Abstract
Distinct pattern of homing receptors determines the tissue preference for T cells to exert their effector functions. This homing competence is mostly determined early during T cell activation of naive T cells. In contrast, mechanisms governing the acquisition of particular homing receptors by T cells of the memory phenotype remain enigmatic. Th2 cell-mediated allergic diseases tend to flare during infections despite that these infections prime APCs to produce the prototypic Th1 cell-differentiating cytokine IL-12. In this study, we investigate the effect of IL-12 on the regulation of cutaneous lymphocyte Ag (CLA) on differentiated Th2 cells and consequences of this expression for allergic inflammation. Upon activation with IL-12, CLA− Th2 cells rapidly up-regulated IL-12Rβ2 chain, α(1-3)-fucosyltransferase VII, and CLA molecules. IL-12-mediated CLA expression on Th2 cells was functional because it mediated rolling of these Th2 cells on E-selectin in vitro and migration into human skin grafts in SCID mice. CLA induction occurred immediately after exposure to IL-12 and was independent of IFN-γ expression. In accordance, the transcription factor mediating IFN-γ expression, T-bet, does not directly affect CLA expression. However, CLA expression was further enhanced after IL-12 treatment of T-bet+-transfected Th2 cells in agreement with an increased IL-12 responsiveness of these cells caused by T-bet. The finding that IL-12 conferred skin-homing potential to already differentiated Th2 cells before inducing a switch in their cytokine production profile may explain the observed exacerbation of allergic skin diseases following bacterial infections.
It is currently believed that the acquisition of tissue-specific homing receptors by T cells occurs in the lymph nodes shortly after the initial activation of naive T cells and that it is regulated by tissue-draining dendritic cells (DC)3 (1, 2, 3, 4). Recent mouse studies have shown that DC can also induce the expression of tissue-specific homing receptors in effector/memory T cells (2, 4), but the factors controlling this particular process are not known. In contrast, it is known that infections may lead to relapses or exacerbations of Th2 cell-mediated allergic diseases such as atopic dermatitis (AD) (5, 6, 7). In humans, T cell migration to the skin requires expression of the E-selectin ligand cutaneous lymphocyte Ag (CLA), and expression of E-selectin ligands is often used to prove effective cutaneous imprinting (1, 2, 4). CLA+ memory Th2 cells can be found and isolated from inflamed skin of AD patients. In addition, CLA+ memory Th2 cells, which are capable of homing to human skin in vivo, are abundant in the peripheral blood of these individuals and their frequency correlates with disease severity (8, 9). Moreover, targeting E-selectin binding abrogates cutaneous migration of skin-specific memory T cells (10, 11, 12). It is not known whether human CLA−, fully differentiated effector/memory T cells are still sensitive to this cutaneous imprinting and consequently obtain the potential to migrate to the skin in vivo.
Allergic diseases such as asthma or AD can be exacerbated by microbial infections, and microbial products activate the innate immune system through interaction with receptors that recognize pathogen-associated molecular patterns (5, 6, 7, 13, 14). In response to pathogen-associated molecular pattern activation, innate immune cells produce, among others, proinflammatory cytokines such as IL-1, IL-6, IL-12, and type I and II IFN, which largely influence the cytokine phenotype secreted by the adaptive immune Th cells. Particularly IL-12 potently drives naive T cells into Th1 cell differentiation (15, 16, 17), but it is believed to have little effect on differentiated Th2 cells due to their low or absent IL-12R expression (18, 19, 20). In this study, we investigated whether IL-12 is capable of instructing skin homing in established human CLA− Th2 cells and whether the induction of this tissue-homing phenotype occurs in the context of a switch from Th2 to Th1 cytokine production. We found that culture of former CLA− Th2 cells with IL-12 was sufficient to induce CLA expression and to confer these cells with skin-homing potential, as demonstrated by migration studies in SCID mice transplanted with human skin. CLA up-regulation was achieved before induction of IFN-γ and was not directly inducible by the transcription factor T-bet. However, IL-12-mediated expression of CLA was much higher in T-bet-transfected Th2 cells. Our findings that IL-12 conferred skin homing to Th2 cells before a switch in their cytokine profile may explain the observed exacerbations of the Th2-associated inflammatory skin disease AD following bacterial infections.
Materials and Methods
Generation, characterization, and retroviral infection of Th2 cell lines and clones
Skin-derived Th2 cells were obtained from AD patients allergic to house dust mite, as described (10). Culture medium (Yssel’s medium, 8% Ultra Low Ig FCS (Invitrogen Life Technologies), supplemented with 100 U/ml human rIL-2) was exchanged every 2–3 days. Th cell lines and clones from PBMCs of healthy donors were generated by activation with PHA and irradiated allogenic PBMCs. Th cell clones were isolated from skin-derived and PBMC-derived Th cell lines by limiting dilution and maintained by repeated stimulation with irradiated allogenic human PBMCs and PHA. Th2 cell clones were characterized based on cytokine production, as detected by standard ELISA and intracytoplasmic FACS analysis, as described (10, 2122, 23
Neuraminidase treatment and analysis of Th2 cells
Newly synthesized CLA was analyzed by treating CLA+ cells with neuraminidase and by analyzing CLA and CD15 expression in the FACS before and 24 h after treatment (24, 25). Briefly, 5 × 106 Th2 cells are suspended in 500 μl of 0.1 U/ml neuraminidase solutions at 37°C for 40 min (Sialidase from Clostridium perfringens; Roche Boehringer Mannheim Diagnostics).
RT-PCR and quantitative real-time RT-PCR analysis
Aliquots of 5 × 10610). FucTVII primers and probes were titrated prior to use in conjunction with the HPRT primers and probe and showed RNA concentration-dependent signals, which were expressed as percentage of HPRT using the ΔCT method (sequence detector 3.1; Applied Biosystems) based on threshold cycle differences.
Analysis of intracellular calcium mobilization
Th2 cells (5 × 106/ml) were loaded with Fluo-4/acetoxymethyl (AM) ester (60 min at 37°C using 0.02% Pluronic, 2 μM Fluo-4 in HBSS; Molecular Probes). Th2 cells were washed with indicator-free medium thereafter and incubated for at least 15 min at room temperature for de-esterification of intracellular AM esters. Intracellular Ca2+ mobilization was monitored by an increase of fluorescence intensity on FL1 in a FACScan (3000 events every 5 s) before and after adding PBS, CCL22, or control chemokines (300 ng/ml).
Transwell migration assay
IL-4- and IL-12-treated T cells were labeled with calcein/AM (1 μl/ml) for 30 min at 37°C. T cells were washed and plated (5 × 105
Transplantation of human adult skin on SCID mice and recruitment of human Th2 cells to human skin grafts of SCID-hu Skin mice
Transplantation of human adult skin onto SCID mice (SCID-hu Skin mice) and recruitment experiments were conducted, as described (10). Briefly, SCID mice were transplanted with two pieces of human skin at their back replacing ∼0.8-mm2 mouse skin each. SCID-hu Skin mice were used for experiments after 6–8 wk when human skin pieces resembled normal human skin. At experimental day 0, SCID-hu Skin mice were reconstituted (i.p.) with 1.5 × 108
Stable tethering of human Th2 cells on E-selectin
Laminar flow assays were performed, as described previously (26). Briefly, polystyrene dishes were coated with up to 20 μg/ml E-selectin-Cκ (3 h, room temperature, in TBS (pH 8.5)), washed with PBS, blocked with 3% milk powder in PBS for 1 h, and washed. The dishes were incorporated as the lower wall of a parallel plate flow chamber. For analysis of tethering, cells were diluted 1/5 in HBSS with cations and perfused through the flow chamber at 2 × 106/ml over a range of shear stresses (3–0.2 dyn/cm2). For comparison of tethering onto different substrate densities, cells were perfused for 3 min at 0.5 dyn/cm2 over up to 20 μg/ml E-selectin-Cκ, and the fraction of tethered cells was determined between 2 and 3 min. At 0.5 dyn/cm2 shear stress, no tethering or rolling cells could be determined at and below 1 μg/ml E-selectin-Cκ density. IL-4- and IL-12-treated resting Th2 cells or IL-12-treated Th2 cells with or without anti-IFN-γ mAb treatment were washed and stored at 10 × 106/ml in HBSS Ca and Mg free, 10 mM HEPES on ice. Th2 cells (2 × 106/ml) were perfused into the chamber, and the fraction of cells that came into close proximity with the substrate and tethered stably (rolling for >1 s after initial attachment) was determined.
Statistical analysis
Statistical analysis was performed using Student’s t test, and differences were considered significant at p < 0.05.
Results
IL-12 up-regulates FucTVII in Th2 cells
Th2 cell clones were generated from AD skin (ADTh2) or from PBMC (pTh2) of 10 different donors. Th2 cell clones had a characteristic Th2 cytokine phenotype (their levels of IL-4 and IFN-γ production following stimulation were >2.0 ng/ml and <0.3 ng/ml, respectively) and were CCR4+, independently of their activation state. All ADTh2 cell clones expressed high levels of CLA (range 300–800 mean fluorescence intensity), whereas pTh2 cell clones expressed low levels of CLA or were CLA− (Fig. 1⇓A). As previously described (18, 19, 20), IL-12Rβ2 was absent in resting Th2 cells independently of their origin, but following activation all pTh2 and ADTh2 cell clones up-regulated this receptor chain (Fig. 1⇓B, data not shown). CLA+ ADTh2 cells and CLA− pTh2 cells were subsequently analyzed for the expression of α(1-3)-FucTVII. FucTVII is an essential part of the machinery for CLA synthesis, which is responsible for the synthesis of the sialyl LewisX saccharides on CD15 backbones, resulting in the expression of the E-selectin-binding receptor CLA. Correlating with the strong expression of CLA, FucTVII was expressed in ADTh2 cells (Fig. 1⇓B). In contrast, FucTVII mRNA transcripts were absent in resting and activated CLA− pTh2 cells (Fig. 1⇓B). In contrast, FucTVII could be induced in these CLA− pTh2 cells by IL-12 treatment (Fig. 1⇓B). Because these Th2 cells produced large amounts of IL-4, and IL-4 has strong antagonist effects on IL-12-mediated Th1 cell differentiation, we investigated whether the expression of FucTVII induced by IL-12 was also regulated by IL-4. Exogenous IL-4 did not induce FucTVII in pTh2 and did not significantly reduce the expression levels of this enzyme in CLA+ ADTh2 or pTh1 (Fig. 1⇓C). Furthermore, IL-12-mediated expression of FucTVII by Th2 cells was not affected by the addition of a neutralizing anti-IL-4 Ab (Fig. 1⇓D).
Up-regulation of FucTVII expression in Th2 cells by IL-12. A, CLA expression was analyzed in Th2 cell clones derived from PBMC (pTh2) or AD skin (ADTh2). All ADTh2 cell clones expressed high levels of CLA, whereas CLA− Th2 cell clones were selected from PBMC. One representative clone of each is shown. B, Resting CLA− FucTVII− pTh2 cell clones and resting CLA+ FucTVII+ ADTh2 cell clones were stimulated with anti-CD3 plus anti-CD28 mAbs, and IL-12, IL-4, or no cytokine (−) was added, as indicated. RNA was prepared from aliquots harvested at the indicated time points, and RT-PCR using specific primers for FucTVII, CCR4, and IL-12Rβ2 was performed. C, After stimulation and expansion with IL-2 alone (−) or in addition with IL-4 or IL-12, FucTVII expression of CLA− pTh2 cells, ADTh2 cells, and PBMC-derived control Th1 cells was analyzed by quantitative real-time RT-PCR. FucTVII expression is shown as percentage of HPRT. Representative clones of >10 analyzed are shown. D, ADTh2 and pTh2 cells were stimulated ± IL-4 or IL-12 and with or without blocking anti-IL-4 Abs. Resting Th2 cells were harvested, and mRNA was extracted and analyzed for the expression of FucTVII, as in C.
IL-12 induced sustained CLA expression and mediated E-selectin binding on Th2 cells
Once we demonstrated that IL-12 alone is sufficient to promote FucTVII expression on differentiated Th2 cells, we searched whether this induction will also result in the appearance of CLA on the surfaces of these pTh2 cells. As illustrated in Fig. 2⇓A, CLA started to be expressed in pTh2 cell clones cultured with IL-12 already 4 days after activation. Additional expansion of pTh2 cells in the presence of IL-12 resulted in further increase in the percentage of CLA+ cells (at least 20–25% of the former CLA− pTh2; Fig. 2⇓, A and B). Expression levels of FucTVII mRNA and CLA protein in these IL-12-treated pTh2 cells remained stable for up to 9 wk under expansion conditions without IL-12 supplementation (data not shown). To test whether the machinery to synthesize CLA was stably up-regulated, these IL-12-expanded Th2 cells were treated with neuraminidase, an enzyme that cleaves CLA from the cell surface of the Th cells, revealing the P-selectin glycoprotein ligand 1 backbone of CD15 (24). Exposure to neuraminidase resulted in the loss of CLA expression and the emergence of CD15 (Fig. 2⇓C, 2 h). Subsequent culture in medium for 24 h resulted in the reappearance of CLA (Fig. 2⇓C, 24 h). In contrast to untreated cells, CLA could be detected simultaneously with CD15, demonstrating re-expression of newly synthesized CLA on previously neuraminidase-treated Th2 cells (Fig. 2⇓C).
IL-12 induced sustained CLA expression on Th2 cells. A, CLA expression on pTh2 cells was analyzed 4 and 21 days after stimulation in the presence of IL-4 or IL-12, as indicated. Numbers in the upper right quadrant indicate percentages of CD4+ T cells expressing CLA. B, CLA expression on pTh2 and ADTh2 cells was compared 3 wk after stimulation and expansion in the presence of IL-4 or IL-12, as indicated. Numbers in the upper right quadrant indicate percentages of CD3+ T cells expressing CLA. Representative experiments of >10 are shown. C, IL-12-induced up-regulation of the machinery for CLA synthesis was analyzed by neuraminidase treatment and monitoring CLA de novo expression by determination of CD15 and CLA expression before (untreated) and 2 and 24 h after cleavage of CLA. One of five similar experiments is shown. D, CLA− pTh2 cells were stimulated and expanded for 8 wk either with IL-4 or IL-12 and analyzed in a dynamic flow chamber assay for stable tethering on E-selectin. Th2 cells were perfused into the chamber, and the fraction of cells that tethered stably (rolling for >1 s after initial attachment) was determined. Shear stress (left) and dose (right)-dependent stable tethering are demonstrated with an E-selectin-coating concentration at 20 μg/ml (left) and applied shear stress at 0.5 dyn/cm2 (right).
The sustained up-regulation of CLA on IL-12-expanded pTh2 cells had functional consequences on the adhesion properties of these cells. IL-12-treated pTh2 cells showed significantly more stable tethers on E-selectin than IL-4-treated pTh2 cells (Fig. 2⇑D). After tethering, slow rolling, typical for and only mediated by selectins, was also observed. At moderate shear stress (0.2 dyn/cm2), the frequency of tethered IL-12-treated pTh2 cells was ∼4-fold higher than that of IL-4-treated or untreated pTh2 cells (Fig. 2⇑D, left). Tethering of IL-12-treated pTh2 cells occurred only at shear stresses below 1.2 dyn/cm2 (Fig. 2⇑D, left), consistent with the reported shear stress tethering pattern described for E-selectin, which is lower than the equivalent pattern described for L-selectin (27). Tethering of IL-12-treated pTh2 cells correlated with the density of E-selectin coated to the dishes. Lowering the density of coated E-selectin reduced the tethering frequency and increased the velocity of rolling. Coating densities below 0.3 μg/ml did not mediate significant tethering or rolling (Fig. 2⇑D, right).
IL-12 instructed skin homing of human Th2 cells
For effective in vivo skin homing of human Th cells, CLA expression is mandatory, but not the exclusive requirement. CCR4 is believed to be a dominant chemokine receptor that mediates T cell adhesion and extravasation into the skin (2, 4, 10, 28), and all Th2 cells analyzed in this study were CCR4+. Importantly, CCR4 expression (Fig. 3⇓A) and CCR4 responsiveness of pTh2 cell clones were not altered by the different culture conditions, as indicated by intracellular Ca2+ mobilization assays (Fig. 3⇓B, left) and Transwell migration assays (Fig. 3⇓B, right), thus allowing us to study skin homing via CCR4.
IL-12 instructed skin homing of human Th2 cells. A, pTh2 cells were expanded in the presence of either IL-4 or IL-12, and analyzed for the expression of the skin-homing receptor CCR4 by FACS. B, The responsiveness to the CCR4 ligand CCL22 was determined by measuring intracellular Ca2+ mobilization (left) and in vitro Transwell migration (right). After culturing with IL-4 or IL-12, pTh2 cells were treated with CCL22, and intracellular Ca2+ mobilization was determined by change of fluorescence intensity (B, left) after 15 s (bold line, arrow) and 120 s (dotted line). These pTh2 cells were further analyzed for in vitro Transwell migration. Migrated cells in response to CCL22 are shown as percentage of input cells (B, right). C and D, These pTh2 cells cultured in the presence of either IL-4 or IL-12 were adoptively transferred into SCID mice previously transplanted with human skin. PBS or CCL22 was injected into the skin pieces (intracutaneously) according to the protocol previously described (10 ). In addition, the different groups of mice received two injections of isotype control mAbs (control mAb) and one group blocking anti-E-selectin mAbs (anti-CD62E mAb), as indicated. After four injections of PBS or CCL22, skin pieces were removed, cellular suspensions were prepared, and CCL22- or PBS-mediated recruitment of Th2 cells was analyzed by FACS. C, Cellular suspensions were stained with anti-CD3 FITC and anti-CD4 PE Abs and analyzed by FACS gating on the mononuclear living cell fraction. Representative FACS dot blot results of cellular suspensions of single skin pieces are shown, one of each group of mice. D, Skin homing of pTh2 cells previously expanded in the presence of IL-4 (□) or IL-12 (▪) is expressed as the mean percentage of human CD3+ and CD4+ cells of single cellular skin suspensions. As indicated, mice received either control mAbs or blocking anti-CD62E mAbs. One representative experiment of four is shown.
Because exposure of pTh2 cells to IL-12 resulted in functional E-selecting binding in vitro, we wanted to test whether IL-12 conferred these pTh2 with the capability to migrate to human skin in vivo. This potential was assessed using the SCID-hu Skin model, as previously described (10). In the SCID-hu Skin mouse, the recruitment of adoptively transferred human Th2 cells into human skin grafts previously transplanted onto SCID mice is analyzed. Intracutaneous injection of the human skin grafts with the CCR4 ligand CCL22, but not with control PBS, recruited significant numbers of IL-12-treated pTh2 cells to the human skin of SCID-hu Skin mice (Fig. 3⇑, C and D). This response was E-selectin dependent, because it could be completely abrogated by systemic treatment with blocking anti-E-selectin mAbs. In contrast, CCR4+, CLA−, pTh2 cells, obtained from cell cultures containing IL-4, were unable to migrate to the human skin in vivo, despite that they displayed similar chemotactic responses toward CCL22 in vitro (Fig. 3⇑).
IL-12 induced up-regulation of CLA on Th2 cells independently of a Th1 phenotype
IL-12-mediated induction of IFN-γ production required sustained exposure to IL-12 for 6–8 wk and induced up to 25% IFN-γ producers among ADTh2 and pTh2 cell clones (Fig. 4⇓ and data not shown), confirming data by others (29, 30). Analyzing CLA expression in these pTh2 cells in more detail revealed that CLA expression was independent of IL-4 or IFN-γ production. Activated pTh2 cells were stained for surface CLA and intracytoplasmic IL-4 and IFN-γ expression. Gating on IL-4+ or IFN-γ+ cells demonstrated unequivocal CLA expression levels in both subsets (Fig. 4⇓A). Likewise, gating on CLA− and CLA+ pTh2 cells after expansion in IL-12 substantiated that IL-4 and IFN-γ producers were evenly distributed among the different CLA-expressing subsets (Fig. 4⇓B). Furthermore, experiments neutralizing IFN-γ by blocking Abs confirmed that the presence of IFN-γ had no influence on CLA expression during culture with IL-12. Following 8 wk of exposure to IL-12, CLA expression was identical independent of the presence of IFN-γ (Fig. 4⇓C, left). Moreover, CLA function was also identical because these cells unequivocally tethered on E-selectin (Fig. 4⇓C, right). These data demonstrate that IL-12-mediated regulation of IFN-γ production in Th2 cells is dissociated from IL-12-mediated CLA induction and that there is no direct effect of IFN-γ on CLA expression.
CLA expression and stable tethering on E-selectin induced by IL-12 were independent of IFN-γ secretion. Former CLA− pTh2 cells were analyzed after 8 wk of stimulation and expansion with IL-4 or IL-12. A, These pTh2 cells, previously cultured in the presence of either IL-4 (left) or IL-12 (right), were activated, and CLA and cytokine expression was determined by FACS analysis. CLA expression in IL-4 producers (upper panels) and IFN-γ producers (lower panels) is shown after gating on CD4+, IL-4+, or IFN-γ+ cells. CLAhigh cells could be detected exclusively in pTh2 cells that had been expanded in IL-12 (right), and CLA was uniformly expressed by the IL-4+ and IFN-γ+ subsets. B, IL-4 secretion (left) and IFN-γ secretion (right) were analyzed after activation of pTh2 cells previously expanded with IL-12. Cytokine-producing CLA− (upper panels) and CLA+ cells (lower panels) are shown as percentage of total. One of three independent and similar experiments is shown (A and B). C, pTh2 cells were expanded with IL-12 in the presence or absence of anti-IFN-γ mAbs, as in A, and CLA expression (mean of three samples) was determined by FACS analysis (left). Functional relevance of CLA expression was demonstrated in a dynamic flow chamber assay for stable tethering on E-selectin. Unequivocal stable tethering on E-selectin was determined for these pTh2 cells independently of anti-IFN-γ mAbs. One of three similar experiments is shown.
The underlying mechanism of IFN-γ production in Th cells is the induction of the transcription factor T-bet. Thus, to understand better the relation of the different IL-12-mediated pathways, we investigated the role of the Th1 phenotype-driving factor T-bet on CLA expression. Polarized human Th2 cells were retrovirally transfected with a bicistronic vector encoding GFP and T-bet. This procedure resulted in a productive infection of up to 10% of the cells, as assessed by GFP expression. Upon activation, up to 75% of GFP+, T-bet-transfected Th2 cells produced IFN-γ (data not shown). T-bet-transfected cells were kept in culture together with uninfected cells to avoid any potential skewing from the autocrine release of Th1 cytokines (22). Expression of CLA was not induced by transfection with T-bet in the absence of IL-12 (Fig. 5⇓A). In contrast, the presence of IL-12 induced higher expression levels of CLA in the T-bet-transfected subset (Fig. 5⇓A). This is in agreement with an increased responsiveness to IL-12 by T-bet-expressing cells due to up-regulation of IL-12Rβ2 (22, 31). To assess the regulation of the machinery for CLA synthesis, CLA re-expression after cleavage by neuraminidase was analyzed (Fig. 5⇓B). Again, T-bet itself did not modulate CLA re-expression in the absence of IL-12, while IL-12 had induced the intracellular machinery for the synthesis of CLA to much higher levels in T-bet-transfected Th2 cells (∗, p < 0.007; Fig. 5⇓B). Together, these data indicate that T-bet participates in the up-regulation of CLA+ glycoproteins by IL-12 through sustained IL-12Rβ2 expression. Accordingly, when Th2 cell cultures with IL-12 were prolonged for 15–20 wk, expression levels of IL-12Rβ2, FucTVII, and CLA were increasingly detected to be higher in IFN-γ producers than in IL-4 producers (data not shown).
T-bet did not directly influence CLA expression, but enhanced IL-12-mediated up-regulation of CLA. CLAlow pTh2 cells were transfected with T-bet-GFP, activated, and expanded, as described, either in the presence (IL-12) or absence of IL-12 (anti-IL-12 mAbs). A, Resting GFP-T-bet− (□) and GFP-T-bet+ pTh2 cells (▪) from the same cell populations were analyzed for CLA surface expression, revealing IL-12-mediated up-regulation of CLA. B, Th cell populations from A were treated with neuraminidase and analyzed for CLA surface expression 2 and 24 h after treatment. Resting T-bet− and T-bet+ pTh2 cells that had been expanded in the absence of IL-12 redisplayed identical CLA expression levels 24 h after neuraminidase treatment. In contrast, IL-12-induced CLA resynthesis was significantly increased in T-bet+ pTh2 cells compared with T-bet− pTh2 cells (p < 0.007).
Discussion
The data of this study show that IL-12 instructs skin homing of human Th2 cells. Interestingly, IL-12, a single cytokine that is secreted during infections, is capable of converting human CLA− Th2 effector/memory cells to CLA+ Th2 cells that home to the skin in vivo. It is believed that programming of T cells to exhibit a tissue-specific homing receptor repertoire is achieved during initial activation of naive T cells in the local lymphoid organs, such as the draining lymph nodes. This process of regional imprinting was demonstrated recently for gut and skin-homing T cells and can be mediated by tissue-draining DC (1, 2, 3, 4, 32). The exact factors regulating imprinting for the skin are still not identified. Our data indicate that IL-12 is one important factor mediating the expression of skin-specific homing receptors and that this mechanism also operates in already differentiated Th2 cells. Indeed, systemic or local infections act as flare factors for Th2-mediated skin inflammation, and our data demonstrate that one underlying mechanism may be IL-12-mediated cutaneous imprinting of human Th2 cells (5, 33, 34). Especially exacerbations of AD, which is induced by skin-homing Th2 cells, can be triggered by local infections (Staphylococcus aureus colonization) and also by infections at sites distant from skin, such as viral upper respiratory tract infections (6, 7, 34, 35). Subsequent development of the three major atopic diseases, allergic asthma, allergic rhinitis, and AD, in one single patient is referred to as atopic march. Based on our results, it is tempting to speculate that allergen-specific Th2 cells following contact with Ag in the airways may change tissue selectivity upon infection-driven exposure to IL-12. Consequently, the coexpression of CLA could determine whether allergen-specific CCR4+ Th2 cells can also home to the skin and induce AD (28, 36, 37, 38, 39, 40).
The IL-12R is composed of the constitutively expressed IL-12Rβ1 chain and the inducible IL-12Rβ2 chain and resting memory Th2 cells have little or no expression of the β2 chain, which suggests that these memory Th2 cells may be refractory to IL-12-mediated CLA induction (18, 19, 20). In this study, we showed that despite the absence of the IL-12Rβ2 chain on resting Th2 cells (data not shown), these Th2 cells immediately up-regulate IL-12Rβ2 chain upon activation, and consequently respond to IL-12 as IL-12R ligation induced the expression of the E-selectin ligand CLA and α(1-3)-FucTVII. FucTVII mediates the final and most selective step of posttranslational glycosylations of E-selectin ligands, and it was shown that FucTVII−/− Th1, and Th2 cells are unable to home to the skin (9, 41, 42, 43, 44, 45). Even though there is evidence that alternative factors may regulate cutaneous imprinting (4, 46), our data demonstrate that IL-12 is sufficient to potently regulate skin homing in T cells. Thus, our studies underline that IL-12 induces FucTVII and CLA not only in naive T cells, but also in fully differentiated Th2 cells, and demonstrate for the first time that this CLA expression is functionally relevant and mediates homing into human skin (47, 48).
The role of the cytokines IL-4 and IL-12 in CLA regulation seems to contrast their role during Th cell polarization. For Th cell polarization, it is the absence or presence of IL-4 that dominates the ultimate Th cytokine phenotype (49, 50). In contrast, addition or blocking of IL-4 during culture of fully differentiated Th2 cells had only minor effects on the expression of CLA compared with IL-12. Thus, in the context of CLA regulation, IL-12 dominates IL-4. Moreover, a reversal of IL-12-mediated CLA induction was not achieved with IL-4 and anti-IL-12 mAbs, nor did expansions without specific cytokine culture conditions during subsequent cycles of activation significantly alter the expression level of FucTVII mRNA or CLA protein (data not shown). Our findings on human CLA regulation contrast findings with naive Th cells, in which IL-4 was described to strongly inhibit or even reverse the expression of E-selectin ligands. However, this effect was demonstrated in studies using naive Th cells that were polarized in vitro with the respective cytokines (47, 48, 51). Studies performed with ex vivo CD45RO+ T cells like ours also showed much weaker or no CLA suppression by IL-4 (48).
Among the different biological functions of IL-12, its role in lymphocyte development and the initiation of IFN-γ production is studied best. Although IL-12 alone cannot activate resting T cells, stimulation of T cells in the presence of IL-12 induces IFN-γ production (17, 52). We therefore examined the effects of IL-12 on the Th cell phenotype during the process of CLA regulation. Interestingly, even though Th2 cells have been cultured in polarizing conditions and a strong up-regulation of CLA by IL-12 could be achieved, these Th2 cells remained IL-4 producers and only some Th2 cells became IFN-γ producers. Moreover, IL-12-mediated CLA expression was not a function of IFN-γ production and also independent of secreted IFN-γ, confirming data by others that E-selectin ligand expression and cytokine synthesis are independently regulated by IL-12 (51, 53).
Only upon sustained exposure to IL-12, CLAhigh cells were more prominent among IFN-γ producers. Transfecting Th2 memory cells with the Th1-specific transcription factor T-bet did not up-regulate CLA in the absence of IL-12, although these T-bet-transfected Th2 cells produced large amounts of IFN-γ. Only in the presence of IL-12, T-bet-transfected cells expressed higher amounts of CLA and more rapidly re-expressed surface CLA after neuraminidase cleavage. This is in accordance with an increased responsiveness of T-bet-expressing cells to IL-12 via up-regulation of the IL-12Rβ2 chain (22, 31, 54). Thus, increased responsiveness to IL-12 explains why up-regulation of CLA in T-bet+ Th cells only occurs in the presence of IL-12. These findings provide a basis why a correlation of FucTVII, CLA, and E-selectin binding with the Th1 phenotype could be demonstrated, despite the fact that IL-12 induces CLA independent of the Th phenotype (43, 47, 48, 51, 53, 55).
In summary, we show that ex vivo human Th2 cells were responsive to IL-12, up-regulated CLA expression, and hence adhesion to E-selectin, resulting in skin homing in vivo. CLA induction by IL-12 occurred independently of IFN-γ production. However, the expression of T-bet significantly enhanced IL-12 responsiveness, and thus, IL-12-mediated up-regulation of CLA. Together, our data provide a basis to explain why CLA and FucTVII expression were often found in correlation with the Th1 phenotype, although CLA expression is independent of the Th cytokine phenotype. In addition, our findings that IL-12 conferred skin-homing potential to Th2 cells before changing their cytokine production profile may help to understand exacerbations of AD following bacterial infections.
Acknowledgments
We thank Dr. Laurie Glimcher for providing the T-bet plasmid, and Sandra Fassl for her technical support.
Disclosures
The authors have no financial conflict of interest.
Footnotes
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.
↵1 This work was partially supported by grants from the Deutsche Forschungsgemeinschaft (DFG Bi 696/3-1; SFB 685 A6), Landesstiftung Baden-Württemberg (P-LS-AL/17), and the medical faculty, University of Tübingen (F126810) (to T.B.).
↵2 Address correspondence and reprint requests to Dr. Tilo Biedermann, Department of Dermatology, Eberhard-Karls-University, Liebermeisterstrasse 25, 72076 Tuebingen, Germany; E-mail address: tilo.biedermann{at}med.uni-tuebingen.de or Dr. José M. Carballido, Novartis Institutes for Biomedical Research, Brunnerstrasse 59, 1235 Vienna, Austria; E-mail address: jose.carballido{at}novartis.com
↵3 Abbreviations used in this paper: DC, dendritic cell; AD, atopic dermatitis; AM, acetoxymethyl; CLA, cutaneous lymphocyte Ag; FucTVII, fucosyltransferase VII; HPRT, hypoxanthine phosphoribosyltransferase; pTh2, Th2 cells from PBMC.
- Received August 10, 2005.
- Accepted June 26, 2006.
- Copyright © 2006 by The American Association of Immunologists
References
In this issue
