IFN-β Inhibits T Cell Activation Capacity of Central Nervous System APCs

Mice

All mice used for cell cultures were of the B10.RIII strain. For EAE experiments, IFN-β^−/− mice generated as described previously (28) were backcrossed for 12 generations to the B10.RIII strain, and these mice were compared with their IFN-β^+/+ littermates. All mice were bred and kept at the conventional animal facility at Lund University (Lund, Sweden). All experiments were performed in accordance with the ethical committee in Malmö-Lund, Sweden.

Glial cell cultures

Primary mixed glial cell cultures were established from the forebrain and cerebellum of 1- to 2-day-old B10.RIII mice. The tissues were carefully dissected out and freed of meninges before being placed in HBSS solution supplemented with 1 mM pyruvate and 11 mM glucose. Tissue was then chopped into smaller pieces using a razor blade and incubated with 1% trypsin (Sigma-Aldrich) for 10 min at 37°C. Thereafter, DNase and FCS were added to a final concentration of 0.12 and 0.5%, respectively, and incubated for 8 min. Finally the solution was mechanically dissociated to single cell levels using a fire-constricted Pasteur pipette. The cells were seeded at a concentration of 1.5 × 10⁵ cells/cm² in a 1:1 mix of DMEM and F12 medium (Invitrogen Life Technologies) supplemented with 0.16 μM/ml penicillin, 0.03 μM/ml streptomycin, and 5% FCS. Medium was changed every third to fourth day. After 7–10 days in culture, oligodendrocytes were detached from the cell monolayer by shaking on an orbital shaker. For expansion, the cells were split using a trypsin and EDTA solution (Sigma-Aldrich) upon reaching confluence. The cells were used for T cell coculture after two to three passages and consisted of a mixture of 20–30% Mac-1⁺ microglia and 70–80% glial fibrillary acidic protein (GFAP⁺) astrocytes. Pure microglial and astrocytic cultures were established as described (29). Briefly, after 8 days in culture, the mixed glial cells were vigorously shaken at 900-1000 rpm for 3 h on an orbital shaker. For microglial cultures, the floating cells were collected, washed, and reseeded in 96-well plates (Nunc) at a concentration of 1 × 10⁴ cells/well. After adhering overnight, nonadherent or loosely attached cells were washed away. The adherent cells represented >97% pure microglial cells as determined by Mac-1⁺ staining, with <3% being GFAP⁺. For astrocyte cultures, the still adherent cells were trypsinized and reseeded in flasks and left to adhere for 30 min. Floating or loosely attached cells were recovered by mild shaking by hand and the adhesion process was repeated, this time for 1 h. Cells in the supernatant were thereafter collected, washed, and reseeded in 96-well plates (Nunc) at a concentration of 4 × 10⁴ cells/well. These cells were >96% pure astrocytes as determined by GFAP⁺ staining, <4% were Mac-1⁺ cells. Both these cultures were used within 24–48 h after establishment for coculture with T cells.

Establishment of myelin-specific T cell lines

Two T cell lines were generated by immunizing three 8- to 12-wk-old male mice in the hind paws and tail base with a total of 200 μl of a 1:1 emulsion of 250 μg of either myelin basic protein (MBP)_89–101 or myelin oligodendrocyte glycoprotein peptide 35–55 in PBS and CFA containing 100 μg of Mycobacterium tuberculosis H37Ra (Difco). Draining lymph nodes were collected day 10 postimmunization and the cells went through several rounds of Ag-specific restimulation as described (30). This generated T cell lines of the Th1 phenotype that produced large amounts of IFN-γ and IL-2 and very low amounts of IL-4 upon Ag stimulation. In the experiments in our study, both MBP_89–101-specific and myelin oligodendrocyte glycoprotein peptide (35–55)-specific T cell lines have been used with very similar results. Hence, no distinction will be made between them in the following sections. In all experiments, the T cell lines had been subjected to from four to eight stimulation rounds.

Coculture of T cells and glial cells

After the last trypsinization, glial cells were seeded in 96-well plates (Nunc) at 4–5 × 10⁴ cells/well. After 24–48 h of adherence time, the cells were either left untreated or treated for 24 h with 100 U/ml IFN-β or IFN-γ. The cultures were then washed three times and T cells were added in a 1:1 ratio. The T cell lines were established as previously discussed, and were, in a first set of experiments, in a resting state (kept in medium supplemented with IL-2 for a minimum of 2 wk) when added to the differently treated glial cultures with or without the addition of Ag. These cells were referred to as resting T cells. The concentration of the antigenic peptides was 20 μg/ml if not stated otherwise. These cocultures were pulsed with [³H]thymidine after 48 h and harvested after 24 h of further incubation on glass fiber filters. [³H]Thymidine incorporation was measured in a beta scintillation counter (Matrix 96 Direct Beta counter; Packard Instrument). Just before harvesting, supernatants were collected for cytokine analysis. In parallel, some T cells were recollected after a total of 72 h in culture, stained for different activation markers, and analyzed in a four-color BD FACSort (BD Biosciences).

In a second set of experiments, the T cells were first activated for 48 h with splenic APCs as described. They are referred to as activated T cells, which were washed thoroughly to remove Ag before being added to untreated or IFN-β- or IFN-γ-pretreated glial cells. These cultures were [³H]thymidine pulsed after 24 h in glial cocultures and harvested after 24 h of further incubation. As before, supernatants were collected and FACS staining experiments were performed in parallel.

Transwell cocultures of T cells and glial cells

Mixed glial cultures were seeded at 2.5 × 10⁵ cells/well in 24-well plates (Nunc). After 24–48 h of adherence, 2.5 × 10⁵ T cells activated for 48 h with splenic APCs (activated T cells) were added to the cultures in cell culture inserts (Transwells) with a pore size of 0.4 μm (BD Falcon). This method allowed soluble molecules, but not cells, to pass between the glia surface and the Transwell insert. As controls, glia and T cells plated at the same density were cultured in 24-well plates without inserts, thus allowing cell-cell contact. After 24 h in coculture, the cells were [³H]thymidine pulsed and after an additional 24 h of incubation, the cells were harvested and supernatants were collected.

Supernatant-mediated inhibition of T cell proliferation

Resting T cells were cocultured with glial cells in cell-cell contact cultures with or without Ag (as described) and supernatants were collected under sterile conditions. To investigate whether the supernatants could inhibit T cell proliferation, resting T cells were activated with Ag and splenic APCs as described. This activation was performed in new medium or medium conditioned with different ratios of the harvested supernatants. The proliferation index was calculated as proliferation in cultures with conditioned medium divided by the control cultures containing 100% fresh medium.

FACS staining and evaluation

After 24 h of treatment with IFN-γ, IFN-β, or a combination of these two cytokines (100 U/ml IFN-γ for 12 h after which 100 U/ml IFN-β was added for the remaining 12 h), glial cell cultures were washed once with PBS and detached and dissociated using a trypsin and EDTA solution (Sigma-Aldrich). Activated or resting T cells were after 48 or 72 h in coculture, respectively, collected through vigorous flushing with the pipette. After washing in FACS buffer (2% FCS in PBS, for FACS of glial cells this buffer also contained 1 mM EDTA), cells were first incubated with anti-Fc receptor Ab (24.G.2, our hybridoma collection) at 10 μg/ml. Thereafter, glial cells were incubated with biotinylated FITC-marked or PE-marked Abs against Mac-1 (M1/70), MHC class II (Y3P or 7.16.17), CD1d (1B1), B7-1 (16-10A1), B7-2 (GL1), CD40 (3/23), ICAM-1 (3E2), VCAM-1 (429), Fas ligand (Kay-10), TGF-β1 (A75-3.1), and GFAP (Zymed Laboratories). For astrocytic GFAP staining and intracellular TGF-β1 staining, cells were first fixed using 3% paraformaldehyde and thereafter permeabilized using 0.3% saponin in FACS buffer. T cells were stained with biotinylated FITC- or PE-marked Ab against CD4 (Gk1.5), TCR αβ (H57-597), CD28 (3751), CTLA-4 (UC10), CD40L (MR1), CD25 (7D4), CD69 (H1.2F3), and CD44 (Caltag Laboratories). After washing, cells were incubated with streptavidin-PE or allophycocyanin diluted 1/400 (BD Pharmingen). Annexin V and propidium iodide (BD Pharmingen) staining was used to detect apoptotic and dead cells. In both types of staining, all Abs were allowed to bind for 20 min on ice. All Abs were used at a concentration of 1–5 μg/ml and were purchased from BD Pharmingen unless stated otherwise. The cells were analyzed using a four-color FACSort (BD Biosciences). Positivity was evaluated by comparison of negative samples stained with irrelevant isotype-matched controls and streptavidin-PE or allophycocyanin. Astrocytes were gated on GFAP, microglia on Mac-1, and T cells on CD4 and TCR expression. For FACS-based proliferation studies, T cells were CFSE labeled at a concentration of 5 nM before addition to glial cells. Cells were collected after 72 h and counterstained with allophycocyanin-labeled anti-CD4 (Gk1.5). T cells labeled just before FACS analysis were used as positive controls.

ELISA

Primary Abs: anti-IFN-γ (5 μg/ml, clone R6A2; our hybridoma collection), and anti-IL-2 and anti-TNF-α (2 and 6 μg/ml, clones JES6-1A12 and G281-2626, respectively; BD Pharmingen). Secondary biotinylated Abs: anti-IFN-γ (0.6 μg/ml, clone Ani8; our hybridoma collection), and anti-IL-2 and anti-TNF-α (both used at 1 μg/ml, clones JES6-5H4 and MP6-XT3, respectively; BD Pharmingen). ELISA were performed as previously described (25)

NO assay

Nitrite was measured as the stable end product of NO and levels were quantified by a colorimetric assay based on the Griess reaction (31).

Induction and clinical evaluation of active EAE

MBP_89–101 peptide (VHFFKNIVTPRTP-COOH; from Å. Engström, University of Uppsala, Uppsala, Sweden) was used to induce EAE in IFN-β^−/− and IFN-β^+/+ littermates. Mice were immunized s.c. at the base of the tail with 100 μl of emulsion of 250 μg of MBP_89–101 peptide in PBS and CFA containing Mycobacterium tuberculosis H37Ra (Difco). In addition, each animal received 400 ng of pertussis toxin (from Bordetella pertussis; Sigma-Aldrich) dissolved in 100 μl of PBS; this mix was given i.p. at the day of immunization and 2 days later.

Mice were sacrificed at day 13 postimmunization for immunohistochemistry during the acute phase of the EAE (25).

Immunohistochemistry

Brain and spinal cord were dissected out on day 13 postimmunization and immediately embedded in OTC compound (Sakura Finetek) and snap-frozen in isopentane on dry ice. Tissues were cryosectioned in 10-μm slices and kept at −70°C until staining. For single staining, the sections were incubated with biotinylated anti-CD1d (1B1; BD Pharmingen) after which ExtrAvidin-peroxidase (Sigma-Aldrich) and diaminobenzidine (Saveen Biotech) were used for detection. Hematoxylin was used for background staining. For double staining, the tissues were again incubated with biotinylated anti-CD1d with streptavidin-Cy3 being used for detection. Tissues were simultaneously stained with FITC-labeled anti-NK1.1 (PK136; our own hybridoma collection).

Statistical analysis

When adding results from several experiments together, differences in proliferation, cytokine, and NO production were analyzed using a paired sign t test. When comparing these parameters within single experiments, the Student t test was used. Differences in FACS marker expression were analyzed with the Student paired t test. Values of p ≤ 0.05 were considered significant.

The intrinsic capacity of glial cells as APCs is to prevent further expansion of T cells while maintaining T cell effector function

We were at first interested in determining whether there are differences between astrocytes and microglia in their Ag-presenting capacity, as there has been a great deal of controversy in the literature (32, 33, 34). As depicted in Fig. 1⇓, there was no major qualitative differences observed in between these two CNS-specific APCs in their Ag-presenting capacity, measured by inducing T cell proliferation (Fig. 1⇓A) and effector functions, determined by production of IFN-γ (Fig. 1⇓, B and C), TNF-α, and NO (data not shown). Furthermore, as shown in Fig. 1⇓, B and C, there were no discrepancies observed between these two APC populations with regard to effect of cytokine treatment (IFN-γ and IFN-β). Hence, we have used mixed culture of astrocytes and glial cells and referred to them as glial cells. Interestingly, our analysis of the glial cells’ Ag-presenting capacity revealed that glial cells do not allow Ag-specific T cell expansion, although they do induce Ag-specific T cells effector functions, measured by production of large amount of IL-2, IFN-γ, TNF-α, and NO (Fig. 2⇓). This result strongly indicates that the intrinsic function of CNS resident glial cells is to permit activated T cells to exert their effector functions once entering CNS, while still preventing them from further expansion in situ.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Astrocytes and microglia fail to induce T cell proliferation while sustaining T cell IFN-γ production. Glial cells cannot induce an Ag-specific T cell proliferation. A, CFSE-labeling of T cells in coculture with pure untreated astrocytes and Ag (solid black histogram) or with pure microglia and Ag (gray solid histogram) is shown. None of these cocultures show proliferating T cells. The hatched histogram shows the positive control of nondividing T cells, that is, T cells CFSE labeled just before FACS analysis. Both astrocytes (B) and microglia (C) do induce an Ag-specific IFN-γ response. Also, both cell types respond similarly to IFN-β or -γ treatment. The IFN-γ response seen in microglial cultures without exogenously added Ag could be explained by the presence of small amounts of endogenous myelin. No IFN-γ was observed in any cultures of astrocytes or microglial cells only. Data show one representative experiment each, and error bars represent SD of duplicate samples. ∗, p < 0.05; ∗∗, p < 0.01.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Glial cells inhibit T cell expansion but induce T cell effector functions. A, Mixed glial cultures Ag-specifically inhibit T cell proliferation but induce Ag-specific IL-2 (B), IFN-γ (C), TNF-α (D), and NO production (E). Data represent three to nine individual experiments taken together, and error bars represent SEM.

IFN-β negatively regulates the Ag-presenting capacity of glial cells by partly rescuing T cell expansion

It is commonly believed that development of a chronic, autoimmune CNS inflammation is the result of activation of autoreactive T cells in the peripheral lymphoid organs, which thereafter enter the bloodstream and transverse the blood brain barrier. It is also generally believed that these T cells require reactivation in situ to exert their deleterious effects (35, 36). Therefore, the function of CNS resident APC populations (glial cells) could be detrimental in the outcome of T cell reactivation in CNS. Thus, we investigated the effects of IFN-β on these cells and compared it with the known effect of a proinflammatory cytokine, IFN-γ. We chose to study both cytokine effects on resting T cells, thus determining whether glial cells could function as professional APCs, as well as on already activated T cells. These T cells were activated using conventional irradiated splenic APCs and Ag (see Materials and Methods) for 48 h and thereafter washed and cocultured with glial cells. In this way, one could mimic interaction of activated T cells entering the CNS with residual glial cells. Because there were no differences observed between specific T cell expansion using CSFE labeling compared with DNA synthesis and thymidine up-take (see Figs. 1⇑A and 2A), we use proliferation measured by thymidine up-take as read out. Surprisingly, treatment of glial cells with the proinflammatory cytokine IFN-γ leads to significant Ag-specific inhibition of T cells proliferation (Fig. 3⇓A). Similar findings were observed when activated T cells were used (Fig. 3⇓B). Importantly, none of the differently treated glial cells induced significant T cell death as measured by annexin V and propidium iodide staining (data not shown). Because these findings challenge several earlier reports regarding the APC role of glial cells in induction of T cell proliferation (32, 37, 38), we investigated what could be the reason for this discrepancy. As shown in Fig. 3⇓C, irradiation of glial cells before coculture with resting T cells results in entirely different findings, as irradiated glial cells could induce T cell proliferation. Furthermore, IFN-β treatment does not have any significant effect on T cell proliferation under this condition (Fig. 3⇓C). Very similar results were also achieved when using mitomycin-treated glial cells (data not shown). Hence, irradiation and/or mitomycin exposure completely alters the APC function of glial cells; it renders them capable of inducing Ag-specific proliferation. It might be assumed that this finding could partly be due to the prevention of T cell expansion as a result of glial cell proliferation and crowding in nonirradiated cultures. However, this was addressed by using half the number of glial and T cells for coculture and very similar results were achieved (data not shown). In addition, as shown earlier in Fig. 1⇑A, CSFE labeling of T cells before coculture with glial cells (which exclusively address T cell proliferation) verified the inability of glial cells to induce T cell proliferation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Glia-mediated T cell proliferation arrest requires living glia. A, Addition of Ag to cocultures of nonirradiated glial cells and resting T cells gives a significantly suppressed T cell proliferation in IFN-γ-treated cultures. IFN-β treatment results in higher proliferation compared with untreated glial cells. B, The proliferation is also suppressed in cocultures of activated T cells and IFN-γ-treated glial cells (nonirradiated). C, Lethal irradiation of the glial cells just before T cell addition results in a totally different proliferation pattern. Untreated or IFN-β-treated glial cells then induce a high Ag-specific proliferation and no suppression is seen in IFN-γ-treated cultures. Data in A and B show the mean ± SEM of eight to nine experiments. Data in C show one representative experiment in which irradiated glial cells were used. Error bars represent SD of triplicate samples. ∗, p < 0.05; ∗∗, p < 0.01.

IFN-β negatively regulates the Ag-presenting capacity of glial cells leading to inhibition of T cell effector function

As stated previously, our interest was to determine the effect of IFN-β-treated glial cells on regulation and maintenance of T cell effector function. Thus, the capacity of resting and activated T cells to produce IFN-γ and TNF-α (two proinflammatory cytokines produced by these encephalitogenic Th1 T cell lines) was measured in response to coculture with IFN-β- or IFN-γ-treated glial cells. Neither pure glial cells (astrocytes and microglia) nor mixed glial cultures produced any measurable IFN-γ or TNF-α when cultured alone in the absence of T cells, regardless of exogenous cytokine stimulation (detection level 5–10 pg/ml) (data not shown). Thus, the measured IFN-γ and TNF-α in the supernatant of cocultures most likely represent the cytokine production by T cells. As depicted in Fig. 4⇓, A and C, IFN-β reduced Ag-specific IFN-γ and TNF-α production by T cells. This result was in contrast to IFN-γ, which as expected, enhanced the proinflammatory cytokine production of T cells. In cocultures using activated T cells, glial cells induced an enhanced T cell cytokine production, and this was totally prevented in IFN-β-treated cultures (Fig. 4⇓, B and D). Because it has been suggested that IFN-β can alter the phenotypic response by driving Th1 toward a Th2 phenotype, IL-4 was measured in these cultures. However, we found no increase in IL-4 production when the glial cells were pretreated with IFN-β, rather a tendency was observed toward a similar down-modulating effect resulting in lower amounts of IL-4 (data not shown). Hypothetically, elevated levels of the immunosuppressive cytokine TGF-β1 could possibly mediate the observed effects of IFN-β pretreatment. However, no TGF-β1 was detected in any supernatant, as measured by ELISA (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

IFN-β treatment of glial cells reduces T cell production of IFN-γ and TNF-α. T cells cocultured with IFN-β-treated glial cultures induce a lower Ag-specific T cell IFN-γ (A) and TNF-α (C) response compared with untreated controls. Pretreatment with IFN-γ enhanced the cytokine response. B and D, When T cells preactivated with splenic APCs and Ag are added to glial cultures, untreated glial prolong the T cell cytokine production compared with T cells cultured alone. IFN-β pretreatment of the glial cells totally inhibits this enhancement and restores the cytokine production to values indistinguishable from T cells cultured alone. No IFN-γ or TNF-α was observed in cultures of glial cells only. Data show the mean values of eight to nine individual experiments, and error bars represent SEM of these experiments. ∗, p < 0.05; ∗∗, p < 0.01.

IFN-β treatment prevents glia-induced T cell up-regulation of CD25

Aside from cytokine production, other markers generally used to define T cell activation were also examined after coculture with IFN-β or -γ treated or untreated glial cells. CD44 expression was high in all cultures probably reflecting a memory phenotype of the encephalitogenic cell lines. CD69, in contrast, was expressed at relatively low levels in all cultures (data not shown). It is possible that investigation of the cells at a different time point would reveal distinctions that were not observed in this study as CD69 is defined as an early activation marker (39). CTLA-4 expression was very low on T cells from all cultures, whereas CD28 and CD40L were not conclusively up- or down-regulated upon coculture with differently treated glial cells (data not shown). When investigating the activation marker CD25 (the IL-2R α-chain), data well in accordance with the cytokine results were obtained. In cocultures of resting T cells and IFN-β pretreated glial cells, a down-regulation of T cell CD25 expression was observed compared with T cell cocultures with untreated glial cells. IFN-γ pretreatment increased the CD25 T cell expression. This observation was true for cultures without Ag addition as shown in Fig. 5⇓A. In the presence of Ag, a very high CD25 up-regulation was observed in all cultures and the effects of either IFN-β or IFN-γ treatment were no longer observed (data not shown). Cocultures of activated T cells with glial cells resulted in up-regulated T cell CD25 expression compared with T cells cultured alone. These data again suggest that glial cells can perpetuate the T cell effector phase and thus support the findings of cytokine production by T cells (Fig. 4⇑, B and D). In accordance, IFN-β treatment of the glial cells totally inhibited the up-regulation of CD25, while IFN-γ treatment enhanced this up-regulation even further (Fig. 5⇓B). It is important to note that no IFN-β is left in the cultures when the T cells are added, and IFN-β treatment alone does not induce an autocrine loop of IFN-β production (which was measured by ELISA, data not shown). Hence, IFN-β is not acting on the T cells, but on the glial cells in the cultures. Taken together, these data suggest that IFN-β inhibits the capacity of potential CNS APCs to function as Ag-presenting cells activating resting T cells. In addition, it prevents their capacity to sustain an ongoing T cell effector phase after Ag-specific activation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

IFN-β treatment of glial cells inhibits the up-regulation of CD25 expression on T cells. A, IFN-β treatment of glial cells reduces the CD25 expression on subsequent cocultures with resting T cells in the absence of Ag, whereas IFN-γ treatment increases this compared with untreated glial cells. Addition of Ag induces a very strong CD25 up-regulation on the T cells with no difference between treatments (data not shown). B, Coculture of activated T cells with untreated glial cells induces up-regulation of the T cell activation marker CD25 compared with T cells culture alone. IFN-β pretreatment of the glial cells totally inhibits this up-regulation, whereas IFN-γ pretreatment increases the CD25 expression even further. Data show the mean value of three experiments. Fluorescence geometric mean values ± SEM are shown in parentheses. ∗, p < 0.05; ∗∗, p < 0.01.

The effects of glial cells on T cell responses is cell-cell contact dependent

Ag-specific activation of resting T cells is obviously dependent on MHC class II and TCR interaction, and therefore cell-cell contact. Nevertheless, we investigated whether the effects of glial cells on preactivated T cells were mediated through a membrane-bound ligand or through a soluble mediator, hence whether a Transwell system was used. As demonstrated in Fig. 6⇓, A and B, both the effects of untreated glial cells on preactivated T cell proliferation and cytokine production, as well as the relative effect of IFN-β or -γ treatment, are inhibited once the cells are physically separated by a membrane. This result strongly indicates that glial cells require direct contact with T cells to exert their function. Still, the possibility remained that this cell-cell interaction resulted in the production of a soluble mediator that could in return affect/inhibit T cell proliferation. To test this hypothesis, T cells were activated, this time with splenic APCs in the presence of different amounts of conditioned medium from glial and T cell cocultures with or without Ag. The results obtained with supernatants from Ag-containing cultures are depicted in Fig. 6⇓C and show that supernatant recovered from glia-T cell cocultures inhibit T cell proliferation in a concentration-dependent manner. However, supernatants from untreated, IFN-β- or IFN-γ-treated cultures inhibited T cell proliferation to the same extent. This outcome suggests that the differences observed between IFN-β- and IFN-γ-treated glial cells were dependent on the expression of one or more cell surface molecules yet undefined and not on soluble factors.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Glial cell-mediated perpetuation of activated T cell cytokine responses requires cell-cell contact. A, The IFN-γ-mediated inhibition of proliferation requires cell-cell contact as demonstrated by the fact that inhibition is blocked in cultures where the T cells and glial cells are separated by a membrane. Error bars represent SD of triplicate samples. B, When activated T cells and glial cells are allowed cell-cell contact, an increase of the cytokine secretion by activated T cells is observed. This increase is totally repressed in IFN-β-treated cultures and enhanced in IFN-γ-treated cultures. When the T cells and glial cells are separated by a membrane, inhibiting direct cell-cell contact but allowing soluble molecules to pass freely, this cytokine production enhancement is blocked. Error bars represent SD of triplicate samples. C, Different concentrations of supernatants from cocultures containing glial cells, T cells, and Ag were added to T cells stimulated by splenic APCs, and proliferation was measured. Data show that there is a soluble mediator released upon cell-cell contact in glia-T cell cocultures that can inhibit proliferation to a certain extent in a concentration-dependent manner. However, the differently treated supernatants have a similar inhibitory capacity. Error bars represent SD of triplicate samples. ∗∗, p < 0.01; ∗∗∗, p < 0.001.

Glial cells inability to induce T cell proliferation is not due to high NO production or deficiency in IL-2 production

NO has previously been shown to inhibit T cell proliferation (40). Furthermore, it is well established that T cells require IL-2 for successful proliferation (41) and it has been debated whether glial cells are capable of inducing an IL-2 response (33, 42, 43, 44). As shown in Fig. 2⇑B, the glial cells induced a clear Ag-specific IL-2 response, and the T cell proliferation arrest could thus not be explained by deficiencies in IL-2 production. To investigate whether the high NO production correlated with the observed inhibition of T cell proliferation, the kinetics of these parameters were examined. Fig. 7⇓A demonstrates that proliferation increases over time and after 96 h in the IFN-γ-treated glial cell culture, the Ag-specific inhibition of proliferation is in fact overcome. Next, Fig. 7⇓B shows that the NO production is also increasing over time and reaches a peak at 96 h, correlating with the highest level of proliferation. These data suggest that the inhibition of proliferation is not due to high levels of NO production.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Proliferation arrest is not caused by high NO production. A, The kinetics of cell proliferation in glial cell (G) and T cell (T) cocultures is shown. Proliferation is expressed as a percentage of the proliferation in cultures of glial cells only. The inhibition of proliferation is Ag-dependent. The inhibition gets less efficient over time and is finally overcome at 96 h in IFN-γ-treated cultures. The kinetics of glial cell proliferation alone are not affected by the treatments (data not shown). B, All glial cultures induce an Ag-specific NO production, which increases over time. The figure shows that the highest NO levels coincide with the highest T cell proliferation. Error bars represent SD of triplicate samples in all panels.

IFN-β treatment does not decreased MHC class II or costimulatory expression on glial cells

The data presented in Fig. 6⇑C suggested that differences in expression of membrane-bound ligands could be responsible for the observed differences between T cell cocultures of untreated, IFN-β-treated, or IFN-γ-treated glia. T cells generally require signals through both MHC class II and costimulatory molecules for full activation and an effective Ag-specific response. FACS studies were performed to investigate whether IFN-β treatment caused a reduced expression of these molecules on glial cells, which could explain the achieved results. Surprisingly, no decrease was observed in any of the markers investigated in this model as result of IFN-β treatment. Microglia and astrocytes were studied separately in these experiments based on Mac-1 or GFAP expression, respectively. MHC class II was not expressed in untreated astrocytes, IFN-γ strongly up-regulated the molecule, and IFN-β alone had no effect. In accordance with previously published data (45, 46, 47), IFN-β down-regulated an IFN-γ-mediated MHC class II expression on the astrocytes (Fig. 8⇓A). The microglia population expressed low levels of MHC class II and treatment with both IFNs resulting in a MHC class II up-regulation (see Fig. 8⇓A). The costimulatory molecules B7-1, B7-2, and CD40 were also studied. Astrocytes were shown to be highly positive for B7-1, whereas negative for B7-2 and CD40 (see Fig. 8⇓, B–D). None of these molecules was affected by IFN-γ or IFN-β treatment alone, although the combination of both cytokines generated a tendency toward up-regulation of B7-2 and CD40. In the microglia population, which was weakly positive for all these molecules when untreated, both cytokines again generated higher expression of these three costimulatory molecules. An additive effect was observed with respect to B7-2 expression, in which the combination of both IFNs resulted in an even higher expression compared with any cytokine alone (see Fig. 8⇓C). ICAM-1 and VCAM-1 were also investigated. ICAM-1 was shown to be relatively highly expressed in both astrocytes and microglia, but the expression was further up-regulated by both IFN-γ and IFN-β on microglia (see Fig. 8⇓E). The pattern was similar for microglial VCAM-1 expression (see Fig. 8⇓F), whereas astrocytes expressed very high levels of VCAM that were further up-regulated by the combination treatment with both IFNs. Because glial cells have been shown to induce T cell apoptosis in a Fas ligand-dependent manner (48) and to secrete the immunosuppressive cytokine TGF-β1 (49, 50), these molecules were also studied. However, no or very low expression was observed in glial cells and there was no marked effect as a result of pretreatment with IFNs (data not shown). These results demonstrate that the suppressive effects that IFN-β treatment of glial cells exerts on both T cell activation and effector phase maintenance is not likely to be mediated through down-regulation of any of the molecules involved in classical T cell activation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

IFN-β treatment of glial cells does not induce down-regulation of MHC class II or costimulatory molecules. Mixed glial cells consisted of astrocytes and microglia distinguished by expression of GFAP and Mac-1, respectively. A, Astrocytes were negative for MHC class II when untreated or IFN-β-treated, but MHC class II was strongly up-regulated upon IFN-γ stimulation. IFN-β reduced the IFN-γ-mediated up-regulation. Untreated microglia expressed low amounts of MHC class II and this effect was further enhanced when treated with either IFN-β or IFN-γ. B, Untreated astrocytes expressed high levels of B7-1, and the expression remained unchanged upon treatment. Microglia expressed medium levels of B7-1 and the expression increased upon treatment with IFN-β or IFN-γ. C, Astrocytes were negative for B7-2 in all conditions. Untreated microglia expressed medium levels of B7-2, whereas the expression was significantly up-regulated after treatment with IFN-β, IFN-γ, or both in combination. D, Astrocytes were negative for CD40 in all cultures although a nonsignificant tendency toward CD40 up-regulation was seen in cultures treated with both IFN-β and IFN-γ. Untreated microglia expressed low levels of CD40 and the expression was increased in cultures treated with IFN-β or IFN-γ. E and F, Astrocytes express very high levels of ICAM and VCAM, respectively, which is further increased by combination treatment with both IFN-γ and IFN-β. Untreated microglia expressed only low levels but up-regulated their expression upon treatment with either IFN-β or IFN-γ.

IFN-β and IFN-γ elevate CD1d expression on glial cells

We also investigated the expression of the nonclassical MHC-like molecule CD1d because we have reported that this molecule plays a crucial role in the regulation of CNS inflammation (30). Interestingly, we found that both astrocytes and microglia express relatively high levels of CD1d and this was elevated by IFN-β as well as by IFN-γ (Fig. 9⇓A). We continued to investigate how expression of CD1d is regulated on glial cells in vivo and whether lack of endogenous IFN-β influenced CD1d expression in the CNS. As shown in Fig. 9⇓B, CD1d (stained in red) is indeed up-regulated in glial cells in the inflamed CNS of EAE mice during the acute phase of the disease. The CD1⁺ cells are mainly astrocytes (Fig. 9⇓B), as CD1 expression colocalized with GFAP expression (data not shown). Moreover, this reaction was associated with the recruitment of NK1.1⁺ CD3⁺ cells to the CNS. Fig. 9⇓B shows NK1.1⁺ infiltrating cells (stained in green) in very close proximity to CD1-expressing astrocytes. Subsequent double staining with anti-CD3 confirmed the presence of NK1.1⁺ T cells (data not shown). However, no difference was found between IFN-β^−/− mice and their IFN-β^+/− littermates with respect to the regulation of CD1d expression in glial cells during the course of EAE (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9.

Glial cells express CD1d NK1.1⁺ infiltrating cells in close vicinity of CD1⁺ glial cells in the CNS of EAE mice. A, Untreated astrocytes and microglia expressed medium levels of CD1d but the expression increased upon treatment with either IFN-β or IFN-γ. Data show the mean value ± SEM for 10 experiments. Significant difference between untreated and IFN-treated cultures is shown ∗, p < 0.05; ∗∗; p < 0.01; and ∗∗∗, p < 0.001. B, NK1.1⁺ infiltrating cells (green) in close interaction with CD1⁺ glial cells (red). The image is taken around an infiltrate from the cerebellum of an EAE-affected B10.RIII wild-type mouse on day 13 after immunization.