IL-10, a cytokine with important anti-inflammatory properties, is generated within the CNS during neuroinflammation. The mechanism for its production is poorly understood. Since infiltrating lymphocytes come into close proximity with the macrophage-like cells of the CNS, the microglia, we have used an in vitro human microglia-T cell coculture system to address the mechanisms of IL-10 production. We demonstrate that microglia or activated T cells alone secrete negligible amounts of IL-10, but that their coculture results in significant IL-10 production, which was effected by both cell types. IL-10 generation was cell contact dependent, and treatment with anti-CD40, CTLA-4-Fc, or anti-CD23 decreased the IL-10 content in microglia-T cell cocultures. The combination of anti-CD40 and CTLA-4-Fc reduced IL-10 levels to the negligible amounts seen with T cells or microglia in isolation. By also measuring TNF-α levels, specificity of cytokine regulation was observed; while anti-CD40 and CTLA-4-Fc reduced IL-10 and TNF-α levels, anti-CD23 did not affect TNF-α while attenuating IL-10 generation. Anti-very late Ag-4, which decreased TNF-α levels, did not affect IL-10. These results implicate the CD40, B7, and CD23 pathways in IL-10 production following microglia-T cell encounter and have relevance to the regulation of an anti-inflammatory response within the CNS.
Multiple sclerosis (MS)3 is an inflammatory disease of the CNS that leads to demyelination and loss of neurological functions. The infiltration into the CNS of activated T lymphocytes, the majority of which do not appear to be Ag specific, is considered a key event in the pathogenesis of MS or experimental allergic encephalomyelitis (EAE), an animal model of MS. The mechanisms through which T cells play an etiologic role in MS remain unclear, although once infiltrated, T cells are found in close proximity to the macrophage-like cells of the CNS, the microglia. We have previously reported that T lymphocytes interact with microglia to generate the production of TNF-α through a mechanism that involves VLA-4 on T cells with VCAM-1 on microglia (1). This result is of pathological relevance, since TNF-α has been reported to cause apoptosis of oligodendrocytes (2, 3, 4), the cells that are lost in MS. Of clinical significance is the observation that IFN-β-1b, a recombinant and modified form of human IFN-β that is effective in the treatment of MS (5), inhibits TNF-α production, probably through the down-regulation of VLA-4 on the surface of T cells (1).
IL-10 is an 18-kDa cytokine produced by a variety of cells, including monocytes/macrophages, T cells, B cells, and mast cells. In the CNS, potential sources of IL-10 include the microglia (6) and astrocytes (7). IL-10 has important anti-inflammatory properties. First, IL-10 inhibits the production of proinflammatory cytokines by many cell types, including those of the mononuclear phagocytic lineage; indeed, IL-10 was shown to inhibit the production of TNF-α and IL-12 produced by monocytes, macrophages, and microglia (8, 9, 10, 11, 12). Also, IL-10 plays a role in causing T cells to undergo anergy (inactivation or unresponsiveness) (13). Other anti-inflammatory functions of IL-10 include its inhibitory effect on the process of Ag presentation. Treatment of macrophages/microglia with IL-10 down-regulated the expression of molecules essential for the presentation of Ags, such as MHC class II (9) and the costimulatory molecules B7-1 and B7-2 (14). Finally, the role of IL-10 as an anti-inflammatory molecule is supported by the phenotype of IL-10-deficient mice; these mice develop chronic colitis, which appears to be mediated by the proinflammatory Th1 cells (15, 16, 17).
Given its anti-inflammatory role, the production of IL-10 within the CNS will probably have a favorable impact on inflammatory diseases of the CNS. Indeed, recent evidence suggests that the induction of IL-10 production may partly account for the therapeutic effect of IFN-β in MS (18), since patients treated with IFN-β have elevated IL-10 levels in their serum (19, 20) and cerebrospinal fluid, even after 2 yr of treatment, which correlated with a favorable therapeutic response (21). In EAE, an animal model of MS, the expression of IL-10 in the brains of mice afflicted with the disease is elevated during the recovery phase of the disease (22). IL-10 was shown to prevent EAE in rats (23), although this was not confirmed (24). Nonetheless, in mice genetically deficient for IL-10, the development of EAE following immunization with myelin oligodendrocyte glycoprotein was accelerated compared with that in wild-type controls, and these mice did not spontaneously recover from EAE, unlike the wild-type controls (25). Another group demonstrated that IL-10-deficient mice were more susceptible and developed a more severe EAE than IL-4-deficient or wild-type mice; furthermore, IL-10 transgenics were resistant to the development of EAE (26).
The mechanisms by which IL-10 is produced within the CNS are unclear. We postulated that T lymphocytes could be an important trigger of IL-10 production by microglia, since the infiltration of T cells into the CNS is a key pathogenic event in several neuroinflammatory disorders, including MS. In this study we investigated whether and how IL-10 is generated from the interaction of T lymphocytes with microglia in vitro. This report demonstrates that IL-10 is produced as a result of human microglia-T cell interactions, and that this is due to a contact-dependent mechanism involving the B7 molecules, CD23 and CD40.
Materials and Methods
Cell culture and treatment
Microglia and T cells were isolated, and cocultures were performed as described previously (1). In brief, adult human microglia were isolated from cortical tissue obtained from patients undergoing surgery to treat intractable epilepsy or brain tumor as previously described (27, 28). After purification of the microglia population (>95% purity) from the primary culture of mixed glial cells, 2.5 × 104 microglia were plated per well of a 96-well plate.
Mononuclear cells (MNC) were isolated from the blood of healthy individuals using Ficoll Paque (Pharmacia, Piscataway, NJ) as described previously (1). After two washes, cells were grown in the serum-free medium, AIM V (Life Technologies, Gaithersburg, MD) and were activated with 1 ng/ml of an anti-CD3 Ab (OKT3) for a period of 72 h. Flow cytometric analysis of the MNC population after the activation period indicated that CD3+ cells constituted about 90% of the total cell population, with approximately 60% CD4+ and 30% CD8+. B lymphocytes (CD19+) and NK cells (CD56+) consisted of 5–6% of the total MNC population, and no monocytes (CD14+) were detected. Henceforth, given that the majority of cells in the MNC population are T cells, it will be referred to as T lymphocytes. T cells were counted, and 5.0 × 104 cells/well were added to the microglia.
We urge caution in interpreting the results of this study, as we used an enriched T cell (90%) population as well as an enriched microglia population (95%) rather than pure cultures of T cells and microglia.
In the experiment involving purified CD4+ or CD8+ T cells, the same number of purified cells (5.0 × 104) was added for coculture with microglia. To purify CD4 or CD8 populations, anti-CD3-activated cells were incubated with magnetic beads coated with a mAb against CD4 or CD8 (Dynal, Great Neck, NY) for a period of 30 min at 4°C under constant agitation. Rosetted CD4+ or CD8+ T cells were isolated using a magnet. To detach beads from purified cells, cells were incubated for 16–20 h at 37°C in a CO2 incubator, and detached beads were then removed by placing the tube on a magnet.
Where indicated, microglia were treated with culture medium containing various concentrations (see Results) of anti-CD40, anti-CD23, or CTLA-4-Fc for a period of 30 min at room temperature before their coculture with T cells. In the VLA-4 function-blocking experiment, T cells were pretreated with 25 μg/ml of anti-VLA-4 or IgG1 isotype control for a period of 30 min at 4°C under constant agitation. Cells were then centrifuged for 2 min at 3000 rpm before resuspending them for coculture with microglia.
Reagents and mAbs
Anti-CD3 (OKT3) was provided by Dr. Jack P. Antel (Montreal, Canada). Recombinant TIMP1 (rTIMP1) and BB-94, a TNF-α-converting enzyme (TACE) inhibitor, were provided by Dr. Dylan Edwards (Norwich, U.K.). LPS was purchased from Sigma-Aldrich Canada (Oakville, Canada). Anti-VLA-4 (HP2/1 Ab) was obtained from Serotec (Raleigh, NC), while IgG1 isotype control was purchased from Chemicon International (Temecula, CA). Anti-CD23 was obtained from Dako (Copenhagen, Denmark). Anti-CD40 was purchased from Genzyme (Cambridge, MA), and anti-TNF-α and CTLA-4-Fc were obtained from R&D Systems (Minneapolis, MN). The mAb against CD80 (B7-1), BB1, as well as the rat anti-human IL-10-IgG2a-PE and its isotype control, rat-IgG2a-PE, were purchased from PharMingen Canada (Mississauga, Canada), as was anti-CTLA-4-Ig2a-PE. Finally, anti-CD40L-IgG1-PE, anti-CD3-IgG1-PE, anti-CD14-IgG1-PE, mouse IgG1, and IgG2a isotype control conjugated with PE used for flow cytometry analysis were obtained from Becton Dickinson Canada (Mississauga, Canada).
Detection of cytokine mRNA by RT-PCR
The levels of transcripts encoding human TNF-α, IL-10, and β-actin were determined using semiquantitative RT-PCR. Total RNA was isolated using Trizol (Life Technologies, Burlington, Canada) from microglia or T cells. RNA (0.5 μg) was reverse transcribed and amplified in a single-step process as previously described (1). The following sequence of primers was used in the RT-PCR experiments: TNF-α, 5′-GAGTGACAAGCCTGTAGCCCATGTTGTAGCA-3′ (sense) and 5′-GCAATGATCCCAAAGTAGACCTGCCCAGACT-3′ (antisense); IL-10, 5′-ATGCCCCAAGCTGAGAACCAAGACCCA-3′ (sense) and 5′-TCTCAAGGGGCTGGGTCAGCTATCCCA-3′ (antisense); and β-actin, 5′-GCCCTGGACACCAACTATTGC-3′ (sense) and 5′-GCTGCACTTGCAGGAGCGCAC-3′ (antisense). Thirty-five cycles of amplification were used for TNF-α and IL-10 transcripts, and 25 cycles were used for β-actin; these were in the linear range of amplification. cDNA products were run on a 1.5% agarose gel containing ethidium bromide and were visualized under UV light. The identity of the PCR products was confirmed by purifying and sequencing the products; sequence analysis was performed by BLAST search.
Cytokine protein quantification
TNF-α and IL-10 protein levels in the conditioned medium of microglia-T cell cocultures were measured using ELISA kits from BioSource International (Montreal, Canada). Assays were performed following detailed instructions by the manufacturer. Unless otherwise stated, all conditioned media were collected after 24 h of microglia-T cell cocultures.
For intracellular staining of IL-10, cells were treated the last 4 h of culture with Golgi Stop, a protein transport inhibitor, obtained from PharMingen Canada (Mississauge, Canada). At the end of the culture period, activated T cells and microglia were collected for flow cytometric analyses. Cells were stained with primary Abs, anti-CD3 IgG1-FITC in the case of T cells and anti-CD14 IgG1-FITC for the staining of microglia, or with IgG1-FITC isotype control for a period of 30 min at 4°C. Cells were then washed twice with PBS containing 3% FCS. To allow intracellular staining to occur, cells were fixed into 100 μl of Cytofix/Cytoperm solution (PharMingen Canada) for 20 min at 4°C. After two washes using Perm/Wash solution (PharMingen Canada) to maintain cell permeability, cells were stained with anti-IL-10-IgG2a-PE or with an appropriate isotype control for 30 min at 4°C. Staining was analyzed by flow cytometry using an argon laser FACS equipped with CONSORT 30 and LYSYS II software (Becton Dickinson); data were collected for 15,000 cells/condition.
In another series of experiments, T cells were stained using primary Abs (anti-CD40L IgG1-PE, anti-CD28-IgG1-PE, anti-CTLA-4 IgG2a-PE, or appropriate IgG isotype controls) for a period of 30 min at 4°C. They were then washed twice and resuspended in PBS before staining was analyzed by flow cytometry.
Live microglia cells were seeded in 16-well Lab-Tek (Life Technologies) chambers and were incubated with mouse anti-human CD40, mouse anti-human CD80, or mouse anti-human CD23 (5 μg/ml each) for a period of 1 h at room temperature or with the diluting medium of the Ab as a control. Cells were then washed in PBS followed by an incubation for 1 h with goat anti-mouse rhodamine (10 μg/ml), fixed for 10 min with 4% paraformaldehyde, and viewed using an immunofluorescence microscope.
Since all experiments involved multiple groups, statistical analyses (compared with controls) were conducted using one-way ANOVA with Bonferroni’s post-ANOVA comparisons.
IL-10 is produced in microglia-T cell coculture
Microglia or T cells in isolation secrete negligible amounts of IL-10 into the conditioned medium. In contrast, their coculture resulted in significant levels of IL-10 (Fig. 1 A). As previously reported (1), TNF-α was also produced in microglia-T cell cocultures and was assayed so as to serve as a positive control for microglia-T cell interaction.
A time-course assay was performed to determine the temporal production of IL-10 compared with TNF-α in human microglia-T cell cocultures (Fig. 1 B). TNF-α levels were elevated by 4 h after microglia and T cells were cocultured and became significantly elevated (p < 0.001) by 6 h after coculture compared with microglia alone. On the other hand, IL-10 levels, which was first detected 6 h after coculture, did not become elevated above control levels until 24 h after coculture (p < 0.001). Thereafter, the production of both cytokines reached levels of saturation.
Microglia and T cells both produce the IL-10 generated from their interaction
Since T cells and microglia (29) are both potential producers of IL-10, semiquantitative RT-PCR was used to examine levels of mRNA for IL-10 in T cells and microglia. Loosely adherent T cells were separated from microglia by several washes of culture medium and were collected as previously described (1). The removal of T cells was verified by microscopy. Total RNA from the T cells and the adherent microglia was collected 6 h after the cells were cocultured, a time point at which T cells remain loosely adherent and could be separated from adherent microglia and at which point levels of IL-10 mRNA become elevated (even though protein levels only rise several hours later). Results obtained from RT-PCR analyses (n = 3) confirmed the results shown in Fig. 2, since they revealed that both cell types produced IL-10 following their coculture (data not shown).
To resolve at the single cell level the source of IL-10 generated in microglia-T cell cocultures, intracellular staining for IL-10 was performed. T cells (CD3+) alone (Fig. 2,A) and microglia (CD14+) alone (Fig. 2,B) did not stain positive for IL-10 supporting the ELISA results shown in Fig. 1,A. However, when cells were cocultured, both cell types were found to be positive for IL-10. In the case of T cells, 23% of CD3+ cells present in the coculture were positive for IL-10 staining (Fig. 2,C). On the other hand, most microglia were found to produce IL-10, since 89% of CD14+ cells were positive for IL-10 staining (Fig. 2 D). No Ab cross-reactivity was detected, since all IgG isotype control stains were negative (data not shown).
The level of IL-10 in microglia-T cell coculture is not dependent on TNF-α
Because the increase in TNF-α resulting from the interaction of microglia and activated T cells occurs before that in IL-10 (Fig. 1,B), and TNF-α was shown to enhance the production of IL-10 in human monocytes (30), we investigated whether the production of IL-10 observed in microglia-T cell cocultures was dependent on TNF-α. TNF-α is initially produced as a 26-kDa pro form and is converted to its 17-kDa secreted form by TACE, a member of the adamylysin subfamily of metalloproteinase (31). The TACE inhibitor BB-94 has been shown to inhibit TNF-α secretion (32). Two approaches were used to test whether TNF-α was responsible for the production of IL-10 in microglia-T cell cocultures. First, TNF-α secretion was blocked using BB-94 (10 μm); second, the neutralization of both secreted and membrane-associated TNF-α was achieved using an Ab against TNF-α (5 μg/ml). Treatments with both BB-94 and anti-TNF-α did not affect the level of IL-10 secreted, suggesting that TNF-α is not responsible for the production of IL-10 in microglia-T cell cocultures (Fig. 3,A). As expected, BB-94 completely inhibited TNF-α secretion, but did not inhibit TNF-α mRNA transcript levels (Fig. 3,B), confirming that its effect on TNF-α secretion is not due to nonspecific cytotoxicity. As a negative control, TIMP1, a natural inhibitor of matrix metalloproteinases (33) with no activity on TACE, did not affect the level of TNF-α or IL-10 secreted into the culture medium (Fig. 3 A).
LPS is a potent inducer of TNF-α and IL-10 production (6). LPS was used to enhance IL-10 and TNF-α levels in microglia-T cell cocultures. Again, BB-94 in LPS-treated cells completely blocked TNF-α secretion, but did not affect IL-10 levels (Fig. 3 A). Collectively, these results demonstrate that the level of IL-10 is not dependent on TNF-α.
IL-10 generated from microglia-T cell-culture is dependent on a cell-to-cell contact mechanism
To ascertain whether the increase in IL-10 in microglia-T cell cocultures was due to soluble factors or cell contact interactions, the conditioned medium collected from cultures of activated T cells was added to microglia. Under this condition, IL-10 protein was not detected by ELISA (Fig. 4 A), suggesting that soluble factors play a minor role, if any, in the induction of IL-10 production in microglia-T cell cocultures. These data were supported by cell culture insert experiments, in which activated T cells were placed in a culture insert (Becton Dickinson, Bedford, MA) and incubated in close proximity but not contacting the microglia. No IL-10 was generated under this condition. IL-10 was produced only when the two cell types were allowed to contact each other, suggesting that a contact-dependent mechanism is involved in the production of IL-10 in microglia-T cell cocultures. Although not formally evaluated in this study, it is unlikely that IL-10 production generated from this allogeneic interaction between microglia and T cells is MHC restricted, since levels of TNF-α generated from both microglia-T cells allogeneic and syngeneic interactions were shown to be similar in our previous study (1).
It has been reported that the production of TNF-α generated from microglia-T cell interactions is partly dependent on the VLA-4/VCAM-1 interaction (1). In this study we confirm the involvement of VLA-4 in the generation of TNF-α in microglia-T cell coculture, since an Ab to the α-chain of VLA-4, anti-CD49d, decreased TNF-α levels (Fig. 3,B). In contrast, anti-CD49d did not affect IL-10 levels, suggesting that IL-10 production is not VLA-4 dependent (Fig. 4 B).
Anti-CD40, CTLA-4-Fc, and CD23 decrease levels of IL-10 in microglia-T cells cocultures
Given that the production of IL-10 in microglia-T cell interaction is cell contact dependent (Fig. 4 A), we sought to elucidate the identity of the cell surface molecules involved. We focused on the CD40, CTLA-4, and CD23 pathways, since the respective ligand-receptor pairs are found on microglia and T cells (see below). First, the contribution of CD40/CD40L interactions, which are known to play a crucial role in macrophage-T cell interactions (34), was studied. CD40 is a molecule expressed on macrophages, B cells, dendritic cells, and endothelial cells (35); recently, murine microglia were also found to express CD40 (35). On the other hand, CD40L (or CD154) is expressed on CD4+ T cells and, to a lesser extent, on CD8+ T cells; NK cells can also express CD40L (34).
Flow cytometric analysis confirmed the presence of CD40L on the surface of T cells (20 ± 3% of the total T cell population) 72 h after their activation. CD40L was expressed by CD4+ T cells (72 ± 9% of CD40L-positive cells) and, to a lesser extent, on CD8+ T cells (40 ± 11% of CD40L-positive cells; mean of three experiments involving three different blood donors). The presence of CD40 on the surface of microglia was confirmed by immunocytochemistry (Fig. 6 B).
Whether CD40/CD40L interaction had a role in IL-10 production in microglia-T cell coculture was examined by the treatment of microglia with an Ab against CD40. Fig. 5 A demonstrates that anti-CD40 inhibited the levels of IL-10 detected in the culture medium in a concentration-dependent manner. This inhibition was not specific to IL-10, since the production of TNF-α was also inhibited (data not shown). As a negative control to verify that the anti-CD40 effect was not due to cytotoxicity, LPS-treated microglia were treated with anti-CD40 (10 μg/ml). The LPS-induced increased level of IL-10 was not affected by anti-CD40 treatment (data not shown). Moreover, microglia treatment with an IgG1 isotype control did not affect levels of IL-10 generated in microglia-T cell cocultures (data not shown).
Next, we examined the possible role of the costimulatory pathway, CD28-CTLA-4/B7, in the production of IL-10 resulting from microglia-T cell interactions. CD28 has been reported to be constitutively expressed on 80% of CD4+ T cells and 50% of CD8+ T cells and becomes up-regulated following T cell activation (36). On the other hand, CTLA-4 is found at very low levels on resting T cells, and its expression is up-regulated after activation of T cells. CD28 and CTLA-4 share the same receptors, the B7 molecules (B7-1 and B7-2). Binding through CD28 provides a positive signal for T cell activation, while CTLA-4-dependent interactions lead to the inhibition of T cell functions. CTLA-4 has a higher affinity and avidity for B7 molecules than CD28, and its expression is up-regulated as cell activation progresses to act as an inhibitor of T cell activation (37, 38). The presence of B7 molecules on human microglia has been shown in vitro (27) as well as in vivo on activated microglia and infiltrating macrophages within active MS lesions (39). B7-2 appears to be expressed constitutively on human microglia and is thought to play a role in the initiation phase of the inflammation (40). In contrast, B7-1 is expressed at low levels and is believed to be involved in the progression of inflammatory responses, as it becomes up-regulated during inflammatory conditions such as MS (41).
We confirmed the expression of CTLA-4 and CD28 on activated T cells by flow cytometry. CTLA-4 (14 ± 2.9% of the total cell population) was equally expressed by CD4+ (60 ± 6% of CTLA-4-positive cells) and CD8+ T cells (51 ± 9% of CTLA-4-positive cells; mean of four experiments involving four different blood donors). After 72 h of activation, CD28 was expressed by 77 ± 7.2% of the total cell population, and CD4+ T cells constituted 65 ± 3% of all CD28-positive cells, while 34 ± 0.5% of CD28-positive cells were CD8+ T cells (mean of three different experiments involving three different blood donors). The presence of the CD28/CTLA-4 receptor, B7-1 (CD80), on the surface of human microglia was confirmed by immunocytochemistry (Fig. 6 C). The constitutive expression of B7-2 (CD86) was previously reported by others (40).
To test whether the CD28-CTLA-4/B7 interaction plays a role in the production of IL-10 in microglia-T cell cocultures, microglia were treated with various concentrations of recombinant human CTLA-4-Fc chimera protein, which binds both B7-1 and B7-2 with high affinity. Fig. 5,B shows that IL-10 levels in T cell-microglia cocultures were reduced in a concentration-dependent manner by treatment with CTLA-4-Fc. Levels of TNF-α were also inhibited by CTLA-4-Fc treatment (see Fig. 8). The CTLA-4-Fc reduction of IL-10 levels is unlikely to be the result of nonspecific cytotoxicity, since CTLA-4-Fc did not affect the IL-10 up-regulation in LPS-treated microglia (data not shown).
CD23 is a molecule that plays an important role in allergy and inflammation. It is the low affinity IgE Fc receptor (FcεRII) expressed on monocytes/macrophages, but it is also viewed as an adhesion molecule because of its ability to interact with CD21 on B cells and with CD11b or CD11c on activated T cells (42). Since human microglia were found to be positive for CD23 expression (Fig. 6,D), the role played by CD23 in the production of IL-10 generated in T cell-microglia coculture was investigated; also, the cross-linking of CD23 on the surface of macrophages has been shown to induce the production of IL-10 through a mechanism dependent on cAMP (43). CD23-dependent interactions were blocked by treating microglia with anti-CD23 (10 μg/ml); this inhibited levels of IL-10 generated in microglia-T cell interactions. Importantly, the inhibition by anti-CD23 was specific for IL-10, since TNF-α levels were not affected (Fig. 5 C). Treatment with anti-CD23 on microglia alone did not affect its cytokine levels or its morphology.
CD4+ and CD8+ T cells are both responsible for the production of IL-10 and TNF-α in microglia-T cell cocultures
To determine whether CD4+ or CD8+ T cells were responsible for IL-10 production when cocultured with microglia, purification of T cell subpopulations was performed using magnetic beads. Coculture of microglia with either purified CD4+ or CD8+ cells triggered the production of IL-10 (Fig. 5,D). Moreover, treatment with anti-CD40 (5 μg/ml), CTLA-4-Fc (5 μg/ml), or anti-CD23 (5 μg/ml) had a similar inhibitory effect on CD4+- and CD8+-dependent IL-10 production (Fig. 5 D).
Morphological changes in microglia-T cell coculture
Microglia acquire various morphology in vitro as well as in vivo. In general, cultured resting human microglia tend to be bipolar (elongated) or ramified (Fig. 7,A) and become ameboid (rounded) when they are activated; in a previous study (27) we provided micrograph of various morphologies of human adult microglia in culture. On the other hand, in culture, activated T cells are found as single cells or as homotypic aggregates (Fig. 7,B). When both microglia and T cells were cultured together, aggregates of T cells were found attached to the microglia and bipolar, i.e., ramified microglia became ameboid in shape (Fig. 7,C). However, in the presence of anti-CD40 or CTLA-4-Fc, microglia retained their bipolar/ramified morphology even though T cells were still adherent on microglia (Fig. 7,D). It is noteworthy, though, that fewer T cells were clustered (i.e., activated) around microglia cells in cultures treated with anti-CD40 or CTLA-4-Fc (Fig. 7 D) compared with similar cultures in the absence of these inhibitors. This is probably the result of a decrease in costimulatory processes necessary for further T cell activation in which B7 and CD40 play important roles. Taken together, these morphological results confirm the cytokine data showing that microglia become activated when in contact with T cells, but that this activation is attenuated by anti-CD40 or CTLA-4-Fc.
In contrast to anti-CD40 or CTLA-4-Fc, the anti-CD23 Ab, which reduced IL-10 but not TNF-α, did not fully prevent the ameboid transformation of microglia in contact with T cells. Indeed, a range of morphology from ramified to ameboid was observed (results not shown) in T cell-microglia cultures.
Combined treatment with anti-CD40 and CTLA-4-Fc augments the inhibition of both IL-10 and TNF-α production in microglia-T cell cocultures
The combination of anti-CD40 (5 μg/ml) and CTLA-4-Fc (5 μg/ml) treatments augmented the activity of each to inhibit the production of IL-10 (Fig. 8,A). Indeed, the level of IL-10 in the culture medium approached the negligible amount found in control microglia culture. Additional blockage with anti-CD23 (i.e., anti-CD40, CTLA-4-Fc, and anti-CD23) did not further reduce the production of IL-10 (Fig. 8 A).
The TNF-α level in microglia-T cells cocultures was also significantly blocked in a combinational manner by the coadministration of anti-CD40 (5 μg/ml) and CTLA-4-Fc (5 μg/ml; Fig. 8 B).
IL-10 is recognized as a potent anti-inflammatory cytokine due to its ability to inhibit the production of proinflammatory cytokines and inflammatory mediators (8, 12), Ag presentation (9, 14), Th1 differentiation, T cell activation (13), and the production of specific Ab (44). Thus, it is likely that the production of IL-10 is critical for shutting down inflammatory reactions involved in chronic neuroinflammatory diseases such as MS.
The mechanisms involved in the regulation of IL-10 expression are not very well understood, although recombinant HIV-1 Nef protein, rIFN-β, and LPS are known to be inducers of IL-10 production (6, 45). Furthermore, the production of IL-10 by macrophages appears to be induced through the Fcγ receptor (46).
This report investigates novel mechanisms (Fig. 9) by which IL-10 may be generated, particularly in the context of the CNS. Activated T lymphocytes infiltrate the CNS during neuroinflammation and are then found in close proximity to the microglia. We show that the interaction of microglia with T cells leads to the production of IL-10, and that blockade of the CD40/CD40L, CD28-CTLA-4/B7, and CD23 pathways results in the inhibition of IL-10 levels, suggesting that these pathways play a role in the anti-inflammatory response. Importantly, combinational blockade of the CD40/CD40L and CD28-CTLA-4/B7 pathways reduced IL-10 production by microglia-T cell interactions almost down to the negligible levels seen with microglia or T cells in isolation, highlighting the important contributions of these two pathways in regulating IL-10 levels. While anti-CD23 also reduced IL-10 levels, its addition to the anti-CD40 and CTLA-4-Fc combination did not further augment the effect of the latter.
Some selectivity of ligand-receptor pairs in microglia-T cell interactions was revealed by the results of this study. While the CD40/CD40L and CD28-CTLA-4/B7 pathways regulate both IL-10 and TNF-α, the VLA-4/VCAM-1 interaction was specific for TNF-α; in contrast, the CD23 system affected IL-10, but not TNF-α (Fig. 9).
It is well established that the interaction of CD40 with its ligand CD40L plays an important role during inflammation and cell-mediated immunity. Of relevance to neuroinflammation, CD40 expression was elevated in the brains of MS patients and in mice undergoing chronic EAE, and this elevation correlated with disease activity, suggesting that CD40/CD40L interactions may play a role in the pathogenesis of these diseases (47, 48). The interaction of CD40L with CD40 has been shown to induce the production of cytokines such as TNF-α and IL-12. In addition, Stout et al. (49) reported that T cells isolated from CD40L-deficient mice fail to induce macrophages to produce TNF-α. While these studies have shown that the CD40/CD40L interaction plays an important role in the proinflammatory process, its role during the anti-inflammatory or Th2-type response has not been well characterized. The results of this study provide the first direct evidence that CD40/CD40L interaction plays a role during the anti-inflammatory response by regulating IL-10 production.
This study also demonstrates a role for the CD28-CTLA-4/B7 pathway in regulating IL-10 production in microglia-T cell coculture, since inhibition of the B7-dependent interactions leads to a decrease in IL-10 production. Other laboratories have provided evidence for the CD28-CTLA-4/B7 pathway in the regulation of IL-10 levels by other cell types. First, blockade of this pathway using CTLA-4-Fc was shown to inhibit the in vivo production of IL-10 from activated lung CD3+ T cells by 70–80% (50). Second, the production of IL-10 in vitro by anti-CD3-activated CD4+ T cells was shown to occur only when CD28 and CD40L were cross-linked simultaneously (51).
A specific role for CD23 in the production of IL-10 generated from microglia-T cell interactions is also suggested by the results of this study, since anti-CD23 treatment specifically inhibited IL-10, but not TNF-α. This study did not address the nature of the ligands for CD23, but CD11b and CD11c are obvious candidates, since they are found on activated T cells. The present report is the first to show that CD23 is expressed by cells of the CNS, namely the microglia, which suggests a novel role for CD23 in the regulation of immune functions of the CNS.
As TNF-α and IL-10 are both generated in response to microglia-T cell interactions, and given that there is selectivity in the ligand-receptor pairs in regulating their expression as the results of this study indicate, it is of interest to determine whether the production of TNF-α and IL-10 can be selectively regulated. In the context of MS the elevated secretion of IL-10 is probably beneficial given its anti-inflammatory role while the generation of TNF-α may exert a deleterious effect given that this is a proinflammatory cytokine that can also induce apoptosis of oligodendrocytes (2, 3, 4). Thus, it is of interest that the clinically useful MS drug, IFN-β, decreases TNF-α levels in microglia-T cell interaction (1), but elevates IL-10 in the same coculture system (S. Chabot and V. W. Yong, unpublished observations). Whether IFN-β differentially affects the CD40, B7, CD23, or VLA4 pathways is being investigated.
In summary, the results of this study demonstrate that IL-10 is produced as a consequence of direct microglia-T cell interaction, an observation that is relevant to the regulation of an anti-inflammatory response within the CNS.
We thank Fiona Yong for her computer skills in generating many of the figures and Lori Robertson from the Flow Cytometry Laboratory for her technical expertise. We thank Tanna Lowe for secretarial assistance.
This work was supported by Medical Services, Inc. (Alberta, Canada). S.C. is supported by a studentship from the Multiple Sclerosis Society of Canada. V.W.Y. is a Senior Scholar of the Alberta Heritage Foundation for Medical Research and a Scientist of the Medical Research Council of Canada.
Abbreviations used in this paper: MS, multiple sclerosis; EAE, experimental allergic encephalomyelitis; VLA-4, very late Ag-4; MNC, mononuclear cells; TACE, TNF-α-converting enzyme; CD40L, CD40 ligand.