Fas system up-regulation in experimental autoimmune encephalomyelitis

Journal of the Neurological Sciences 170 (1999) 96–104 www.elsevier.com / locate / jns

Fas system up-regulation in experimental autoimmune encephalomyelitis Jean-Christophe Ouallet*, Nicole Baumann, Yannick Marie, Henri Villarroya ´ ˆ ˆ ` , Paris, France Institut National de la Sante´ et de la Recherche Medicale ( INSERM), Unit 495, Hopital de la Salpetriere Received 31 March 1999; received in revised form 29 June 1999; accepted 29 June 1999

Abstract Experimental autoimmune encephalomyelitis (EAE) is a T-cell-mediated disorder characterized by infiltration of the central nervous system (CNS) by mononuclear cells and macrophages, and serves as a model for multiple sclerosis. In acute monophasic and relapsing remitting forms of EAE, the CNS inflammatory infiltrates are cleared within a few days and, simultaneously, animals recover from their clinical disability. The mechanisms for rapid disappearance of the inflammatory cells are not fully understood. Fas and Fas-ligand (Fas-L) molecules are thought to play an important role in the deletion of autoimmune reactive T cells through apoptosis. However, recent observations in transgenic lpr and gld mice show that mutations inactivating Fas and Fas-L respectively ameliorate signs of EAE despite persistence of immune cell infiltrates into the CNS. In the current study, the expression of Fas and Fas-L was investigated by immunochemistry and in situ hybridization during the course of EAE in DA rats that were actively immunized with syngenic spinal cord homogenate. CNS apoptotic cells were simultaneously examined using terminal transferase dUTP nick end-labeling techniques. During the acute phase of the disease, a significant proportion of CNS CD41 cells (80%) and macrophages (50%) expressed Fas and Fas-L (80 and 60%, respectively). Simultaneously, about 20% of CD41 cells and 30% of macrophages were found to be apoptotic. Some astrocytes and neurons also expressed Fas and Fas-L, although they did not appear to be apoptotic. These results further support a role for Fas-mediated lymphocyte and macrophage apoptosis in this model of CNS autoimmune disease but they also suggest a more complex role for Fas / Fas-L interactions in CNS autoimmunity, including resident cells.  1999 Elsevier Science B.V. All rights reserved. Keywords: Multiple sclerosis; Immunity; CD95; Neuron; Glia

1. Introduction The Fas antigen, also referred to as APO-1 and CD95, is a transmembrane type 1 molecule that belongs to the tumor necrosis factor receptor / nerve growth factor (TNF / NGF) receptor superfamily and is known to induce apoptosis when it binds to its ligand Fas-L [1]. Under physiological conditions, Fas is broadly expressed by many tissues, including lung, thymus, heart and liver, and low levels are found in brain [2]. Fas-L is mainly expressed in activated T lymphocytes and certain cell types of the eye and testis [3,4]. The Fas / Fas-L system has been described as being *Corresponding author. University of California, San Francisco, Department of Neurology, S-258, 505 Parnassus Avenue, San Francisco, CA 94143-0435, USA. Tel.: 11-415-502-1330; fax: 11-415-502-1331. E-mail address: [email protected] (J.-C. Ouallet)

involved in a number of regulatory mechanisms within the immune system, including apoptosis-mediated immunologic tolerance and T-cell- and natural killer (NK)-cell cytotoxicity [5]. Experimental autoimmune encephalomyelitis (EAE) is a T-cell-mediated demyelinating disease, which serves as an animal model for multiple sclerosis (MS). Migration of immune cells into the central nervous system (CNS) is a key event in the pathogenesis of inflammation in EAE. However, mechanisms by which these cells reach or leave established inflammatory infiltrates are poorly understood. Pender et al. [6] and Nguyen et al. [7] first drew attention to apoptosis as a possible mechanism of lymphocyte and macrophage removal in inflammatory brain lesions of EAE. Here, we have examined Fas–Fas-L expression in the CNS and its relationships with apoptosis of immune cells

0022-510X / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 99 )00157-4

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

and the clinical course of EAE in the DA rat. We attempted to determine if the Fas system was involved in CNS inflammation and at the time of immune cell apoptosis. Our findings indicate that the Fas / Fas-L system is overexpressed for a very short period just before and during the time of clinical recovery. Apoptosis of CD41 cells and macrophages dramatically involves the CNS at the same period. However, Fas and Fas-L proteins are also expressed in non-apoptotic glial cells and neurons, and may play a more complex (non-apoptotic) role in CNS autoimmunity.

2. Methods

2.1. Animals, induction and clinical assessment of EAE Male Dark Agouti (DA) rats, 10–11 weeks old, were purchased from CERJ (St Genest, France). They were kept for a two-week adaptation period before use in experiments. All rats were provided with food and water ad libitum. They were weighed and examined for clinical symptoms daily. Rats were injected subcutaneously at the tail base under light anesthesia with 0.05 ml of a mixture containing 15 mg / ml whole myelin in isotonic saline and an equal volume of complete Freund’s adjuvant (CFA; Difco) containing 2 mg / ml Mycobacterium tuberculosis H37RA (Difco). Purified whole myelin was prepared from syngenic spinal cords using a sucrose gradient, as described elsewhere [8], and was then lyophilized and stored at 2208C until use. Three animals were sacrificed at 5, 9, 11, 15, 25 and 45 days post-inoculation (dpi). Age–strainmatched control animals included naive, unimmunized animals (n56).

2.2. Tissue sampling and processing Animals were euthanized under anesthesia with acepromazin and sodium thiopental, and were perfused through the left cardiac ventricle with a solution of 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4. Following perfusion, the spinal cords were carefully removed and stored in the same fixative for 30 min, then transferred to a solution of 30% sucrose in phosphate-buffered saline (PBS) and kept overnight at 48C. The spinal cords were split into three identical size regions (cervical, thoracic and lumbar) and were cryopreserved in tissue-tek OCT compound (Miles, USA). Sections (10 mm) were attached to gelatin-coated slides and stored at 2208C until use.

2.3. In situ terminal transferase-mediated dUTP nick end-labeling ( TUNEL) Spinal cord sections (10 mm thick) were incubated with methanol for 30 min at room temperature. Slides were rinsed with PBS and incubated in permeabilization solution

97

(0.1% Triton X-100 in 0.1% sodium citrate) for 2 min on ice (48C). Slides were then rinsed twice with PBS and incubated with 50 ml of TUNEL (45 ml of terminal transferase solution and 5 ml of fluorescein labeled nucleotide solution; Boehringer Mannheim, Germany) reaction mixture for 60 min at 378C. The sections were rinsed in PBS, mounted and examined under a Leitz fluorescence microscope. Control sections without terminal transferase solution were always negative. In double-labeling immunofluorescence, for identification of the type of cells undergoing apoptosis, a standard immunocytochemistry study was performed immediately after incubation with the TUNEL reaction mixture (see below). Comparing the number of apoptotic cells determined either by morphological criteria or by in situ nick end-labeling techniques has previously been shown to have a correlation of nearly 1:1 in this model [21].

2.4. In situ hybridization ( ISH) and immunohistochemistry In situ hybridization was performed using a mixture of three digoxigenin-labeled single-strand DNA probes (RandD Systems, Oxon, UK) that recognize sequences in the positions 39 3070, 39 4519 and 39 6491 of the rat Fas gene. Hybridized probes were detected using a nonradioactive ‘nucleic acid detection kit’ (Boehringer-Mannheim), according to the manufacturer’s instructions. The probe cocktail was at a final concentration of 2 or 5 ng ml 21 . Control procedures, either digestion of cellular signal with RNase A (40 mg ml 21 for 30 min at 378C) or competition with excess unlabeled probe cocktail (40-fold) in the hybridization mixture abolished all specific mRNA signal. Absence of signal was also observed by omitting the probe mixture during hybridization. For protein immunochemistry, the Fas antigen was identified by a rabbit anti-rat Fas M20 antibody (TEBUSanta Cruz Biotechnology, Santa Cruz, CA, USA). This antibody recognizes amino acids 308–327 in the carboxyterminal cytoplasmic portion of Fas. The Fas-L protein was identified by a rabbit anti-rat FAS-L N 20 antibody (TEBU-Santa Cruz Biotechnology). This antibody recognizes amino acids 2–19 in the amino-terminal cytoplasmic portion of Fas-L. Neurons were identified using a mouse anti-MAP2 2a12b monoclonal antibody (Sigma, St Quentin Fallavier, France). Astrocytes were identified using a mouse anti-GFAP GA5 monoclonal antibody (Serotec, Oxford, UK). Oligodendrocytes were identified using a mouse anti-CNPase monoclonal antibody (Sigma). CD41 cells were identified using a mouse anti-CD4 MRC OX-38 monoclonal antibody (Serotec). Macrophages were identified with a mouse anti-ED1 monoclonal antibody (Serotec). In double-labeling immunofluorescence studies, the secondary antibodies, donkey rhodamine-labeled antimouse immunoglobulins, donkey biotinylated anti-rabbit immunoglobulins and streptavidin conjugated to fluores-

98

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

cein-labeled [Jackson ImmunoResearch (IR), Pennsylvania, USA], were used. The sections were examined under a Leitz fluorescence microscope. For peroxidase staining, a biotin–streptavidin two-step labeling procedure was used: donkey anti-rabbit immunoglobulins (Jackson IR) were visualized with streptavidin conjugated to peroxidase (Jackson IR). Visualization was performed with diaminobenzidine (Dako) containing 0.5% NiCl 2 . The sections were dehydrated in ethanol, clarified in xylene and examined under a Leitz light microscope. Control sections, which were labeled either with anti-FAS or anti-FAS-L antibodies, presented identical patterns of staining to those obtained using double-labeled sections. Control sections without primary antibodies were always negative. Antibody concentration, incubation time and other experimental conditions, were identical for all animals used in this study. For each experiment, controls included naive animals of the same strain and age that were sacrificed at the same time point and under the same conditions as rats with the disease. Hence, working procedures for each spinal cord section from rats with EAE were performed in the same slides with spinal cord section from control naive animals. The percentage of double-labeled cells was determined by counting 20 cells per field at a magnification of 100 in three different areas of each spinal cord section.

3. Results

3.1. Clinical course of EAE Animals developed weight loss by five–six dpi. Ascending paralysis started from 9 dpi (limp tail) and peaked at 10–12 dpi (hind limb paralysis). Eleven to thirteen days after EAE induction, the animals were in the critical phase of the disease and presented total hind limb paralysis, and experienced a weight loss of 16–20%. In most animals, clinical recovery occurred by 15–20 dpi.

Fig. 1. Apoptotic cell death in the CNS of DA rats with EAE. Cervical spinal cord, 11 days after immunization. TUNEL-positive cells in the white matter. Magnification 3500.

identified, no glial cells or neurons displaying apoptotic features were observed. In particular, we failed to observe any apoptotic CNPase-positive oligodendrocytes. TUNELpositive cells represented 1% of macrophages (ED1-positive cells) present in the infiltrates at nine dpi, 5% at 11 dpi and 30% at 15 dpi. TUNEL-positive CD41 cells represented 5% of the total CD41 cells at nine dpi, 20% at 11 dpi and only 5% at 15 dpi. Thus, while apoptosis appeared to involve mostly CD41 cells by day 11 pi, apoptosis of macrophages appeared to be delayed (by day 15 pi). In control, unimmunized animals and before nine dpi or after 15 dpi in immunized animals, no TUNEL-positive cells were observed.

3.2. Assessment of apoptosis Considerable lymphocytic and macrophage infiltration was present in all levels of the spinal cord examined between 9 and 15 dpi. During this time, intense TUNEL staining was observed in the cervical (Fig. 1), thoracic and lumbar spinal cord. TUNEL-positive cells were encountered throughout the spinal cord although most of them were found in the white matter. There was no perivascular pattern of TUNEL-positive cells. Few apoptotic cells were found in meningeal spaces. Fewer TUNEL-positive cells were detected at nine dpi and mainly in the periphery of the spinal cord’s white matter. Double staining indicated that the apoptotic cells mostly comprised CD41 cells and macrophages (Fig. 2). Even though approximately 30% of the cells staining with the TUNEL technique were not

Fig. 2. Macrophage apoptotic cell death. Double labeling with TUNEL (A) and ED1 (B) showing apoptotic macrophages (arrows), 15 days after immunization, in the lumbar spinal cord. Magnification 3300.

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

99

3.3. Time course of Fas expression Fas expression was predominantly detected eleven days after EAE induction, at the time when the animals were in the critical phase of clinical disease and when typical EAE inflammatory infiltrates were present in the spinal cord’s white matter. At this time, Fas expression was quantitatively identical in cervical, thoracic or lumbar spinal cord. Cells expressing Fas were identified by double labeling immunofluorescence as CD41 cells, macrophages (Fig. 3), astrocytes and neurons. In the gray matter, a faint immunoreactivity for Fas was observed, and this was confirmed by low levels of Fas mRNA expression. Some labeled cells with the morphology of neurons were observed mainly in the ventral horn of the spinal cord. These cells were identified as neurons by double-labeling immunochemistry with the anti-MAP2 antibody (Fig. 4). In contrast, intense expression of Fas was detected in the periphery of inflammatory infiltrates in the white matter. These Faspositive cells were mainly identified as CD41 cells: 80% of CD41 cells were Fas1 at 11 dpi (versus ,5% at nine dpi and 10% at 15 dpi). Some Fas-positive cells were also found in the meningeal infiltrates. When studying Fas mRNA expression by in situ hybridization, all types of Fas-immunoreactive cells displayed a strong positive signal (Fig. 5). In the parenchyma, more labeled cells were identified at a distance from vessels at 11 than at nine dpi.

Fig. 4. Fas expression on neurons. Double labeling immunofluorescence with (A) Fas antibody and (B) identification of neurons using MAP2 antibody. Lumbar spinal cord, 11 days after immunization. Magnification 3600.

Numerous macrophages expressed Fas by double-labeling immunofluorescence: 50% of ED1-positive cells were found to be Fas-positive at 11 dpi (versus 5% at nine dpi and 10% at 15 dpi) throughout sections of the spinal cord.

Fig. 3. Fas expression in rats with EAE. Spinal cord, thoracic transverse section, 15 days after immunization. Double labeling immunofluorescence with (A) Fas antibody and (B) identification of macrophages (arrows) using ED1 antibody. Fas expression is observed in white matter cells but also in large motor neuron cells (arrowheads) and in interneurons throughout the gray matter. Magnification 3300.

Fig. 5. Fas mRNA expression along the spinal cord in rats with EAE. In situ hybridization (ISH) with Fas antisense oligonucleotide probes double-labeled with digoxigenin in transverse sections of the spinal cord (final probe concentration, 5 ng ml 21 ). Fas messengers are very prevalent in immune cells, glial cells, motoneurons and interneurons. Thoracic spinal cord section, 11 days after immunization. Magnification 3160.

100

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

Fig. 6. Fas-L expression in white matter of the lumbar spinal cord, nine days after immunization. Fas-L-positive cell agglutination around a lesion (arrow). Magnification 3600.

3.4. Time course of Fas-L expression Fas-L expression occurred mainly at nine and 11 dpi. At this time, most Fas-L-positive cells were identified in the white matter infiltrates (Fig. 6). Eighty percent of CD41 cells were Fas-L-positive at nine and 11 dpi versus 20% at 15 dpi. Sixty percent of ED1-positive macrophages were Fas-L-positive at nine and 11 dpi, versus 15% at 15 dpi. Fas-L expression was identical in the cervical, thoracic and lumbar spinal cord. The most prominent Fas-L staining was localized around inflammatory lesions. Similar to Fas positivity, some neurons (Fig. 7) and astrocytes (Fig. 8) also showed Fas-L immunoreactive staining and a few Fas-L-labeled cells were also found in the meningeal areas. At 15 dpi, many Fas-L immunoreactive cells were identified as being GFAP-positive astrocytes in the white matter. There appeared to be less Fas-L expression in neurons. The animals sacrificed 25 and 45 days after EAE induction had completely recovered with no evidence of persisting paralysis or weight loss, after the initial four days of total hind limb paralysis, which occurred between days 10 and 13 post-inoculation. At this time, no immune cells were found throughout the parenchyma of the spinal

Fig. 7. Fas-L expression in a thoracic transverse section of gray matter from the spinal cord, 11 days after immunization. Note the neuronal staining (arrows) in the ventral horn of the spinal cord. Magnification 3300.

Fig. 8. Fas-L expression on astrocytes. Lumbar spinal cord, 15 days after immunization following double-labeling immunofluorescence for the detection of (A) Fas-L and (B) GFAP (arrows). Magnification 3600.

cord. At 25 dpi, only slight Fas and Fas-L glial and neuronal expression was detected. At 45 dpi, only faint Fas staining, but no Fas-L staining, was present and these finding were similar to those found for controls, although we did observe some demyelinated lesions in the white matter, mainly in the dorsal column, though without inflammatory infiltrates. In control animals, only faint expression of Fas was found: some astrocytes and neurons showed a slight Fas protein signal, using immunohistochemistry, and a Fas mRNA signal, using in situ hybridization. The ventral horn region of the spinal cord showed mainly neuronal cytoplasmic Fas expression (Fig. 9). These data were similar to those obtained in unimmunized

Fig. 9. Fas expression in control animal. A slight Fas signal was detected on glial cells and neurons. Thoracic spinal cord. Magnification 3200.

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

Fig. 10. Fas-L expression in a control animal. No Fas-L signal was detected. Cervical spinal cord. Magnification 3300.

animals and five days after EAE induction in immunized animals. Expression of Fas-L was not observed either in unimmunized animals (Fig. 10) or in immunized animals at day five, before onset of the clinical disease. It is of special interest that no oligodendrocytes (identified using double immunostaining with anti-CNPase antibody) were immunoreactive for Fas or Fas-L at any time after EAE induction (5, 9, 11, 15, 25 and 45 dpi).

4. Discussion Our results indicate the presence of Fas and Fas-L expression and its correlation with apoptosis of immune cells infiltrating the CNS during a very short period in rat EAE (Fig. 11). These findings suggest that Fas / Fas-Ldriven immune cell apoptosis may contribute, at least in part, to the clearance of CNS inflammation during EAE. Moreover, we have shown that Fas and Fas-L expression is upregulated not only on inflammatory immune cells, but also on resident astrocytes and even on neurons, although these cells are not apoptotic. Several reports [9,10] have demonstrated the occurrence of T-cell apoptosis in EAE. Investigating T-cell apoptosis in inflammatory brain and spinal cord lesions of rats suffering from acute EAE, Schmied et al. [9] found that, in adoptively transferred EAE, 64% of all cells undergoing apoptosis were T lymphocytes expressing the a / b T-cell receptor. In these studies, programmed cell death of T cells reached maximal levels at day seven, when the animals were already recovering from the disease. Apoptosis of infiltrating T cells was associated with increased circulating corticosterone levels at the onset of remission and the adrenocortical response appears to be important in this process [11]. Very few apoptotic cells were detected outside the CNS, in spleen and mesenteric lymph nodes [9,11], and, thus, apoptosis of T-cells appeared to be restricted mainly to the CNS. Several authors have shown that autoreactive T-cell apoptosis occurred mainly during

101

the stage of maximum clinical disease and first days of recovery [10,12]. The current data show that, in addition to T-cells apoptosis, apoptosis of macrophage may also play an important role in the disease process, and CD41 T-cell death occurs significantly earlier than macrophage apoptosis. In contrast to previous findings [9], we did not detect any oligodendrocyte apoptosis in our study. This discrepancy may be due to differences in the disease model (adoptive transfer versus active immunization), strain-specific characteristics or to the method used to identify oligodendrocytes. It is possible that oligodendrocytes express low levels of Fas in some circumstances, although they seem to resist Fas-L-mediated apoptosis [13]. Fas (APO-1, CD95) was initially described as a cellsurface molecule that could mediate apoptotic cell death of activated normal human T cells [14]. However, Fas is broadly expressed in the body on many different kinds of cells. For example, Fas is naturally expressed on some endothelial cells [15]. T cells upregulate expression of Fas antigen and its ligand within 24 h of T-cell receptor stimulation [16]. Immunological analyses have indicated that the Fas system is involved in the clonal deletion of autoreactive T cells in the periphery, in the down-regulation of the immune reaction and cytotoxic T-lymphocytemediated cytotoxicity. CD8 cytotoxic T cells may specifically kill CD4 encephalitogenic T cells in EAE [3] and this process is known to involve the Fas / Fas-L apoptotic pathway. Moreover, CD81 T cells clearly decrease the propensity of developing relapses in EAE [17]. Macrophage apoptosis might also be caused by cytotoxic T cells, which kill their targets by inducing apoptosis. Encephalitogenic myeline basic protein-specific VB8.21 Tcells, assessed by flow cytometry, have a high frequency of Fas and Fas-L expression [18] and are highly vulnerable to apoptosis [12]. In adoptive SJL / J mice with EAE, a positive correlation was shown between disease activity and up-regulation of the Fas system using flow cytometry and reverse transcriptase PCR [19]. In lpr and gld mice carrying mutations of Fas and Fas-L, respectively, fewer apoptotic cells were detected in inflammatory lesions than in wild-type lesions of similar severity [20]. Hence, taken together, these results imply that the Fas / Fas-L system may play a role in the elimination of self-reactive immune cells in EAE. Otherwise, it is clear that the lpr or gld mutations ameliorate clinical signs of EAE [20,13] although the numbers of TH1 cells and macrophages infiltrating the CNS are not altered [20] and, in some animals, the acute disease takes a more prolonged course [13]. In these lpr and gld mice, maturation of the immune system is strongly modified in the periphery, with accumulation of CD4 2 CD8 2 T cells and this may account for the relative resistance to induction of EAE. Fas ligation seems to exhibit a costimulatory effect on the proliferation of freshly isolated CD4-positive T cells in the process of CD3-mediated triggering [21] and enhances production of interleukin-2 (IL2) [22]. Thus, Fas ligation

102

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

Fig. 11. Fas (A) and Fas-L (B) expression in CD41 cells and ED1-positive cells (macrophages) and its correlation with apoptosis (C) and clinical course (D) in DA rats with acute EAE; Fas-, Fas-L- and TUNEL-positive cells were counted per field at a magnification of 100 (each time point, average of three animals studied). Neurological manifestations were scored on a scale of zero to five (average of 18 rats studied): 05no clinical disease; 15weight loss; 25flaccid tail; 35hind limb paresis; 45complete hind limb paresis; 55moribund or dead. Fas / Fas-L system is upregulated during a short period coinciding with acute clinical disease and the presence of inflammatory infiltrates in the CNS.

does not always activate programmed cell death and it is now well known that, on resting T lymphocytes, it can co-stimulate proliferation with the T-cell receptor / CD3 complex [22]. Fas ligation may also induce Fas-L expression or IL8 expression without cell death [23]. These processes could be explained by the role of proteins interacting with the cytosolic domain of Fas having an inhibitory effect on Fas apoptotic signal transduction [24]. Our findings that the Fas system may be expressed by resident CNS cells even though these cells are not apoptotic also emphasizes this probable complex, non-apoptotic role of the Fas / Fas-L system.

It was recently shown in multiple sclerosis (MS) that Fas is expressed by some oligodendrocytes in chronic demyelinating lesions [25,26]. Surprisingly, this was correlated with oligodendrocyte cytolytic, and not apoptotic, death. Few endothelial cells and few infiltrating lymphocytes were also Fas-immunoreactive. Immunohistochemical studies showed that microglia and infiltrating lymphocytes displayed intense immunoreactivity to Fas ligand. Another recent immunohistochemical study detected large numbers of Fas-L-bearing cells in chronic lesions in MS. The type of cells involved were mainly resident glial cells [27]. Fas immunoreactivity was also found around MS

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

103

Fig. 11. (continued)

lesions. The expression of the Fas / Fas-L system on resident CNS cells may be understood in relation to the role of glial cells in immunomodulation throughout the CNS in inflammatory diseases. Recently, Villarroya et al. [28] also demonstrated tumor necrosis factor-a (TNFa) expression in central neurons of spinal cord from Lewis rats after EAE induction. This draws new attention to neurons that may play a significant role in immune events in the CNS. Ichikawa et al. [29] showed increased Fas antigen on T cells of the cerebrospinal fluid (CSF) in MS. In systemic lupus erythematosus (SLE), Jodo et al. [30] found highly soluble Fas in serum of SLE patients and a correlation with the activity of the disease. These results emphasize the probable but complex role of this system in active autoimmune diseases with no systematic correlation between Fas / Fas-L expression and apoptosis.

The Fas / Fas-L system is strongly and quickly upregulated in the CNS at the time of acute disease and recovery in EAE, including resident CNS cells. Immune cell clearance and apoptosis may not be the only key to understanding the role of Fas in the CNS. Despite the negative results in lpr and gld mice with EAE, evidence for Fas / Fas-L upregulation during active inflammation in the CNS may still be of great interest in understanding mechanisms for CNS immunity.

Acknowledgements Acknowledgements to Claude Genain and Hans Christian Von Buedingen for their valuable assistance in reviewing the manuscript. This study was supported in part by a

104

J.-C. Ouallet et al. / Journal of the Neurological Sciences 170 (1999) 96 – 104

grant from La Fondation pour La Recherche Medicale, France.

References [1] Suda T, Takahashi T, Golstein P, Nagata S. Molecular cloning and expression of the Fas ligand, a novel member of the TNF family. Cell 1993;75:1169. [2] Watanabe-fukunaga R, Brannan CI, Itoh N, Yonehara S, Copeland NG, Jenkins NA, Nagata S. The cDNA structure, expression, and chromosomal assignment of the mouse Fas antigen. J Immunol 1992;148:1274–7. [3] Griffith TS, Brunner T, Fletcher SM, Green DR, Ferguson TA. Fas ligand-induced apoptosis as a mechanism of immune privilege. Science 1995;270:1189–93. [4] Suda T, Okazaki T, Naito Y, Yokota T, Arai N, Ozaki S, Nakao K, Nagata S. Expression of the Fas ligand in cells of T cell lineage. J Immunol 1995;154:3806–13. [5] Oshimi Y, Oda S, Honda Y, Nagata S, Miyazaki S. Involvement of Fas ligand and Fas-mediated pathway in the cytotoxicity of human natural killer cells. J Immunol 1996;157:2909–15. [6] Pender MP, Nguyen KB, McCombe PA, Kerr JF. Apoptosis in the nervous system in experimental allergic encephalomyelitis. J Neurol Sci 1991;104:81–7. [7] Nguyen KB, McCombe PA, Pender MP. Macrophage apoptosis in the CNS in EAE. J Autoimmun 1994;7:145–52. [8] Pelletier L, Rossert J, Pasquier R, Villarroya H, Belair MF, Vial MC, Oriol R, Druet P. Effect of HgCl 2 on experimental allergic encephalomyelitis in Lewis rats. HgCl 2 -induced down-modulation of the disease. Eur J Immunol 1988;18:243–7. [9] Schmied M, Breitschopf H, Gold R, Zischler H, Rothe G, Wekerle H, Lassmann H. Apoptosis of T lymphocytes in experimental autoimmune encephalomyelitis. Am J Pathol 1993;143:446–52. [10] Tabi Z, McCombe A, Pender MP. Apoptotic elimination of Vb 8.2 ] cells from the CNS during recovery from EAE induced by the 1 passive transfer of V b 8.2 encephalitogenic T cells. Eur J ] Immunol 1994;24:2609–17. [11] Smith T, Schmied M, Hewson AK, Lassmann H, Cuzner L. Apoptosis of T cells and macrophages in the CNS of intact and adrenalectomized Lewis rats during EAE. J Autoimmun 1996;9:167–74. [12] Ishigami T, White CA, Pender MP. Soluble antigen therapy induces apoptosis of autoreactive T cells preferentially in the target organ rather than in the peripheral lymphoid organs. Eur J Immunol 1998;28:1626–35. [13] Malipiero U, Frei K, Spanaus KS, Agresti C, Lassmann H, Hahne M, Tschopp J, Eugster HP, Fontana A. Myelin oligodendrocyte glycoprotein-induced autoimmune encephalomyelitis is chronic / relapsing in perforin knockout mice, but monophasic in Fas- and Fas ligand-deficient lpr and gld mice. Eur J Immunol 1997;27:3151–60. [14] Lynch DH, Ramsdell F, Alderson MR. Fas and FasL in the

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25]

[26]

[27]

[28]

[29] [30]

homeostatic regulation of immune responses. Immunol Today 1995;16:569–74. Dong C, Wilson JE, Winters GL, McManus BM. Human transplant coronary disease: pathological evidence for Fas-mediated apoptotic cytotoxicity in allograft arteriopathy. Lab Invest 1996;74:921–31. Dhein J, Walczac H, Baumler C, Debatin KM, Krammer PH. Autocrine T-cell suicide mediated by APO-1 (Fas / CD95). Nature 1995;373:438. Koh DR, Fung-Leung WP, Ho A, Gray D, Acha-Orbea H, Mak TW. Less mortality but more relapses in experimental allergic encephalomyelitis in CD82 / 2 mice. Science 1992;256:1210–3. White CA, McCombe PA, Pender MP. The roles of Fas, Fas ligand and bcl-2 in T cell apoptosis in the central nervous system in experimental autoimmune encephalomyelitis. J Neuroimmunol 1998;82:47–55. Bonetti B, Pohl J, Gao YL, Raine CS. Cell death during autoimmune demyelination: effector but not target cells are eliminated by apoptosis. J immunol 1997;159:5733–41. Sabelko KA, Selly KA, Nahm MH, Cross AH, Russel JH. Fas and Fas ligand enhance the pathogenesis of experimental allergic encephalomyelitis, but are not essential for immune privilege in the central nervous system. J Immunol 1997;159:3096–9. Miyawaki T, Uehara T, Nibu R, Suji TT, Yachie A, Yonehara S, Taniguchi N. Differential expression of apoptosis-related Fas antigen on lymphocyte subpopulations in human peripheral blood. J Immunol 1992;149:3753–8. Alderson MR, Armitage RJ, Maraskovsky E, Tough TW, Roux E, Schooley K, Ramsdell F, Lynch DH. Fas tranduces activation signals in normal human T lymphocytes. J Exp Med 1993;178:2231. Abreu-Martin MT, Vidrich A, Lynch DH, Targan SR. Divergent induction of apoptosis and IL-8 secretion in HT-29 cells in response to TNFa and ligation of Fas antigen. J Immunol 1995;155:4147–54. Sato T, Irie S, Kitada S, Reed JC. FAP-1: a protein tyrosine phosphatase that associates with Fas. Science 1995;268:411–5. D’Souza SD, Bonetti B, Balasingam V, Cashman NR, Barker PA, Troutt AB, Raine CS, Antel JP. Multiple sclerosis: Fas signaling in oligodendrocyte cell death. J Exp Med 1996;184:2361–70. Bonetti B, Raine CS. Multiple sclerosis: oligodendrocytes display cell death-related molecules in situ but do not undergo apoptosis. Ann Neurol 1997;42:74–84. Dowling P, Shang G, Raval S, Menonna J, Cook S, Husar W. Involvement of the CD95 (APO-1 / Fas) receptor / ligand system in multiple sclerosis brain. J Exp Med 1996;184:1513–8. ´ ´ Villarroya H, Marie Y, Ouallet JC, Le Saux F, Tchelingerian JL, Baumann N. Expression of TNF a in central neurons of Lewis rats spinal cord after EAE induction. J Neurosci Res 1997;49:1–8. Ichikawa H, Ota K, Iwata M. Increased Fas antigen on T cells in multiple sclerosis. J Neuroimmunol 1996;71:125–9. Jodo S, Kobayashi S, Kayagaki N, Ogura N, Feng Y, Amasaki Y, Fujisaku A, Azuma M, Yagita H, Okumara K, Koike T. Serum levels of soluble Fas /Apo 1 (CD95) and its molecular structure in patients with systemic lupus erythematosus (SLE) and other autoimmune diseases. Clin Exp Immunol 1997;107:89–95.