Induction of experimental autoimmune encephalomyelitis in the ...

International Immunology, Vol. 14, No. 8, pp. 849±856

ã 2002 The Japanese Society for Immunology

Induction of experimental autoimmune encephalomyelitis in the absence of c-Jun N-terminal kinase 2 Kirsty Nicolson1, So®a Freland1, Catherine Weir1, Brett Delahunt2, Richard A. Flavell3 and B. Thomas BaÈckstroÈm1 1Malaghan

Institute of Medical Research and 2Department of Pathology, Wellington School of Medicine, Wellington South, New Zealand 3Section of Immunobiology, Yale University School of Medicine and Howard Hughes Medical Institute, New Haven, CT 06520, USA Keywords: autoimmunity, experimental autoimmune encephalomyelitis, mitogen-activated protein kinases, signal transduction, Th1/Th2 cells Abstract Experimental autoimmune encephalomyelitis (EAE) is a CD4+ T cell-dependent, organ-speci®c autoimmune model commonly used to investigate mechanisms involved in the activation of autoreactive Th1 cells. Mitogen-activated protein kinases such as c-Jun N-terminal kinase (Jnk) 1 and 2 play an important role in the differentiation of naive precursors into Th1 or Th2 effector cells. To investigate the role of Jnk2 on autoimmunity, Jnk2±/± and wild-type mice were immunized with the myelin oligodendrocyte glycoprotein (MOG) 35±55 peptide and the onset of EAE studied. Surprisingly, Jnk2±/± mice were as susceptible to EAE as wild-type mice, regardless of whether low or high antigen doses were used to induce disease. In vitro stimulation of lymph node cells from Jnk2±/± and wild-type mice resulted in comparable proliferation in response to MOG35±55, Mycobacterium tuberculosis and concanavalin A. MOG35±55-speci®c T cells lacking Jnk2 showed a Th1 cytokine pro®le with IFN-g, but no IL-4 or IL-5 production. No differences in the types of in®ltrating cells or myelin destruction in the central nervous system were found between Jnk2±/± and wild-type mice, indicating that lack of Jnk2 does not alter the effector phase of EAE. Our results suggest that, despite involvement in Th1/Th2 differentiation in vitro, Jnk2 is necessary neither for the induction nor effector phase of MOG35±55-induced EAE and nor is it required for antigen-speci®c IFN-g production. Introduction Multiple sclerosis (MS) is one of the major human demyelinating diseases affecting the central nervous system (CNS). MS has been postulated to be mediated by autoreactive T cells, and it is known that both genetic and environmental factors contribute to the pathogenesis of MS (1). The animal model of experimental autoimmune encephalomyelitis (EAE) is often used to develop potential therapies for MS. Although no single animal model mimics a human autoimmune disease completely, EAE resembles the immunopathology found in the human disease (2). EAE can be induced by immunization with self-antigens derived from CNS myelin components, such as myelin basic protein, proteolipid protein, myelin-associated glycoprotein, myelin oligodendrocyte glycoprotein (MOG) or

2¢,3¢-cyclic nucleotide 3¢-phosphodiesterase and also nonmyelin antigens such as the astroglial S-100b protein (3±6). EAE is considered to be a T cell-dependent autoimmune disease model, since the majority of the encephalitogenic T cells are CD4+ lymphocytes recognizing self-peptides in the context of MHC class II molecules (5). Naive CD4+ T cells develop into either Th1 or Th2 effector cells upon activation depending on the cytokines present, and the quality and quantity of both T cell and co-stimulatory receptor signaling (7±10). While encephalitogenic CD4+ T cells involved in the autoimmune EAE model have been shown to be of the Th1 phenotype (11), Th2 cells that respond to non-self-antigens have been shown to modify the cytokine phenotype of the

The ®rst two authors contributed equally to this work Correspondence to: T. BaÈckstroÈm; E-mail: [email protected] Transmitting editor: D. Tarlinton

Received 22 February 2002, accepted 26 May 2002

850 EAE in Jnk2±/± mice autoreactive T cell which can lead to signi®cant protection against EAE (12). Conversely, autoreactive Th2 cells can, under special circumstances, induce rather than protect against EAE (13,14). In humans, MS cannot simply be attributed to diseasecausing Th1 cells. Instead, certain cytokines, such as IFN-g, IL-4, IL-5 and IL-10, play an important role in the pathogenesis of MS and their relative roles may change over the course of the disease (15). Taken together, the Th1/Th2 paradigm in MS and EAE has been challenged (16), and a more comprehensive understanding of the molecular mechanisms involved in T cell differentiation is necessary. In the EAE model, the cytokines IL-12 and IFN-g play an important role in disease development. For example, both IFN-g and IFN-g receptor gene-de®cient mice are more susceptible to EAE than their wild-type counterparts (17). The CNS of affected IFN-g and IFN-g receptor gene-de®cient mice shows a large neutrophilic in®ltration, whereas wild-type mice with similar clinical symptoms show predominantly mononuclear cell in®ltrates (17,18). This indicates that the clinical symptoms of EAE can be due to CNS in®ltration of at least two different cell populations, i.e. neutrophils or mononuclear cells. Further, IL-12 is essential for EAE disease development, since it is involved in maintaining and controlling Th1 lineage commitment and autoimmunity (19). Neutrophilic in®ltration into the CNS might be IL-12 dependent, since blocking antibodies to IL-12 also inhibit EAE in IFN-g genede®cient mice (20). In addition to cytokines and other extracellular stimuli involved in Th1/Th2 differentiation, recent research has unraveled some of the intracellular signaling molecules involved. The role of the mitogen-activated protein kinase (MAPK) signal transduction pathways on the regulation of cytokines have been characterized. In particular, the c-jun N-terminal kinase family (Jnk) 1, 2 and 3 are activated upon stress and in¯ammatory cytokine responses (21,22), and Jnk1 and 2 both have been shown to play an important role in Th1/Th2 differentiation. For example, in vitro activated naive Jnk1±/± T cells preferentially develop into Th2 cells under conditions in which wild-type cells exhibit a Th1 phenotype (23). In this regard, Jnk1-de®cient mice are susceptible to Leishmania major infection due to an enhanced antigen-speci®c Th2 response, which might down-modulate the protective Th1 response found in wild-type animals (24). On the other hand, naive cells lacking Jnk2 show decreased IFN-g production upon TCR stimulation which leads to impaired Th1 but not Th2 differentiation (25). Taken together, the Jnk1 and 2 signaling pathways appear to be involved in Th1/Th2 commitment following antigen-speci®c triggering of the TCR. The focus of this study is to determine whether Jnk2 plays a role in the generation of autoreactive T cells in the EAE/ MOG35±55 model. We show that wild-type and Jnk2±/± mice developed EAE with similar severity and day of onset. Although the level of antigen-speci®c IFN-g production was lower, lymphocytes from Jnk2±/± mice still produced IFN-g and proliferated to a similar extent in response to the MOG35±55 peptide. In addition, the cellular in®ltrate into the CNS was indistinguishable from that of wild-type mice. Together, our data indicate that the Jnk2 signaling pathway is dispensable for the generation of autoreactive T cells responsible for EAE.

We hypothesize that Jnk1 or other MAPKs may compensate for the lack of Jnk2 in the generation of autoimmune Th1 cells involved in the immunopathology of EAE. Methods Mice All mice were bred and maintained in the animal facility of the Wellington School of Medicine. All animal experimental procedures used in this study were approved by the Wellington School of Medicine Animal Ethics Committee and carried out in accordance with the guidelines of the University of Otago, New Zealand. The generation of Jnk2 gene-de®cient mice has been described elsewhere (25). These mice were back-crossed onto C57BL/6J (H-2b) 3 times. Jnk2+/± littermates were then intercrossed to generate Jnk2+/+ (wild-type) and Jnk2±/± mice. EAE induction and clinical evaluation Female wild-type and Jnk2±/± mice (8±12 weeks old) were immunized s.c. over the left and right ¯anks with indicated amounts of the MOG35±55 peptide emulsi®ed in complete Freund's adjuvant (CFA; Difco, Detroit, MI), containing 400 mg of Mycobacterium tuberculosis. Pertussis toxin (List Biological, Campbell, CA), 200 ng, was given i.p. on days 0 and 2 after immunization. Clinical signs of EAE were assessed daily using a standard scale ranging from 0±5 as follows: 0 = no clinical signs, 1 = loss of tail tonicity, 2 = ¯accid tail, 3 = affected hind leg or legs, 4 = both hind legs paralyzed, and 5 = hind body and hind legs paralyzed, moribund state, at which point the mice were culled in accordance with the ethical approval by the local animal committee. The MOG35±55 peptide used corresponded to amino acid residues 35±55 of the mouse sequence (MEVGWYRSPFSRVVHLYRNGK) and was synthesized by Mimotopes (Clayton, Australia), with a purity of >97%. In vitro proliferation of lymph node cells Draining lymph nodes from MOG35±55 peptide-immunized animals, as described above, were harvested 10 days postimmunization and single-cell suspensions prepared. Lymphocytes, 4 3 105 cells/well in 200 ml, were incubated in 96-well ¯at-bottom plates with the indicated amounts of antigen in IMDM medium supplemented with 2 mM glutamine, 100 U/ml penicillin G, 100 mg/ml streptomycin sulfate and 5% FCS (all from Gibco/BRL, Auckland, New Zealand). Cultures were incubated for 72 h, 0.5 mCi of [3H]thymidine added to each well and plates harvested after an additional 16 h of culture. For cultures receiving concanavalin A (Con A; Sigma, St Louis, MO), cells were incubated for 40 h, 0.5 mCi of [3H]thymidine added to each well and harvested after an additional 8 h of culture. In vitro determination of cytokine levels by ELISA Draining lymph node cells were harvested and prepared as described above. Cultures were stimulated with the MOG35± 55 peptide, M. tuberculosis or Con A at various concentrations. Supernatants were collected after 72 h of culture for MOG35±55 and M. tuberculosis, and 40 h for Con A. Samples

EAE in Jnk2±/± mice were stored at ±20°C until tested. Cytokine concentrations were determined by sandwich ELISA. Brie¯y, 96-well immunoplates were coated overnight at 4°C with 1±2 mg/ml of the capture antibody. The plates were washed and incubated with 1% BSA in PBS. Culture supernatants and standards were added to the plates, and incubated at room temperature for 2 h. The plates were then washed and incubated with 0.5±1 mg/ml of biotinylated anti-cytokine antibodies at room temperature for 2 h. After washing, plates were incubated with streptavidin-conjugated horseradish peroxidase (Amersham Pharmacia, Uppsala, Sweden) and the reaction developed using the substrate TMB (Sigma). The following antibody pairs were used; IFN-g, AN18 and XMG-D6±biotin; tumor necrosis factor (TNF)-a, TN3-19.12 (gift from Dr R. Schreiber) and polyclonal rabbit anti-mouse/rat TNF-a±biotin; IL-4, 11B11 and BVD6-24G2±biotin. The antibodies were either puri®ed from cultured supernatants or purchased from PharMingen (San Diego, CA). Standard curves for each assay were generated using recombinant mouse cytokines. Histology Wild-type and Jnk2±/± mice were immunized with 50 mg of MOG35±55 peptide, as described above. Brains and spinal cords were removed 20 days later and ®xed with 10% buffered formalin (Sigma). Paraf®n-embedded sections (6 mm thick) were stained with either H & E or Luxol fast blue to assess cellular in®ltration and demyelination respectively. Results Jnk2 is dispensable for the development of EAE Previous in vitro studies have shown that differentiation of precursor CD4+ T cells into Th1, but not Th2 cells, is impaired in cells lacking Jnk2 (25). In this study, we used mice lacking the Jnk2 gene to address the question whether the clinical outcome of EAE is dependent on the MAPK signaling pathway through Jnk2. We immunized wild-type and Jnk2±/± mice with 50 mg of the MOG35±55 peptide and observed mice daily for clinical signs of EAE. Surprisingly, Jnk2±/± mice were susceptible to MOG-induced EAE and developed disease similar to that seen in wild-type mice (Fig. 1). The day of onset and disease severity were not signi®cantly different comparing wild-type with Jnk2±/± mice (Table 1). Wild-type mice showed a slightly higher grade of disease severity towards the end of the experiment, in comparison with Jnk2±/± mice (Fig. 1). However this difference was not statistically signi®cant (P = 0.12, Table 1). It has been shown that in vitro cultures of cells lacking Jnk2 still produce IFN-g upon activation, although less than cell cultures from wild-type mice (25). Thus, since IFN-g production is dose dependent, it is possible that a role for Jnk2 was not detected because of the high immunizing dose used. Therefore, we investigated the role of Jnk2 in EAE using lower immunizing doses. Wild-type and Jnk2±/± mice were immunized with 50, 10 or 3 mg MOG35±55 and observed over a period of 35 days. Although the severity of disease decreased with a decrease in antigen dose, no differences in clinical scores or day of onset between wild-type and Jnk2±/± mice were observed at any of the doses administered (Fig. 2).

851

Fig. 1. Clinical course of MOG35±55-induced EAE in wild-type and Jnk2±/± mice. Mice were immunized with 50 mg MOG35±55 peptide in CFA on day 0 and scored for clinical signs of EAE as described in Methods. Results are plotted as the mean clinical score 6 SEM for all of the animals in each group (wild-type = 16 and Jnk2±/± = 15) versus day post-immunization. The graph shows pooled data from four different experiments.

Table 1. Effect of Jnk2 on EAE induced with MOG35±55a

Incidence Day of onset Maximum score Disease at day 35

Wild-type

Jnk2±/±

Pb

16/16 13.3 6 0.61 3.4 6 0.31 3.1 6 0.37

15/15 13.8 6 0.76 2.9 6 0.32 2.6 6 0.36

NS (P = 0.55) NS (P = 0.16) NS (P = 0.12)

aSummary of data collected from the experiments shown in Fig. 1. Wild-type and Jnk2±/± mice were immunized with 50 mg of MOG35± 55 as described in Methods. Data are displayed as mean 6 SEM of the total mice in each group. bThe signi®cance of difference was calculated using the Mann± Whitney U-test.

In addition, immunizations with doses of up to 200 mg of MOG35±55 generated EAE with similar clinical scores as found with 50 mg of peptide in both wild-type and Jnk2±/± mice (data not shown). Taken together, these results indicate that Jnk2 is not essential for the development of EAE. Jnk2±/± MOG35±55-speci®c lymphocytes show no defect in in vitro proliferation Although no differences in clinical score of wild-type and Jnk2±/± mice were revealed, we further investigated whether disease-inducing T cells responded normally to re-stimulation in vitro. To study in vitro proliferation of antigen-speci®c Jnk2±/± cells, mice were immunized with 50 mg of MOG35±55 peptide and 10 days later [3H]thymidine uptake by lymphocytes was assayed. The results indicated that cells from wild-type and Jnk2 ±/± mice showed similar responses to the MOG35±55 peptide (Fig. 3A). Cell cultures stimulated with either M. tuberculosis (immunization control for components in CFA) or Con A (a T cell mitogen) showed no difference in the degree of proliferation (Fig. 3B and C). Comparing cell cultures from wild-type and Jnk2±/± mice immunized with 10 or 3 mg MOG35± 55 peptide showed no difference in responses to MOG35±55,

852 EAE in Jnk2±/± mice showed that there was no signi®cant difference in antigenspeci®c IFN-g production between cell cultures from Jnk2±/± or littermate control mice (P = 0.08 at 3 mg/ml of MOG35±55 peptide, using the Mann±Whitney U-test). Further, Jnk2 wildtype and Jnk2±/± cells produced antigen-speci®c TNF-a, but undetectable amounts of IL-4 (Fig. 4), indicating that MOG35± 55 immunization induced Th1 and not Th2 cells, regardless of whether the cells possessed a functional Jnk2 signaling pathway or not. Therefore, Jnk2 is not required for generation of MOG35±55-speci®c Th1 cells. Comparable CNS in¯ammation and demyelination in wild-type and Jnk2±/± mice Since mice de®cient of IFN-g have been shown to develop EAE with increased polymorphonuclear instead of mononuclear cellular in®ltrates, we determined the subsets of in®ltrating cells in the CNS and the level of demyelination. Wild-type and Jnk2±/± mice were injected with 50 mg of MOG35±55 in CFA, and brains and spinal cords removed 20 days postimmunization. Paraf®n sections stained with H & E revealed no phenotypical or cellular difference in in®ltrating cells in the CNS tissue of Jnk2 wild-type and Jnk2±/± mice (Fig. 5A). Several in¯ammatory foci were found especially in the white matter of the spinal cord in both wild-type and Jnk2±/± mice, which consisted mainly of mononuclear cells. A similar degree of demyelination was also found in the CNS of MOG35±55immunized mice in the absence of Jnk2, using Luxol fast bluestained spinal cord sections (Fig. 5B). Neither cellular in®ltrates nor demyelination were found in CFA-immunized control mice (Fig. 5A and B). Further, FACS analysis of brain and spinal cord cell suspensions did not reveal any difference in in®ltrating T cells, B cells, Mac-1+ or Gr-1+ cells between wild-type and Jnk2±/± mice (data not shown). These results indicate that clinical EAE can be induced in the absence of the Jnk2 signaling pathway, with a similar degree of demyelination and no difference in the levels or subtypes of mononuclear cell in®ltrates in the CNS. Fig. 2. Comparison of EAE development using different doses of MOG35±55 in wild-type and Jnk2±/± mice. Mice were immunized with either 50 (A), 10 (B) or 3 (C) mg MOG35±55 in CFA on day 0 and scored for clinical signs of EAE as described in Methods. Results are plotted as the mean clinical score for all of the animals in each group (n = 3) versus day post-immunization. The graph shows one of two independent experiments with similar results.

M. tuberculosis and Con A (data not shown). Therefore, these experiments revealed that lack of Jnk2 did not compromise the ability of antigen-speci®c cells to proliferate in vitro. Jnk2±/± MOG35±55-speci®c lymphocytes display a Th1 phenotype Since Jnk2-de®cient T cells have been shown to display a defect in the generation of Th1 cells, we investigated the in vitro cytokine pro®le after MOG35±55 challenge. Primed lymph node cells were stimulated as described in Fig. 3 and supernatants tested for cytokine content. Although cell cultures from Jnk2±/± mice showed lower IFN-g production in response to MOG35±55, M. tuberculosis or Con A compared to littermate control mice, Jnk2±/± lymphocytes produced a considerable amount of IFN-g (Fig. 4). Statistical analyses

Discussion The involvement of Jnk2 in the generation of Th1 cells in vivo and in autoimmunity has not been fully elucidated. We have studied its role in the Th1-dependent MOG35±55 EAE model, using Jnk2±/± and littermate control mice. In contrast to what would have been predicted from previous data regarding Jnk2 function, we found that mice lacking the Jnk2 gene develop a similar degree of EAE compared to control mice following immunization with either 50 mg or as low as 3 mg of peptide (Figs 1 and 2). Immunizations with doses of up to 200 mg of MOG35±55 also generated EAE with similar clinical scores in Jnk2±/±, littermate control and pure C57BL/6 mice housed in the same animal facility (data not shown), indicating that neither antigen dose nor the mixed genetic background of the mice used could explain the lack of a role for Jnk2 on EAE. These results were unexpected, since Yang et al. have shown that Th1 development in vitro is impaired in lymphocytes isolated from Jnk2-de®cient mice (25). Jnk2±/± CD4+ T cells show a reduced IFN-g production which is rescued by exogenous IFN-g. This indicates that decreased autocrine IFN-g production leads to a reduction in the degree of Th1

EAE in Jnk2±/± mice

853

Fig. 3. Proliferation of antigen-speci®c cells from MOG35±55-primed wild-type and Jnk2±/± mice. Mice were immunized with 50 mg MOG35±55, and draining lymph node cells harvested and prepared 10 days later as described in Methods. Cultures were stimulated with indicated concentrations of the MOG35±55 peptide (A), M. tuberculosis (B) or Con A (C) and T cell proliferation determined. The graph shows one of ®ve independent experiments with similar results and the results are plotted as mean c.p.m. 6 SEM.

Fig. 4. Cytokine production by antigen-speci®c cells from MOG35±55-primed wild-type and Jnk2±/± mice. Mice were immunized with 50 mg MOG35±55 and draining lymph node cells were harvested 10 days later. Cultures were stimulated with the MOG35±55 peptide, M. tuberculosis or Con A, and the levels of IFN-g, TNF-a and IL-4 production were determined by sandwich ELISA. The results are plotted as mean concentration 6 SEM from three separate experiments.

differentiation and that defective Th1 development would be predicted to play a role in the immunopathology of the EAE model. However, since some IFN-g is still produced by in vitro activated CD4+ T cells from Jnk2±/± mice (25), this reduced

level could be suf®cient to generate encephalitogenic Th1 cells in vivo. This could explain the fact that Jnk2±/± mice develop EAE with similar severity as wild-type mice (Figs 1 and 2). Interestingly, primary in vitro cultures from Jnk2±/± MOG35±

854 EAE in Jnk2±/± mice

Fig. 5. Immunopathology of MOG35±55-primed wild-type and Jnk2±/± mice. Mice were immunized with 50 mg MOG35±55 peptide or PBS in CFA as in Fig. 1. Spinal cords were removed 20 days later, ®xed in 10% formalin and embedded in paraf®n. Six-micrometer sections were stained with H & E (A) or Luxol fast blue (B). MOG35±55-immunized wild-type and Jnk2±/± mice with a disease score of 3±4 were used and PBS control immunized mice showed no clinical signs. Representative sections from one of two experiments with similar results are shown. Original magni®cations: 3200.

55-immunized mice also show antigen-speci®c IFN-g production (Fig. 4), although at reduced levels. However, this reduction was not statistically signi®cant compared to the wildtype cells (P = 0.08). Since we used whole lymph node cell preparations in contrast to highly puri®ed T cells, it is dif®cult to assess whether the decreased MOG35±55-speci®c IFN-g production in Jnk2±/± lymphocytes is comparable with those found by Yang et al. (25). The expected phenotype of Jnk2±/± mice is speculative, since the role of IFN-g in EAE and MS is unclear. For example, IFN-g has been shown to be both an inhibitory and activating molecule. In clinical trials in which MS patients were treated with IFN-g, exacerbations of clinical symptoms were observed (26), indicating that IFN-g production directly correlates with disease severity. Conversely, IFN-g and IFN-g receptor genede®cient mice are more susceptible to EAE than their wild-type counterparts (17). In this regard, Tran et al. found that the CNS of IFN-g-de®cient mice with EAE had a large neutrophilic in®ltration which led to severe clinical symptoms (18). Wildtype mice also had severe clinical symptoms, but much reduced neutrophil in®ltrates. We found that the degree of mononuclear cell in®ltrate, myelin destruction and phenotype of in®ltrating cells in the CNS of MOG35±55-immunized Jnk2±/± is similar to that of wild-type mice (Fig. 5 and data not shown). Taking into consideration that IFN-g-de®cient mice develop EAE with neutrophilic in®ltration and that Jnk2±/± Th1 cells show impaired IFN-g production and Th1 development in vitro (25), we expected to observe an immunopathology similar to IFN-gde®cient mice or reduced disease due to poor development of Th1 cells in our EAE model. Nonetheless, we found no difference in the degree of disease or CNS in®ltrates after

immunization with the MOG35±55 peptide, indicating that IFN-g or the generation of Th1 cells in vivo is not limited in Jnk2±/± mice. Further, the proin¯ammatory cytokine TNF-a has been shown to in¯uence the clinical outcome of EAE (27±31). However, MOG35±55-speci®c cells from wild-type and Jnk2±/± mice show similar levels of TNF-a upon activation (Fig. 4), and no difference in disease (Figs 1 and 2). These results indicate that antigen-speci®c TNF-a production is independent of the Jnk2 gene and therefore no differences in TNF-a-mediated EAE symptoms are found. The role of Jnk2 in vivo has not been studied extensively using Th1-dependent models. However, Jnk1±/± mice have been shown to be susceptible to L. major infection, a disease in which differentiated CD4+ Th1 cells provide protection (24). The lack of a protective immune response in these mice is characterized by an enhanced Th2-speci®c response (IL-5 and IL-13 production), which possibly acts to down-modulate the protective Th1 response normally found in wild-type mice. This indicates that Jnk1 is necessary to generate a protective Th1 response to L. major by inhibiting an otherwise dominating non-protective Th2 immune response. In the case of autoimmune responses to the MOG35±55 peptide in our EAE model, lack of the Jnk2 gene is not associated with the appearance of a Th2 response. In fact, peripheral Jnk2±/± T cells produce the Th1-type cytokine IFN-g, and the Th2-type cytokines IL-4 and IL-5 are not detected in our system (Fig. 4 and data not shown). Together, data from these experiments suggest that Jnk1 and 2 are involved in specialized and distinct functions in vivo. The question arises, what mechanisms could explain the observation that EAE is not altered by the lack of Jnk2

EAE in Jnk2±/± mice activation? Differentiated Th1 cells lacking Jnk2 can produce a small amount of IFN-g (25). Therefore, low levels of IFN-g may still be suf®cient to drive the differentiation of MOG35±55speci®c precursor T cells into competent disease-causing Th1 cells in vivo, resulting in no difference in EAE between Jnk2 wild-type or gene-de®cient mice. Furthermore, recent results have unexpectedly shown that CD8+ T cells play an important role in the pathogenesis of both EAE and MS (32±34). In the MOG35±55 EAE model, antigen-speci®c CD8+ T cells cause a severe permanent disease with a signi®cant percentage of neutrophils (>30%) in the CNS of diseased mice (34). The EAE generated by CD8+ T cells in normal mice is reminiscent of the immunopathology generated in IFN-g and IFN-g receptor knockout mice (17,18,35), indicating the possibility that EAE caused by neutrophils could be IFN-g independent. Interestingly, CD8+ T cells lacking the Jnk2 gene have been shown to hyperproliferate due to increased IL-2 production (36). Because CD8+ T cells can also cause EAE (33,34), a possibility exists that increased CD8+ T cell activity could have compensated for a defective Th1 development, and no obvious difference in the severity of EAE between Jnk2 wildtype and gene-de®cient mice would have been seen. Although an interesting possibility, no difference in the CD4/ CD8 ratio or neutrophilic in®ltration in the CNS was found in our experiments (Fig. 5 and data not shown). Taken together, our results indicate that autoimmune Th1cells can be generated and cause EAE in the absence of the MAPK gene Jnk2. Since Jnk1 and 2 become activated through a similar signaling pathway, Jnk1 could compensate for the loss of function of Jnk2 (manuscript in preparation) and, therefore, no differences in EAE would be found. However, it is important to investigate the consequence of differential activation of Jnk1 and 2, since potential future treatment could evolve from the identi®cation of a speci®c MAPK signaling pathway involved in MS.

Acknowledgements We thank members of the Malaghan Institute of Medical Research for critical reading of the manuscript, and Ann Thornton and Joan Nicol for technical laboratory assistance. We thank Dr Mercedes Rincon for supplying the Jnk2±/± mice and for sharing unpublished data with us. This work was supported by the Swedish Foundation for International Co-operation in Research and Higher Education (S. F.), New Zealand Lottery Grants Board and the Wellcome Trust, UK (K. N.). Work by R. A. F. was supported by grants from the NIH. R. A. F. is an investigator of the Howard Hughes Medical Institute. B. T. B is the recipient of the Wellcome Trust Senior Research Fellowship in Medical Science, New Zealand.

Abbreviations CFA CNS Con A EAE Jnk MAPK MOG MS TNF

complete Freund's adjuvant central nervous system concanavalin A experimental autoimmune encephalomyelitis c-Jun N-terminal kinase mitogen-activated protein kinase myelin oligodendrocyte glycoprotein multiple sclerosis tumor necrosis factor

855

References 1 Martin, R., McFarland, H. F. and McFarlin, D. E. 1992. Immunological aspects of demyelinating diseases. Annu. Rev. Immunol. 10:153. 2 Taneja, V. and David, C. S. 2001. Lessons from animal models for human autoimmune diseases. Nat. Immunol. 2:781. 3 Mendel, I., Kerlero de Rosbo, N. and Ben-Nun, A. 1995. A myelin oligodendrocyte glycoprotein peptide induces typical chronic experimental autoimmune encephalomyelitis in H-2b mice: ®ne speci®city and T cell receptor V beta expression of encephalitogenic T cells. Eur. J. Immunol. 25:1951. 4 Hartung, H. P. and Rieckmann, P. 1997. Pathogenesis of immunemediated demyelination in the CNS. J. Neural. Transm. Suppl. 50:173. 5 Wekerle, H., Kojima, K., Lannes-Vieira, J., Lassmann, H. and Linington, C. 1994. Animal models. Ann. Neurol. 36 (Suppl.):S47. 6 Weir, C. R., Nicolson, K. and BaÈckstroÈm, B. T. 2002. Experimental autoimmune encephalomyelitis induction in naive mice by dendritic cells presenting a self-peptide. Immunol. Cell Biol. 80:14. 7 Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A. and Coffman, R. L. 1986. Two types of murine helper T cell clone. I. De®nition according to pro®les of lymphokine activities and secreted proteins. J. Immunol. 136:2348. 8 Hosken, N. A., Shibuya, K., Heath, A. W., Murphy, K. M. and O'Garra, A. 1995. The effect of antigen dose on CD4+ T helper cell phenotype development in a T cell receptor-alpha betatransgenic model. J. Exp. Med. 182:1579. 9 Constant, S. L. and Bottomly, K. 1997. Induction of Th1 and Th2 CD4+ T cell responses: the alternative approaches. Annu. Rev. Immunol. 15:297. 10 Dong, C. and Flavell, R. A. 2000. Cell fate decision: T-helper 1 and 2 subsets in immune responses. Arthritis Res. 2:179. 11 Cua, D. J., Hinton, D. R. and Stohlman, S. A. 1995. Self-antigeninduced Th2 responses in experimental allergic encephalomyelitis (EAE)-resistant mice. Th2-mediated suppression of autoimmune disease. J. Immunol. 155:4052. 12 Falcone, M. and Bloom, B. R. 1997. A T helper cell 2 (Th2) immune response against non-self antigens modi®es the cytokine pro®le of autoimmune T cells and protects against experimental allergic encephalomyelitis. J. Exp. Med. 185:901. 13 Lafaille, J. J., Keere, F. V., Hsu, A. L., Baron, J. L., Haas, W., Raine, C. S. and Tonegawa, S. 1997. Myelin basic protein-speci®c T helper 2 (Th2) cells cause experimental autoimmune encephalomyelitis in immunode®cient hosts rather than protect them from the disease. J. Exp. Med. 186:307. 14 Wensky, A., Garibaldi Marcondes, M. C. and Lafaille, J. J. 2001. The role of IFN-gamma in the production of Th2 subpopulations: implications for variable Th2-mediated pathologies in autoimmunity. J. Immunol. 167:3074. 15 Pelfrey, C. M., Rudick, R. A., Cotleur, A. C., Lee, J. C., TaryLehmann, M. and Lehmann, P. V. 2000. Quanti®cation of selfrecognition in multiple sclerosis by single-cell analysis of cytokine production. J. Immunol. 165:1641. 16 Laman, J. D., Thompson, E. J. and Kappos, L. 1998. Balancing the Th1/Th2 concept in multiple sclerosis. Immunol. Today 19:489. 17 Willenborg, D. O., Fordham, S., Bernard, C. C., Cowden, W. B. and Ramshaw, I. A. 1996. IFN-gamma plays a critical downregulatory role in the induction and effector phase of myelin oligodendrocyte glycoprotein-induced autoimmune encephalomyelitis. J. Immunol. 157:3223. 18 Tran, E. H., Prince, E. N. and Owens, T. 2000. IFN-gamma shapes immune invasion of the central nervous system via regulation of chemokines. J. Immunol. 164:2759. 19 Chang, J. T., Shevach, E. M. and Segal, B. M. 1999. Regulation of interleukin (IL)-12 receptor beta2 subunit expression by endogenous IL-12: a critical step in the differentiation of pathogenic autoreactive T cells. J. Exp. Med. 189:969. 20 Segal, B. M., Dwyer, B. K. and Shevach, E. M. 1998. An interleukin (IL)-10/IL-12 immunoregulatory circuit controls susceptibility to autoimmune disease. J. Exp. Med. 187:537.

856 EAE in Jnk2±/± mice 21 Davis, R. J. 2000. Signal transduction by the JNK group of MAPK. Cell 103:239. 22 Rincon, M., Conze, D., Weiss, L., Diehl, N. L., Fortner, K. A., Yang, D., Flavell, R. A., Enslen, H., Whitmarsh, A. and Davis, R. J. 2000. Do T cells care about the mitogen-activated protein kinase signalling pathways? Immunol. Cell Biol. 78:166. 23 Dong, C., Yang, D. D., Wysk, M., Whitmarsh, A. J., Davis, R. J. and Flavell, R. A. 1998. Defective T cell differentiation in the absence of Jnk1. Science 282:2092. 24 Constant, S. L., Dong, C., Yang, D. D., Wysk, M., Davis, R. J. and Flavell, R. A. 2000. JNK1 is required for T cell-mediated immunity against Leishmania major infection. J. Immunol. 165:2671. 25 Yang, D. D., Conze, D., Whitmarsh, A. J., Barrett, T., Davis, R. J., Rincon, M. and Flavell, R. A. 1998. Differentiation of CD4+ T cells to Th1 cells requires MAP kinase JNK2. Immunity 9:575. 26 Panitch, H. S., Hirsch, R. L., Schindler, J. and Johnson, K. P. 1987. Treatment of multiple sclerosis with gamma interferon: exacerbations associated with activation of the immune system. Neurology 37:1097. 27 Dal Canto, R. A., Shaw, M. K., Nolan, G. P., Steinman, L. and Fathman, C. G. 1999. Local delivery of TNF by retrovirustransduced T lymphocytes exacerbates experimental autoimmune encephalomyelitis. Clin. Immunol. 90:10. 28 Kassiotis, G., Pasparakis, M., Kollias, G. and Probert, L. 1999. TNF accelerates the onset but does not alter the incidence and severity of myelin basic protein-induced experimental autoimmune encephalomyelitis. Eur. J. Immunol. 29:774. 29 Korner, H., Riminton, D. S., Strickland, D. H., Lemckert, F. A., Pollard, J. D. and Sedgwick, J. D. 1997. Critical points of tumor necrosis factor action in central nervous system autoimmune in¯ammation de®ned by gene targeting. J. Exp. Med. 186:1585. 30 Liu, J., Marino, M. W., Wong, G., Grail, D., Dunn, A., Bettadapura,

31

32

33 34

35

36

J., Slavin, A. J., Old, L. and Bernard, C. C. 1998. TNF is a potent anti-in¯ammatory cytokine in autoimmune-mediated demyelination. Nat. Med. 4:78. Riminton, S. D., Korner, H., Strickland, D. H., Lemckert, F. A., Pollard, J. D. and Sedgwick, J. D. 1998. Challenging cytokine redundancy: in¯ammatory cell movement and clinical course of experimental autoimmune encephalomyelitis are normal in lymphotoxin-de®cient, but not tumor necrosis factor-de®cient, mice. J. Exp. Med. 187:1517. Babbe, H., Roers, A., Waisman, A., Lassmann, H., Goebels, N., Hohlfeld, R., Friese, M., Schroder, R., Deckert, M., Schmidt, S., Ravid, R. and Rajewsky, K. 2000. Clonal expansions of CD8+ T cells dominate the T cell in®ltrate in active multiple sclerosis lesions as shown by micromanipulation and single cell polymerase chain reaction. J. Exp. Med. 192:393. Huseby, E. S., Liggitt, D., Brabb, T., Schnabel, B., Ohlen, C. and Goverman, J. 2001. A pathogenic role for myelin-speci®c CD8+ T cells in a model for multiple sclerosis. J. Exp. Med. 194:669. Sun, D., Whitaker, J. N., Huang, Z., Liu, D., Coleclough, C., Wekerle, H. and Raine, C. S. 2001. Myelin antigen-speci®c CD8+ T cells are encephalitogenic and produce severe disease in C57BL/ 6 mice. J. Immunol. 166:7579. Willenborg, D. O., Fordham, S. A., Staykova, M. A., Ramshaw, I. A. and Cowden, W. B. 1999. IFN-gamma is critical to the control of murine autoimmune encephalomyelitis and regulates both in the periphery and in the target tissue: a possible role for nitric oxide. J. Immunol. 163:5278. Conze, D., Krahl, T., Kennedy, N., Weiss, L., Lumsden, J., Hess, P., Flavell, R. A., Le Gros, G., Davis, R. J. and RincoÂn, M. 2002. cJun NH2-terminal kinase (JNK)1 and JNK2 have distinct roles in CD8+ T cell activation. J. Exp. Med. 195:811.