The Role of Stat4 in Species-Specific Regulation of Th Cell Development by Type I IFNs

The discovery of polarized subsets of CD4⁺ T cells that differ in their cytokine secretion pattern and effector functions has provided a basis for the diversity of T cell-dependent immune responses (1). The two subsets of differentiated CD4⁺ T cells, referred to as Th1 and Th2 cells, protect against different microbial pathogens by producing cytokines able to mobilize different mechanisms of defense. Th1 cells secrete IFN-γ and TNF-β and are the mediators of phagocyte-dependent immune reactions, whereas Th2 cells that secrete IL-4 and IL-5 are responsible for phagocyte-independent host defense (2). Uncontrolled Th1 and Th2 responses can cause chronic inflammatory autoimmune diseases and allergies, respectively (3).

Th1 and Th2 cells develop from naive CD4⁺ T cells. The differentiation process is initiated by ligation of the TCR and directed by cytokines present during the initiation of a T cell response (4). IL-4 promotes Th2 development (5, 6, 7), whereas IL-12 is a potent inducer of Th1 cells (8, 9, 10, 11). IL-4 activates Stat6 in Th2 cells (12, 13), and IL-12 induces tyrosine phosphorylation of Stat4 in developing and differentiated Th1 cells (14, 15, 16, 17, 18, 19). The activation of these two STAT factors is essential for T cell subset development, since Stat6-deficient T lymphocytes fail to differentiate into Th2 cells in response to IL-4 (20, 21, 22), and the analysis of Stat4^−/−-deficient mice revealed that Stat4 is essential for Th1 cell differentiation (23, 24).

Regulation of the IL-12 signaling pathway is crucial for the development of Th cells. The IL-12Rβ2 subunit, a ligand-binding and signal-transducing component of the IL-12R (25), is expressed on human Th1, but not Th2, cells (18, 19). Triggering of the Ag receptor on naive T cells is sufficient for the initial expression of functional IL-12Rs, which are quickly lost during differentiation of human and mouse cells along the Th2 pathway (18, 19). In addition to TCR-mediated regulation, the IL-12Rβ2 subunit can be up-regulated by IL-12 (19).

IFNs have also been shown to play an important role in Th1 cell development. Recently, we have shown that human cord blood leukocytes primed in the presence of IFN-α, but not IFN-γ, develop into IFN-γ-producing Th1 cells even when cultured in the presence of IL-4 and neutralizing anti-IL-12 Abs. The Th1 cells generated in this manner express the IL-12Rβ2 mRNA and are responsive to IL-12 (19). In contrast to human cells, IFN-α/β are unable to induce Th1 development and prime mouse T cells to respond to IL-12 (26). In the present study we investigated the molecular basis for the species-specific effect of IFN-α/β on the selective induction of Th1-type immune responses. Our results indicate that IFN-α/β, by phosphorylating Stat4 in human, but not in mouse, T cells, promote species-specific induction of Th1-type immune responses. Furthermore, since T cell responsiveness to IFN-α is not lost during Th2 differentiation, we explored the effect of IFN-α-induced Stat4 activation on the phenotype of fully differentiated Th2 cells.

Human neonatal leukocytes were isolated from freshly collected, heparinized, neonatal blood by Ficoll-Paque (Pharmacia Biotech, Uppsala, Sweden) density gradient centrifugation. Monocytes were removed by one round of plastic adherence, and CD4⁺ T cells were isolated by negative selection with an Ab mixture and magnetic activated cell sorting, according to a protocol supplied by the manufacturer (CD4⁺ T cell isolation kit, Miltenyi Biotec, Bergisch Gladbach, Germany). Neonatal T cell preparations were >97% CD45 RA⁺ and >99% CD4⁺. Th1 and Th2 cell lines were generated by stimulating neonatal CD4⁺ T cells with irradiated (6000 rad), autologous monocytes and 2 μg/ml PHA (Wellcome, Beckenham, U.K.) in the presence of 2.5 ng/ml IL-12 (Hoffmann-La Roche, Nutley, NJ) and 200 ng/ml neutralizing anti-IL-4 Abs (18500D, PharMingen, San Diego, CA) for Th1 cultures, or 1 ng/ml IL-4 (PharMingen, San Diego, CA) and 2 μg/ml neutralizing anti-IL-12 Abs 17F7 and 20C2 (provided by M. Gately, Hoffmann-La Roche, Nutley, NJ) for Th2 cultures, respectively. To test the effect of IFNs on Th cell development, 1000 U/ml of IFN-α (Roferon A, Hoffmann-La Roche, Basel, Switzerland), IFN-γ (Hoffmann-La Roche, Basel, Switzerland), or IFN-β (Frone, Ares Serono, Geneva, Switzerland) was added at the time of priming to cultures containing IL-4 and neutralizing anti-IL-12 Abs (Th2-inducing conditions). Cells were washed on day 4 and expanded in complete RPMI 1640 medium (Life Technologies, Milan, Italy) supplemented with 5% FetalClone I (HyClone, Logan, UT), 2 mM l-glutamine, 1 mM sodium pyruvate, and 100 U/ml penicillin-streptomycin containing 100 U/ml IL-2 (Hoffmann-La Roche, Nutley, NJ).

Human PBMC from healthy donors were isolated from buffy coats by Ficoll-Paque density gradient centrifugation. Monocytes were depleted by two rounds of plastic adherence, and B cells were depleted by adherence to nylon wool as previously described (27). CD45RA⁺ T cells were isolated by two rounds of immunomagnetic negative selection with a mixture of mAbs as described previously (28). The purity of the CD3/CD45RA⁺ T cells using this procedure was typically >98% as determined by flow cytometry. Purified CD45RA⁺ T cells were stimulated with plate-bound anti-CD3 mAb (TR66) (29) in the presence of IL-12 (2.5 ng/ml) and neutralizing anti-IL-4 mAb (200 ng/ml), or IL-4 (1 ng/ml) for the generation of Th1 and Th2 lines, respectively. To test whether IFNs exert their functional effects directly on T cells, IFN-α, IFN-β, or IFN-γ (1000 U/ml) was added at the time of stimulation to cultures containing IL-4 (1 ng/ml). Cells were washed on day 3 and expanded in complete medium containing 100 U/ml IL-2.

T cell clones GL93 and D4.11 have been described previously (30, 31). Briefly, Lol p1-specific T cell clones were generated from PBMC of two Lol p1 allergic subjects as previously described (32). The three Lol p1-specific T cell clones that we selected for this study had a polarized cytokine profile. Two (D4.11 and E4.1) produced IL-4 but not IFN-γ and were categorized as Th2 clones, and one (ET3.22) that was able to produce IFN-γ but not IL-4 was categorized as a Th1 clone. Th1 clone GL93 is specific for the hepatitis virus δ Ag and has been described previously (30). To analyze cytokine production, 10⁵/0.2 ml T cells were stimulated with a combination of anti-CD3 mAb (0.1 μg/ml; CLB-T3/4E, CLB, Amsterdam, The Netherlands) and anti-CD28 mAb (1 μg/ml; PharMingen) in the presence or the absence of IL-12 (2.5 ng/ml) or IFN-α (1000 U/ml). Cell-free supernatants were collected after 24 h, and IFN-γ was measured by a specific sandwich ELISA following a protocol provided by the manufacturer (Genzyme, Cambridge, MA).

Single cell analysis of IFN-γ and IL-4 production was performed as described previously (33). Briefly, T cell lines were collected 7 days after priming and washed, and 10⁶ cells were restimulated with PMA (50 ng/ml; Sigma, St. Louis, MO) and ionomycin (1 μg/ml; Sigma) for 2 h at 37°C in complete medium. Brefeldin A (10 μg/ml; Sigma) was added to the cultures, and the cultures were incubated for an additional 2 h. Then the cells were fixed with 4% paraformaldehyde and permeabilized with saponin. Fixed cells were stained with anti-human IFN-γ-FITC (PharMingen, San Diego, CA) and anti-human IL-4-PE (PharMingen) following a protocol provided by the manufacturer and analyzed with a FACScan flow cytometer (Becton Dickinson, Mountain View, CA).

Lewis rats were immunized and boosted with purified recombinant huIL-12Rβ2-IgG1 fusion protein consisting of the extracellular domain of human IL-12Rβ2 and the Fc portion of human IgG1 (D. H. Presky et al., manuscript in preparation). Splenocytes from the immunized rat were isolated and fused to the SP2/0 mouse myeloma cell line, and Ab-producing hybridomas were produced by standard methods (34). Hybridoma supernatants were screened by flow cytometry using a human IL-12Rβ2-expressing Ba/F3 cell line (25), and the mAb 2B6 bound to the human IL-12Rβ2-expressing transfectants, but not the parental Ba/F3 cell line. Purified 2B6, a rat IgG2a, was prepared by sequential caprylic acid and ammonium sulfate precipitation as described previously (35).

Cells were stained with 1 μg/ml of purified rat anti-human IL-12Rβ1 mAb (2B10) (36), rat anti-human IL-12Rβ2 mAb (2B6), and isotype-matched control Abs at 4°C for 45 min. After being washed with cold FACS-buffer (PBS, 2% FCS, and 0.1% sodium azide), the cells were incubated as before with 2 μg/ml of biotinylated polyclonal anti-mouse or anti-rat IgG Ab (Jackson ImmunoResearch Laboratories, West Grove, PA). Cells were washed in FACS buffer and incubated with streptavidin-phycoerythrin (1/100; Jackson ImmunoResearch Laboratories). After two additional washes, the cells were resuspended in 0.5 ml of FACS buffer and analyzed with a FACScan flow cytometer (Becton Dickinson).

T cell lines generated from human cord blood leukocytes were harvested on day 9 after priming, and mouse T cells were harvested 6 days after stimulation. Cells (10⁷) were washed and incubated 30 min at 37°C in 1 ml of complete medium with or without IL-12 (2.5 ng/ml) or IFN-α, IFN-β, or IFN-γ (1,000 U/ml). Cells were washed once in PBS before lysing the cell pellet in 250 μl of IP buffer (10 mM Tris-HCl (pH 7.4); 150 mM NaCl; 1 mM EDTA (pH 8.0); 1 mM EGTA (pH 8.0); 1% Nonidet P-40; 0.25% sodium deoxycholate, 10 μg/ml aprotinin, leupeptin, and NaF; 1 mM 4-(2-aminoethyl)-benzene sulfonyl fluoride hydrochloride (AEBSF), and sodium orthovanadate). The lysate was incubated 30 min on a shaker at 4°C, and insoluble debris was removed by centrifugation (13,000 rpm, 4°C, 30 min). Stat4 was immunoprecipitated with rabbit polyclonal anti-Stat4 antisera (Santa Cruz Biotechnology, Santa Cruz, CA) and resolved by SDS-PAGE. Following transfer to nitrocellulose, blots were probed with anti-phosphotyrosine Ab 4G10 (Upstate Biotechnology, Lake Placid, NY), and immunoreactive bands were visualized using the ECL Western blotting system (Amersham, Italy), according to the company’s protocols. To control for equal protein loading, blots were stripped and reprobed with anti-Stat4 antisera.

Mel-14^high CD4⁺ T cells were purified from spleen and lymph nodes of DO11.10 TCR-transgenic mice (37) by positive selection using anti-CD4⁻FITC (PharMingen) and an anti-FITC multisort kit (Miltenyi Biotec, Bergisch Gladbach, Germany) followed by positive selection of Mel-14^high cells using anti-CD62L microbeads (Miltenyi Biotec). Naive CD4⁺ T cells (2.5 × 10⁵ cells/well) were stimulated with 0.3 μM OVA peptide 323–339 and mitomycin C-treated BALB/c splenocytes (5 × 10⁶ cells/well) as APC in a total volume of 2 ml in 24-well plates in the presence of 100 pg/ml IL-12 (provided by M. Gately, Hoffmann-La Roche, Nutley, NJ) and 10 μg/ml anti-IL-4 mAb (11B11) to promote Th1 phenotype development or in the presence of 20 ng/ml IL-4 and 10 μg/ml anti-IL-12 mAb 10F6 (provided by M. Gately, Hoffmann-La Roche, Nutley, NJ) to promote Th2 phenotype development as described previously (16, 38). Cells were expanded in complete medium supplemented with recombinant human IL-2 (10 ng/ml; provided by M. Gately, Hoffmann-La Roche, Nutley, NJ) and harvested on day 5 or 7. The Th phenotype was determined by intracellular staining for IFN-γ and IL-4 production after restimulation with PMA/ionomycin as described previously (38).

RNase protection assays to analyze IL-12Rβ2 transcripts in mouse Th cells were performed as previously described (39). Briefly, a 332-bp SacI DNA fragment from the mouse IL-12Rβ2 subunit (25) was subcloned into pSPUTK (Stratagene, La Jolla, CA). The construct was linearized with EcoRI, and radiolabeled antisense transcripts were synthesized with SP6 polymerase and a commercial kit, according to the manufacturer’s protocol (Promega, Madison, WI). RNA was extracted from cells using Ultraspec total RNA extraction reagent (Biotecx, Houston, TX) as previously described (40). The antisense RNA probes were hybridized to 10 μg of total RNA, and RNase protection assays were performed with a commercial kit (Ambion, Austin, TX), according to the company’s protocol. Products were resolved on 6% denaturing polyacrylamide gels, and the protected fragments were visualized by autoradiography. The radioactivity present in the protected fragments was also quantitated using a MolecularImager (Bio-Rad, Richmond, CA). An 18S RNA probe was used as a control for equal RNA loading.

Mouse Th1 and Th2 lines generated from DO.11.10-transgenic CD4⁺ T cell cultures were harvested 5 days after stimulation and incubated for 3 days with or without IFN-α (1000 U/ml). Cells were stained with biotinylated anti-clonotypic mAb KJ1-26 (41) followed by incubation with streptavidin-FITC (PharMingen) and anti-H-2K^d-PE (PharMingen) or an isotype-matched control mAb. Cells were analyzed with a FACScan flow cytometer (Becton Dickinson).

To examine the roles of IFN-α, IFN-β, and IFN-γ in the differentiation of human Th cell subsets, we generated T cell lines from CD4⁺, CD45RA⁺ T cells isolated from cord blood. Naive T cells were stimulated with irradiated allogeneic monocytes and PHA in the presence of IL-12 and neutralizing anti-IL-4 mAb or IL-4 and neutralizing anti-IL-12 mAb, respectively. IFN-α, IFN-β, or IFN-γ were added at the time of priming to cultures containing IL-4 and neutralizing anti-IL-12 mAbs (Th2-inducing conditions). The Th phenotype was determined 7 days after stimulation. After restimulation with PMA and ionomycin, IFN-γ and IL-4 productions were determined at the single cell level by intracellular staining (Fig. 1). Neonatal T cells primed in the presence of IL-12 and neutralizing anti-IL-4 mAbs developed mostly into Th1 cells (55% IFN-γ-producing cells vs 4% IL-4 producers and 8% double producers; Fig. 1,A), whereas priming in the presence of IL-4 and neutralizing anti-IL-12 mAbs induced development of a polarized Th2 population (0.5% IFN-γ-producing cells vs 23% IL-4 producers; Fig. 1,B). Addition of IFN-α (Fig. 1,C) or IFN-β (Fig. 1,D) at the time of priming under Th2-inducing conditions (IL-4 and anti-IL-12 mAbs) resulted in a marked increase in cells producing IFN-γ (11 and 20% compared with 0.5%) and a reduction (two- to fourfold) in cells producing IL-4 (12 and 5% compared with 23%). IFN-γ, in contrast, was less effective in inducing Th1 cell development (Fig. 1 E): only a marginal increase in cells producing IFN-γ (2.9% compared with 0.5%) and a minor decrease in IL-4-producing cells (20% compared with 23%) were observed in cultures that had been stimulated in the presence of IL-4, neutralizing anti-IL-12 mAbs, and IFN-γ. Priming in the presence of IFN-α, neutralizing anti-IL-4 mAb, and anti-IL-12 mAb also resulted in the development of Th1 cells, indicating that IFN-α is directly implicated in Th1 development rather than through the action of IL-4 (data not shown). These results confirm and extend our previous finding that cord blood leukocytes primed with PHA in the presence of IL-4, anti-IL-12 mAbs, and IFN-α secrete predominantly IFN-γ upon restimulation (19).

We next analyzed whether IFN-α/β promote Th1 development by acting directly on naive T cells or whether the presence of APC is required. Purified CD45RA⁺ T cells isolated from peripheral blood were stimulated with plate-bound anti-CD3 mAb in the presence of IL-12 and neutralizing anti-IL-4 mAb to induce Th1 development or in the presence of IL-4 to promote Th2 development, respectively. IFN-α, IFN-β, or IFN-γ was added at the time of CD3 stimulation to cultures containing IL-4. After 3 days of culture, the cells were washed and expanded for an additional 6 days in medium containing IL-2. The IL-12 responsiveness of the cell lines was used as a readout to determine the Th phenotype. As expected, IL-12 induced tyrosine phosphorylation of Stat4 in cultures that have been stimulated in the presence of IL-12 (Fig. 2,A, lane 2), whereas it was unable to activate this signaling pathway in cultures that had been primed in the presence of IL-4 (lane 4). Moreover, addition of IFN-α or IFN-β at priming even in the presence of the Th2-inducing cytokine IL-4 resulted in the development of IL-12-responsive T cells (lanes 6 and 8), whereas IFN-γ did not significantly increase IL-12 responsiveness under these conditions (lane 10). In addition, IL-12 induced tyrosine phosphorylation of Stat4 in cultures that had been stimulated in the presence of IFN-α and neutralizing anti-IL-4 mAb (data not shown). The ability of cells primed in the presence of IFN-α/β to signal in response to IL-12 correlates with IL-12Rβ2 expression. Surface IL-12R expression was monitored using mAbs directed against the β1 and β2 subunits of the IL-12R. Fig. 2 B shows that while IL-12Rβ1 was expressed to the same extent regardless of the cytokines present at priming, IL-12Rβ2 was selectively expressed by T cells primed in the presence of IL-12 or IFN-α/β. In conclusion, these data indicate that in addition to IL-12, IFN-α and IFN-β act directly on human T cells to promote the development of Th1-type cells.

Wenner et al. have shown that in contrast to human cells, IFN-α/β could not induce Th1 development and IL-12 responsiveness in mouse T cells (26). IFN-γ, however, induced IL-12-responsive mouse T cells by up-regulating the IL-12Rβ2 subunit (18). To address the apparent differences in IFN action in the two species, we compared side by side the functional effects of IFN-α/β on human and mouse Th subset differentiation. Mouse OVA-specific Th1 and Th2 lines were generated by stimulating purified naive CD4⁺ T cells from DO11.10 TCR-transgenic mice specific for OVA peptide 323–339 with OVA peptide presented by mitomycin C-treated BALB/c splenocytes in the presence of IL-12 and neutralizing anti-IL-4 mAb or IL-4 and neutralizing anti-IL-12 mAbs, respectively (16, 38). At priming, IFN-α or IFN-γ was added to cultures containing IL-4 and neutralizing anti-IL-12 mAbs. The Th phenotype of these lines was analyzed by staining for intracellular cytokines after stimulation with PMA and ionomycin. As expected, priming in the presence of IL-12 and neutralizing anti-IL-4 mAb induced Th1 phenotype development, whereas priming in the presence of IL-4 and neutralizing anti-IL-12 mAbs promoted differentiation of Th2 cells (data not shown). Addition of IFN-α or IFN-γ at priming to cultures containing IL-4 and neutralizing anti-IL-12 mAbs had no significant effect on the Th2 cytokine profile of the resulting population (data not shown).

We next analyzed expression of the IL-12Rβ2 subunit and IL-12 responsiveness in these cells. As Abs against the mouse IL-12Rβ1 and -β2 subunits are not yet available, transcripts encoding the IL-12Rβ1 and -β2 subunits were analyzed by RNase protection assays. Consistent with previous observations (18, 39), no significant differences in IL-12Rβ1 mRNA expression were detected in mouse Th cell lines (data not shown). Since IL-12 up-regulates the expression of IL-12Rβ2 transcripts in activated human and mouse T cells (19, 39), we used induction of IL-12Rβ2 mRNA by IL-12 as a readout for IL-12 responsiveness. IL-12Rβ2 is expressed in mouse Th1 cells (Fig. 3, lane 1) and culturing Th1 cells for 20 h in the presence of IL-12 strongly up-regulates the expression of IL-12Rβ2 (lane 2), indicating the presence of a functional IL-12R. T cells primed under Th2-inducing conditions in the absence (lane 3) or the presence (lane 5) of IFN-α did not express IL-12Rβ2 transcripts, and IL-12 treatment did not induce IL-12Rβ2 mRNA expression (lanes 4 and 6). In contrast, priming in the presence of IFN-γ resulted in the development of cells that express IL-12Rβ2 transcripts and are IL-12 responsive (lanes 7 and 8). These cells, when restimulated in the presence of IL-12, secreted IFN-γ (data not shown).

The data obtained with mouse T cells diverge from the findings in the human system, where priming of naive T cells in the presence of IFN-α/β results in the development of IFN-γ-producing cells that express the IL-12Rβ2 subunit (Figs. 1 and 2) (19).

To determine whether the distinct functional effects of IFN-α/β on human and mouse Th cell development originate from different uses of signaling molecules, we analyzed tyrosine phosphorylation of Stat4 in response to IFNs and IL-12 in differentiating human and mouse Th1 and Th2 cells. In agreement with our previous findings (19) and those of other laboratories (16, 17, 18), IL-12 induces tyrosine phosphorylation of Stat4 in human Th1, but not Th2, cells (Fig. 4 A,lanes 2 and 7). Interestingly, IFN-α and IFN-β were very efficient in activating Stat4 in human Th1 cells, and unlike IL-12, they also induced tyrosine phosphorylation of Stat4 in Th2 lines (lanes 3, 4, 8, and 9). IFN-γ, in contrast, did not induce activation of Stat4 in Th1 or Th2 cells (lanes 5 and 10).

We next examined whether IFN-α induces tyrosine phosphorylation of Stat4 in mouse Th1 and Th2 cells. DO11.10 Th1 and Th2 lines were exposed to IL-12, IFN-α, or IFN-γ. As shown in Fig. 4,B, Stat4 is activated by IL-12 in Th1 cells, but neither IFN-α nor IFN-γ induces Stat4 tyrosine phosphorylation in mouse Th1 or Th2 cells. Notably, both cell types respond to IFN-α by up-regulating class I molecules (Fig. 4 C), thus indicating that mouse T cells also express a functional IFN-α receptor. These results provide clear evidence for a species-specific difference in IFN-α-induced activation of the Stat4 molecule.

The analysis of Stat4- and Stat6-deficient mice has demonstrated the crucial role of STAT proteins in the initiation of transcription programs leading to the differentiation of naive T cells into polarized Th1 and Th2 subsets, respectively (20, 21, 22, 23, 24). An important question is whether activation of a specific STAT protein in differentiated Th1 and Th2 cells is sufficient to alter their cytokine secretion pattern. Since IFN-α, in contrast to IL-12, induces Stat4 phosphorylation in differentiating human Th1 and Th2 cells (Fig. 4,A) as well as in established human Th1 and Th2 clones (data not shown), we tested whether stimulation of Th2 lines and clones in the presence of IFN-α results in IFN-γ production. Restimulation of Th2 lines (Fig. 5,A) and clones (Fig. 5 B) in the presence of IFN-α resulted in only a modest increase in IFN-γ production. Consistent with a previous report, we detected high amounts of IFN-γ in the culture supernatant after restimulation of Th2 cells in the presence of IL-12 (42). Thus, activation of Stat4 does not appear to be sufficient to induce IFN-γ production by differentiated Th2 cells despite the fact that it has been shown to bind to a tandem site present in the first intron of the human IFN-γ gene and thus has been implicated in the regulation of the IFN-γ gene (43).

The functions of IFN-α/β and IFN-γ in inducing differentiation of human and mouse naive T cells to Th1-type effector cells have remained controversial. Stimulation of resting human T cells in the presence of IFN-α increases the frequency of IFN-γ-secreting CD4⁺ T cells (44), and allergen-specific T cell clones generated in the presence of IFN-α from the peripheral blood of atopic patients demonstrate a shift toward the Th0/1 phenotype (45). In addition, human cord blood leukocytes primed in the presence of IFN-α develop into IFN-γ-producing Th1 cells, even when cultured in the presence of the Th2-inducing cytokine IL-4 and neutralizing anti-IL-12 mAbs. The Th1 cells generated in this manner express the IL-12Rβ2 mRNA and are responsive to IL-12 (19). In contrast, IFN-γ, but not IFN-α, has been shown to be an important cofactor for Th1 development in mice (26, 46, 47). IFN-γ can rescue IL-12Rβ2 mRNA expression even in the presence of IL-4 and restore IL-12 responsiveness in early developing mouse Th2 cells (18).

The data presented in this report provide a molecular basis for the previously described effect of IFN-α/β in human Th1 development and help to explain the divergent function of type I IFNs when comparing human and mouse Th cell development. Consistent with our previous observation (19), IFN-α/β induce Th1 development in human naive T cells, even in the presence of the Th2-inducing cytokine IL-4 (Fig. 1). Importantly, this is a direct effect on T cells, as it occurs in the absence of APCs or exogenous factors other than IFN-α/β. Th1 cells generated by IFN-α/β in the absence of IL-12 are comparable to IL-12-generated Th1 cells in terms of cytokine production, IL-12 responsiveness, and cell surface expression of the IL-12Rβ2 subunit. Together, these findings provide clear evidence that human Th1-type immune responses can develop in the absence of IL-12.

To analyze the functions of IFN-α/β and IFN-γ in mouse T cell differentiation, we generated mouse OVA-specific Th1 and Th2 cell lines from DO11.10 TCR-transgenic mice (16, 37). Neither IFN-α nor IFN-γ, when added at priming to cultures containing IL-4 and neutralizing anti-IL-12 mAbs, had a significant effect on the cytokine secretion profile of the resulting cell lines (data not shown). These data are in agreement with previous findings (18, 26). Importantly, priming of naive T cells in the presence of IFN-γ, but not IFN-α, results in the expression of IL-12Rβ2 transcripts. These cells are IL-12 responsive, as demonstrated by IL-12-induced up-regulation of IL-12Rβ2 transcripts (Fig. 3) and Stat4 phosphorylation (data not shown). Restimulation of these cultures in the presence of IL-12 results in the production of IFN-γ (18) (data not shown). These data suggest that IFN-γ contributes to Th1 development in mice, whereas IFN-α has no effect, unlike previous suggestions made from the analysis of transgenic mice expressing IFN-α under the control of the rat insulin promoter (48). In conclusion, these data indicate that IFN-α/β act directly on human, but not mouse, T cells to induce Th1 development, whereas IFN-γ acts in both species in at least two ways; on the one hand, IFN-γ induces Th1 development by enhancing IL-12-production by human and mouse phagocytic cells (49), and on the other hand, it up-regulates the expression of functional IL-12Rs on CD4⁺ T cells, rendering the cells more responsive to IL-12 (18, 50). Our data indicate that the effect of IFN-γ on the induction of IL-12 responsiveness is much stronger in mouse than in human T cells, as the expression of IL-12Rβ2 is up-regulated in response to IFN-γ at much higher levels in mouse rather than in human T cells. Together, these findings indicate that in mice IFN-γ provides a signal for initiating Th1 development, the full progression of which requires IL-12-induced signaling. On the contrary, IFN-α/β are able to induce Th1 development in the absence of IL-12 in human, but not in mouse, cells.

To explain the divergent effects of IFN-α/β and IFN-γ on Th1 development in human compared with mouse T cells, we analyzed their ability to activate different signaling pathways. IFN-α and IFN-β induced a strong and rapid tyrosine phosphorylation of Stat4 in human Th1 and Th2 cells. Our findings are consistent with a recent report showing that IFN-α efficiently induces phosphorylation of Stat4 in mitogen-activated PBMC (51). Thus, both IL-12 and IFN-α/β activate Stat4, a protein critically involved in the generation of Th1 responses (23, 24). In contrast, IFN-α treatment does not induce tyrosine phosphorylation of Stat4 in Th1 and Th2 cells generated from DO11.10 TCR transgenic mice, confirming that IFN-α is unable to activate Stat4 in a mouse Th1 clone (15). These data indicate a striking and to date unparalleled species-specific activation of a STAT factor and provide a molecular basis for the finding that IFN-α/β are powerful inducers of Th1 development only in human cells. Given the importance of Stat4 activation for Th1 phenotype development, it is conceivable that activation of Stat4 in response to type I IFNs produced, for example, in the course of a viral infection may induce Th1 development in human T cells independently of IL-12.

The analysis of Stat4- and Stat6-deficient mice has revealed that activation of Stat4 and Stat6 is necessary for the development of Th1 and Th2 responses, respectively (20, 21, 22, 23, 24), but is activation of Stat4 and Stat6 sufficient to alter the cytokine secretion pattern characteristic for differentiated Th1 and Th2 cells, respectively? Our finding that IFN-α/β induce Stat4 tyrosine phosphorylation and DNA binding to a similar extent in human Th1 and Th2 cells might suggest that IFN-α/β, in addition to the induction of Th1 differentiation of naive T cells, have similar functional effects on established Th2 cells. Alternatively, activation of Stat4 may only be required at priming of T cells or at early stages of Th subset development to induce differentiation toward the Th1 phenotype. We have analyzed whether IFN-α treatment of established Th2 lines and clones can, at least transiently, restore IFN-γ production. Our finding that restimulation of Th2 lines and clones in the presence of IFN-α does not significantly increase IFN-γ secretion indicates that activation of Stat4, which is essential to initiate the genetic program leading to the development of Th1 immune responses, is not sufficient to induce IFN-γ production in differentiated Th2 cells. It will be interesting to analyze whether activation of Stat4 in Th2 cells via IFN-α/β can induce other Th1-specific functions. In conclusion, our findings suggest that the loss of IL-12 responsiveness along Th2 differentiation may be important to establish a hierarchy between the Th2- and Th1-inducing signals of IL-4 and IL-12, respectively, but is not required to maintain the polarized phenotype of Th2 cells. Once effector T cells have reached their developmental end point, other factors may be implicated in maintaining the polarized phenotype. Recent reports have demonstrated the differential expression of c-maf and GATA-3 transcription factors in Th1 and Th2 cells and their contributions to the maintenance of Th phenotypes (52, 53).

We thank M. Gately for IL-12 and anti-IL-12 mAbs, U. Gubler for preparing huIL-12Rβ2-IgG1 expression plasmids, P. Panina-Bordignon and V. Barnaba for providing T cell clones, P. Vigano for cord blood samples, and E. Bianchi for critical reading of the manuscript.