistently infected glioma cell line Coexpression of class I major histocompatibility antigen and viral RNA in central nervous system of mice infected with Theiler's virus: A model for multiple sclerosis Characterization of the inXammatory response in the central nervous system of mice susceptible or resistant to demyelination by Theiler's virus Theiler's virus infection in mice: An unusual biphasic disease process leading to demyelination Characterization of the TO strains of Theiler's mouse encephalomyelitis viruses The 5' noncoding sequences from a less virulent Theiler's virus dramatically attenuate GDVII neurovirulence Chronic neurologic disease in Theiler's virus infection of SJL/J mice Theiler's virus-induced demyelination: Prevention by immunosuppression Susceptibility of inbred mice to chronic central nervous system infection by Theiler's murine encephalomyelitis virus Genetic analysis of susceptibility to Theiler's virus-induced demyelinating disease in mice Attenuation of neurovirulence of Theiler's murine encephalomyelitis virus strain GDVII is not suYcient to establish persistence in the central nervous system The predominant virus antigen burden is present in macrophages in Theiler's murine encephalomyelitis virus-induced demyelinating disease Theiler's murine encephalomyelitis virus group includes two distinct genetic subgroups that diVer pathologically and biologically The controversy surrounding the pathogenesis of the multiple sclerosis lesion Evidence for a ''dying-back'' gliopathy in demyelinating disease Three-dimensional structure of Theiler murine encephalomyelitis virus (BeAn strain) The structure of a highly virulent Theiler's murine encephalomyelitis virus (GDVII) and implications for determinants of viral persistence Neither B cells nor T cells are required for CNS demyelination in mice persistently infected with MHV-A59 Axonal loss results in spinal cord atrophy, electrophysiological abnormalities and neurological deWcits following demyelination in a chronic inXammatory model of multiple sclerosis Quantitative assessment of neurologic deWcits in a chronic progressive murine model of CNS demyelination Functional evidence for epitope spreading in the relapsing pathology of experimental autoimmune encephalomyelitis IdentiWcation of a locus on mouse chromosome 3 involved in diVerential susceptibility to Theiler's murine encephalomyelitis virus-induced demyelinating disease Central nervous system remyelination clinical application of basic neuroscience principles Spontaneous CNS remyelination in beta 2 microglobulin-deWcient mice following virus-induced demyelination Class II-restricted T cell responses in Theiler's murine encephalomyelitis virus (TMEV)-induced demyelinating disease. III. Failure of neuroantigen-speciWc immune tolerance to aVect the clinical course of demyelination Persistent infection with Theiler's virus leads to CNS autoimmunity via epitope spreading Spontaneous remyelination following extensive demyelination is associated with improved neurological function in a viral model of multiple sclerosis CD4(þ) and CD8(þ) T cells make discrete contributions to demyelination and neurologic disease in a viral model of multiple sclerosis Myelin-speciWc tolerance attenuates the progression of a virusinduced demyelinating disease: Implications for the treatment of MS Relationship of Theiler's murine encephalomyelitis viruses to the cardiovirus genus of picornaviruses The immune system preferentially clears Theiler's virus from the gray matter of the central nervous system Short-term treatment with interferon-alpha/beta promotes remyelination, whereas long-term treatment aggravates demyelination in a murine model of multiple sclerosis Absence of spontaneous central nervous system remyelination in class II-deWcient mice infected with Theiler's virus Theiler's virus persistence and demyelination in major histocompatibility complex class II-deWcient mice Multiple sclerosis clinical trials: Old and new challenges Multiple sclerosis Cytotropism of Theiler's murine encephalomyelitis viruses in oligodendrocyte-enriched cultures Molecular cloning and sequence determination of the DA strain of Theiler's murine encephalomyelitis viruses Pathology of infection with the JHM virus IFN-gamma is required for viral clearance from central nervous system oligodendroglia Contributions of Fas-Fas ligand interactions to the pathogenesis of mouse hepatitis virus in the central nervous system DiVerential pathogenesis between mouse strains resistant and susceptible to Theiler's virus-induced demyelination Major histocompatibility complex-conferred resistance to Theiler's virus-induced demyelinating disease is inherited as a dominant trait in B10 congenic mice Insights into Theiler's virus neurovirulence based on a genomic comparison of the neurovirulent GDVII and less virulent BeAn strains Analysis of the complete nucleotide sequence of the picornavirus Theiler's murine encephalomyelitis virus indicates that it is closely related to cardioviruses Three-dimensional model of the capsid proteins of two biologically diVerent Theiler virus strains: Clustering of amino acid diVerence identiWes possible locations of immunogenic sites on the virion Oligodendrocytes and their myelin-plasma membrane connections in JHM mouse hepatitis virus encephalomyelitis Class I-deWcient resistant mice intracerebrally inoculated with Theiler's virus show an increased T cell response to viral antigens and susceptibility to demyelination Heparan sulfate mediates infection of high-neurovirulence Theiler's viruses Absence of neurological deWcits following extensive demyelination in a class I-deWcient murine model of multiple sclerosis Virus-induced demyelination in mice: ''dying back'' of oligodendrocytes Demyelination induced by Theiler's virus: InXuence of the H-2 haplotype Abrogation of resistance to Theiler's virus-induced demyelination in H-2 b mice deWcient in beta 2-microglobulin Partial suppression of Theiler's virusinduced demyelination in vivo by administration of monoclonal antibodies to immune-response gene products (Ia antigens) Susceptibility to Theiler's virus-induced demyelination. Mapping of the gene within the H-2D region Persistent infection of oligodendrocytes in Theiler's virus-induced encephalomyelitis InXuence of deletion of T cell receptor V beta genes on the Theiler's virus model of multiple sclerosis Theiler's murine encephalomyelitis: A model of demyelination and persistence of virus Role of T cell receptor V beta genes in Theiler's virus-induced demyelination of mice The balance between persistent virus infection and immune cells determines demyelination Pathogenesis of early and late disease in mice infected with Theiler's virus, using intratypic recombinant GDVII/DA viruses Theiler's virusassociated antigens on the surfaces of cultured glial cells DA strain of Theiler's murine encephalomyelitis virus induces demyelination in nude mice Mechanism of Theiler's virus-induced demyelination in nude mice Role of macrophages during Theiler's virus infection Viruses and autoimmunity: An aVair but not a marriage contract Picornaviridae: The Viruses and Their Replication Characteristics of cloned cerebrovascular endothelial cells following infection with Theiler's virus. II. Persistent infection Role of viruses in the etiopathogenesis of multiple sclerosis Infection of macrophages by Theiler's murine encephalomyelitis virus is highly dependent on their activation or diVerentiation state Clinical viral infections and multiple sclerosis Central-peripheral distal axonopathy: The pathology of dyingback polyneuropathies Chronic central nervous system demyelination in mice after JHM virus infection CTL eVector function within the central nervous system requires CD4 þ T cells Mouse hepatitis virus-speciWc cytotoxic T lymphocytes protect from lethal infection without eliminating virus from the central nervous system Activation of astrocytes in the spinal cord of mice chronically infected with a neurotropic coronavirus L* protein of the DA strain of Theiler's murine encephalomyelitis virus is important for virus growth in a murine macrophage-like cell line Crystallization and preliminary X-ray diVraction studies of Theiler's virus (GDVII strain) Axonal transection in the lesions of multiple sclerosis High numbers of viral RNA copies in the central nervous system of mice during persistent infection with Theiler's virus Polyreactive antibodies to glatiramer acetate promote myelin repair in murine model of demyelinating disease Human monoclonal antibodies reactive to oligodendrocytes promote remyelination in a model of multiple sclerosis Adoptive transfer of EAE-like lesions from rats with coronavirus-induced demyelinating encephalomyelitis Comparative analysis of coronavirus JHM-induced demyelinating encephalomyelitis in Lewis and Brown Norway rats The eVect of L3T4 T cell depletion on the pathogenesis of Theiler's murine encephalomyelitis virus infection in CBA mice Role of spleen macrophages in innate and acquired immune responses against mouse hepatitis virus strain A59 Receptor for mouse hepatitis virus is a member of the carcinoembryonic antigen family of glycoproteins EVective clearance of mouse hepatitis virus from the central nervous system requires both CD4 þ and CD8 þ T cells Characterization of brain-inWltrating mononuclear cells during infection with mouse hepatitis virus strain JHM Virusspecific intracytoplasmic inclusions in mouse brain produced by a newly isolated strain of Theiler virus. I. Virologic and morphologic studies Macrophage inWltration, but not apoptosis, is correlated with immunemediated demyelination following murine infection with a neurotropic coronavirus Depletion of blood-borne macrophages does not reduce demyelination in mice infected with a neurotropic coronavirus Infectious origins of, and molecular mimicry in, Guillain-Barre and Fisher syndromes Chimeric cDNA studies of Theiler's murine encephalomyelitis virus neurovirulence Astrocytes, not microglia, are the main cells responsible for viral persistence in Theiler's murine encephalomyelitis virus infection leading to demyelination Role of sialyloligosaccharide binding in Theiler's virus persistence Sialylation of the host receptor may modulate entry of demyelinating persistent Theiler's virus