Macrophages and neutrophils are the targets for immune suppression by glucocorticoids in contact allergy

Glucocorticoids (GCs) are widely used in the treatment of allergic skin conditions despite having numerous side effects. Here we use Cre/loxP-engineered tissue- and cell-specific and function-selective GC receptor (GR) mutant mice to identify responsive cell types and molecular mechanisms underlying the antiinflammatory activity of GCs in contact hypersensitivity (CHS). CHS was repressed by GCs only at the challenge phase, i.e., during reexposure to the hapten. Inactivation of the GR gene in keratinocytes or T cells of mutant mice did not attenuate the effects of GCs, but its ablation in macrophages and neutrophils abolished downregulation of the inflammatory response. Moreover, mice expressing a DNA binding–defective GR were also resistant to GC treatment. The persistent infiltration of macrophages and neutrophils in these mice is explained by an impaired repression of inflammatory cytokines and chemokines such as IL-1β, monocyte chemoattractant protein-1, macrophage inflammatory protein-2, and IFN-γ–inducible protein 10. In contrast TNF-α repression remained intact. Consequently, injection of recombinant proteins of these cytokines and chemokines partially reversed suppression of CHS by GCs. These studies provide evidence that in contact allergy, therapeutic action of corticosteroids is in macrophages and neutrophils and that dimerization GR is required.

Contact dermatitis is a common condition caused by exposure to low molecular weight compounds such as metals or poison ivy. It is usually treated by topical or systemic application of glucocorticoids (GCs) (1). Such treatment effectively ameliorates the ongoing inflammatory response, but the long-term use of corticosteroids can cause adverse effects including skin atrophy or osteoporosis (2). While the general antiinflammatory capacity of GCs is well documented, their target cells and the transcriptional mechanisms underlying the treatment of allergic conditions by GCs remain largely elusive.

Contact hypersensitivity (CHS), a T cell–dependent immune response, is a classic mouse model of human allergic contact dermatitis. It involves 2 phases known as the sensitization phase and the challenge phase (3). In the sensitization phase, the hapten is cross-linked to epidermal proteins and taken up by resident epidermal Langerhans cells (LCs), dermal DCs, and tissue macrophages. Subsequently, these cells — triggered by cytokines such as TNF-α and IL-1β — migrate to the draining lymph nodes and mature into potent antigen presenting cells (APCs) (4). In turn, naive T cells are activated and differentiate into antigen-specific Th1 and cytotoxic T cells under the influence of polarizing signals such as IL-12 (5). These T cells are then designated as “sensitized.” In parallel, B-1 cells in the peritoneal cavity are stimulated by IL-4–secreting NKT cells to produce antigen-specific IgM. In addition, NK cells appear to play a role in sensitization (6–8). In the challenge phase, initiated by the reexposure to the same hapten, cross-linking of IgM molecules causes the release of the complement factor C5a (9). This acts on platelets, skin resident mast cells, and macrophages (10) and thereby results in the activation of the microvasculature (11), allowing for the penetration of a first wave of leukocytes within 2 hours (12). In response to chemokines such as IFN-γ–inducible protein 10 (IP-10, also known as CXCL10), sensitized T cells enter the dermis at the site of hapten exposure and become restimulated by resident APCs (13). These T cells then release proinflammatory mediators, which trigger resident myeloid cells to secrete chemokines such as macrophage inflammatory protein-2 (MIP-2, also known as CXCL2) and monocyte chemoattractant protein-1 (MCP-1, also known as CCL2), leading to a massive second wave of leukocyte infiltration (4).

GCs are known to suppress a plethora of proinflammatory genes encoding cytokines, chemokines, cell adhesion molecules, and other mediators, thereby interfering with the inflammatory response (14). These effects are mediated by the GC receptor (GR), a transcription factor belonging to the nuclear receptor superfamily. The GR controls transcription by different mechanisms (15). One mechanism requires binding of receptor homodimers to GC-responsive elements (GREs) in gene regulatory regions. In a second mechanism the GR also has the ability to modulate the activity of other transcription factors such as AP-1, NF-κB, and STAT5 independent of dimerization and subsequent DNA binding of the receptor (reviewed in refs. 15, 16). We previously demonstrated that AP-1– and NF-κB–dependent proinflammatory genes were effectively repressed in mice carrying a point mutation that impairs dimerization-induced DNA binding of the GR (GR^dim) (17, 18). In addition, phorbol ester–induced irritative inflammation was effectively suppressed by GCs in these animals (17). Here we asked whether intervention with GCs of a T cell–dependent delayed-type hypersensitivity reaction requires DNA binding of the GR.

To identify the phase, cell types, and mechanisms of GR action that are involved in the repression of an allergic skin condition by GCs, we induced CHS in various GR-mutant mouse strains. Our results demonstrate that only the challenge phase is responsive to GCs, that myeloid cells are the primary targets of these actions, and that these effects require DNA binding by the GR. The analysis of macrophages in cell culture further indicates that downregulation of cytokines such as IL-1β and chemokines such as MCP-1, MIP-2, and IP-10 is impaired. Application of these proteins in part counteracts the antiinflammatory effects of GCs in CHS. Thus, our results shed new light on the mechanisms of these drugs in the treatment of allergic skin conditions.

GCs inhibit CHS only during the challenge phase. To determine the antiinflammatory effect of GCs during CHS, mice were either sensitized with the contact sensitizer 2,4-dinitrofluorobenzene (DNFB) or treated with vehicle alone (Figure 1A). One group of mice was additionally treated with dexamethasone. At day 5, all animals were challenged by application of DNFB on the dorsal ear, and 24 hours later the inflammatory response was determined. As expected, dexamethasone suppressed the DNFB-induced ear swelling to levels of that of ears in vehicle-sensitzed animals, confirming its therapeutic efficacy (Figure 1A).

Figure 1

Dexamethasone treatment suppresses CHS but is ineffective when applied during the sensitization phase. (A) Beginning at day –1 and continuing for 7 days, mice were treated with either PBS (rows 1 and 2) or dexamethasone (Dex; row 3). On day 0, mice were treated on the shaved back skin either with vehicle (row 1) or with DNFB (rows 2 and 3). All mice were challenged with DNFB at the ear on day 5, the effect of which was determined 24 hours later. *P < 0.01 (n = 6). (B and C) Wild-type mice were injected i.p. with either dexamethasone or vehicle. Twenty-four hours later, they were topically treated on the shaved back skin with the fluorophore FITC. After 18 hours, cells from regional lymph nodes were analyzed by flow cytometry. (B) FACS analysis of lymph node cells was performed to determine the level of expression of CD11c and loading with FITC. (C) Numbers of FITC-positive cells after treatment with either FITC alone or FITC in combination with dexamethasone. *P < 0.01. (D) Mice were treated as described in A except that the dexamethasone treatment was terminated at day 1 after DNFB sensitization (Dex 48 h; row 3) (n = 6).

To determine the phase of CHS that is responsive to GC-mediated inhibition, we first studied the sensitization phase. At this stage, hapten-loaded DCs migrate to the lymph nodes, which is essential for T cell priming. To monitor this process, we applied FITC as a contact sensitizer. Dexamethasone treatment strongly reduced the number of FITC-loaded CD11c⁺ cells found in the draining lymph nodes after 19 hours (Figure 1, B and C). However, the numbers of T and B cells were also reduced (data not shown). These data are comparable with earlier studies that describe a rapid effect of GCs on DC migration (19).

We then tested whether the reduction of hapten-loaded DCs in lymph nodes during the sensitization phase had any effect on the inflammatory response (Figure 1D). To this end, dexamethasone treatment was started 1 day before the first application of DNFB and lasted 48 hours until day 1 as specified in Methods. On day 5 after sensitization, the mice were re-challenged with DNFB, and 24 hours later the swelling was determined. Interestingly, systemic treatment with dexamethasone during the sensitization phase did not prevent the inflammatory response evoked by the second application of DNFB (Figure 1D). Similarly, topical treatment during the sensitization phase with fluocinonide-containing ointment, a commercially available synthetic GC preparation, was unable to inhibit the swelling response (data not shown). Thus, the diminished DC and T cell numbers in the draining lymph nodes after dexamethasone treatment suffice to allow effective T cell priming by the hapten.

Since GCs apparently do not alter the sensitization phase of CHS, we assumed that the challenge phase was responsive to the treatment. Therefore, we applied dexamethasone only at day 5, 1 hour prior to the second DNFB application (Figure 2A). GC application at this time point effectively prevented the inflammatory response. A similar inhibition, in concordance with the clinical practice, was also achieved by topical application of fluocinonide ointment (data not shown).

Figure 2

Dexamethasone inhibits edema formation during the challenge phase of CHS. (A) Mice were treated on day 0 as described in Figure 1A with either vehicle alone (row 1) or DNFB on the shaved back skin (rows 2 and 3). All mice were treated with DNFB on the ear on day 5 to evoke a challenge response. Mice were additionally treated with PBS (rows 1 and 2) or dexamethasone (row 3) starting 1 hour before the reapplication of DNFB at day 5 and dexamethasone was supplied in the drinking water until day 6. *P < 0.01 (n = 5). (B) Animals were treated with vehicle (rows 1 and 4) or oxazolone (Ox) on the back skin on day 0 (rows 2, 3, 5, and 6). On day 5, mice were mock treated with vehicle (rows 1, 2, 4, and 5) or with dexamethasone (rows 3 and 6). One hour later, all mice were challenged on the ear with oxazolone. The ear swelling was determined 2 hours (rows 1–3) or 24 hours (rows 4–6) after challenge. *P < 0.01 (n = 6).

The effector phase of CHS consists of 2 waves of leukocyte infiltration (20). An early wave occurs 2 hours after application of the contact sensitizer and is elicited by IgM and C5a, which activate mast cells. Those in turn act on vascular endothelial cells, leading to the migration of sensitized T cells to the site of hapten application (9, 12). This early invasion is essential for a second wave of leukocyte infiltration, which leads to the development of the full inflammatory response (12). Since the initial swelling is comparably weak, we used the strong contact sensitizer oxazolone to dissect the effects of dexamethasone on the 2 phases. GC treatment of oxazolone-sensitized mice 1 hour before the rechallenge with oxazolone did not substantially affect the early response measured after 2 hours (Figure 2B, compare rows 2 and 3). In contrast, the late phase of the CHS response measured after 24 hours was strongly repressed by this treatment (Figure 2B, compare rows 5 and 6). In conclusion, GCs repress CHS during the late challenge phase without interfering with the early wave of leukocyte infiltration.

Immunosuppressive functions of GCs require the presence of the GR in myeloid cells. Next, we wished to identify the cell types sensitive to GC action by the use of conditional GR mutant mice. We crossed K14CreERT2 transgenic mice (21) with GR^flox mice (22) to study whether effects on keratinocytes contribute to the observed suppression of CHS. After we induced recombination of the GR^flox allele with tamoxifen, these GR^K14Cre mice selectively lacked the receptor in keratinocytes (data not shown). Administration of DNFB induced a massive swelling response, which was effectively repressed by dexamethasone irrespective of the genotype (Figure 3A). Thus the presence of the GR in keratinocytes is not essential for the antiinflammatory effect of dexamethasone in the treatment of CHS.

Figure 3

GC-mediated suppression of CHS requires GR actions in myeloid cells. (A) Swelling response of GR^flox and GR^K14CreERT2 mice (shown as % increase in thickness; *P < 0.01; n = 5–6). All mice were sensitized at day 0 with vehicle (Veh) or DNFB and challenged with DNFB in the presence or absence of dexamethasone. (B) Swelling response of GR^flox and GR^lckCre mice treated as described in A. (C) FACS analysis of BM cells from GR^LysMCre mice crossed with RA/EG mice containing a tk-loxP-EGFP reporter gene (GR^LysMCre;RA/EG mice) to determine the expression of EGFP and GR-1. Numbers indicate the percentage of cells in each quadrant. (D) PCR analysis of genomic DNA isolated from GR-1⁺ granulocytes positive or negative for EGFP from GR^LysMCre;RA/EG mice in comparison with genomic DNA derived from GR^flox/GR^null heterozygous mice (50% per allele). (E) Swelling response of GR^flox and GR^LysMCre mice treated as described in A.

Because CHS is a T cell–dependent hypersensitivity response, we next analyzed the capacity of GCs to repress inflammation in mice carrying a T cell–specific deletion of the GR gene (GR^lckCre mice). These animals were generated by crossing GR^flox mice (22) with lckCre transgenic mice (23), which display an effective and selective disruption of the GR gene only in the T cell lineage (24). Unexpectedly, the dexamethasone-mediated suppression of CHS, induced by either DNFB or oxazolone, was completely unaffected in GR^lckCre mice (Figure 3B and data not shown). Consequently, the control of T cell functions by GCs is also dispensable for the inhibitory action of dexamethasone on CHS.

Since the second wave of leukocyte infiltration is primarily mediated by neutrophils and macrophages, we studied mice that lack the GR in myeloid cells (GR^LysMCre). The animals were generated by crossing GR^flox mice with lysozyme M Cre knockin mice (25). To determine the recombination efficiency in various myeloid cell types, GR^LysMCre mice were mated with tk-loxP–enhanced GFP (tk-loxP-EGFP) reporter (RA/EG) animals (26). Cells from bone marrow, spleen, and lymph nodes were isolated and analyzed by flow cytometry (Figure 3C). Almost complete recombination of the loxP sites was achieved in neutrophils (more than 90%) and macrophages (70%), whereas recombination in LCs (43%), DCs (16%), and mast cells (26%) was less effective (Table 1). T and B cells were largely unaffected in these mice (data not shown). Effective disruption of the GR gene in neutrophils was further demonstrated by the complete conversion of the GR^flox allele into the GR^null allele in FACS-sorted EGFP⁺GR-1⁺ granulocytes (Figure 3D). Whereas DNFB-induced swelling was normally repressed by dexamethasone in control mice, the same treatment completely failed to suppress the inflammatory response in GR^LysMCre mice (Figure 3E). Histological analysis demonstrated the persistence of massive leukocyte infiltration in the mutant mice despite treatment with GCs (data not shown). These results demonstrate that GR-controlled functions in macrophages and neutrophil granulocytes are essential for the suppression of CHS by GCs, indicating that these cells are the primary targets of dexamethasone-mediated therapy.

Table 1

Recombination efficiency of a tk-flox-EGFP allele in LysM-Cre knockin mice

DNA binding of the GR is required for the suppression of CHS by GCs. To define the mechanism of GR action in the suppression of CHS, we took advantage of the function-selective GR^dim mutation. We have previously demonstrated that interference with dimerization-induced DNA binding in GR^dim mice does not impair the antiinflammatory activity of GCs in irritative inflammation (17). To determine whether GCs are similarly able to suppress a T cell–dependent allergic inflammation in GR^dim mice, we induced CHS with DNFB (Figure 4A) and oxazolone (Figure 4B), followed by administration of dexamethasone during the challenge phase. Unexpectedly, the inflammatory response was hardly affected by dexamethasone treatment in GR^dim mice in contrast to wild-type mice (Figure 4, A and B). Histological examination of inflamed ears after oxazolone treatment revealed a strong infiltration by leukocytes in wild-type and GR^dim mice. As expected, ears of wild-type mice showed only few infiltrating cells when treated with oxazolone and dexamethasone. However, in GR^dim ears dexamethasone did not prevent a persistent accumulation of leukocytes (Figures 5 and 6). The number of CD68⁺ macrophages located in the dermis was not diminished in ear skin of GR^dim mice (Figure 5). Similarly, massive infiltration by GR-1⁺ neutrophils of dermis and epidermis after oxazolone treatment was reduced by dexamethasone only in the skin of wild-type but not GR^dim mice (Figure 6). These results indicate that dimerization of the GR and GRE-dependent transcription (27) are essential for the suppression of an allergic inflammation.

Figure 4

GC-mediated suppression of CHS is impaired in GR^dim mice. (A) Swelling response observed in wild-type and GR^dim mice sensitized with vehicle or DNFB. Mice were challenged with DNFB in the presence or absence of dexamethasone. *P < 0.01 (n = 5–6). (B) Swelling response in wild-type and GR^dim mice sensitized with vehicle or oxazolone. Mice were challenged with oxazolone in the presence or absence of dexamethasone. *P < 0.01 (n = 5–6).

Figure 5

Reduction of macrophages by dexamethasone is impaired in the skin of GR^dim mice. Confocal images of ear sections from wild-type and GR^dim mice sensitized with oxazolone followed by administration of oxazolone alone or together with dexamethasone and then stained with CD68 (green). Nuclei are stained in blue (DAPI). Bottom row shows magnification of boxed regions in the third row. Scale bars: 50 μm.

Figure 6

Reduction of neutrophil granulocytes by dexamethasone is hampered in the skin of GR^dim mice. Confocal images of ears of the same specimens used in Figure 5, stained with antibodies against GR-1 (green). Nuclei are stained in blue (DAPI). Bottom row shows magnification of boxed regions in the third row. Scale bars: 50 μm.

Similar impairment of GC suppression of leukocyte infiltration in the CHS response in GR^LysMCre mice and GR^dim mice indicates that factors that fail to be regulated in myeloid cells from GR^dim mice may represent essential targets of GC action in CHS. Therefore, we determined the level of release of proinflammatory cytokines and chemokines by BM-derived macrophages (BMDMs) of wild-type and GR^dim mice. TNF-α, IL-1β, MCP-1, MIP-2, and IP-10 are important mediators of leukocyte migration in CHS (4). Dexamethasone repressed LPS-induced TNF-α release to a similar extent in wild-type and GR^dim cells (Figure 7A). In contrast, inhibition of IL-1β and MCP-1 release by GCs was significantly impaired in GR^dim cells (Figure 7, B and C). Importantly, the synthesis of the MIP-2 and IP-10 was hardly repressed by dexamethasone in GR^dim macrophages (Figure 7, D and E). It is noteworthy that protein levels were regulated to a similar extent as the mRNA levels of the respective genes (Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI28034DS1). These data suggest that impaired downregulation of IL-1β, MCP-1, MIP-2, and IP-10 may contribute to the persistent inflammation during CHS in dexamethasone-treated GR^dim mice.

Figure 7

GC-mediated repression of cytokines and chemokines in LPS-activated BMDMs of wild-type and GR^dim mice. BMDMs were treated with LPS or LPS and dexamethasone, and cytokine release in the supernatant was determined for TNF-α after 8 hours (A), for IL-1β after 8 hours (B), for MCP-1 after 24 hours (C), for MIP-2 after 8 hours (D), and for IP-10 after 24 hours (E). For IL-1β, BMDMs were also treated with ATP in the last 30 minutes. Statistical analysis was performed for at least 3 independent experiments. *P < 0.05.

The functional relevance of the factors analyzed for the repression of CHS by GCs was confirmed by local administration of recombinant proteins to dexamethasone-treated mice during the challenge phase (Figure 8). The antiinflammatory effect of dexamethasone was significantly impaired by exogenous IL-1β, MCP-1, MIP-2, and IP-10 (P < 0.05; Figure 8, A–D) but not by application of TNF-α (P = NS; Figure 8E). Administration of cytokines or chemokines tested individually did not evoke a complete reversal of GC-induced repression of CHS, which suggests that they might act in synergy or that other proteins might contribute. However, our findings do indicate that these proinflammatory mediators participate in the suppressive activity of GCs in the treatment of allergic skin conditions.

Figure 8

Cytokines and chemokines with impaired downregulation in GR^dim macrophages interfere with GC-mediated suppression of CHS. Mice were sensitized and challenged with oxazolone. During the challenge phase they were also treated either with dexamethasone (Ox+Dex) alone or with recombinant IL-1β (A), MCP-1 (B), MIP-2 (C), IP-10 (D), or TNF-α (E) at the concentration indicated in Methods. *P < 0.05 (n = 6).

GCs applied at pharmacological doses are strong inhibitors of contact dermatitis, which makes them the treatment of choice for this condition. Here we describe molecular and cell type–specific mechanisms underlying the beneficial effects of GCs, exemplified by CHS, a classic rodent model of human allergic dermatitis. We show that the immune response elicited by reexposure to hapten is not suppressed by GCs during the sensitization phase but is strongly reduced when applied during the challenge phase. In this phase the presence of the GR in keratinocytes and T cells is dispensable, whereas GR expression is required in myeloid cells for effective immunosuppression. These GC actions depend on DNA binding of the GR and involve repression of IL-1β, MCP-1, MIP-2, and IP-10.

In agreement with an earlier study (19), we observed that GCs applied during the sensitization phase of CHS suppressed the accumulation of APCs in the draining lymph nodes but not the development of an inflammatory response. Thus, GCs apparently do not interfere with T cell priming, which indicates that a reduced number of APCs in the lymph nodes still allows for the generation of sufficient antigen-specific T cells capable of initiating an immune response. This is in agreement with the previous observation that fewer than 50,000 DCs are required to support CHS (28). In contrast, systemic as well as topical application of GCs during the challenge phase prevents an inflammatory response. However, this effect was limited to the late inflammatory reaction that consists of the influx of myeloid and T cells after cytokine and chemokine release, whereas the early swelling response was unaffected.

Given the well documented role of T cells in CHS response, our observation that GCs strongly repress inflammation in mice lacking the GR in T cells was unexpected. Sensitized T cells were previously found to transfer antigen specificity to recipient animals (29). In addition, T cell–depleted, CD8-knockout, and CD4-knockout mice exhibit an impaired CHS response (5, 30). While these experiments point to a central role of T cells in CHS, recent investigations showed that skin inflammation does not necessarily depend on T cells. In a psoriasis model established in keratinocyte-specific IKK2-knockout mice, the inflammatory response was retained after genetic ablation of T cells (31). In Rag2^–/– and SCID mice, hapten-specific CHS reactions occur normally as in wild-type mice (8). In our model, at least inhibition of T cell activity by GCs is not essential to block CHS. This is supported by the observation that GR^LysMCre mice, which carry a functional GR in T cells — but not in myeloid cells — are resistant to GC treatment. Keratinocytes participate in the pathogenesis of skin diseases (reviewed in ref. 32), e.g., by the release of TNF-α necessary for DC mobilization (33, 34). However, tamoxifen-treated GR^K14Cre mice, which lack the GR in the entire epidermis still exhibit a significant reduction of the swelling response after GC application. Keratinocytes are a source of CHS-mediating cytokines, but they are not a critical target for GCs in CHS. In contrast, we observed a complete block of GC actions on CHS in mice lacking the GR in myeloid cells (GR^LysMCre). These animals primarily lack the GR in macrophages and neutrophils, while DCs and mast cells are less affected by Cre-induced recombination. Although 42% of the LCs showed Cre-induced recombination, LCs should not be a major target for GCs because depletion of LCs by GCs or genetic ablation of LCs enhanced CHS responses (35, 36). Thus, the persistent activity of macrophages and neutrophils in the skin after GC treatment is the most likely reason for the impaired immunosuppression in GR^LysMCre mice. Interestingly, macrophages were recently shown to be required for neutrophil infiltration and inflammatory lesions in 2 models of psoriasis (37, 38). In one of them, namely mice with a hypomorphic CD18 allele, dexamethasone treatment was able to resolve this inflammation (39). These studies support our finding that GC action on macrophages is crucial for curing skin conditions.

Interestingly, repression of CHS responses via GCs in myeloid cells requires binding of the GR, as we demonstrate here. This is in contrast to our previous findings in an irritative inflammatory model of GR^dim mice, in which inflammation was resolved by dexamethasone (17). In irritative skin inflammation, the lack of the GR in myeloid cells does not abrogate the antiinflammatory activity by GC. Thus, the requirement of GR DNA binding for the antiinflammatory activities of GCs appears to be important of certain immunological processes that involve myeloid cells as targets for GCs.

Infiltration of the skin by CD68⁺ macrophages and GR-1⁺ neutrophils in GR^dim mice was not prevented by dexamethasone. This suggests that the persistent infiltration of leukocytes is likely due to inflammatory mediators that fail to be regulated by a dimerization- and DNA binding–deficient GR. The fact that both GR^LysMCre and GR^dim mice are resistant to GC-mediated repression of CHS suggests that a functional DNA-binding competent GR in myeloid cells is required for this effect. Since the cellular composition of dexamethasone-treated wild-type and GR^dim skin is different, it is not feasible to quantitatively determine the levels of inflammatory mediators per macrophage in tissue samples. Therefore, BMDMs from GR^dim mice were studied ex vivo as a well defined macrophage model (40). Although these cells do not represent skin resident macrophages in every respect, these cells show the classic features of activated macrophages when stimulated with LPS, such as the expression of proinflammatory mediators (40) known to be involved in CHS (4). Our experiments revealed that suppression of several of these mediators by dexamethasone was impaired in GR^dim macrophages. Moreover, when applied in vivo, these factors indeed partially counteracted dexamethasone-mediated repression of CHS. Collectively, using ex vivo–generated macrophages with a DNA binding–impaired GR, we identified proinflammatory genes, which are GC targets participating in the efficacy of steroid treatment of CHS. TNF-α release was effectively repressed by dexamethasone in GR^dim macrophages, and administration of recombinant TNF-α to the foot pads of mice did not significantly counteract GC-mediated suppression of CHS. Thus, TNF-α does not seem to be the important GC target, but it could still act in synergy with other chemokines or cytokines.

In contrast, IL-1β release was not effectively repressed in GR^dim macrophages. IL-1β is known to play a pivotal role in inflammatory responses (41), although the requirement for it in CHS depends on the specific hapten. While IL-1β–knockout mice fail to mount a hypersensitivity response to low doses of trinitrochlorobenzene (42), they strongly respond to oxazolone (43). On the other hand, we found that exogenous application of IL-1β to oxazolone-treated mice slightly impaired the dexamethasone-mediated suppression. Studies with IL-1β–knockout mice treated with oxazolone and dexamethasone could clarify this question. We also demonstrate that dexamethasone repression of MCP-1, a potent chemoattractant molecule, was significantly impaired in GR^dim macrophages. While MCP-1–knockout animals are able to mount a normal CHS response, they fail to recruit leukocytes to the site of inflammation (44). This is in agreement with our finding that the number of macrophages in the inflamed ears of GR^dim mice was not reduced after dexamethasone treatment. The treatment with recombinant MCP-1 led to a partial relief of dexamethasone repression. Thus, inhibition of MCP-1 release is presumably relevant for the antiinflammatory activity of GCs. Downregulation of MIP-2 protein secretion by GCs was completely abolished in GR^dim BMDMs. MIP-2 is an attractant molecule for neutrophils, and neutralization of MIP-2 leads to an attenuation of virally induced hypersensitivity response (45). In addition, mice lacking the receptor for MIP-2 exhibit an impairment of CHS (46). Administration of recombinant MIP-2 to oxazolone- and dexamethasone-treated ears also causes a slight attenuation of dexamethasone-mediated suppression of CHS. Therefore, inhibition of MIP-2 could participate in diminishing the CHS response by GC via reducing infiltration of neutrophils. Similarly, IP-10 was hardly downregulated in GR^dim cells by dexamethasone. IP-10 is also known to be required for a complete induction of CHS because the IP-10–knockout mice show a reduced swelling response (13, 47). IP-10 is induced during the challenge phase in keratinocytes and monocytic cells in the skin (48). Activated infiltrating T cells express CXCR3, the receptor for IP-10 (49), thus linking increased expression of IP-10 to the recruitment of sensitized T cells to the site of inflammation. In line with our finding that GCs fail to repress IP-10 in GR^dim cells, administration of recombinant IP-10 partially reversed the suppressive effect of dexamethasone on the CHS response. Thus, IP-10 could also be a crucial target for GC-mediated suppression of CHS.

In summary, several cytokines and chemokines participate in the impaired suppression of macrophage and/or neutrophil function in GR^dim mice. Consistently, administration of each recombinant mediator to oxazolone-treated mice only partially reversed the suppressive effect of dexamethasone. Our data suggest that there is not a single target of the antiinflammatory action of GCs in CHS but that a network of inflammatory mediators is controlled by the GR dimer.

We have identified the critical cell types for GC action and determined that DNA binding by the GR is involved in the molecular mechanism that mediates the therapeutic effects of GCs in the treatment of allergic skin conditions. Our results indicate that the most promising strategy for the future will be to target genes in myeloid cells, in particular chemoattractant proteins and cytokines, which interfere with leukocyte migration to the site of hapten exposure. Interestingly, low molecular weight antagonists for chemokine receptors, such as CCR2-binding (50) MCP-1 or IL-1β receptor inhibitor (41), are being tested in animal models and in clinical trials and could serve as promising alternatives for steroid therapy in contact dermatitis.

View Supplemental data

We would like to acknowledge Falk Weih and Helen Morrison for critical reading of the manuscript; Stefan Berger for designing the GR^flox PCR; Bernd Arnold for providing the RA/EG mice; Andrea Takacs, Katrin Buder, Susanne Ostermay, and Sabine Hübner for excellent technical assistance; and Sabine Matz for maintaining the mouse colonies. This work was supported by the Deutsche Forschungsgemeinschaft through Collaborative Research Centres SFB488, SFB405, and SFB636 and grants FOR Ot 165/2-2, GRK 791/1.02, GRK 484, Sachbeihilfe Tu-220/3, and Schu 51/7-2; by the Fonds der Chemischen Industrie; by the European Community through grant QLG1-CT-2001-01574 and LSHM-CT-2005-018652 (CRESCENDO); by the Bundesministerium für Bildung und Forschung (BMBF) through Nationales Genomforschungsnetz (NGFN) grants FZK 01GS01117, 01GS0477, and KGCV1/01GS0416; by the German-Polish cooperation project 01GZ0310 and 0313074C (systems biology); by the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO; grant VID 917.76.365); by the Landsteiner Foundation for Blood Transfusion Research (LSBR, 0414); and by the Austrian Fonds zur Förderung der wissenschaftlichen Forschung (grant SFB F28).

Address correspondence to: Günther Schütz, German Cancer Research Center, Division of Molecular Biology of the Cell I, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany. Phone: 49-6221-423411; Fax: 49-6221-423470; E-mail: g.schuetz@dkfz.de. Or to: Jan P. Tuckermann, Leibniz Institute for Age Research, Fritz-Lipmann-Institute, Division of Tissue-Specific Hormone Action, Beutenberg Str. 11, D-07745 Jena, Germany. Phone: 49-3641-656134; Fax: 49-3641-656133; E-mail: jan@fli-leibniz.de.

Nonstandard abbreviations used: BMDM, BM-derived macrophages; CHS, contact hypersensitivity; DNFB, 2,4-dinitrofluorobenzene; EGFP, enhanced GFP; GC, glucocorticoid; GR, GC receptor; GRE, GC-responsive element; IP-10, IFN-γ–inducible protein 10; LC, Langerhans cell; MCP-1, monocyte chemoattractant protein-1; MIP-2, macrophage inflammatory protein-2.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.117:1381–1390 (2007). doi:10.1172/JCI28034

1. Leung, D.Y., Boguniewicz, M., Howell, M.D., Nomura, I., Hamid, Q.A. 2004. New insights into atopic dermatitis. J. Clin. Invest. 113:651-657.
   View this article via: JCI PubMed Google Scholar
2. Barnes, P.J. 1998. Anti-inflammatory actions of glucocorticoids: molecular mechanisms. Clin. Sci. (Lond.). 94:557-572.
   View this article via: PubMed Google Scholar
3. Grabbe, S., Schwarz, T. 1998. Immunoregulatory mechanisms involved in elicitation of allergic contact hypersensitivity. Immunol. Today. 19:37-44.
   View this article via: CrossRef PubMed Google Scholar
4. Wang, B., et al. 2003. Cytokine knockouts in contact hypersensitivity research. Cytokine Growth Factor Rev. 14:381-389.
   View this article via: PubMed Google Scholar
5. Wang, B., et al. 2000. CD4+ Th1 and CD8+ type 1 cytotoxic T cells both play a crucial role in the full development of contact hypersensitivity. J. Immunol. 165:6783-6790.
   View this article via: PubMed Google Scholar
6. Campos, R.A., et al. 2003. Cutaneous immunization rapidly activates liver invariant Valpha14 NKT cells stimulating B-1 B cells to initiate T cell recruitment for elicitation of contact sensitivity. J. Exp. Med. 198:1785-1796.
   View this article via: CrossRef PubMed Google Scholar
7. Campos, R.A., et al. 2006. Interleukin-4-dependent innate collaboration between iNKT cells and B-1 B cells controls adaptative contact sensitivity. Immunology. 117:536-547.
   View this article via: CrossRef PubMed Google Scholar
8. O’Leary, J.G., Goodarzi, M., Drayton, D.L., von Andrian, U.H. 2006. T cell– and B cell–independent adaptive immunity mediated by natural killer cells. Nat. Immunol. 7:507-516.
   View this article via: CrossRef PubMed Google Scholar
9. Tsuji, R.F., et al. 2002. B cell–dependent T cell responses: IgM antibodies are required to elicit contact sensitivity. J. Exp. Med. 196:1277-1290.
   View this article via: CrossRef PubMed Google Scholar
10. Tsuji, R.F., et al. 2000. Early local generation of C5a initiates the elicitation of contact sensitivity by leading to early T cell recruitment. J. Immunol. 165:1588-1598.
    View this article via: PubMed Google Scholar
11. McHale, J.F., Harari, O.A., Marshall, D., Haskard, D.O. 1999. Vascular endothelial cell expression of ICAM-1 and VCAM-1 at the onset of eliciting contact hypersensitivity in mice: evidence for a dominant role of TNF-alpha. J. Immunol. 162:1648-1655.
    View this article via: PubMed Google Scholar
12. Hwang, J.M., Yamanouchi, J., Santamaria, P., Kubes, P. 2004. A critical temporal window for selectin-dependent CD4+ lymphocyte homing and initiation of late-phase inflammation in contact sensitivity. J. Exp. Med. 199:1223-1234.
    View this article via: CrossRef PubMed Google Scholar
13. Dufour, J.H., et al. 2002. IFN-gamma–inducible protein 10 (IP-10; CXCL10)-deficient mice reveal a role for IP-10 in effector T cell generation and trafficking. J. Immunol. 168:3195-3204.
    View this article via: PubMed Google Scholar
14. Tuckermann, J.P., Kleiman, A., McPherson, K.G., Reichardt, H.M. 2005. Molecular mechanisms of glucocorticoids in the control of inflammation and lymphocyte apoptosis. Crit. Rev. Clin. Lab. Sci. 42:71-104.
    View this article via: CrossRef PubMed Google Scholar
15. Beato, M., Herrlich, P., Schutz, G. 1995. Steroid hormone receptors: many actors in search of a plot. Cell. 83:851-857.
    View this article via: CrossRef PubMed Google Scholar
16. Herrlich, P. 2001. Cross-talk between glucocorticoid receptor and AP-1. Oncogene. 20:2465-2475.
    View this article via: CrossRef PubMed Google Scholar
17. Reichardt, H.M., et al. 2001. Repression of inflammatory responses in the absence of DNA binding by the glucocorticoid receptor. EMBO J. 20:7168-7173.
    View this article via: CrossRef PubMed Google Scholar
18. Tuckermann, J.P., et al. 1999. The DNA binding-independent function of the glucocorticoid receptor mediates repression of AP-1–dependent genes in skin. J. Cell Biol. 147:1365-1370.
    View this article via: CrossRef PubMed Google Scholar
19. Cumberbatch, M., Dearman, R.J., Kimber, I. 1999. Inhibition by dexamethasone of Langerhans cell migration: influence of epidermal cytokine signals. Immunopharmacology. 41:235-243.
    View this article via: CrossRef PubMed Google Scholar
20. Askenase, P.W. 2001. Yes T cells, but three different T cells (alphabeta, gammadelta and NK T cells), and also B-1 cells mediate contact sensitivity. Clin. Exp. Immunol. 125:345-350.
    View this article via: CrossRef PubMed Google Scholar
21. Indra, A.K., et al. 2000. Targeted somatic mutagenesis in mouse epidermis. Horm. Res. 54:296-300.
    View this article via: CrossRef PubMed Google Scholar
22. Tronche, F., et al. 1999. Disruption of the glucocorticoid receptor gene in the nervous system results in reduced anxiety. Nat. Genet. 23:99-103.
    View this article via: CrossRef PubMed Google Scholar
23. Gu, H., Marth, J.D., Orban, P.C., Mossmann, H., Rajewsky, K. 1994. Deletion of a DNA polymerase beta gene segment in T cells using cell type-specific gene targeting. Science. 265:103-106.
    View this article via: CrossRef PubMed Google Scholar
24. Baumann, S., et al. 2005. Glucocorticoids inhibit activation-induced cell death (AICD) via direct DNA-dependent repression of the CD95 ligand gene by a glucocorticoid receptor dimer. Blood. 106:617-625.
    View this article via: CrossRef PubMed Google Scholar
25. Clausen, B.E., Burkhardt, C., Reith, W., Renkawitz, R., Forster, I. 1999. Conditional gene targeting in macrophages and granulocytes using LysMcre mice. Transgenic Res. 8:265-277.
    View this article via: CrossRef PubMed Google Scholar
26. Constien, R., et al. 2001. Characterization of a novel EGFP reporter mouse to monitor Cre recombination as demonstrated by a Tie2 Cre mouse line. Genesis. 30:36-44.
    View this article via: CrossRef PubMed Google Scholar
27. Reichardt, H.M., et al. 1998. DNA binding of the glucocorticoid receptor is not essential for survival. Cell. 93:531-541.
    View this article via: CrossRef PubMed Google Scholar
28. Macatonia, S.E., Knight, S.C., Edwards, A.J., Griffiths, S., Fryer, P. 1987. Localization of antigen on lymph node dendritic cells after exposure to the contact sensitizer fluorescein isothiocyanate. Functional and morphological studies. J. Exp. Med. 166:1654-1667.
    View this article via: CrossRef PubMed Google Scholar
29. Hauser, C. 1990. Cultured epidermal Langerhans cells activate effector T cells for contact sensitivity. J. Invest. Dermatol. 95:436-440.
    View this article via: CrossRef PubMed Google Scholar
30. Biedermann, T., et al. 2000. Mast cells control neutrophil recruitment during T cell-mediated delayed-type hypersensitivity reactions through tumor necrosis factor and macrophage inflammatory protein 2. J. Exp. Med. 192:1441-1452.
    View this article via: CrossRef PubMed Google Scholar
31. Pasparakis, M., et al. 2002. TNF-mediated inflammatory skin disease in mice with epidermis-specific deletion of IKK2. Nature. 417:861-866.
    View this article via: CrossRef PubMed Google Scholar
32. Albanesi, C., Scarponi, C., Giustizieri, M.L., Girolomoni, G. 2005. Keratinocytes in inflammatory skin diseases. Curr. Drug Targets Inflamm. Allergy. 4:329-334.
    View this article via: CrossRef PubMed Google Scholar
33. Piguet, P.F., Grau, G.E., Hauser, C., Vassalli, P. 1991. Tumor necrosis factor is a critical mediator in hapten induced irritant and contact hypersensitivity reactions. J. Exp. Med. 173:673-679.
    View this article via: CrossRef PubMed Google Scholar
34. Enk, A.H., Katz, S.I. 1992. Early molecular events in the induction phase of contact sensitivity. Proc. Natl. Acad. Sci. U. S. A. 89:1398-1402.
    View this article via: CrossRef PubMed Google Scholar
35. Grabbe, S., Steinbrink, K., Steinert, M., Luger, T.A., Schwarz, T. 1995. Removal of the majority of epidermal Langerhans cells by topical or systemic steroid application enhances the effector phase of murine contact hypersensitivity. J. Immunol. 155:4207-4217.
    View this article via: PubMed Google Scholar
36. Kaplan, D.H., Jenison, M.C., Saeland, S., Shlomchik, W.D., Shlomchik, M.J. 2005. Epidermal langerhans cell-deficient mice develop enhanced contact hypersensitivity. Immunity. 23:611-620.
    View this article via: CrossRef PubMed Google Scholar
37. Stratis, A., et al. 2006. Pathogenic role for skin macrophages in a mouse model of keratinocyte-induced psoriasis-like skin inflammation. J. Clin. Invest. 116:2094-2104.
    View this article via: JCI PubMed Google Scholar
38. Wang, H., et al. 2006. Activated macrophages are essential in a murine model for T cell–mediated chronic psoriasiform skin inflammation. J. Clin. Invest. 116:2105-2114.
    View this article via: JCI PubMed Google Scholar
39. Bullard, D.C., et al. 1996. A polygenic mouse model of psoriasiform skin disease in CD18-deficient mice. Proc. Natl. Acad. Sci. U. S. A. 93:2116-2121.
    View this article via: CrossRef PubMed Google Scholar
40. Lang, R., Patel, D., Morris, J.J., Rutschman, R.L., Murray, P.J. 2002. Shaping gene expression in activated and resting primary macrophages by IL-10. J. Immunol. 169:2253-2263.
    View this article via: PubMed Google Scholar
41. Dinarello, C.A. 1996. Biologic basis for interleukin-1 in disease. Blood. 87:2095-2147.
    View this article via: PubMed Google Scholar
42. Shornick, L.P., et al. 1996. Mice deficient in IL-1beta manifest impaired contact hypersensitivity to trinitrochlorobenzone. J. Exp. Med. 183:1427-1436.
    View this article via: CrossRef PubMed Google Scholar
43. Zheng, H., et al. 1995. Resistance to fever induction and impaired acute-phase response in interleukin-1 beta-deficient mice. Immunity. 3:9-19.
    View this article via: CrossRef PubMed Google Scholar
44. Lu, B., et al. 1998. Abnormalities in monocyte recruitment and cytokine expression in monocyte chemoattractant protein 1–deficient mice. J. Exp. Med. 187:601-608.
    View this article via: CrossRef PubMed Google Scholar
45. Tumpey, T.M., Fenton, R., Molesworth-Kenyon, S., Oakes, J.E., Lausch, R.N. 2002. Role for macrophage inflammatory protein 2 (MIP-2), MIP-1alpha, and interleukin-1alpha in the delayed-type hypersensitivity response to viral antigen. J. Virol. 76:8050-8057.
    View this article via: CrossRef PubMed Google Scholar
46. Cattani, F., et al. 2006. The role of CXCR2 activity in the contact hypersensitivity response in mice. Eur. Cytokine Netw. 17:42-48.
    View this article via: PubMed Google Scholar
47. Dilulio, N.A., et al. 1999. Groalpha-mediated recruitment of neutrophils is required for elicitation of contact hypersensitivity. Eur. J. Immunol. 29:3485-3495.
    View this article via: CrossRef PubMed Google Scholar
48. Mitsui, G., et al. 2003. Kinetic profiles of sequential gene expressions for chemokines in mice with contact hypersensitivity. Immunol. Lett. 86:191-197.
    View this article via: CrossRef PubMed Google Scholar
49. Qin, S., et al. 1998. The chemokine receptors CXCR3 and CCR5 mark subsets of T cells associated with certain inflammatory reactions. J. Clin. Invest. 101:746-754.
    View this article via: JCI PubMed Google Scholar
50. Brodmerkel, C.M., et al. 2005. Discovery and pharmacological characterization of a novel rodent-active CCR2 antagonist, INCB3344. J. Immunol. 175:5370-5378.
    View this article via: PubMed Google Scholar
51. Bauer, A., et al. 1999. The glucocorticoid receptor is required for stress erythropoiesis. Genes Dev. 13:2996-3002.
    View this article via: CrossRef PubMed Google Scholar
52. Robbiani, D.F., et al. 2000. The leukotriene C(4) transporter MRP1 regulates CCL19 (MIP-3beta, ELC)-dependent mobilization of dendritic cells to lymph nodes. Cell. 103:757-768.
    View this article via: PubMed CrossRef Google Scholar