T Cells +CD4 Mrna and High Levels of IL-10-Producing and Is

Home , Cetus

The Inﬂammatory Response to Nonfatal Sindbis Virus Infection of the Nervous System Is More Severe in SJL Than in BALB/c Mice and Is Associated with Low Levels of IL-4 mRNA and High Levels of IL-10-Producing CD4؉ T Cells1

Jennifer F. Rowell and Diane E. Grifﬁn2

SJL mice are susceptible to inflammatory autoimmune diseases of the central nervous system (CNS), while BALB/c mice are relatively resistant. To understand differences in immune responses that may contribute to autoimmune neurologic disease, we compared the responses of SJL and BALB/c mice to infection with Sindbis virus, a virus that causes acute nonfatal encephalomyelitis in both strains of mice. Clearance of virus was similar, but SJL mice developed a more intense inflammatory response in the brain and spinal cord and inflammation persisted for several weeks. Analysis of lymphocytes isolated from brains early after infection showed an absence of NK cells in SJL mice, while both strains of mice showed CD4؉ and CD8؉ T cells. During the second week after infection, CD4؉ T cells increased in SJL mice and the proportion of CD8؉ T cells decreased, while the opposite pattern was seen in BALB/c mice. Expression of IL-10 mRNA was higher and IL-4 mRNA was lower in the brains of infected SJL than in BALB/c mice, while expression of the mRNAs of IL-6, IL-1␤, TNF␣, and the Th1 cytokines IL-2, IL-12, and IFN-␥ was similar. Lymphocytes isolated from the CNS of SJL mice produced large amounts of IL-10. CNS lymphocytes from both strains of mice produced IFN-␥ in response to stimulation with Sindbis virus, but not in response to myelin basic protein. These data suggest that IL-10-producing CD4؉ T cells are differentially recruited to or regulated within the CNS of SJL mice compared with BALB/c mice infected with Sindbis virus, a characteristic that may be related to low levels of IL-4, and is likely to be involved in susceptibility of SJL mice to CNS inflammatory diseases. The Journal of Immunology, 1999, 162: 1624–1632.

he AR339 strain of Sindbis virus (SV)3 is an alphavirus tion is first detectable 3–4 days after infection when mononuclear that causes acute, nonfatal, and usually asymptomatic en- cells start to cross the cerebrovascular endothelium to form T cephalomyelitis in weanling mice. After intracerebral in- perivascular cuffs. The inflammatory response is maximal after oculation, the virus replicates rapidly in the brain and spinal cord, 7–10 days and is usually resolved within 2–3 wk (2). primarily in neurons. A systemic and local immune response is Inflammation in the central nervous system (CNS) is rigorously by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. induced and virus replication is soon controlled with levels of in- controlled by several known mechanisms and probably additional fectious virus becoming undetectable within 7–8 days after infec- mechanisms yet to be identified. The blood brain barrier is a phys- tion. Elimination of infectious virus correlates with the appearance ical barrier that prevents entry of most immune cells. Activated T of neutralizing Ab and the development of a virus-specific mono- cells can pass through the blood brain barrier, but quickly exit the nuclear inflammatory response in the brain and spinal cord (1, 2). CNS in the absence of Ag recognition (3, 4). T cell stimulation by Adoptive transfer experiments using persistently infected immu- Ag present in the CNS is further inhibited due to low levels of nodeficient SCID mice have shown that Ab specific for the E2 MHC molecules, costimulatory molecules, and adhesion mole- glycoprotein of SV mediates clearance of infectious virus, even in cules, although cellular damage and inflammatory mediators can the absence of SV-specific cell-mediated immunity (1), so the role https://www.jimmunol.org increase expression (5, 6). of the inflammatory response in recovery is not clear. Inflamma- SV infection results in varying levels of CNS inflammation de- pending on the strain of mouse infected. Compared with BALB/c mice, SJL mice display more extensive inflammation and this in- Department of Molecular Microbiology and Immunology, Johns Hopkins University School of Hygiene and Public Health, Baltimore, MD 21205 flammation may persist for several weeks to months (7). SJL mice Received for publication June 17, 1998. Accepted for publication October 26, 1998. are also susceptible to several inflammatory autoimmune diseases

Downloaded from The costs of publication of this article were defrayed in part by the payment of page of the CNS such as experimental autoimmune encephalitis (EAE) charges. This article must therefore be hereby marked advertisement in accordance (8, 9) and Theiler’s murine encephalomyelitis virus (TMEV)-in- with 18 U.S.C. Section 1734 solely to indicate this fact. duced demyelination (10), while BALB/c and many other strains 1 This work was supported by Research Grant NS29234, a Javits Neuroscience In- vestigator Award from the National Institutes of Health (D.E.G.), and by a postdoc- of mice are relatively resistant to these diseases. Both of these toral fellowship from the National Multiple Sclerosis Society (J.F.R.). murine demyelinating diseases serve as models for the human de- 2 Address correspondence and reprint requests to Dr. Diane E. Griffin, Department of myelinating disease multiple sclerosis, and more complete under- Molecular Microbiology and Immunology, Johns Hopkins School of Public Health, standing of the inflammatory process may enhance our general 615 N. Wolfe Street, Baltimore, MD 21205. E-mail address: dgriffi[email protected] understanding of autoimmune diseases of the CNS. 3 Abbreviations used in this paper: SV, Sindbis virus; CNS, central nervous system; EAE, experimental autoimmune encephalomyelitis; GAPDH, glyceraldehyde phos- EAE is induced by immunization with whole spinal cord or phate dehydrogenase; IP, IFN-␥-inducible protein; MBP, myelin basic protein; MIP, components of myelin such as myelin basic protein (MBP) or pro- macrophage inflammatory protein; PE, phycoerythrin; PFU, plaque-forming units; PI, postinfection; TMEV, Theiler’s murine encephalomyelitis virus; TUNEL, terminal teolipid protein together with CFA (8, 11). Paralysis is accompa- deoxynucleotidyltransferase-mediated UTP nicked end labeling. nied by inflammation consisting of T cells, macrophages and B

Table I. Abs used for analysis of inﬂammatory cells and for ELISA

Determinant Recognizeda Clone Conjugate Source Method

Macrophage/Monocyte F480 None Hybridoma supernatant IHCb CD45R/B220 RA3-6B2 None, FITC PharMingen, San Diego, CA IHC, FCc CD4 (L3T4) RM4-5 None, FITC PharMingen IHC, FC CD8a (Ly-2) 53-6.7 None, FITC PharMingen IHC, FC CD3e 145-2C11 PE PharMingen FC NK cells DX5 FITC PharMingen FC ␣␤ T cell receptor H57-597 FITC PharMingen FC ␥␦ T cell receptor GL3 FITC PharMingen FC IgM R6-60.2 FITC PharMingen FC Ig polyclonal HRPd Dako, Carpinteria, CA ELISA IgG1 G1-6.5 HRP PharMingen ELISA IgG2a R19-15 HRP PharMingen ELISA IL-10 (capture) JES5-2A5 None PharMingen ELISA IL-10 (detection) SXC-1 Biotin PharMingen ELISA

a All Abs are speciﬁc for mouse determinants. b IHC, immunohistochemistry. c FC, ﬂow cytometry. d HRP, horseradish peroxidase.

ϩ cells. The disease is mediated by CD4 Th1-type T cells that se- erate to large inflammatory foci in Ͼ50% of ϫ10 fields. Scores were in- crete inflammatory cytokines such as IL-2 and IFN-␥ (12). The creased by up to 1 point with abundant parenchymal cellularity. inflammation persists for several days, then resolves. Cryopreserved sections were cut from OCT-embedded tissue and used for immunohistochemistry. Sections were blocked with 0.5% dry milk in TMEV infects neurons and oligodendroglia of the brain and PBS, and endogenous peroxidase activity was quenched with either 1%

spinal cord. Resistant strains of mice clear virus during the acute H2O2 in absolute methanol or 0.3% H2O2 in 0.1% sodium azide. Sections phase of the infection. In SJL mice, virus clearance is incomplete were incubated overnight at 4°C, with primary Ab diluted in PBS contain- resulting in persistent infection of macrophages and glial cells. In ing 0.5% dry milk, washed, and incubated with biotinylated secondary Ab ϩ (Vector Laboratories, Burlingame, CA) in PBS/0.5% dry milk. The signal these mice, TMEV-specific CD4 Th1-type T cells mediate a was amplified using the Vectastain Elite ABC kit (Vector) and detected chronic inflammatory process that results in demyelination and with 3,3Ј-diaminobenzidine (Sigma, St. Louis, MO). For detection of CD8, paralysis (13, 14). the signal was further amplified using TSA-Indirect Tyramide Signal Am- The differences in the development and regulation of immune plification (DuPont NEN, Boston, MA). The primary Abs used in this study responses within the CNS of SJL mice compared with BALB/c are listed in Table I. Sections were examined under code for the level of inflammation, the proportion of positively staining cells in inflammatory mice that increase susceptibility to inflammatory diseases involve lesions, and the relative number of positively staining cells in the paren- genes primarily outside the H-2 region and are only partially de- chyma. Immunohistochemical staining was scored on coded slides using a by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. fined (10). To gain a better understanding of these differences, we scale of 0–3 in which 0 indicates no positive cells, 1 indicates 5–20%, 2 have examined the CNS inflammatory response in SJL and indicates 20–60%, and 3 indicates 60–100% positive inflammatory cells. Scores were adjusted by up to 1 point when the numbers of positively BALB/c mice during acute encephalitis induced by infection with staining cells in the parenchyma were considerably higher or lower than SV, a virus that is not associated with late neurologic disease. The average. immune response to SV results in clearance of virus in both strains of mice, but the inflammatory response to SV is more severe in Isolation of lymphocytes from brain and spleen SJL mice and SJL mice have lower levels of IL-4 mRNA, fewer ϩ Brain lymphocytes were isolated as previously described (15). Briefly, SV- NK cells, and more IL-10-producing CD4 T lymphocytes in re- infected mice were anesthetized with methoxyflurane and perfused with https://www.jimmunol.org sponse to SV infection, suggesting altered immune regulation in PBS. Brains were removed, homogenized through a mesh screen, and col- the CNS. lected in HBSS containing 0.05% collagenase D (Boehringer Mannheim, Indianapolis, IN) and 10 ␮g/ml DNase I (Sigma). The brain homogenate was mixed at room temperature for 20 min, then allowed to settle for 20 Materials and Methods min. The supernatant fluid was collected and layered onto a mixture of 75% Animals and viruses Ficoll-Paque (Pharmacia, Piscataway, NJ) and 25% RPMI 1640 with 10% FBS. After centrifugation at 500 ϫ g for 30 min, the overlying media and SJL/J and BALB/cJ mice were purchased from The Jackson Laboratories interface of tissue debris were removed, and cells in the remaining gradient Downloaded from (Bar Harbor, ME). Four- to ten-week-old female mice were inoculated media were washed with PBS/2% FBS. The pelleted cells were resus- intracerebrally with 1000 plaque-forming units (PFU) of the AR339 strain pended in PBS/2% FBS for flow cytometry or RPMI 1640/10% FBS for of SV in 0.03 ml HBSS containing 1% FBS. Virus was grown and assayed culture. using BHK-21 cells. For isolation of spleen lymphocytes, spleens were removed from per- Histology and immunohistochemistry fused mice, homogenized through a mesh screen, and collected in PBS. The homogenate was layered onto Lympholyte M density separation media At various times after infection, mice were anesthetized with methoxyflu- (Cedarlane, Westbury, NY). After centrifugation, cells were collected from rane and perfused with PBS. Brains, spinal cords, and spleens were re- the interface, washed, and resuspended as described for brain lymphocytes. moved. All or part of each brain and spinal cord was either placed in 4% buffered paraformaldehyde for subsequent embedding in paraffin or rapidly Flow cytometry frozen, together with the spleen, in OCT compound (Tissue-Tek; Miles, Elkhart, IN). Sections of paraffin-embedded tissue were stained with he- For flow cytometry, 3–5 ϫ 105 isolated lymphocytes in 100 ␮l PBS/2% matoxylin and eosin or with luxol fast blue, coded, and examined for in- FBS were incubated for1hat4°Cwith FITC- or phycoerythrin (PE)- flammation and demyelination. Inflammation was scored on coded slides conjugated Abs as listed in Table I at 0.1 ␮g/ml. Conjugated isotype- using a scale of 0–3 in which 0 indicates no detectable inflammation, 1 matched control Abs at the same concentration were used to determine indicates one or two small inflammatory foci per section, 2 indicates mod- background staining. Stained cells were washed, resuspended in 0.5 ml erate inflammatory foci in up to 50% of ϫ10 fields, and 3 indicates mod- PBS/2% FBS, and analyzed using a FACScalibur flow cytometer (Becton 1626 VIRAL ENCEPHALITIS IN SJL MICE

Dickinson, Mountain View, CA). Debris was excluded and intact lympho- diluted with binding buffer and analyzed immediately by flow cytometry as cytes were included in the analysis using forward and side light scattering described. gates, and 5,000–10,000 gated events were collected for each sample. Per- centages of positively staining cells were calculated by establishing quad- Plaque assay rants on log-scale scatterplots of FITC vs PE fluorescence, which separated Brains were removed from perfused mice, frozen on dry ice, and stored at background staining from positive staining. Ϫ80°C. Thawed brains were homogenized in cold PBS to make 10% (w/v) Cytokine mRNA expression homogenates, and dilutions were prepared immediately in DMEM containing 1% FBS and plaqued on BHK-21 cells. Brains from perfused mice were removed, frozen on dry ice, and subse- quently homogenized in PBS. RNA was extracted from 1/4 of whole brain ELISA and plaque neutralization assay for serum Ab homogenate using RNA STAT-60 (Tel-Test “B”, Friendswood, TX). Blood was collected from anesthetized mice, and serum was isolated using cDNA was synthesized from 10% of total RNA as previously described microtainer serum separation tubes (Becton Dickinson). For ELISA, en- (16) using avian myeloblastosis virus reverse transcriptase (Boehringer zyme immunoassay microtitration plates (Costar, Cambridge, MA) were Mannheim). Then, 5% of the cDNA was used for each PCR reaction that ␮ ␮ ␮ coated with polyethylene glycol-precipitated SV at 3 g/ml. Serial dilu- contained 200 M of each dNTP, 1 M of each specific primer, 2.5 U Taq tions of serum ranging from 1:10 to 1:3000 were added to coated plates. polymerase (Boehringer Mannheim), and buffer as supplied by the manu- Specific Ab was detected with horseradish peroxidase-conjugated anti- facturer. PCR was performed using a 9600 thermal cycler (Perkin-Elmer, mouse Ig, IgG1, or IgG2a (see Table I) followed by O-phenylenediamine Cetus, Norwalk, CT). Glyceraldehyde-3-phosphate dehydrogenase dihydrochloride (Sigma). Relative titers were calculated from dilutions in (GAPDH) cDNA detection was used to control for varying amounts of the linear range using a hyperimmune serum as a standard. input RNA. The optimal number of cycles needed for amplification in the For neutralization assays, serial dilutions of serum were incubated with linear range was determined for each cytokine and for GAPDH. Standards a known amount of SV for 30 min at 37°C. PFU/ml were measured on were included with each PCR reaction to ensure that amplification was ␤ BHK-21 cells, and dilutions of serum needed for 50% plaque reduction occurring in the linear range. Primers and probes for GAPDH, IL-1 , IL-2, were determined. IL-4, IL-6, IL-10, TNF␣, and IFN-␥ and primers for the chemokines macrophage inflammatory protein-1␣ (MIP-1␣), MIP-1␤, and IFN-␥-inducible Statistical analysis protein-10 (IP-10) have been described previously (16, 17). Primers and probe for IL-12 were: sense, 5Ј-TCTGCAGAGAAGGTCACACTG; anti- Student’s two-tailed test was used to assess significance of differences. sense, 5Ј-GACTTCGGTAGATGTTTCCTC; probe, 5Ј-CCTGATGAA GAAGCTGGTGCTGTAGTTCTC. For cytokines, PCR products were Results identified and quantitated relative to GAPDH as previously described (16) Histopathology during SV infection except that probes were labeled with digoxigenin and detected with alka- line phosphatase-labeled Ab to digoxigenin and the chemiluminescent sub- Inflammation in the brains and spinal cords of SV-infected SJL strate disodium 3-(4-methoxyspirodecan)phenyl phosphate (Boehringer and BALB/c mice was assessed by routine histology and by im- Mannheim). For chemokines, PCR products were visualized on gels by staining with ethidium bromide. munohistochemistry at various times after infection. Total inflammation was quantitated using the number and size of inflammatory Functional assays using brain lymphocytes lesions in each tissue section (Fig. 1A). Inflammation was greater Ͻ ϭ For determination of IL-10 secretion, lymphocytes isolated from brains or in SJL mice (day 5, p 0.002; day 8, p 0.02) than in BALB/c spleens of infected mice were added to 96-well plates at 5 ϫ 105/ml in mice. Inflammatory lesions were first detected on day 3, and the RPMI 1640 plus 10% FBS with or without 5 ␮g/ml ConA (Sigma) or 100 level of inflammation reached a peak on day 5 in both strains. ng/ml phorbol myristic acid (Sigma) plus 2 ␮g/ml ionomycin (Sigma). However, at the time of maximal inflammation, SJL mice had by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. After incubation for 3 days, supernatant fluids were collected and the extensive infiltration of mononuclear cells typically seen during amount of IL-10 was determined by ELISA using capture and detecting Abs shown in Table I following the protocol provided by the manufacturer. fatal infection with a neurovirulent strain of SV (Fig. 1B), while rIL-10 (PharMingen, San Diego, CA) was used to generate a standard BALB/c mice showed only moderate inflammation typical of non- curve. fatal infection (Fig. 1C) (16). In BALB/c mice, the inflammatory For determination of Ag specificity, mononuclear cells were isolated response was essentially resolved 2–3 wk postinfection (PI), but from spleens of uninfected mice and T cells were depleted by treatment with rabbit anti-Thy1 antiserum (1:20; Cedarlane) at 4°C followed by in- remained at moderate levels in SJL mice until 4–5 wk PI in most cubation at 37°C with Low-Tox-M rabbit complement (1:10; Cedarlane). mice and for several weeks longer in a few mice. By 2 wk PI, Lysed cells were removed using Lympholyte M as described above. Lym- inflammation in BALB/c mice consisted mostly of increased cel- phocytes isolated from brains of infected mice (5 ϫ 105/ml) were com-

https://www.jimmunol.org lularity in the parenchyma. In contrast, SJL mice continued to have ϫ 6 bined with autologous T cell-depleted spleen cells (2.5 10 /ml) and significant perivascular cuffing for the duration of the inflamma- incubated with or without UV-inactivated SV equivalent to 2 ϫ 107 or 8 ϫ 107 PFU/ml or with human MBP at 25 ␮g/ml. After incubation for 3 days, tory response. No demyelination was observed in either strain at supernatant fluids were collected and IFN-␥ production was analyzed by any time point. ELISA (Endogen, Woburn, MA). Immunohistochemistry was used to identify the immune cells present in inflammatory lesions in SJL and BALB/c mice, and the Detection of apoptosis ϩ relative numbers of B cells, macrophages, CD4 T cells, and Downloaded from Brains and spinal cords removed from perfused mice were fixed in 4% CD8ϩ T cells were determined (Fig. 2). Cells expressing the B cell paraformaldehyde and embedded in paraffin. Deparaffinized sections were marker B220 were present at similar levels in SJL mice and subjected to terminal deoxynucleotidyltransferase-mediated UTP nicked end-labeling (TUNEL) as described previously (18). This technique detects BALB/c mice, and the levels remained moderate through 4 wk PI. endonucleolytically cleaved chromosomal DNA characteristically found in Macrophages, identified by reactivity with the Ab F4/80, increased the nuclei of apoptotic cells. Staining was quantitated on coded slides by in number during the first week PI, and levels were similar in SJL scoring the proportion of positive cells in inflammatory loci on a scale of and BALB/c mice. Cells expressing CD8 were also present at sim- 0–3 as for immunohistochemistry and the relative number of positive cells ilar levels in both strains. In contrast, cells expressing CD4 were in the parenchyma on a scale of 0–3 in which 0 indicates no positive cells and 3 indicates the maximum number seen among all sections. These two more abundant in SJL mice than in BALB/c mice. Many inflam- scores were averaged for each section, then adjusted by up to 1 point if matory lesions in brain and spinal cord sections from SJL mice background staining was unusually high or low. consisted of 60–80% CD4ϩ T cells (Fig. 3A), while CD4ϩ T cells For annexin V staining, lymphocytes were isolated from brain and made up only 20–30% of cells in lesions from BALB/c mice (Fig. stained with PE-conjugated anti-CD3e for flow cytometry as described. ϩ Cells were then washed with PBS and incubated with FITC-conjugated 3B). The levels of CD4 T cells in BALB/c mice were similar to annexin V diluted 1:10 in manufacturer-provided binding buffer from an those seen in C57BL/6 mice and BALB/cBy mice (data not ϩ Apoptosis Detection Kit (R & D Systems, Minneapolis, MN). Cells were shown). The representation of CD4 T cells was notably higher in The Journal of Immunology 1627

SJL mice than in BALB/c mice at all time points with detectable inﬂammation and remained at near maximal levels from 7 to 25 days PI (Fig. 2).

Isolation of lymphocytes from brains of infected mice Lymphocytes were isolated from the brains of SV-infected SJL and BALB/c mice by density gradient separation and characterized by flow cytometry. This allowed for a more quantitative characterization of the types of immune cells present in the brains of these mice. Isolated cells were double-stained for expression of CD3 together with CD4, CD8, B220, ␣␤ or ␥␦ TCR, or a pan NK cell marker. The proportion of isolated lymphocytes expressing each marker was determined at time points between 4 and 14 days PI (Fig. 4). Because debris from brain tissue remained in the cell preparations, only cells with light scattering characteristics of lymphocytes were analyzed. At most time points, 70–80% of the cells were identifiable. However, on day 4 PI only 50% of the cells were identifiable due either to the presence of a larger proportion of unidentified cells or to the presence of proportionately more debris within the light scattering gate because fewer lymphocytes were isolated on day 4 PI. The percentage of CD3ϩ cells increased during the first week PI and remained elevated through the second week. At early time points, the percentage of lymphocytes that were CD3ϩ was higher in SJL mice than in BALB/c mice (day 7, p ϭ 0.001) (Fig. 4A). By day 10, the proportion of CD3ϩ cells was similar in both strains. Parallel to this observation, the proportion of lymphocytes that were NK cells was higher in BALB/c mice (day 6, p ϭ 0.002) (Fig. 4A), consistent with earlier studies showing that SJL mice have low, poorly inducible NK activity (19). In contrast to what was by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. https://www.jimmunol.org Downloaded from

FIGURE 1. Inflammation in the CNS of SV-infected SJL and BALB/c mice. A, Quantitation of inflammation in hematoxylin and eosin-stained sections of brains and spinal cords from SJL and BALB/c mice. Inflammation was FIGURE 2. Identification of inflammatory cells by immunohistochem- scored under code on a scale of 0–3: 0, no detectable inflammation; 1, one or istry. Brain and spinal cord sections from infected SJL and BALB/c mice two small inflammatory foci per section; 2, moderate-sized inflammatory foci were stained for the immune cell markers indicated, coded, and represen- in up to 50% of ϫ10 fields; 3, moderate to large inflammatory foci in Ͼ50% tation of each type quantitated. Quantitation was on a scale of 0–3: 0, no of ϫ10 fields. Scores were increased by up to 1 point for abundant parenchy- positive cells; 1, 5–20% of inflammatory cells positive; 2, 20–60% posi- mal cellularity. Between 3 and 14 days PI, each point represents the average tive; and 3, 60–100% positive. Scores were adjusted by up to 1 point when score for three to six mice from two separate experiments, and SEs are shown the numbers of positively staining cells in the parenchyma were substan- (day 5, p Ͻ 0.002; day 8, p ϭ 0.02). After 14 days PI, each point represents tially higher or lower than average. Through day 14 PI, each point repre- the average score for two mice. B and C, Representative hematoxylin and sents the average score for three mice, and, after day 14, each point rep- eosin-stained sections of brains from an SJL (B) and BALB/c (C) mouse at day resents the average score for two mice. For CD4ϩ cells: day 5, p ϭ 0.02; 6 PI (magnification, ϫ400. day 8, p ϭ 0.1, day 14, p ϭ 0.005. 1628 VIRAL ENCEPHALITIS IN SJL MICE

seen with immunohistochemistry, the percentage of B220-expressing cells was very low in both strains at all times examined (data not shown). Staining for IgM resulted in similarly low levels of positive cells, indicating that B cells may not be isolated efﬁciently by the technique used. The composition of the CD3ϩ cells was further studied by staining for CD4 or CD8 (Fig. 4B). The proportion of CD4ϩ and CD8ϩ cells was similar in the two mouse strains through day 7 PI. After day 7, the proportion of CD4ϩ cells increased in SJL mice, while the proportion of CD8ϩ cells increased in BALB/c mice. A late increase in the proportion of CD8ϩ T cells in the brains of BALB/c mice was also seen in earlier ﬂow cytometry, but not in immunocytochemical, studies (15, 20). Staining of CD3ϩ cells for ␣␤ TCR and ␥␦ TCR showed that in both strains of mice the majority of CD3ϩ cells expressed ␣␤ TCR with Ͻ10% expressing ␥␦ TCR (data not shown).

Cytokine mRNA expression by resident and inflammatory cells in the CNS FIGURE 4. Identification by flow cytometry of lymphocytes isolated ϩ from brains of SV-infected SJL and BALB/c mice. A, Isolated lymphocytes To characterize further the CD4 T cells present at increased lev- were labeled with PE-conjugated anti-CD3e and FITC-conjugated anti- els in SJL mice and to determine whether early responses of res- NK. The percentages of events staining positively with each Ab are shown. ident CNS to infection were different, cytokine mRNA expression B, Isolated lymphocytes were labeled with PE-conjugated anti-CD3e and was analyzed by semiquantitative RT-PCR of RNA isolated from either FITC-labeled anti-CD4 or FITC-labeled anti-CD8. The percentages ϩ brains of infected mice (Fig. 5). Early expression of IL-1␤, IL-6, of CD3 events staining positively for CD4 or CD8 are shown. Each point ϩ IL-12, and TNF␣ mRNAs, proinflammatory cytokines likely to be represents the average of results from two to four mice. For CD4 T cells: ϭ ϭ ϩ ϭ important for recruitment of inflammatory cells, was identical. Ex- day 10, p 0.001; day 14, p 0.005. For CD8 T cells: day 10, p ϭ pression of the mRNAs of three chemokines, believed to be im- 0.001; day 14, p 0.005. portant in the recruitment of T cells and the development of EAE,

MIP-1␣, MIP-1␤, and IP-10 (21, 22), was measured, and no differences were seen (data not shown). At days 5 and 8 PI, after initiation of the inﬂammatory responses, expression of IL-10 was higher in SJL mice (day 5, p ϭ 0.002). Expression of IL-4 was consistently higher in BALB/c mice (day 1, p ϭ 0.001; day 3, p ϭ by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. 0.01). Expression of IL-12, IFN-␥, and TNF␣ mRNAs increased to similar levels in both strains. Because the presence of cytokine mRNA may not reﬂect syn- thesis of protein, IL-10 expression was examined further by mea- suring IL-10 production by lymphocytes isolated from brains of infected mice at day 10 PI and stimulated with phorbol ester and ionomycin (Fig. 6). Secretion of IL-10 by lymphocytes from BALB/c brains was undetectable, while secretion by lymphocytes from SJL brains was high and similar to secretion by lymphocytes https://www.jimmunol.org isolated from spleens of both strains of mice.

Ag specificity of brain lymphocytes Earlier studies had suggested that autoimmune T cells recognizing MBP were activated in SJL mice during SV infection (7). To determine whether the increased numbers of CD4ϩ T cells seen in Downloaded from the CNS of SJL mice were virus-specific or recognized the au- toantigen MBP, we tested the ability of CNS lymphocytes to rec- ognize SV proteins and MBP (Fig. 7). Lymphocytes isolated from the brains of mice 10 days after infection were cultured with autologous T cell-depleted spleen cells in the presence or absence of UV-inactivated SV or MBP to stimulate T cells through the MHC class II pathway. Ag recognition was measured by secretion of IFN-␥. Both SJL and BALB/c lymphocytes secreted IFN-␥ in response to SV in a dose-dependent manner, and neither responded to MBP. No IFN-␥ secretion was detected in the absence of Ag. SJL lymphocytes secreted higher levels of IFN-␥ in response to FIGURE 3. CD4ϩ T cells in brains of SJL and BALB/c mice at day 5 SV than BALB/c lymphocytes, approximately proportionate to the ϩ PI. Immunoperoxidase staining with Ab to L3T4(CD4) in SJL (A) and higher levels of CD4 T cells in the SJL brain lymphocyte BALB/c (B) (magnification, ϫ160). population. The Journal of Immunology 1629

FIGURE 5. Expression of cytokines in brains of SV-infected SJL and BALB/c mice. The relative levels of cytokine mRNAs as determined by semiquantitative RT-PCR. Each point represents the average for three mice.

Apoptosis of inflammatory cells portion of CD3ϩ cells from the brains of both mouse strains T cells and macrophages present in CNS inflammation during EAE stained with annexin V. undergo apoptosis during the recovery phase of the disease (23, Virus replication and clearance 24). To determine whether the increased inflammation seen in SV- infected SJL mice was due to reduced death of infiltrating inflam- Differences in virus replication or clearance could change the de- matory cells, apoptotic cells were quantitated by two methods dur- gree or persistence of the inflammatory response. To assess this ing the second week PI. First, brain and spinal cord sections from possibility, infectious virus in brain homogenates was measured by SV-infected SJL and BALB/c mice were stained using the TUNEL plaque assay at various times after infection (Fig. 9A). The levels technique (Fig. 8A). In tissue sections from both strains, a small of infectious virus and the time course of viral clearance were identical between the two strains of mice.

by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. proportion of cells in inflammatory lesions stained positively. In addition, TUNEL-positive cells with lymphocyte morphology Although antiviral Ab has been shown to be sufficient for clear- were scattered throughout the parenchyma. TUNEL staining was ance of SV in SCID mice (1), virus-specific T lymphocytes might scored on coded sections by identifying the proportion of positive compensate for an inadequate Ab response and contribute to virus cells in inflammatory foci together with the relative numbers of clearance. To determine whether the Ab response to SV differs positive cells in the parenchyma and was similar during this period between SJL and BALB/c mice, levels of SV-specific Ab in the of time in the two mouse strains. sera of infected mice were measured by ELISA (Fig. 9B). In both Preapoptotic T cells were identified by annexin V staining of strains, Ab levels increased during the first 1–2 wk PI and re- lymphocytes isolated from the brains of infected mice (Fig. 8B). mained high thereafter. There were no differences in Ab titers at https://www.jimmunol.org Annexin V binds to phosphotidylserine expressed on the cell sur- any time point. The levels of SV-specific IgG1 and IgG2a were face during the early stages of apoptosis, and the level of binding can be measured using flow cytometry (25). A similar small pro- Downloaded from

FIGURE 7. Ag specificity of CD4ϩ T cells isolated from brains of SV- infected mice. Lymphocytes isolated at 10 days PI were incubated with T FIGURE 6. Production of IL-10 by brain lymphocytes. Lymphocytes cell-depleted autologous spleen cells alone or with 108 PFU (1ϫ)or4ϫ isolated from brains at day 10 PI were stimulated in vitro with PMA and 108 PFU (4ϫ) UV-inactivated SV or 25 ␮g MBP. After 3 days, supernatant ionomycin, and IL-10 was measured in supernatant fluids collected at 72 h. fluids were analyzed for the presence of IFN-␥ by ELISA. Each bar rep- Each bar represents cells from one spleen or four brains, each divided into resents cells from two or three brains combined and divided into three three wells (for brain p ϭ 0.001). wells. 1630 VIRAL ENCEPHALITIS IN SJL MICE

FIGURE 9. Virus clearance and Ab response to virus. A, Levels of infectious virus in brain homogenates at various times PI. Each point represents the average of results from three mice. Infectious virus was undetectable at day 8 PI. B, SV-speciﬁc serum Ab titers as detected by ELISA. Titers were determined relative to a standard hyperimmune serum. Each point through day 14 PI represents the average titer for three mice, and points after day 14 PI represent the average titer for two mice. Ab was undetectable in BALB/c mice and barely detectable in SJL mice at day 3 PI.

FIGURE 8. Apoptosis of inflammatory cells in the CNS of SV-infected mice. A, TUNEL staining of inflammatory cells in brain and spinal cord but SJL CNS T cells also produced large amounts of IL-10. These sections at various times PI. Apoptotic cells were identified by TUNEL studies demonstrate that CNS immune responses in SJL mice have staining and quantitated as described in Materials and Methods. Each point inherent differences from those in BALB/c mice and provide in- represents the average score for three mice. B, Annexin V staining of lym- sight into the pathogenesis of the inflammatory diseases to which phocytes isolated from brains of infected mice. Lymphocytes were stained SJL mice are susceptible. with FITC-conjugated annexin V and PE-conjugated anti-CD3e. The per- CD4ϩ T cells were much more prevalent and persistent in the ϩ centages of CD3 cells that stained with annexin are shown. Each point brains and spinal cords of SJL mice in response to SV infection. represents the average of results from two mice. That this is a characteristic of SJL mice is suggested by the fact that prolonged CNS inflammation, without evidence of persistent by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. infection, has also been reported in SJL mice infected with Semliki also determined by ELISA, and in both strains the majority of Ab Forest virus, an alphavirus related to SV (26). The phenotypes of was of the IgG2a isotype (data not shown). Levels of neutralizing the cells responding to Semliki Forest virus infection were not Ab in the serum were determined by plaque neutralization, and characterized, but studies of inflammatory cells in the brains of again, no differences were seen (data not shown). ϩ TMEV-infected SJL mice have also shown fewer CD8 and more activated perivascular CD4ϩ T cells than in other strains of mice Discussion (14, 27). Furthermore, CNS mononuclear inflammatory cells in SJL mice are susceptible to several inflammatory diseases of the SJL mice infected with SV are more likely to show evidence of

https://www.jimmunol.org CNS, while other strains of mice such as BALB/c are relatively proliferation than in BALB/c mice (28). Together, these data sug- resistant (8–10). Susceptibility to induction of these autoimmune gest that alterations in recruitment of CD4ϩ T cells into the CNS CNS diseases may be a consequence of inadequate regulation of and regulation of their numbers after entry may contribute signif- immune responses and inflammatory processes in the CNS. We icantly to the generation of inflammatory disease in the brain and sought to better understand this regulation by studying the re- spinal cord. sponse to SV, a well-characterized viral infection of the CNS that The initial phase of inflammation induced by SV infection fol- results almost invariably in asymptomatic encephalitis. SV infec- lowed a similar time course. However, a considerable proportion Downloaded from tion of SJL mice resulted in more extensive and longer lasting of the lymphocytes found in BALB/c brains early in the inflam- mononuclear cell inflammation than occurred in BALB/c mice, matory response were NK cells, while NK cells were nearly un- even though there were no differences in virus growth, virus clear- detectable in SJL brains. The response to TMEV in SJL mice is ance, or Ab production. However, the characteristics of the cellular also deficient in NK cell activity, and this is postulated to contrib- infiltration and cytokines produced were distinct. In SJL mice, the ute to the defect in early virus clearance (29). Although NK cells initial inflammatory response included very few NK cells and, over and NK activity are prevalent in the CNS of BALB/c mice early time, CD4ϩ T cells continued to accumulate in perivascular re- after SV infection, there is no evidence that they are necessary for, gions, and inflammation was prolonged. In BALB/c mice, NK or contribute to, SV clearance from the CNS (30, 31). NK cells cells were abundant early, CD4ϩ T cells were the predominant T secrete IFN-␥ and TNF␣, cytokines that may play a role in the cell early but were in the minority by 10 days PI, and inflammation initiation and control of T cell responses (32). However, analysis was essentially resolved within 4 wk. BALB/c mice had consis- of cytokine mRNAs in the brain early after infection revealed no tently higher levels of IL-4 mRNA in the CNS than SJL mice. T differences in IFN-␥ or TNF␣. It is possible that the absence of NK cells isolated from the CNS of both strains of mice produced com- cells in SJL mice may be involved in shifting the immune response parable amounts of IFN-␥ in response to Ag stimulation in vitro, to one dominated by CD4ϩ T cells. The Journal of Immunology 1631

We also found no differences in mRNA expression for IL-1␤ or to play an important role in regulation of inflammation in immune IL-6 or the chemokines MIP-1␣, MIP-1␤, or IP-10. The only dif- privileged sites (53). Apoptosis of T cells in inflammatory lesions ference was lower IL-4 mRNA levels in SJL mice. This difference of EAE along with increased production of IL-10 has been asso- has also been observed in TMEV-infected mice (33) and may be ciated with decreasing inflammation and recovery from disease due to the known SJL deficiency in CD4ϩ T cells expressing (23, 24, 52). IL-10 promotes activation-induced death of T cells NK1.1, a subset that produces IL-4 efficiently early after TCR and can prevent or treat autoimmune diseases, including EAE (54– cross-linking (34, 35). Numerous studies have demonstrated the 56). Expression of IL-10 by T cells in SJL brains may, therefore, ability of IL-4 to promote the development of Th2 responses while be a response to the high levels of inflammation and represent an inhibiting the development of Th1 responses (36). IL-4 knockout attempt to down-regulate that process. mice have an increased susceptibility to EAE with more intense We found no evidence that SJL lymphocytes in the CNS exhib- CNS inflammation, suggesting that IL-4 also down-modulates in- ited increased apoptosis during the recovery phase, suggesting that flammatory processes in the CNS (37). Therefore, low levels of the increased IL-10 produced did not effectively control the in- IL-4 in the CNS may play an important role in the predisposition flammatory process. Thus, the enhanced inflammation seen in SJL of SJL mice to increased and prolonged inflammation. mice is most likely to result from increased entry of activated In EAE and TMEV-induced demyelination, disease-mediating T SV-specific IL-10-producing CD4ϩ T cells into the CNS of SJL cells are of the Th1 subtype, secreting inflammatory cytokines mice from the periphery and perhaps local expansion of this pop- such as IL-2 and IFN-␥ (12, 13). A role for antiviral CD4ϩ Th1 ulation after entry. This increase in CNS inflammation is correlated responses has also been suggested by an increased proportion of with low levels of IL-4. TMEV-specific Abs of the IgG2a rather than the IgG1 subclass (38). We measured cytokine expression to determine whether the References CD4ϩ T cells present in the CNS of SJL mice during SV infection were also of the Th1 subtype. A comparison of the overall levels 1. Levine, B., J. M. Hardwick, B. D. Trapp, T. O. Crawford, R. C. Bollinger, and D. E. Griffin. 1991. Antibody-mediated clearance of alphavirus infection from of CNS cytokine mRNA expression during the peak of T cell in- neurons. Science 254:856. filtration showed that expression of IL-2 and IFN-␥ mRNAs, cy- 2. McFarland, H. F., D. E. Griffin, and R. T. Johnson. 1972. Specificity of the inflammatory response in viral encephalitis I: adoptive immunization of immu- tokines typically expressed by Th1 cells, was not elevated in SJL nosuppressed mice infected with Sindbis virus. J. Exp. Med. 136:216. mice. However, expression of Th2 cytokines IL-4 and IL-10 was 3. Wekerle, H., C. Linington, H. Lassman, and R. Meyermann. 1986. Cellular im- different. IL-10 mRNA and protein were increased while IL-4 mune reactivity within the CNS. Trends Neurosci. 9:271. 4. Irani, D. N., and D. E. Griffin. 1996. Regulation of lymphocyte homing into the mRNA was decreased. The study of CNS cytokine mRNAs in brain during viral encephalitis at various states of infection. J. Immunol. 156: response to TMEV infection has also shown lower levels of IL-4 3850. mRNA and higher levels of IL-10 mRNA during TMEV-induced 5. Wong, G. H., P. F. Bartlett, I. Clark-Lewis, F. Battye, and J. W. Schrader. 1984. Inducible expression of H-2 and Ia antigens on brain cells. Nature 310:688. disease in SJL mice compared with BALB/c mice (33). 6. Neumann, H., A. Cavalie, D. E. Jenne, and H. Wekerle. 1995. Induction of MHC The increased expression of IL-10 mRNA in SV-induced en- class 1 genes in neurons. Science 269:549. cephalitis was due to T cells because large amounts of IL-10 were 7. Mokhtarian, F., D. Grob, and D. E. Griffin. 1989. Role of the immune response in Sindbis virus-induced paralysis of SJL/J mice. J. Immunol. 143:633. produced in culture by lymphocytes isolated from brains of in- 8. Tuohy, V. K., R. A. Sobel, and M. B. Lees. 1988. Myelin proteolipid protein- fected SJL, but not BALB/c, mice. Isolated lymphocytes also se- induced experimental allergic encephalomyelitis: variations of disease expression by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. crete IFN-␥ in response to stimulation with SV in culture, and both in different strains of mice. J. Immunol. 140:1868. strains of mice produced antiviral Ab primarily of the IgG2a sub- 9. Bernard, C. C. 1976. Experimental autoimmune encephalomyelitis in mice: ge- ϩ netic control of susceptibility. J. Immunogenet. 3:263. class. Thus, the CD4 T cells present at increased levels in the 10. Melvold, R. W., D. M. Jokinen, R. L. Knobler, and H. L. Lipton. 1987. Variations CNS of SJL mice during SV infection do not appear to be pre- in genetic control of susceptibility to Theiler’s murine encephalomyelitis virus (TMEV)-induced demyelinating disease. I. Differences between susceptible dominantly Th1 or Th2 cells. Although we do not know whether SJL/J and resistant BALB/c strains map near the T cell ␤-chain constant gene on IFN-␥ and IL-10 are produced by the same population of T cells in chromosome 6. J. Immunol. 138:1429. SJL mice, CD4ϩ T cells secreting this combination of cytokines 11. Fritz, R. B., C. H. J. Chou, and D. E. McFarlin. 1983. Relapsing murine experimental allergic encephalomyelitis induced by myelin basic protein. J. Immunol. have been reported by murine clones suppressing autoimmune di- 130:1024. 12. Ando, D. G., J. Clayton, D. Kono, J. L. Urban, and E. E. Sercarz. 1989. Enceph- https://www.jimmunol.org abetes (39) and by human lines and clones responding to chronic infection or primed by IL-12 (40–42). alitogenic T cells in the B10.PL model of experimental allergic encephalomyelitis (EAE) are of the Th-1 lymphokine subtype. Cell. Immunol. 124:132. The effect of IL-10 on inflammatory processes is complex. Al- 13. Gerety, S .J., M. K. Rundell, M. C. Dal Canto, and S. D. Miller. 1992. Class though generally regarded as a Th2 cytokine that down-regulates II-restricted T cell responses in Theiler’s murine encephalomyelitis virus-induced Ag presentation and production of proinflammatory cytokines by demyelinating disease. J. Immunol. 152:919. 14. Pope, J. G., W. J. Karpus, C. VanderLugt, and S. D. Miller. 1996. Flow cyto- macrophages (43), IL-10 also up-regulates macrophage production metric functional analyses of central nervous system-infiltrating cells in SJL/J of nitric oxide and expression of Fc␥R (44, 45). IL-10 inhibits the mice with Theiler’s virus-induced demyelinating disease. J. Immunol. 156:4050. Downloaded from ability of microglial cells to express MHC class II, to produce 15. Irani, D. N., and D. E. Griffin. 1991. Isolation of brain parenchymal lymphocytes for flow cytometric analysis: application to acute viral encephalitis. J. Immunol. IL-12, to present Ag to T cells and to activate astrocytes (46–48). Meth. 139:223. IL-10 down-regulates T cell responses in vivo. When given sys- 16. Wesselingh, S. L., B. Levine, R. J. Fox, S. Choi, and D. E. Griffin. 1994. Intra- cerebral cytokine mRNA expression during fatal and nonfatal alphavirus enceph- temically IL-10 synergizes with IL-4 to inhibit delayed-type hy- alitis suggests a predominant type 2 T cell response. J. Immunol. 152:1289. persensitivity responses (49, 50). IL-10 also stimulates B cell pro- 17. Pearlman, E., J. Lass, D. S. Bardenstein, E. Diaconu, F. E. Hazlett, J. Albright, liferation and induces Ig secretion (51). Because the presence of A. W. Higgins, and J. W. Kazura. 1997. IL-12 exacerbates helminth-mediated Ab is critical for control of SV replication (1), inducing Ab pro- corneal pathology by augmenting inflammatory cell recruitment and chemokine ϩ expression. J. Immunol. 158:827. duction through IL-10 secretion may be a major role for CD4 T 18. Lewis, J., S. L. Wesselingh, D. E. Griffin, and J. M. Hardwick. 1996. Alphavirus- cells in the CNS of infected mice. induced apoptosis in mouse brains correlates with neurovirulence. J. Virol. 70: 1828. During monophasic inflammatory processes, IL-10 tends to be 19. Kaminsky, S. G., I. Nakamura, and G. Cudkowicz. 1993. Selective defect of expressed at the highest levels late in inflammation. An increase in natural killer and killer cell activity against lymphomas in SJL mice: low respon- IL-10 expression in the CNS is observed during and after recovery siveness to interferon inducers. J. Immunol. 130:1980. 20. Moench, T. R., and D. E. Griffin. 1984. Immunocytochemical identification and from EAE (52). Apoptosis of activated cells is believed to be es- quantitation of mononuclear cells in cerebrospinal fluid, meninges, and brain sential for down-regulation of most immune responses and appears during acute viral encephalitis. J. Exp. Med. 159:77. 1632 VIRAL ENCEPHALITIS IN SJL MICE

21. Karpus, W. J., N. W. Lukacs, B. McRae, R. M. Strieter, S. L. Kunkel, and 39. Hans, H. S., H. S. Jun, T. Utsugi, and J. W. Yoon. 1996. A new type of CD4ϩ S. D. Miller. 1995. An important role for the chemokine macrophage inflamma- suppressor T cell completely prevents spontaneous autoimmune diabetes and tory protein-1␣ in the pathogenesis of the T cell-mediated autoimmune disease, recurrent diabetes in syngeneic islet-transplanted NOD mice. J. Autoimmun. experimental autoimmune encephalomyelitis. J. Immunol. 155:5003. 9:331. 22. Ransohoff, R. M., T. A. Hamilton, M. Tani, M. H. Stoler, H. E. Shick, 40. Gerosa, F., C. Paganin, D. Peritt, F. Paiola, M. T. Scupoli, M. Aste-Amezaga, J. A. Major, M. L. Estes, D. M. Thomas, and V. K. Tuohy. 1993. Astrocyte I. Frank, and G. Trinchieri. 1996. Interleukin-12 primes human CD4 and CD8 T expression of mRNA encoding cytokines IP-10 and JE/MCP-1 in experimental cell clones for high production of both interferon-␥ and interleukin-10. J. Exp. autoimmune encephalomyelitis. FASEB J. 7:592. Med. 183:2559. 23. Pender, M. P., K. B. Nguyen, P. A. McCombe, and J. F. R. Kerr. 1991. Apoptosis 41. Windhagen, A., D. E. Anderson, A. Carrizosa, R. Williams, and D. A. Hafler. in the nervous system in experimental allergic encephalomyelitis. J. Neurol. Sci. 1996. IL-12 induces human T cells secreting IL-10 with IFN-␥. J. Immunol. 104:81. 157:1127. 24. Nguyen, K. B., P. A. McCombe, and M. P. Pender. 1994. Macrophage apoptosis 42. Pohl-Koppe, A., K. E. Balashov, A. C. Steere, E. L. Logigian, and D. A. Hafler. in the central nervous system in experimental autoimmune encephalomyelitis. 1998. Identification of a T cell subset capable of both IFN-␥ and IL-10 secretion J. Autoimmun. 7:145. in patients with Borrelia burgdorferi infection. J. Immunol. 160:1804. 25. Fadok, V., D. Voelker, P. A. Campbell, J. J. Cohen, D. L. Bratton, and P. M. Henson. 1992. Exposure of phosphatidylserine on the surface of apoptotic 43. Bogdan, C., J. Paik, Y. Vodovotz, and C. Nathan. 1992. Contrasting mechanisms ␤ lymphocytes triggers specific recognition and removal by macrophages. J. Im- for suppression of macrophage cytokine release by transforming growth factor- munol. 148:2207. and interleukin-10. J. Biol. Chem. 267:23301. 26. Smyth, J. M., B. J. Sheahan, and G. J. Atkins. 1990. Multiplication of virulent and 44. Corradin, S. B., N. Fasel, Y. Buchmuller-Rouiller, A. Ransijn, J. Smithe, and demyelinating Semliki Forest virus in the mouse central nervous system: conse- J. Mauel. 1993. Induction of macrophage nitric oxide production by interferon-␥ quences in BALB/c and SJL mice. J. Gen. Virol. 71:2575. and tumor necrosis factor-␣ is enhanced by interleukin-10. Eur. J. Immunol. 27. Lindsley, M. D., and M. Rodriguez. 1989. Characterization of the inflammatory 23:2045. response in the central nervous system of mice susceptible or resistant to demy- 45. Lehmann, J., D. Seegert, I. Strehlow, C. Schindler, M. Lohmann-Matthes, and elination by Theiler’s virus. J. Immunol. 142:2677. T. Decker. 1994. IL-10-induced factors belonging to the p91 family of proteins 28. Irani, D. N. 1998. The susceptibility of mice to immune-mediated neurologic bind to IFN-␥-responsive promoter elements. J. Immunol. 153:165. disease correlates with the degree to which their lymphocytes resist the effects of 46. Frei, K., H. Lins, C. Schwerdel, and A. Fontana. 1994. Antigen presentation in brain-derived gangliosides. J. Immunol. 161:2746. the central nervous system: the inhibitory effect of IL-10 on MHC class II ex- 29. Paya, C. V., A. K. Patick, P. J. Leibson, and M. Rodriguez. 1989. Role of natural pression and production of cytokines depends on the inducing signals and the killer cells as immune effectors in encephalitis and demyelination induced by type of cell analyzed. J. Immunol. 152:2720. Theiler’s virus. J. Immunol. 143:95. 47. Balasingam, V., and V. W. Yong. 1996. Attenuation of astroglial reactivity by 30. Hirsch, R. L. 1981. Natural killer cells appear to play no role in the recovery of interleukin-10. J. Neurosci. 16:2945. mice from Sindbis virus infection. Immunology 43:81. 31. Griffin, D. E., and J. L. Hess. 1986. Cells with natural killer activity in the 48. Aloisi, F., G. Penna, J. Cerase, B. M. Iglesias, and L. Adorini. 1997. IL-12 cerebrospinal fluid of normal mice and athymic nude mice with acute Sindbis production by central nervous system microglia is inhibited by astrocytes. J. Im- virus encephalitis. J. Immunol. 136:1841. munol. 159:1604. 32. Trinchieri, G. 1995. Natural killer cells wear different hats: effector cells of innate 49. Li, L., J. F. Elliott, and T. R. Mosmann. 1994. IL-10 inhibits cytokine production, resistance and regulatory cells of adaptive immunity and of hematopoiesis. Sem. vascular leakage, and swelling during T helper 1 cell-induced delayed-type hy- Immunol. 7:83. persensitivity. J. Immunol. 153:3967. 33. Sato, S., S. L. Reiner, M. A. Jensen, and R. P. Roos. 1997. Central nervous 50. Pwrie, F., S. Menon, and R. L. Coffman. 1993. Interleukin-4 and interleukin-10 system cytokine mRNA expression following Theiler’s murine encephalomyelitis synergize to inhibit cell-mediated immunity in vivo. Eur. J. Immunol. 23:3043. virus infection. J. Neuroimmunol. 76:213. 51. Moore, K. W., A. O’Garra, R. Malefyt, P. Vieira, and T. R. Mosmann. 1993. 34. Yoshimoto, T., A. Bendelac, J. Hu-Li, and W. E. Paul. 1995. Defective IgE Interleukin-10. Annu. Rev. Immunol. 11:165. ϩ ϩ production by SJL mice is linked to the absence of CD4 , NK1.1 T cells that 52. Kennedy, M. K., D. S. Torrance, K. S. Picha, and K. M. Mohler. 1992. Analysis promptly produce interleukin 4. Proc. Natl. Acad. Sci. USA 92:11931. of cytokine mRNA expression in the central nervous system of mice with ex- 35. Beutner, U., P. Launois, T. Ohteki, J. A. Louis, and H. R. MacDonald. 1997. perimental autoimmune encephalomyelitis reveals that IL-10 mRNA expression Natural killer-like T cells develop in SJL mice despite genetically distinct defects correlates with recovery. J. Immunol. 149:2496. in NK1.1 expression and in inducible interleukin-4 production. Eur. J. Immunol. 53. D’Orazio, T. J., and J. Y. Niederkorn. 1998. A novel role for TGF-␤ and IL-10 27:928.

by guest on September 30, 2021. Copyright 1999 Pageant Media Ltd. in the induction of immune privilege. J. Immunol. 160:2089. 36. Paul, W. E., and R. A. Seder. 1994. Lymphocyte responses and cytokines. Cell 76:241. 54. Georgescu, L., R. K. Vakkalanka, K. B. Elkon, and M. K. Crow. 1997. Interleu- 37. Falcone, M., A. J. Rajan, B. R. Bloom, and C. F. Brosnan. 1998. A critical role kin-10 promotes activation-induced cell death of SLE lymphocytes mediated by for IL-4 in regulating disease severity in experimental allergic encephalomyelitis Fas ligand. J. Clin. Invest. 100:2622. as demonstrated in IL-4-deﬁcient C57BL/6 and BALB/c mice. J. Immunol. 160: 55. Mignon-Godefroy, K., O. Rott, M. Brazillet, and J. Charreire. 1995. Curative and 4822. protective effects of IL-10 in experimental autoimmune thyroiditis (EAT). J. Im- 38. Peterson, J. D., C. Waltenbaugh, and S. D. Miller. 1992. IgG subclass responses munol. 154:6634. to Theiler’s murine encephalomyelitis virus infection and immunization suggest 56. Rott, O., B. Fleischer, and E. Cash. 1994. Interleukin-10 prevents experimental a dominant role for Th1 cells in susceptible mouse strains. Immunology 75:652. allergic encephalomyelitis in rats. Eur. J. Immunol. 24:1434. https://www.jimmunol.org Downloaded from