Interleukin-4 supports interleukin-12-induced proliferation and interferon-γ secretion in human activated lymphoblasts and T helper type 1 cells

Abstract

Interleukin-12 (IL-12) and IL-4 are known to differentially promote T helper (Th) cell differentiation. While IL-12 induces interferon-γ (IFN-γ) production and maturation of Th1 cells, IL-4 is thought to antagonize IL-12 and to favour Th2 development. Here we studied the combined action of various concentrations of common γ-chain (γ_c-chain) cytokines, including IL-4 and the Th1 cytokine IL-12, in human activated lymphoblasts and Th1 cells. IL-4 and IL-7 potentiated IL-12-induced proliferation at every concentration tested (1–10 ng/ml) without increasing rescue from apoptosis, indicating that proliferation was directly affected by these cytokine combinations. With regards to cytokine secretion, IL-2 together with IL-12 initiated tumour necrosis factor-α synthesis, enhanced IFN-γ production, and shedding of soluble IL-2 receptor α as expected. Importantly, combining IL-4 with IL-12 also enhanced IFN-γ secretion in lymphoblasts and a Th1 cell line. Investigating signal transduction in lymphoblasts induced by these cytokines, we found that not only IL-2 but also IL-4 enhances signal transducer and activator of transcription 3 (STAT3) tyrosine phosphorylation by IL-12. Tyrosine phosphorylations of janus kinase 2 (JAK-2), tyrosine kinase 2 (TYK2), extracellular signal-regulated kinase (ERK) and STAT4, STAT5 and STAT6 were not potentiated by combinations of these cytokines, suggesting specificity for increased STAT3 phosphorylation. In conclusion, two otherwise antagonizing cytokines co-operate in activated human lymphoblasts and Th1 cells, possibly via STAT3 as a converging signal. These data demonstrate that IL-4 can directly enhance human Th1 cell function independently of its known actions on antigen-presenting cells. These findings should be of importance for the design of cytokine-targeted therapies of human Th-cell-driven diseases.

Introduction

T helper cell differentiation is orchestrated by signals from the T-cell receptor (TCR), costimulatory molecules and cytokine receptors.¹ Cytokines can modulate T helper type 1 (Th1)/Th2 cell differentiation via chromatin remodelling of Th cell loci.^1–3 Interleukin-12 (IL-12) is considered a prototypical cytokine for Th1 development, although first identified as a natural killer (NK) cell stimulatory factor.^1,4–8 It is mostly produced by antigen-presenting cells^5–7,9 and induces production of interferon-γ (IFN-γ) in T and NK cells, which in turn augments synthesis of IL-12 in monocytes and polymorphonuclear cells.⁵ IL-12 is a central player in the initiation of Th1 responses against some types of infections^5,7 but newly emerging cytokines and receptor systems add more complexity to this system.^5,6,10,11

In quiescent peripheral T cells or Th1 clones, IL-12 synergizes with B7 for optimal proliferation and IFN-γ production.^12,13 IL-12 also promotes the development of Th1 cells through IFN-γ-dependent and -independent mechanisms and inhibits the generation of IL-4-producing Th2 cells.^14–16 On the other hand, IL-4 has been demonstrated to inhibit IFN-γ production¹⁷ and is thought to skew Th cell differentiation towards Th2.¹

Thus the current paradigm is that IL-12 and IL-4 compete for the promotion of Th1 versus Th2 development at least in quiescent T cells. This concept has been validated in multiple in vivo models of autoimmunity and infection, but has been challenged by contrasting findings in some autoimmunity and infection models.^18,19 In line with this, IL-12 can increase IL-4 production under certain circumstances^20,21 and human IL-4 can also induce antigen-presenting cells to produce IL-12.^22,23 Furthermore, IFN-γ enhances priming of CD4 T cells for IL-4 production and some Th1-polarized cells have been shown to coexpress IFN-γ and IL-4.^26,27 In addition, IL-4 and IL-12 can co-operate in Th1, NK and dendritic cells in the murine system.^20,28–31 This co-operation between Th cell cytokines might explain some of the paradoxical effects seen in several Th1-polarized mouse models of autoimmunity.¹⁸ However, neither the mechanism responsible for synergy nor its role in the human system has been established for T cells to date. It has been speculated that modulation of cytokine signalling might be involved in Th1/Th2-positive cross-regulation, because cytokine receptor levels were unchanged in murine Th cells.³⁰ Supporting this idea, enhanced signal transduction has recently been demonstrated for IL-4/IL-12 in murine NK cells.³¹

One key signalling pathway downstream of cytokine receptors is the Janus kinase (JAK)/signal tranducer and activator of transcription (STAT) pathway.³² Signal transduction through the IL-12 receptor-β1/β2 heterodimer induces phosphorylation of JAK2 and tyrosine kinase 2 (TYK2) as well as STAT1, STAT3, STAT4 and STAT5,^5,6,33–38 which are also phosphorylated in Th1 cells after differentiation.³⁹ IL-2 signalling predominantly induces phosphorylation of STAT5A/B and extracellular signal-regulated kinase (ERK), while IL-4 specifically triggers STAT6 activation.^{5,6,33–36,38} Importantly, both IL-2 and IL-4 also induce STAT3 phosphorylation in preactivated T cells.^32,37,40

Activated T lymphoblasts also differ from quiescent cells in various other ways. In addition to acquisition of effector function, activated cells readily proliferate in response to common γ-chain (γ_c-chain) cytokines without the need for TCR triggering.^37,40 Furthermore, CD45 signals in quiescent T cells inhibit proliferation and cytokine production, whereas the same signals in activated lymphocytes decrease proliferation without changing cytokine production.³⁷

To further dissect differences of activated human lymphoblasts, we were studying the combinatorial effects of γ_c-chain cytokines on IL-12-induced effector functions in activated human lymphoblasts. In this paper we demonstrate that IL-12 and IL-4 have additive effects on proliferation and cytokine production in human T lymphoblasts and Th1 cells. While activation of several major signalling molecules was not potentiated, STAT3 tyrosine phosphorylation was selectively enhanced, pointing towards a possible molecular explanation for the observed loss of antagonization between otherwise opposing Th cytokines.

Materials and methods

Generation of activated lymphoblasts

Peripheral blood mononuclear cells (PBMC) were isolated from heparinized peripheral blood by Ficoll–Hypaque (BAG, Lich, Germany) density gradient centrifugation, washed twice with phosphate-buffered saline (PBS), and resuspended in RPMI-1640 (Gibco BRL, Eggenstein, Germany) supplemented with 4 mm l-glutamine, 10 mm HEPES buffer, 100 U/ml penicillin, 0·1 mg/ml streptomycin (all BioWitthaker, Verviers, Belgium) and 10% (v/v) heat-inactivated fetal calf serum (Gibco BRL). For generation of lymphoblasts, freshly isolated PBMC were activated with 1 μg/ml phytohaemagglutinin (PHA; Sigma Chemical, Deisenhofen, Germany) for 5 days with medium being replenished every 2–3 days. Thereafter, cells were expanded with 10 U/ml IL-2 (10 U/ml, Boehringer Mannheim, Mannheim, Germany) for another 2 days. Resulting lymphoblasts were > 95% CD3⁺ (with 60–80% CD4⁺ cells). Finally, lymphoblasts were extensively washed and activated as indicated. TCR/CD28 and IL-2 together (similar to PHA/IL-2 stimulation) have been demonstrated to drive Th1 development in human naive T cells.⁴¹ Similarly, our PHA/IL-2-generated lymphoblasts exhibited a predominant Th1 phenotype (1% IL-4 positive, 2% IL-4/IFN-γ positive, and 38% IFN-γ positive by intracellular cytokine staining).

Isolation of CD4⁺ lymphoblasts with magnetic beads

CD4⁺ T cells were isolated from PHA/IL-2-generated lymphoblasts with a negative CD4⁺ T-cell isolation kit from Miltenyi Biotec (Bergisch Gladbach, Germany) according to the manufacturer's protocol and as previously described.⁴² Purity was routinely confirmed by flow cytometric analysis (> 95% CD4-positive).

Cell line, antibodies and reagents

Human Th1 cell line HG 163.12 was kindly provided by Dr Alla Skapenko. This line is TCRα/β-positive and CD3/CD4-positive, derived from a healthy normal donor via repetitive activation with anti-CD3 and syngeneic feeder cells (Epstein–Barr virus-transformed B cells) plus Th1-inducing conditions (IL-12 and anti-IL-4). Large amounts of IFN-γ and IL-2, but no IL-4 are produced. The cell line was maintained in culture medium as used for lymphoblast generation. A proliferative phase with high-dose IL-2 (5000 U/ml; Aldesleukin, Chiron, Munich, Germany), PHA (1 μg/ml) and feeder cells (two-fold excess) was alternated weekly with a quiescent phase of low-dose IL-2 only (2000 U/ml). Medium was replenished every 3–4 days.

IL-2 (10 U/ml) was from Boehringer Mannheim. IL-4, IL-7 and IL-15 (10 ng/ml) were purchased from PeproTech EC (London, UK). Hybridoma cells producing CD3 monoclonal antibody (mAb) OKT-3 [immunoglobuin G2a (IgG2a), 1 μg/ml] were from the American Type Culture Collection (Rockville, MD). Antibodies were purified from hybridoma cell supernatant using the FPLC unit LCC 500 plus (Pharmacia, Erlangen, Germany) employing HiTrap protein A–Sepharose columns from Pharmacia. pERK-1/2 mAb (E-4) and ERK mAb (K-23) were obtained from Santa Cruz Biotechnology (Heidelberg, Germany). JAK2, pJAK2, TYK2, pSTAT-3, pSTAT-5 and STAT-5 mAbs were purchased from Upstate Biotechnology, Lake Placid, NY. pSTAT-6 was from New England Biolabs, Beverly, MA; mAbs recognizing STAT-3 and STAT-6 were obtained from Transduction Laboratories, Lexington, KY, STAT4 and pSTAT4 mAbs were from Zymed, San Francisco, CA, and pTYK2 mAbs were from Calbiochem, Bad Soden, Germany.

Cell proliferation assays

Uptake and incorporation of [³H]thymidine into genomic DNA was used to quantify cellular proliferation. PHA-activated blasts (1 × 10⁵) were incubated for 72 hr in a total volume of 150 μl medium in a 96-well round-bottom microtitre plate (Costar, Cambridge, MA) with the indicated reagents in triplicates at 37° and 5% CO₂. Samples were pulsed for the last 6 hr with 0·5 μCi/well [³H]thymidine (2 Ci/mmol, Amersham, Braunschweig, Germany) and harvested onto glass-fibre filters. Radioactivity was measured by liquid scintillation counting. Proliferation of purified CD4⁺ T cells was similarly performed, but with slight modifications: 5 × 10⁴ CD4 T cells were incubated in flat-bottom microtitre plates and incubated as above. Plates were coated with anti-CD3 antibodies to cross-link T-cell receptors in addition to cytokine-mediated stimulation. This combination of stimuli ensured optimal IFN-γ secretion, which was concomitantly measured (see below).

Quantification of apoptosis

Microscopic examination of the cell cultures prone to undergo apoptosis revealed morphological changes like zeiosis. For quantification of apoptosis, DNA staining with propidium iodide (Sigma Chemical) and flow cytometry analysis were performed as described elsewhere.⁴³ In brief, 4 × 10⁵ cells were pelleted with 200 g and gently resuspended in 150 μl hypotonic fluorochrome solution of 50 μg/ml phosphatidylinositol in 0·1% (w/v) sodium citrate plus 0·1% (v/v) Triton X-100 (Sigma Chemical). After a minimum period of 6 hr in the dark at 4°, samples were analysed on a FACScan (Coulter, Hialeah, FL). The percentage of apoptotic cells was calculated as follows: % cells with subdiploid DNA content/% all cells positive for propidium iodide staining × 100.

Measurement of cytokines in culture supernatants

Supernatants were collected after 16–18 hr of restimulation for PHA-activated lymphoblasts and after 9 hr for the Th1 cell line HG 163.12. IFN-γ levels in culture supernatants were measured using an enzyme-linked immunosorbent assay kit from R & D systems (Wiesbaden, Germany) following the manufacturer's recommendations. IL-1β, IL-6, IL-8, TNF-α and soluble CD25 were measured with an automated immunoanalyser (Immulite™, DPC Biermann, Bad Nauheim, Germany). Supernatants of CD4⁺ lymphoblasts isolated by magnetic antibody cell sorting were also measured for IFN-γ levels. Cells were stimulated in 96-well flat-bottom microtitre plates (5 × 10⁴ CD4⁺ T cells per well in 200 μl volume). Plate-bound anti-CD3 stimulation was added to maximize cytokine secretion from CD4⁺ lymphoblasts according to initial titration experiments (data not shown). Supernatants were collected after 48 hr of restimulation.

Western blot analysis

IL-2-expanded PHA-activated blasts were extensively washed and aliquots of 1 × 10⁶ cells were rested for 6–8 hr in medium before restimulation. Thereafter, cytokines were added for an incubation period of 10–20 min in concentrations as indicated in the figures. Cells were lysed in RIPA lysis buffer containing 50 mm Tris–HCl pH 7·6, 150 mm NaCl, 5 mm ethylenediaminetetraacetic acid, 0·5% (v/v) nonidet P-40, 1% (v/v) Triton X-100, 1 mm sodium vanadate, 1 mm phenylmethylsulphonyl fluoride, 10 μg/ml aprotinin, 10 μg/ml pepstatin, 10 μg/ml leupeptin, and 0·25% sodium deoxycholate at 4° for 1 hr (all reagents from Sigma Chemical). Supernatants were clarified by centrifugation for 10 min at 10 000 g at 4°. Protein concentration was determined using a Bradford method protein assay kit (Bio-Rad, Munich, Germany): 20 μg total protein lysates from each sample were diluted in an equal volume of 2 × sodium dodecyl sulphate (SDS) sample buffer, boiled for 5 min and resolved by SDS–polyacrylamide gel electrophoresis. Immunoblots were performed by semi-dry transfer from the proteins onto nitrocellulose membranes. Unspecific binding sites were blocked in freshly prepared 5% non-fat dry milk in PBS/0·1% Tween-20 for at least 30 min at room temperature. 0·1–1 μg/ml of the specific antibody as indicated was incubated in blocking buffer overnight at 4°. Subsequently, 0·2–1 μg/ml horseradish peroxidase-conjugated goat anti-mouse (or anti-rabbit) IgG was added in blocking buffer for 1·5 hr at room temperature. Detection was performed using enhanced chemiluminescence (Amersham). After detection of phosphorylated forms of proteins, antibodies were washed off the membranes by incubation in 62·5 mm NaCl, 100 mm 2-mercaptoethanol and 2% SDS for 30 min at 50–60°. After washing in PBS/0·1% Tween-20, membranes were blocked as described above, and staining with antibodies specific for all forms of signalling molecules was performed as described below. Densitometric analysis was performed with a densitometer from the Alpha Innotech Corporation (San Leandro, CA).

Statistical analysis

Student's t-test was performed where appropriate and P-values are indicated in the figures or figure legends.

Results

IL-4 and IL-7, but not IL-2 or IL-15, co-operate with IL-12 to induce proliferation in activated human lymphoblasts at various concentrations

PHA-activated and IL-2-expanded lymphoblasts were generated as described above. After extensive washing, cells were stimulated with various concentrations of γ_c-chain cytokines IL-2, IL-4, IL-7 and IL-15 (1–10 U or ng/ml) alone or in combination with different concentrations of IL-12 (1–10 ng/ml). As shown in Fig. 1(a), IL-2 induced DNA synthesis in a concentration-dependent manner. IL-12 induced less proliferation with no differences between the various concentrations. However, 1 ng/ml of IL-12 was sufficient to induce proliferation, indicating that these lymphoblasts possibly resemble highly sensitive Th1 cells in vivo.

Figure 1.

Proliferation of PHA-activated lymphoblasts in response to IL-12 alone or in combination with IL-2 (a; n = 10), IL-15 (b; n = 8), IL-4 (c; n = 8) or IL-7 (d; n = 7). PHA-activated, IL-2-expanded lymphoblasts were generated as described in the Materials and methods. Cells were restimulated after extensive washing with varying concentrations of IL-12 alone or together with one of the γ_c-chain cytokines. Doses ranged from 1 to 10 ng/ml (or U/ml for IL-2) as depicted in the figure. Proliferation was measured as described in the Materials and methods after 3 days of restimulation. Each data-point on a line represents an average of multiple, independent experiments from different normal donors. SEM ranged from 206 to 3601 c.p.m. Concentration-dependent responses of each γ_c-chain cytokine alone (Medium) or combined with the highest concentration of IL-12 (10 ng/ml) are plotted with solid lines, all others with interrupted lines. Statistical significances were calculated for highest proliferation rates of solid lines, i.e. γ_c-chain cytokine alone at 10 ng/ml (or U/ml) versus combination with IL-12 at 10 ng/ml.

Interestingly, combinations of IL-12 and IL-2 were partially antagonistic with IL-12 (> 1 ng/ml) suppressing optimal proliferation at higher IL-2 concentrations (Fig. 1a). When IL-12 and IL-15 were combined (Fig. 1b), IL-12 showed additive effects with IL-15 only at the lowest concentration, but did not suppress IL-15-mediated proliferation at higher doses in contrast to IL-2. Compared to IL-2 or IL-15 alone, the proliferative capacity of lymphoblasts after isolated IL-4 (Fig. 1c) or IL-7 (Fig. 1d) stimulation was inferior. Surprisingly, however, under these circumstances, the addition of IL-12 had clear additive effects without a distinct dose–response in the tested range (Fig. 1c,d). Similar effects on IL-4/IL-12-induced proliferation have been found with lymphoblasts generated with tetanus toxoid stimulation (data not shown).

Protection from apoptosis induced by growth factor withdrawal is reduced by IL-2, but unaltered by IL-4 in combination with IL-12

As activated lymphocytes are prone to undergo apoptosis, we next wondered whether the variable influence of IL-12 in combination with the different γ_c-chain cytokines (Fig. 1) was also reflected in a differential prevention of apoptosis. Apoptotic cells were quantified as described in the Materials and methods after a 3-day culture under the various conditions. Figure 2(a) indicates that IL-2 inhibited apoptosis dose dependently. IL-12 was much weaker in this perspective with no significant dose-dependency. In parallel to decreased proliferation, IL-12 prevented optimal anti-apoptotic effects of IL-2 (Fig. 2a). This was also seen in the combination of IL-12 with IL-15 (data not shown). IL-4 exerts anti-apoptotic effects as well (Fig. 2b) with no clear dose-dependency. Interestingly, in contrast to IL-2, the addition of IL-12 to IL-4 did not obviously alter the percentages of apoptotic lymphoblasts (Fig. 2b) despite its additive effects in proliferation assays (see Fig. 1c). This has been similarly seen in combinations of IL-12 with IL-7 (data not shown). Increased proliferation is therefore the result of enhanced cell division as opposed to enhanced survival in this case.

Figure 2.

Apoptosis of PHA-activated lymphoblasts in response to withdrawal of IL-12 alone or in combination with IL-2 (a; n = 10) or IL-4 (b; n = 8). PHA-activated, IL-2-expanded lymphoblasts were generated as described in the Materials and methods. After extensive washing cells were left in medium alone or cultured with IL-12 alone or together with one of the γ_c-chain cytokines. The rate of apoptosis was measured as described in the Materials and methods after a culture period of 3 days. Each data-point on a line represents an average of multiple, independent experiments from different normal donors. SEM are depicted as error bars.

IL-12-induced IFN-γ production is enhanced by γ_c-chain cytokines including IL-4

To investigate whether the various combinations of cytokines caused differential cytokine secretion from these lymphoblasts, we stimulated them under conditions as detailed in the figures. To yield optimal cytokine secretion we added the CD3 mAb OKT-3 to restimulation cultures. As shown previously for γ_c-chain cytokines alone,³⁷ stimulation of lymphoblasts with any of the γ_c-chain cytokines with or without IL-12 could not induce cytokine secretion significantly above background levels in the absence of OKT-3 (data not shown). IL-12 alone, on the other hand, was capable of inducing IFN-γ secretion even without further TCR support (see below).

Supernatants were analysed for concentrations of IFN-γ and the following other cytokines for comparison: IL-1β, IL-6, IL-8, TNF-α, and soluble CD25 (sCD25). Only TNF-α and sCD25 were significantly elevated under these experimental conditions in addition to IFN-γ. As shown in Fig. 3(a), neither γ_c-chain cytokines nor IL-12 could increase TNF-α secretion above background level (OKT-3 alone), with IL-4, IL-7 and IL-15 diminishing TNF-α concentrations even below background. On the other hand, IL-12 plus IL-2 increased TNF-α secretion, whereas addition of IL-12 to IL-4, IL-7 or IL-15 eliminated the inhibitory effects of each of these γ_c-chain cytokines alone (Fig. 3a). The only cytokines that increased the soluble form of the IL-2Rα (sCD25) were IL-12 and, as expected, especially IL-2 (Fig. 3b). IL-2 plus IL-12 also showed additive effects on shedding of CD25 (Fig. 3b). Addition of the other γ_c-chain cytokines to IL-12 had no significant effect (Fig. 3b).

Figure 3.

TNF-α (a; n = 3), soluble CD25 (b; n = 3) and IFN-γ (c, d, e, f; n = 4) production of PHA-activated lymphoblasts in response to restimulation with IL-12 alone, γ_c-chain cytokines alone or in combination. PHA-activated, IL-2-expanded lymphoblasts were generated as described in the Materials and methods. After extensive washing cells were restimulated with 1 or 10 ng/ml (or U/ml for IL-2) of cytokines (or cytokine combinations) with or without soluble OKT-3 as depicted in the figure. Cytokine production or receptor shedding were measured as described in the Materials and methods after a culture period 16–18 hr. Error bars represent SEM.

As mentioned above, IL-12 stimulation led to significant IFN-γ secretion in lymphoblasts even without TCR support (21 pg/ml on average; Fig. 3c). This was seen less after IL-2 stimulation alone, and was absent under conditions with isolated IL-4, IL-7, or IL-15 activation (Fig. 3c). Combining IL-12 and IL-2 induced copious amounts of IFN-γ to be secreted by preactivated T cells even in the absence of OKT-3 (Fig. 3c) and despite its inhibitory effects on proliferation (see above and Fig. 1a). The other γ_c-chain cytokines also augmented IL-12-induced IFN-γ secretion albeit not as strongly as IL-2 (Fig. 3c). Interestingly, IL-4 co-operated in a dose-dependent manner with IL-12 (36 and 57 pg/ml of IFN-γ secretion at doses of 1 and 10 ng/ml of IL-4, respectively; see also Fig. 3c and e). When TCR stimulation was added to cytokines, IL-12 especially (and to a lesser extent IL-2) induced IFN-γ production above levels seen without CD3 support (Fig. 3d). Notably, combining any of the γ_c-chain cytokines with IL-12/OKT-3 also enhanced IFN-γ secretion (Fig. 3d) indicating their Th1-potentiating capacity, including Th2 cytokine IL-4 (Fig. 3d and f).

To confirm the co-operation observed between γ_c-chain cytokine IL-4 and IL-12 in human lymphoblasts, we also tested IFN-γ secretion in a human Th1 cell line under these conditions. This cell line has a high constitutive expression of IFN-γ (around 300 pg/ml in average) because of its activated Th1-type nature (Fig. 4). IL-12 alone did not increase this high baseline, but TCR stimulation did, which was strongly enhanced by IL-12. Interestingly especially IL-4 co-operated with IL-12/OKT-3 in these cells, although IL-4 alone strongly suppressed constitutive IFN-γ secretion (average values: 164 pg/ml of IFN-γ for IL-4 concentrated 1 ng/ml; 84 pg/ml of IFN-γ for IL-4 concentrated 10 ng/ml; see Fig. 4). Taken together with the data on primary human lymphoblasts, these results strongly support the observed additive effects between IL-4 and IL-12, while IL-4 alone still acts as a classical Th2 cytokine, antagonizing constitutive IFN-γ production. Finally, we also isolated CD4⁺ T lymphoblasts and examined them for IFN-γ secretion after restimulation with IL-12 ± IL-4 (1 and 10 ng/ml). This experiment should fully exclude any influence of possible small fractions of other contaminating cell types. Highly pure CD4⁺ T cells were negatively selected with magnetic beads after generation of PHA/IL-2-expanded lymphoblasts. Restimulation of CD4⁺ T cells revealed again the additive effects of IL-12 in combination with IL-4 in addition to plate-bound anti-CD3 for optimal cytokine secretion (Table 1). Interestingly, this effect was abrogated with additional costimulation with soluble anti-CD28 antibodies (Table 1).

Figure 4.

IFN-γ production of human Th1 cell line HG 163.12 in response to stimulation with IL-12 alone, γ_c-chain cytokines alone or in combination. Cells were stimulated with 1 or 10 ng/ml (or U/ml for IL-2) of cytokines (or cytokine combinations) with or without soluble OKT-3 as depicted in the figure. Cytokine production was measured as described in the Materials and methods after a culture period of 9 hr. Each column represents an average of three independent experiments. Error bars represent SEM. P-values for OKT-3 + IL-12 alone versus combinations with IL-2 or IL-4 are depicted in the figure.

Table 1.

IFN-γ secretion (in pg/ml) from purified CD4⁺ lymphoblasts

	IL-12 alone	+ IL-4 (1 ng/ml)	+ IL-4 (10 ng/ml)
OKT-3	1885	2215	2065
OKT-3/IL-12	3890	5450	5090
OKT-3/CD28	1665	2315	2185
OKT-3/CD28/IL-12	5180	5625	5085

CD4⁺ lymphoblasts were stimulated with plate-bound anti-CD3⁺-soluble anti-CD28 and IL-12 (10 ng/ml) alone or in combination with IL-4 (1 and 10 ng/ml). CD4⁺ T cells were isolated from PHA/IL-2-expanded lymphoblasts as described in the Material and methods section and supernatants were measured by enzyme-linked immunosorbent assay as described in the Material and methods. Values represent averages of duplicates.

IL-12-induced STAT3 tyrosine phosphorylation is potentiated by IL-2 or IL-4

Next, we were interested which STAT molecules were activated by combined IL-12/IL-4 signalling in activated lymphoblasts. To test this, we performed immunoblot analyses using antibodies specific for tyrosine-phosphorylated forms of TYK-2, JAK-2, ERK, STAT3, STAT4, STAT5 and STAT6 proteins. Cytokine concentrations were lowered to 0·1 ng/ml (or U/ml) to better distinguish the enhancing effects of combinations. However, we did not find additive effects of TYK-2, JAK-2, ERK, STAT4 or STAT6 phosphorylations (data not shown and Fig. 5a for ERK and STAT4). As shown in Fig. 5(a), weak phosphorylation of STAT3 was induced at low doses of IL-2 or IL-4, respectively. Low-level phosphorylation was also seen at high IL-12 concentrations (10 ng/ml). Interestingly, combinations of IL-2 or IL-4 with IL-12 enhanced tyrosine phosphorylation of STAT3 (Fig. 5a). Densitometric analysis of IL-4/IL-12-induced and IL-2/IL-12-induced enhancements of STAT3 phosphorylation are shown in Fig. 5(b,c). Finally, IL-12 signals induced phosphorylation of STAT5 to a much lower extent as compared to low IL-2 signals (data not shown). However, because of the strength of the IL-2 effects, no enhancement could be seen with IL-12/IL-2 activation (data not shown). In summary, IL-2 and IL-4 potentiated IL-12-induced STAT3 phosphorylation. While we have not examined IL-12/IL-7-activated STAT3, we would expect similar enhancement considering the demonstrated additive actions on proliferation and IFN-γ secretion.

Figure 5.

Signal transduction of PHA-activated lymphoblasts in response to restimulation with IL-12 alone, γ_c-chain cytokines alone or in combination. PHA-activated, IL-2-expanded lymphoblasts were generated as described in the Materials and methods. After extensive washing cells were rested, then restimulated with 0·1 or 1 ng/ml IL-4 or 0·01 or 1 U/ml IL-2 alone or in combination with IL-12 (10 ng/ml). Cells were lysed after 10–20 min of stimulation and protein lysates were separated by SDS–PAGE and blotted on nitrocellulose membranes as described in the Materials and methods. Signalling proteins were detected in tyrosine phosphorylated and non-phosphorylated forms as described in Materials and methods. Blots (a) are representative of at least three independent experiments. Total STAT4 was not equally loaded in lane 4, thereby accounting for the slightly increased phospho-STAT4 in lane 4. Only pERK-1 (p44) is displayed, although ERK-2 (p42) was equally phosphorylated by IL-2 (1 U/ml). Optical densitometry for phospho-STAT3 was performed on all blots with lysates from IL-4/-12-stimulated cells (b) and from IL-2/-12-stimulated cells (c). Each column represents an average of three independent experiments. Error bars represent SEM. Statistical significances were calculated for relevant columns. P-values are as follows for (b): **P = 0·008 (IL-4 0·1 ng/ml versus IL-4/-12); **P = 0·01 (IL-12 10 ng/ml versus IL-4/-12). P-values are as follows for (c): **P = 0·001 (IL-2 0·01 ng/ml versus IL-2/-12); *P = 0·03 (IL-12 10 ng/ml versus IL-2/-12).

Discussion

The differential roles of IL-12 versus IL-4 have generally been attributed to mutually antagonistic effects in Th1 differentiation versus Th2 differentiation. However, this concept might be oversimplified and only appropriate for quiescent lymphocytes. Synergism between these cytokines has been described in the murine system for Th and NK cells.^29–31 However, the existence of this cross-regulatory pathway in humans is not known to our knowledge. In addition, the precise mechanisms of cross-regulation in Th cells still have to be elucidated, although flexibility in histone acetylation of cytokine loci might be important.⁴⁴

In this paper, we demonstrated that IL-12 and IL-4 have additive effects on proliferation, cytokine production and STAT3 phosphorylation in human T-lymphoblasts. Additive IL-12/IL-4-induced IFN-γ secretion has also been confirmed in a Th1 cell line and CD4⁺ lymphoblasts. Reduction of apoptosis in lymphoblasts did not parallel the proliferation rates in the IL-4/IL-12 and IL-7/IL-12 experiments, indicating that regulation of programmed cell death and cellular proliferation seem not to be tightly linked in activated human lymphocytes. This was also noted when cytokine signalling was inhibited by the protein tyrosine phosphatase CD45.³⁷ On the other hand, combining IL-2 with IL-12 reduced proliferation and induced cell death at certain concentrations. In contrast, cytokine secretion and STAT3 phosphorylation were also greatly enhanced with this combination despite its anti-proliferative/pro-apoptotic effects. In a study on a specific subset of CD8⁺ T cells (CD8⁺ CD18^bright T cells) with distinct signalling capacities, enhanced proliferation was observed when IL-2 and IL-12 were combined.⁴⁵ This differs from our findings in lymphoblasts, where IL-2 plus IL-12 reduced proliferation compared to IL-2 alone. These differences are most likely the result of the different activation state, cell type and experimental setting used in their study.

While IL-2/IL-12 strongly supported cytokine secretion, but not proliferation, IL-12 plus IL-4 enhanced both proliferation and IFN-γ production. These biological responses were associated with increased STAT3 tyrosine phosphorylations. We hypothesize that this additive effect on STAT3 activation could possibly be responsible for the enhanced functional capacities of Th1 cells under these conditions. While we have not tested STAT4 serine phosphorylation, this event was shown by Morinobu et al. not to be involved in proliferation⁴⁶ and so was unlikely to explain the additive effects seen in proliferation. We, however, cannot exclude the possibility that other newly discovered pathways, which are independent of STAT4, could also be engaged.⁴⁷

It has recently been shown that murine NK cells also secrete IFN-γ in response to IL-4 and IL-12.³¹ This has been associated with enhanced STAT4 tyrosine phosphorylation. The differences from our findings can most likely be attributed to the different cell types studied as well as to species-specific differences as it is known for other cytokine-activated STATs.⁴⁸ While STAT4 is well known for induction of IFN-γ in Th1 cells, IL-4 does not activate this member of the STAT family.³² Enhancement of IL-12-induced STAT4 phosphorylation directly by IL-4 would therefore be unlikely in our system and has not been observed. On the other hand, the precise role of IL-4-triggered STAT3 in activated lymphoblasts has not been investigated to our knowledge. After activation by IL-6 signalling, however, STAT3 has been demonstrated to initiate T-cell proliferation.³² A similar effect might be responsible for the observed co-operation between IL-4-induced and IL-12-induced proliferation described in this paper. It would therefore be interesting to examine IL-4 together with IL-6 or type I IFNs in the future.

Taking our data and those on murine Th cells together, the paradigm of Th1 and Th2 cytokine-specific cellular effects seems not to prevail in activated T cells during the cytokine-dependent maintenance phase of T-cell expansion. The data suggest that differentiation to Th2 cells under the influence of IL-4 predominates at the initiation of the immune response in quiescent T cells. Later, IL-4 can have pro-Th1 effects in activated human T cells. These results may have important implications for the design of therapeutic trials with IL-4: the fact that IL-4 does not restrict, but supports the expansion and cytokine secretion of activated T lymphocytes, clearly limits the therapeutic efficacy of IL-4 in ongoing Th1-associated inflammatory diseases. In support of this hypothesis, it has been recognized that the therapeutic efficacy of IL-4 in patients with rheumatoid arthritis, a chronic inflammatory Th1-associated disease, is at best limited.⁴⁹ In contrast, continuous IL-12 production is important – at least in certain autoimmune models in vivo – for optimal proliferation and cytokine production of Th1 cells in response to antigens.⁵⁰ From these perspectives, inhibition of IL-12 might be more promising for therapy of Th1-associated chronic inflammatory diseases than treatment with IL-4. This is supported by our findings that IL-12 – especially in combination with IL-2 and CD3 signals – induces maximal secretion of TNF-α from activated human lymphocytes. While therapy with TNF-α blockers like infliximab, etanercept or adalimumab has been proven to be very effective in many Th1-prone autoimmune diseases (for review see ref. 51 or many others), a substantial proportion of patients still do not respond. Here, blockade of IL-12 might represent a reasonable alternative, which can reduce TNF-α secretion at least from activated lymphocytes (as suggested indirectly by our data), but might also limit Th1 responses independent of TNF-α effects. First clinical trials with biological agents against IL-12 are in progress, e.g. for multiple sclerosis⁵² or rheumatoid arthritis. On the other hand, if correct timing and dosing of IL-4 administration are applied, this cytokine might still be a useful therapeutic agent for some Th1-like autoimmune diseases, as demonstrated in a recent trial on psoriasis.⁵³ Interestingly, although IL-4 was shown in this study to be of clinical benefit, increased levels of IFN-γ-secreting T cells have been noted in the peripheral blood of patients.⁵³ This phenomenon might also be partly explained by our in vitro data presented here.

In conclusion, we provided evidence that the Th2 cytokine IL-4 co-operates with the prototypical Th1 cytokine IL-12 for proliferation and cytokine production in human activated T lymphoblasts and Th1 cells. This was associated with enhanced phosphorylation of STAT3, but not of other STATs downstream of these two otherwise opposing cytokines. Our work underscores that the Th1/Th2 dichotomy may not hold true in activated lymphoblasts. Furthermore, our data support the existence of positive Th1/Th2 cross-regulation in humans and have important implications for the design of antagonizing cytokine therapies in any T helper cell driven, chronic disease state.

Abbreviations

ERK

extracellular signal-regulated kinase

γ_c-chain

common γ-chain

JAK

Janus kinase

STAT

signal transducer and activator of transcription

TYK

tyrosine kinase