Increased Epitope-Specific CD8+ T Cells Prevent Mu
http://www.100md.com
病菌学杂志 2005年第6期
Department of Microbiology
Department of Pathology and Laboratory Medicine, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
ABSTRACT
CD8+ T cells are important for clearance of neurotropic mouse hepatitis virus (MHV) strain A59, although their possible role in A59-induced demyelination is not well understood. We developed an adoptive-transfer model to more clearly elucidate the role of virus-specific CD8+ T cells during the acute and chronic phases of infection with A59 that is described as follows. C57BL/6 mice were infected with a recombinant A59 virus expressing the gp33 epitope, an H-2Db-restricted CD8+ T-cell epitope encoded in the glycoprotein of lymphocytic choriomeningitis virus, as a fusion with the enhanced green fluorescent protein (RA59-gfp/gp33). P14 splenocytes (transgenic for a T-cell receptor specific for the gp33 epitope) were transferred at different times pre- and postinfection (p.i.). Adoptive transfer of P14 splenocytes 1 day prior to infection with RA59-gfp/gp33, but not control virus lacking the gp33 epitope, RA59-gfp, reduced weight loss and viral replication and spread in the brain and to the spinal cord. Furthermore, demyelination was significantly reduced compared to that in nonrecipients. However, when P14 cells were transferred on day 3 or 5 p.i., no difference in acute or chronic disease was observed compared to that in nonrecipients. Protection in mice receiving P14 splenocytes prior to infection correlated with a robust gp33-specific immune response that was not observed in mice receiving the later transfers. Thus, an early robust CD8+ T-cell response was necessary to reduce virus replication and spread, specifically to the spinal cord, which protected against demyelination in the chronic phase of the disease.
INTRODUCTION
The capacity of the host immune response to control viral replication and spread throughout the central nervous system (CNS) is a critical determinant of mouse hepatitis virus (MHV) pathogenesis. However, the degree to which the immune response, specifically activated virus-specific CD8+ T cells, may contribute to neurological disease during the acute and chronic phases of MHV infection has not been clearly defined. MHV strain A59 induces CNS disease in susceptible mice, providing an animal model to study virus-induced acute encephalitis, as well as primary demyelination.
MHV is a member of the coronavirus family, which contains very large positive-stranded RNA genomes of approximately 30 kb. These viruses infect many vertebrate hosts and induce a variety of diseases ranging in severity. For example, while human coronaviruses 229E and OC43 are responsible for causing the common cold in humans, the recently identified coronavirus severe acute respiratory syndrome is responsible for very severe and potentially lethal respiratory disease (7). The outcome of MHV-induced disease is dependent on several factors, including the age and strain of the mouse, the strain of MHV, and the route of virus inoculation. A59 is a dual-tropic virus infecting both the liver and the CNS. Within the CNS, A59 infects primarily neurons but also glial cells and endothelial cells (14, 25). Following intracranial (i.c.) infection with A59, 4-week-old C57BL/6 (B6) mice develop mild to moderate encephalitis and moderate hepatitis with virus titers peaking between days 3 and 5 postinfection (p.i.) (12). Infectious virus is cleared within the first 10 to 14 days; however, at this time mice begin to develop primary demyelination, either subacute or accompanied by hind limb paralysis (13, 30). At the onset of demyelination, infectious virus is no longer detectable in the CNS but viral RNA can be detected for up to a year p.i. (8).
Virus clearance requires both CD8+ and CD4+ T cells (26, 28, 29). Infections of ?2-microglobulin knockout (?2M–/–) mice revealed that CD8+ T cells responding to the endogenous H-2Kb epitope within the spike protein S598-605 (S598) are necessary for clearance of A59 as these mice are extremely sensitive to even low doses of virus and experience delayed kinetics of viral clearance (9). CD4+ T cells are necessary for proper CD8+ T-cell activation, survival, and retention in the infected CNS (26, 27). Clearance of infectious virus is mediated by both cytolytic and cytokine-mediated mechanisms; while oligodendrocytes require gamma interferon (IFN-), astrocytes and microglia are targeted by perforin and the mechanisms controlling virus replication in neurons are largely unknown but may include noncytolytic mechanisms such as the effects of cytokines and antibody (22).
Demyelination is a complex process, and while the precise mechanisms of this pathology are unclear, on the basis of studies with the highly neurovirulent JHM strain of MHV, MHV-induced demyelination is thought to be primarily immune mediated (10, 33). Demyelination can be completely eliminated in JHM-infected, recombinase-activating gene-deficient (RAG–/–) mice that lack functional T and B cells, and this can be reversed upon transfer of splenocytes from immunocompetent mice (35). It has also been shown by depletion and transfer studies in the JHM model that either CD4+ or CD8+ T cells can induce demyelination (35). On the contrary, A59-induced demyelination has been shown to develop in the absence of B and T cells (18). Furthermore, depletion of CD4+ or CD8+ T cells after the acute stage of the infection does not reduce demyelination (30). Thus, the role of CD8+ T cells in promoting A59-induced demyelination has not been clearly defined, and it appears that the two closely related strains of MHV may induce demyelination via unique mechanisms.
Our hypothesis is that A59-induced neurovirulence is determined by the extent of virus replication and spread of antigen throughout the CNS; thus, we predicted that epitope-specific CD8+ T cells would impede virus growth and lessen disease severity in the acute phase of infection. To test this hypothesis and to better understand whether virus-specific CD8+ T cells are involved in MHV strain A59-induced CNS disease, we have established an adoptive-transfer model. In this model, mice are infected with RA59-gfp/gp33, a recombinant form of A59 that expresses the H-2Db-restricted CD8+ T-cell epitope gp33-41 (referred to as gp33); this epitope is derived from the glycoprotein of lymphocytic choriomeningitis virus (LCMV). As a source of transferred cells, we used the P14 transgenic mouse, which produces CD8+ T cells specific for this epitope (3). Thus, we adoptively transferred splenocytes from P14 mice into B6 recipients 1 day prior to infection with RA59-gfp/gp33 or during the acute phase of infection, on day 3 or 5 p.i., to determine the kinetics of CD8+ T-cell-mediated protection and/or pathogenesis. Overall, we observed less disease and lower virus titers, as well as reduced spread of virus in the CNS of those animals receiving epitope-specific CD8+ T cells 1 day before infection with RA59-gfp/gp33; this correlated with a robust gp33-specific CD8+ T-cell response within the infected brain. We also observed reduced spinal cord demyelination at 4 weeks p.i. in the animals that received the P14 cells 1 day prior to infection. Interestingly, adoptive transfer of P14 splenocytes into acutely infected B6 mice, on either day 3 or 5 p.i., did not cause either an increase or a decrease in viral titers, virus spread, or demyelination compared to those in RA59-gfp/gp33-infected mice that did not receive P14 cells. Thus, we demonstrated that virus spread to the spinal cord white matter occurred very rapidly and protection from demyelination correlated with the recruitment of high numbers of virus-specific CD8+ T cells that inhibited virus spread to the spinal cord during the acute phase of infection.
MATERIALS AND METHODS
Mice and viruses. Four-week-old male mice were used in all experiments; B6 or B6-LY5.2/Cr (CD45.1) mice were obtained from the National Cancer Institute, and ?2M–/– mice (B6 background) were obtained from Jackson Laboratories. P14 mice (3) were bred at the University of Pennsylvania. Recombinant MHV strain A59 expressing enhanced green fluorescent protein (EGFP) (referred to here as RA59-gfp) was derived from MHV strain A59 by targeted recombination as previously described (6). Selection of RA59-gfp expressing the LCMV gp33 epitope as a fusion protein with EGFP, referred to as RA59-gfp/gp33, is described elsewhere (5). Briefly, a double-stranded DNA fragment containing the coding region for the first nine complementary synthetic oligonucleotides and 63 amino acids of MHV ORF4a (including the initiation AUG codon), followed by the coding region for the 9 amino acids of gp33, was generated from complementary synthetic oligonucleotides 63 bases in length. This fragment was ligated into pEGFP (Clontech) such that the ORF4a/gp33 fragment was adjacent to and in frame with the gene for EGFP. The fragment containing ORF4a/gp33/EGFP was cleaved from this plasmid with SalI and NotI and inserted into pMH54-EGFP (6) to replace most of gene 4. Targeted recombination of donor RNA transcribed from pMH54-EGFP containing the gp33 sequences and the recipient virus fMHV was carried out with feline FCWF cells (5). Recombinant viruses were selected by two rounds of plaque purification on L2 cells, and sequencing was performed to verify the presence of gp33 (5).
Isolation of mononuclear cells from spleens for adoptive transfer. Spleens were removed from P14 mice. Suspensions were prepared by homogenizing spleens in a nylon bag (mesh opening size of 64 μm) with a syringe plunger in RPMI 1640 medium supplemented with 1% fetal calf serum. Red blood cells were lysed with 0.83% ammonium chloride, and the lymphocyte suspension was washed twice in 1x phosphate-buffered saline (PBS) and resuspended in 1x PBS for transfer.
Intravenous (i.v.) injection and i.c. inoculation of mice. Adoptive transfer of 2 x 107 cells in 0.5 ml of 1x PBS to 4-week-old B6 mice was achieved through i.v. injection via the tail. Mice received P14 cells 1 day prior to infection with A59-gfp/gp33 or A59-gfp. For i.c. infections, mice were anesthetized with isoflurane, and 30 μl containing 105 PFU of virus diluted in 1x PBS-0.75% bovine serum albumin was injected into the left cerebrum.
Virus replication in mice. To measure in vivo virus replication, mice were sacrificed on days 3, 5, 7, 10, and 28 p.i. Mice were perfused with 10 ml of 1x PBS, and their brains were removed. The left half of each brain was placed in 3 ml of gel saline (an isotonic saline solution containing 0.167% gelatin), weighed, and stored frozen at –80°C. Brains were subsequently homogenized, and standard plaque assays were performed with L2 mouse fibroblast monolayers (12). The right half of each brain was placed in 10% phosphate-buffered formalin to fix for histology and viral antigen staining.
Histologic and immunohistochemical analyses. The right halves of brains and entire spinal cords from animals sacrificed at days 3, 5, and 7 p.i. were fixed in formalin, embedded in paraffin, sectioned, and stained for viral antigen or inflammation. Antigen staining was performed by the avidin-biotin-immunoperoxidase technique (Vector Laboratories, Burlingame, Calif.) by using diaminobenzidine tetrahydrochloride as the substrate and a 1:20 dilution of rabbit antinucleocapsid monoclonal antibody (kindly provided by Julian Leibowitz). All slides were read in a blinded manner, and groups of at least four mice were examined in two separate experiments for each virus and treatment. Hematoxylin-and-eosin (H-and-E) staining for inflammation was carried out, and analysis was performed in a blinded manner by a neuropathologist. The encephalitis scores used to classify inflammation were as follows: 0, no specific pathological changes within the brain parenchyma; 1, few perivascular cuffs with minimal extension into the parenchyma and some vacuolation and cell loss; 2, same as 1 except more areas affected; 3, neuropil vacuolation and/or cell loss; 4, widespread vacuolation and/or cell loss.
Isolation of mononuclear cells from the brain for analysis. Mononuclear cells from the brain were prepared as previously described (5, 24) on days 7, 10, 12, and 28 p.i. Cells harvested from five or six brains per group were pooled. Cells were passed through a 30% Percoll gradient and then passed through a cell strainer (70-μm pore diameter; Becton Dickinson). The cell suspension was layered atop a 2-ml Lymphocyte-M (Cedarlane Laboratories) cushion, and viable cells were removed from the interface, washed with 1x PBS, and counted.
Intracellular IFN- staining and flow cytometry analysis. Intracellular IFN- secretion was assayed in response to stimulation with specific peptides as previously described (19, 25). Brain-derived mononuclear cells or splenocytes (106) were cultured with 10 U of human recombinant interleukin-2-1 μl of brefeldin A (Golgiplug; PharMingen) per ml either with or without 1 μg of specific peptide per ml in a total volume of 200 μl of RPMI 1640 medium supplemented with 5% fetal calf serum for 5 h at 37°C. Cells were then stained for surface expression of CD8, CD4, and/or CD45.2 with monoclonal antibodies specific for CD8a (clone 53-6.7), CD4 (clone RM4-5), and CD45.2 (clone 104) (PharMingen). After surface staining, intracellular IFN- was detected by first fixing and then permeabilizing cells with the Cytofix/Cytoperm kit (PharMingen) and stained with a fluorescein isothiocyanate-conjugated monoclonal rat anti-mouse IFN- antibody (clone XMG 1.2; PharMingen). Cells were analyzed with a FACScan flow cytometer (Becton Dickinson). The total number of cells positive for IFN- per mouse was determined by multiplying the fraction of cells positive for IFN- by the total number of live cells isolated per brain.
Demyelination. Demyelination was analyzed for both quantity and severity. Analysis of demyelination was performed on spinal cords harvested from animals 28 days p.i. Five to eight mice were examined in each of two separate experiments. After sacrifice, mice were perfused with 10 ml of 1x PBS and their spinal cords were removed. Spinal cords were sectioned into five regions representing the cervical through lumbar regions and embedded in paraffin for sectioning. Cross sections were stained with a myelin-specific dye, luxol fast blue. To determine the percentage of demyelinated spinal cord, quadrants of spinal cord were counted. At least 10 sections of cord were counted for each animal. A description of the pathology score used to analyze the severity of demyelination is as follows: 0, no demyelination; 1, small foci of demyelination seen in less than 1 section per slide (containing a total of 10 sections of cord); 2, one to three small areas of demyelination per slide; 3, at least one large region of demyelination per slide and all sections containing some demyelination; 4, ample demyelination involving all sections with one to three large areas of demyelination; 5, all sections involved with more than three large regions of demyelination, confluent and involving the medulla oblongata.
RESULTS
Reduced weight loss in early transfer recipients infected with RA59-gfp/gp33. In order to assess the role of epitope-specific CD8+ T cells during the acute phase of infection with A59, we developed an adoptive-transfer system with splenocytes from P14 mice, which produce gp33-specific CD8 T cells (3), and a recombinant form of strain A59 that expresses the gp33 epitope as a fusion protein with EGFP (RA59-gfp/gp33). Figure 1A shows schematic diagrams of the genomes of RA59-gfp/gp33 and control virus RA59-gfp (5). Nonessential gene 4 was replaced with a small piece of ORF4a and the gfp/gp33 fusion. Elimination of expression of ORF4 has previously been shown to have no effect on virulence of MHV in the CNS (21). As shown in Fig. 1B, in the adoptive-transfer model, 4-week-old B6 mice received 2 x 107 P14 splenocytes i.v. 1 day prior to i.c. infection or on day 3 or 5 p.i. with 105 PFU of either RA59-gfp/gp33 or control virus RA59-gfp, which does not express the gp33 epitope. Control animals that did not receive an adoptive transfer were also infected with either RA59-gfp/gp33 or RA59-gfp. To assess clinical symptoms, mice were weighed prior to infection and on various days p.i. Compared to mock-infected animals, mice infected with RA59-gfp/gp33 lost significant weight by day 7 p.i. (Fig. 2A). Animals that received gp33-specific CD8+ T cells prior to infection with the gp33-expresssing virus lost significantly less weight by day 7 p.i. (P < 0.001) than did mice receiving a transfer on day 3 or 5 p.i. and those animals not receiving a transfer (Fig. 2A). RA59-gfp-infected mice lost similar amounts of weight with or without a transfer of P14 cells (Fig. 2B). Thus, adoptive transfer of epitope-specific cells prior to, but not during, infection with MHV protected against the severe weight loss associated with the acute phase of infection.
Virus titers, antigen spread, and encephalitis in the brain after transfer of epitope-specific CD8+ T cells. First, we measured the effect of the transferred epitope-specific CD8+ T cells on viral titers in the brains of infected mice. Compared to mice that did not receive P14 cells, those that received the transfer prior to infection had a significant (P < 0.001) reduction in virus titers of approximately 20-fold on day 5 p.i. (Fig. 3A). In addition, animals that received the P14 cell transfer on day 3 p.i. had titers similar to those of the nonrecipients and their titers were significantly greater than those of the animals that received the transfer prior to infection (P < 0.05). However, on day 7 p.i., the titers in all groups were low and no significant differences were observed (Fig. 3A). Control mice infected with RA59-gfp that received gp33-specific CD8+ T cells exhibited no difference in titers compared to those of nonrecipients, as expected, demonstrating that the CD8+ T-cell response is specific. Thus, the transfer of naive, epitope-specific CD8+ T cells was capable of protecting against virus replication when the transfer was done i.v. prior to infection. In all RA59-gfp/gp33- and RA59-gfp-infected animals, with or without having received the P14 cells, titers were undetectable on days 10 and 28 p.i. (data not shown).
We have previously observed that spread of viral antigen in the brain, compared with viral titers, is a more accurate indication of the severity of CNS damage (25, 31). In order to assess virus spread, we examined viral antigen in sagittal sections of brains that were harvested at different times p.i. with a monoclonal antibody directed at the viral nucleocapsid protein. The peak of virus antigen expression was on day 5 p.i. in the brain; virus antigen staining was very low on day 3 p.i., despite high titers, and was almost undetectable on day 7 in all of the sections examined from each group (data not shown). Several regions of the brain that were consistently involved during the infection included the basal ganglia, the olfactory bulb, the subiculum, the hypothalamus, and the hindbrain, including the medulla and pons. Correlating with the difference in titers observed on day 5, dramatically less antigen staining was observed throughout the brain, as represented in the subiculum and hindbrain, in mice that received gp33-specific CD8+ T cells prior to infection (Fig. 4B and E). Only small foci of antigen-positive cells existed in examined sections from animals that received the transfer prior to infection. However, animals that received the transfer on day 3 p.i. had an amount of viral antigen similar to that in mice that did not receive a transfer, as shown in the subiculum and the hindbrain (Fig. 4). As expected, control animals infected with RA59-gfp had similar levels of antigen with or without having received the transfer (data not shown).
Encephalitis was determined by H-and-E staining of sagittal brain sections and spinal cord cross sections taken on day 7 p.i., the typical peak of inflammation. In agreement with observations of reduced weight loss in RA59-gfp/gp33-infected mice that received the adoptive transfer prior to infection (Fig. 2A), we observed a significant reduction in inflammation (Table 1). H-and-E-stained sections from mice that received the transfer prior to infection revealed a reduction in infiltrating inflammatory cells, little to no pathological changes within the parenchyma, and minimal perivascular cuffing compared to all other groups of infected mice (Table 1).
Early transfer of epitope-specific CD8+ T cells blocks virus spread to the spinal cord. Previous studies with the JHM strain of MHV demonstrated that the virus spreads from the brain into the spinal cord between days 7 and 10 p.i. (23). However, virus spread from the brain to the spinal cord during the acute phase of infection with the A59 strain of MHV has not been examined. Thus, we examined spinal cords harvested from infected mice at days 5 and 7 p.i. to determine the kinetics of viral antigen spread. Immunohistochemical analysis with the antinucleocapsid antibody was performed on cross sections of formalin-fixed spinal cord. Robust viral antigen expression was detected on day 5 p.i. in RA59-gfp/gp33-infected animals that did not receive the adoptive transfer (Fig. 5A and B) or received the day 3 p.i. adoptive transfer (Fig. 5E and F), but no antigen was detected in animals that received the transfer prior to infection (Fig. 5C and D). On day 7 p.i., viral antigen was still detectable in all of the animals except those that received the transfer prior to infection (data not shown). Virus antigen was detected in both the gray and white matter on day 5 p.i. (Fig. 5), whereas on day 7 p.i. virus antigen was only detected in the white matter (data not shown). Thus, corresponding to the detection of virus antigen in the brain, virus spread to the spinal cord is reduced only when gp33-specific CD8+ cells are transferred prior to infection.
Early recruitment and activation of gp33-specific CD8+ T cells correlates with reduced CNS disease. To evaluate the degree to which the recombinant virus could elicit an immune response to the introduced gp33 epitope and thus how well the transferred cells engrafted, we sought to determine the activation status of virus-specific CD8+ T cells in animals that had or had not received the transfer of P14 cells. In order to accurately assess the role of the transferred epitope-specific CD8+ T cells at the site of infection, we obtained mononuclear cells from the brains of infected mice on day 7 p.i., the typical peak of lymphocyte infiltration in the brain after i.c. infection with A59. More than half (51.7%) of the CD8+ T cells harvested from the brains of RA59-gfp/gp33-infected mice that received the adoptive transfer of gp33-specific CD8+ T cells were specific for gp33 on day 7 p.i. (Fig. 6A). In contrast, only 14.8% of the brain-derived CD8+ T cells were activated and specific for the gp33 epitope in mice that did not receive the adoptive transfer; this reflected the normal primary immune response and is similar to previously reported results (5). Interestingly, mice that received the adoptive transfer on either day 3 or 5 p.i. had percentages of gp33-specific CD8+ T cells similar to those of animals that did not receive the transfer (15.7 and 14.5%, respectively).
In order to understand how the response to an immunodominant epitope may affect the response to other, less dominant epitopes, we analyzed the response to the endogenous epitope in the spike protein S598. Animals infected with RA59-gfp, lacking the gp33 epitope, contained a higher percentage of CD8+ T cells specific for the endogenous S598 epitope than did the RA59-gfp/gp33-infected animals (Fig. 6B). RA59-gfp/gp33-infected mice that received the transfer prior to infection exhibited the lowest percentage of S598-specific CD8+ T cells (1.2%), perhaps because of the very high number of activated gp33-specific CD8+ T cells (Fig. 6A).
In order to determine if the lower numbers of gp33-specific CD8+ T cells observed in the mice that received the adoptive transfer on day 3 or 5 p.i., compared to the group that received the transfer prior to infection, was due to the kinetics of activation and recruitment into the brain, we analyzed brain-derived mononuclear cells for gp33-specific CD8+ T cells on days 10 and 12 p.i. (which corresponds to day 7 posttransfer for each transfer recipient group). To distinguish between the transferred P14 cells and the endogenous gp33-specific CD8+ T cells, we used B6-LY5.2/Cr mice (CD45.1) as recipients; thus, the CD45.2-positive cells reflect the adoptively transferred P14 cells. We observed an overall increase in the percentage of gp33-specific CD8+ T cells on day 12 (Fig. 7, bottom row) compared to day 10 (Fig. 7, top row), and this reflects a decrease in the total number of CD8+ T cells in the brain and retention of virus-specific CD8+ T cells. As expected, no CD45.2 cells were detected in animals that did not receive the adoptive transfer (Fig. 7). The majority of the IFN--secreting, gp33-specific CD8+ T cells were P14 in origin (CD45.2+) in the mice that received the transfer prior to infection. Interestingly, a significantly smaller percentage of the gp33-CD45.2+ gp33-gp33-specific CD8+ T cells were observed in the mice that received the adoptive transfer on day 3 p.i. than in the mice that received the transfer prior to infection (2.24% compared with 33.7%). Furthermore, P14-derived gp33-specific CD8+ cells were completely undetectable on both days 10 and 12 p.i. in mice that received the adoptive transfer on day 5 p.i. Thus, activation and trafficking of naive, virus-specific CD8+ T cells appeared to be limited to the first 3 days of the infection.
Demyelination is reduced in animals protected from acute disease. A focus of this study was to understand how a strong CD8+ T-cell response could shape the outcome of the chronic disease induced by A59. Earlier depletion studies indicated that when CD8+ T cells were depleted after resolution of acute A59 infection, demyelination was nearly the same as in nondepleted mice, which suggested the possibility that CD8+ T cells could initiate disease during the acute phase of infection (30). Mice infected with RA59-gfp/gp33 or RA59-gfp received transfers the day prior to infection or on either day 3 or day 5 p.i. and were sacrificed 4 weeks p.i., at which time their spinal cords were removed, fixed, and stained with a myelin-specific dye, luxol fast blue. Demyelination was assayed by counting quadrants of cross-sectioned spinal cord. We also used a pathology score to describe the severity of demyelination of the spinal cord; the scores ranged from 0, no demyelination evident, to 5, pathology involving severe, confluent demyelination of nearly the entire spinal cord (see Materials and Methods) (Fig. 8).
Mice protected from acute disease via adoptive transfer of gp33-specific CD8+ T cells 1 day prior to infection had significantly less demyelination, as measured by both the percentage of quadrants containing demyelination and the severity of demyelination, than did those that were not protected from acute disease (non-transfer recipients and day 3 and 5 p.i. transfer recipients) (Fig. 8A). The levels of demyelination observed in animals infected with RA59-gfp did not differ, regardless of the adoptive transfer of P14 cells, as expected (Fig. 8B). Thus, protection from demyelination correlated with reduced acute disease, including reduced virus spread.
We observed that the percentage of demyelinated spinal cord and the severity of spinal cord demyelination were greater in the RA59-gfp/gp33-infected mice than in RA59-gfp-infected animals. We reasoned that this could be due to either a small difference in the kinetics of virus replication and spread or possibly to a pathogenic immune response elicited by the gp33-expressing virus. Infections of ?2M–/– mice were carried out to determine whether a detrimental major histocompatibility complex class I-restricted immune response was elicited in RA59-gfp/gp33-infected mice. ?2M–/– mice were infected with 250 PFU of either RA59-gfp/gp33 or RA59-gfp, as these mice are highly susceptible to MHV infection. Surprisingly, despite the absence of detectable differences in either virus titers in the brain or virus antigen detection in the brain during the first week of infection (data not shown), RA59-gfp/gp33 still induced significantly more demyelination than did RA59-gfp (Fig. 9). This suggests that demyelination induced by RA59-gfp/gp33 may be more severe because of an intrinsic property of the virus other than immune recognition of the gp33 epitope expressed by RA59-gfp/gp33; however, we cannot completely rule out the possibility that CD8+ T cells can contribute to demyelination in these mice, as they do exist in low numbers.
Retention of epitope-specific CD8+ T cells within the brain after CNS infection. CD8+ T cells have been shown to persist in an activated, IFN--secreting state within the brains of mice after infection with MHV (15), and it has been shown that this is dependent upon the persistence of viral RNA (16). To more clearly understand how the acute phase of infection, as shaped by the transfer of P14 cells, affected the retention of CD8+ T cells during the chronic phase of infection and to determine if retained epitope-specific CD8+ T cells correlated with demyelination, brains were removed 4 weeks p.i. for isolation and analysis of localized inflammatory cells. Brain-derived cells were analyzed for intracellular IFN- secretion in response to in vitro peptide stimulation. The total numbers of CD8+ T cells in the brains of all of the animals were similar, although the total number of virus-specific, IFN--secreting CD8+ T cells was about 10-fold higher in RA59-gfp/gp33-infected animals than in RA59-gfp-infected animals, likely because of the response to the immunodominant gp33 epitope in the former (data not shown). In RA59-gfp/gp33-infected animals, the levels of activated gp33-specific IFN--secreting CD8+ T cells were similar but highest in the mice that received the transfer prior to infection (Fig. 10). Thus, we observed that a similar, high number of gp33-specific, IFN--secreting CD8+ T cells was retained in all groups. The percentage of gp33-specific CD8+ T cells did not appear to correlate with the gp33-specific CD8+ T-cell response observed during the acute phase of infection (on day 7, 10, or 12 p.i.; Fig. 6 and 7). A clear correlation between the presence of activated gp33-specific CD8+ T cells and demyelination could not be made; however, it was noted that the day –1 transfer recipients had the lowest levels of demyelination and the highest percentage of gp33-specific CD8+ T cells, whereas the lowest percentage of gp33-specific CD8+ T cells was observed in animals that demonstrated high levels of severe demyelination (no-transfer group and day 3 and 5 p.i. transfer groups; Fig. 8 and 10).
DISCUSSION
Infection with the A59 strain of MHV provides a model for both viral encephalitis and demyelination, but despite extensive study, the mechanism of A59-induced demyelination is still not fully understood. The findings presented here have highlighted several factors that are important in understanding the pathogenesis of A59-induced demyelination, including the relationship between virus spread during the acute phase of infection and its outcome in terms of demyelination, as well as how the kinetics of CD8+ T-cell activation and recruitment to the brain affect virus spread within the CNS.
We have previously examined the role of CD8+ T-cell-mediated protection from A59 infection by a vaccination strategy (5); however, the adoptive-transfer system with splenocytes derived from the P14 mouse has allowed us to go a step further and determine the effects of T cells, specific for one epitope, on pathogenesis. It has also allowed the kinetic analysis of activation and recruitment of naive CD8+ T cells to the infected CNS. In this study we assessed protection from CNS disease achieved by adoptive transfer of naive gp33-specific splenocytes into mice prior to or during the acute phase of infection with a gp33-expressing recombinant form of strain A59. Epitope-specific CD8+ T cells inhibited virus growth and viral antigen spread in the brain and reduced disease severity and inflammation only when transferred prior to infection. This correlated with an increased activated gp33-specific CD8+ T-cell response in the brain. In animals protected from acute disease, there was a dramatic reduction in the level and severity of demyelination at 1 month p.i. This is in agreement with other studies of JHM infection that demonstrated virus spread within the CNS as a major determinant in the development of chronic demyelination (15). Similarly, studies using the transfer of nucleoprotein-specific cytotoxic T lymphocytes was effective at reducing JHM replication within the CNS (28).
Priming of virus-specific CD8+ T cells has been shown to occur very rapidly after infection with a variety of pathogens, and here we show that the window of naive CD8+ T-cell priming, in the context of a virus infection within the CNS, is limited to the first 3 days of infection. Wong and Pamer recently demonstrated a feedback mechanism during infection with Listeria monocytogenes by which antigen-presenting cells are eliminated by effector cytotoxic T lymphocytes, thereby limiting the CD8+ T-cell response (34). It is possible that this accounts for the absence of CD45.2-positive, gp33-specific CD8+ T cells observed in the brains of RA59-gfp/gp33-infected mice that had received an adoptive transfer on day 5 p.i.
CD8+ T-cell expansion is thought to be initiated in the cervical lymph nodes, the major site of T-cell priming during CNS infections, and then expansion continues in the spleen, followed by rapid accumulation of virus-specific T cells in the CNS during MHV infection (17). Norbury et al. showed that virus-specific CD8+ T cells are primed by dendritic cells, but not macrophages, in vivo after infection with vaccinia virus (20). We do not know which antigen-presenting cells are responsible for priming naive CD8+ T cells after MHV infection of the CNS. Analysis of splenocytes on days 10 and 12 p.i. revealed the presence of transferred P14 cells in this organ in all transfer recipients regardless of the day p.i. on which the cells were transferred (data not shown), although the percentage of activated spleen-derived P14 cells was dependent upon the day of transfer. CD8+ CD45.2+ T cells were present in the spleens of both transfer groups (days 3 and 5 p.i.); however, only cells harvested from mice that received the transfer on day 3 p.i. secreted IFN- in response to gp33 peptide (data not shown), which suggests that failure of the day 5 p.i. transfer to become activated was due to a deficit in priming.
The increased virus-specific CD8+ T-cell response observed in protected mice limited virus replication and spread within the CNS of RA59-gfp/gp33-infected animals. At day 5 p.i., virus titers and viral antigen staining were reduced in these animals. In some CNS infections, such as Borna disease virus and i.c. infection of mice with LCMV, CD8+ T cells can play a detrimental role during the infection; however, virus-specific CD8+ T cells primarily control virus replication and protect against virus spread, CNS damage, and encephalitis in MHV strain A59-infected mice (2, 4). Thus, during the acute phase of MHV infection of the CNS, virus-specific CD8+ T cells play a crucial role in protection from both acute and chronic disease and a stronger CD8+ T-cell response is beneficial to the host.
Virus antigen was detected throughout the brain and in the gray and white matter of the spinal cord with similar kinetics, peaking on day 5 p.i. and diminishing by day 7 p.i. Virus antigen was still detected in the spinal cords on day 7 p.i. but was present only in the white matter. On the basis of studies of JHM, virus spread from the brain into the spinal cord is thought to occur in a retrograde fashion via transneuronal spread; the rapid appearance of virus in the spinal cord, as observed here, is consistent with this. Whether the CD8+ T-cell-mediated block in virus spread to the spinal cord reflected an overall diminished load of virus within the CNS or whether CD8+ T cells mediated clearance of particular cell types is not known.
Clearance of infectious virus from the CNS requires both perforin- and cytokine-mediated mechanisms. Interestingly, our vaccination studies with perforin-deficient mice with recombinant L. monocytogenes expressing the gp33 epitope, followed by a challenge with RA59-gfp/gp33, revealed that perforin was necessary for protection, as determined by virus titers in the brain (5). This is interesting since the majority of A59-infected cells are positive for the neuronal marker dendrite-specific microtubule-associated protein 2 (25). Perforin is required for clearance from microglia and astrocytes, which express major histocompatibility complex class I; thus, these cell types may support the majority of virus replication. It is unclear how adoptively transferred P14 CD8+ T cells mediated protection from virus spread via transneuronal pathways, but it likely involved cytokine-mediated effects and is of current interest.
Interestingly, we found that the retention of IFN--secreting gp33-specific CD8+ T cells at 4 weeks p.i. was similar in all RA59-gfp/gp33-infected animals regardless of transfer of gp33-specific CD8+ T cells or the amount of virus replication and spread that occurred during the acute phase of infection. In addition, both recipients and nonrecipients had similar percentages of S598-specific CD8+ T cells despite having very different levels of S598-specific CD8+ T cells at the peak of infiltration during the acute phase of infection. Thus, the level of replication and spread during the acute phase of infection appeared to have little if any effect on the retention of CD8+ T cells during the chronic phase of infection. It has been shown in several viral infections of the CNS that CD8+ T cells are maintained within the brain after the acute phase of infection is resolved (11, 32). The function of these retained virus-specific CD8+ T cells is not clear, but as van der Most et al. (32) suggest in a model of dengue virus infection of the mouse CNS, the brain may retain local, activated, virus-specific CD8+ T cells with an effector memory phenotype as an in situ rapid response force.
Factors contributing to the maintenance of virus-specific CD8+ T cells in the CNS are also largely unknown. Retention of CD8+ T cells within the CNS of JHM-infected animals is thought to be largely dependent on the presence of viral RNA and possibly very low levels of antigen expression (16). In JHM-infected mice that recognize both H-2Db- and H-2Dd-restricted epitopes, the epitope specificity of CD8+ T cells changes from the acute to the chronic phase of infection with a decrease in H-2Dd-restricted nucleocapsid-specific CD8+ T cells and an increase in H-2Db spike-specific CD8+ T cells suggestive of altered antigen presentation during the chronic phase of infection (1). However, in the present study animals that were protected from acute disease via adoptive transfer exhibited minimal virus replication and spread, and although possible, it is unlikely that persistent viral RNA was maintaining the population of activated CD8+ T cells in these mice.
The severity of demyelination in the chronic phase of infection did not correspond to the levels of activated virus-specific CD8+ T cells. At 4 weeks p.i., total numbers and percentages of epitope-specific CD8+ T cells in the brain were similar in transfer recipients and nonrecipients infected with either RA59-gfp/gp33 or RA59-gfp. However, demyelination was much less severe in animals protected from virus replication and spread, P14 splenocyte recipients infected with RA59-gfp/gp33, despite a strong immune response during the acute phase of infection. This supports previous findings that CD8+ T cells likely do not play a direct role in the process of demyelination (9, 15, 30).
The outcome of the acute stage of A59-induced CNS disease depends on the rapid recruitment and activation of CD8+ T cells, and a good prognosis is linked with limited spread of the virus within the CNS, specifically to the spinal cord. Preventing virus spread to the spinal cord reduced levels of demyelination seen in the chronic stage of disease, suggesting that the presence of viral RNA and/or low levels of viral antigen are required for demyelination to occur. This study also demonstrated that the initial priming event after infection with a neurotropic pathogen was critical in establishing a long-lasting protective immune response in the brain. Thus, the immediate host immune response is critical and determines the outcome of A59 infection.
ACKNOWLEDGMENTS
This work was supported by NIH grants AI 47800, AI 60021, NS 30606 (S.R.W.), and AI 45025 (H.S.), and K.C.M. was supported by NIH training grant AI 007324.
We thank Devon J. Shedlock and Lauren A. Zenewicz for providing P14 mice.
REFERENCES
Bergmann, C. C., J. D. Altman, D. Hinton, and S. A. Stohlman. 1999. Inverted immunodominance and impaired cytolytic function of CD8+ T cells during viral persistence in the central nervous system. J. Immunol. 163:3379-3387.
Bilzer, T., and L. Stitz. 1994. Immune-mediated brain atrophy. CD8+ T cells contribute to tissue destruction during Borna disease. J. Immunol. 153:818-823.
Brandle, D., K. Burki, V. A. Wallace, U. H. Rohrer, T. W. Mak, B. Malissen, H. Hengartner, and H. Pircher. 1991. Involvement of both T cell receptor V alpha and V beta variable region domains and alpha chain junctional region in viral antigen recognition. Eur. J. Immunol. 21:2195-2202.
Buchmeier, M. J., R. M. Welsh, F. J. Dutko, and M. B. Oldstone. 1980. The virology and immunobiology of lymphocytic choriomeningitis virus infection. Adv. Immunol. 30:275-331.
Chua, M., K. MacNamara, L. SanMateo, H. Shen, and S. R. Weiss. 2004. Effects of an epitope-specific CD8+ T-cell response on murine coronavirus central nervous system disease: protection from virus replication and antigen spread and selection of epitope escape mutants. J. Virol. 78:1150-1159.
DasSarma, J., E. Scheen, S.-H. Seo, M. Koval, and S. R. Weiss. 2002. Enhanced green fluorescent protein expression may be used to monitor murine coronavirus spread in vitro and in the mouse central nervous system. J. Neurovirol. 8:1-11.
Drosten, C., S. Gunther, W. Preiser, S. van der Werf, H. R. Brodt, S. Becker, H. Rabenau, M. Panning, L. Kolesnikova, R. A. Fouchier, A. Berger, A. M. Burguiere, J. Cinatl, M. Eickmann, N. Escriou, K. Grywna, S. Kramme, J. C. Manuguerra, S. Muller, V. Rickerts, M. Sturmer, S. Vieth, H. D. Klenk, A. D. Osterhaus, H. Schmitz, and H. W. Doerr. 2003. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N. Engl. J. Med. 348:1967-1976.
Fleming, J. O., J. J. Houtman, H. Alaca, D. Hinze, D. McKenzie, J. Aiken, T. Bleasdale, and S. Baker. 1993. Persistence of viral RNA in the central nervous system of mice inoculated with MHV-4. Adv. Exp. Med. Biol. 342:327-332.
Gombold, J., R. M. Sutherland, E. Lavi, Y. Paterson, and S. R. Weiss. 1995. Mouse hepatitis virus A59-induced demyelination can occur in the absence of CD8+ T cells. Microb. Pathog. 18:211-221.
Haring, J. S., L. L. Pewe, and S. Perlman. 2002. Bystander CD8 T cell-mediated demyelination after viral infection of the central nervous system. J. Immunol. 169:1550-1555.
Hawke, S., P. G. Stevenson, S. Freeman, and C. R. Bangham. 1998. Long-term persistence of activated cytotoxic T lymphocytes after viral infection of the central nervous system. J. Exp. Med. 187:1575-1582.
Hingley, S. T., J. L. Gombold, E. Lavi, and S. R. Weiss. 1994. MHV-A59 fusion mutants are attenuated and display altered hepatotropism. Virology 200:1-10.
Houtman, J. J., and J. O. Fleming. 1996. Pathogenesis of mouse hepatitis virus-induced demyelination. J. Neurovirol. 2:361-376.
Lavi, E., P. S. Fishman, M. K. Highkin, and S. R. Weiss. 1988. Limbic encephalitis after inhalation of a murine coronavirus. Lab. Investig. 58:31-36.
Marten, N. W., S. A. Stohlman, R. D. Atkinson, D. R. Hinton, J. O. Fleming, and C. C. Bergmann. 2000. Contributions of CD8+ T cells and viral spread to demyelinating disease. J. Immunol. 164:4080-4088.
Marten, N. W., S. A. Stohlman, and C. C. Bergmann. 2000. Role of viral persistence in retaining CD8+ T cells within the central nervous system. J. Virol. 74:7903-7910.
Marten, N. W., S. A. Stohlman, J. Zhou, and C. C. Bergmann. 2003. Kinetics of virus-specific CD8+ T-cell expansion and trafficking following central nervous system infection. J. Virol. 77:2775-2778.
Matthews, A. E., E. Lavi, S. R. Weiss, and Y. Paterson. 2002. Neither B cells nor T cells are required for CNS demyelination in mice persistently infected with MHV-A59. J. Neurovirol. 8:257-264.
Murali-Krishna, K., J. D. Altman, M. Suresh, D. J. Sourdive, A. J. Zajac, J. D. Miller, J. Slansky, and R. Ahmed. 1998. Counting antigen-specific CD8 T cells: a reevaluation of bystander activation during viral infection. Immunity 8:177-187.
Norbury, C. C., D. Malide, J. S. Gibbs, J. R. Bennink, and J. W. Yewdell. 2002. Visualizing priming of virus-specific CD8+ T cells by infected dendritic cells in vivo. Nat. Immunol. 3:265-271.
Ontiveros, E., L. Kuo, P. S. Masters, and S. Perlman. 2001. Inactivation of expression of gene 4 of mouse hepatitis virus strain JHM does not affect virulence in the murine CNS. Virology 289:230-238.
Parra, B., D. R. Hinton, N. W. Marten, C. C. Bergmann, M. T. Lin, C. S. Yang, and S. A. Stohlman. 1999. IFN- is required for viral clearance from central nervous system oligodendroglia. J. Immunol. 162:1641-1647.
Perlman, S., G. Jacobsen, and A. Afifi. 1989. Spread of a neurotropic murine coronavirus into the CNS via the trigeminal and olfactory nerves. Virology 170:556-560.
Pewe, L., S. B. Heard, C. Bergmann, M. O. Dailey, and S. Perlman. 1999. Selection of CTL escape mutants in mice infected with a neurotropic coronavirus: quantitative estimate of TCR diversity in the infected central nervous system. J. Immunol. 163:6106-6113.
Phillips, J. J., M. M. Chua, G. F. Rall, and S. R. Weiss. 2002. Murine coronavirus spike glycoprotein mediates degree of viral spread, inflammation, and virus-induced immunopathology in the central nervous system. Virology 301:109-120.
Stohlman, S. A., C. C. Bergmann, D. J. Cua, M. T. Lin, S. Ho, W. Wei, and D. R. Hinton. 1998. Apoptosis of JHMV-specific CTL in the CNS in the absence of CD4+ T cells. Adv. Exp. Med. Biol. 440:425-430.
Stohlman, S. A., C. C. Bergmann, M. T. Lin, D. J. Cua, and D. R. Hinton. 1998. CTL effector function within the central nervous system requires CD4+ T cells. J. Immunol. 160:2896-2904.
Stohlman, S. A., C. C. Bergmann, R. C. van der Veen, and D. R. Hinton. 1995. Mouse hepatitis virus-specific cytotoxic T lymphocytes protect from lethal infection without eliminating virus from the central nervous system. J. Virol. 69:684-694.
Sussman, M. A., R. A. Shubin, S. Kyuwa, and S. A. Stohlman. 1989. T-cell-mediated clearance of mouse hepatitis virus strain JHM from the central nervous system. J. Virol. 63:3051-3056.
Sutherland, R. M., M. M. Chua, E. Lavi, S. R. Weiss, and Y. Paterson. 1997. CD4+ and CD8+ T cells are not major effectors of mouse hepatitis virus A59-induced demyelinating disease. J. Neurovirol. 3:225-228.
Tsai, J. C., L. de Groot, J. D. Pinon, K. T. Iacono, J. J. Phillips, S. H. Seo, E. Lavi, and S. R. Weiss. 2003. Amino acid substitutions within the heptad repeat domain 1 of murine coronavirus spike protein restrict viral antigen spread in the central nervous system. Virology 312:369-380.
van der Most, R. G., K. Murali-Krishna, and R. Ahmed. 2003. Prolonged presence of effector-memory CD8 T cells in the central nervous system after dengue virus encephalitis. Int. Immunol. 15:119-125.
Wang, F. I., S. A. Stohlman, and J. O. Fleming. 1990. Demyelination induced by murine hepatitis virus JHM strain (MHV-4) is immunologically mediated. J. Neuroimmunol. 30:31-41.
Wong, P., and E. G. Pamer. 2003. Feedback regulation of pathogen-specific T cell priming. Immunity 18:499-511.
Wu, G. F., A. A. Dandekar, L. Pewe, and S. Perlman. 2000. CD4 and CD8 T cells have redundant but not identical roles in virus-induced demyelination. J. Immunol. 165:2278-2286.(Katherine C. MacNamara, M)
Department of Pathology and Laboratory Medicine, School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania
ABSTRACT
CD8+ T cells are important for clearance of neurotropic mouse hepatitis virus (MHV) strain A59, although their possible role in A59-induced demyelination is not well understood. We developed an adoptive-transfer model to more clearly elucidate the role of virus-specific CD8+ T cells during the acute and chronic phases of infection with A59 that is described as follows. C57BL/6 mice were infected with a recombinant A59 virus expressing the gp33 epitope, an H-2Db-restricted CD8+ T-cell epitope encoded in the glycoprotein of lymphocytic choriomeningitis virus, as a fusion with the enhanced green fluorescent protein (RA59-gfp/gp33). P14 splenocytes (transgenic for a T-cell receptor specific for the gp33 epitope) were transferred at different times pre- and postinfection (p.i.). Adoptive transfer of P14 splenocytes 1 day prior to infection with RA59-gfp/gp33, but not control virus lacking the gp33 epitope, RA59-gfp, reduced weight loss and viral replication and spread in the brain and to the spinal cord. Furthermore, demyelination was significantly reduced compared to that in nonrecipients. However, when P14 cells were transferred on day 3 or 5 p.i., no difference in acute or chronic disease was observed compared to that in nonrecipients. Protection in mice receiving P14 splenocytes prior to infection correlated with a robust gp33-specific immune response that was not observed in mice receiving the later transfers. Thus, an early robust CD8+ T-cell response was necessary to reduce virus replication and spread, specifically to the spinal cord, which protected against demyelination in the chronic phase of the disease.
INTRODUCTION
The capacity of the host immune response to control viral replication and spread throughout the central nervous system (CNS) is a critical determinant of mouse hepatitis virus (MHV) pathogenesis. However, the degree to which the immune response, specifically activated virus-specific CD8+ T cells, may contribute to neurological disease during the acute and chronic phases of MHV infection has not been clearly defined. MHV strain A59 induces CNS disease in susceptible mice, providing an animal model to study virus-induced acute encephalitis, as well as primary demyelination.
MHV is a member of the coronavirus family, which contains very large positive-stranded RNA genomes of approximately 30 kb. These viruses infect many vertebrate hosts and induce a variety of diseases ranging in severity. For example, while human coronaviruses 229E and OC43 are responsible for causing the common cold in humans, the recently identified coronavirus severe acute respiratory syndrome is responsible for very severe and potentially lethal respiratory disease (7). The outcome of MHV-induced disease is dependent on several factors, including the age and strain of the mouse, the strain of MHV, and the route of virus inoculation. A59 is a dual-tropic virus infecting both the liver and the CNS. Within the CNS, A59 infects primarily neurons but also glial cells and endothelial cells (14, 25). Following intracranial (i.c.) infection with A59, 4-week-old C57BL/6 (B6) mice develop mild to moderate encephalitis and moderate hepatitis with virus titers peaking between days 3 and 5 postinfection (p.i.) (12). Infectious virus is cleared within the first 10 to 14 days; however, at this time mice begin to develop primary demyelination, either subacute or accompanied by hind limb paralysis (13, 30). At the onset of demyelination, infectious virus is no longer detectable in the CNS but viral RNA can be detected for up to a year p.i. (8).
Virus clearance requires both CD8+ and CD4+ T cells (26, 28, 29). Infections of ?2-microglobulin knockout (?2M–/–) mice revealed that CD8+ T cells responding to the endogenous H-2Kb epitope within the spike protein S598-605 (S598) are necessary for clearance of A59 as these mice are extremely sensitive to even low doses of virus and experience delayed kinetics of viral clearance (9). CD4+ T cells are necessary for proper CD8+ T-cell activation, survival, and retention in the infected CNS (26, 27). Clearance of infectious virus is mediated by both cytolytic and cytokine-mediated mechanisms; while oligodendrocytes require gamma interferon (IFN-), astrocytes and microglia are targeted by perforin and the mechanisms controlling virus replication in neurons are largely unknown but may include noncytolytic mechanisms such as the effects of cytokines and antibody (22).
Demyelination is a complex process, and while the precise mechanisms of this pathology are unclear, on the basis of studies with the highly neurovirulent JHM strain of MHV, MHV-induced demyelination is thought to be primarily immune mediated (10, 33). Demyelination can be completely eliminated in JHM-infected, recombinase-activating gene-deficient (RAG–/–) mice that lack functional T and B cells, and this can be reversed upon transfer of splenocytes from immunocompetent mice (35). It has also been shown by depletion and transfer studies in the JHM model that either CD4+ or CD8+ T cells can induce demyelination (35). On the contrary, A59-induced demyelination has been shown to develop in the absence of B and T cells (18). Furthermore, depletion of CD4+ or CD8+ T cells after the acute stage of the infection does not reduce demyelination (30). Thus, the role of CD8+ T cells in promoting A59-induced demyelination has not been clearly defined, and it appears that the two closely related strains of MHV may induce demyelination via unique mechanisms.
Our hypothesis is that A59-induced neurovirulence is determined by the extent of virus replication and spread of antigen throughout the CNS; thus, we predicted that epitope-specific CD8+ T cells would impede virus growth and lessen disease severity in the acute phase of infection. To test this hypothesis and to better understand whether virus-specific CD8+ T cells are involved in MHV strain A59-induced CNS disease, we have established an adoptive-transfer model. In this model, mice are infected with RA59-gfp/gp33, a recombinant form of A59 that expresses the H-2Db-restricted CD8+ T-cell epitope gp33-41 (referred to as gp33); this epitope is derived from the glycoprotein of lymphocytic choriomeningitis virus (LCMV). As a source of transferred cells, we used the P14 transgenic mouse, which produces CD8+ T cells specific for this epitope (3). Thus, we adoptively transferred splenocytes from P14 mice into B6 recipients 1 day prior to infection with RA59-gfp/gp33 or during the acute phase of infection, on day 3 or 5 p.i., to determine the kinetics of CD8+ T-cell-mediated protection and/or pathogenesis. Overall, we observed less disease and lower virus titers, as well as reduced spread of virus in the CNS of those animals receiving epitope-specific CD8+ T cells 1 day before infection with RA59-gfp/gp33; this correlated with a robust gp33-specific CD8+ T-cell response within the infected brain. We also observed reduced spinal cord demyelination at 4 weeks p.i. in the animals that received the P14 cells 1 day prior to infection. Interestingly, adoptive transfer of P14 splenocytes into acutely infected B6 mice, on either day 3 or 5 p.i., did not cause either an increase or a decrease in viral titers, virus spread, or demyelination compared to those in RA59-gfp/gp33-infected mice that did not receive P14 cells. Thus, we demonstrated that virus spread to the spinal cord white matter occurred very rapidly and protection from demyelination correlated with the recruitment of high numbers of virus-specific CD8+ T cells that inhibited virus spread to the spinal cord during the acute phase of infection.
MATERIALS AND METHODS
Mice and viruses. Four-week-old male mice were used in all experiments; B6 or B6-LY5.2/Cr (CD45.1) mice were obtained from the National Cancer Institute, and ?2M–/– mice (B6 background) were obtained from Jackson Laboratories. P14 mice (3) were bred at the University of Pennsylvania. Recombinant MHV strain A59 expressing enhanced green fluorescent protein (EGFP) (referred to here as RA59-gfp) was derived from MHV strain A59 by targeted recombination as previously described (6). Selection of RA59-gfp expressing the LCMV gp33 epitope as a fusion protein with EGFP, referred to as RA59-gfp/gp33, is described elsewhere (5). Briefly, a double-stranded DNA fragment containing the coding region for the first nine complementary synthetic oligonucleotides and 63 amino acids of MHV ORF4a (including the initiation AUG codon), followed by the coding region for the 9 amino acids of gp33, was generated from complementary synthetic oligonucleotides 63 bases in length. This fragment was ligated into pEGFP (Clontech) such that the ORF4a/gp33 fragment was adjacent to and in frame with the gene for EGFP. The fragment containing ORF4a/gp33/EGFP was cleaved from this plasmid with SalI and NotI and inserted into pMH54-EGFP (6) to replace most of gene 4. Targeted recombination of donor RNA transcribed from pMH54-EGFP containing the gp33 sequences and the recipient virus fMHV was carried out with feline FCWF cells (5). Recombinant viruses were selected by two rounds of plaque purification on L2 cells, and sequencing was performed to verify the presence of gp33 (5).
Isolation of mononuclear cells from spleens for adoptive transfer. Spleens were removed from P14 mice. Suspensions were prepared by homogenizing spleens in a nylon bag (mesh opening size of 64 μm) with a syringe plunger in RPMI 1640 medium supplemented with 1% fetal calf serum. Red blood cells were lysed with 0.83% ammonium chloride, and the lymphocyte suspension was washed twice in 1x phosphate-buffered saline (PBS) and resuspended in 1x PBS for transfer.
Intravenous (i.v.) injection and i.c. inoculation of mice. Adoptive transfer of 2 x 107 cells in 0.5 ml of 1x PBS to 4-week-old B6 mice was achieved through i.v. injection via the tail. Mice received P14 cells 1 day prior to infection with A59-gfp/gp33 or A59-gfp. For i.c. infections, mice were anesthetized with isoflurane, and 30 μl containing 105 PFU of virus diluted in 1x PBS-0.75% bovine serum albumin was injected into the left cerebrum.
Virus replication in mice. To measure in vivo virus replication, mice were sacrificed on days 3, 5, 7, 10, and 28 p.i. Mice were perfused with 10 ml of 1x PBS, and their brains were removed. The left half of each brain was placed in 3 ml of gel saline (an isotonic saline solution containing 0.167% gelatin), weighed, and stored frozen at –80°C. Brains were subsequently homogenized, and standard plaque assays were performed with L2 mouse fibroblast monolayers (12). The right half of each brain was placed in 10% phosphate-buffered formalin to fix for histology and viral antigen staining.
Histologic and immunohistochemical analyses. The right halves of brains and entire spinal cords from animals sacrificed at days 3, 5, and 7 p.i. were fixed in formalin, embedded in paraffin, sectioned, and stained for viral antigen or inflammation. Antigen staining was performed by the avidin-biotin-immunoperoxidase technique (Vector Laboratories, Burlingame, Calif.) by using diaminobenzidine tetrahydrochloride as the substrate and a 1:20 dilution of rabbit antinucleocapsid monoclonal antibody (kindly provided by Julian Leibowitz). All slides were read in a blinded manner, and groups of at least four mice were examined in two separate experiments for each virus and treatment. Hematoxylin-and-eosin (H-and-E) staining for inflammation was carried out, and analysis was performed in a blinded manner by a neuropathologist. The encephalitis scores used to classify inflammation were as follows: 0, no specific pathological changes within the brain parenchyma; 1, few perivascular cuffs with minimal extension into the parenchyma and some vacuolation and cell loss; 2, same as 1 except more areas affected; 3, neuropil vacuolation and/or cell loss; 4, widespread vacuolation and/or cell loss.
Isolation of mononuclear cells from the brain for analysis. Mononuclear cells from the brain were prepared as previously described (5, 24) on days 7, 10, 12, and 28 p.i. Cells harvested from five or six brains per group were pooled. Cells were passed through a 30% Percoll gradient and then passed through a cell strainer (70-μm pore diameter; Becton Dickinson). The cell suspension was layered atop a 2-ml Lymphocyte-M (Cedarlane Laboratories) cushion, and viable cells were removed from the interface, washed with 1x PBS, and counted.
Intracellular IFN- staining and flow cytometry analysis. Intracellular IFN- secretion was assayed in response to stimulation with specific peptides as previously described (19, 25). Brain-derived mononuclear cells or splenocytes (106) were cultured with 10 U of human recombinant interleukin-2-1 μl of brefeldin A (Golgiplug; PharMingen) per ml either with or without 1 μg of specific peptide per ml in a total volume of 200 μl of RPMI 1640 medium supplemented with 5% fetal calf serum for 5 h at 37°C. Cells were then stained for surface expression of CD8, CD4, and/or CD45.2 with monoclonal antibodies specific for CD8a (clone 53-6.7), CD4 (clone RM4-5), and CD45.2 (clone 104) (PharMingen). After surface staining, intracellular IFN- was detected by first fixing and then permeabilizing cells with the Cytofix/Cytoperm kit (PharMingen) and stained with a fluorescein isothiocyanate-conjugated monoclonal rat anti-mouse IFN- antibody (clone XMG 1.2; PharMingen). Cells were analyzed with a FACScan flow cytometer (Becton Dickinson). The total number of cells positive for IFN- per mouse was determined by multiplying the fraction of cells positive for IFN- by the total number of live cells isolated per brain.
Demyelination. Demyelination was analyzed for both quantity and severity. Analysis of demyelination was performed on spinal cords harvested from animals 28 days p.i. Five to eight mice were examined in each of two separate experiments. After sacrifice, mice were perfused with 10 ml of 1x PBS and their spinal cords were removed. Spinal cords were sectioned into five regions representing the cervical through lumbar regions and embedded in paraffin for sectioning. Cross sections were stained with a myelin-specific dye, luxol fast blue. To determine the percentage of demyelinated spinal cord, quadrants of spinal cord were counted. At least 10 sections of cord were counted for each animal. A description of the pathology score used to analyze the severity of demyelination is as follows: 0, no demyelination; 1, small foci of demyelination seen in less than 1 section per slide (containing a total of 10 sections of cord); 2, one to three small areas of demyelination per slide; 3, at least one large region of demyelination per slide and all sections containing some demyelination; 4, ample demyelination involving all sections with one to three large areas of demyelination; 5, all sections involved with more than three large regions of demyelination, confluent and involving the medulla oblongata.
RESULTS
Reduced weight loss in early transfer recipients infected with RA59-gfp/gp33. In order to assess the role of epitope-specific CD8+ T cells during the acute phase of infection with A59, we developed an adoptive-transfer system with splenocytes from P14 mice, which produce gp33-specific CD8 T cells (3), and a recombinant form of strain A59 that expresses the gp33 epitope as a fusion protein with EGFP (RA59-gfp/gp33). Figure 1A shows schematic diagrams of the genomes of RA59-gfp/gp33 and control virus RA59-gfp (5). Nonessential gene 4 was replaced with a small piece of ORF4a and the gfp/gp33 fusion. Elimination of expression of ORF4 has previously been shown to have no effect on virulence of MHV in the CNS (21). As shown in Fig. 1B, in the adoptive-transfer model, 4-week-old B6 mice received 2 x 107 P14 splenocytes i.v. 1 day prior to i.c. infection or on day 3 or 5 p.i. with 105 PFU of either RA59-gfp/gp33 or control virus RA59-gfp, which does not express the gp33 epitope. Control animals that did not receive an adoptive transfer were also infected with either RA59-gfp/gp33 or RA59-gfp. To assess clinical symptoms, mice were weighed prior to infection and on various days p.i. Compared to mock-infected animals, mice infected with RA59-gfp/gp33 lost significant weight by day 7 p.i. (Fig. 2A). Animals that received gp33-specific CD8+ T cells prior to infection with the gp33-expresssing virus lost significantly less weight by day 7 p.i. (P < 0.001) than did mice receiving a transfer on day 3 or 5 p.i. and those animals not receiving a transfer (Fig. 2A). RA59-gfp-infected mice lost similar amounts of weight with or without a transfer of P14 cells (Fig. 2B). Thus, adoptive transfer of epitope-specific cells prior to, but not during, infection with MHV protected against the severe weight loss associated with the acute phase of infection.
Virus titers, antigen spread, and encephalitis in the brain after transfer of epitope-specific CD8+ T cells. First, we measured the effect of the transferred epitope-specific CD8+ T cells on viral titers in the brains of infected mice. Compared to mice that did not receive P14 cells, those that received the transfer prior to infection had a significant (P < 0.001) reduction in virus titers of approximately 20-fold on day 5 p.i. (Fig. 3A). In addition, animals that received the P14 cell transfer on day 3 p.i. had titers similar to those of the nonrecipients and their titers were significantly greater than those of the animals that received the transfer prior to infection (P < 0.05). However, on day 7 p.i., the titers in all groups were low and no significant differences were observed (Fig. 3A). Control mice infected with RA59-gfp that received gp33-specific CD8+ T cells exhibited no difference in titers compared to those of nonrecipients, as expected, demonstrating that the CD8+ T-cell response is specific. Thus, the transfer of naive, epitope-specific CD8+ T cells was capable of protecting against virus replication when the transfer was done i.v. prior to infection. In all RA59-gfp/gp33- and RA59-gfp-infected animals, with or without having received the P14 cells, titers were undetectable on days 10 and 28 p.i. (data not shown).
We have previously observed that spread of viral antigen in the brain, compared with viral titers, is a more accurate indication of the severity of CNS damage (25, 31). In order to assess virus spread, we examined viral antigen in sagittal sections of brains that were harvested at different times p.i. with a monoclonal antibody directed at the viral nucleocapsid protein. The peak of virus antigen expression was on day 5 p.i. in the brain; virus antigen staining was very low on day 3 p.i., despite high titers, and was almost undetectable on day 7 in all of the sections examined from each group (data not shown). Several regions of the brain that were consistently involved during the infection included the basal ganglia, the olfactory bulb, the subiculum, the hypothalamus, and the hindbrain, including the medulla and pons. Correlating with the difference in titers observed on day 5, dramatically less antigen staining was observed throughout the brain, as represented in the subiculum and hindbrain, in mice that received gp33-specific CD8+ T cells prior to infection (Fig. 4B and E). Only small foci of antigen-positive cells existed in examined sections from animals that received the transfer prior to infection. However, animals that received the transfer on day 3 p.i. had an amount of viral antigen similar to that in mice that did not receive a transfer, as shown in the subiculum and the hindbrain (Fig. 4). As expected, control animals infected with RA59-gfp had similar levels of antigen with or without having received the transfer (data not shown).
Encephalitis was determined by H-and-E staining of sagittal brain sections and spinal cord cross sections taken on day 7 p.i., the typical peak of inflammation. In agreement with observations of reduced weight loss in RA59-gfp/gp33-infected mice that received the adoptive transfer prior to infection (Fig. 2A), we observed a significant reduction in inflammation (Table 1). H-and-E-stained sections from mice that received the transfer prior to infection revealed a reduction in infiltrating inflammatory cells, little to no pathological changes within the parenchyma, and minimal perivascular cuffing compared to all other groups of infected mice (Table 1).
Early transfer of epitope-specific CD8+ T cells blocks virus spread to the spinal cord. Previous studies with the JHM strain of MHV demonstrated that the virus spreads from the brain into the spinal cord between days 7 and 10 p.i. (23). However, virus spread from the brain to the spinal cord during the acute phase of infection with the A59 strain of MHV has not been examined. Thus, we examined spinal cords harvested from infected mice at days 5 and 7 p.i. to determine the kinetics of viral antigen spread. Immunohistochemical analysis with the antinucleocapsid antibody was performed on cross sections of formalin-fixed spinal cord. Robust viral antigen expression was detected on day 5 p.i. in RA59-gfp/gp33-infected animals that did not receive the adoptive transfer (Fig. 5A and B) or received the day 3 p.i. adoptive transfer (Fig. 5E and F), but no antigen was detected in animals that received the transfer prior to infection (Fig. 5C and D). On day 7 p.i., viral antigen was still detectable in all of the animals except those that received the transfer prior to infection (data not shown). Virus antigen was detected in both the gray and white matter on day 5 p.i. (Fig. 5), whereas on day 7 p.i. virus antigen was only detected in the white matter (data not shown). Thus, corresponding to the detection of virus antigen in the brain, virus spread to the spinal cord is reduced only when gp33-specific CD8+ cells are transferred prior to infection.
Early recruitment and activation of gp33-specific CD8+ T cells correlates with reduced CNS disease. To evaluate the degree to which the recombinant virus could elicit an immune response to the introduced gp33 epitope and thus how well the transferred cells engrafted, we sought to determine the activation status of virus-specific CD8+ T cells in animals that had or had not received the transfer of P14 cells. In order to accurately assess the role of the transferred epitope-specific CD8+ T cells at the site of infection, we obtained mononuclear cells from the brains of infected mice on day 7 p.i., the typical peak of lymphocyte infiltration in the brain after i.c. infection with A59. More than half (51.7%) of the CD8+ T cells harvested from the brains of RA59-gfp/gp33-infected mice that received the adoptive transfer of gp33-specific CD8+ T cells were specific for gp33 on day 7 p.i. (Fig. 6A). In contrast, only 14.8% of the brain-derived CD8+ T cells were activated and specific for the gp33 epitope in mice that did not receive the adoptive transfer; this reflected the normal primary immune response and is similar to previously reported results (5). Interestingly, mice that received the adoptive transfer on either day 3 or 5 p.i. had percentages of gp33-specific CD8+ T cells similar to those of animals that did not receive the transfer (15.7 and 14.5%, respectively).
In order to understand how the response to an immunodominant epitope may affect the response to other, less dominant epitopes, we analyzed the response to the endogenous epitope in the spike protein S598. Animals infected with RA59-gfp, lacking the gp33 epitope, contained a higher percentage of CD8+ T cells specific for the endogenous S598 epitope than did the RA59-gfp/gp33-infected animals (Fig. 6B). RA59-gfp/gp33-infected mice that received the transfer prior to infection exhibited the lowest percentage of S598-specific CD8+ T cells (1.2%), perhaps because of the very high number of activated gp33-specific CD8+ T cells (Fig. 6A).
In order to determine if the lower numbers of gp33-specific CD8+ T cells observed in the mice that received the adoptive transfer on day 3 or 5 p.i., compared to the group that received the transfer prior to infection, was due to the kinetics of activation and recruitment into the brain, we analyzed brain-derived mononuclear cells for gp33-specific CD8+ T cells on days 10 and 12 p.i. (which corresponds to day 7 posttransfer for each transfer recipient group). To distinguish between the transferred P14 cells and the endogenous gp33-specific CD8+ T cells, we used B6-LY5.2/Cr mice (CD45.1) as recipients; thus, the CD45.2-positive cells reflect the adoptively transferred P14 cells. We observed an overall increase in the percentage of gp33-specific CD8+ T cells on day 12 (Fig. 7, bottom row) compared to day 10 (Fig. 7, top row), and this reflects a decrease in the total number of CD8+ T cells in the brain and retention of virus-specific CD8+ T cells. As expected, no CD45.2 cells were detected in animals that did not receive the adoptive transfer (Fig. 7). The majority of the IFN--secreting, gp33-specific CD8+ T cells were P14 in origin (CD45.2+) in the mice that received the transfer prior to infection. Interestingly, a significantly smaller percentage of the gp33-CD45.2+ gp33-gp33-specific CD8+ T cells were observed in the mice that received the adoptive transfer on day 3 p.i. than in the mice that received the transfer prior to infection (2.24% compared with 33.7%). Furthermore, P14-derived gp33-specific CD8+ cells were completely undetectable on both days 10 and 12 p.i. in mice that received the adoptive transfer on day 5 p.i. Thus, activation and trafficking of naive, virus-specific CD8+ T cells appeared to be limited to the first 3 days of the infection.
Demyelination is reduced in animals protected from acute disease. A focus of this study was to understand how a strong CD8+ T-cell response could shape the outcome of the chronic disease induced by A59. Earlier depletion studies indicated that when CD8+ T cells were depleted after resolution of acute A59 infection, demyelination was nearly the same as in nondepleted mice, which suggested the possibility that CD8+ T cells could initiate disease during the acute phase of infection (30). Mice infected with RA59-gfp/gp33 or RA59-gfp received transfers the day prior to infection or on either day 3 or day 5 p.i. and were sacrificed 4 weeks p.i., at which time their spinal cords were removed, fixed, and stained with a myelin-specific dye, luxol fast blue. Demyelination was assayed by counting quadrants of cross-sectioned spinal cord. We also used a pathology score to describe the severity of demyelination of the spinal cord; the scores ranged from 0, no demyelination evident, to 5, pathology involving severe, confluent demyelination of nearly the entire spinal cord (see Materials and Methods) (Fig. 8).
Mice protected from acute disease via adoptive transfer of gp33-specific CD8+ T cells 1 day prior to infection had significantly less demyelination, as measured by both the percentage of quadrants containing demyelination and the severity of demyelination, than did those that were not protected from acute disease (non-transfer recipients and day 3 and 5 p.i. transfer recipients) (Fig. 8A). The levels of demyelination observed in animals infected with RA59-gfp did not differ, regardless of the adoptive transfer of P14 cells, as expected (Fig. 8B). Thus, protection from demyelination correlated with reduced acute disease, including reduced virus spread.
We observed that the percentage of demyelinated spinal cord and the severity of spinal cord demyelination were greater in the RA59-gfp/gp33-infected mice than in RA59-gfp-infected animals. We reasoned that this could be due to either a small difference in the kinetics of virus replication and spread or possibly to a pathogenic immune response elicited by the gp33-expressing virus. Infections of ?2M–/– mice were carried out to determine whether a detrimental major histocompatibility complex class I-restricted immune response was elicited in RA59-gfp/gp33-infected mice. ?2M–/– mice were infected with 250 PFU of either RA59-gfp/gp33 or RA59-gfp, as these mice are highly susceptible to MHV infection. Surprisingly, despite the absence of detectable differences in either virus titers in the brain or virus antigen detection in the brain during the first week of infection (data not shown), RA59-gfp/gp33 still induced significantly more demyelination than did RA59-gfp (Fig. 9). This suggests that demyelination induced by RA59-gfp/gp33 may be more severe because of an intrinsic property of the virus other than immune recognition of the gp33 epitope expressed by RA59-gfp/gp33; however, we cannot completely rule out the possibility that CD8+ T cells can contribute to demyelination in these mice, as they do exist in low numbers.
Retention of epitope-specific CD8+ T cells within the brain after CNS infection. CD8+ T cells have been shown to persist in an activated, IFN--secreting state within the brains of mice after infection with MHV (15), and it has been shown that this is dependent upon the persistence of viral RNA (16). To more clearly understand how the acute phase of infection, as shaped by the transfer of P14 cells, affected the retention of CD8+ T cells during the chronic phase of infection and to determine if retained epitope-specific CD8+ T cells correlated with demyelination, brains were removed 4 weeks p.i. for isolation and analysis of localized inflammatory cells. Brain-derived cells were analyzed for intracellular IFN- secretion in response to in vitro peptide stimulation. The total numbers of CD8+ T cells in the brains of all of the animals were similar, although the total number of virus-specific, IFN--secreting CD8+ T cells was about 10-fold higher in RA59-gfp/gp33-infected animals than in RA59-gfp-infected animals, likely because of the response to the immunodominant gp33 epitope in the former (data not shown). In RA59-gfp/gp33-infected animals, the levels of activated gp33-specific IFN--secreting CD8+ T cells were similar but highest in the mice that received the transfer prior to infection (Fig. 10). Thus, we observed that a similar, high number of gp33-specific, IFN--secreting CD8+ T cells was retained in all groups. The percentage of gp33-specific CD8+ T cells did not appear to correlate with the gp33-specific CD8+ T-cell response observed during the acute phase of infection (on day 7, 10, or 12 p.i.; Fig. 6 and 7). A clear correlation between the presence of activated gp33-specific CD8+ T cells and demyelination could not be made; however, it was noted that the day –1 transfer recipients had the lowest levels of demyelination and the highest percentage of gp33-specific CD8+ T cells, whereas the lowest percentage of gp33-specific CD8+ T cells was observed in animals that demonstrated high levels of severe demyelination (no-transfer group and day 3 and 5 p.i. transfer groups; Fig. 8 and 10).
DISCUSSION
Infection with the A59 strain of MHV provides a model for both viral encephalitis and demyelination, but despite extensive study, the mechanism of A59-induced demyelination is still not fully understood. The findings presented here have highlighted several factors that are important in understanding the pathogenesis of A59-induced demyelination, including the relationship between virus spread during the acute phase of infection and its outcome in terms of demyelination, as well as how the kinetics of CD8+ T-cell activation and recruitment to the brain affect virus spread within the CNS.
We have previously examined the role of CD8+ T-cell-mediated protection from A59 infection by a vaccination strategy (5); however, the adoptive-transfer system with splenocytes derived from the P14 mouse has allowed us to go a step further and determine the effects of T cells, specific for one epitope, on pathogenesis. It has also allowed the kinetic analysis of activation and recruitment of naive CD8+ T cells to the infected CNS. In this study we assessed protection from CNS disease achieved by adoptive transfer of naive gp33-specific splenocytes into mice prior to or during the acute phase of infection with a gp33-expressing recombinant form of strain A59. Epitope-specific CD8+ T cells inhibited virus growth and viral antigen spread in the brain and reduced disease severity and inflammation only when transferred prior to infection. This correlated with an increased activated gp33-specific CD8+ T-cell response in the brain. In animals protected from acute disease, there was a dramatic reduction in the level and severity of demyelination at 1 month p.i. This is in agreement with other studies of JHM infection that demonstrated virus spread within the CNS as a major determinant in the development of chronic demyelination (15). Similarly, studies using the transfer of nucleoprotein-specific cytotoxic T lymphocytes was effective at reducing JHM replication within the CNS (28).
Priming of virus-specific CD8+ T cells has been shown to occur very rapidly after infection with a variety of pathogens, and here we show that the window of naive CD8+ T-cell priming, in the context of a virus infection within the CNS, is limited to the first 3 days of infection. Wong and Pamer recently demonstrated a feedback mechanism during infection with Listeria monocytogenes by which antigen-presenting cells are eliminated by effector cytotoxic T lymphocytes, thereby limiting the CD8+ T-cell response (34). It is possible that this accounts for the absence of CD45.2-positive, gp33-specific CD8+ T cells observed in the brains of RA59-gfp/gp33-infected mice that had received an adoptive transfer on day 5 p.i.
CD8+ T-cell expansion is thought to be initiated in the cervical lymph nodes, the major site of T-cell priming during CNS infections, and then expansion continues in the spleen, followed by rapid accumulation of virus-specific T cells in the CNS during MHV infection (17). Norbury et al. showed that virus-specific CD8+ T cells are primed by dendritic cells, but not macrophages, in vivo after infection with vaccinia virus (20). We do not know which antigen-presenting cells are responsible for priming naive CD8+ T cells after MHV infection of the CNS. Analysis of splenocytes on days 10 and 12 p.i. revealed the presence of transferred P14 cells in this organ in all transfer recipients regardless of the day p.i. on which the cells were transferred (data not shown), although the percentage of activated spleen-derived P14 cells was dependent upon the day of transfer. CD8+ CD45.2+ T cells were present in the spleens of both transfer groups (days 3 and 5 p.i.); however, only cells harvested from mice that received the transfer on day 3 p.i. secreted IFN- in response to gp33 peptide (data not shown), which suggests that failure of the day 5 p.i. transfer to become activated was due to a deficit in priming.
The increased virus-specific CD8+ T-cell response observed in protected mice limited virus replication and spread within the CNS of RA59-gfp/gp33-infected animals. At day 5 p.i., virus titers and viral antigen staining were reduced in these animals. In some CNS infections, such as Borna disease virus and i.c. infection of mice with LCMV, CD8+ T cells can play a detrimental role during the infection; however, virus-specific CD8+ T cells primarily control virus replication and protect against virus spread, CNS damage, and encephalitis in MHV strain A59-infected mice (2, 4). Thus, during the acute phase of MHV infection of the CNS, virus-specific CD8+ T cells play a crucial role in protection from both acute and chronic disease and a stronger CD8+ T-cell response is beneficial to the host.
Virus antigen was detected throughout the brain and in the gray and white matter of the spinal cord with similar kinetics, peaking on day 5 p.i. and diminishing by day 7 p.i. Virus antigen was still detected in the spinal cords on day 7 p.i. but was present only in the white matter. On the basis of studies of JHM, virus spread from the brain into the spinal cord is thought to occur in a retrograde fashion via transneuronal spread; the rapid appearance of virus in the spinal cord, as observed here, is consistent with this. Whether the CD8+ T-cell-mediated block in virus spread to the spinal cord reflected an overall diminished load of virus within the CNS or whether CD8+ T cells mediated clearance of particular cell types is not known.
Clearance of infectious virus from the CNS requires both perforin- and cytokine-mediated mechanisms. Interestingly, our vaccination studies with perforin-deficient mice with recombinant L. monocytogenes expressing the gp33 epitope, followed by a challenge with RA59-gfp/gp33, revealed that perforin was necessary for protection, as determined by virus titers in the brain (5). This is interesting since the majority of A59-infected cells are positive for the neuronal marker dendrite-specific microtubule-associated protein 2 (25). Perforin is required for clearance from microglia and astrocytes, which express major histocompatibility complex class I; thus, these cell types may support the majority of virus replication. It is unclear how adoptively transferred P14 CD8+ T cells mediated protection from virus spread via transneuronal pathways, but it likely involved cytokine-mediated effects and is of current interest.
Interestingly, we found that the retention of IFN--secreting gp33-specific CD8+ T cells at 4 weeks p.i. was similar in all RA59-gfp/gp33-infected animals regardless of transfer of gp33-specific CD8+ T cells or the amount of virus replication and spread that occurred during the acute phase of infection. In addition, both recipients and nonrecipients had similar percentages of S598-specific CD8+ T cells despite having very different levels of S598-specific CD8+ T cells at the peak of infiltration during the acute phase of infection. Thus, the level of replication and spread during the acute phase of infection appeared to have little if any effect on the retention of CD8+ T cells during the chronic phase of infection. It has been shown in several viral infections of the CNS that CD8+ T cells are maintained within the brain after the acute phase of infection is resolved (11, 32). The function of these retained virus-specific CD8+ T cells is not clear, but as van der Most et al. (32) suggest in a model of dengue virus infection of the mouse CNS, the brain may retain local, activated, virus-specific CD8+ T cells with an effector memory phenotype as an in situ rapid response force.
Factors contributing to the maintenance of virus-specific CD8+ T cells in the CNS are also largely unknown. Retention of CD8+ T cells within the CNS of JHM-infected animals is thought to be largely dependent on the presence of viral RNA and possibly very low levels of antigen expression (16). In JHM-infected mice that recognize both H-2Db- and H-2Dd-restricted epitopes, the epitope specificity of CD8+ T cells changes from the acute to the chronic phase of infection with a decrease in H-2Dd-restricted nucleocapsid-specific CD8+ T cells and an increase in H-2Db spike-specific CD8+ T cells suggestive of altered antigen presentation during the chronic phase of infection (1). However, in the present study animals that were protected from acute disease via adoptive transfer exhibited minimal virus replication and spread, and although possible, it is unlikely that persistent viral RNA was maintaining the population of activated CD8+ T cells in these mice.
The severity of demyelination in the chronic phase of infection did not correspond to the levels of activated virus-specific CD8+ T cells. At 4 weeks p.i., total numbers and percentages of epitope-specific CD8+ T cells in the brain were similar in transfer recipients and nonrecipients infected with either RA59-gfp/gp33 or RA59-gfp. However, demyelination was much less severe in animals protected from virus replication and spread, P14 splenocyte recipients infected with RA59-gfp/gp33, despite a strong immune response during the acute phase of infection. This supports previous findings that CD8+ T cells likely do not play a direct role in the process of demyelination (9, 15, 30).
The outcome of the acute stage of A59-induced CNS disease depends on the rapid recruitment and activation of CD8+ T cells, and a good prognosis is linked with limited spread of the virus within the CNS, specifically to the spinal cord. Preventing virus spread to the spinal cord reduced levels of demyelination seen in the chronic stage of disease, suggesting that the presence of viral RNA and/or low levels of viral antigen are required for demyelination to occur. This study also demonstrated that the initial priming event after infection with a neurotropic pathogen was critical in establishing a long-lasting protective immune response in the brain. Thus, the immediate host immune response is critical and determines the outcome of A59 infection.
ACKNOWLEDGMENTS
This work was supported by NIH grants AI 47800, AI 60021, NS 30606 (S.R.W.), and AI 45025 (H.S.), and K.C.M. was supported by NIH training grant AI 007324.
We thank Devon J. Shedlock and Lauren A. Zenewicz for providing P14 mice.
REFERENCES
Bergmann, C. C., J. D. Altman, D. Hinton, and S. A. Stohlman. 1999. Inverted immunodominance and impaired cytolytic function of CD8+ T cells during viral persistence in the central nervous system. J. Immunol. 163:3379-3387.
Bilzer, T., and L. Stitz. 1994. Immune-mediated brain atrophy. CD8+ T cells contribute to tissue destruction during Borna disease. J. Immunol. 153:818-823.
Brandle, D., K. Burki, V. A. Wallace, U. H. Rohrer, T. W. Mak, B. Malissen, H. Hengartner, and H. Pircher. 1991. Involvement of both T cell receptor V alpha and V beta variable region domains and alpha chain junctional region in viral antigen recognition. Eur. J. Immunol. 21:2195-2202.
Buchmeier, M. J., R. M. Welsh, F. J. Dutko, and M. B. Oldstone. 1980. The virology and immunobiology of lymphocytic choriomeningitis virus infection. Adv. Immunol. 30:275-331.
Chua, M., K. MacNamara, L. SanMateo, H. Shen, and S. R. Weiss. 2004. Effects of an epitope-specific CD8+ T-cell response on murine coronavirus central nervous system disease: protection from virus replication and antigen spread and selection of epitope escape mutants. J. Virol. 78:1150-1159.
DasSarma, J., E. Scheen, S.-H. Seo, M. Koval, and S. R. Weiss. 2002. Enhanced green fluorescent protein expression may be used to monitor murine coronavirus spread in vitro and in the mouse central nervous system. J. Neurovirol. 8:1-11.
Drosten, C., S. Gunther, W. Preiser, S. van der Werf, H. R. Brodt, S. Becker, H. Rabenau, M. Panning, L. Kolesnikova, R. A. Fouchier, A. Berger, A. M. Burguiere, J. Cinatl, M. Eickmann, N. Escriou, K. Grywna, S. Kramme, J. C. Manuguerra, S. Muller, V. Rickerts, M. Sturmer, S. Vieth, H. D. Klenk, A. D. Osterhaus, H. Schmitz, and H. W. Doerr. 2003. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N. Engl. J. Med. 348:1967-1976.
Fleming, J. O., J. J. Houtman, H. Alaca, D. Hinze, D. McKenzie, J. Aiken, T. Bleasdale, and S. Baker. 1993. Persistence of viral RNA in the central nervous system of mice inoculated with MHV-4. Adv. Exp. Med. Biol. 342:327-332.
Gombold, J., R. M. Sutherland, E. Lavi, Y. Paterson, and S. R. Weiss. 1995. Mouse hepatitis virus A59-induced demyelination can occur in the absence of CD8+ T cells. Microb. Pathog. 18:211-221.
Haring, J. S., L. L. Pewe, and S. Perlman. 2002. Bystander CD8 T cell-mediated demyelination after viral infection of the central nervous system. J. Immunol. 169:1550-1555.
Hawke, S., P. G. Stevenson, S. Freeman, and C. R. Bangham. 1998. Long-term persistence of activated cytotoxic T lymphocytes after viral infection of the central nervous system. J. Exp. Med. 187:1575-1582.
Hingley, S. T., J. L. Gombold, E. Lavi, and S. R. Weiss. 1994. MHV-A59 fusion mutants are attenuated and display altered hepatotropism. Virology 200:1-10.
Houtman, J. J., and J. O. Fleming. 1996. Pathogenesis of mouse hepatitis virus-induced demyelination. J. Neurovirol. 2:361-376.
Lavi, E., P. S. Fishman, M. K. Highkin, and S. R. Weiss. 1988. Limbic encephalitis after inhalation of a murine coronavirus. Lab. Investig. 58:31-36.
Marten, N. W., S. A. Stohlman, R. D. Atkinson, D. R. Hinton, J. O. Fleming, and C. C. Bergmann. 2000. Contributions of CD8+ T cells and viral spread to demyelinating disease. J. Immunol. 164:4080-4088.
Marten, N. W., S. A. Stohlman, and C. C. Bergmann. 2000. Role of viral persistence in retaining CD8+ T cells within the central nervous system. J. Virol. 74:7903-7910.
Marten, N. W., S. A. Stohlman, J. Zhou, and C. C. Bergmann. 2003. Kinetics of virus-specific CD8+ T-cell expansion and trafficking following central nervous system infection. J. Virol. 77:2775-2778.
Matthews, A. E., E. Lavi, S. R. Weiss, and Y. Paterson. 2002. Neither B cells nor T cells are required for CNS demyelination in mice persistently infected with MHV-A59. J. Neurovirol. 8:257-264.
Murali-Krishna, K., J. D. Altman, M. Suresh, D. J. Sourdive, A. J. Zajac, J. D. Miller, J. Slansky, and R. Ahmed. 1998. Counting antigen-specific CD8 T cells: a reevaluation of bystander activation during viral infection. Immunity 8:177-187.
Norbury, C. C., D. Malide, J. S. Gibbs, J. R. Bennink, and J. W. Yewdell. 2002. Visualizing priming of virus-specific CD8+ T cells by infected dendritic cells in vivo. Nat. Immunol. 3:265-271.
Ontiveros, E., L. Kuo, P. S. Masters, and S. Perlman. 2001. Inactivation of expression of gene 4 of mouse hepatitis virus strain JHM does not affect virulence in the murine CNS. Virology 289:230-238.
Parra, B., D. R. Hinton, N. W. Marten, C. C. Bergmann, M. T. Lin, C. S. Yang, and S. A. Stohlman. 1999. IFN- is required for viral clearance from central nervous system oligodendroglia. J. Immunol. 162:1641-1647.
Perlman, S., G. Jacobsen, and A. Afifi. 1989. Spread of a neurotropic murine coronavirus into the CNS via the trigeminal and olfactory nerves. Virology 170:556-560.
Pewe, L., S. B. Heard, C. Bergmann, M. O. Dailey, and S. Perlman. 1999. Selection of CTL escape mutants in mice infected with a neurotropic coronavirus: quantitative estimate of TCR diversity in the infected central nervous system. J. Immunol. 163:6106-6113.
Phillips, J. J., M. M. Chua, G. F. Rall, and S. R. Weiss. 2002. Murine coronavirus spike glycoprotein mediates degree of viral spread, inflammation, and virus-induced immunopathology in the central nervous system. Virology 301:109-120.
Stohlman, S. A., C. C. Bergmann, D. J. Cua, M. T. Lin, S. Ho, W. Wei, and D. R. Hinton. 1998. Apoptosis of JHMV-specific CTL in the CNS in the absence of CD4+ T cells. Adv. Exp. Med. Biol. 440:425-430.
Stohlman, S. A., C. C. Bergmann, M. T. Lin, D. J. Cua, and D. R. Hinton. 1998. CTL effector function within the central nervous system requires CD4+ T cells. J. Immunol. 160:2896-2904.
Stohlman, S. A., C. C. Bergmann, R. C. van der Veen, and D. R. Hinton. 1995. Mouse hepatitis virus-specific cytotoxic T lymphocytes protect from lethal infection without eliminating virus from the central nervous system. J. Virol. 69:684-694.
Sussman, M. A., R. A. Shubin, S. Kyuwa, and S. A. Stohlman. 1989. T-cell-mediated clearance of mouse hepatitis virus strain JHM from the central nervous system. J. Virol. 63:3051-3056.
Sutherland, R. M., M. M. Chua, E. Lavi, S. R. Weiss, and Y. Paterson. 1997. CD4+ and CD8+ T cells are not major effectors of mouse hepatitis virus A59-induced demyelinating disease. J. Neurovirol. 3:225-228.
Tsai, J. C., L. de Groot, J. D. Pinon, K. T. Iacono, J. J. Phillips, S. H. Seo, E. Lavi, and S. R. Weiss. 2003. Amino acid substitutions within the heptad repeat domain 1 of murine coronavirus spike protein restrict viral antigen spread in the central nervous system. Virology 312:369-380.
van der Most, R. G., K. Murali-Krishna, and R. Ahmed. 2003. Prolonged presence of effector-memory CD8 T cells in the central nervous system after dengue virus encephalitis. Int. Immunol. 15:119-125.
Wang, F. I., S. A. Stohlman, and J. O. Fleming. 1990. Demyelination induced by murine hepatitis virus JHM strain (MHV-4) is immunologically mediated. J. Neuroimmunol. 30:31-41.
Wong, P., and E. G. Pamer. 2003. Feedback regulation of pathogen-specific T cell priming. Immunity 18:499-511.
Wu, G. F., A. A. Dandekar, L. Pewe, and S. Perlman. 2000. CD4 and CD8 T cells have redundant but not identical roles in virus-induced demyelination. J. Immunol. 165:2278-2286.(Katherine C. MacNamara, M)