IL-1β-dependent activation of dendritic epidermal T cells in contact hypersensitivity

Abstract

Substances that penetrate the skin surface can act as allergens and induce a T cell-mediated inflammatory skin disease called contact hypersensitivity (CHS). IL-17 is a key cytokine in CHS and was originally thought to be produced solely by CD4⁺ T cells. However, it is now known that several cell types including γδ T cells can produce IL-17. Here, we determine the role of γδ T cells, especially the dendritic epidermal T cells (DETC), in CHS. By use of a well-established model for CHS where dinitroflourobenzen (DNFB) is used as allergen, we found that γδ T cells are important players in CHS. Thus, an increased number of IL-17 producing DETC appear in the skin following exposure to DNFB in WT mice, and DNFB-induced ear-swelling is reduced by approximately 50% in TCRδ^−/− mice compared to WT mice. In accordance, DNFB-induced ear-swelling was reduced by approximately 50% in IL-17^−/− mice. We show that DNFB triggers DETC activation and IL-1β production in the skin, and that keratinocytes produce IL-1β when stimulated with DNFB. We find that DETC activated in vitro by incubation with anti-CD3 and IL-1β produce IL-17. Importantly, we demonstrate that the IL-1 receptor antagonist anakinra significantly reduces CHS responses as measured by decreased ear-swelling, inhibition of local DETC activation and a reduction in the number of IL-17⁺ γδ T cells and DETC in the draining lymph nodes. Taken together, we show that DETC become activated and produce IL-17 in an IL-1β-dependent manner during CHS suggesting a key role for DETC in CHS.

Introduction

Contact hypersensitivity (CHS) that manifests as redness, swelling and itching of the skin is a T cell-mediated inflammatory disease induced by exposure of the skin to contact allergens. IL-17 is an important cytokine in the pathogenesis of various autoimmune and allergic diseases, and IL-17^−/− mice show a strong reduction in ear-swelling compared to wild type (WT) mice in experimental models of CHS (1). In line with this, several studies have suggested the involvement of IL-17 in CHS by finding Th17 and IL-17 producing CD8⁺ T cells in allergic reactions in both humans and mice (2–5). However, recent studies have found that Th17 and CD8⁺ T cells are not the only sources of IL-17. Among the newly identified important sources of IL-17 during early inflammatory responses are the γδ T cells and the NKT cells (6,7).

Splenic γδ T cells can produce IL-17 following stimulation with IL-1β and IL-23 independently of TCR stimulation (7). This suggests that these IL-17-producing γδ T cells play an innate role in the immune response (7). Interestingly, both IL-1β and IL-23 are up-regulated in the skin following exposure to contact allergens (3,5,8). Recently it was shown that dermal γδ T cells can be activated in a similar fashion as splenic γδ T by IL-1β and IL-23, and it was suggested that the major source of IL-17 during psoriatic skin inflammation is the dermal γδ T cells (9). Dermal γδ T cells are however not the only subtype of γδ T cells found in the skin. Murine epidermis contains a large number of a special subpopulation of Vγ3⁺ γδ T cells, (Garman nomenclature) (10) called dendritic epidermal T cells (DETC) due to their dendritic morphology (11). The role of DETC as IL-17 producing cells is still debated (7,9,12–15). Interestingly, it was recently shown that a subset of DETC can produce IL-17 (14). This DETC subset was shown to be critical for the epidermal barrier function and for effective wound healing (14).

Several studies have found that γδ T cells can have both a pro- and anti-inflammatory roles in contact hypersensitivity (16–23). In adoptive transfer experiments, γδ T cells have been found to assist αβ T cells in mediating the CHS response in a non-allergen and non-MHC specific manner (21). The role of γδ T cell help in the CHS response was further investigated, and DETC were identified as the γδ T cell assisting the αβ T cells (17,18). Furthermore, in vitro studies found that keratinocytes could induce DETC proliferation in a non-MHC restricted way following exposure to various contact allergens, whereas exposure to irritants could not (20). Thus, these studies indicated that DETC play an inflammatory role in the immune response to contact allergens. However, other studies found that DETC have a regulatory role in CHS by down-regulating the response in an allergen-specific, MHC-independent pathway (16,19,22,23). Thus, the role of DETC in CHS is unclear and seems to depend on the experimental settings, including the presence or absence of keratinocytes.

The aims of this study were to investigate the role of γδ T cells especially DETC in CHS and to determine the ability of DETC to produce IL-17. By use of a well-established model for CHS where dinitroflourobenzen (DNFB) is used as allergen, we found that γδ T cells as well as IL-17 are required in CHS as seen by reduced ear-swelling in TCRδ^−/− mice and IL-17^−/− mice compared to WT mice following DNFB exposure (24). We found that the numbers of IL-17 producing DETC were significantly increased in the skin of WT mice following exposure to DNFB. Furthermore, analysis of epidermal ear-sheets demonstrated that DETC became activated following exposure to 2,4-dinitrobenzenesulphonic acid (DNSB). Purified DETC did however not become activated when stimulated with the allergen, which indicated that a different route of activation than direct allergen recognition is required for DETC activation. We found that IL-1β is produced in the skin following exposure to DNFB and that keratinocytes up-regulate IL-1β mRNA following DNSB treatment. We show that TCR stimulation and IL-1β act in synergy in mediating DETC activation and IL-17 production. Furthermore, i.d. injection with anti-CD3 in combination with IL-1β induced an increased ear-thickness compared to mice treated with PBS. Interestingly, DETC activated in vitro with anti-CD3 and IL-1β induced an increased ear-thickness following i.d. injection compared to mice injected with unstimulated DETC. Finally, treatment with the IL-1 receptor antagonist anakinra inhibited ear-swelling, local DETC activation and the numbers of DETC and IL-17⁺ γδ T cells in the dLN after DNFB exposure. We suggest a model where allergen exposure induces IL-1β production and expression of the still unknown TCR-ligand for DETC in keratinocytes. Together, these molecules induce DETC activation and IL-17 production required for the CHS reaction.

Material and Methods

Mice

Female C57Bl/6 (WT) mice were purchased from Taconic (Ry, Denmark). B6.129P2-Tcrd^tm1Mom/J (TCRδKO) mice were purchased from Jackson (San Diego, California). IL-17^−/− mice were provided by Y. Iwakura (University of Tokyo, Tokyo, Japan)(1) and K. Ley (La Jolla Institute for Allergy and Immunology, La Jolla, California, USA). Mice were housed in specific pathogen-free facilities at either the Department of Experimental Medicine, Panum Institute, The University of Copenhagen, in accordance with national animal protection guidelines (license number: 2007/561-1357) or at The Scripps Research Institute, in according to The Scripps Research Institute IACUC guidelines.

Materials and reagents

2,4-dinitrofluorobenzene (DNFB), 2,4-dinitrobenzenesulphonic acid (DNSB), Phorbol 12-myristate 13-acetate (PMA), concavalin A (Con A), indomethacin, ionomycin and monensin were purchased from Sigma Aldrich (Brøndby, Denmark). rmIL-1β, rmIL-23, IFN-γ, IL-10, TGFβ ELISA kits, fixable viability dye eFluor780, and intracellular staining kit was purchased from eBioscience (San Diego, CA, USA). Mouse TH1/TH2 9-plex and mouse IL-17A Ultra-sensitive kits were purchased from Meso Scale Discovery. Anti-CD16/CD32 (2.4G2), anti-γδ (GL3), anti-Vγ3 (536), anti-CD3ε (145–2C11), anti-IL17 (TC11-18H10), anti-IFN-γ (XMG1.2) and isotype control (R3–34) were all purchased from BD Pharmingen. Anti-CD25 (PC61) was purchased from Biolegend, San Diego, CA, USA. TaqMan Reverted Aid First Strand cDNA synthesis Kit (K1622), TaqMan Universal PCR master mix (4326708) and primers GAPDH (Mm999999_g1) and IL-1β (Mm01336189) were purchased from Applied Biosystems.

Cells and culture conditions

The murine keratinocyte cell line PAM2.12 was cultured in Dulbecco’s Modified Eagle Medium (DMEM medium) at 37°C, 5% CO₂ (Sigma Aldrich). DMEM was supplemented with 10% FBS, 0.5 IU/L penicillin, 500 mg/L streptomycin, 1% L-glutamine, 2-Mercaptoethanol, 0.63 mM HEPES (Gibco, San Diego, CA, USA), 1 mM Na Pyruvate (Gibco, USA). Medium was changed every other day and cells were split when 80% confluent. Freshly isolated DETC were cultured in flat-bottomed 96 well plates in RPMI 1640 at 37°C, 5% CO₂. RPMI 1640 was supplemented with 10% FBS 0.5 IU/L penicillin, 500 mg/L streptomycin, 1% L-glutamine, 2-Mercaptoethanol, 0.63 mM HEPES, 1 mM Na Pyruvate, 1 µM non-essential amino acids (Gibco), 5 U/ml IL-2 (Invitrogen, USA). Every other day half of the medium was gently removed and fresh medium was added.

Contact hypersensitivity

To induce CHS, mice were painted with 25 µl 0.15% DNFB in a 1:4 olive olie:acetone (OOA) mixture on the dorsal side of both ears for three consecutive days (0–2). On day 23, mice were challenged on the dorsal side of both ears with 25µl 0.15% DNFB in OOA. Control mice were exposed to the vehicle on day 23. Mice were euthanized 24 hours after challenge and ear thickness was measured on both ears using an engineer’s micrometer (Mitutoyo, Tokyo, Japan). Draining lymph nodes were harvested and single cell suspensions were prepared for further analysis.

In vivo blocking of IL-1β

C57Bl/6 mice were sensitized for three consecutive days (day 0–2) and challenged on day 23. Anakinra (200 µl of 150 µg/µl) or PBS was administered i.p. 12 hours and immediately before challenge. 24 hours post challenge ear thickness was measured and dLN were collected for FACS analysis.

Cell sorting

To produce epidermal cell suspensions, the hair was first removed from euthanized naïve C57Bl/6 mice. Skin pieces were then floated with the dermis-side down in 0.3% Trypsin/GNK at 37°C overnight. The next day the epidermis was peeled from the dermis and treated with Trypsin/GNK with 0.1% DNase (Sigma Aldrich). Lymphocytes were purified using Lympholyte M (Gibco, USA) and plated in complete DETC medium containing 2 µg/ml ConA and 2 µg/ml indomethocin. Every second day half of the medium was removed and fresh medium added. Cells were expanded in culture for 3 weeks prior to the experiments. Cultured epidermal cells were collected and non-specific binding was blocked by anti-CD16/CD32 (2.4G2). Cells were stained with anti-Vγ3 (536), anti-CD3 (145–2C11) and anti-γδ (GL3). DETC (CD3⁺γδ⁺Vγ3⁺) cells were sorted on an Aria to a purity of more than 98%.

Intracellular cytokine staining

Single cell suspensions from the draining lymph nodes were adjusted and plated at 1 × 10⁶ cells/well and re-stimulated with PMA (1.25 µg/ml) andionomycin (625 ng/ml) in complete RPMI including monensin (2.08 µg/ml) for 5 hours at 37°C. Subsequently, cells were treated with anti-CD16/CD32 (2.4G2) to block non-specific binding, washed and stained with Anti-γδ (GL3) and anti-Vγ3 (536). Cells were fixed and permeabilized with eBioscience intracellular staining kit. Cells were subsequently stained with anti-IL17 (TC11-18H10) or anti-IFN-γ (XMG1.2). Single cell suspension from epidermis was obtained as described in Cell sorting. Cells were rested O/N to allow re-expression of trypsin-sensitive surface markers. The following day cells were incubated in the presence of PMA (1.25 µg/ml) and ionomycin (625 ng/ml) in complete RPMI including monensin (2.08 µg/ml) for 5 hours at 37°C. Subsequently, cells were treated with anti-CD16/CD32 (2.4G2) to block non-specific binding, washed and stained with anti-CD3 (145–2C11), anti-γδ (GL3), anti-Vγ3 (536) and fixable viability dye eFluor780. Cells were fixed and permeabilized with eBioscience intracellular staining kit. Cells were subsequently stained with anti-IL17 (TC11-18H10), anti-IFN-γ (XMG1.2) or anti-IL-10 (JE55-16E3). Cell samples were analyzed on a FACSCaliburor Fortessa (BD, Brøndby, DK) with Cellquest Pro or FACS Diva software. Data was analyzed using Flowjo.

Stimulation and preparation of ear sheets

Ear sheets were prepared by separating the dorsal and ventral sides of the ears. Subsequently ear-pieces were floated dermis side down for 20 h (DNSB stimulation) or 24 h (cytokine or anti-CD3 stimulation) at 37°C in DMEM medium containing either DMSO, 0.1% DNSB, 10 ng/ml IL-1β, or 1 µg/ml anti-CD3. Subsequently epidermal sheets were prepared as previously described (25). Sheets were stained with anti-γδ TCR (GL3) for 1 h at RT, washed and mounted on slides using DAKO fluorescent mounting medium (Carpenteria, San Diego, California, USA). Samples were imaged directly after staining with a Nikon Eclipse E800 microscope. Digital images were collected with an AxioCamHRc camera and AxioCam software (Zeiss, Oberkochen, Germany) and image processing was performed using Adobe Photoshop and InDesign. Changes in the DETC morphology after anti-CD3 and cytokine treatment were assessed by counting the number of dendrites on each cell (more than two dendrites = resting; one or two dendrites = partially activated; no dendrites = activated). A minimum of 200 cells from at least 3 experiments was counted for each treatment.

IL-1β expression profile in mouse skin and keratinocytes

Mice were exposed once to 0.15% DNFB in OOA or as a control to OOA. At the indicated times post allergen exposure, mice were euthanized and 8 mm punch biopsies were immediately taken from both ears. Biopsies were split in half and snap-frozen and stored at −80°C for later protein or RNA purification. Biopsies were individually crushed and lysed simultaneously using Precellys technology (Bertin Technologies, France). For mRNA analysis, RNA was purified using the mirVana mRNA purification kit (Life Technologies, Germany) and reverse-transcribed using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA, USA). The samples were amplified by real-time RT-PCR using Applied Biosystems validated gene expression assays and PRISM 7900HT sequence detection system (SDS 2.3). Fold changes of mRNA expression were calculated by the comparative Ct method and normalized to GAPDH using the RealTimeStatMiner software (Integromics, Granada, Spain). For protein analysis, total amounts of protein in each sample were quantified by the Pierce BCA Protein Assay Kit (Thermo Scientific, Rockford, IL) and total protein concentration of each sample was standardized before assessment for cytokines expression. Cytokines from homogenized ear biopsies were measured by Mouse TH1/TH2 9-Plex Ultra-Sensitive Kits on a meso scale discovery platform. PAM2.12 cells were plated at 2.5 × 10⁵ cells/well in 6 well plates and incubated overnight at 37°C. The following day cells were stimulated for the indicated time and subsequently harvested using 0.5% Trypsin/EDTA (Gibco, USA). RNA extraction was performed using the standard protocol RNeasy mini kit (Qiagen, USA) and reverse transcribed as described for tissue lysates.

DETC stimulation

Following cell sorting, DETC were rested O/N and re-stimulated to determine IL-17 production. In brief, 1.0 × 10⁵ DETC/well were stimulated with rmIL-1β, rmIL-23 (10 ng/ml), anti-CD3 (1 µg/ml) or combinations of these. For IL-17A detection, cells were incubated at 37°C and supernatant was collected at 12, 24 and 48 hours and analyzed using the mouse IL-17A Ultra-Sensitive kits on a meso scale discovery platform (MSD). For IFN-γ, IL-10 and TGF-β detection supernatant was collected at 24 h and analyzed by ELISA. For FACS analysis of sorted DETC, cells were re-stimulated for 24 hours and stained with anti-Vγ3 (536), anti-γδ (GL3), Anti-CD25 (PC61) and anti-IL17 (TC11-18H10) as described in Intracellular cytokine staining.

Intradermal (i.d.) injections of anti-CD3, IL-1β and DETC

C57Bl/6 mice were injected i.d. in the ears with 20 µl of 1 µg/ml anti-CD3, isotype control, 10 ng/ml IL-1β alone or in combination. 24 hours after the injections, ear thickness was measured to evaluate the degree of inflammation. For cell transfer experiments, sorted DETC were either left untreated or stimulated with 1 µg/ml anti-CD3 and 10 ng/ml IL-1β for 24 hours prior to i.d. injection. 20 µl fresh medium containing 1 × 10⁴ DETC was injected i.d. in the ears, and the ear thickness was measured 24 hours after the injection.

Results

γδ T cells and IL-17 are involved in the pathogenesis of CHS

To investigate the role of γδ T cells and IL-17 in CHS, we compared CHS responses to DNFB in WT, TCRδ^−/− and IL-17^−/− mice as measured by ear-thickness. We found that the ear-swelling was significantly reduced in both TCRδ^−/− mice and IL-17^−/− mice compared to WT mice (Figure 1A and B). Interestingly, suppression of the CHS response was similar in IL-17^−/− and TCRδ^−/− mice. The important role of IL-17 in CHS is in good agreement with a previous study (1); however, it is still unknown whether skin-derived γδ T cells contribute to the production of IL-17 during CHS. To investigate this, we isolated and re-stimulated cells from the dLN and skin from DNFB-exposed and control WT mice with PMA and ionomycin, and determined intracellular IL-17. A significantly increased number of γδ T cells and IL-17⁺ γδ T cells were found in the dLN of DNFB-exposed mice compared to control mice (Figure 1C and D). To analyze whether skin derived γδ T cells contributed to the increased number of γδ T cells in the dLN, we determined the number of DETC in the dLN. A ten-fold increase in the number of DETC was found in DNFB-exposed mice compared to control mice (Figure 1E). However, the DETC did not seem to produce IL-17A and IFN-γ in the dLN (Figure 1F). In contrast, when analyzing IL-17A and IFN-γ production by DETC isolated from the skin, an increased number of IL-17A⁺ and IFN-γ⁺ DETC was found in mice exposed to DNFB compared to control mice (Figure 1G and H). Taken together, these results show that both γδ T cells and IL-17 are involved in the pathogenesis of CHS, and that an increased number of DETC that produces IL-17A and IFN-γ is fund in skin exposed to DNFB. Furthermore, the numbers of γδ T cells including DETC and IL-17⁺ γδ T cells increase in the dLN during CHS. However, DETC do not seem to produce IL-17A and IFN-γ in the dLN.

Figure 1. γδ T cells are involved in the pathogenesis of CHS.

(A) C57BL/6, IL-17^−/− and TCRδ^−/− mice were split into two groups with 4–5 mice pr. group. Mice were either exposed to vehicle or DNFB on the dorsal side of the ears for three consecutive days (day 0–2) and then re-exposed to DNFB on day 23. Ear thickness was measured 24 hours after the final exposure and is shown as relative to vehicle-treated mice. (B) Shows suppression of the ear inflammation in DNFB-treated IL-17^−/− and TCRδ^−/− mice compared to WT. The suppression was calculated as [(Ear-thickness_{(IL-17−/−, DNFB)} −(Ear-thickness_{(IL-17−/−, vehicle)})/(Ear-thickness_{(WT, DNFB)} −(Ear-thickness_{(WT, vehicle)})] * 100%. (C) The total number of γδ T cells (GL3⁺ cells) (D), IL-17 producing γδ T cells (GL3⁺IL-17⁺ cells), and (E) DETC (GL3⁺Vγ3⁺ cells)in dLNs of C57BL/6 mice exposed to vehicle or DNFB. (F) Representative FACS plots of Vγ3⁺IL-17A⁺ cells and Vγ3⁺IFN-γ⁺ cells in the dLN. (G) Percentages of Vγ3⁺IL-17A⁺ cells and Vγ3⁺IFN-γ⁺ cells in epidermis from C57BL/6 mice exposed to vehicle or DNFB. (H) Representative FACS plots of Vγ3⁺IL-17A⁺ and Vγ3⁺IFN-γ⁺ cells in epidermis from C57BL/6 mice exposed to vehicle or DNFB. Student’s t-test was performed (**p<0.01; ***p<0.001). Data are shown as mean + SEM of two independent experiments comprising 4–5 mice pr. group.

Exposure of the skin to allergens leads to DETC activation in the epidermis

As we found that DETC from the skin produce IL-17A and IFN-γ following exposure of the skin to DNFB, we wanted to study whether exposure of the skin to DNFB induced DETC activation in situ. It has previously been shown that DETC change their morphology and round up during their activation in the epidermis (11). To investigate whether exposure of the skin to allergens activates DETC in situ, we treated epidermal ear-sheets with 0.1% DNSB or vehicle for 20 h and subsequently determined the morphology of the DETC. We found that DETC retained their classical dendritic morphology in epidermal ear-sheets treated with vehicle, whereas DNSB treatment induced rounding up of the DETC (Figure 2A). To investigate whether DETC activation was caused by direct recognition of the allergen, we purified DETC and treated them for 24 h with non-supplemented medium, or medium supplemented with either 0.1% DNSB, anti-CD3 or anti-CD3 plus IL-1β. DETC activation was accessed by CD25 up-regulation (Figure 2B and C). DNSB did not activate DETC, whereas stimulation through the TCR with anti-CD3 did (Figure 2B and C). Interestingly, addition of IL-1β seemed to increase the CD25 expression on the DETC although this was not significant. Taken together, these results indicate that DETC are activated in the epidermis when allergens penetrate the skin, and furthermore, that they are not directly but rather indirectly activated by the allergens via allergen-induced activation of other cells in the skin.

Figure 2. Exposure of the skin to allergens leads to DETC activation in the epidermis.

(A) Ears sheets from untreated mice were treated for 20 h with vehicleor 0.1% DNSB. The epidermal ear-sheets were subsequently isolated and stained with anti-γδ TCR (green), 20× magnification. Data are representative of 5 separate experiments. (B and C) DETC were isolated by sorting CD3⁺ GL3⁺, Vγ3⁺ cells and subsequently rested O/N. The DETC were then stimulated for 24 h as indicated and CD25 expression was determined by FACS. (B) FACS histograms of CD25 expression on untreated DETC and DETC treated with anti-CD3, anti-CD3 + IL-1β or DNSB. (C) CD25 expression on DETC treated with anti-CD3, anti-CD3 + IL-1β or DNSB normalized to CD25 expression on untreated DETC. Values are shown as mean + SEM of 4 individual experiments.

DETC produce IL-17A when stimulated with IL-1β, IL-23 and anti-CD3

It has been reported that whereas spleen and dermal γδ T cells have the ability to produce IL-17A independently of TCR stimulation following treatment with IL-1β and IL-23, DETC do not have this ability (7,9). To investigate this further, we purified DETC from the epidermis (Figure 3A) and treated them for the indicated periods with IL-1β, IL-23 andanti-CD3 either alone or in or combination. We found that DETC treated with IL-1β, IL-23 or anti-CD3 alone produce low amounts of IL-17A (Figure 3B). However, IL-23 and anti-CD3 in combination resulted in a 50% increase in IL-17A production compared to cells treated with anti-CD3 or IL-23 alone. Even more impressive, combined IL-1β and anti-CD3 treatment resulted in a 3-fold increase in the IL-17A secretion compared to cells stimulated with anti-CD3 or IL-1β alone (Figure 3B). No additional effect was seen when IL-23 was added together with IL-1β. To further validate that DETC have the capacity to produce IL-17A, we purified DETC, stimulated them for 24 hours with anti-CD3 alone or in combination with IL-1β, and measured intracellular IL-17A (Figure 3C). We found that DETC can produce IL-17A following TCR stimulation. Again, when IL-1β was added in combination with anti-CD3 a marked increase in IL-17A⁺ DETC was observed (Figure 3C). As we found an increased number of IL-17A⁺ and IFN-γ⁺ DETC in the skin following exposure of the skin to DNFB, we wanted to investigate the effect of anti-CD3 and IL-1β on IFN-γ production by DETC. Furthermore, as the role of γδ T cells during CHS has been suggested to be anti-inflammatory we investigated if treatment with anti-CD3 and IL-1β lead to induction anti-inflammatory cytokines IL-10 and/or TGFβ(17,18,22,23). Treatment with anti-CD3 induced the production of both IFN-γ and IL-10 (Figure 3D). Addition of IL-1β increased the production of IFN-γ, whereas it did not seem to have an effect on the IL-10 production (Figure 3D). Interestingly, neither anti-CD3 alone or in combination with IL-1β lead to production of TGFβ (Figure 3D). Treatment with DNSB did not induce production of any of the cytokines (Figure 3D). Taken together, these results indicate that DETC have the capacity to produce IL-17A, and that TCR and IL-1β signaling in combination is highly potent in inducing IL-17A production in DETC. The same was seen for IFN-γ production, whereas IL-10 production only depended on TCR signaling. Finally, DNSB treatment did not induce cytokine production by DETC, further supporting that DNFB-induced activation of DETC in situ is mediated by an indirect pathway.

Figure 3. DETC produce IL-17A when stimulated with IL-1β, IL-23 and anti-CD3.

(A) Representative FACS plot of DETC purity after cell sorting on CD3⁺GL3⁺Vγ3⁺ cells from short term epidermal cultures. (B) DETC were sorted, rested overnight and then treated with IL-1β, IL-23, anti-CD3 or combination of these for the time indicated. Subsequently, supernatants were collected and the concentration of IL-17A determined. Values are shown as mean + SEM and data are representative of two independent experiments. (C) IL-17A profile of sorted DETC cultivated for 24 h in medium without additions, medium with anti-CD3 or with both anti-CD3 and IL-1β. Data is representative of two separate experiments. (D) DETC were sorted, rested O/N and then treated with anti-CD3, anti-CD3 + IL1β or DNSB for 24 h. Subsequently the supernatants were collected and the concentration of IFN-γ, IL-10 and TGF-β was determined. Student’s t-test was performed (**p<0.01; ***p<0.001). Values are shown as mean + SEM of 4 individual experiments.

Allergens activate keratinocytes to produce IL-1β that activate DETC

It has been shown that IL-1β is produced in the skin following allergen exposure (8). Based on this observation and our findings described above, we hypothesized that activation of DETC in the epidermis following allergen exposure is mediated by IL-1β derived from keratinocytes activated by the allergen. To determine if IL-1β was produced in the skin in our experimental setup, we treated mice with DNFB on the dorsal side of the ears. After 6, 24 and 48 h the IL-1β mRNA levels in the ears were determined by real-time RT-PCR. IL-1β was rapidly up-regulated after exposure to DNFB (Figure 4A). The protein levels of IL-1β in ear tissue extracts were in good agreement with the mRNA results (Figure 4B). To determine if allergens directly induce IL-1β in keratinocytes, we stimulated the PAM2.12 keratinocyte cell line for various periods with DNSB. We found that exposure to DNSB induced a significant up-regulation of IL-1β mRNA (Figure 4C). IL-1β mRNA peaked after 24 hours of DNSB treatment with an approximate 25-fold up-regulation compared to cells treated with vehicle (Figure 4C). To determine the role of IL-1β in DETC activation, epidermal ear-sheets were treated with IL-1β or anti-CD3 for 24 h and DETC morphology determined. Non-treated ear-sheets contained mostly highly dendritic DETC (more than two dendrites), whereas the majority of DETC had a partial retraction of dendrites following treatment of the ear-sheets with IL-1β, suggesting that IL-1β induced partial activation of the DETC (Figure 4D). Full DETC activation, as indicated by a rounded morphology, was seen in approximately 80% of the DETC following treatment with anti-CD3 (Figure 4D). These results suggested that DETC become activated in situ when stimulated with IL-1β or anti-CD3. To investigate this further, we injected mice with anti-CD3 alone or in combination with IL-1β i.d. in the ears and measured the ear thickness 24 h later. We found that anti-CD3 alone and in combination with IL-1β induced significant ear-swelling (Figure 4E). To analyze whether DETC might contribute to this anti-CD3-induced ear-swelling, we injected purified DETC that had been activated with anti-CD3 and IL-1β in vitro or were untreated i.d. in the ears of untreated mice. We found that injection of activated DETC induced increased ear swelling compared to injections of untreated DETC (Figure 4F). Taken together, these experiments suggested that keratinocytes are directly activated by DNFB and start to produce IL-1β, that together with the still un-identified DETC TCR ligand induce DETC activation, IL-17 production and skin inflammation.

Figure 4. Allergens induce keratinocytes to produce IL-1β that activate DETC.

(A) Fold change of IL-1β mRNA in ear biopsies from mice exposed to DNFB for 6, 24 and 48 h compared to mice exposed to vehicle. The mRNA level of IL-1β was determined by real-time RT-PCR. The relative expression was determined compared to GAPDH. Students’t-test was applied on the ΔCT values obtained from the real-time RT-PCR (vehicle-treated compared to DNFB treated-treated) (**=p<0.01). (B) The protein levels of IL-1β in ear biopsies from mice exposed to vehicle or DNFB for 6, 24 and 48 h. (C) Fold change of IL-1β mRNA in PAM2.12 cells treated with 0.1% DNSB for the indicated time compared to PAM2.12 cells treated with vehicle. The relative expression was determined compared to GAPDH. Values are shown as mean ± SEM for 2 separate experiments, n=4. (D) Epidermal ear-sheets were treated with IL-1β or anti-CD3 for 24 h and the morphology of the DETC was subsequently determined by fluorescence microscopy. Inserts show examples of fluorescence microscopy images of DETC stained with anti-γδTCR antibodies in red. A minimum of 200 cells from at least 3 experiments were counted and characterized based on their number of dendrites (more than two dendrites = resting DETC; one or two dendrites = partially activated DETC; no dendrites = activated DETC).(E and F) Ear thickness of C57BL/6 mice 24 h after i.d. injection with (E) anti-CD3, anti-CD3 + IL1β, PBS or isotype control; (F) untreated DETC or DETC pre-activated with anti-CD3 and IL1β. Student’s t-test was performed (**p<0.01). Values are shown as mean + SEM of 2 individual experiments, n=4.

IL-1β is required for DNFB-induced DETC activation in vivo

To determine the role of IL-1β in CHS and DETC activation in vivo, CHS responses in WT mice treated with the IL-1 receptor antagonist anakinra or PBS were investigated. We found that mice treated with anakinra had significantly reduced ear-swelling compared to mice treated with PBS (Figure 5A and B). Furthermore, anakinra treatment lead to reduced numbers of total γδ T cells, IL-17⁺γδ T cell and DETC in the dLN (Figure 5C–E). To investigate whether the reduced number of DETC in the dLN from mice treated with anakinra could be explained by inhibition of DETC activation in the epidermis, we determined the DETC activation in situ as measured by DETC morphology in mice treated with either anakinra or PBS. DETC were clearly activated in the epidermis from PBS-treated mice as judged by the round shape of the DETC (Figure 5F). In contrast, DETC retained their dendritic morphology in the epidermis from anakinra-treated mice, indicating that DETC are in a low/none activation state (Figure 5F). Taken together, these results indicate that IL-1β is one of the key cytokines for a full blown CHS response and for the activation and migration of DETC to the dLN during CHS.

Figure 5. IL-1β is required for DNFB-induced DETC activation in vivo.

Mice were sensitized and challenged with DNFB as described in materials and methods. Before challenge mice were either given anakinra or PBS i.p. (A) Ear thickness measured 24 h after DNFB exposure.(B) Percent suppression of inflammation in mice treated with anakinra compared to mice treated with PBS. (C) Total number of γδ T cells (GL3⁺), (D) IL-17A producing γδ T cells (GL3⁺IL-17⁺) and (E) DETC (GL3⁺Vγ3⁺) in the dLN of mice treated with PBS or anakinra before challenge with DNFB. Data represent two individual experiments with 4 mice pr. group. Student’s t-test was performed. (F) Epidermal ear sheets from mice treated with PBS or anakinra before challenge with DNFB were isolated and stained with anti-γδ TCR (green), 20× magnification. Data are representative of 2 separate experiments.

Discussion

In this study, we show that both γδ T cells and IL-17 are involved in the pathogenesis of CHS, and that γδ T cells, IL-17⁺γδ T cells and DETC accumulate in the dLN following exposure of the skin to allergens. We found that exposure of the skin to DNFB leads to local production of IL-1β and activation of DETC in the epidermis as seen by the rounding of the DETC and increased numbers of DETC producing IL-17A and IFN-γ. We demonstrate that DETC have the ability to produce IL-17A, IFN-γ and IL-10 following TCR stimulation, and that the TCR-induced IL-17A and IFN-γ production is significantly enhanced by IL-1β. We show that anti-CD3 and IL-1β injected i.d. into the ears induce significant ear-swelling, and furthermore, that DETC activated in vitro induce an increased ear-swelling following i.d. injection into the ears. Finally, we found that blocking of IL-1 receptor in vivo resulted in decreased ear-swelling and DETC activation, and significantly lowered the numbers of γδ T cells, IL-17⁺γδ T cells and DETC in the dLN.

In agreement with Nakae et al, we found that IL-17 is important for the CHS response as mice lacking IL-17 had an approximately 50% reduced response to DNFB as measured by changes in ear-thickness compared to WT mice (1). Furthermore, we found a strong reduction in ear-swelling following allergen exposure in TCRδ^−/− mice compared to WT mice on the C57Bl/6 background. In contrast, Girardi et al. found that γδ T cells did not have an effect on the CHS response in mice on the C57Bl/6 background (26). However, the protocols used in the two studies differ significantly. Girardi et al. sensitized the mice on the abdominal skin and challenged on the ears already 5 days after sensitization. We sensitized for three consecutive days on the ears and challenged after a minimum of 21 days on the ears (26). It is well known that primary adaptive responses during CHS induction and infectious diseases peak between day 5 and 10 depending on the presence of allergen and type of infection (27,28). We have previously shown that the primary response is ongoing in the draining lymph nodes for at least 10 days when using our CHS protocol (24). Thus, most likely the primary response is still ongoing at day 5 in the protocol used by Girardi et al., and a challenge at this time would primarily be driven by the activated CD4⁺ and CD8⁺ effector T cells. In contrast, our model is likely to be more dependent on early responders like the DETC. Although γδ T cells are thought to be more innate than αβ T cells, it has been shown that they can develop a memory-like phenotype and that they become more efficient during secondary antigen exposure (29,30). Furthermore, existence of memory-like γδ T cells in CHS is further supported by previous studies, which found that DETC are required for adaptive transfer of CHS (18).

Conflicting results have been reported concerning the role of γδ T cells as pro- or anti-inflammatory in CHS. Adaptive transfer experiments have indicated that γδ T cells are required for optimal transfer of CHS from sensitized to allergen-inexperienced mice (17,18,21). Interestingly, DETC were shown to be the γδ T cell subset required for the full transfer of CHS (17,18). In contrast, transfer of DETC primed with allergen in vitro induced tolerance towards the specific allergen (17,18,22,23). The divergent results obtained in these experiments might be explained by the different conditions during priming of DETC in vivo and in vitro. Based on the changes in DETC morphology, we found that DETC become activated following exposure of ears/ear-sheets to DNFB/DNSB, and that the number of DETC increases in the draining lymph nodes following DNFB exposure of the skin. We found that highly purified DETC from naïve mice produce IL-17, IFN-γ and IL-10 when stimulated with anti-CD3. This TCR-induced production of IL-17 and IFN-γ was further increased by IL-1β, whereas the IL-10 production was unaffected. From this it might be suggested that DETC play a pro-inflammatory role in the allergic response as long as IL-1β is present in the skin, but that they later on when the IL-1β concentration wane can change to an anti-inflammatory role.

The ability of DETC to produce IL-17 is in agreement with previous studies, which found that DETC produce IL-17 following skin infection with Staphylococcus aureus, when stimulated with PMA and ionomycin in the presence of IL-1β, IL-23 or a combination of the two, and following skin wounding(12,14). In contrast, other studies using either mycobacterium or a psoriasis model found that DETC were unable to produce IL-17 and that it is the dermal γδ T cells that are the primary source of IL-17 in the skin(7,9,13,15). The different results obtained concerning the ability of DETC to produce IL-17 might be explained by the requirement for TCR stimulation. Thus, DETC might require TCR triggering whereas dermal γδ T cells seem to be able to produce IL-17 by a cytokine-dependent, TCR-independent pathway as reported for spleen γδ T cells (7,13). Cai et al. used an IL-23-induced psoriasis model to investigate the main IL-17-producing cell types within the skin, and it is likely that this model did not induce any stress to the keratinocytes because IL-23 was injected i.d. (9). A recent study, using this model characterized a subtype of γδ T cells in mouse epidermis that was distinct from DETC as they did not express Vγ3 (31). This γδ T cell subtype was suggested to be the main IL-17 and IL-22 producing cell-type in the skin when the IL-23 induced psoriasis model was used(31). Interestingly, in a CHS model using DNFB as allergen the same group was unable to detect this γδ T cell subtype in the epidermis (31). This indicated that different γδ T cell subtypes exist in the epidermis. It is also likely that different types of skin infections have different effects on the keratinocytes (12,32). Interestingly, DETC have higher levels of TCR expression at their surface than dermal γδ T cells, which might indicate that DETC are more dependent on TCR signaling than dermal γδ T cells (9,32).

Taken together, we show that γδ T cells play a pro-inflammatory role during CHS. We believe that our data support the hypothesis that DETC produce IL-17 when allergens penetrate the skin, and that this is important for the immune response to allergens in the skin. Based on our study and a previous study that demonstrated up-regulation of a still uncharacterized ligand for the DETC TCR at the cell surface of stressed keratinocytes (33), we propose the following model for cross-talk between DETC and keratinocytes: Exposure of the skin to allergens leads to keratinocyte stress with rapid production of IL-1β and up-regulation of the still un-identified TCR ligand on the keratinocyte cell surface. Together, IL-1β and the TCR ligand provide the necessary signals required for full activation and IL-17 production by DETC and the development of CHS. Interestingly, IL-1β is rapidly up-regulated in the skin of allergic patients after exposure to contact allergens (34). Although DETC is a unique cell population in murine skin, other subset of γδ T cells seems to play a role as IL-17 producing cells in human inflammatory skin diseases like psoriasis and contact allergy (35,36). Finally, our study suggests that anakinra might be efficient in the treatment of patients with severe contact dermatitis in part by inhibiting IL-17 production from γδ T cells. However, this needs further investigations.