Costimulation via TCR and IL-1 Receptor Reveals a Novel IL-1α-Mediated Autocrine Pathway of Th2 Cell Proliferation

Immunologic reagents

Purified recombinant murine (rm)³ IL-2 and rmIL-4 were kindly provided by Dr. E. Schmitt, University of Mainz, Germany. Recombinant human (rh)IL-1α, rhIL-1β, and rmIL-1α were provided by Dr Haag (Boehringer, Mannheim, Germany), Dr. D. Boraski (Sclavo, Siena, Italy), and Dr. I. G. Otterness (Pfizer, Groton, MA), respectively. Soluble rhIL-1RII (solIL-1RII) was provided by Behringwerke, Marburg, Germany. Purified anti-IL-4 mAb was obtained by passage of 11B11 hybridoma (10) cell supernatants through protein G columns (Pharmacia, Freiburg, Germany). Anti-mouse IL-1α mAb (11) was purchased from Genzyme (Rüsselsheim, Germany). Polyclonal anti-human (h)IL-1α antiserum was purchased from Endogen (distributed by Biozol, Eching, Germany). Anti-murine CD3 mAb was purified from hybridoma 145-2C-11 (12) cell supernatants using protein A columns (Pharmacia).

T cells

Cloned D10.G4.1 Th2 cells (13) were restimulated (1 × 10⁶ cells/well) every 3 wk in 12-well flat-bottom culture plates (Costar, Cambridge, MA) with 100 μg/ml conalbumin (Sigma, Deisenhofen, Germany) and 5 × 10⁶ irradiated (25 Gy) splenocytes from C3H/HeJ mice (Charles River, Sulzfeld, Germany) in 3 ml Click’s/RPMI (Seromed, Berlin, Germany) medium supplemented as described (14). One week later, the cells were restimulated with fresh medium containing rmIL-2 and rmIL-4 (each at 1 ng/ml). Additionally, 5 to 6 days later, the cells were harvested, extensively washed, and used for the experiments. At this stage, D10G.4.1 cells had a purity of 99.9% as revealed by anti-CD4 staining and FACS analysis.

For the in vivo induction of Th2 cells, BALB/c mice (Charles River) at 6 wk of age were infected with 2 × 10⁷ Leishmania major parasites, as described (14). Three weeks later, the lesion-draining popliteal lymph nodes (LNs) were removed and single-cell suspensions were obtained. Selection of CD4⁺ LN cells was performed using the MACS system and anti-CD8, anti-B220, and anti-Mac-1 mAb conjugated to magnetic beads (all from Miltenyi Biotec, Bergisch Gladbach, Germany) in accordance with the manufacturer’s specifications (15). Enrichment of Th2 cells within the MACS-sorted cells was accomplished by a novel method described in detail elsewhere (16). Briefly, the method uses FACS sorting of cells according to their capacity to extrude the fluorescent dye Fluo-3. MACS-purified LN cells were loaded with Fluo-3 AM (Sigma) and incubated to allow secretion of the dye as described (16). Thereafter, cells were stained with either anti-CD4 Tri-Color mAb (Medac, Hamburg, Germany) or anti-CD4 mAb conjugated to phycoerythrin (PharMingen, Hamburg, Germany). CD4⁺ cells showing decreased Fluo-3 fluorescence (∼3% of total MACS-sorted cells) were sorted using an EPICS XL cell sorter and Elite software (Coulter, Krefeld, Germany). We have shown in the article cited above (16) that within the LNs of L. major-infected BALB/c mice, this procedure leads to strong selection for T cells of the Th2 phenotype. After sorting, the dye-extruding cells used in the experiments described therein contained 99.9% CD4⁺ T cells and 99.3% dye-extruding cells.

“Stimulator cell” lysates and proliferation assays

Stimulator D10.G4.1 cells were cultured (1 × 10⁶/well in 2 ml of medium) in 12-well plates with or without rhIL-1β (300 U/ml) or rhIL-1α (10 U/ml) and immobilized (8) anti-CD3 mAb (coating concentration, 2 μg/ml). After 24 h, the cells were harvested, washed, and added (2 × 10⁴/well) in a volume of 40 μl of fresh medium into 96-well flat-bottom microtiter plates (Costar) either coated or uncoated with anti-CD3 mAb (2 μg/ml). Ex vivo-isolated stimulator Th2 cells (2 × 10⁴/well) were prepared by activating the EPICS-sorted cells described above in a total volume of 200 μl medium in 96-well flat-bottom microtiter plates with or without immobilized anti-CD3 mAb and rhIL-1α, as described for D10.G4.1 stimulator cells. After 24 h, these plates were washed extensively. To obtain cell lysates, all microtiter plates containing activated stimulator cells (D10.G4.1 or ex vivo-purified cells) were shock frozen at −70°C and thawed.

To all plates containing stimulator cell lysates, resting D10.G4.1 cells (2 × 10⁴/well) were added as “responder cells.” Where indicated, wells also received anti-IL-4 mAb (20 μg/ml), solIL-1RII (2 μg/ml), anti-hIL-1α (30 μg/ml), or anti-mIL-1α mAb (5 μg/ml) in a total volume of 200 μl of medium. Responder cells were cultured for 48 h, pulsed with 1.85 × Bq/well [³H]thymidine (New England Nuclear, Dreieich, Germany) for 16 h, and processed for β-scintillation counting.

Flow cytometry

D10.G4.1 cells (1 × 10⁶/well) were cultured for 4 h in the presence of Brefeldin A (10 μg/ml, Sigma) in 12-well plates (2 ml medium/well) with or without rhIL-1β (300 U/ml) and immobilized anti-CD3 mAb (coating concentration 2 μg/ml). Thereafter, cells were harvested, washed in PBS, fixed in 75 μl of solution A (Fix & Perm kit, Medac) for 15 min at 25°C, washed again, and incubated in 75 μl of solution B containing digoxigenin (DIG)-conjugated polyclonal rabbit anti-mouse IL-1α Ab (1:100) (17) for 15 min at 25°C. After washing, cells were incubated for additional 15 min at 25°C in 75 μl of solution B containing FITC-conjugated sheep anti-DIG Fab fragments (2 μg/ml) (Boehringer). After washing, cell fluorescence was analyzed in a FACScan using Lysis II software (Becton Dickinson, Heidelberg, Germany). In control samples, anti-IL-1α Ab were preincubated for 1 h with rmIL-1α (10 μg/ml) or rmIL-2 (10 μg/ml) before staining.

Northern blot analysis

D10.G4.1 cells (1 × 10⁶/well) were activated for 6 h in 12-well plates with or without rhIL-1β and anti-CD3 mAb, as described above. Peritoneal exudate cells were stimulated with LPS (10 μg/ml) for 4 h. After stimulation, total RNA was prepared using the RNeasy Total RNA Kit (Qiagen, Hilden, Germany). Equal amounts of RNA (20 μg/lane) were separated on 1.2% agarose gels, and a Northern blot was performed to detect mRNA for mIL-1α and β-actin as described (18). The autoradiographs were scanned using an Epson GT 8000 and ScanPack software (Biometra, Göttingen, Germany).

Th2 cells activated via TCR and exogenous IL-1 mediate proliferation of resting Th2 cells independently of IL-4 and exogenous IL-1

Recently, we demonstrated that the majority of Th2 cell clones proliferate independently of IL-4 when they are simultaneously costimulated via TCR and IL-1R (5). In the current paper, we investigated whether this type of Th2 cell proliferation was induced directly by exogenous IL-1 or whether costimulation via TCR and exogenous IL-1 led to endogenous synthesis of a mitogenic molecule by the Th2 cells. To test for this possibility, we stimulated Th2 cells in a first culture phase via TCR (using immobilized anti-CD3 mAb) and IL-1R (using rhIL-1β) for 24 h. Thereafter, the cells (called stimulator cells) were harvested, washed, lysed, and tested for their ability to induce an IL-4-independent proliferation of responder Th2 cells. For this purpose, we cocultured the stimulator cell lysates with fresh, viable Th2 responder cells of the same Th2 cell clone. Residual rhIL-1β from the first culture phase and endogenous IL-4 were neutralized by solIL-1RII and anti-IL-4 mAb, respectively. As a control, responder cells were cultured in parallel without stimulator cell lysates but in the presence of anti-CD3 mAb and rhIL-1β. The results of a representative experiment are depicted in Figure 1⇓.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Effect of lysates of activated Th2 cells on the proliferation of resting responder Th2 cells. Stimulator D10.G4.1 Th2 cells were activated as indicated on the y axis, washed, and added into 96-well microtiter plates either coated (A) or uncoated (B) with anti-CD3 mAb. To obtain cell lysates, the plates were shock frozen immediately. After thawing, resting D10.G4.1 Th2 responder cells and, where indicated, anti-IL-4 mAb, solIL-1RII, or anti-IL-4 mAb and solIL-1RII were added. Where indicated, responder cells were stimulated in the absence of lysates, but the presence of rhIL-1β (300 U/ml), with (A) or without (B) immobilized anti-CD3 mAb. After 48 h of culture, the responder cells were pulsed with [³H]thymidine, further incubated for 16 h, and processed for β-scintillation counting. Data represent cpm ± SD of triplicate cultures.

In the absence of stimulator cell lysates, responder cell proliferation depended on the presence of anti-CD3 (“no lysates” in Fig. 1⇑, A vs B) as well as on exogenous rhIL-1β and was blocked by solIL-1RII. This finding demonstrates the efficacy of the solIL-1RII preparation. As reported (5), the proliferation was independent of endogenous IL-4, because it was unchanged by anti-IL-4 (Fig. 1⇑) and cyclosporin A (5). In contrast, the same responder cells, when coincubated with stimulator cell lysates, proliferated independently of exogenous rhIL-1β and IL-4, because proliferation was not blocked by solIL-1RII or by a combination of solIL-1RII and anti-IL-4. However, this type of proliferation occurred only when the lysates were obtained from stimulator cells, which themselves had been incubated with a combination of anti-CD3 mAb and exogenous rhIL-1β. Therefore, after costimulation by anti-CD3 mAb and exogenous rhIL-1β, but not by either reagent alone, stimulator cells expressed a novel proliferative activity that was independent of IL-4 and of exogenous rhIL-1β.

Not only the production of this activity but also the proliferative response to it was dependent on costimulation. This is shown by the finding that responder cells were only able to respond to this activity when they were simultaneously triggered via the TCR (compare Fig. 1⇑, A and B). In the absence of the TCR signal, the proliferation of the responder cells was entirely blocked by anti-IL-4.

The novel costimulatory activity for Th2 cell proliferation is cell associated

The uncharacterized activity could have been represented by a cell-associated or a released molecule. To differentiate between these two possibilities, stimulator cells were again activated by anti-CD3 mAb and exogenous rhIL-1β. Cell lysates and supernatants were obtained and compared for their capacity to mediate responder cell proliferation independently of IL-4 and exogenous rhIL-1β. The results demonstrate (Fig. 2⇓) that upon simultaneous triggering of their TCR, responder cells were able to detect the activity (black bars) only within the lysates, but not within the supernatants of stimulator cells. In contrast, in the absence of the TCR signal, responder cells readily proliferated in response to IL-4 present within lysates or supernatants of stimulator cells. Thus, the activity was cell associated and not released by the cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Fragments but not culture supernatants of anti-CD3- and rhIL-1β-activated Th2 cells costimulate the rhIL-1β- and IL-4-independent proliferation of responder Th2 cells. Stimulator D10.G4.1 Th2 cells were activated with rhIL-1β and immobilized anti-CD3 mAb. After 24 h, culture supernatants and stimulator cells were harvested and added separately into 96-well microtiter plates (A, 2 × 10⁴ cells per well; B, 40 μl supernatant/well) with or without immobilized anti-CD3 mAb, and the plates were shock frozen and thawed. D10.G4.1 responder cells and, where indicated, anti-IL-4 mAb or solIL-1RII were added, and the test was processed as described for Figure 1⇑.

Endogenous IL-1α costimulates Th2 cell proliferation independently of IL-4 and exogenous rhIL-β

Three main features of this activity are very reminiscent of IL-1α: first, the proliferative response to IL-1 also requires costimulation via the TCR (5, 6, 7); second, in a recent report, Th2 cells were also shown to produce IL-1α only after costimulation via two different signals, namely via TCR and CD28 (9); third, IL-1α has been reported to be produced in T cells in a cell associated, but not released form (19). Therefore, we investigated whether the activity was mediated via endogenous IL-1α. To do this, we used several IL-1 antagonists (11, 20), the specificity of which is depicted in Figure 3⇓: anti-mIL-1α mAb specifically blocked the proliferative response of anti-CD3-triggered responder cells to rmIL-1α, but not to rhIL-1α or rhIL-1β. In contrast, solIL-1RII blocked the activity of rhIL-1β, but not of rhIL-1α or rmIL-1α, whereas anti-hIL-1α blocked the activity of rhIL-1α, but not of rhIL-1β or rmIL-1α.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Specificity of anti-IL-1 agents. D10G4.1 Th2 cells were cultured with immobilized anti-CD3 mAb and either rmIL-1α (100 pg/ml) or rhIL-1α (10 U/ml) or rhIL-1β (300 U/ml). Where indicated, cultures also received solIL-1RII (2 μg/ml), anti-hIL-1α (30 μg/ml), or anti-mIL-1α (5 μg/ml). After 48 h of culture, the responder cells were pulsed with [³H]thymidine, further incubated for 16 h, and processed for β-scintillation counting. Data represent cpm ± SD of triplicate cultures.

Using these specific antagonists for IL-1 activity, we next tested whether the novel costimulatory activity for Th2 cell proliferation could be blocked by anti-mIL-1α. In these experiments, we also asked whether this activity was inducible not only by IL-1β but also by IL-1α. Again, stimulator cells were incubated with anti-CD3 mAb with or without rhIL-1β or rhIL-1α, and lysates were obtained and incubated with responder cells in the presence of anti-CD3 mAb (Fig. 4⇓). As before, only lysates from stimulator cells that had been triggered by anti-CD3 mAb and exogenous IL-1 were able to induce responder cell proliferation independently of IL-4 and of the exogenously added IL-1. Clearly, the cells were also able to express the novel activity when stimulated via TCR plus rhIL-1α instead of rhIL-1β. Most interestingly, however, the activity was completely blocked by anti-mIL-1α mAb. This result demonstrates that costimulation of Th2 cells via TCR plus exogenous IL-1α or IL-1β leads to autocrine production of cell-associated IL-1α by Th2 cells and an IL-4-independent proliferative response of the Th2 cells to IL-1α. This result also established for Th2 cells that, in the presence of a TCR signal, IL-1α is able to amplify its own production.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Th2 cells produce IL-1α after combined stimulation via TCR and IL-1R. Stimulator D10.G4.1 Th2 cells were activated with immobilized anti-CD3 mAb and rhIL-1α or rhIL-1β as indicated. Lysates were obtained and added into microtiter plates (all coated with anti-CD3) as described for Figure 1⇑. D10.G4.1 responder cells and, where indicated, anti-IL-4 mAb, solIL-1RII, anti-hILα Ab, or anti-mIL-1α mAb were added. The test was further processed as described for Figure 1⇑.

So far, only stimulator cells derived from a well-characterized Th2 cell clone had been used for analysis. Next, we asked whether ex vivo-isolated Th2 cells were also able to produce IL-1α after costimulation via TCR and exogenously supplied rhIL-1β. To induce differentiation of Th2 cells in vivo, we infected BALB/c mice with the protozoan parasite L. major. In this mouse strain, infection with L. major leads to drastic enlargement of the draining lymph node and expansion of parasite-specific Th2 cells at a frequency of about 1 in 100 to 1000 (reviewed in 21 . These Th2 cells were enriched as described in Materials and Methods. When such Th2 cells were stimulated by anti-CD3 and/or exogenous rhIL-1β and lysed, the lysates also contained an activity that was refractory to blockade with solIL-1RII or anti-IL-4 mAb, but was blocked by anti-mIL-1α mAb (Fig. 5⇓). As for cloned Th2 stimulator cells (Fig. 1⇑A), this activity only emerged when the ex vivo-purified Th2 cells had been incubated simultaneously with immobilized anti-CD3 mAb and exogenous rhIL-1β. Thus, endogenous production of IL-1α after costimulation via TCR and exogenous IL-1 is not restricted to long-term cultured Th2 cell clones, but can readily be detected in freshly ex vivo-isolated Th2 cells as well.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Ex vivo-isolated Th2 cells produce IL-1α after combined stimulation via TCR and IL-1R. Ex vivo-isolated Th2 cells were obtained as described in Materials and Methods. They were activated (2 × 10⁴/well) with immobilized anti-CD3 mAb as described above. After 24 h, the plates containing the cells were extensively washed and, to obtain cell lysates, were shock-frozen at −70°C and thawed. D10.G4.1 responder cells and, where indicated, anti-IL-4 mAb, solIL-1RII, or anti-mIL-1α mAb were added. The test was further processed as described for Figure 1⇑.

Combined stimulation of Th2 cells via TCR and IL-1R induces the expression of IL-1α protein and mRNA

So far, only the functional activity of IL-1α has been identified. The next set of experiments was performed to detect the expression of IL-1α at the protein and mRNA level. For detection at the protein level, we adapted the method of intracellular cytokine staining (22) and FACS analysis for IL-1α. Again, cloned Th2 cells were stimulated with or without anti-CD3 mAb or exogenous rhIL-1β. Then, 24 h later, cells were processed for intracellular staining of IL-1α. A representative result is shown in Figure 6⇓: IL-1α protein was detected in individual cells, provided that they had been triggered by a combination of anti-CD3 mAb plus rhIL-1β but not by either of the reagents alone. Thus, staining at the single-cell level perfectly confirmed the functional data described above. The staining was specific for IL-1α for several reasons: first, an affinity purified antiserum was used that reacts specifically with murine IL-1α, but not with IL-1β (17); second, no staining was obtained when this antiserum was omitted and only the FITC-conjugated secondary Ab was applied (Fig. 6⇓B); and third, staining was blocked by preincubation of anti-IL-1α antiserum with rmIL-1α, but not with rmIL-2 (Fig. 6⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Costimulation via TCR and rhIL-1β induces expression of IL-1α protein in Th2 cells. D10.G4.1 cells were stimulated for 4 h, as indicated. Thereafter, the cells were fixed, permeabilized, and stained intracellularly using rabbit anti-mouse IL-1α-DIG Ab and sheep anti-DIG-FITC Fab fragments (A). B, Control staining was done, as indicated, using either sheep anti-DIG-FITC Fab fragments alone or rabbit anti-mouse IL-1α-DIG Ab preincubated with rmIL-1α or rmIL-2 followed by sheep anti-DIG-FITC Fab fragments.

For analysis at the mRNA level, cloned Th2 cells were stimulated as before, and total RNA was isolated and tested by Northern blotting for the presence of mRNA for IL-1α. The results of a representative experiment are shown in Figure 7⇓. As with protein expression, the results obtained for mRNA expression entirely correlated with the functional data reported above; a signal for IL-1α became visible only when the Th2 cells had been incubated with the combination of anti-CD3 plus exogenous rhIL-1β. Surprisingly, after adjustment for the strength of the β-actin signal, the amount of IL-1α mRNA was in the same order of magnitude as that obtained from LPS-triggered peritoneal macrophages, i.e., professional producers of IL-1α. Thus, appropriate stimulation identifies Th2 cells as a major source of this cytokine.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Costimulation via TCR and rhIL-1β induces expression of IL-1α mRNA in Th2 cells. Total RNA was prepared from D10.G4.1 cells that had been stimulated for 6 h as indicated. The RNA together with control RNA from LPS-activated peritoneal exudate cells was analyzed by Northern blotting as described in Materials and Methods. Scans of the autoradiographs are depicted.

Conclusions

Our results establish for the first time a self-amplifying autocrine loop for Th2 cell proliferation; triggering of their TCR by an APC in conjunction with small amounts of IL-1α or IL-1β leads to amplification of IL-1α production by the Th2 cells and their IL-4 independent proliferation in response to it (Fig. 8⇓). These findings help to explain the differences in the literature with regard to IL-1α production by Th2 cells. While it was reported in several studies that a signal via the TCR alone is sufficient to induce IL-1α production by these cells (19, 23), a recent study did not confirm this result (9). Rather, in that study, a costimulatory signal delivered by CD28 was found to be strictly required for production of IL-1α. However, the amounts of IL-1α mRNA produced appeared to be quite low, because they were only detectable using the very sensitive PCR technology. In our study, using the same Th2 cell clone, we detected substantial amounts of IL-1α mRNA by Northern blotting, i.e., without requiring amplification. In the cited studies, Th2 cells have probably been analyzed in different states of activation. TCR triggering of Th2 cells that are not sufficiently resting may by itself lead to production of small amounts of IL-1α in the absence of a costimulus. Such IL-1α then self-amplifies further production of IL-1α, particularly when the cell density is high. If the cells are resting, however, they are critically dependent on costimulation, e.g., via CD28, to produce IL-1α. Yet, the amounts of IL-1α detected after costimulation via TCR and CD28 are very small and may serve only as a starter for production of higher amounts by the self-enhancing loop described in our article. In the absence of a B7-CD28 interaction, small amounts of IL-1 derived from other cellular sources, such as APC or other already-activated Th2 cells, may serve to costimulate IL-1α production of TCR-triggered Th2 cells and initiate their autocrine proliferative loop.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

A model that illustrates a novel pathway of Th2 cell proliferation. Triggering of Th2 cells via Ag-peptide presented by an APC (Signal 1) and IL-1 or B7 expressed by the APC (Signal 2) leads to the self-amplifying production of IL-1α by Th2 cells and IL-4-independent proliferation in response to IL-1α.

The novel demonstration of this autocrine loop of Th2 cell proliferation may have important clinical implications. In the presence of the TCR signal and absence of IL-1, Th2 cell proliferation was blocked in response to IL-4 (5), the generally accepted autocrine lymphokine for Th2 cell proliferation. IL-4 may therefore have limited importance for Th2 cell expansion in the presence of high loads of Ag and when differentiation of Th2 cells, mediated by IL-4 (24, 25), has already been established. This conclusion is supported by the fact that therapy with anti-IL-4 or soluble IL-4R has little effect on the fatal course of established leishmaniasis in Th2-prone BALB/c mice (21, 26). We suggest that, under these conditions, Th2 proliferation results mainly from the autocrine loop mediated via IL-1α. Therefore, therapy of an established Th2 response may necessitate the application of reagents that neutralize IL-4 as well as IL-1α. Studies to test this hypothesis in murine leishmaniasis have been initiated in our laboratory.