News and Publications

Division of Virology

Michael B.A. Oldstone, M.D., Division Head

The Division of Virology continues as a focus for intellectual enrichment and exchange with other scientists at TSRI who are studying animal and plant virology and viral pathogenesis. Our areas of interest are (1) developing transgenic models of human and plant disease; (2) studying the interactions between viruses and host cells; (3) determining the molecular basis of viral persistence; (4) analyzing interactions between viruses and immune cells that result in immunosuppression or autoimmunity; (5) mapping neural tracts and studying interactions between viruses and CNS cells; (6) analyzing mechanisms whereby viruses cause acute, chronic-degenerative, and demyelinating diseases; (7) developing antiviral therapies and vaccines; and (8) targeting drugs into the CNS by using viruses as carriers.

Viral-Immunobiology Laboratory

M.B.A. Oldstone, D. Berger, P. Borrow, W. Cao, J.C. de la Torre, C.F. Evans, J.E. Gairin, S. Gonzalez, A. Holz, D. Homann, R. Kiessling, K.J. Lee, H. Lewicki, D. Naniche, M. Manchester, I. Novella, J. Patterson, N. Sevilla, A. Tishon, M.G. von Herrath, M. Watanabe

The goal of the Viral-Immunobiology Laboratory is understanding the basic mechanisms by which viruses persist, escape immune recognition, and cause disease. Integral parts of this goal are understanding how viruses infect cells; defining the cellular receptors used by viruses, and mapping the travel of viruses into the cell and their subsequent uncoating, replication, assembly, exit, and spread. Because the immune system has evolved to recognize and abort these foreign substances, we are keen to evaluate the immune response against viruses, understand how viruses subvert this response to provide a selective advantage for their survival, and correct the immunodeficiency that occurs in persistent infections to allow termination of viral persistence.

Other major interests of this laboratory are determining how viruses alter differentiation of cells and thereby disturb homeostasis and how viruses induce autoimmune disease. Because different viruses have a number of unique life histories, we have focused our resources on 3 negative-stranded RNA viruses: lymphocytic choriomeningitis virus, measles virus, and Borna disease virus.

Uncoding the Cellular Receptor for Lymphocytic Choriomeningitis Virus and Lassa Fever Virus

W. Cao, P. Borrow, H. Lewicki, M.B.A. Oldstone

Viral infection starts with the specific attachment of the virus to a component of the cell surface. This interaction largely determines the tropism and host range of infection and plays a part in the overall pathogenesis of the virus. Lymphocytic choriomeningitis virus (LCMV) is a natural rodent virus capable of infecting a wide range of mammalian cell types. It is also the prototype virus for the arenavirus family, which includes Lassa fever virus, a major pathogen in humans.

A virus overlay protein blot assay indicated that a cell-surface glycoprotein (120--140 kD) interacts specifically with the purified LCMV strain Armstrong clone 13 or with Lassa fever virus. This glycoprotein was biochemically purified from a mouse fibroblast cell line and analyzed with mass spectrometry to determine the internal sequence. Peptides analyzed matched the sequence of -dystroglycan, the -subunit of dystrophin-associated proteins. This putative cellular receptor is abundant on cells susceptible to LCMV infection and sparse on cells refractory to infection with this virus, a situation that matches the tropism of LCMV.

In studies done in collaboration with K. Campbell, University of Iowa, and E. Rostov and S. Nichols, Centers for Disease Control and Prevention, placing purified -dystroglycan protein on membranes allowed specific binding of both LCMV and Lassa fever virus (Fig. 1). However, LCMV did not bind to the ₂-subunit of voltage-gated calcium channel, a control glycoprotein purified from the same cellular source. Soluble -dystroglycan at nanomolar concentrations specifically inhibited infection by LCMV. Disruption of the gene for -dystroglycan in stem cells ordinarily permissive to LCMV prevented LCMV binding and infection.

Using Antiviral Vaccination Strategy and Immunocytotherapy to Treat a Persistent Viral Infection

D.P. Berger, D. Homann, M.G. von Herrath, A. Tishon, M.B.A. Oldstone

Two basic ingredients allow persistent viral infection to occur: (1) an immune response that is ineffectual in recognizing and clearing virus or virus-infected cells and (2) one or more unique components or strategies of the virus that favor nonlytic replication. Viral persistence is an evasion of the host's immunologic surveillance system. Termination of persistent viral infection in the lymphocytic choriomeningitis virus (LCMV) model is totally dependent on MHC class I--restricted CD8⁺ cytotoxic T lymphocytes (CTLs) and CD4⁺ helper T cells.

In H-2^d mice, the CTL response to LCMV is directed to a single immunodominant domain of the viral nucleoprotein (amino acid residues 118--126, RPQASGVYM). In collaboration with A. Sette and R.W. Chesnut, Epimmune, La Jolla, California, we used a peptide consisting of the specific MHC I sequence RPQASGVYM conjugated to a general class II padre-helper epitope and a lipid tail to inoculate adult H-2^d mice persistently infected since birth with LCMV. The inoculated mice had a primary CD8⁺ MHC class I--restricted anti-LCMV CTL response. These CTLs recognized and lysed target cells infected with LCMV or coated with the nucleoprotein peptide but were unable to clear persistent LCMV infection.

Other mice persistently infected with LCMV were treated with combinatorial therapy that consisted of the antiviral agent ribavirin and repeated peptide inoculations. Ribavirin reduced the viral load, and repeated peptide injections led to the generation of specific anti-LCMV CD8⁺ CTLs. However, the level of virus in the serum and organs was not further decreased. Affinity analysis indicated that the CTLs in uninfected peptide-inoculated controls had high affinity for LCMV nucleoprotein, whereas the CTLs in persistently infected mice had low affinity for the protein. These data indicate that in mice persistently infected with virus in utero or at birth, complete tolerance does not occur and antiviral CTLs can be generated. However, the CTLs are of low affinity, and manipulation is required to increase the affinity to allow an opportunity for effector CTLs to clear the persistent infection.

Clearance of persistent LCMV can be accomplished by using immunocytotherapy. Clearance depends on MHC class I--restricted CD8⁺ CTLs and CD4⁺ helper T cells. The precise role of CD4⁺ T cells in maintaining CD8⁺ responses in chronic viral infections and the number of CD4⁺ T cells necessary for successful adoptive therapy were defined. We isolated donor splenocytes from control, uninfected mice (naive cells) or from mice immune to LCMV (primed cells), used magnetic beads or magnetic separation columns to deplete or enrich the splenocytes for various subpopulations of T cells, and then transferred the subpopulations of T cells to H-2^b mice persistently infected with LCMV.

Splenocytes depleted of CD4⁺ T cells (CD8⁺ T-cell preparations) consisted of 12% CD8⁺ cells and less than 0.5% CD4⁺ cells. Splenocytes enriched for CD4⁺ T cells consisted of 94% CD4⁺ cells, less than 0.01% CD8⁺ cells, and less than 0.1% B cells. Adoptive transfer of 3 x 10⁶ primed CD8⁺ T cells and various numbers of primed CD4⁺ T cells indicated that at least 4 x 10⁵ primed CD4⁺ T cells are necessary for sustained viral clearance. In contrast, if 3 x 10⁶ primed CD8⁺ cells and up to 3 x 10⁶ naive CD4⁺ T cells were given, clearance of persistent LCMV infection was not achieved. Other studies showed that depletion of B cells did not influence clearance of virus by CD8⁺ and CD4⁺ T cells.

In vitro stimulation of primed CD8⁺ T cells cultured with or without naive or primed CD4⁺ cells revealed different cytokine profiles. Secretion of IL-2 was 5-fold higher and secretion of IFN- and IL-6 was greater in cultures of CD8⁺ cells plus primed CD4⁺ cells than in cultures of CD8⁺ cells plus naive CD4⁺ T cells or in cultures of CD8⁺ T cells alone.

Transgenic Models of Human Disease

J. Patterson, M. Manchester, J.E. Gairin, M.G. von Herrath, D. Homann, M.B.A. Oldstone

MEASLES VIRUS INFECTION

Infection with measles virus still kills more than 1 million persons each year. Death is associated with profound immunosuppression. In addition, encephalomyelitis develops in roughly 1 in 1000 persons with measles virus infection and a chronic CNS disease termed subacute sclerosing panencephalitis in about 1 in 100,000 to 1 in 1 million. The recent discovery that the human membrane glycoprotein CD46 is the measles virus receptor enabled us to establish transgenic mice in which (1) the gene for CD46 was either transcriptionally regulated by a neuron-specific promoter and expressed only in neurons (NSE model) or (2) the genomic CD46 was used to express the 4 major isoforms of the CD46 molecule throughout the body (genomic model) (Fig. 1).

Intracerebral inoculation of mice with the Edmonston strain of measles virus leads to replication of the virus in neurons, with lesions throughout the cortex, hippocampus, and thalamus (Fig. 1). Signs of illness include weight loss, tremors, paralysis, seizures, and death approximately 3 weeks (NSE model) or 15 days (genomic model) after infection. Marked T-cell infiltration, especially CD4 T cells, occurs. In collaboration with M. Billeter, University of Zurich, Switzerland, we used reverse genetics to manipulate specific measles virus genes. We found that removal of the V gene led to attenuation and decreased viral spread in the genomic model (Fig. 1).

Studies completed in the NSE model also showed reactive astrocytosis, microglial activation, and upregulation of MHC class I molecules on endothelial cells. High levels of measles virus mRNA were detected in brain tissue, and apoptosis of neurons occurred. The chemokines RANTES, Mip-1, Mip-1ß, and IP-10 were upregulated in the brain.

The cytokine IFN-/ß is induced during viral infection and is thought to be important for controlling a variety of viral infections, including measles virus infection. IFN-/ß is also thought to be a key player in driving the conversion from acute to persistent infection. To determine the role of interferon in controlling measles virus infection in vivo, we crossed CD46 transgenic mice with mice lacking the receptor for interferon IFN-/ß. When the cross-bred mice were infected with measles virus, weight loss, paralysis, and death occurred much more rapidly (mean for death, 7.5 ± 1.5 days after infection) than in CD46 transgenic control mice that had the receptor, which died about 28 days after infection. Lymphocytic infiltration, chemokine upregulation, and viral load were equivalent in both groups of mice. However, upregulation of IL-1, IL-6, TNF-, and IFN- was significantly greater in the mice that lacked the receptor for IFN-/ß than in the controls. These results suggest that IFN-/ß is important for prolonging survival in measles virus infection, possibly by modulating the expression of deleterious proinflammatory cytokines.

AUTOIMMUNE DIABETES

Viral infections have been implicated in the pathogenesis of autoimmune insulin-dependent diabetes mellitus (IDDM) through 2 essential mechanisms that can lead to activation of autoreactive lymphocytes. One is bystander activation, which is mediated by inflammatory cytokines or chemokines; the other is molecular mimicry, which involves direct cross-reactivity between self and viral epitopes. Using the well-established RIP-LCMV-NP model of virally induced slow-onset IDDM, we tested whether sequential infections with 2 different viruses could enhance autoimmune disease.

RIP-LCMV-NP transgenic mice express the nucleoprotein of lymphocytic choriomeningitis virus (LCMV) in the ß cells in the pancreas and in the thymus. Because of negative thymic selection, only a few autoreactive cytotoxic T lymphocytes (CTLs) specific for the LCMV nucleoprotein are generated, and therefore IDDM develops slowly, within 2--4 months after infection with LCMV.

Administration of Pichinde virus, which does not cross-react at the CTL level with LCMV, 1 month after the triggering LCMV infection accelerated the development of IDDM. The acceleration of disease was associated with activation of LCMV memory CTLs by the Pichinde virus infection. Thus, a direct association of enhanced clinical autoimmune diabetes (IDDM) with infection by an unrelated virus (Pichinde) can amplify autoreactive primary primed (LCMV) lymphocytes and enhance inflammation caused by autoimmune disease.

The in vivo potential of a designed MHC class I--restricted blocking peptide to decrease antiviral (self) CTL responses responsible for causing IDDM was evaluated. In vitro the blocking peptide selectively aborted CTL killing restricted to the D^b allele not only for CTLs specific for LCMV but also for CTLs specific for influenza virus and simian virus 40. In addition, in vitro proliferation and expansion of memory CTLs specific for LCMV were diminished, resulting in a more than 10-fold decrease in the secretion of IFN-. In vivo, a 2-week course of daily treatment with the blocking peptide decreased the anti-LCMV D^b-restricted CTL response enough to prevent LCMV-induced IDDM in the RIP-LCMV-GP transgenic mouse model. In this system, LCMV viral glycoprotein is expressed constitutively as a transgene in pancreatic ß-cells under control of the rat insulin promoter. When these mice are infected with LCMV, IDDM develops that is completely dependent on MHC class I D^b--restricted antiself (viral) CTLs.

Precursor numbers of splenic anti-LCMV CTLs in peptide-treated RIP-GP mice without IDDM were reduced 26- to 46-fold in multiple experiments, although the cytokine profile of the mice was not altered. Although D^b-restricted CTL responses were eliminated by treatment with the blocking peptide, K^b CTLs were generated that were able to control the LCMV infection. Thus, in vivo treatment with a single MHC class I blocking peptide, designed to prevent a specific MHC antigen-presentation pathway, can prevent MHC-linked autoimmune disease. The peptide directly affects the expansion of autoreactive MHC-restricted CTLs and limits the generation of the number of autoaggressive cells required to cause disease.

Autoimmune CNS Disease

A. Holz, M.B.A. Oldstone

The myelin-associated/oligodendrocyte basic protein (MOBP) was recently cloned from myelin. Its expression is restricted to oligodendrocytes of the CNS. A single copy of the gene for MOBP is generated by alternative splicing of several small and highly basic myelin proteins. The kinetics of the developmental expression of MOBP directly coincide with the appearance of compact myelin, and MOBP proteins are localized in the myelin sheath along the major dense line, suggesting that MOBP functions in the signaling of myelin sheath compaction.

We expressed recombinant MOBP protein and found that it produces a disease similar to experimental allergic encephalomyelitis in SJL/J mice. Using overlapping MOBP peptides, we broadly mapped the encephalitogenic domain to a single 24 amino acid site. Experimental allergic encephalomyelitis induced by MOBP causes a clinical phenotype and an intense infiltration of mononuclear cells into the CNS (Fig. 1). Thus, MOBP can cause CNS disease. The involvement of MOBP in human diseases such as multiple sclerosis is currently under investigation in collaboration with R. Martin, National Institutes of Health.

PUBLICATIONS

Borrow, P., Tough, D.F., Eto, D., Tishon, A., Grewal, I.S., Sprent, J., Flavell, R.A., Oldstone, M.B.A. A role for CD4⁺ T cells in the generation of virus-specific CD8⁺ memory cells demonstrated in CD40L-deficient mice. J. Virol., in press.

Grewal, I.S., Borrow, P., Pamer, E.G., Oldstone, M.B.A., Flavell, R.A. The CD40/CD154 system in anti-infective host defense. Curr. Opin. Immunol., in press.

Homann, D., Tishon, A., Berger, D.P., Weigle, W.O., von Herrath, M.G., Oldstone, M.B.A. Evidence for an underlying CD4 helper and CD8 T cell defect in B cell-deficient mice: Failure to clear persistent virus infection after adoptive immunotherapy with virus-specific memory cells from µMT/µMT mice. J. Virol., in press.

Oldstone, M.B.A. HIV versus cytotoxic T lymphocytes: The war being lost. N. Engl. J. Med. 337:1306, 1997.

Oldstone, M.B.A. Molecular mimicry and immune mediated diseases. FASEB J., in press.

Oldstone, M.B.A. Viral persistence: Mechanisms and consequences. Curr. Opin. Microbiol., in press.

Oldstone, M.B.A. Viruses, Plagues and History. Oxford University Press, New York, 1998.

Raeber, A.J., Race, R.E., Brandner, S., Priola, S.A., Sailer, A., Bessen, R.A., Mucke, L., Manson, J., Aguzzi, A., Oldstone, M.B.A., Weissmann, C., Chesebro, B. Astrocyte-specific expression of hamster prion protein (PrP) renders PrP knockout mice susceptible to hamster scrapie. EMBO J. 16:6057, 1997.

Steinman, L., Oldstone, M.B.A. More mayhem from molecular mimics. Nature Med. 3:1321, 1997.

von Herrath, M.G., Coon, B., Lewicki, H., Mazarguil, H., Gairin, J.E., Oldstone, M.B.A. In vivo treatment with MHC class I--restricted blocking peptide can prevent virus-induced autoimmune diabetes. J. Immunol., in press.

von Herrath, M.G., Holz, A., Homann, D., Oldstone, M.B.A. Role of viruses in type I diabetes. Semin. Immunol. 10:87, 1998.

von Herrath, M.G., Homann, D., Gairin, J.E., Oldstone, M.B.A. Pathogenesis and treatment of virus-induced autoimmune diabetes: Novel insights gained from the RIP-LCMV transgenic mouse model. Biochem. Soc. Trans. 25:630, 1997.

von Herrath, M.G., Oldstone, M.B.A. Interferon-* is essential for destruction of ß cells and development of insulin-dependent diabetes mellitus. J. Exp. Med. 185:531, 1997.