CMR FigSearch
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Oleszak, E. L.
Right arrow Articles by Platsoucas, C. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Oleszak, E. L.
Right arrow Articles by Platsoucas, C. D.

 Previous Article  |  Next Article 

Clinical Microbiology Reviews, January 2004, p. 174-207, Vol. 17, No. 1
0893-8512/04/$08.00+0     DOI: 10.1128/CMR.17.1.174-207.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.

Theiler's Virus Infection: a Model for Multiple Sclerosis

Emilia L. Oleszak,1,2* J. Robert Chang,3 Herman Friedman,4 Christos D. Katsetos,3,5 and Chris D. Platsoucas3

Department of Anatomy and Cell Biology,1 Fels Institute for Cancer Research and Molecular Biology,2 Department of Microbiology and Immunology,3 Temple University School of Medicine, and Department of Pediatrics (Neurology) and Pathology, St. Christopher's Hospital for Children and Department of Pediatrics, Drexel University College of Medicine, Philadelphia, Pennsylvania,5 Department of Medical Microbiology and Immunology, University of South Florida, Tampa, Florida4

SUMMARY
INTRODUCTION
    TMEV Subgroups and Different Strains of the Virus
    TMEV Infection: Early Acute Disease and Late Chronic Demyelinating Disease
    Determinants of Viral Persistence
TMEV-INDUCED DEMYELINATING DISEASE AND MULTIPLE SCLEROSIS
    Susceptibility of Mice to TMEV-Induced Demyelinating Disease and of Humans to MS is MHC Dependent
    Comparative Neuropathology of TMEV-Induced Demyelinating Disease in Mice and MS in Humans
        Neuropathology of TMEV-induced disease.
            (i) Inflammatory demyelination.
            (ii) Remyelination.
            (iii) Axonal damage.
        Neuropathology of MS.
            (i) Acute (fresh) lesions.
            (ii) Chronic active lesions.
            (iii) Chronic inactive lesions.
            (iv) ''Shadow plaques.''
            (v) Inflammatory demyelination.
            (vi) Oligodendroglial damage and abortive remyelination.
            (vii) Axonal damage.
            (viii) Vascular pathology and hypoxic-ischemic damage.
    Viral Etiology of MS
SUSCEPTIBILITY TO TMEV INFECTION IS GENETICALLY CONTROLLED
TMEV AND MACROPHAGES
ADHESION MOLECULES
CD4 AND CD8 T CELLS
CYTOTOXIC T LYMPHOCYTES
CLONALLY EXPANDED T CELLS ARE PRESENT IN THE CNS OF TMEV-INFECTED MICE AND IN THE CENTRAL NERVOUS SYSTEM OF PATIENTS WITH MULTIPLE SCLEROSIS
    TMEV-Infected Mice
    Oligoclonal T Cells Are Present in Brain Plaques from Patients with MS
COSTIMULATORY MOLECULES
    Role in the Development of TMEV-Induced Disease
    Role in MS
EPITOPE SPREADING, MOLECULAR MIMICRY, AND AUTOIMMUNITY
    TMEV-Induced Disease in Susceptible Mice
    MS in Humans
ROLE OF CYTOKINES
    TMEV-Induced Disease in Mice
        TGF-ß.
        IL-1, IL-6, and TNF-{alpha}.
        IFN-{gamma}.
        IL-2.
        IL-12.
        IL-10.
        IL-4.
    Cytokines in CNS Lesions of Patients with MS
ROLE OF NITRIC OXIDE IN TMEV-INDUCED DISEASE AND IN MULTIPLE SCLEROSIS LESIONS
APOPTOSIS OF T CELLS IN THE CENTRAL NERVOUS SYSTEM OF TMEV-INFECTED MICE AND IN PATIENTS WITH MULTIPLE SCLEROSIS
    Apoptosis of Infiltrating T Cells in the CNS of TMEV-Infected Mice
    Apoptosis of T Cells in Patients with MS
TMEV AND ANTIBODY RESPONSES
    IVIg Treatment of MS
CONCLUDING REMARKS
ACKNOWLEDGMENTS
REFERENCES

   SUMMARY
Top
 Next
 References
 
Both genetic background and environmental factors, very probably viruses, appear to play a role in the etiology of multiple sclerosis (MS). Lessons from viral experimental models suggest that many different viruses may trigger inflammatory demyelinating diseases resembling MS. Theiler's virus, a picornavirus, induces in susceptible strains of mice early acute disease resembling encephalomyelitis followed by late chronic demyelinating disease, which is one of the best, if not the best, animal model for MS. During early acute disease the virus replicates in gray matter of the central nervous system but is eliminated to very low titers 2 weeks postinfection. Late chronic demyelinating disease becomes clinically apparent approximately 2 weeks later and is characterized by extensive demyelinating lesions and mononuclear cell infiltrates, progressive spinal cord atrophy, and axonal loss. Myelin damage is immunologically mediated, but it is not clear whether it is due to molecular mimicry or epitope spreading. Cytokines, nitric oxide/reactive nitrogen species, and costimulatory molecules are involved in the pathogenesis of both diseases. Close similarities between Theiler's virus-induced demyelinating disease in mice and MS in humans, include the following: major histocompatibility complex-dependent susceptibility; substantial similarities in neuropathology, including axonal damage and remyelination; and paucity of T-cell apoptosis in demyelinating disease. Both diseases are immunologically mediated. These common features emphasize the close similarities of Theiler's virus-induced demyelinating disease in mice and MS in humans.


   INTRODUCTION
Top
 Previous
 Next
 References
 
Theiler's murine encephalomyelitis virus (TMEV or Theiler's virus) was first reported by Theiler in 1937 [461] and is a single-stranded RNA virus that belongs to the Picornaviridae family. TMEV is responsible for causing neurological and enteric diseases in susceptible strains of mice, such as SJL (reviewed by Dal Canto et al. [86], Monteyne et al. [299], Oleszak et al. [328], Tsunoda and Fujinami [475], and Lipton and Jelachich [250]).

TMEV is a member of the Cardiovirus genus. Its genome consists of single-stranded RNA of positive polarity (322, 341, 354) comprising approximately 8,100 nucleotides. The genomic organization of TMEV follows that of standard picornavirus genomic format (L-4-3-4). It codes for 12 proteins arranged in the order 5'-L, VP4, VP2, VP3, VP1, 2A, 2B, 2C, 3A, 3B, 3C, 3D-3'. The 76-amino-acid long L protein is a zinc-binding metalloprotein (74), but its exact function is not fully known. VP4, VP2, VP3, and VP1 are capsid proteins. Proteins 2A, 2B, 2C, 3A, 3B, 3C, and 3D are required, directly or indirectly, for viral RNA replication.

TMEV Subgroups and Different Strains of the Virus

Two major subgroups of TMEV have been reported, and they are distinguished primarily on the basis of their different neurovirulence, antigenicity, and other characteristics (86, 87, 234, 248). The first subgroup includes the GDVII and FA strains, which are extremely neurovirulent variants that induce only acute encephalitis and do not persist in the very few animals that survive the infection. The second subgroup is known as Theiler's original (TO) and includes the BeAn and DA strains. Members of the two subgroups, particularly in the GDVII, BeAn, and DA strains, have been sequenced and extensively characterized (127, 128, 246). Although the capsid proteins of the BeAn and DA strains have 93% amino acid homology, it is well established that the disease induced by the BeAn strain in SJL mice is different from the disease induced by the DA strain. Early acute disease (see below) is more attenuated in BeAn-infected mice in comparison to the distinct grey matter disease induced by the DA strain of TMEV (86, 87). Although both BeAn and DA strains induce late chronic demyelinating disease (see below), the kinetics of the disease caused by the two strains is also different. BeAn-infected SJL mice develop clear clinical signs, such as waddling gait and hind leg paralysis 30 to 40 days postinfection (p.i.) or 50 days p.i. (175), depending on the dose of the virus and the age of the animals. In contrast, DA-infected SJL mice develop such signs much later, at approximately 140 to 180 days p.i. The differences and similarities in the neurological disease induced by the DA and the BeAn strains of TMEV in SJL mice are summarized in Table 1.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Comparison of neurological disease induced by DA and BeAn strains of TMEV in SJL mice

 
TMEV Infection: Early Acute Disease and Late Chronic Demyelinating Disease

Intracranial (i.c.) inoculation of susceptible strains of mice with the DA strain induces biphasic disease, consisting of early acute disease, which occurs within 3 to 12 days p.i., followed by late chronic demyelinating disease, which develops at 30 to 40 days p.i. and eventually causes the death of the animals (reviewed by Dal Canto et al. [86, 88], Monteyne et al. [299], Oleszak et al. [328], Theiler [461], and Tsunoda and Fujinami [475]). In contrast, resistant strains of mice, such as C57BL/6 (B6), develop only early acute disease, clear the virus completely in about 3 weeks p.i., and do not develop late chronic demyelinating disease (245, 260).

Early acute disease resembles polioencephalomyelitis and is associated with replication of the virus in the central nervous system (CNS) gray matter (106 to 107 PFU/g of CNS tissue) and with destruction of neurons to a variable degree. During acute early disease, both susceptible and resistant strains of mice exhibit extensive mononuclear cell infiltrates in the CNS, consisting of T cells (of both CD4 and CD8 phenotypes), cells of the monocyte/macrophage lineage, few B lymphocytes, and few plasma cells (reviewed by Rodriguez et al. [384], Oleszak et al. [328], Begolka et al. [28], Drescher et al. [109], Murray et al. [308], and Pope et al. [356]). Early acute disease is not always clinically apparent (245, 260, 447, 461), and the severity of this phase of TMEV pathogenesis depends on the strain and the dose of the virus. Viral titers in TMEV-infected susceptible mice are greatly reduced by 12 days p.i. However, TMEV-infected susceptible strains of mice fail to completely clear the virus, which persists in monocytes/macrophages, microglia, astrocytes, and oligodendrocytes (77, 180, 364, 447, 475).

At 30 to 40 days p.i., TMEV DA-infected susceptible mice develop late chronic demyelinating disease with extensive demyelinating lesions of the white matter and mononuclear cell infiltrates in the spinal cord, consisting primarily of CD4+ and CD8+ T cells, some monocytes/macrophages, and few B cells and plasma cells (reviewed by Rodriguez et al. [384], Oleszak et al. [328], Begolka et al. [28], Drescher et al. [109], Murray et al. [308], and Pope et al. [356]). Late chronic demyelinating disease leads to progressive spinal cord atrophy and axonal loss with ensuing neurological deterioration (disruption in motor coordination, hind limb paralysis, spasticity, ataxia, and incontinence) (109, 206, 207, 250, 283, 284). Low-grade viral replication, at the level of 101 to 103 PFU/g of CNS tissue, has been well documented in macrophages, oligodendrocytes, and astrocytes in the white matter of the spinal cords of TMEV-infected SJL mice with late chronic demyelinating disease (109, 250, 475). This persistent infection has been observed in these mice for as long as 2 years p.i. (244). In BeAn-infected SJL mice, 20 to 30 copies of viral RNA/µg of total RNA from infected spinal cords were detected at 4 months p.i. by real-time reverse transcription-PCR (472). Few copies of viral RNA could still be detected in infected SJL mice at 1 year p.i. (472). It is not fully understood why the virus persists in the CNS of SJL mice whereas B6 mice are able to clear the infection. Possible mechanisms of viral persistence in SJL mice are discussed later in this review.

The immune responses to the virus of TMEV-infected (DA strain) sensitive (SJL) and resistant (B6) strains of mice are summarized in Fig. 1.



View larger version (60K):
[in this window]
[in a new window]
  		FIG. 1. Comparison of the immune responses to TMEV of sensitive (SJL) and resistant (B6) strains of mice infected i.c. with the virus. Differences in the immune responses of these strains of mice to TMEV are summarized during early acute disease (polioencephalomyelitis). Resistant strains of mice completely clear the virus and do not develop demyelinating disease. In contrast, sensitive strains of mice fail to completely clear the virus and develop persistent viral infection associated with low virus titers and late chronic demyelinating disease, characterized by massive mononuclear cell inflammatory infiltrates and demyelinating lesions.

 
Determinants of Viral Persistence

Viral persistence appears to be a prerequisite to developing late chronic demyelinating disease. However, different mechanisms may be involved in these two processes. We have provided evidence suggesting that persistence of TMEV in the CNS is not sufficient to produce demyelinating disease (331, 345). We have isolated two plaque size variants of the DA strain of TMEV (331). One variant produced small plaques, had a significantly higher growth rate at 37°C, and reached 130- to 500-fold higher titer at 37 than at 39°C. The other variant produced large plaques and had a lower growth rate. Although both plaque size variants and the DA strain were capable of establishing persistent infections, only the small-plaque variant and the DA strain were able to induce demyelinating disease in SJL mice (331). We have also reported persistent infection with the DA strain (106 to 108 PFU/ml) in the G26-20 glioma cell line in vitro (345). A TMEV variant isolated from the G26-20 glioma cells produced smaller plaques in vitro than did the wild-type DA TMEV strain. i.c. infection of SJL mice with this TMEV variant did not result in viral persistence or in late chronic demyelinating disease. However, total-body irradiation (300 rads) of SJL mice infected with this TMEV variant resulted in viral persistence without the development of late chronic demyelinating disease (345). These results demonstrate that different mechanisms may lead to the development of viral persistence and demyelinating disease.

Many studies have been carried out to identify the molecular determinants of persistence. Chimeric viruses between GDVII (an extremely neurovirulent variant which induces only acute encephalitis and does not persist) and DA or BeAn (which cause persistence and demyelination) strains have been constructed. The major finding of these studies is that the viral capsid plays a major role in persistence (3, 178, 280, 454). Within the capsid, several amino acids have been identified as being involved in establishing persistent infection (178, 409, 426, 524). It has been suggested that persistence depends on a conformational determinant within the capsid requiring homologous sequences in both the VP2 puff and VP1 loop, which are in close contact on the virion surface (3, 252, 477). These residues, located around the edge of the "pit," may be in close proximity to a putative receptor-binding site (47). However, conflicting results have been obtained about whether these determinants for persistence and demyelination can be contributed only by the DA and BeAn strains (60, 127, 280, 392) or whether these determinants are also present within the capsid of GDVII virus (3, 127, 392). In general, a chimeric GDVII virus, whose capsid had been replaced by that of the DA or the BeAn strain, was attenuated whereas a recombinant DA or BeAn virus with a GDVII capsid was virulent. However, mere attenuation of the neurovirulence of GDVII is not sufficient to established persistence (179, 252). The significance of conformational differences via interaction of VP2 puff B and VP1 loop II between GDVII and DA viruses has been recently illustrated by generation of specific mutants (477). Thus, the DA virus mutant with a puff B similar to that of GDVII induced both acute and late chronic demyelinating disease similar to that induced by wild-type DA, while the DapBL2M virus with VP1 loop II of GDVII and an additional mutation in VP2 puff B caused only prolonged gray matter disease without demyelination (477). Whether mutations within VP2 puff B and VP1 of the capsid modulate tropism of the virus by altering the affinity of the capsid for the receptor on different cells is not fully determined (186).

The major gene controlling viral persistence is the H-2D gene (55, 379, 386). Additional loci controlling viral persistence have been identified, and they are located close to the Ifng locus (in the telomeric region of chromosome 10). They include two loci designated Tmevp2 and Tmevp3 (37, 56). A third locus close to the locuse encoding myelin basic protein (MBP) (in the telomeric region of chromosome 18) has also been found (37, 56). Although Tmevp2 and Tmevp3 are found close to the Ifng locus, genetic analysis revealed that the IFN-{gamma} gene was excluded from the chromosomal regions containing the Tmevp2 and Tmevp3 loci (37). Additional studies have shown that the difference in the Th1-Th2 cytokine balance between TMEV-susceptible SJL mice and two lines of resistant congenic mice is not due to the Ifng locus. Therefore, the Ifng locus does not seem to be responsible for the differences in susceptibility between SJL mice and the resistant congenic mice (298).


   TMEV-INDUCED DEMYELINATING DISEASE AND MULTIPLE SCLEROSIS
Top
 Previous
 Next
 References
 
As discussed above, i.c. infection of susceptible strains of mice with TMEV results in biphasic disease of the CNS, consisting of early acute disease and late chronic demyelinating disease that appears at 30 to 40 days p.i. This late chronic demyelinating disease in the CNS of TMEV-infected susceptible strains of mice is one of the best, if not the best, experimental animal models of multiple sclerosis (MS). Close similarities between MS in humans and TMEV-induced late chronic demyelinating disease in susceptible strains of mice are discussed in the following section.

Susceptibility of Mice to TMEV-Induced Demyelinating Disease and of Humans to MS is MHC Dependent

Susceptibility of mice to TMEV-induced demyelinating disease is associated with the H-2D class I haplotypes s, q, r, p, and f (14, 22, 76, 78, 121, 251, 345, 361, 379, 380, 383, 385, 390). Susceptibility of humans to MS is associated with the DRBI*1501, DQA1*0102, and DQB1*0602 molecular haplotypes and either the DPB1*0401 or DPB1*0402 molecular haplotypes. These molecular haplotypes correspond to the cellular types DR2 and Dw2 (441, 444, 457, 485). Approximately 57% of Caucasian patients with MS are DR2+/Dw2+ whereas this cellular type is expressed in only 30% of the normal Caucasian population. The remaining patients with MS are DR2- Dw2- and represent a patient population that has not been sufficiently studied.

Comparative Neuropathology of TMEV-Induced Demyelinating Disease in Mice and MS in Humans

Neuropathology of TMEV-induced disease. As discussed above, TMEV infection of SJL mice results in early acute disease and late chronic demyelinating disease of the CNS. Early acute disease is characterized by variably intense and multifocal inflammation involving cerebral and spinal cord gray matter (247, 250, 386, 475). The inflammatory infiltrates are mononuclear and consist of lymphocytes (predominantly T cells) and monocytes/macrophages. Although there is lymphocytic infiltration of the leptomeninges and the cerebral cortex, the bulk of the inflammation is concentrated in the subcortical gray matter, especially in the regions of the diencephalon (thalamus, hypothalamus, and subthalamus), the hippocampus (stratum pyramidale), and the basal ganglia (pallidum and caudoputamen). In the spinal cord, the inflammation is concentrated predominantly in the anterior horns of the gray matter, although infiltration of the leptomeninges is also present. During early acute disease, the white matter is unaffected throughout the neuraxis.

The inflammatory mononuclear infiltrates have a distinctive perivascular predilection, in that they often traverse the walls of small and medium-sized parenchymal blood vessels, leading to vasculitis (243). Besides dense perivascular cuffing, there is evidence of sparse inflammatory infiltration in the nearby gray matter neuropil. A significant number of perivascular T lymphocytes exhibit apoptotic features (326). Foci of ischemic-type coagulative necrosis are detected in the diencephalon and hippocampus near dense perivascular inflammation, suggesting vasculitis (C. D. Katsetos, C. D. Platsoucas, and E. L. Oleszak, unpublished data).

As early as 30 days p.i., mononuclear inflammatory infiltrates of the spinal white matter consisting predominantly of T cells and monocytes/macrophages have been documented; they coincide with the onset of late chronic demyelinating disease. Histologically, demyelination is characterized by vacuolar change of the white matter, overt myelin loss, and appearance of myelin-laden phagocytic macrophages within the lesions. Axonal swellings (spheroids) are detected within demyelinating lesions during the advanced stages of chronic disease (283, 284). The demyelinating process is multifocal and involves anterior, posterior, and lateral columns. Although there is an apparent preferential susceptibility of the thoracic segments, demyelination is present throughout the rostrocaudal length of the spinal cord (386). The demyelinating lesions are accompanied by predominantly perivascular infiltrates consisting of lymphocytes and monocytes/macrophages (109, 250). There is widespread inflammatory infiltration of the spinal leptomeninges. Perivascular, leptomeningeal, and/or neuropil-infiltrating lymphocytes from the chronic lesions lack apoptotic features, in contrast to the prominent apoptosis of lymphocytes during early acute disease (326).

(i) Inflammatory demyelination. The immunological mechanisms of TMEV-induced demyelination are discussed elsewhere in this review. Briefly, both CD4+ and CD8+ T cells infiltrating the CNS of TMEV-infected mice contribute to demyelination. Autoimmune responses to myelin antigens generated during epitope spreading may also play a role in propagating the disease (see below). Additional immunological, metabolic, and toxic factors (discussed in the section on the pathogenesis of demyelination in human MS), acting on myelin sheaths, myelin-forming oligodendrocytes, and/or the axons themselves, may also be involved in TMEV demyelination (see below). A pattern of oligodendroglial damage, which is similar to that encountered in MS (174, 267), is a form of retrograde ("dying back") degeneration beginning in the most distal cell processes of oligodendrocytes near nerve axons in TMEV-infected mice (378, 386). A comparison of the neuropathological features of DA strain-induced neurological disease in mice (early acute and late chronic demyelinating disease) and human MS is presented in Table 2.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Aspects of comparative neuropathology

 
(ii) Remyelination. Variable degrees of remyelination are present within TMEV demyelinating lesions (284, 407). The frequency of medium or large remyelinated axons within a single lesion may be a powerful indicator of axonal preservation (407).

(iii) Axonal damage. In the advanced chronic phase of the demyelinating disease, that is, between 100 and 200 days p.i., there is a marked increase in spinal cord parenchymal atrophy without a concomitant increase in spinal cord demyelination (283). A morphological and ectrophysiologic study by McGavern et al. (283) has shown a statistically significant loss of medium-sized and large myelinated axons only after the demyelinating phase of the disease was established. The authors (283) speculate that following myelin denudation, the naked axons are vulnerable to further inflammation and ultimately succumb to secondary damage (283). This hypothesis is consistent with our findings that axonal swellings occur next to areas of inflammation in advanced-stage chronic demyelinating lesions (Katsetos et al., unpublished). It is unclear whether the axonal damage in TMEV infection is the result of direct (inflammatory) injury or of delayed Wallerian or dying-back type degeneration, accounting for the delayed development of axonal loss (283). In contrast, Tsunoda and Fujinami (474) have recently advanced the assertion that in TMEV-infected mice, axonal injury accompanied by oligodendrocyte apoptosis precedes demyelination, implying that axonal injury can trigger demyelination. The pathogenesis of axonal damage in experimental demyelinating disease should be the subject of further studies. Improvement of our understanding of the mechanisms of axonal damage may have potential therapeutic implications.

Neuropathology of MS. MS is presently defined as an inflammatory demyelinating disease of the CNS. MS lesions are heterogeneous and characterized by inflammation, demyelination, and variable degrees of axonal damage (228, 466). Historically, the morphological hallmark of the disease is the MS plaque; however, the MS plaque represents an end-point anatomical lesion. Accordingly, MS lesions should be viewed in the context of plaque evolution. MS plaques reflect a continuum of immunological activity encompassing variable degrees of inflammation together with new or residual neural damage and secondary reactive cellular changes in the affected brain tissue. Morphologically, plaques are grouped as acute (active), chronic active, chronic inactive, and "shadow plaque" types (157, 466). The cellular heterogeneity and the age of the plaques may be part of a temporospatial gradient of the same pathological process, or, conversely, it may reflect distinct morphological patterns of divergent immunological mechanisms (157, 228). Moreover, it should be emphasized that although the inciting pathological process may be fundamentally inflammatory, consistent with immune-mediated responses, the clinically defined neuropathological lesions seem complex and variegated, involving a combination of immunological, metabolic, and/or other neurotoxic pathways (227, 228).

(i) Acute (fresh) lesions. Acute lesions are the precursor lesions of the MS plaques. They are typified by perivascular infiltration of inflammatory cells (predominantly T lymphocytes and monocytes/macrophages), edema, myelin swelling, and activation of endothelial cells (157, 228, 360, 367, 466). Variable although often pronounced depletion of oligodendrocytes is present (340, 358, 359, 466). Plasma cells are infrequent. In general, there is preservation of axons, but variable degree of axonal damage may be present (466, 467).

(ii) Chronic active lesions. Chronic active lesions are by definition older lesions (plaques) exhibiting areas of active inflammation and demyelination, typically at their margins (the interface with normal tissue). They are characterized by perivascular lymphocytic infiltrates, ongoing myelin breakdown, myelin phagocytosis by foamy macrophages, reduction in the number of oligodendrocytes, and reactive astrocytosis. The lymphocytic infiltrates may extend beyond the margin of overt demyelination (157, 360, 367, 466). There is sparing of axons, but variable degrees of axonal damage may also be present (466, 467).

(iii) Chronic inactive lesions. Chronic inactive lesions are older, "burnt-out" or quiescent lesions, tantamount to glial scars. They are sharply delimited from the adjoining normal myelinated tissue. There is astrocytic gliosis, loss of oligodendrocytes, and variable axonal loss (157, 360, 367, 466, 467). Scant perivascular lymphocytes (cuffs), monocytes, and/or plasma cells may be focally persistent. The walls of the blood vessels may be sclerotic and hyalinized, denoting an antecedent inflammatory vascular injury. A thin rim of perivascular collagenous fibrosis may be present in long-standing lesions. There is disruption of the blood-brain barrier (223).

(iv) "Shadow plaques." Shadow plaques consist of variably sized and ill-defined zones of partially demyelinated or incompletely remyelinated tissue surrounding (and occasionally "overshadowing") the principal plaque (368). Their occurrence in chronic MS is unpredictable, ranging from absent to frequent. There are no known clinical correlates, but shadow plaques may underlie a distinctive pathogenetic pathway (157).

Inflammatory demyelinating lesions are typically disseminated throughout the CNS white matter, but are more frequently present in the optic nerves, brain stem, cerebellum (including the cerebellar peduncles), and spinal cord. Distinctive anatomical correlates are encountered in certain variants, such as the neuromyelitis optica or Devic type of MS (see below). As a rule, involvement of these frequently affected sites leads to neurological symptoms and signs. In the cerebral hemispheres, the lesions exhibit predominantly, but not exclusively, a periventricular distribution. Lesions involving convolutional white matter subjacent to the cerebral cortex typically spare subcortical myelinated fibers (U fibers). Occasionally, white matter lesions extend into the contiguous cortex or deep grey nuclei (basal ganglia and thalami) (reviewed by Prineas and McDonald [360]). Cortical involvement by MS lesions has been described as an additional contributor of neurological disability (353). Exceptionally, atypical solitary, mass-like lesions resembling tumors are encountered (202).

The vast majority of MS cases (~80 to 85%) manifest as the classical (Charcot) type, also known as remitting-relapsing and secondary progressive MS. The remitting-relapsing phase of the disease lasts for 10 to 15 years and is frequently followed by a phase of progressive neurological disability referred to as secondary progressive MS (reviewed by Trapp et al. [466]). The plaque evolution, described above, mirrors the protean activity of lesions in classical MS. Although inflammatory demyelination is undoubtedly a major component of the MS lesion, it does not explain the nature of the neurological disability, which seems to be related more to the degree of underlying axonal damage (213, 262, 466) (see below). Balo's concentric sclerosis is an "anatomical subtype" characterized by a distinctive topographic distribution of MS lesions in which bands of demyelinated white matter alternate with ribbons of unaffected white matter in a concentric fashion (82, 157). A minority of MS cases (~10%) exhibit an unremitting, albeit variably severe, course of clinical progression and are designated primary progressive MS (52, 185, 281).

The acute (Marburg) type of MS may be viewed as the fulminant end of the nosological spectrum of primary progressive MS. It is characterized by a rapidly and relentlessly progressive neurological deterioration, often leading to death within 1 year from the onset of the illness (157, 360). Although most cases arise and progress de novo, acute MS may also be superimposed on instances of remitting-relapsing MS (157). Myelinoclastic diffuse sclerosis (Schilder's disease) has certain similarities to and pathological features in common with the acute (Marburg) type of MS (334, 360) and is another relatively rare form of acute MS.

The neuromyelitis optica (Devic) type of MS is typified by combined involvement of the optic nerve and the spinal cord. The disease is often fulminant, leading to partial or total loss of vision and extensive spinal cord involvement. Besides demyelination, areas of confluent parenchymal necrosis, consistent with spinal cord ischemic infarction, are often present, culminating in atrophy and cavitation (cystic necrosis) of the affected segments of the spinal cord (157, 360). The active lesions are accompanied by inflammatory infiltrates, consisting predominantly of perivascular T lymphocytes and monocytes/macrophages (157). From the standpoint of preferential spinal cord involvement, Devic disease shows a remarkable similarity to the myelopathic pattern encountered in the late chronic demyelinating disease of TMEV infection.

(v) Inflammatory demyelination. Morphologically, the two integral components of MS lesions are perivascular inflammation and demyelination. It has long been hypothesized that inflammatory demyelination is the result of immune-mediated responses to myelin antigens either in the myelin sheaths of axons and/or at the level of myelin-forming oligodendrocytes (reviewed by Prineas and McDonald [360] and Lassmann [228]). Destruction of myelin and oligodendrocytes is not uniform in MS plaques (157, 360). In spite of decades of intensive research, the mechanism(s) of demyelination currently remains unresolved. However, there is clear evidence of early disruption of the blood-brain barrier and infiltration of the brain substance by blood-borne monocytes and T lymphocytes (228, 360, 466). Importantly, blood-brain barrier disruption relates to the onset of clinical symptoms, but a correlation between symptoms and inflammatory demyelination remains circumstantial.

Four fundamentally distinct patterns of demyelination (I to IV) have been proposed on the basis of myelin protein loss, topography and spatial extension of the lesions/plaques, patterns of oligodendrocyte damage, and immunopathological features of complement activation (264).

The current view is that inflammation composed predominantly of lymphocytes and monocytes/macrophages can cause demyelination by direct and/or indirect mechanisms. Lymphocytes contribute to the pathological process through cellular and humoral immunological responses (presumptive direct mechanisms) or by the production of cytokines (indirect mechanisms). Patterns I and II exhibit remarkable similarities to either T-cell-mediated or T-cell-plus antibody-mediated autoimmune encephalomyelitis (264). It is claimed that the other two patterns (III and IV) are consistent with a primary oligodendrocyte pathology (dystrophy) reminiscent of direct virus- or toxin-induced demyelination, as opposed to autoimmune mechanisms (264).

Monocytes/macrophages participate in the demyelinating process in a dual manner. Besides their traditional phagocytic role (ingestion and removal of myelin debris), cells of the monocyte/macrophage lineage, including blood-borne and activated/transformed resident microglia of the CNS, are potent effectors of axonal myelin and oligodendrocyte damage (466). Monocytes contribute to demyelination through the production of cytokines, nitric oxide, and proteases and/or by directly targeting oligodendrocytes at the border of MS lesions (352, 360, 466). Activated CNS resident microglia play a role in the early stages of demyelination through cell-to-cell contact with myelin internodes of the axons at the edges of active and chronic active MS lesions (reviewed by Prineas and McDonald [360] and Trapp et al. [466]).

Collectively, the formation and evolution of MS plaques require the interaction of complex immunological and metabolic factors including the effect of cytotoxic T cells, antibodies, toxic metabolites derived from activated monocytes/macrophages, and metabolic defects of oligodendrocytes (18, 227, 263, 264, 313). The pathogenesis of myelin destruction in MS appears to be complex and heterogeneous and reflects different patterns of demyelination. Furthermore, MS plaques may represent a common morphological end point of divergent immunological pathways involving myelin and axonal damage (263, 265, 348).

(vi) Oligodendroglial damage and abortive remyelination. Over the years, various theories have been advanced concerning the nature of oligodendrocyte damage in MS lesions. The prevailing, albeit circumstantial, view is that this damage is incurred through a variety of immunological mechanisms, including anti-myelin oligodendrocyte glycoprotein (MOG) antibodies, production of proinflammatory cytokines by monocytes/macrophages and lymphocytes, T-cell-mediated injury (through CD8+ class I major histocompatibility complex [MHC]-restricted cytotoxicity), immunoglobulins and components of activated complement, apoptosis, and a variety of other oligodendrogliotoxic factors (136, 227, 263, 313, 466). As mentioned above, patterns III and IV of demyelinating injury in MS are attributed to some form of oligodendroglial dystrophy (264). In pattern IV there is extensive degeneration and death of oligodendrocytes in the white matter around active lesions (264). In pattern III there is preferential loss of myelin proteins in the distal-most (periaxonal) cell processes of oligodendrocytes, which is associated with oligodendrocyte apoptosis (174, 227, 264). This distinctive mechanism has previously been defined as distal or dying-back (oligodenro)gliopathy (267). Lassmann has recently reappraised the latter pattern in the context of MS, suggesting that it resembles the oligodendroglial pathology incurred during early hypoxic-ischemic demyelination of the white matter (227).

In new MS lesions, both oligodendrocytes and myelin are actively destroyed, a process that in many cases ceases within few weeks. Remyelination frequently ensues following recruitment and repopulation of the plaque by oligodendrocytes (358-360). Evidence of oligodendrocyte precursor cells migrating toward demyelinated lesions has recently been found (69). As a rule, new lesions undergo variable degrees of remyelination, which is, however, either interrupted or confounded by recurrent activity (360).

There is clear ultrastructural evidence of attempted, but generally abortive, remyelination, which is particularly well illustrated in the paradigm of shadow plaques (360, 368). Why remyelination is not distributed uniformly within a lesion or why it is not seen in all cases of MS is not known. However, this adds to the notion that MS may be the clinicopathological end point of divergent pathogenetic mechanisms (157, 263, 340).

(vii) Axonal damage. Axonal damage in the form of axonal swellings and transection was described in the early literature but was underrated or dismissed as merely an epiphenomenon (reviewed by Kornek et al. [213]). In recent years there has been a critical reappraisal of axonal damage in MS, so much so that it has emerged as a major component of the disease (117, 466, 467). Importantly, axonal injury correlates with certain parameters of functional magnetic resonance imaging (MRI) (reduction of N-acetylaspartate) and neurological disability in MS (43, 213, 262, 263). Axonal pathology, evidenced by amyloid precursor protein staining, is more prominent in active MS lesions than in chronic inactive plaques (213). Because axonal injury is likely to be irreversible, early neuroprotection is a paramount therapeutic target.

Several unanswered questions are at issue. First, it is unclear whether axonal damage is secondary to or is sustained concomitantly with myelin damage (90), or whether in fact, it occurs independent of demyelination. Bitsch et al. (41) showed that axonal injury in MS, as determined by immunoreactivity for amyloid precursor protein, is partly independent of demyelination but relates to the number of monocytes/macrophages and CD8+ T cells in MS lesions. Second, although it appears that axonal pathology relates, in part, to the activity of the disease, its temporospatial profile needs further elucidation. Third, the pathogenesis of axonal damage is essentially unknown and may not be entirely immune mediated (263). Finally, the highly variable degree of axonal injury among MS patients may be consistent with the divergent pathogenetic mechanisms of MS.

(viii) Vascular pathology and hypoxic-ischemic damage. Another long-standing yet not universally embraced aspect of MS neuropathology is the distinctive angiocentric distribution of demyelinating lesions, particularly around postcapillary venules and veins (12, 148, 157, 268). Recently, Lassmann (227) revisited Putnam's hypothesis (363) of hypoxic-ischemic type injury as a component of MS lesions. Disturbances in oxidative metabolism resembling hypoxia-ischemia may be the consequence of vascular factors and/or the production of toxic metabolites typically associated with hypoxia-ischemia. Inflammatory damage of the vessel wall, endothelium, and blood-brain barrier by T cells and monocytes is tantamount to a vasculitic state, reminiscent of the angiocentric T-cell infiltrates in human immunodeficiency virus type 1-associated CNS disease (197). The latter, compounded by edema and disturbance of the cerebral microcirculation, may culminate in variably pronounced hypoxic-ischemic injury causing damage to myelin, axons, and oligodendrocytes (227).

Whereas overt ischemic damage has been demonstrated in severe cases of acute MS, Balo's concentric sclerosis, and neuromyelitis optica (82, 266, 490), it is thought that most active MS lesions may represent a form of "sublethal" hypoxic injury reminiscent of ischemic white matter damage. Evidence of hypoxia-like metabolic tissue injury in MS due to the liberation of excitotoxins and reactive oxygen species lends further support to this hypothesis (44, 93, 227, 256, 334). Whether hypoxic-ischemic injury is associated with any viral infection(s) of the CNS remains to be determined.

Viral Etiology of MS

Viral etiology and genetic background appear to play a substantial role in the susceptibility of mice to TMEV infection and of humans to MS. Genetic factors play a role in determining susceptibility to MS. It is well documented that about 10% of patients with MS have first- and second-degree relatives with MS (112). As previously discussed, the frequency of certain MHC class II genes is higher in certain patients with MS than in the general population in the same area (441, 444, 457, 485). However, the concordance rate in monozygotic twins is about 25%, clearly indicating that MS is not purely a genetic disease (reviewed in reference 412).

The role of environmental factors in MS has been suggested by numerous studies. Epidemiological studies that examined the migration of populations between low- and high-incidence zones indicate that there is a profound North-South gradient in disease incidence (39). Further, these studies strongly suggest that an infectious agent(s), possibly a virus(es), may be involved and that the infectious agent is likely to be acquired before 13 to 15 years of age (219). In addition, studies of MS "epidemics" occurring in areas with previously low incidence of MS, preceded by the introduction of new infectious agents (such as in the Faroe Islands and Sardinia), strongly argue for the role of an environmental factor(s) in triggering the disease (29, 218, 220-222, 395).

A number of viruses have been isolated from the CNS of patients with MS; however, it is not clear whether they are endogenous viruses of the CNS or pathogens triggering the disease. Paramyxoviruses are particularly interesting since the "epidemic of MS" in the Faroe Islands followed the introduction of canine distemper virus (CDV) into the island by British troops. It has been suggested that dogs infected with CDV may have transmitted the disease to humans. Patients with MS have higher titer of measles virus than does the control population (4). Simian virus-5 (another member of the Paramyxoviridae family), which infects both humans and dogs, has also been implicated in MS (119). The list of viruses associated with MS includes coronaviruses, retroviruses, endogenous retroviruses, and several members (Epstein-Barr virus [EBV] and human herpesvirus 6 and 7) of the herpesvirus family (118, 300, 404). Recently, virus-like structures have been isolated from the brains of patients with MS and from the brains of cats (498). However, it appear unlikely that MS is triggered by a single virus. Results obtained with viral experimental models suggest that many different viruses may trigger inflammatory demyelinating diseases resembling MS. Among these viruses are mouse hepatitis virus (a coronavirus), Semliki Forest virus (SFV) (an alphavirus), visna virus (a lentivirus), and CDV (a morbillivirus) (116). However, late chronic demyelinating disease induced by TMEV is an excellent model of MS because of its histopathological and immunological similarities as well as its similar genetic characteristics to MS.


   SUSCEPTIBILITY TO TMEV INFECTION IS GENETICALLY CONTROLLED
Top
 Previous
 Next
 References
 
Evidence has accumulated demonstrating that several loci in the mouse genome are responsible for the genetic control of susceptibility to TMEV infection, viral persistence, and the development of TMEV-induced late chronic demyelinating disease. The loci involved include the H-2D region of MHC class I, the Cß gene segment of the T-cell receptor (TCR), and a third locus mapped on chromosome 3 (22, 24, 47, 55, 458). Viral persistence is also genetically controlled, as discussed earlier in this review (37, 56). Welsh and coworkers (406, 494, 496) reported that the establishment of persistent infection in the CNS of TMEV-infected mice appears to be associated with the increased expression of H-2 class I on cerebrovascular endothelial cells. These cerebrovascular endothelial cells may be potentially responsible for presenting antigen to T cells (406, 494, 496). The genetics of TMEV infections has been reviewed elsewhere by ourselves (328, 384) and others (37, 38, 47, 86, 250, 299, 475) and is not discussed further in this review.


   TMEV AND MACROPHAGES
Top
 Previous
 Next
 References
 
Neurons serve as the major targets of infection for the GDVII and DA strains of TMEV during the early phase of TMEV-induced disease (polioencephalomyelitis) (21). However, during late chronic demyelinating disease, macrophages (77, 88, 253, 398) and, to a lesser extent, glial cells (77, 88, 253, 337) serve as the site of TMEV DA persistence. The presence of active viral replication in macrophages is indicated from (i) the immunohistochemical detection of virus associated with cytoskeletal changes in macrophages (88), (ii) the presence of a large TMEV antigen burden (253), and (iii) detection of the viral genome within macrophages (253). The localization of the virus in the cytoplasm of macrophages, but not in phagolysosomes, suggests infection, and not phagocytosis, as the mode of viral entry into macrophages (253).

Recent studies demonstrate that the presence of the L* viral protein allows the DA strain of TMEV to infect and persist in macrophages (320, 483). In vitro studies of TMEV-macrophage interactions show that TMEV preferentially infects activated macrophages (180, 181). Undifferentiated M1 cells were not susceptible to the BeAn strain of TMEV, whereas M1 cells treated with IFN-{gamma} became susceptible to TMEV infection as well as apoptosis (181). In addition, while microglia may play a role in TMEV pathogenesis, viral persistence seems to depend more on blood-borne macrophages. The depletion of blood-borne macrophages by treating SJL mice with dichloromethylene diphosphate leads to a loss of viral persistence as well as a lack of demyelination (398). Time course studies characterizing the persistently infected cells demonstrate that there is an increase in the number of infected cells that are F4/80 positive while the number of infected oligodendrocytes and astrocytes remain constant (398). These results, taken together, suggest that circulating macrophages cross the blood-brain barrier and serve as hosts for persistent TMEV DA infection. It is, however, not yet clear whether (i) the macrophages are sufficiently activated for TMEV infection before crossing the blood-brain barrier, or if such activation of macrophages occurs locally within the CNS, and (ii) what role the L* protein of TMEV-DA plays in infection and persistence.

While TMEV persistence contributes to demyelination, it is thought that autoimmune mechanisms are responsible for triggering the actual demyelinating process itself. The presence of activated macrophages (both infected and uninfected) found in demyelinating lesions suggest that these cells also contribute to the demyelinating process (357). These macrophages express MHC class II molecules and B7-1 and B7-2 costimulatory molecules (357), suggesting that they are presenting antigens to the CD4+ T- cells that are also in close proximity to the demyelinating lesions (357). Recent studies show that macrophages isolated from animals infected with TMEV are able to present to T cells self-antigens, such as proteolipid proteins, in addition to viral antigens (199), suggesting that macrophages can serve as a possible link between immune responses directed against the virus and self.

In addition to activating CD4+ T cells (357), macrophages and microglia are known to produce proteolytic factors that degrade MBP (257). While B6-derived macrophages and microglia infected with the BeAn strain of TMEV had no effect on MBP degradation, macrophages and microglia from SJL mice infected with BeAn released factors that degraded MBP (257). In addition, these factors were found to be cytotoxic to the E20.1 oligodendrocyte cell line (257). Activated macrophages also produce tumor necrosis factor alpha (TNF-{alpha}) (138, 455). In transgenic mice where TNF-{alpha} expression is induced in the CNS, apoptosis of oligodendrocytes is seen (7), while in TMEV-infected mice, TNF-{alpha} has been implicated in the loss of myelin (364). Not surprisingly, high levels of TNF-{alpha} have been found in TMEV-infected SJL mice during the chronic demyelinating disease phase by ourselves and others (70, 170). In addition, within the microenvironment of demyelinating lesions in TMEV-infected mice, the presence of foamy (myelin-laden) macrophages has been documented (327). Foamy macrophages are also found in demyelinating lesions of patients with MS (316, 334).


   ADHESION MOLECULES
Top
 Previous
 Next
 References
 
Interactions of lymphocytes with the endothelium are required for lymphocyte trafficking into the CNS. Adhesion molecules play a critical role in determining such interactions and are pivotal in lymphocytes infiltrating the CNS through the blood-brain barrier (103). One such molecule is the intercellular cell adhesion molecule 1 (ICAM-1), a glycoprotein that interacts with many ß2-integrins such as lymphocyte function-associated antigen-1 (LFA-1) (274) on T- cells and CD11b/CD18 (102) on monocytes. Endothelial cells (of the brain-blood barrier) can upregulate ICAM-1 expression upon stimulation by cytokines (189, 349, 464) and/or chemokines (50, 94). However, the involvement of ICAM-1 in TMEV pathogenesis is unclear. In one study, administration of anti-ICAM-1 antibody or simultaneous administration of anti-ICAM-1 antibody with anti-LFA-1 antibody has been shown to be effective in decreasing both the inflammatory response and the demyelinating disease in TMEV-infected mice (171). However, although anti-ICAM-1 antibody treatment has been shown to increase the severity of EAE after induction by adoptive transfer, anti-ICAM-1 did not significantly affect the development of TMEV-induced disease (397). More convincingly, while ICAM-1-/- mice (in the H-2b background) develop greater inflammation after TMEV infection than does the wild type, ICAM-1-/- mice are still able to clear the virus and refrain from developing the demyelinating disease (111).

Another important pair of adhesion molecules mediating the interactions between lymphocytes and endothelial cells is L-selectin (on lymphocytes) and E-selectin (on endothelial cells) (456). However, studies in L-selectin-/- mice showed that while the levels of CD8+ T-cells infiltrating the CNS decreased and the levels of B cell infiltrating the CNS increased, no significant changes in TMEV-induced disease with respect to viral persistence or demyelination was observed (518).

It should be noted that the above studies were focused on single adhesion molecules involved in the interactions of the blood-brain barrier and the pathogenic lymphocytes. Other adhesion molecules not yet studied in TMEV pathogenesis may account for the lymphocyte-endothelium interactions required for lymphocytic trafficking into the CNS. Trafficking mechanisms that depend on other adhesion molecules may be responsible for the results obtained with the adhesion molecule-deficient mice experiments described above (111, 518). For example, very late antigen 4/vascular cell adhesion molecule 1 (VCAM-1) interactions are thought to be pivotal for T-cell migrations into the CNS in experimental autoimmune encephalomyelitis (EAE) (511).

Immunohistochemical analysis of post mortem CNS samples from MS patients showed increased microglia-associated expression of VCAM-1 at the edges of active demyelinating lesions but not in inactive lesions (352). In addition, certain studies have reported an increased presence of soluble ICAM-1 (sICAM-1) in the blood and CSF of MS patients (282, 473), especially in patients with progressive MS (204). Although the significance of the increased expression of sICAM-1 is unknown, it appears that ICAM-1 alone can induce VLA-4 expression in T cells by signaling through CD11a (435). Therefore, although the role of sICAM-1 in MS is unclear, one can envision its involvement in both lymphocyte trafficking and activation.


   CD4 AND CD8 T CELLS
Top
 Previous
 Next
 References
 
Both CD4+ and CD8+ T cells appear to play important roles in the pathogenesis of both early acute disease and late chronic demyelinating disease induced by TMEV. Although the specificity of infiltrating T cells during early acute disease is against the virus, there is limited information about the antigen recognized by infiltrating T cells during late chronic demyelinating disease (190-192, 198, 199, 240, 241). These T cells may recognize host or viral antigens. Identification of the antigenic specificity or specificities of T cells infiltrating the CNS of mice with TMEV-induced late chronic demyelinating disease or of T cells infiltrating brain plaques of patients with MS is, in our view, the most important scientific problem that needs to be resolved in the pathogenesis of demyelinating diseases of the CNS. Both CD4 and CD8 cells contribute to myelin damage (389).

Several lines of evidence demonstrate that CD4+ T cells play an important role in the development of demyelinating disease: (i) treatment of TMEV-infected susceptible SJL mice with anti-CD4 (495) or anti-I-A antibodies (126, 382) resulted in decreased demyelinating disease; (ii) CD4+-mediated delayed-type hypersensitivity (DTH) to TMEV antigens has been associated with myelin damage (291); (iii) while the wild-type B6 mice clear the virus and are resistant to demyelinating disease, TMEV infection of CD4-/- B6 mice results in viral persistence and demyelinating disease (122, 308, 318); (iv) TMEV infection of susceptible SJL mice with genetically deleted CD4 molecules significantly increased the severity of demyelinating disease (308); and (v) the specificity of certain CD4+ T cells generated during the course of TMEV infection induced by either the BeAn or the DA strain of TMEV has been described. These T cells recognize three predominant viral peptides (VP1233-250, VP274-86, and VP324-37) (137, 509, 510). It is not fully understood how deficiency in class II gene products contributes to persistence of TMEV and the level of demyelination. The lack of "help" from CD4+ T cells could affect the generation of CD8+ T-cell responses. T-cell help by CD4+ cells is required for antibody production. It has been demonstrated that the levels of specific antiviral antibodies in the class II-deficient mice or mice treated with antibodies to class II gene products is very low in comparison to those in control/immunocompetent TMEV-infected mice (122, 126, 318, 382, 495). Lack of neutralizing antiviral antibodies in CD4+ T-cell-deficient mice may lead to neurological impairement and demyelinating disease by reduction of the clearance of the virus, resulting in an increase in the viral load in oligodendrocytes (122, 126, 318, 382, 495).

Also, several lines of evidence demonstrate that class I-restricted CD8+ T cells play an important role in the development of demyelinating disease: (i) depletion of CD8+ T lymphocytes using an anti-CD8 monoclonal antibody (MAb) greatly reduced myelin destruction in the CNS of TMEV-infected animals (393); (ii) the extent of demyelination was proportional to the expression of H-2 class I expression in the CNS of TMEV-infected mice (393); (iii) TMEV infection of class II-deficient mice results in demyelination (308); (iv) TMEV infection of resistant B6 mice with genetically deleted CD8 resulted in viral persistence and demyelinating disease, suggesting that at least for B6 mice, intact class I- and class II-restricted immune responses are essential for viral clearance (290, 376); (v) demyelination in resistant B6 mice deficient for either the CD4 or the CD8 molecule is preceded by deficient viral clearance (290, 308, 376); (vi) motor functions are preserved in TMEV-infected mice by blocking viral peptide-specific CD8+ T cells with free peptide (188); and (vii) CD8-deficient mice infected with TMEV (BeAn strain) displayed enhanced susceptibility to TMEV infection and increased pathological changes during demyelination (28). The requirement for CD8+ cytotoxic T lymphocytes (CTL) for clearing the virus may be responsible for these results. In contrast, TMEV DA infection of SJL susceptible strains of mice with genetically deleted CD8 molecule does not have any effect on the level of myelin damage (308).

It has been suggested that certain CD8+ T cells may also play a regulatory role in TMEV-infected SJL mice (27, 28, 121, 153, 361). Recent studies by Karls et al. (193) indicated that the susceptibility of BALB/cAnNCr mice to TMEV-induced demyelinating disease is due to the defective function of regulatory CD8+ T cells, which do not receive an activation signal from CD4+ T cells at early stages of infection.


   CYTOTOXIC T LYMPHOCYTES
Top
 Previous
 Next
 References
 
It appears that the development of viral persistence in the CNS of TMEV-infected mice precedes the development of demyelinating disease (308, 318). The differences in the clearance of the virus between TMEV-infected susceptible strains (such as SJL) and resistant strains (such as B6) leads to demyelinating disease in susceptible strains of mice and to complete recovery in resistant strains of mice. The mechanisms responsible for these differences are not fully understood. Lack of complete viral clearance in TMEV-infected SJL mice results in viral persistence and demyelinating disease which is immunologically mediated. TMEV-infected SCID mice survive TMEV-induced early acute encephalitic disease and do not develop demyelinating disease (389). However, reconstitution of these SCID mice with adoptive transfer of H-2-matched splenocytes from immunocompetent mice, results in the development of demyelinating disease in the CNS.

CTL appear to play an important role in the pathogenesis of TMEV-induced disease. Several lines of evidence are supporting the significance of CD8+ CTL in clearing TMEV, including the following: (i) resistance to demyelinating disease in B6 mice is determined by the H-2Db locus (22); (ii) passive transfer of CD8+ T lymphocytes from resistant TMEV-infected BALB/cByJ mice to susceptible BALB/cAnNcr animals protected the latter against demyelinating disease but only when the transfer of the cells was carried out early enough to allow clearance of TMEV; and (iii) DBA/2 mice can be protected from TMEV-induced demyelinating disease by in vivo administration of interleukin-2 (IL-2), a growth factor for T cells that may act by recruiting virus-specific CTL (226).

Resistant B6 mice are capable of generating a strong CTL response specific for the VP2 protein of TMEV as early as 3 days p.i. (96, 242, 244). A predominant H-2Db-restricted VP2121-130 viral peptide in BeAn- or DA-infected mice recognized by CTL has been identified (322, 354). Two other H-2Db-restricted TMEV epitopes (VP165-173 and VP110-120) recognized by CTL have been also described (270). In contrast, only a weak TMEV-specific CTL response is generated in the CNS of TMEV-infected SJL mice (96, 242, 244). Although the appearance of CD8+ T cells infiltrating the CNS of TMEV-infected SJL mice at 11 days p.i. has been reported (243), the function of these CD8+ T cells is not known. They may not be mature CTLs, and they may not exhibit cytolytic activity. TMEV-specific CTL activity appears much later in these mice and remains very low well into the late chronic demyelinating disease phase (up to 180 days p.i., the latest time p.i. when CTL activity has been tested [96]). Furthermore, TMEV-specific CTL precursor frequency in B6 mice has been reported to be relatively high, at 1 in 7,200 and at 1 in 9,000 splenocytes at 10 and 21 days p.i., respectively (96). In contrast, the frequency of TMEV-specific CTL precursors in susceptible SJL mice is low and has been reported to be 1 in 125,000 and 1 in 50,000 splenocytes at 10 and 21 days p.i., respectively (96). The activity of these CTL was determined using appropriate fibroblast cell lines (KSSV [H-2s] and C57SV [H-2b]) as target cells (241). These results demonstrate that anti-CTL cytotoxic responses appear considerably later in SJL than in B6 mice and remain low throughout the TMEV infection.

The mechanisms that are prohibiting SJL mice, in contrast to B6 mice, from generating a sufficient antiviral CTL response to clear the viruses are not fully understood. We have demonstrated (70) that during early acute disease, transforming growth factor ß (TGF-ß) transcripts were expressed in the brains of TMEV-infected SJL mice at levels 9 to 10 times higher than those found in the brains of TMEV-infected B6 mice. In addition, TGF-ß protein was found only in the CNS of TMEV-infected SJL mice but not in B6 mice. TGF-ß was produced by infiltrating mononuclear cells and was found primarily in the leptomeninges of the mesial temporal lobe. TGF-ß exhibits strong immunosuppressive properties, which are discussed in the cytokine section of this review (see below). TGF-ß is able to greatly inhibit the immune response (reviewed by Roberts and Sporn [377] and Kulkarni and Letterio [215]).

It should be mentioned that the failure to generate a CTL response sufficient to clear the virus in TMEV-infected SJL mice does not appear to be an intrinsic defect of SJL mice. SJL mice were capable of clearing SFV (also a model of virus-induced encephalitis and demyelination) (107, 297). It is not known whether SFV peptides are present in the context of H-2Ds or H-2Ks. SFV-infected B6 mice exhibited higher viral titers in the brain, developed more severe early acute disease, and produced lower levels of proinflammatory (Th1) cytokine transcripts (TNF-{alpha}, IL-6, and IL-1) and higher levels of anti-inflammatory (Th2) cytokine transcripts (IL-4) than did SFV-infected SJL mice (297).

Kang et al. (191) reported that SJL mice infected with the BeAn strain of TMEV in fact generated a CD8+ cytotoxic TMEV-specific T-cell response directed against the following three H-2Ks-restricted viral epitopes: a dominant epitope (VP3159-166) and two subdominant epitopes (VP111-20 and VP3173-181) of the BeAn virus capsid protein (191). The TMEV-BeAn strain is different from the TMEV DA strain, as discussed above (Table 1). The two strains of the virus were discussed in the Introduction. These TMEV-specific CTL produced IFN-{gamma}. A large number of 20-mer peptides with sequences overlapping the major capsid proteins of TMEV BeAn (L1, VP1, VP2, VP3, and VP4) were examined for IFN-{gamma} production by an enzyme-linked immunospot (ELISPOT) assay (191). It was determined that CD8+ CTL infiltrating the CNS of TMEV BeAn-infected SJL mice recognize the following three H-2Ks-restricted viral epitopes: VP111-20 VP3159-166, and VP3173-181 (191). These CTL have fully cytotoxic potential and lysed target cells pulsed with appropriate peptides. These T-cell responses were H-2Ks and not H-2Ds restricted (191).

In a separate report, Kang et al. (190) compared the avidity and the viral epitopes recognized by CD8+ T cells infiltrating the CNS of SJL mice infected with either the BeAn or the DA strain of TMEV. SJL mice infected with either TMEV BeAn or DA exhibited similar CD4+ T-cell responses to UV-inactivated TMEV BeAn or DA and to the major T-helper epitopes VP1233-250 VP274-86 and VP324-37 (137, 190, 509, 510). These VP1 and VP2 T-helper epitopes are identical between the two strains, whereas there is a single amino acid difference in VP3. CD8+ CTL infiltrating the CNS of TMEV BeAn-infected SJL mice recognize the following three H-2Ks-restricted viral epitopes: VP111-20, VP3159-166 and VP3173-181 (190). However, CD8+ CTL infiltrating the CNS of TMEV DA-infected SJL mice recognize only one, VP111-20, of these three H-2Ks-restricted viral epitopes. The peptide sequence differences between the BeAn and DA strains in two of the three predominant and intermediate epitopes do not permit the induction of CD8+ T cell responses in TMEV DA-infected SJL mice (190). Similar results were obtained with IFN-{gamma} determinations by ELISPOT assay and intracellular cytokine staining (190). Although the res