Role of the Stress Kinase Pathway in Signaling Via the T Cell Costimulatory Receptor 4-1BB

Mice, cell lines, Abs, and reagents

CD28^−/− mice backcrossed onto the C57BL/6 (H-2b) background (n = 10) were obtained from Dr. Tak Mak (Amgen Institute, Toronto, Ontario, Canada) (5). C57BL/6 mice were obtained from Charles River Laboratories (St. Constant, Quebec, Canada). CD28^−/− or C57BL/6 mice were used at 8–10 wk of age.

The autoreactive T cell hybrid, C8.A3, was originally obtained from Dr. Laurie Glimcher (Harvard Medical School, Boston, MA). This T cell hybridoma responds to A^k complexed to an unidentified peptide. Although this T hybridoma can respond to anti-CD3 alone, the response of C8.A3 T cells to MHC II/peptide requires costimulation (42, 43). The BALB/c B cell lymphoma, K46J (44), expresses 4-1BB ligand constitutively, undetectable levels of B7-1, and low levels of B7-2 (43). K46J73.35 was obtained by transfecting the K46J parental line with truncated α- and β-chains of A^k (45). C8.A3 cells respond to K46J73.35 without added Ag, but fail to respond to the parental K46J cells, which are used in these studies as a negative control. Cells were maintained in RPMI 1640 containing 10% FCS (Cansera, Rexdale Ontario, Canada), 50 μM 2-ME, MEM nonessential amino acids (Life Technologies, Gaithersburg, MD), antibiotics, pyruvate, and l-glutamine, as previously described (46).

The anti-CD3-producing hybridoma 145-2C11 (47) was obtained from Dr. Jeff Bluestone (University of Chicago, Chicago, IL). The 12CA5 hybridoma (48) specific for influenza hemagglutinin (HA) was obtained from Dr. Bob Phillips (University of Toronto, Toronto, Ontario, Canada). The hybridomas N418 (anti-CD11c), Y-3P (anti-A^b), RA3-6B2 (anti-B220), TIB-128 (anti-MAC-1), M1/69 (anti-heat-stable Ag), and RG7/7.6H2 (anti-rat Ig κ-chain) were obtained from the American Type Culture Collection (Manassas, VA). The ICAM-1-specific YN-1 rat-mouse hybridoma was kindly provided by Fumeo Takei (University of British Columbia, Vancouver, B.C., Canada). The anti-CD28-secreting hybridoma 37.51.1 (49) was provided by Dr. J. Allison (University of California, Berkeley, CA). Abs were purified from hybridoma supernatants using protein G- or protein A-Sepharose (Pharmacia Biotech, Piscataway, NJ), according to the manufacturer’s instructions. 3T3 cells secreting 4-1BB linked to alkaline phosphatase (AP) were provided by Dr. Byoung Kwon (Indiana University, Indianapolis, IN). 4-1BB.AP was purified on anti-AP-Sepharose, as previously described (50). AP from human placenta was obtained from Sigma (St. Louis, MO). The generation and purification of recombinant soluble 4-1BB ligand (s4-1BBL) from baculovirus-infected insect cells were previously described (11).

TNF-α and monoclonal anti-4-1BB (1AH2, a rat IgG) were purchased from PharMingen (SanDiego, CA). The anti-TRAF2 Ab and the anti-JNK1 Ab (used in JNK/SAPK in vitro kinase assay) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-JNK/SAPK and anti-phospho-JNK/SAPK Thr¹⁸³/Tyr¹⁸⁵ were purchased from New England Biolabs (Beverly, MA). Goat anti-rabbit Ig HRP, goat anti-mouse Ig HRP, and sorbitol were purchased from Sigma.

Expression vectors and transfections

HA epitope-tagged expression plasmids for ASK-1 and catalytically inactive ASK-1 K709E have been previously described (31). C8.A3 (2 × 10⁷) T cells in PBS/135 mM sucrose were transiently transfected with 50 μg of plasmid DNA at 500 V, 25 μF, 13 ohms, and 3 pulses using the BTX Electrocell Manipulator 600 (BTX, San Diego, CA). Plasmids encoding GST-SEK1 K129R (51) and GST-c-Jun (5-89) (52) bacterial fusion proteins were previously described. Fusion proteins were produced in the pLysS (BL21DE3) strain of Escherichia coli using the pGEX expression system (Promega, Madison, WI). Proteins were affinity purified on glutathione-Sepharose beads and eluted with 10 mM reduced glutathione, 50 mM Tris, pH 8. Eluted proteins were dialyzed against a buffer containing 50 mM Tris, pH 8, 150 mM NaCl, 5 mM EDTA, 0.1% Triton X-100, 0.1% 2-ME, and 50% glycerol.

Kinase assays

C8.A3 T cells were incubated in low serum-containing medium (2% FCS) overnight before performing JNK kinase assays. C8.A3 cells were treated with either TNF-α or sorbitol at 37°C. C8.A3 cells were incubated with anti-4-1BB (1AH2) or anti-ICAM-1 (YN-1) Abs for 5 min at 4°C, and the receptors were aggregated with anti-rat Ig at 37°C. Cells were lysed with ice-cold lysis buffer (10 mM NaCl, 20 mM PIPES, pH 7, 0.5% Nonidet P-40, 5 mM EDTA, 0.05% 2-ME, 100 μM Na₃VO₄, 50 mM NaF, 20 μg/ml leupeptin, and 1 mM benzamidine). Lysates were adjusted to equal protein concentrations by the method of Lowry. Lysates were immunoprecipitated with either anti-JNK (SAPK/JNK assay) or anti-HA (for ASK-1 assay) (2 h rotating at 4°C). Immune complexes were collected with 30 μl protein A-Sepharose beads (2 h rotating at 4°C). Immune complexes were washed three times with PBS-Triton (PBST) buffer (150 mM NaCl, 16 mM Na₂HPO₄, 4 mM NaH₂PO₄, 0.1% Triton X-100, 100 mM Na₃VO₄, 50 mM NaF, 20 μg/ml leupeptin, and 1 mM benzamidine). Kinase assays were performed in 20 μl of kinase buffer (10 μM ATP, 10 mM MgCl₂, 50 mM Tris-Cl, pH 7.5, 1 mM EGTA, pH 7.5) in the presence of 1.2 μCi [³²P]γ-ATP (Amersham Life Science, Arlington Heights, IL) and 5 μg of either GST-c-Jun (5-89) or GST-SEK1 K129R as the in vitro substrate (30°C for 30 min). The reaction was stopped by the addition of 2× SDS sample buffer. Samples were boiled and phosphoproteins were separated by SDS-PAGE and visualized by autoradiography.

Immunoprecipitation and Western blotting

C8.A3 T cells were incubated in low serum-containing medium (2% FCS) overnight before activation and lysis, which were conducted as described above for the kinase assays. Epitope-tagged proteins were immunoprecipitated by incubating cell lysates with 2 μg anti-HA Ab (12CA5). Endogenous TRAF2 and JNK were immunoprecipitated by incubating with 2 μg of anti-TRAF2 Ab or anti-JNK Ab, respectively, and harvested with 30 μl of protein A-Sepharose. The immune complexes were washed four times with PBST. Precipitated proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes (Millipore, Bedford, MA). For analysis of protein-protein interactions, immune complexes were analyzed by Western blot with the indicated Abs. Bound Abs were detected with either goat anti-rabbit Ig HRP or goat anti-murine Ig HRP by chemoluminescence, according to the manufacturer’s protocol (ECL system; Amersham Life Science).

Verification of endogenous expression of ASK-1

RNA was extracted from C8.A3 T cells using QIAGEN RNeasy Kit (Qiagen, Santa Clarita, CA). Single-stranded cDNA was synthesized from 3 μg of total RNA using the First Strand cDNA Synthesis Kit (Pharmacia Biotech). PCR was performed with the primers 5′-CG GGA TCC ATG AGC ACG GAG GCG GAC GAA GGC AT-3′ and 5′-CC ATC GAT GTA ACA TAG TAG AGA ACA TCC-3′ synthesized by Life Technologies (31) and 1 μg of template cDNA, using an initial 5-min denaturation at 94°C, followed by 35 cycles of each of the following: 94°C for 30 s, 55°C for 30 s, 72°C for 1 min, and a 72°C hold for 7 min.

Primary T cell isolation

APC were depleted from spleen cell suspensions in HBSS (Life Technologies)/2.5% FCS/50 μM 2-ME, with a mixture of Abs, anti-class II, anti-B220, anti-heat-stable Ag, (M1/69), anti-MAC-1, and anti-CD11c (N418), each at a final concentration of 10 μg/ml at 4°C for 30 min. A 1/10 dilution of baby rabbit complement (Cedarlane Labs, Hornby, Ontario, Canada) was added and the cultures were incubated at 37°C for 40 min. To remove adherent cells, the cell suspensions were passed through a Sephadex G10/nylon wool column, and then centrifuged through Percoll gradients consisting of 60, 70, and 80% Percoll layers. Small (high density) resting T cells were isolated from the 70/80% interface and used for subsequent T cell stimulation experiments. For JNK/SAPK assays, primary T cells were isolated.

T cell activation assays

For stimulation of the T cell hybrid C8.A3, monoclonal anti-CD3 (145-2C11) was immobilized on the surface of 96-well plates (Nunc, Gaithersburg, MD) by incubation in PBS overnight at 4°C. C8.A3 T cells (5 × 10⁴) were incubated with immobilized anti-CD3 (0.1 μg/ml) or with 5 × 10⁴ irradiated B cell lymphomas (K46J or K46J73.35) in the presence or absence of 4-1BB.AP or AP (10 μg/ml) overnight. Small high density T cells (1 × 10⁵) from CD28^−/− or CD28^+/+ H-2^b mice were stimulated with either immobilized anti-CD3 or anti-CD3 plus anti-CD28 or plus s4-1BBL. Where indicated, anti-CD3 and costimulatory ligands were immobilized on the wells simultaneously. After 48 h of culture, supernatants were collected and assayed for IL-2 content using the IL-2-dependent line, CTLL. CTLL proliferation was measured by adding 1 μCi of [³H]thymidine (Amersham Life Science) to the wells for the last 6 h of the 24-h stimulation. The data for the IL-2 assays are expressed as mean +/− SD of triplicate cultures.

JNK/SAPK activation after 4-1BB receptor ligation

To assess the activation of the JNK/SAPK cascade following 4-1BB aggregation, we utilized a T cell hybrid, C8.A3, which was previously demonstrated to be responsive to 4-1BB signaling. C8.A3 T cells express low levels of 4-1BB constitutively and further up-regulate 4-1BB expression following TCR ligation (43). Resting C8.A3 T cells were treated with anti-4-1BB plus second step or, as a negative control, with anti-ICAM-1 plus second step. Previous results have shown that engagement of 4-1BB by Ab without aggregation fails to induce TRAF2 recruitment; therefore, second step aggregation was used in all experiments (11). As a positive control for JNK/SAPK activation, C8.A3 T cells were treated with either sorbitol to induce osmotic shock, or TNF-α. JNK/SAPK was immunoprecipitated from the lysates, and the immune complexes were subject to Western blot analysis with either JNK/SAPK-specific or phospho-JNK/SAPK-specific Abs. Phosphorylated JNK/SAPK was detected in C8.A3 cells stimulated with sorbitol, TNF-α, or following aggregation of 4-1BB, but not following ICAM-1 aggregation (Fig. 1⇓A). JNK/SAPK protein was detected in all of the C8.A3 immunoprecipitates regardless of stimulation (Fig. 1⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

JNK/SAPK phosphorylation following 4-1BB aggregation on the surface of C8.A3 T cells. C8.A3 T cells (4 × 10⁶) were stimulated with either anti-4-1BB (5 μg/ml) or anti-ICAM (5 μg/ml) Abs for 5 min at 4°C, followed by anti-rat Ig (20 μg/ml) for 5, 10, or 20 min at 37°C. C8.A3 T cells were also separately stimulated with either TNF-α (5 ng/ml) or sorbitol (0.4 M) for 5, 10, or 20 min at 37°C. JNK/SAPK protein was isolated from the lysates by immunoprecipitation with anti-JNK/SAPK Ab and protein A-Sepharose. Immunoprecipitates (2 × 10⁶ cell equivalents/lane) were separated by SDS-PAGE and detected by immunoblotting with either A, Ab specific for phosphorylated JNK/SAPK, or B, anti-JNK/SAPK Ab. This experiment is representative of six separate experiments.

To verify that phosphorylated JNK/SAPK was functional, we used an in vitro kinase assay. C8.A3 T cells were stimulated with either TNF-α, sorbitol, anti-4-1BB, or anti-ICAM-1, as described above. JNK/SAPK was immunoprecipitated and protein kinase activity was determined in an immune complex kinase assay with GST-c-Jun as a substrate. Phosphorylated GST-c-Jun could be detected following treatment with sorbitol, TNF-α, as well as 4-1BB ligation, but not after anti-ICAM-1 treatment (Fig. 2⇓A). Fig. 2⇓B indicates that there were equivalent amounts of JNK1 present in all of the immunoprecipitates. A decline in JNK phosphorylation was not observed in Fig. 1⇑ most likely due to the different sensitivities of the assays, a reflection of the in vitro system, and the use of different Abs (see Materials and Methods).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

JNK/SAPK activation following 4-1BB aggregation on T cells. A, C8.A3 T cells (8 × 10⁶) were stimulated with either anti-4-1BB (5 μg/ml) or anti-ICAM (5 μg/ml) Abs at 4°C for 5 min, followed by anti-rat Ig (20 μg/ml) for 5, 10, or 15 min at 37°C. C8.A3 cells were also stimulated with either TNF-α (5 ng/ml) or sorbitol (0.4 M) for 5, 10, or 15 min at 37°C. JNK/SAPK was immunoprecipitated from the lysates, and its activity was assayed in an in vitro kinase activity using 5 μg/ml GST-c-Jun (5-89) as a substrate. B, JNK/SAPK was immunoprecipitated using anti-JNK1 Ab and harvested with protein A-Sepharose. Immunoprecipitates were separated by SDS-PAGE and detected by immunoblotting with anti-JNK1. This experiment is representative of six independent experiments. C, A total of 6 × 10⁶ primary T cells was stimulated with sorbitol (0.4 M) for 0 or 0.5 h as a positive control. T cells were stimulated with immobilized reagents, as follows: either anti-CD3 (1 μg/ml), s4-1BBL (5 μg/ml), or anti-CD3 (1 μg/ml) plus s4-1BBL (5 μg/ml) for 0, 0.5, 24, 48, and 70 h. JNK/SAPK was immunoprecipitated from the lysates and its activity was assayed in an in vitro kinase assay with 5 μg/ml of GST-c-Jun (5-89) as a substrate. D, JNK/SAPK was immunoprecipitated using anti-JNK1 Ab and harvested with protein A-Sepharose. Immunoprecipitates were separated by SDS-PAGE and detected by immunoblotting with anti-JNK1. This experiment is representative of three independent experiments.

To verify that the results obtained in a T cell hybridoma were also valid for primary T cells, T cells were isolated from the spleen of mice, as described in Materials and Methods. The T cell hybridoma, C8.A3, expresses low levels of 4-1BB, and the anti-4-1BB Ab can induce signaling events. However, primary resting T cells do not express 4-1BB constitutively and respond poorly to this Ab, but can respond after anti-CD3 treatment to induce higher levels of receptor (15). In contrast to the results with anti-4-1BB Ab, we have found that primary T cells respond without prior stimulation when anti-CD3 is used in conjunction with 4-1BB ligand. Therefore, for experiments with primary T cells, we used s4-1BB ligand immobilized on plastic to signal via 4-1BB (11). Phosphorylated GST-c-Jun could be detected following T cell treatment with either sorbitol or anti-CD3, but not with s4-1BBL alone. However, the combination of anti-CD3 plus s4-1BBL resulted in greater JNK/SAPK activation than observed with anti-CD3 alone (Fig. 2⇑C). Fig. 2⇑D demonstrates the presence of equivalent amounts of JNK1 protein in the immunoprecipitates. This result is consistent with the observation that primary resting T cells require anti-CD3 treatment to induce 4-1BB expression (53). In contrast, C8.A3 T cells, which express a low level of 4-1BB constitutively, showed JNK/SAPK activation in response to 4-1BB ligation alone (Figs. 1⇑ and 2⇑A). The above data indicate that 4-1BB aggregation on the surface of a T cell hybridoma or on primary T cells leads to activation of the JNK/SAPK pathway.

Induction of ASK-1-TRAF2 association by 4-1BB receptor ligation

As discussed above, the MAP kinase kinase kinase, ASK-1, binds to TRAF2 after TNFR engagement, thereby providing the link between the TNFR and JNK/SAPK cascade (33, 34). To investigate the potential role of ASK-1 in 4-1BB signaling, we again used the T cell hybridoma, C8.A3. RT-PCR analysis indicated that C8.A3 T cells express ASK-1 message (data not shown). However, since Abs to ASK-1 were unavailable, we used epitope-tagged transfected ASK-1 to assess the role of ASK-1 in C8.A3 T cell activation. C8.A3 cells were transiently transfected with vector control, HA-ASK-1, or a kinase dead mutant (HA-ASK-1 K709E). Contrary to results in other cell types, the overexpression of ASK-1 in C8.A3 T cells did not result in noticeable cell death, as measured by trypan blue exclusion (data not shown). Transfected T cells were stimulated with TNF-α, anti-4-1BB, or anti-ICAM-1, as above. HA-ASK-1 was immunoprecipitated with anti-HA Ab, and the immune complexes were subjected to Western blot analysis with the anti-TRAF2 Ab. HA-ASK-1 and HA-ASK-1 K709E were found to associate with TRAF2 after treatment of cells with TNF-α or anti-4-1BB, but not after treatment with anti-ICAM-1 (Fig. 3⇓A). Western blot analysis of extracts indicated similar levels of TRAF2 and HA-ASK-1 in each lysate (Fig. 3⇓, B and C). Previous results have shown that TRAF2 is upstream of ASK-1 in the TNFR signaling cascade (33, 34) and that TRAF2 can associate with 4-1BB upon receptor aggregation (11). Therefore, the present results suggest that following 4-1BB aggregation, TRAF2 is recruited to the receptor complex and can interact with ASK-1.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Association of TRAF2 and ASK-1 following 4-1BB aggregation. C8.A3 T cells (2 × 10⁷) were transiently transfected with either vector (pcDNA3) control, HA-ASK-1, or HA-ASK-1 K709E. Thirty-eight hours following the transfection, C8.A3 T cells (3 × 10⁷) were stimulated with either anti-4-1BB (5 μg/ml) or anti-ICAM-1 (5 μg/ml) for 5 min at 4°C, followed by anti-rat Ig (20 μg/ml) for 15 min at 37°C. C8.A3 cells were also separately stimulated with TNF-α (5 ng/ml) for 15 min at 37°C. Lysates were subjected to immunoprecipitation with anti-HA Ab and harvested with protein A-Sepharose. A, Immunoprecipitates were separated by SDS-PAGE and detected by Western blotting with anti-TRAF2. For detection of TRAF2 and HA-ASK-1 directly in the lysates, ∼1 × 10⁶ cell equivalents were separated by SDS-PAGE and TRAF2, and HA-ASK-1 present in the lysates were detected by Western blot analysis with anti-TRAF2 Ab (B) or anti-HA Ab (C), respectively. Data shown are representative of five separate experiments.

Activation of ASK-1 is induced by 4-1BB aggregation

To determine whether ASK-1 could be activated following 4-1BB aggregation, C8.A3 T cells were transiently transfected with either vector control, HA-ASK-1, or HA-ASK-1 K709E. Following stimulation of the C8.A3 T cells, ASK-1 was immunoprecipitated with anti-HA Ab. The immune complexes were subject to an in vitro kinase assay using the kinase dead GST-SEK1 K129R as the substrate. GST-SEK1 K129R could be phosphorylated by HA-ASK-1 immunoprecipitated from C8.A3 T cells stimulated with TNF-α or anti-4-1BB treatment, but not after anti-ICAM-1 treatment (Fig. 4⇓A). It should be noted that there are background bands in the ICAM-1 lane close to the GST-SEK1 band. However, careful examination indicates that these bands migrate at a distinct position from the GST-SEK1 band, and therefore we do not believe this indicates that ICAM activates ASK-1. As expected, C8.A3 T cells expressing the kinase dead ASK-1 variant did not show GST-SEK1 K129R phosphorylation after any of the treatment regimens (Fig. 4⇓A). Western blot analysis indicates that each cell extract contained similar levels of HA-ASK-1 or HA-ASK-1 K709E (Fig. 4⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Activation of ASK-1 following 4-1BB aggregation. C8.A3 T cells (2 × 10⁷) were transiently transfected with vector (pcDNA3) control, HA-ASK-1, or HA-ASK-1 K709E. Thirty-eight hours following the transfection, transfected C8.A3 T cells (8 × 10⁶) were stimulated with anti-4-1BB (5 μg/ml) or anti-ICAM-1 (5 μg/ml) Abs for 5 min at 4°C, followed by anti-rat Ig (20 μg/ml) for 5, 10, or 20 min at 37°C. C8.A3 cells were also stimulated with TNF-α (5 ng/ml) for 5, 10, or 20 min at 37°C. A, The HA-tagged proteins were immunoprecipitated using 12CA5 Ab, and ASK-1 and ASK-1 K709E were assayed for in vitro kinase activity using the substrate GST-SEK1 K129R (74 kDa). B, HA-ASK-1 or HA-ASK-1 K709E present in the lysate were detected by Western blot analysis with anti-HA Ab (12CA5). This result is representative of six individual experiments.

Dominant-negative ASK-1 interferes with 4-1BB-mediated costimulation in C8.A3 cells

The above studies suggest that ASK-1 can be activated following 4-1BB aggregation on the surface of C8.A3 T cells. To determine the importance of the JNK/SAPK cascade in 4-1BB-mediated costimulation, we took advantage of the ability of the dominant-negative form of ASK-1 to interfere with this pathway. C8.A3 cells were transiently transfected with either vector control, HA-ASK-1, or HA-ASK-1 K709E. The C8.A3 T cell hybrid is costimulation dependent when stimulated using APC expressing the appropriate MHC/peptide complex (A^k plus an unidentified self peptide). In contrast, when stimulated with anti-CD3 immobilized on plastic, the C8.A3 T cell response is costimulation independent. The transfected C8.A3 T cells were stimulated with either immobilized anti-CD3, or an APC (K46J73.35)-expressing transfected A^k. This APC expresses 4-1BB ligand, but little or no B7 family molecules (43). The parental cell line (K46J) lacking the transfected A^k molecule was used as a control. Following an overnight stimulation, supernatants were collected and assayed for their ability to induce proliferation of the IL-2-dependent cell line, CTLL. Mock-transfected, HA-ASK-1-transfected, and HA-ASK-1 K709E-transfected C8.A3 cells secreted similar amounts of IL-2 in response to stimulation with immobilized anti-CD3 (Fig. 5⇓A), indicating that there was no effect of either HA-ASK-1 or HA-ASK-1 K709E transfection on costimulation-independent IL-2 production. However, when C8.A3 T cells were stimulated with APC expressing A^k and 4-1BBL, the HA-ASK-1 K709E-transfected C8.A3 T cells secreted significantly less IL-2 than vector control or HA-ASK-1-transfected C8.A3 T cells (Fig. 5⇓B). A soluble form of the receptor, 4-1BB alkaline phosphatase (4-1BB.AP), was previously shown to specifically interfere with stimulation of C8.A3 T cells in response to Ag/MHC presented on K46J lymphomas (43). Fig. 5⇓C confirms that 4-1BB-AP, but not the AP control, inhibits IL-2 production in this assay. These data are consistent with a role for ASK-1 in providing the link between TRAF2 and the JNK/SAPK cascade following 4-1BB aggregation, and suggest that activation of the JNK/SAPK and/or p38 MAP kinase pathways is critical to 4-1BB-mediated costimulation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Role of ASK-1 in IL-2 secretion by C8.A3 T cells. C8.A3 T cells were transiently transfected with vector (pcDNA3) control, HA-ASK-1, or HA-ASK-1 K709E. Following a 38-h incubation, C8.A3 cells (5 × 10⁴) were stimulated with immobilized anti-CD3 (0.1 μg/ml), irradiated K46J cells, or irradiated K46J A^k cells (5 × 10⁴). After an overnight stimulation, culture supernatants were removed and serial dilutions of supernatant were incubated with an IL-2-dependent cell line, CTLL, for 24 h. [³H]Thymidine was added to the wells for the last 6 h of the CTLL culture. A, Unstimulated and anti-CD3 (0.1 μg/ml)-treated transfected C8.A3 cells. B, Transiently transfected C8.A3 T cells stimulated with APC-expressing transfected A^k. The negative control is vector mock-transfected C8.A3 cells stimulated with APC lacking the A^k molecule. C, Transiently transfected C8.A3 cells stimulated with APC expressing A^k in the presence or absence of either 4-1BB.AP or AP (10 μg/ml). Note that for clarity, the scale of the y-axis scale is not the same in each panel. This experiment is representative of four individual experiments.

Osmotic shock can mimic the costimulatory signal required for IL-2 production by T cells

Hyperosmotic shock, such as caused by sorbitol treatment, results in the activation of the stress kinase cascade (54). In mouse wild-type and MEKK1^−/− embryonic stem cell clones, osmotic shock results in a dramatic increase in JNK/SAPK activation with only modest activation of the ERK and p38 pathways (55). Given that CD28 and 4-1BB signaling have in common their ability to activate the JNK/SAPK cascade and to costimulate IL-2 production, we reasoned that an unrelated stimulus such as hyperosmotic shock might do the same thing. For these studies, we used primary resting T cells responding to anti-CD3, plus or minus a secondary stimulus. To rule out possible effects of B7-CD28 interaction either from B7 on contaminating APC or through T cell-T cell interaction, we used CD28^−/− as well as CD28^+/+ mice as a source of T cells. Fig. 6⇓ shows IL-2 secretion by purified CD28^+/+ or CD28^−/− T cells responding to immobilized anti-CD3 in the presence of either sorbitol, anti-CD28, or s4-1BBL. In a separate experiment, a concentration of 0.4 M of sorbitol was found to give an optimal costimulatory effect, with higher concentrations becoming inhibitory (data not shown). Previous experiments had shown that costimulation with immobilized s4-1BB ligand in this system saturates at ∼5 μg/ml, whereas costimulation with immobilized anti-CD28 saturates at between 5 and 10 μg/ml (data not shown). Neither anti-CD3 (Fig. 6⇓), sorbitol, anti-CD28, nor s4-1BBL alone (data not shown) resulted in IL-2 secretion. In contrast, anti-CD3 in combination with either sorbitol, anti-CD28, or s4–1BBL leads to significant IL-2 secretion. These data suggest that sorbitol is able to partially replace the costimulatory signal for IL-2 secretion provided by CD28 or 4-1BB, most likely due to its ability to activate the JNK/SAPK pathway, demonstrating a critical role of the SAPK in T cell costimulation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Role of the stress kinase cascade in IL-2 secretion. Resting T cells were isolated from spleens of either C57BL/6 CD28^+/+ mice or CD28^−/− mice, as described in Materials and Methods. T cells were incubated with immobilized anti-CD3 (1 μg/ml) in the presence or absence of sorbitol (0.4 M), immobilized anti-CD28 (1 or 10 μg/ml), or immobilized s4-1BBL (1 or 5 μg/ml), as indicated in the figures. After 48 h of culture, supernatants were analyzed for IL-2 production using CTLL cells. Note that for clarity, the scale of the y-axis scale is not the same in each panel. Similar results were obtained from three independent experiments, with two wild-type and two CD28^−/− mice analyzed separately in each experiment.