T cell recognition of rat myelin basic protein as a TCR antagonist inhibits reciprocal activation of antigen-presenting cells and engenders resistance to experimental autoimmune encephalomyelitis

2.1 The self-antigen RMBP inhibits antigenic activation of 3H3 T cells

Previous studies have shown that RMBP antagonized GPMBP-induced expression of B7 molecules on 3H3 T cells 7. TCR antagonism rather than competition for the MHC glycoprotein peptide-binding groove accounted for the inhibitory action of RMBP. This conclusion was based on the observation that RMBP antagonized antigen-induced IL-2 production (Fig. 1A) and B7.1 expression (Fig. 1B) by MBP-specific 3H3 T cells but not by conalbumin-specific 9H8 T cells even though both 3H3 and 9H8 T cells were I-A restricted (Fig. 1C).

2.2 Antagonism of 3H3 T cells results in the generation of tolerogenic APC

Previous studies have indicated that recognition of the partial agonist RMBP by 3H3 T cells may differentially regulate expression of MHC class II glycoproteins and costimulatory molecules on T cells to generate an I-A⁺, B7^low phenotype 7 and possibly non-activated tolerogenic APC. To test whether recognition of RMBP as a TCR antagonist promotes regulatory activity in EAE, 3H3 T cells were cultured with irradiated splenocytes (irrSPL) for 2 days in serum-free medium in the presence of RMBP. Adoptive transfer of these cells into naive Lewis rats resulted in reduction in the intensity of EAE when challenged 1 week later with GPMBP in CFA (Table 1). Reduction in disease severity was dependent on the presence of RMBP because equivalent cultures lacking RMBP did not exhibit tolerogenic activity.

2.3 T cell recognition of RMBP as a partial agonist negatively regulates antigenic activation of splenic APC

T cell-dependent activation of B cell and MΦ APC is critical for humoral and cell-mediated effector responses. Experiments were therefore performed to determine whether the feedback activation of splenic APC was regulated by RMBP-mediated antagonism. Splenic APC were approximately 60% I-A⁺ (HIS-28), 20% ED3⁺ (MΦ subset marker), 30–35% OX41⁺ I-A⁺/OX42⁺ I-A⁺ (MΦ markers) and 10% OX33⁺ I-A⁺ (B cell marker). During a 2-day culture with splenic APC and 3H3 T cells, a relatively low concentration of GPMBP (0.10 μM) induced a substantial loss of I-A molecules from the APC (Fig. 2). A relatively high concentration of RMBP (3.2 μM) did not affect quantitative expression of I-A on splenic APC, whereas the same concentration of RMBP effectively antagonized GPMBP-catalyzed loss of I-A from splenic APC. Similarly, RMBP antagonized GPMBP-induced expression of B7.1 and B7.2 molecules on splenic APC (Fig. 2). Induction and antagonism of B7.1 expression was primarily apparent on OX41⁺ and OX42⁺ APC (data not shown). These results indicate that TCR antagonism regulates B7 expression on splenic APC and may thereby control the extent of an immune response via provision of signal 1 (MHC/TCR ligation) without signal 2 (B7/CD28 costimulation).

2.4 TCR antagonism inhibits agonist-induced factors that promote survival of MΦ 

In 3-day cultures of 3H3 T cells and irrSPL, GPMBP not only stimulated T cell activation but also resulted in increased yields of splenic APC (Fig. 3). Thus, splenic APC represented a greater percentage of the cell population in cultures of activated T cells than in non-stimulated cultures. The mechanism underlying enhanced yields of splenic APC (mostly MΦ) in antigen-stimulated cultures is not currently known but may reflect antigen-induced elaboration of M-CSF and increased survival or perhaps proliferation of MΦ. Like inhibition of B7.1 and B7.2 expression, RMBP strongly antagonized GPMBP-induced increases in the yield of splenic APC (Fig. 3, bottom left).

Two approaches were used to assess mechanisms responsible for antigen-induced enrichment of splenic APC. First, 3H3 T cells were cultured for 2 days with splenic APC in the presence of no antigen, RMBP, GPMBP, or both antigens. Supernatants from these cultures were then tested for the maintenance of PEC MΦ (Fig. 3, bottom right). Culture supernatants from GPMBP-stimulated cultures maintained high levels of MΦ viability, whereas supernatants from T cells cultured with no antigen, RMBP, or both antigens exhibited reduced activity. These data indicate that the balance of agonist versus antagonist directly regulated production of soluble factors that promote MΦ survival. Second, counts of total viable cells recovered from culture were multiplied by percentages of PKH26^– APC assessed by flow cytometry. These analyses revealed that GPMBP engendered an approximate doubling of the number of viable splenic APC recovered from culture (data not shown). These data indicate that TCR antagonism may directly affect accumulation of effector cells such as MΦ APC.

2.5 The self-antigen RMBP is a partial agonist that antagonizes MΦ production of NO

To test whether RMBP regulates an MΦ effector response implicated in EAE, we focused on T cell-dependent NO production (Fig. 4). Splenic MΦ were isolated by plastic adherence, were rested for 7 days in a source of M-CSF, and then were pulsed with designated concentrations of GPMBP for 45 min. 3H3 T cells were cultured for 2 days with GPMBP-pulsed splenic MΦ in the presence of designated concentrations of RMBP. MΦ pulsed with GPMBP concentrations as low as 10 nM were sufficient to stimulate NO production. This response was entirely dependent upon the presence of T cells (data not shown). These data indicated that the NO production assay was a very sensitive assay for measurement of T cell-dependent MΦ activation. RMBP behaved as a classical mixed agonist/antagonist in NO production assays. This finding provides unique evidence that TCR antagonists directly regulate T cell-dependent MΦ responses.

4.1 Animals and reagents

Lewis rats (Harlan-Sprague Dawley, Indianapolis, IN) were maintained at East Carolina University School of Medicine. MBP was purified from rat or guinea pig spinal cords (Rockland). The anti-I-A OX6 IgG1 8, anti-SIRP OX41 IgG1 9, anti-CD11 OX42 IgG1 9, anti-rat MΦ ED3 IgG2a 10, and anti-B cell CD45 OX33 IgG1 11 mAb were produced as culture supernatants and were concentrated by ultrafiltration through Amicon spiral wound membranes (100-kDa exclusion). Anti-B7.1 IgG1 (3H5), anti-B7.2 IgG1 (24F), and anti-RT1B IgG2a (HIS28) were purchased from PharMingen. Aminoguanidine and conalbumin were from Sigma.

4.2 Derivation and culture of Lewis rat T cell lines

The CD4^– subclone GP2.3H3.16 (referred to as 3H3) was derived from Lewis rats sensitized with an emulsion of GPMBP in CFA 2. The CONAL.9H8 clone was derived from Lewis rats immunized with conalbumin in CFA. Clones were maintained in complete RPMI medium (cRPMI) supplemented with IL-2 (CM). The cRPMI medium (cRPMI) consisted of 10% heat-inactivated fetal bovineserum (Summit, Boulder, CO), 2 mM glutamine, 100 μg/ml streptomycin, 100 U/ml penicillin (Whittaker Bioproducts, Walkersville, MD), and 50 μM 2-ME (Sigma). Rat IL-2 was expressed from a recombinant rat IL-2 baculovirus.

4.3 Preparation of splenic APC

SPL were obtained from naïve Lewis rats and were purified on discontinuous Percoll gradients of 40%, 50%, and 60%. The splenic APC cell fraction was collected at the 40/50% interface and was largely comprised of cells expressing high levels of I-A, B7.1, and B7.2.

4.4 In vitro proliferation, IL-2 production, and NO production

Responder T cells (2.5×10⁴/well) were cultured with irrSPL (5×10⁵/well) or splenic APC (10⁵/well) with or without antigen in cRPMI and were pulsed with 1 μCi [³H]thymidine (6.7 Ci/mmol, NEN) during the last day of a 2–3-day assay. Supernatants (100–120 μl/well) were transferred to replicate plates to measure IL-2 bioactivity, and T cells were harvested onto filters to measure [³H]thymidine incorporation by scintillation counting (Wallac). CTLL cells (10⁴/50 μl cRPMI/well) were cultured with 100–120 μl of supernatant for 24–48 h, and 10 μl of a 2.9 mg/ml 3-(4,5-dimethylthiazol-2-yl)-5-3 (3-carboxy-methoxyphenyl). 2-(4-sulfophenyl)-2H-tetrazolium (MTS) (Promega) and 0.1 mg/ml phenazine methosulfate (PMS, Sigma) (MTS/PMS) solution were added to each well. Plates were read the next day at 492 nm on an Anthos ELISA Reader (ACCSales, Chapel Hill, NC). Antigen-stimulated IL-2 production was calculated as the mean absorbance from experimental cultures minus the mean absorbance from control unstimulated cultures. The production of NO was measured by formation of the stable decomposition product nitrite in cell-free supernatants (50 μl) after mixing with an equal volume of the Griess reagent 12.

4.5 Flow cytometric analysis

3H3 or 9H8 T cells were labeled with the PKH26 lipophilic dye (Sigma) according to manufacturer's directions. Splenic APC (5×10⁵/ml) or irrSPL (5×10⁶/ml) and PKH26-labeled T cells (5×10⁵/ml) were cultured for 2–3 days in cRPMI. T cells were stained with a primary mAb (2.5 μg/ml or 1/20 titration of a concentrated supernatant) and an FITC-conjugated goat anti-mouse IgG (H+L). Washing and staining of T cells were performed at 4 °C, and normal rat serum was used to block Fc receptors. Dead cells were excluded from analysis by forward versus side scatter profiles. Data were acquired with a Becton Dickinson FACScan cytometer and were analyzed with the CellQuest software program.

4.6 Assay of M-CSF-like activity

3H3 T cells and splenic APC were cultured for 2 days with no antigen, GPMBP, RMBP, or both antigens. Supernatants from these cultures were tested for M-CSF-like activity in a subsequent culture of peritoneal exudate cell (PEC) MΦ indicator cells. To generate indicator cells for the bioassay, PEC were collected from the peritoneal cavity 2 days after i.p. injection of heat-killed C. parvum (200 μg/5 ml HBSS) into naïve Lewis rats. After 5 days of culturing MΦ with T cell supernatants, 10 μl of an MTS/PMS solution were added to each well to assess MΦ viability. Plates were read after 24 h at 492 nm on an ELISA Reader. Metabolic activity of surviving MΦ was determined as mean absorbance values. Production of MCSF-like cytokines was determined asmean absorbance for experimental cultures minus mean absorbance for control non-stimulated cultures.

4.7 Clinical assessment of EAE

The following scale was used to assign intensity of EAE; paralysis in the distal tail, 0.25; limp tail, 0.5; ataxia, 1.0; hind leg paresis, 2.0; full hind leg paralysis, 3.0. The mean maximal intensity scores were assigned to each group based on the average maximum score among afflicted rats within a group. The cumulative score for each rat consisted of the sum of daily scores for each rat. The mean cumulative score for a group was calculated by averaging the cumulative scores for all rats within a group.