Theiler's Virus Infection of the Mouse, or: Of the Importance of Studying Animal Models

Virology 301, 1–5 (2002) doi:10.1006/viro.2002.1607

MINIREVIEW Theiler’s Virus Infection of the Mouse, or: Of the Importance of Studying Animal Models Michel Brahic Unite´ des Virus Lents, URA CNRS 1930, Institut Pasteur, 75724 Paris Cedex 15, France Received May 28, 2002; accepted June 5, 2002

To virologists interested in solving problems of pathogenesis, Theiler’s virus has it all. A natural pathogen of mouse that can be found in feral populations worldwide and in laboratory colonies kept under poor hygiene conditions, it is the infectious component of a multifactorial disease that resembles multiple sclerosis (MS). Theiler’s virus is a picornavirus with a life cycle very similar to that of poliovirus. It is transmitted by the fecal–oral route and multiplies in the gut without causing overt symptoms. On rare occasions it reaches the central nervous system (CNS) where it may persist in glial cells of the white matter of spinal cord and cause chronic inflammation and demyelination. Whereas Max Theiler used his virus in the 1930’s as a model for poliomyelitis, present day virologists, following the pioneering work of Daniels, Pappenheimer Richardson, and then Howard Lipton, are more interested in the demyelinating disease and its similarities to MS. Indeed, the infection of the SJL/J mouse by Theiler’s virus and experimental autoimmune encephalomyelitis are the most widely used MS models. The power of the model stems from the fact that the agent is very well characterized, genetically simple and amenable to reverse genetics, and that it infects the mouse, the ideal laboratory animal to study infections and immune-mediated pathologies. The pathogenesis of this disease raises two main questions: (i) How does the agent persist in spite of the immune responses? (ii) What is the mechanism of demyelination? I will examine some aspects of both, from a personal viewpoint that may raise more questions than it will provide answers. Finally, I will emphasize how preventing and treating human infectious diseases can still benefit enormously from studying animal models.

virus goes from there to the CNS, although there is evidence that it uses the neural rather than the hematogenous route. The reason is that all studies have focused on the experimental biphasic CNS disease that follows intracerebral inoculation. The first phase, also called early disease, is an acute meningoencephalomyelitis during which the virus infects neurons. It lasts approximately 2 weeks. The second phase, or late disease, is a life-long persistent infection of glial cells of the white matter of spinal cord accompanied by chronic inflammation and the destruction of myelin around axons. It is this late disease which is studied as a model of MS. There is some discrepancy in the literature regarding the nature of the glial cells that are persistently infected. However, the consensus is that, at least in SJL/J mice, the main viral burden is in infiltrating macrophages and possibly in activated microglial cells (Lipton et al., 1995; Pena Rossi et al., 1997). However, oligodendrocytes, and to some extent astrocytes, are also infected and may even play a key role in the establishment of persistence, as will be argued below. Interestingly, neurons are no longer infected during late disease. Although all mice are susceptible to initial infection and early disease, some clear the infection after approximately 2 weeks. Others, such as inbred SJL/J mice, remain infected for life and develop the late, MS-like, demyelinating disease. Susceptibility/resistance to persistence of the infection and to demyelination is genetically determined. The existence of inbred mouse strains with various degrees of susceptibility to persistent infection—determined by measuring the viral load in CNS— indicates that the trait is multigenic. Besides the H-2D class I gene, which has a major effect, several non-H-2 loci responsible for the extreme susceptibility to persistent infection of the SJL/J strain have been mapped and characterized, including one close to Ifng on chromosome 10 (for a review see Brahic and Bureau, 1998). Viral genetics has helped to study Theiler’s virus’s pathogenesis because of the existence of two groups of strains with strikingly different phenotypes. The majority

GENETIC APPROACHES TO THE PROBLEM OF PERSISTENCE Theiler’s virus’s primary habitat is the gut. Interestingly, virtually nothing is known about the infection at this level. We do not know which cells are infected and how the 1

0042-6822/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.

2

MINIREVIEW

of strains isolated so far, including strains DA and BeAn, cause the biphasic disease described above. However, the GDVII and FA strains are exceptions. They spread very efficiently among neurons, both in vivo and in vitro, and, as a result, kill most of the animals from acute encephalomyelitis. Importantly, they do not persist in rare survivors. Furthermore, attenuated GDVII strains, obtained by introducing mutations in the 5⬘-noncoding region of the genome, do not kill their hosts and, yet, do not persist, indicating that specific positive determinants of persistence are present in the genomes of the DA and BeAn strains (Jarousse et al., 1999; Lipton et al., 1998). Using series of recombinants between the two types of strains and mutant DA or BeAn viruses, several groups showed that only strains that infect white matter glial cells persist and that the ability to persist maps predominantly to the viral capsid, in particular to residues located around the edge of the “pit,” a depression which is thought to be the receptor-binding site (see Brahic and Bureau, 1998 for a review). Unfortunately, the receptor for Theiler’s virus has not yet been identified. However, we know that the presence of sialic acid at the surface of cells is necessary, but not sufficient, for the binding and entry of persisting strains, whereas it is not necessary in the case of the GDVII strain (Zhou et al., 2000). Therefore, the present working hypothesis is that nonpersistent and persistent strains use different receptors, or use the same receptor differently, such that the former infects neurons more efficiently than glial cells and, conversely, the latter infects glial cells more efficiently than neurons. Sparing neurons would allow the host to survive and infecting glial cells would allow the virus to persist. It can be seen that, according to this hypothesis, cell tropism is central to the problem of Theiler’s virus persistence. PERSISTENCE AND THE L* PROTEIN This viral protein is translated from an alternative reading frame starting 13 nucleotides downstream from the first AUG codon of the main viral ORF. It may play a role in persistence because it is expressed by all known persisting strains of Theiler’s virus but not by the GDVII and FA strains, and because it facilitates viral replication in macrophage cell lines (van Eyll and Michiels, 2000). However, conflicting results have been published regarding the persistence of mutant DA viruses which do not express L*. The discrepancy could be due to minor variations in the infectious DA clones used by different groups. Interestingly, the phenotypes of L* mutants was studied in SJL/J mice, a strain highly susceptible to persistence of the infection and whose macrophages have unusual properties. It might be interesting to reexamine this point using less susceptible inbred strains, or susceptible outbred strains such as the Swiss mouse. In conclusion, the biological function of L* is still an open question. The presence of the L* ORF in all persisting

strains argues strongly in favor of some selection pressure to maintain this gene. However, that selection pressure may not be related to persistence in CNS, which is most likely a dead end as far as virus transmission is concerned. The role of the L* protein in the biology of Theiler’s virus may have to be looked for somewhere else, maybe in the gut. PERSISTENCE IN MACROPHAGES AND THE ROLE OF OLIGODENDROCYTES Although infiltrating macrophages carry the main viral burden in persistently infected SJL/J mice, oligodendrocytes may have a key role in establishing persistence. This concept stemmed from an unexpected result obtained with mutant mice. The shiverer mutation is a deletion of a large part of the gene (Mbp) which codes for the myelin basic protein (MBP). rump shaker, on the other hand, is a point mutation in the Plp gene, which codes another myelin structural protein, the proteolipid protein (PLP). C3H mice homozygous for either mutation are strikingly resistant to persistent infection, whereas the parental C3H mice are susceptible. In both cases, resistance is not immune mediated. Interestingly, C3H mice heterozygous for the shiverer mutations are more susceptible to persistent infection than the parental strain (Bihl et al., 1997). These results tell us that myelin, or the oligodendrocyte cell body, plays a central role in allowing the virus to persist. (It has been claimed that the Mbp and Plp genes are expressed at low levels in macrophages. Because this is controversial, and for the sake of simplicity, we will ignore it in the rest of the discussion.) The resistance to persistent infection of the mutant mice is so profound that the step in pathogenesis affected by the mutations must be essential for persistence. What could that step be? Myelin is an extension of the oligodendrocyte cell body wrapped many times around the axons. The cytoplasm is extruded from these extensions during myelination, the so-called “compaction” phenomenon. As a result myelin is a multilamellar structure in which intraand extracellular spaces alternate. Some cytoplasm remains in specialized areas of myelin such as the paranodal loops and the Schmidt–Lanterman incisures. The MBP and PLP proteins are two main constituents of myelin that play important roles in compaction. As a result, myelin is virtually absent in shiverer mice, and abnormal in rump shaker animals, although the oligodendrocyte cell bodies are maintained in both mutants. One can therefore hypothesize that the infection of myelin, which is impossible in the shiverer mouse, is required for persistence. This is congruent with the fact that Theiler’s virus antigens have been found in the paranodal loops and the Schmidt–Lanterman incisures of infected oligodendrocytes (Rodriguez et al., 1983). Why should myelin infection be required for persis-

MINIREVIEW

tence? One view, suggested by Howard Lipton, is that macrophages, the main reservoir during persistence, become infected by phagocytizing infected myelin. Another possibility is that the virus, which uses axonal transport during early disease, infects the myelin sheath from the axon. From myelin, the infection could spread to the oligodendrocyte cell body due to cytoplasm continuity and to other glial cells and macrophages by the release of extracellular particles. Whatever the mechanism will turn out to be, testing such a model should uncover new aspects of picornavirus trafficking within the CNS. WHY PERSISTENCE IN SPINAL CORD AND NOT IN BRAIN? It is striking that Theiler’s virus, although it infects the gray matter of brain, persists in the white matter of spinal cord and not in that of brain. This peculiar distribution must reflect important differences in the biology of glia, possibly oligodendroglia, in brain and spinal cord. It is also possible that monocytes acquire different properties when they infiltrate the brain or the spinal cord and that they become permissive to viral infection only in the latter case. All these questions, which could have implications going far beyond the Theiler’s virus model, remain entirely unanswered and would deserve thorough investigation. WHY GUT AND CNS? Theiler’s virus, similar to several other picornaviruses, infects naive animals by the oral route, lives in the digestive tract for long periods of time without producing symptoms, and is shed in feces. This lifestyle, which ensures efficient perpetuation of the agent in the mouse population, must be the result of a long selection process. On the contrary, there is no obvious selection pressure that would have given the virus the ability to infect the CNS, whether acutely or persistently, since it is most likely a dead end for the virus, as far as transmission is concerned. It is not even a reservoir from which the virus might infect organs with an access to the outside, since no virus has been found in any organ besides CNS during persistent infection, even using highly sensitive RT-PCR techniques (Trottier et al., 2002). Interestingly, many other enteric viruses, including several enteroviruses, are neurotropic. This may indicate that cells in the gut share properties, including surface molecules, that can act as viral receptors, with cells in the CNS. These cells could be enterocytes or immune cells in Peyer’s patches, but another possibility, which has not been much investigated yet, is that enteric viruses replicate in the abundant neurons and glial cells which form the enteric nervous system (Gershon et al., 1993). Adaptation to the CNS, in this case, would have taken place at the periphery, in the “brain-like” system

3

present in the wall of the digestive tract to ensure peristalsis. PERSISTENCE AND AVOIDANCE OF THE INNATE IMMUNE RESPONSES The importance of innate immune responses is particularly well illustrated in the case of Theiler’s virus by the phenotype of mice with an inactivated gene for the interferon-␣/␤ receptor. These mice die of overwhelming encephalomyelitis a few days following intracranial inoculation. At the time of death, viral titers in CNS are extremely high (Fiette et al., 1995). Therefore, in a normal mouse, the virus replicates under very strong pressure from the interferon-␣/␤ response and one should expect that it has developed some sort of countermeasure. A recent article by van Pesch and co-workers demonstrates that the L protein, a 76 amino acid long protein corresponding to the N-terminus of the viral polyprotein, specifically inhibits the transcription of the genes for interferons alpha-4 and beta, the immediate-early interferons (van Pesch et al., 2001). This is most likely the reason the L protein is important for both the neurovirulence of the GDVII strain (Calenoff et al., 1995) and the persistence of the DA strain (van Pesch et al., 2001). Understanding the mechanism by which L prevents transcription of the immediate-early interferon genes could shed new light on the biology of the interferon response. PERSISTENCE AND AVOIDANCE OF THE SPECIFIC IMMUNE RESPONSES The antiinterferon activity of the L protein, although it contributes to Theiler’s virus pathogenicity, is not enough to allow the virus to evade all defense mechanisms and to persist. Indeed, the virus is cleared after the early disease in mice genetically resistant to persistence of the infection. A host of data, which I will not review in detail here, shows that clearance is mediated primarily by class I restricted CD8 ⫹ lymphocytes (for a review of the subject, see Monteyne et al., 1997). In mice susceptible to persistent infection, on the other hand, the balance between viral activity and host defenses tilts in favor of the former. Interestingly, the antiviral CTL response in the spleen of these mice is slow and weak compared to that of resistant mice (Dethlefs et al., 1997a). Therefore, it looks as if the outcome of early disease depended on a race between viral multiplication and CTL responses. However, this may be simplistic since an unpublished observation, quoted in a recent review, suggests that similar levels of virus-specific CD8 ⫹ T lymphocytes are present in the CNS of SJL/J and C57BL/6 mice (Kim et al., 2001). Interestingly, virus-specific CTL are H-2K restricted in SJL/J (Kim et al., 2001) and DBA2 mice (Pena Rossi et al., 1991), which are susceptible to persistent infection, whereas they are mostly H-2D restricted in resistant

4

MINIREVIEW

C57BL/6 mice (Dethlefs et al., 1997b). In C57BL/6 mice knock-out for the H-2D b gene, remaining H-2K b restricted CTL are unable to clear the virus and the mice remain persistently infected (Azoulay-Cayla et al., 2000). Why should H-2D b-, but not H-2K b-restricted CTL clear the infection? Although preliminary data indicate that peptide presentation, rather than other properties of the class I molecule such as pattern of expression, is the cause (Azoulay-Cayla et al., 2001), we are still a long way from a complete mechanistic explanation. Important information is still lacking. For example, most of the experiments on class I restriction of viral clearance have been performed on splenocytes, for technical reasons, although clearance takes place in CNS. In summary, important questions on immune evasion by Theiler’s virus remain entirely open. Solving them might have far-reaching implications for understanding resistance to viral infections and the for the design of effective vaccines. DEMYELINATION, ANTIVIRAL IMMUNE RESPONSES, AND EPITOPE SPREADING There is now considerable evidence that the demyelination observed in SJL/J mice persistently infected with Theiler’s virus is immune mediated, although part of it could be due to direct virus-induced injury to oligodendrocytes. Numerous studies, performed mainly in the laboratory of Stephen Miller, showed that virus-specific CD4 ⫹ T cells with a Th1 phenotype, for which dominant and subdominant epitopes have been identified, are involved in demyelination (Gerety et al., 1994). On the other hand, the presence of numerous myelin-laden activated macrophages in the lesions suggests that macrophages are effectors of myelin destruction. Based on a large body of data, Miller and his collaborators proposed the following scenario: T cells are recruited to the site of persistent infection and activated by the class II restricted presentation of viral epitopes at the surface of macrophages/microglial cells that could be infected or that could have ingested infected cells. Class II restricted presentation and the secretion of IL-12 promotes a Th1 response. Activated Th1 CD4 ⫹ T cells produce proinflammatory chemokines and cytokines which attract and activate more mononuclear inflammatory cells and resident microglia. In the end, myelin destruction is due to the chronic secretion of toxic factors, such as TNF␣ and nitric oxide, by activated inflammatory cells (innocent bystander mechanism). Class II restricted presentation of epitopes from the PLP and MOG myelin proteins, as well as PLP- and MOG-specific Th1 CD4 ⫹ T cells, can be demonstrated in the CNS of SJL/J mice several months after inoculation. Therefore it appears that, late into disease, autoimmune demyelination gets superimposed on the innocent bystander mechanism described above (Miller and Eagar, 2001). The appearance of autoimmunity as a result of

persistent infection and chronic inflammation is an important observation. However, the contribution of autoreactive Th1 cells to Theiler’s virus induced demyelination has not been assessed directly. Miller has argued convincingly that this autoimmune response arises through “epitope spreading” rather than “molecular mimicry.” Epitope spreading may require that myelin-loaded macrophages exit the CNS and migrate to lymph nodes where they could activate autoreactive Th1 cells. Autoimmunity is clearly not required for Theiler’s virus induced demyelination since demyelination is already there before autoreactive T cells can be demonstrated. Therefore the part played by autoimmunity in Theiler’s virus induced demyelination is not entirely clear at present. It would be important to devise a way to eliminate the infection after the onset of autoimmunity to see if demyelination can continue on a purely autoimmune mode. The experiment has not been possible yet for the lack of an efficient antiviral drug. It might be worth exploring the possibility of genetically engineering a virus that would persist in CNS for only a limited period of time or that could be “turned off” at will. MS is considered an autoimmune disease that results from particular infections in genetically susceptible individuals. Therefore the concept of epitope spreading is very relevant to the pathogenesis of this complex human disease. The fact that an infectious agent has not been yet convincingly associated with MS plaques could mean either that the methods used to look for agents were inappropriate or that the agent is no longer there by the time clinical disease has developed. OF THE IMPORTANCE OF STUDYING ANIMAL MODELS There is at present a strong tendency in medicine to encourage only research that deals with a specific, important human disease. This is supposedly because perfect animal models do not exist, and because new and more powerful techniques are constantly being developed and made available for clinical investigations. Although this may be true, clinical investigations, because one cannot manipulate the system and see how it reacts, tend to be mainly descriptive and are not conducive to uncovering new paradigms. Much remains to be learned about basic aspects of host/pathogen relationship and new, unexpected concepts will undoubtedly emerge from well-conceived experimental approaches of infections of laboratory animals, such as inbred mice. Concepts such as MHC restriction of immune responses, or the infectivity of prions proteins, came forth from such studies. More recently, studying the immune responses of insects has considerably enriched our understanding of innate immunity in mammals, a field that has been neglected for a long time. Because life thrives on diversity, one can be certain that a lot more remains in store which could have

MINIREVIEW

profound implications for human medicine. Furthermore, animal models bring big returns for relatively small investments, compared to clinical research in humans. Unfortunately, only a handful of laboratories are now engaged in research on viral diseases of laboratory animals and they have been the same ones, studying the same models, for more than a decade. Lets hope that new generations will soon join in and bring with them totally new systems with which they will surprise us with unexpected turns in the virus/host interplay. ACKNOWLEDGMENTS I thank Jean-Francois Bureau, Daniel Gonzalez-Dunia, Karla Kirkegaard, Ignacio Mena, Thomas Michiels, and Genevie`ve Milon for critical reading of the manuscript and for many helpful suggestions and Mireille Gau for secretarial help.

REFERENCES Azoulay-Cayla, A., Dethlefs, S., Pe´rarnau, B., Larsson-Sciard, E. L., Lemonnier, F. A., Brahic, M., and Bureau, J.-F. (2000). H-2D b⫺/⫺ mice are susceptible to persistent infection by Theiler’s virus. J. Virol. 74, 5470–5476. Azoulay-Cayla, A., Syan, S., Brahic, M., and Bureau, J. F. (2001). Roles of the H-2D b and H-K b genes in resistance to persistent Theiler’s murine encephalomyelitis virus infection of the central nervous system. J. Gen. Virol. 82, 1043–1047. Bihl, F., Pena-Rossi, C., Gue´net, J.-L., Brahic, M., and Bureau, J.-F. (1997). The shiverer mutation affects the persistence of Theiler’s virus in the central nervous system. J. Virol. 71, 5025–5030. Brahic, M., and Bureau, J.-F. (1998). Genetics of susceptibility to Theiler’s virus infection. Bioessays 20, 627–633. Calenoff, M. A., Badshah, C. S., Dal Canto, M. C., Lipton, H. L., and Rundell, K. (1995). The leader polypeptide of Theiler’s virus is essential for neurovirulence but not for virus growth in BHK cells. J. Virol. 69, 5544–5549. Dethlefs, S., Brahic, M., and Larsson-Sciard, E. L. (1997a). An early, abundant cytotoxic T-lymphocyte response against Theiler’s virus is critical for preventing viral persistence. J. Virol. 71, 8875–8878. Dethlefs, S., Escriou, N., Brahic, M., van der Werf, S., and LarssonSciard, E.-L. (1997b). Theiler’s virus and Mengo virus induce crossreactive cytotoxic T lymphocytes restricted to the same immunodominant VP2 epitope in C57BL/6 mice. J. Virol. 71, 5361–5365. Fiette, L., Aubert, C., Mu¨ller, U., Huang, S., Aguet, M., Brahic, M., and Bureau, J.-F. (1995). Theiler’s virus infection of 129Sv mice that lack the interferon ␣/␤ or interferon ␥ receptors. J. Exp. Med. 181, 2069– 2076. Gerety, S. J., Rundell, M. K., Dal Canto, M. C., and Miller, S. D. (1994). Class II-restricted T cell responses in Theiler’s murine encephalo-

5

myelitis virus-induced demyelinating disease. VI. Potentiation of demyelination with and characterization of an immunopathologic CD4 ⫹ T cell line specific for an immunodominant VP2 epitope. J. Immunol. 152, 919–929. Gershon, M. D., Chalazonitis, A., and Rothman, T. P. (1993). From neural crest to bowel: Development of the enteric nervous system. J. Neurobiol. 24, 199–214. Jarousse, N., Viktorova, E. G., Pilipenko, E. V., Agol, V. I., and Brahic, M. (1999). An attenuated variant of the GDVII strain of Theiler’s virus does not persist and does not infect the white matter of the central nervous system. J. Virol. 73, 801–804. Kim, B. S., Lyman, M. A., Kang, B. S., Kang, H. K., Lee, H. G., Mohindru, M., and Palma, J. P. (2001). Pathogenesis of virus-induced immunemediated demyelination. Immunol. Res. 24, 121–130. Lipton, H. L., Pritchard, A. E., and Calenoff, M. A. (1998). Attenuation of neurovirulence of Theiler’s murine encephalomyelitis virus strain GDVII is not sufficient to establish persistence in the central nervous system. J. Gen. Virol. 79, 1001–1004. Lipton, H. L., Twaddle, G., and Jelachich, M. L. (1995). The predominant virus antigen burden is present in macrophages in Theiler’s murine encephalomyelitis virus-induced demyelinating disease. J. Virol. 69, 2525–2533. Miller, S. D., and Eagar, T. N. (2001). Functional role of epitope spreading in the chronic pathogenesis of autoimmune and virus-induced demyelinating diseases. In “Mechanisms of Lymphocyte Activation and Immune Regulation” (S. Gupta, Eds.), pp. 99–107. Kluwer Academic/Plenum Publishers, Dordrecht, The Netherlands. Monteyne, P., Bureau, J.-F., and Brahic, M. (1997). The infection of mouse by Theiler’s virus: From genetics to immunology. Immunol. Rev. 159, 163–176. Pena Rossi, C., Delcroix, M., Huitinga, I., McAllister, A., van Rooijen, N., Claassen, E., and Brahic, M. (1997). Role of macrophages during Theiler’s virus infection. J. Virol. 71, 3336–3340. Pena Rossi, C., McAllister, A., Fiette, L., and Brahic, M. (1991). Theiler’s virus infection induces a specific cytotoxic T lymphocyte response. Cell. Immunol. 138, 341–348. Rodriguez, M., Leibowitz, J. L., and Lampert, P. W. (1983). Persistent infection of oligodendrocytes in Theiler’s virus-induced encephalomyelitis. Ann. Neurol. 13, 426–433. Trottier, M., Schlitt, B. P., and Lipton, H. L. (2002). Enhanced detection of Theiler’s virus RNA copy equivalents in the mouse central nervous system by real-time RT-PCR. J. Virol. Meth. 103, 89–99. van Eyll, O., and Michiels, T. (2000). Influence of the Theiler’s virus L* protein on macrophage infection, viral persistence, and neurovirulence. J. Virol. 74, 9070–9077. van Pesch, V., van Eyll, O., and Michiels, T. (2001). The leader protein of Theiler’s virus inhibits immediate-early Alpha/Beta interferon production. J. Virol. 75, 7811–7817. Zhou, L., Luo, Y., Wu, Y., Tsao, J., and Luo, M. (2000). Sialylation of the host receptor may modulate entry of demyelinating persistent Theiler’s virus. J. Virol. 74, 1477–1485.