In the periphery, IL-18 synergistically induces the expression of the Th1 cytokine IFN-γ in the presence of IL-12 and the Th2 cytokines IL-5 and IL-13 in the presence of IL-2. Although the expression of these cytokines has been described in the thymus, their role in thymic development and function remains uncertain. We report here that freshly isolated thymocytes from C57BL/6 and BALB/c mice stimulated in vitro with IL-2-plus-IL-18 or IL-12-plus-IL-18 produce large amounts of IFN-γ and IL-13. Analysis of the thymic subsets, CD4−CD8− (DN), CD4+CD8+, CD4+CD8−, and CD4−CD8+ revealed that IL-18 in combination with IL-2 or IL-12 induces IFN-γ and IL-13 preferentially from DN cells. Moreover, DN2 and DN3 thymocytes contained more IFN-γ+ cells than cells in the later stage of maturation. Additionally, IL-18 in combination with IL-2 induces CCR4 (Th2-associated) and CCR5 (Th1-associated) gene expression. In contrast, IL-18-plus-IL-12 specifically induced CCR5 expression. The IL-2-plus-IL-18 or IL-12-plus-IL-18 effect on IFN-γ and IL-13 expression is dependent on Stat4 and NF-κB but independent of Stat6, T-bet, or NFAT. Furthermore, IL-12-plus-IL-18 induces significant thymocyte apoptosis when expressed in vivo or in vitro, and this effect is exacerbated in the absence of IFN-γ. IL-12-plus-IL-18-stimulated thymocytes can also induce IA-IE expression on cortical and medullary thymic epithelial cells in an IFN-γ-dependent manner. Thus, the combination of IL-2, IL-12, and IL-18 can induce phenotypic and functional changes in thymocytes that may alter migration, differentiation, and cell death of immature T cells inside the thymus and potentially affect the Th1/Th2 bias in peripheral immune compartments.
The development of T lymphocytes occurs primarily in the thymus, and this process is dependent, at least in part, on cytokines (1, 2, 3). The thymic epithelial cells (TEC)4 are the principal source of these immunoregulatory molecules (3), and in addition to IL-7, TEC can produce proinflammatory cytokines such as IL-1α, IL-1β, and TNF-α. Considering that no inflammatory reactions occur in the thymus, the function of these cytokines in the thymus can differ considerably from that in the peripheral immune system (1, 2, 3). Although known for their role in Th1 cell polarization and for their strong synergism in IFN-γ production, IL-12 and IL-18 have recently been reported to be expressed in the thymus and play a role during thymocyte proliferation, differentiation, and thymus involution (4, 5, 6, 7, 8). Interestingly, IL-18 is not strictly associated with Th1 responses and has also been shown to have the potential of inducing the Th2 cytokines IL-4 (9), IL-5, and IL-13 (10, 11). Our group previously reported that, in NK cell and T cells, IL-18 can synergize with IL-2 for the induction of IL-5 and IL-13 as well for IFN-γ (10). Therefore, IL-18 can induce both Th1 and/or Th2 responses depending on its surrounding cytokine milieu (11, 12).
Based on the fact that IL-2 (13), IL-12 (5, 6, 7), and IL-18 (4, 8) have been previously reported to be expressed and also to play a role in intrathymic T cell development, we investigated the effects of IL-18-plus-IL-12 or IL-18-plus-IL-2 on the induction of Th1 and Th2 cytokines from fresh thymocytes and on the specific T cell populations that arise during thymic ontogeny.
Typically, T cells require prior activation signals to be optimally responsive to cytokine stimulation. In this report, we demonstrate that freshly isolated thymocytes can produce large amounts of IFN-γ and IL-13 after stimulation with IL-18 in combination with IL-2 or IL-12, independent of other external stimuli (PMA, ionomycin, CD3, Con A, TCR engagement, etc.). This cytokine gene induction is effective even at concentrations of IL-12 and IL-18 lower than that present during certain microbial infections (14). Furthermore, IL-2-plus-IL-18 also induces CCR4 (Th2 phenotype) and CCR5 (Th1 phenotype) expression, whereas IL-12-plus-IL-18 triggers the expression of only CCR5. Considering that CCR4 and CCR5 are constitutively expressed in the thymus, the up-regulation of these chemokine receptors expression could alter the natural environment of this organ through changes in migration pattern (15), negative selection (16), as well as potentially releasing cells to the peripheral immune compartment with a Th1 or Th2 pre-established bias.
Interestingly, distinct thymocyte subsets exhibit specific cytokine production profiles, suggesting a different capacity for cytokine expression by these subsets during T cell maturation. Furthermore, using mice deficient in Stat family members, T-bet, the NF-κB p50 subunit, and NFAT, we were able to identify the transcription factors essential for the production of IFN-γ and IL-13 by thymocytes. IL-12-plus-IL-18 induces cell death in adult thymic organ culture (ATOC) but not in suspension cultures, demonstrating that the thymic microenvironment is important during apoptosis induction triggered by these cytokines. This effect is also seen in IL-12-plus-IL-18 cDNA-injected mice in vivo and is exacerbated in the absence of IFN-γ. Finally IL-12-plus-IL-18-stimulated thymocytes induce IA-IE expression in cortical and medullary epithelial cell lines in an IFN-γ-dependent manner, demonstrating that the synergistic effect of IL-2, IL-12, and IL-18 may not only influence Th1/Th2 polarization and apoptosis induction in thymocytes but also may influence the thymic microenvironment in an indirect way.
Materials and Methods
Cytokines and Abs
Recombinant human IL-2 was obtained from Hoffmann-La Roche. rIL-18 was obtained from Medical and Biological Laboratories, and IL-12 was obtained from PeproTech. FITC or PE anti-mCD8, PE or PE-Cy5 anti-mouse (m)CD4, PE anti-mIFN-γ, PE anti-mNK1.1, PE-Cy5 anti-mCD44, allophycocyanin or PE-Cy7 anti-mNK1.1, allophycocyanin-Cy7 anti-mCD25, and FITC anti-mMHC-class II Abs were all purchased from BD Pharmingen.
The IL-2 and IL-13 neutralizing Abs were purchased from BD Pharmingen and R&D Systems, respectively. IL-12 neutralizing Ab was a gift from Dr. A. Sher (National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD). Anti-IL-2 was a gift from Dr. J. Cote-Sierra (National Institute of Allergy and Infectious Diseases, National Institutes of Health).
C57BL/6 mice (B6), IFN-γ−/−, Stat6−/−, T-bet−/−, p50−/−, and BCL-2 transgenic mice on a B6 background, Stat4−/− mice on a 129 background, NFAT1 and NFAT4−/− mice on a BALB/c background, and their control littermates were used in this study. All gene-deficient and transgenic mice were backcrossed for >10 generations with B6, 129, or BALB/c mice and maintained under specific pathogen-free condition and used for experiments at 4–6 wk of age.
Hydrodynamic injection of IL-12 and IL-18 cDNA
The hydrodynamic gene transfer procedure was conducted as described previously (17). In brief, animals were separated into four different groups and injected by tail vein with the following: 15 μg of empty vector control cDNA; 5 μg of IL-12 cDNA (pscIL-12, p40-p35 fusion gene) plus 10 μg of IL-18 cDNA (pDEF pro-IL-18); IL-12 cDNA alone or IL-18 cDNA alone in 1.6 ml of sterile 0.9% sodium chloride solution. All of the plasmids are driven by human elongation 1-α promoter. Thymuses were harvested 24, 48, 72, and 96 h postinjection.
Cell cultures and cell lines
Thymuses, spleens, and lymph nodes (LNs) were smashed, washed, and resuspended in supplemented medium (RPMI 1640 supplemented with 10% heat-inactivated FBS, 100 U/ml penicillin G sodium, 100 μg/ml streptomycin sulfate, 2 mM l-glutamine, 1 mM sodium pyruvate, 1× essential amino acids, and 10 mM 2-ME). Splenocyte cell suspensions were depleted of red cells by treatment with ACK lysis buffer, washed, and resuspended in supplemented medium. Cells were counted and cultured at 2.5 × 106 cells/ml at 37°C with medium alone or in the presence of IL-2 (100 IU/ml), IL-12 (100 ng/ml), IL-18 (100 ng/ml), IL-2-plus-IL-18, or IL-12-plus-IL-18 for 48 h, unless otherwise specified.
The ATOC methods used have been described in detail elsewhere (18). Briefly, thymi were cut into 1-mm3 fragments and cultured on top of Nucleopore filters (0.8 μM; Corning) at the air-medium interface using plastic cylinders instead of sponges. Cultures were maintained in a fully humidified incubator at 37°C with 5% CO2 for 24 h. Thymic lobe fragments were stimulated with the cytokines for 24 h and then dispersed into a cell suspension, counted, and used in the apoptosis assays.
Total RNA was isolated using a single-step phenol/chloroform extraction procedure (TRIzol; Invitrogen Life Technologies). For the RNase protection assay (RPA), 5 μg of total cytoplasmic RNA were analyzed using the RiboQuan kits (BD Pharmingen) and [33P]UTP-labeled riboprobes as described previously (21).
Supernatants from cells treated in vitro were collected and assayed for cytokine production by ELISA. The kits used were as follows: mouse IFN-γ and mouse IL-13 (R&D Systems).
Flow cytometry analysis and cell sorting
Up to six-color analysis was performed on a BD Biosciences FACSort flow cytometer as previously reported (22). Anti-mouse CD16/CD32 mAb (2.4G2; BD Pharmingen) was used to block the nonspecific binding. To detect intracellular expression of cytokines, cells were cultured with the different stimuli, and 5 μg/ml brefeldin A (BD Pharmingen) was added during the last 8 h. Cells were then stained for surface markers, washed, and fixed with Cytofix/Cytoperm buffer (BD Pharmingen) for 15 min at 4°C. Cells were washed with Perm Wash buffer (BD Pharmingen) and incubated with the anti-mouse IFN-γ Ab or isotype-matched Ab for 30 min at 4°C. Following two washings, cells were analyzed in the flow cytometer.
For flow cytometry cell sorting, cells were stained with CD4-PE and CD8-FITC, and CD4−CD8− double-negative (DN), CD4+CD8+ double-positive (DP), CD4+CD8− single-positive (CD4+), and CD4−CD8+ single-positive (CD8+) thymocytes were sorted to a purity of >97–99% using a MoFlo sorter (DakoCytomation).
Testing for significance of differences was assessed by Student’s t test using Microsoft Excel statistical analysis computer program.
Thymocytes can produce large amounts of Th1 and Th2 cytokines
It has been previously reported that, in NK and T cells, the combination of IL-18 with IL-12 or IL-2 synergizes to produce Th1 and Th2 cytokines, respectively (10, 11). In our experiments, injection with a mammalian expression plasmid for IL-12 or IL-18 resulted in high circulating levels of these cytokines whether injected individually or in combination. Interestingly, we found that thymocytes from mice injected with IL-12-plus-IL-18 cDNA respond by up-regulating IFN-γ expression in vivo (Fig. 1, A and B). IFN-γ expression is low at 24 h poststimulation (3-fold increase over control) and increases gradually over time, reaching levels almost 50 times higher than control at 96 h postinjection (Fig. 1, A and B). These findings led us to investigate the capacity of immature T cells in the thymus to produce Th1 and Th2 cytokines. We developed an in vitro model where thymocytes were directly stimulated with IL-2-plus-IL-18 or IL-12-plus-IL-18 and assayed for IFN-γ, IL-13, and IL-5 mRNA expression and compared with those from secondary lymphoid organs such as LNs and spleen (Fig. 1,C). As shown in Fig. 1,C, the combination of IL-2-plus-IL-18 synergizes to induce IFN-γ (Th1), IL-13, and IL-5 (Th2) in all tissues analyzed, with the Th2 cytokines particularly strongly expressed in thymus compared with spleen and LNs (C and D). A neutralizing anti-IL-12 Ab added to the cultures did not change the levels of cytokines produced (data not shown), indicating the synergy between IL-2-plus-IL-18 was not due to the expression of endogenous IL-12. Although IL-12-plus-IL-18 stimulation induces some IL-13 but not IL-5 expression (Fig. 1, C and D), it strongly induces a Th1 response with high levels of IFN-γ mRNA expression in all three tissues analyzed (C and E). Furthermore, the use of a neutralizing anti-IL-2 Ab confirmed that endogenous IL-2 production was not responsible for the expression of IL-13 in the IL-12-plus-IL-18-treated thymic cultures (data not shown).
IL-18 rapidly and differentially synergizes with IL-2 and IL-12 for the production of Th1 and Th2 cytokines in thymocytes
Next, we evaluated how rapidly fresh thymocytes are able to express Th1 and Th2 cytokines upon stimulation with IL-18 in combination with IL-2 or IL-12. As can be seen in Fig. 2, thymocytes produced increasing and significant amounts of IFN-γ and IL-13 mRNA (A) and protein (B and C, respectively) up to 48 h poststimulation. We did not continue the kinetic analysis beyond 48 h because the viability of DP cells is significantly compromised at later time points. We also observed that IL-18 synergizes with both IL-2 and IL-12 to produce both IL-13 and IFN-γ. IL-18-plus-IL-2 treatment is a stronger stimulus for IL-13 and uniquely for IL-5 expression; in contrast, IL-18-plus-IL-12 provides a more rapid and potent stimulus for IFN-γ production (Fig. 2, A–C).
Evaluation of the sensitivity of thymocytes to these combinations of cytokines revealed that thymic cells produced a significant amount of IFN-γ (1000 pg/ml) after IL-18 and IL-12 stimulation with concentrations as low as 1 ng/ml (Fig. 2 D) and IL-13 (10 pg/ml) after addition of 1 ng/ml IL-18 plus 1 U/ml IL-2 (E). Interestingly, levels of IL-12 that range from 0.4 to 15 ng/ml and 1 ng/ml IL-18 can be found in the sera of mice after infection with certain pathogens (14), thus indicating that the sensitivity of thymocytes to cytokine stimulation that we have observed is physiologically relevant.
Thymocytes from BCL-2 transgenic mice, which are more resistant to apoptotic death, produced equivalent amounts of IFN-γ and IL-13 as control mice after IL-2-plus-IL-18 or IL-12-plus-IL-18 stimulation (data not shown). Thus, our results do not represent a selection for thymocyte subsets that preferentially survive under these conditions.
Subpopulations of thymocytes have different and/or exclusive capacities for producing IFN-γ and IL-13 after IL-18-plus-IL-2 or IL-18-plus-IL-12 stimulation
To determine whether the production of IFN-γ and IL-13 comes from thymocytes themselves or from thymic stromal cells (epithelial cells, dendritic cells, macrophages, fibroblast, etc.) present in the bulk thymocytes cultures, we separated thymocytes into four groups based on their expression of CD4 and CD8 and investigated the cytokine expression after IL-2-plus-IL-18 or IL-12-plus-IL-18 stimulation. Fig. 3 shows that DN, CD4+, and CD8+ thymocytes were able to produce IFN-γ to different extents (A). However, only DN and CD4+ cells are able to produce IL-13 (Fig. 3 B).
The synergy between IL-12 and IL-18 in IFN-γ production is partially explained by the mutual up-regulation of the cytokine receptors (12, 23). Thus, we evaluated the surface expression of IL-2R, IL-12R, and IL-18R in the different subsets of thymocytes and found that, in all cases, the expression of these receptors was very low and did not change during any of the treatments used in the experiments. However, at the RNA level, we detected up-regulation of IL-12Rβ1 and -β2 by IL-18, and IL-18R by IL-12 (data not shown), a result that has been reported previously in T cells (11, 23).
Differential capacity for IFN-γ production during T cell development
The high expression of IL-13 and IFN-γ observed in the DN subset (Fig. 3, A and B) could be due to the presence of a small population of NKT cells present in the thymus. Therefore, we evaluated the percentage and type of IFN-γ-producing cells in the different subsets of thymocytes after IL-18-plus-IL-12 stimulation (Tables I and II). A high percentage of NKT cells (68%) was responsible for the production of IFN-γ in the DN subset; however, 34% of the remaining DN thymocytes could still produce IFN-γ (Table I). Interestingly, six-color flow cytometry analysis revealed that, in the DN subset, cells in the DN2 and DN3 stages show higher numbers of IFN-γ+ cells by intracellular staining than in the DN1 and DN4 stages (Table II). Moreover, the number of thymocytes with the capacity to produce IFN-γ decreases from DN (34%) to a very low number when they become DP (5%), and then recovers to some extent when they mature to CD8+ (22%) and CD4+ (10%) (Table I).
|.||DN .||DP .||CD4+ .||CD8+ .||NKT .|
|.||DN .||DP .||CD4+ .||CD8+ .||NKT .|
|.||DN1 (CD44+CD25−) .||DN2 (CD44+CD25+) .||DN3 (CD44−CD25+) .||DN4 (CD44−CD25−) .|
|.||DN1 (CD44+CD25−) .||DN2 (CD44+CD25+) .||DN3 (CD44−CD25+) .||DN4 (CD44−CD25−) .|
Both CCR4 and CCR5 are expressed after IL-2-plus-IL-18 stimulation but only CCR5 after IL-12-plus-IL-18 stimulation in DN, CD4+, and CD8+ cells
It has been reported that Th1 cells preferentially express the chemokine receptor CCR5 (24), whereas Th2 cells express CCR4 (24). To evaluate whether the treatment with the combined cytokines results not only in the ability to produce IFN-γ and IL-13 but also a chemokine receptor pattern that correlates with the Th1 and/or Th2 profiles, we performed a RPA to evaluate the expression of CCR4 and CCR5 mRNA. As shown in Fig. 4,A, IL-2-plus-IL-18 induces the expression of both CCR4 and CCR5, whereas IL-12-plus-IL-18 induces the expression of only CCR5 in thymus, LNs, and spleen cells. The expression of these chemokine receptors is not due to the presence of IL-13 or IFN-γ in the cultures because thymocytes from IFN-γ−/− mice or the addition of IL-13 neutralizing Ab does not change the chemokine receptors pattern seen in C57BL/6 mice after stimulation with the cytokine combinations (Fig. 4,B and data not shown). Moreover, although CCR4 can be expressed by DN, CD4+, and CD8+ cells (Fig. 4 C), CCR5 can be predominantly expressed in DN and CD4+ cells and weakly in CD8+ cells (C).
IL-12-plus-IL-18 stimulation induces significant thymocyte death in vivo and in vitro: protective role of IFN-γ
It has been described that, in the presence of IFN-γ, TCR engagement can cause apoptosis in CD4+ and CD8+ thymocytes (25). To evaluate whether the large amounts of IFN-γ produced in the IL-12-plus-IL-18 thymic cultures correlates with increased cell death, we generated thymic suspension cultures and ATOC where the architecture of the organ is disrupted or intact, respectively, and analyzed apoptosis by annexin V/7-aminoactinomycin D (7AAD) staining.
Fig. 5,A shows that no significant differences in cell death were observed when we compared thymocytes stimulated with individual or combined cytokines in suspension cultures with the control group. However, when apoptosis was assayed in ATOC, we found that IL-12-plus-IL-18 treatment results in significantly increased cell death compared with control cultures (75 vs 55%, respectively; p < 0.05) (Fig. 5,A). Analysis of apoptosis in the different subset of thymocytes in the ATOC demonstrated that IL-12-plus-IL-18 induces cell death preferentially in DP cells with 22% more annexin V+ cells than in the control ATOC (Fig. 5,B). This effect can also be seen in vivo because we observed a marked reduction in the absolute cell number in the thymuses of mice injected with IL-12-plus-IL-18 cDNA compared with control cDNA-injected mice (Fig. 5,C). Moreover, the loss of cells in the IL-12-plus-IL-18-injected mice correlated with an increase in the percentage of annexin V+ cell in the bulk population (Fig. 5 D) that corresponds to a selective apoptosis in the DN and DP subsets (E).
To evaluate the role of IFN-γ in cell death induction in vivo after IL-12-plus-IL-18 cDNA injection and in vitro in ATOC, we performed our studies in IFN-γ-deficient mice. Fig. 5,C shows a significant reduction in the absolute cell numbers in thymocytes coming from either C56BL/6 or IFN-γ−/− mice after cDNA injection. Accordingly, an increased percentage of apoptotic cells can be seen in both strains of mice after IL-12-plus-IL-18 stimulation in ATOC (data not shown) or after IL-12-plus-IL-18 cDNA injection (Fig. 5,D). Finally, IL-12-plus-IL-18 stimulation induces apoptosis preferentially in DP and DN cells in vitro and in vivo in the presence (Fig. 5, B and E, respectively) or absence of IFN-γ (data not shown). Μoreover, IFN-γ seems to play some type of overall protective role because only ∼60% of IFN-γ−/− mice survived 5 days post-IL-12-plus-IL-18 cDNA injection compared with 100% survival in the cDNA-injected control mice (data not shown).
IL-12-plus-IL-18-stimulated thymocytes induce IA-IE expression in cortical and medullary epithelial cell lines in an IFN-γ-dependent manner
In our model, we evaluated whether thymocytes that have been exposed to the different combinations of cytokines have the potential to affect cells from the thymic microenvironment by modulating MHC class II expression in different TEC as has been previously described for human TEC (26). Fig. 6,A shows that a cortical (ANV 41.2) and a medullary (TE-71) TEC lines, when cocultured with thymocytes in the presence of IL-12-plus-IL-18, up-regulate IA-IE expression; moreover, this effect does not need cell-cell interaction because supernatants from IL-12-plus-IL-18-stimulated thymocytes induced IA-IE expression in both ANV 41.2 or TE-71 cells (Fig. 6,B). Finally, IA-IE up-regulation seems to be mediated by IFN-γ produced by thymocytes after IL-12-plus-IL-18 stimulation because ANV 41.2 and TE-71 do not express IL-12 or IL-18 receptors on their cell surface (data not shown) and IFN-γ-deficient thymocytes do not have the capacity to induce IA-IE in these TEC lines (Fig. 6 C).
IFN-γ and IL-13 but not IL-5 expression is Stat4 dependent and Stat6/T-bet independent in response to IL-18-plus-IL-2 or IL-18-plus-IL-12 stimulation
To gain insight into the transcription factors that participate in the expression of IFN-γ, IL-13, and IL-5 in the thymus, we stimulated thymocytes with IL-18-plus-IL-2 or IL-18-plus-IL-12 in Stat4, Stat6, and T-bet knockout (KO) mice. As previously reported, Stat4 activation correlates with the capacity to promote IFN-γ production in response to IL-12 signaling (12, 27). Stat6 is essential for chromatin remodeling at the Th2 cytokine loci and for the production of IL-4, IL-5, and IL-13 (27, 28, 29). T-bet, a member of the T-box family transcription factors, has been shown to be required for the induction of IFN-γ by CD4+ cells (12, 27). However, our analysis demonstrated that neither Stat6 nor T-bet is required for the expression of IL-13, IL-5, or IFN-γ after IL-18-plus-IL-2 or IL-18-plus-IL-12 stimulation because levels of expression in the thymocytes from these KO mice were comparable with those observed in the wild-type mice (data not shown). However, thymocytes from Stat4−/− mice showed a profound decrease in IFN-γ and IL-13 at the mRNA (Fig. 7,A) and protein level (B and C) in response to IL-18-plus-IL-12. Interestingly, when the cells were stimulated with IL-18-plus-IL-2, IFN-γ was considerably reduced, whereas IL-13 was only partially affected in the Stat4−/− mice (Fig. 7).
NF-κB but not NFAT1 or NFAT4 is required for both IL-13 and IFN-γ production
It has been reported that NF-κB and NFAT have an important role in regulating IFN-γ and IL-13 gene expression (11, 27, 30, 31, 32). To evaluate whether these transcription factors are involved in IFN-γ, IL-13, and IL-5 expression after IL-2-plus-IL-18 or IL-12-plus-IL-18 stimulation, we cultured fresh thymocytes from NF-κB p50−/−, NFAT1−/−, NFAT4−/−, and control littermate mice for 48 h with the combined cytokines. As shown in Fig. 8, whereas IFN-γ (B) and IL-13 (C) protein expression is considerably blocked in thymocytes from p50−/− mice after both IL-2-plus-IL-18 or IL-12-plus-IL-18 stimulation, the mRNA levels (A) are only partially affected. However, IL-5 mRNA expression (Fig. 8 A) is completely absent in IL-2-plus-IL-18 cultures. In contrast, thymocytes from NFAT1 or NFAT4−/− mice expressed similar levels of IFN-γ and IL-13 RNA and protein when compared with control mice (data not shown), indicating that NFAT1 and -4 do not play a role in the expression of these cytokines in response to IL-12-plus-IL-18 or IL-2-plus-IL-18.
Recently, it has been demonstrated that the proinflammatory cytokines IL-12 and IL-18 are expressed in the thymus and may play a role during thymic development (4, 5, 6, 7, 8, 13). Our initial experiments using injections of cytokine-inducing cDNAs demonstrated that thymocytes in vivo respond to circulating levels of IL-12 and IL-18 by expressing IFN-γ (Fig. 1Α). These findings are quite interesting considering the thymus is a closed organ, and the entrance of macromolecules including cytokines into the thymus is limited (3).
In vitro, we found that thymocytes rapidly and strongly respond to IL-18 in combination with IL-12 or IL-2 to produce the Th1 cytokine IFN-γ and the Th2 cytokines IL-5 and IL-13. Interestingly, the levels of IFN-γ produced by thymocytes in vitro under these conditions are comparable with those produced by spleen and LN cells in vitro (data not shown). Moreover, the Th2 cytokines, IL-5 and IL-13, are more highly expressed in the thymus than in spleen or LNs. These findings surprised us because thymocytes have been reported to be poor cytokine producers themselves, unless they receive activation signals (e.g., PMA, ionomycin, Con A, anti-CD3 Abs) (3, 13, 33).
The production of IFN-γ by thymocytes in response to IL-2-plus-IL-18 or IL-12-plus-IL-18 is relevant because IL-2, IL-12, and IL-18 are present in the thymus (4, 5, 6, 7, 8) and because IFN-γ plays multiple roles in thymic development (34, 35, 36). IFN-γ can modulate TEC interaction with thymocytes and induce Ag presentation in TEC (34, 36). Moreover, as an effector molecule, IFN-γ induces NO in thymic stromal cells restricting thymocyte development by inducing apoptosis (35). In our model, we found that thymocytes exposed to IL-12-plus-IL-18 produce IFN-γ that in turn can mediate IA-IE expression in cortical and medullary TEC. This is important considering that MHC class II expression by TEC is essential during negative and positive selection (37, 38).
Separation of DN, DP, CD4+, CD8+ cells provided a clearer insight regarding the ability of these subsets to express cytokines following specific stimulation. Our data revealed that DN cells are the main source of IFN-γ and IL-13 after IL-2-plus-IL-18 or IL-12-plus-IL-18 stimulation. However, it has been described that small populations in the thymus, including NKT cells and γ-δ thymocytes, can strongly express cytokines (39). As these cells are present in the thymus in an extremely low percentage, their purification is a significant technical challenge. To overcome this problem, we stimulated thymocytes with IL-12-plus-IL-18 and analyzed IFN-γ expression by intracellular staining. The data demonstrated that, in the DN subset, NKT cells may in fact highly contribute to the production of IFN-γ, because 68% were IFN-γ+. Remarkably, we found 34% of DN NK1.1− cells also produce IFN-γ, especially in the intermediate stage of DN T cell development, when CD25 is expressed (DN2 and DN3, 40 and 30%, respectively) compared with when it is absent (DN1 and DN4 stages, 22 and 23%, respectively). The number of IFN-γ+ cells decreased considerably in the DP stage (5%) as has been previously reported (40). Interestingly, at the final stage of T cell maturation, IL-12-plus-IL-18 induces more CD8+ cells to produce IFN-γ than CD4+ cells (20 vs 10%, respectively).
It has been proposed that IL-12 influences intrathymic T cell development through induction of apoptosis in immature thymocytes (5). In our model, even when the baseline of apoptosis observed in ATOC in unstimulated cultures is very high but expected based on a previous report describing ATOC (18), the IL-12-induced apoptosis observed is significantly increased when IL-18 is added (Fig. 5 A). The apoptotic effect is more prominent after IL-12-plus-IL-18 in vivo expression where more than half of thymocytes (especially DN and DP cells) are lost. Moreover, the normal thymic environment seems to be essential, because when the tissue is disrupted and cells cultured in suspension, thymocytes still produce large amounts of IFN-γ but the apoptotic effect is completely lost.
Because IFN-γ produced after CD3 stimulation has been associated with an increased apoptotic effect in human medullary thymocytes (25), we questioned whether the IL-12-plus-IL-18 effect on increased cell death correlates with augmented levels of IFN-γ. The in vitro (ATOC) and in vivo (cDNA injections) experiments using IFN-γ-deficient mice suggest that IL-12-plus-IL-18 can induce apoptosis in the absence of IFN-γ (Fig. 5,C). IFN-γ seems to have a protective effect considering that the 60% of thymocytes lost after IL-12-plus-IL-18 treatment in C57BL/6 mice (NS vs IL-12-plus-IL-18, p = 0.001) increases to 80% in IFN-γ−/− mice (NS vs IL-12-plus-IL-18, p = 0.0001) (Fig. 5 C). This phenomenon might be explained based on previous reports where it has been proposed that IFN-γ can induce IL-15 which in turn mediates an antiapoptotic effect (41). This regulatory mechanism exerted by IL-15 may be more systemic, because ∼40% of IFN-γ−/− mice died after 5 days post-IL-12-plus-IL-18 injection compared with 100% survival in control C57BL/6 mice (data not shown).
Over the past years, important progress has been made in understanding the molecular events resulting in strong IFN-γ expression in mature T cells in response to IL-12 and IL-18. Considering that the transcription factors that participate in cytokine expression by mature T cells could be differentially expressed in thymocyte subsets, we considered it important to evaluate how this synergy works in immature T cells.
IL-12 induces phosphorylation of specific proteins known as Stats, and it has been demonstrated that the Stat4 is linked to IL-12R signaling and is critical for IFN-γ expression in T cells mediated by IL-12 alone or in synergy with IL-18 (42). In thymocytes, we confirmed that Stat4 is an important mediator in the expression of IFN-γ and surprisingly found that Stat4 is required for IL-13 expression by IL-12-plus-IL-18 stimulation as well as in IL-2-plus-IL-18-treated cultures, in the absence of IL-12.
In contrast to Stat4, our studies in T-bet- and Stat6-deficient mice demonstrate that IFN-γ and IL-13 expression is independent of T-bet and Stat6 in response to IL-2-plus-IL-18 or IL-12-plus-IL-18 (data not shown). Because the role of Stat6 in Th2 differentiation is well established (27, 28, 29), it is interesting to note that IL-13 and IL-5 expression after IL-2-plus-IL-18 treatment represents a novel way to induce these two Th2 cytokines in a Stat6-independent manner.
Our group and other laboratories have reported an important role for the transcription factor NFAT family in the regulation of IFN-γ (30, 32) and IL-13 gene expression (31). NFAT1 has been reported to be a major regulator of IFN-γ in vivo (30) and important during IL-13 transcription after CD3 stimulation in vitro (43). In contrast, NFAT4 has been proposed as an enhancer of IFN-γ and suppressor of Th2 cytokines (44).
Surprisingly, thymocytes from NFAT1- or NFAT4-deficient mice expressed similar levels of IFN-γ and IL-13 in response to IL-2-plus-IL-18 or IL-12-plus-IL-18 compared with control mice (data not shown). Although these results rule out a role for NFAT1 and NFAT4 in regulating IFN-γ and IL-13 expression by thymocytes, a potential role for other NFAT family members awaits further analysis.
NF-κB is an important transcription factor induced following IL-18 signaling and is known to be involved in the regulation of IFN-γ gene expression (11, 27, 32). In our analysis, thymocytes from p50 KO mice demonstrated a dramatic decrease in IFN-γ and IL-13 expression especially at the protein level, indicating a role for the p50 subunit in regulating cytokine gene expression.
Due in part of the physical proximity of the il-4, il-13, and il-5 genes, chromatin remodeling during T cell differentiation may permit transcription of all these genes simultaneously (45). However, it is intriguing to note that, following IL-2-plus-IL-18 stimulation, we observed expression of IL-5 and IL-13 but not IL-4. The control of the il-4 loci is tightly regulated at the chromatin level (45). To address this question, we evaluated the levels of histone acetylation at sites shown previously to correlate with IL-13 and IL-4 expression, respectively (46), in either nontreated or IL-2-plus-IL-18-treated bulk thymocyte cultures or in purified CD4+ cells. The results indicated equivalent levels of histone acetylation that were not affected by the cytokine treatments (data not shown). It remains to be determined whether other mechanisms, such as DNA methylation, are responsible for the differential gene expression observed in our experiments.
An essential event during T cell development is the continuous migration of cells through the thymus and to the peripheral compartments of the immune system (16). This process is mediated in part by the interaction of chemokines with their receptors (15, 16, 47). In this study, we found that stimulation of thymocytes with IL-2-plus-IL-18 induces the expression of CCR4 and CCR5, whereas IL-12-plus-IL-18 triggers the expression of only CCR5. Although CCR5 has been described to be weakly expressed in a minor fraction of thymocytes and is nonfunctional (47), the up-regulation of CCR4 expression by IL-2-plus-IL-18 could be of importance considering that CCR4 is responsible for the retention of mature cells in the thymic medulla (15). In addition, CCR4 coexpression with CD30 has been implicated in the process of negative selection (16). In the peripheral immune tissues, CCR4 is the major chemokine receptor functionally expressed on in vitro-polarized Th2 T cells (24), whereas CCR5 is preferentially expressed on cells with a Th1 polarization (48). Overall, the predominant expression of CCR5 and CCR4 in DN and CD4+ cells correlates well with the higher expression of Th1 and Th2 cytokines in these two subsets, indicating that DN and CD4+ thymocytes are more sensitive to express a Th1 or Th2-like phenotype after IL-12-plus-IL-18 or IL-2-plus-IL-18 stimulation, respectively (Figs. 3 and 4).
In this context, the thymic expression of IL-5, IL-13, and CCR4 (Th2 phenotype, preferentially mediated by IL-2-plus-IL-18) or IFN-γ and CCR5 (Th1 phenotype, mediated by IL-12-plus-IL-18) could be important not only inside the thymus but also in releasing cells to the peripheral immune compartment with a Th1 or Th2 pre-established bias.
The data presented here address the possibility that endogenous or exogenous production of IL-2, IL-12, and IL-18 during host defense and/or normal homeostasis could be responsible for the production of cytokines and induction of phenotypic changes in T cells during thymic maturation. This phenomenon is also important not only because of its potential effects on T cell development but also because it could represent an important mechanism of induction of Th phenotypes from naive precursors early in thymocyte development. Proof of this hypothesis awaits further experimentation.
The authors have no financial conflict of interest.
We thank Drs. John Ortaldo, Pablo Iribarren, Xia Zhang, Deborah Hodge, and Scott Durum, and especially Dr. David M. Reynolds for helpful comments and review of the manuscript. We also thank Della Reynolds for expert technical assistance, Mike Sanford for performing ELISA and RPA analysis, Mehrnoosh Abshari for flow cytometry sorting, and John Wine for his support in animal care and experimentation. The IL-2 neutralizing Abs were kindly provided by Dr. Javier Cote-Sierra and Dr. William Paul (National Institute of Allergy and Infectious Diseases, National Institutes of Health), and the IL-12 neutralizing Ab was a gift from Dr. Alan Sher (National Institute of Allergy and Infectious Diseases, National Institutes of Health). Drs. Terry Fry and Crystal Mackall (National Cancer Institute, National Institutes of Health) kindly provided the BCL-2 transgenic mice. IL-2, IL-12, and IL-18 expression plasmids were a gift from Dr. Mori Watanabe (National Cancer Institute).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. N01-CO-12400. The TE-71 and ANV 41.2 cell lines were created under National Institute of Allergy and Infectious Diseases Grants AI 59575 and AI 24137 (to A.F.).
The publisher or recipient acknowledges right of the U.S. Government to retain a nonexclusive, royalty-free license in and to copyright covering the article. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.
Abbreviations used in this paper: TEC, thymic epithelial cell; ATOC, adult thymic organ culture; m, mouse; LN, lymph node; RPA, RNase protection assay; DN, CD4−CD8− double negative; DP, CD4+CD8+ double positive; CD4+, CD4+CD8− single positive; CD8+, CD4−CD8+ single positive; 7AAD, 7-aminoactinomycin D; KO, knockout.