Mast cell-derived IL-13 downregulates IL-12 production by skin dendritic cells to inhibit the Th1 response to cutaneous antigen exposure.

Abstract

Background:

Atopic dermatitis (AD) is characterized by a skin barrier defect, aggravated by mechanical injury inflicted by scratching, a Th2-dominated immune response and susceptibility to viral skin infections normally restrained by a Th1 response. The signals leading to a Th2-dominated immune response in AD are not completely understood.

Objective:

Determine the role of IL-13 in the initiation of the Th cell response to cutaneously encountered antigens.

Methods:

WT, Il13^−/,⁻ Il1rl1^−/−, Il4ra^−/− and mice with selective deficiency of IL-13 in mast cells (MCs) were studied, dendritic cells (DCs) purified from the draining lymph nodes (dLNs) of tape-stripped and ovalbumin (OVA)-sensitized skin were examined for their ability to polarize naïve OVA-TCR transgenic CD4⁺ T cells. Cytokine expression was examined by RT-qPCR, intracellular flow cytometry and ELISA. Contact hypersensitivity (CHS) to dinitrofluorobenzene (DNFB) was examined.

Results:

Tape stripping caused IL-33-driven upregulation of Il13 expression by skin MCs. MC-derived IL-13 acted on DCs from dLNs of OVA-sensitized skin to selectively suppress their ability to polarize naïve OVA-TCR transgenic CD4⁺ T cells into interferon-γ (IFN-γ) secreting cells. MC-derived IL-13 inhibited the Th1 response in CHS to DNFB. IL-13 suppressed IL-12 production by mouse skin-derived DCs in vitro and in vivo. Scratching upregulated IL13 expression in human skin, and IL-13 suppressed the capacity of lipopolysaccharide-stimulated human skin DCs to express IL-12 and promote IFN-γ secretion by CD4⁺ T cells.

Conclusion:

Release of IL-13 by cutaneous MCs in response to mechanical skin injury inhibits the Th1 response to cutaneous antigen exposure in AD.

Graphical Abstract

CAPSULE SUMMARY.

Mechanical skin injury causes IL-33-driven upregulation of Il13 expression by MCs, which selectively suppresses IL-12 production by skin-derived DCs restraining their ability to polarize naïve CD4⁺ T cells into Th1 cells.

Introduction.

Atopic dermatitis (AD) is characterized by a skin barrier defect, a Th2 dominated immune response to environmental allergens and foods, allergic skin inflammation with dermal infiltration by Th2 cells and eosinophils, and susceptibility to cutaneous viral infection e.g. eczema herpeticum, eczema vaccinatum and molluscum contagiosum^{1, 2}. Mechanical injury inflicted by scratching dry itchy skin aggravates the skin barrier defect in AD. Normal skin is impermeable to cutaneously encountered antigens. Defective skin barrier function allows antigen entry which is essential to drive an immune response. Mutations in genes coding for proteins involved in epidermal barrier function such as FLG, DSG1, SPINK5 and CDSN are associated with AD³. The signals leading to a Th2-dominated immune response to cutaneously encountered antigens, development of allergic skin inflammation and subsequent susceptibility to viral skin infections in AD are not well understood.

Dendritic cells (DCs) are abundant in the skin and in other body interfaces and are critical for the adaptive immune response to antigen. The composition and migration of skin DCs are heterogenous⁴. The skin harbors at steady state, conventional DCs and monocyte-derived DCs in the dermis and Langerhans cells in the epidermis, and plasmacytoid DCs populate inflamed skin⁴. Following its entry through a disrupted skin barrier, antigen is taken up and processed by conventional DCs in the dermis, which rapidly migrate to draining lymph nodes (dLNs) where they drive the expansion of recirculating naïve antigen-specific CD4⁺ T cells and their differentiation into cytokine producing cells^4–6. The migration of conventional DCs is followed later by the migration of Langerhans cells⁷. DCs integrate the cues present in their microenvironment to polarize the T cell response. We and others have previously shown that thymic stromal lymphopoietin (TSLP) released following mechanical skin injury acts on skin DCs to promote their ability to polarize T cells towards a Th2 phenotype^8–11. In addition, we have shown that mechanical skin injury inhibits the ability of skin DCs emigrants to drive Th1 polarization⁸. The molecules that mediate this inhibition remain unknown.

IL-13 is a type 2 cytokine produced by Th2 cells, NKT cells, mast cells (MCs), basophils and type 2 innate lymphoid cells (ILC2s)¹². IL-13 is expressed in AD skin lesions and plays an important role in the pathogenesis of the disease¹³. IL-13 upregulates the expression of Th2 cell-attracting chemokines by stromal cells, impairs epidermal integrity and promotes epidermal hyperplasia^14–16. There is strong evidence that IL-13 acts in the effector phase of the immune response to dampen Th1 cell activation. In a mouse model of asthma DCs from mediastinal LNs of Il13^−/− mice exhibit enhanced capacity to promote IFN-γ production in memory allergen-specific CD4⁺ T cells¹⁷. Furthermore, IL-13 inhibits IFN-γ production and is protective in experimental models of autoimmune disease and graft rejection^18–21.

The role for IL-13 in the initiation of the immune response to cutaneously encountered antigen has not been explored. We demonstrate that mechanical injury to mouse skin promotes IL-33 mediated upregulation of Il13 expression in skin MCs, and that MC-derived IL-13 selectively inhibits the capacity of skin-derived DCs to polarize naïve mouse CD4⁺ T cells to a Th1 phenotype. Furthermore, we demonstrate that scratching induces IL13 upregulation in human skin, and that IL-13 suppresses the ability of human skin DCs to promote IFN-γ secretion by CD4⁺ T cells.

Methods.

Mice.

C57BL/6J WT, BALB/c WT and DO11.10 TCR transgenic mice were purchased from Charles River Laboratory. Rag2^−/− mice on BALB/c background were purchased from Taconic. Rag2^−/−gc^−/− mice on B6 background and Il4/13^flox/flox on BALB/c background were purchased from the Jackson Laboratory. Il13^−/−, Il1rl1^−/− and Kit^W-sh/W-sh mice on BALB/c background mice were previously described^22–24. Mcpt5-cre mice on BALB/c background were crossed with Il13^−/−, then were crossed with Il4/13^flox/flox mice. All mice were kept in a pathogen-free environment and fed an OVA-free diet. All procedures performed on the mice were in accordance with the Animal Care and Use Committee of the Boston Children’s Hospital.

Human Subjects.

After obtaining informed consent, the inner side of the forearm of three healthy non-allergic adult subjects was scratched 30 times with a #11 sterile blade with care not to draw blood. 6 hours later a 4 mm punch biopsy was obtained from the scratched site and another one from a skin site on the contralateral forearm. The samples were kept in RNAlater and stored at −80° C. All procedure performed were in accordance with the Institutional Review Board of the Boston Children’s Hospital

Preparation of skin cell suspensions from human skin and isolation of skin DCs.

Human surgical discards of abdominal skin were obtained from the Human Skin Disease Resource Core at Brigham and Women’s Hospital. To obtain cells all the fat was removed using a scalpel and the skin was chopped into small pieces and digested for 2 hours at 37°C with vigorous shaking in complete RPMI containing Liberase LT (0.2 mg/ml, Roche) and DNAse I (0.2 mg/ml, Sigma). Digested tissue was mechanically disrupted using a plunger, filtered, centrifuged, and resuspended. Skin DCs were sorted using the CD1c (BDCA-1)⁺ dendritic cell isolation kit from Miltenyi Biotec, according to the manufacturer’s instructions.

Functional analysis of human skin DCs.

1×10⁴ skin DCs were stimulated overnight with LPS (100 ng/ml) or medium in the absence or presence of rIL-13 (50 ng/ml). The next day the cells were harvested and kept in lysis buffer and store at −80° C for mRNA analysis. For co-cultures experiments stimulated skin DCs were extensively washed 3 times to remove LPS and rIL-13 and co-cultured with 1×10⁵ CD4⁺ T cells isolated from the blood healthy donors using the CD4⁺ T cell isolation kit from Miltenyi. Cells were cultured for 5 days in complete medium in the presence of T cell activation/expansion kit beads (Miltenyi).

Tape stripping of mouse skin.

The back skin of anesthetized mice was shaved and subjected to tape striping 6 times with a film dressing (TegadermTM, 3M). Mouse skin samples were collected at different time points and kept in RNAlater and stored at −80° C.

RNA extraction and RT-PCR.

RNA was extracted from human and mouse skin using Total RNA Isolation Kit (Ambion). cDNA was prepared using iscript cDNA synthesis kit (Biorad). Quantitative real-time PCR was done with the Taqman gene expression assay, universal PCR master mix and ABI prism 7300 sequence detection system (Applied Biosystems). Fold induction was calculated using delta-delta ct; with normalization to the internal control β−2 microglobulin. An arbitrary unit of 1 was assigned to the mean value of unmanipulated skin samples.

Preparation of skin cell suspensions from mouse skin.

Ears from vehicle or DNFB treated mice were split in half and were finely chopped using scissors and digested for 90 minutes in the media containing Liberase TL (0.2mg/mlRoche) and DNAse II (Sigma), with continuous shaking at 37° C. Digested skin homogenates were passed to 70μm cell strainer, washed and resuspended in PBS and used for flow cytometry.

Isolation and functional analysis of mouse skin dLN DCs.

For in vivo DC priming experiment, ovalbumin (1 mg in 100μL saline) was epicutaneously applied to shaved, tape stripped skin of mice. 24 hours after sensitization, DCs were isolated from axillary LNs using CD11c beads from Miltenyi after enzymatic digestion with 1mg/ml collagenase IV. Naïve CD4⁺ T cells were purified from spleen of DO11.10 mice using the naïve CD4⁺T cells isolation kit II from Milenyi after mechanical dissociation. For proliferation studies, 2.5×10⁴ CD11c⁺ cells were co-cultured 1:1 with celltrace violet stained naïve DO11.10 CD4⁺ T cells without addition of exogenous OVA protein for 3 days, then analyzed by flow cytometry. For polarization studies, 1×10⁵ CD11c⁺ cells were co-cultured 1:1 with naïve DO11.10 CD4⁺ T cells without addition of exogenous OVA protein for 5 days. Cell free supernatant was used to detect mouse IL-5, IL-13, and IFNγ by ELISA Ready SET Go kits (eBioscience).

IL-33 release by epidermis and dermis following tape stripping.

Ears were excised from unmanipulated mice or immediately after tape stripping. Ears were split into dorsal and ventral halves, and skin was floated on 4 mg/ml dispase solution (Roche) for 30 min at 37 °C. Epidermis and dermis were separated and cultured for 3h in complete RPMI. IL-33 in the supernatant was measured using Quantikine ELISA kit (R&D).

BMMCs reconstitution of MC deficient mice.

Bone marrow cells from both femurs and tibias of BALB/c WT or BALB/c Il1rl1^−/− were cultured for 4 weeks in complete RPMI supplemented with IL-3 (10ng/ml) and stem cell factor (20ng/ml) (Peprotech). 1.5×10⁶ BMMCs were injected intradermally into 6 to 8 week-old female Kit^W-sh/W-sh mice. To allow BMMC engraftment, mice were given a 4-week rest before tape stripping treatment.

Flow cytometry.

Cells were preincubated with FcγR-specific blocking mAb (2.4G2) and washed before staining with the following monoclonal antibodies (mAbs): PE-anti-CD124 (mIL4R-M1), FITC-anti-MHCII (M5/114.15.2), APC-anti-CD80 (16–10A1), PerCP-eFluor-710-anti-CD40 (1C10), PE-Cy7-anti-CD86 (GL1), Biotin-anti CD11c (N418), AlexaFluor700-anti-MHCII (M5/114.15.2) and BV605-anti-CD11b (M1/70). Ebioscience Fixable Viability Dye eFluor 506 was used to exclude dead cells. For cytokine staining, cells were stimulated with Ionomycin (0.5ug/ml; Sigma), Phorbol 12,13- dibutyrate (1ug/ml; Sigma), Brefeldin A (eBioscience) and Monensin (eBioscience) in complete RPMI for 3 hours before surface staining. Subsequently, cells were fixed and permeabilized (BD Biosciences Cytofix/Cytoperm) and stained in permeabilization solution with antibodies against IFNγ,IL-13 and IL-17A from eBioscience. Cells were analyzed on LSRFortessa (BD Biosciences), and the data was analyzed with FlowJo software.

DNFB induced Contact Hypersensitivity.

Mice were sensitized with 50 μl 0.5% DNFB (1-fluoro-2,4-dinitrobenzene; Millipore Sigma) in acetone:olive oil (4:1) on shaved abdomens then challenged 5 days later with 25 μl 0.5% DNFB in acetone:olive oil (4:1) on the right ear and 25 μl acetone:olive oil (4:1) (vehicle) on the left ear. Ear thickness was measured with a micrometer before challenge and after 24 hours.

Statistical analysis.

Statistical significance was determined by the two-tailed Student’s t test. A p value <0.05 was considered statistically significant.

Results.

Keratinocyte-derived IL-33 drives IL-13 production by MCs following skin mechanical injury.

Tape stripping increased Il13 mRNA levels in mouse skin in as early as 1 h, with a return to baseline by 12h (Fig. 1A). To determine the source of IL-13 released following skin mechanical injury, we examined Rag2^−/− mice which lack mature T and B cells, Rag2^−/−/gc^−/− mice which lack ILCs in addition to mature T and B cells, and Kit^−Wsh/Wsh mice which lack MCs 3 h after tape stripping. Upregulation of Il13 in the skin was preserved in Rag2^−/− mice (Fig. 1B) and Rag2^−/−/gc^−/− mice (Fig. 1C) but was abolished in Kit^−Wsh/Wsh mice (Fig. 1D), suggesting that MCs are the major source for IL-13 in mechanically injured skin. Kit^−Wsh/Wsh mice have abnormalities in cell linages other than MCs²⁵. To ascertain the role of MCs in IL-13 induction after tape stripping, we reconstituted the skin of Kit^−Wsh/Wsh mice with WT bone marrow-derived MCs (BMMCs). Four weeks after intradermal (i.d.) administration of WT BMMCs, the reconstituted skin was subjected to tape stripping. Flow cytometry revealed absence of MCs in non-reconstituted skin of Kit^−Wsh/Wsh mice. In contrast, MCs were present in BMMC-reconstituted skin of Kit^−Wsh/Wsh mice, but their percentages among skin cells was about half than in WT controls (Fig. 1E). Reconstitution of the skin of Kit^−Wsh/Wsh mice with WT BMMCs partially restored the increase in Il13 mRNA 3 h after tape stripping (Fig. 1F). Together these results indicate that MCs are the major source of Il13 induced in the skin by mechanical injury.

Figure 1. Keratinocyte-derived IL-33 drives IL-13 production by cutaneous MCs following mechanical skin injury.

A. Time course analysis of cutaneous Il13 mRNA levels after tape stripping (T/S). B-D. Relative Il13 mRNA expression in unmanipulated and T/S skin of Rag2^−/− (B), Rag2gc^−/− (C), and Kit^W-sh/W-sh (D) mice and WT controls. E-F. Representative flow cytometry plots (left) and quantitation (right) of CD45⁺CD3⁻G1⁻CD117⁺ MCs (E) and relative cutaneous Il13 mRNA expression in skin WT and Kit^W-sh/W-sh and Kit^W-sh/W-sh mice reconstituted WT BMMCs. (F). G. IL-33 secretion by epidermal and dermal sheets from unmanipulated and T/S ears of WT mice. H. Relative Il13 mRNA expression in unmanipulated and T/S skin of Il1rl1^−/− and WT controls at different time points. I. Relative cutaneous Il13 mRNA expression in non-reconstituted skin of Kit^W-sh/W-sh mice and skin of Kit^W-sh/W-sh mice reconstituted with WT and Il1rl1^−/− BMMCs. Results in A-I are representative of 2 independent experiments with 4–5 mice/group. Columns and bars represent mean±SEM. * p<0.05, ** p<0.005.

IL-33 promotes the production of IL-13 by MCs²⁶, and its level increases in the skin following mechanical injury^{27, 28}. Epidermal sheets, but not dermal sheets, prepared immediately after tape stripping the skin of WT mice demonstrated increased IL-33 secretion (Fig. 1G). We used IL-33 receptor deficient Il1rl1^−/− mice to examine the role of IL-33 in Il13 induction by mechanical skin injury. Induction of Il13 mRNA following tape stripping was abolished in Il1rl1^−/− mice at all the time points analyzed (Fig. 1H). To investigate whether IL-33 targets skin MCs directly to promote their production of IL-13 following mechanical injury, we reconstituted the skin of Kit^−Wsh/Wsh mice with BMMCs derived from Il1rl1^−/− mice or WT controls. Tape stripping failed to increase Il13 expression in skin of Kit^Wsh/Wsh mice reconstituted with Il1rl1^−/− BMMCs (Fig. 1I), but as a previously shown in Fig. 1F, it increased Il13 expression in skin of Kit^−Wsh/Wsh mice reconstituted with WT BMMCs. These results demonstrate that keratinocyte-derived IL-33 acts on skin MCs to promote IL-13 production following mechanical injury.

MC-derived IL-13 inhibits the capacity of skin-derived DCs to promote Th1 polarization of naïve CD4⁺ T cells to cutaneously introduced antigens.

We previously reported that mechanical injury skews skin-derived DCs to promote Th2 polarization and suppress Th1 polarization⁸. To determine the role of IL-13 in this process we made use of Il13^−/− mice. Il13^−/− mice and WT controls were subjected to EC sensitization with OVA and 24 h later skin dLNs were examined by flow cytometry for CD11c⁺MHCII^high, which represent recent skin emigrants²⁹. The number of skin DC emigrants in dLNs as well as expression of the activation markers CD40, CD80, CD86 and PDL2 by these cells were comparable in Il13^−/− mice and WT controls (Fig. 2A, B and Fig. E1A in the Online Repository). Moreover, co-injection of rIL-13 with OVA did not alter the accumulation of CD11c⁺MHCII^high DCs in the dLNs nor the expression of activation markers by these cells (Fig. E1B and C in the Online Repository). These results indicate that IL-13 has no effect on the migration or activation of skin-derived DCs.

Figure 2. MC-derived IL-13 inhibits the capacity of antigen capturing skin-derived DCs to promote Th1 polarization of naïve CD4⁺ T cells.

A-B. Representative flow cytometry plots (left) and quantitation (right) of CD11c⁺MHCII^high skin-derived DCs (A), and mean fluorescence intensity (MFI) of CD40, CD80 and CD86 expression by skin-derived DCs (B) in dLNs of unmanipulated and OVA sensitized skin of Il13^−/− mice and WT controls. C-F. Representative flow cytometry histogram of proliferation (left) and proliferation index (right) (C), IL-5 and IL-13 secretion (D) IFN-γ secretion (E) and representative (left) and quantitative (right) analysis of intracellular IFN-γ expression (F) by naïve CTV loaded CD4⁺ T cells from DO11.10 mice co-cultured with DCs from dLNs of OVA sensitized skin of Il13^−/− mice and WT controls. G. IFNγ− production by naïve CD4⁺ T cells from DO11.10 mice co-cultured with DCs from dLNs of unmanipulated Il4ra^−/− mice and WT controls, in the absence or presence of rIL-13. H. IL-13 and IFNγ− secretion by naïve CD4⁺ T cells from DO11.10 mice co-cultured with DCs from dLNs of OVA sensitized skin of Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− and Il4^flox/+Il13^flox/− controls. Results in A-H are representative of 2 independent experiments with 4–5 mice/group. Columns and bars represent mean±SEM. * p<0.05, ** p<0.005.

We next investigated the role of IL-13 in modulating the ability of skin-derived DCs to cause the proliferation and polarization of antigen-specific naïve CD4⁺ T cells. Il13^−/− mice and WT controls were EC sensitized with OVA. Twenty four hours later, DCs were purified from the skin dLNs and co-cultured with splenic naïve CD4⁺ T cells from DO11.10 transgenic mice which carry a TCR specific for OVA. No exogenous OVA was added. DCs from skin dLNs of Il13^−/− mice and WT controls induced comparable proliferation and secretion of IL-5, IL-13 and IL-17A by naïve CD4⁺ T cells (Fig. 2 C,D and Fig. E1D in the Online Repository). However, DCs from the skin dLNs of Il13^−/− mice caused increased production of IFN-γ by naïve CD4⁺ T cells compared with WT controls as demonstrated by IFN-γ secretion into the supernatants (Fig. 2E), as well as by intracellular staining of the T cells for IFN-γ (Fig. 2F). In contrast, DCs from dLNs of skin co-injected with rIL-13 and OVA caused significantly less IFN-γ production by naïve CD4⁺ T cells compared with DCs from dLNs of skin i.d. injected with OVA alone (Fig. E1E in the Online Repository). DCs from skin dLNs of mice EC sensitized with saline failed to induce proliferation or cytokine secretion by naïve CD4⁺ T cells in the absence of addition of OVA to the cultures (data not shown). These data suggest that IL-13 inhibits the capacity of antigen-bearing skin-derived DCs to promote Th1 polarization.

To investigate whether IL-13 affects antigen presentation by DCs obtained from skin DLNs at later time points after cutaneous antigen exposure we isolated CD11c⁺ cells from DLNs 72 h after EC sensitization with OVA and analyzes their ability to promote polarize naïve CD4⁺ T cells. Neither T cell proliferation nor secretion of IL-13, IL-17A and IFNγ were induced (data not shown). These results suggest that CD11c⁺ cells that migrate to DLNs late after cutaneous sensitization do not support the activation of antigen specific naïve CD4⁺ T cells.

We next investigated whether IL-13 directly modulates the capacity of DCs from skin dLNs to promote Th1 polarization. To this purpose we used Il4ra^−/− mice which lack the IL-4Rα chain common to the IL-4 and IL-13 receptors. DCs were purified from the skin dLNs of unmanipulated WT and Il4ra^−/− mice and co-cultured with DO11–10 T cells and OVA with or without addition of rIL-13. In the absence of IL-13, DCs from WT and Il4ra^−/− mice caused comparable secretion of IFN-γ by naïve CD4⁺ T cells (Fig. 2G). Addition of IL-13 significantly reduced IFN-γ secretion by naïve CD4⁺ T cells co-cultured with DCs from WT mice, but had no effect on IFN-γ secretion by naïve T cells cocultured with IL-4Rα deficient DCs (Fig. 2G). To demonstrate that IL-13 released by MCs inhibits the capacity of skin-derived DCs to induce Th1 polarization, we examined Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− mice which lack IL-13 selectively in MCs. DCs from the dLNs of OVA sensitized skin from Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− mice caused significantly more IFN-γ secretion, but comparable IL-13 and IL-17A secretion, by naïve DO11.10 T cells compared to DCs from dLNs of OVA sensitized skin from Il4^flox/+Il13^flox/− controls (Fig. 2H and Fig. E1F in the Online Repository). These results suggest that IL-13 released by MCs in mechanically injured skin acts directly on skin DCs to inhibit their capacity to drive Th1 polarization.

MC-derived IL-13 controls IL-12 expression in skin-derived DCs.

The cytokine IL-12 is produced by DCs and plays a critical role in DC-mediated Th1 polarization³⁰. IL-12 (p70) is a heterodimer of the p35 chain encoded by Il12a and of the p40 chain encoded by Il12b. To investigate whether IL-13 regulates IL-12 production by DCs, we examined the effect of IL-13 on IL-12 secretion by LPS-stimulated DCs from the skin dLNs of WT mice. LPS stimulation increased IL-12p70 secretion by skin dLN DCs in a dose-dependent manner (Fig. 3A). Notably, rIL-13 decreased the secretion of IL-12p70 by skin dLN DCs (Fig. 3A). rIL-13 had no detectable effect on LPS induction of Il12a (IL-12p35) mRNA; however, it significantly inhibited LPS induction of Il12b (IL-12p40) (Fig. 3B). These in vitro results suggest that IL-13 inhibits LPS driven IL-12p70 production by DCs by downregulating Il12b (IL-12p40) expression.

Figure 3. MC-derived IL-13 inhibits IL-12 expression in skin-derived DCs.

A-B. IL-12p70 secretion (A) and relative Il12a and Il12b mRNA expression (B) in DCs stimulated with LPS in the absence or presence of added rIL-13. C-D. Relative Il12a and Il12b mRNA levels in sorted CD11c⁺ DCs from dLNs of OVA sensitized skin of Il13^−/− mice and WT controls (C) and Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− and Il4^flox/+Il13^flox/− controls (D). Results in A-D are representative of 2 independent experiments with 4–5 mice/group. Columns and bars represent mean±SEM. * p<0.05, ** p<0.005.

To investigate whether IL-13 regulates IL-12 expression in skin-derived DCs in vivo, we measured Il12a and Il12b mRNA levels in DCs sorted from skin dLNs of WT and Il13^−/− mice 24 h after EC sensitization with OVA. Il12b (IL-12p40) but not Il12a (IL-12p35) expression was significantly higher in DCs purified from the dLNs of OVA sensitized skin of Il13^−/− mice compared to WT controls (Fig. 3C). No difference in the expression of Il23a (IL-23p19) or Il6 was observed between DCs from the two strains, however Tnf expression was significantly higher in DCs of Il13^−/− mice compared to WT controls (Fig. E2A in the Online Repository). DCs purified from the dLNs of OVA sensitized skin from Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− mice exhibited significantly higher expression of Il12b and Tnf, but not Il12a, Il23a or Il6, compared to DCs from Il4^flox/+Il13^flox/− littermates (Fig. 3D and Fig. E2B in the Online Repository). These results indicate that IL-13 produced by skin MCs in response to mechanical injury suppresses IL-12 production in skin-derived DCs.

MC-derived IL-13 controls Th1 differentiation during CHS induced by DNFB.

We next investigated whether MC-derived IL-13 regulates Th1 differentiation in a Th1 dominated cutaneous immune response. Contact hypersensitivity (CHS) to DNFB is mediated predominantly by IFN-γ produced by CD4⁺ Th1 and CD8⁺ T cells^31–35. DNFB-painting of ears from non-sensitized WT mice showed upregulation of Il33 and Il13 compared with vehicle treated ears (Fig. E3A and B in the Online Repository). Moreover, Il13 upregulation induced by DNFB painting was abolished in Il1rl1^−/− and significantly reduced in Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− mice (Fig. E3C and E3D in the Online Repository). These results indicate that MCs are an important source of IL-33-dependent IL-13 production in DNFB treated skin. To investigate whether IL-13 regulates CHS to DNFB, Il13^−/− mice and WT controls were sensitized by painting DNFB on the abdomen. Five days later, they were challenged by application of hapten or vehicle control to the ears. Ears were examined 24 hours after challenge for thickness, and intracellular cytokine expression by CD4⁺ T cells. Challenge of WT mice with DNFB, but not vehicle, caused a robust increase in ear thickness 24 hours later, as well as an increase in CD3⁺CD4⁺IFN-γ⁺ Th1 cells as determined by intracellular flow cytometry analysis (Fig. 4A and 4B right panel). The increase in ear thickness following DNFB challenge was significantly greater in Il13^−/− mice compared with WT controls (Fig. 4A). DNFB-challenged ears also showed significantly higher percentage of CD3⁺CD4⁺IFN-γ⁺ cells in Il13^−/− mice compared to WT controls (Fig. 4B). The percentages of CD3⁺CD4⁺IFN-γ⁺ Th1 cells in vehicle challenged ears were comparable among the two strains (Fig. 4B right panel).

Figure 4. MC-derived IL-13 inhibits ear swelling and Th1 cell differentiation in the CHS response to DNFB.

A. Change in ear thickness on day 1 after hapten (+) or vehicle (−) challenge of DNFB sensitized Il13^−/− mice and WT controls. B. Representative flow cytometry analysis (left) and quantitation (right) of the percentage of CD45⁺CD3⁺CD4⁺IFN-γ⁺ Th1 cells among skin cells on day 1 after hapten challenge in ear skin of DNFB sensitized Il13^−/− mice and WT controls. C, D. Change in ear thickness (C) representative flow cytometry analysis (left) and quantitation (right) of CD45⁺CD3⁺CD4⁺IFN-γ⁺ Th1 cells among skin cells in ear skin (D) on Day 3 after hapten or vehicle challenge of DNFB sensitized WT mice treated locally with rIL-13 or saline during sensitization. E, F. Change in ear thickness (E) and representative flow cytometry analysis (left) and quantitation (right) of CD45⁺CD3⁺CD4⁺IFN-γ⁺ Th1 cells among skin cells in ear skin (F) on day 1 after hapten or vehicle challenge of DNFB sensitized Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− and Il4^flox/+Il13^flox/− controls. Results in A-F are representative of 2 independent experiments with 4–5 mice/group. Columns and bars represent mean±SEM. * p<0.05, ** p<0.005.

We next investigated whether intradermal rIL-13 injection during the sensitization phase decreases ear swelling and Th1 polarization following hapten challenge. WT mice were i.d. injected in the abdominal skin with rIL-13 or saline 30 min before DNFB sensitization. There was a significant decrease in swelling and in the accumulation of CD3⁺CD4⁺IFN-γ⁺ cells, but not CD3⁺CD4⁺IL-17A⁺ cells, in DNFB challenged ears from mice i.d. injected prior to sensitization with rIL13 compared to saline injected controls (Fig. 4 C,D and data not shown).

To investigate whether MC-derived IL-13 is important for the inhibition of Th1-dominated CHS, we examined the CHS response of Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− mice to DNFB. Ear swelling in response to DNFB challenge as well as the accumulation of CD3⁺CD4⁺INF-γ⁺ Th1 cells were significantly greater in Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− mice compared to Il4^flox/+Il13^flox/− controls (Fig. 4 E, F). The accumulation of CD3⁺CD4⁺IL-13⁺ cells in DNFB challenged ears was comparable between both strains (Fig. 4 F and data not shown). These data suggest that IL-13 derived from skin MCs inhibits the Th1 response to hapten, and thereby decreases the magnitude of ear swelling response in CHS.

IL-13 suppresses the ability of human skin DCs to promote Th1 differentiation.

We investigated whether suppression of Th1 polarization by IL-13 produced in the skin is applicable to humans. We examined whether scratching induces IL13 expression in human skin and evaluated the effect of IL-13 on the ability of human CD1c⁺ skin DCs, which include Langerhans cells and conventional DCs⁴, to promote Th1 cytokine production in vitro. IL13 mRNA level in human skin was significantly upregulated 6 hours after scratching (Fig. 5A). Human dermal CD1c⁺ DCs were found to express on their surface both chains of the IL-13 receptor (IL-4Rα and IL-13Rα) as indicated by flow cytometry using anti-human IL-4Rα and IL-13Rα antibodies (Fig. 5B). Pretreatment of CD1c⁺ DCs isolated from normal human skin with IL-13 significantly suppressed LPS induction of IL12B (IL12p40), but not IL12A (IL12p35) (Fig. 5C). Importantly, CD4⁺ T cells stimulated with anti-CD3+anti-CD28 secreted significantly less IFN-γ upon co-culture with human skin DCs pretreated with LPS+IL-13 compared to co-culture with DCs pretreated with LPS alone (Fig. 5D, left panel). This effect was specific to IFN-γ, because IL-13 secretion was comparable in the two conditions (Fig 5D, right panel) These results strongly suggest that cutaneous IL-13 induced by scratching impairs the capacity of human skin DCs to promote IFN-γ secretion by CD4⁺ T cells.

Figure 5. Scratching upregulates IL13 expression in human skin and IL-13 suppresses the ability of LPS-stimulated human skin DCs to promote IFN-γ secretion by human CD4⁺ T cells.

A. Relative IL13 mRNA expression in unmanipulated and scratched skin (6 hours post scratching) of healthy subjects. B. Flow cytometry analysis of IL-4Rα and IL-13Rα expression by human skin CD1c⁺ DCs from surgical discard specimens. C. Relative IL12A and IL12B mRNA expression in human skin CD1c⁺ DCs induced by LPS in the absence or presence of rIL-13. D. IFN-γ and IL-13 secretion by CD4⁺ T cells from healthy donors co-cultured with skin CD1c⁺ DCs from heterologous healthy donors that were untreated or treated with LPS in the absence or presence of added rIL-13. Results in A-D are representative of one experiment with 3–5 donors/group. Columns and bars represent mean±SEM. * p<0.05.

Discussion.

We demonstrate that mechanical skin injury drives IL-33-dependent production of IL-13 by cutaneous MCs, which suppresses the capacity of skin DCs to elicit a Th1 response to cutaneously encountered antigens. This may contribute to the preferential induction of a Th2 immune response to these antigens in patients with AD and to the susceptibility of these patients to cutaneous viral infections.

Mechanical skin injury inflicted by tape stripping upregulated Il13 expression in cutaneous MCs. This upregulation was dependent on the alarmin IL-33 that was released by epidermal cells in response to tape stripping, as evidenced by the failure of tape stripping to upregulate Il13 expression in the skin of MC deficient Kit^Wsh/Wsh mice. This failure was reversed by reconstituting the skin of these mice with WT BMMCS but not BMMCS that lack IL-33R expression. We found no role for ILC2s in Il13 upregulation in mechanically injured skin as there was normal upregulation of Il13 in the tape stripped skin of Rag2^−/−gc^−/− mice which lack ILCs. This is in contrast to the major role ILC2s play in the induction of Il13 expression in the lungs following intranasal rIL-33 treatment and in the IL-33-dependent Il13 upregulation in the lungs following inhalation of papain, HDM or Alternaria^{36, 37}. Thus, IL-33 promotes innate IL-13 production by different tissue resident innate cells in a tissue-specific and context dependent manner.

MCs play an important role in the initiation of the immune response in the skin as MC-derived molecules (e.g. TNFα) promote the migration of skin DCs into draining lymph nodes^{38, 39}. Our results show that MC-derived IL-13 is not required for DC migration from skin to dLNs or the induction of Th2 polarization by these cells. Notably, MC-derived IL-13 inhibits the capacity of skin-derived DCs to promote Th1 polarization of naïve CD4⁺ T cells. This was evidenced in loss of function experiments in which skin-derived DCs from Il13^−/− mice and mice with conditional deletion of Il13 in MCs promoted Th1 polarization, as well as in gain of function experiments in which skin-derived DCs from mice intradermally injected with rIL-13 were impaired in driving Th1 cell differentiation. Thus, MC-derived IL-13 inhibits the ability of skin DCs to drive a Th1 response to cutaneously encountered antigen.

IL-13 likely acted directly on skin DCs in vivo, to inhibit their ability to drive Th1 polarization. Addition of IL-13 in vitro to skin-derived DCs from WT mice, but not from Il4ra^−/− mice, suppressed their capacity to drive Th1 polarization. Importantly, IL-13 suppressed the production of the Th1 polarizing cytokine IL-12 by skin DCs. This was associated with inhibition of expression of Il12b encoding for the IL-12p40 chain. This result is consistent with the observation that rIL-13 inhibits the IL-12p40 and IL-12p70 production by human PMBCs stimulated with LPS⁴⁰. . Importantly, in a gain of function experiment we show that expression of Il12b (IL-12p40), but not Il12a (IL-12p35) was significantly higher in DCs purified from the dLNs of OVA sensitized skin of Il13^−/− and Mcpt5-Cre^Tg/0Il4^flox/+Il13^flox/− mice. PD-L2 inhibits the production of IL-12p70 in mouse bone marrow derived DCs induced by house dust mite antigen⁴¹. Our results indicate that PD-L2 expression by skin-derived DCs was unaffected by IL-13 deficiency. These results suggest that the inhibition of IL-12 expression in skin-derived DCs by MC-derived IL-13 is independent of PD-L2.

We used a Th1-dominated model of skin inflammation, namely CHS to the hapten DNFB, to verify the role of MC-derived IL-13 in suppressing the Th1 response to cutaneously encountered antigens. In this model, Il33 and Il13 mRNA are upregulated in the sensitization phase. Importantly, IL-33 receptor was required for DNFB induced IL-13 and mast cells were an important source of this cytokine, suggesting that cutaneous sensitization with hapten and antigen sensitization use a common mechanism for the induction of IL-13 expression by skin MCs. Consistent with previous results we showed that CHS response to DNFB as measured by ear swelling and the accumulation of Th1 cells in hapten challenged ear skin is exaggerated in Il13^−/− mice⁴². In contrast, this response was diminished by intradermal injection of rIL-13 during hapten sensitization. Importantly, both ear swelling and accumulation of Th1 cells in hapten-challenged ear skin were exaggerated in mice with selective deficiency of IL-13 in MCs, thereby establishing a role of MC-derived IL-13 in the Th1 response to cutaneous hapten exposure.

Our findings in mice regarding the role MC-derived IL-13 plays in inhibiting the induction of a Th1 response to cutaneously encountered antigens are likely applicable to humans. Scratching upregulated IL13 expression in human skin, rIL-13 inhibited IL12B expression by human skin CD1c⁺ DCs and impaired their capacity to promote IFN-γ production by CD4⁺ T cells. IFN-γ plays an important role in controlling viral infections, including those caused by vaccinia and herpes simplex^43–45. Eczema herpeticum has been associated with IFNG and IFNGR1 SNPs. Moreover, AD patients suffering this complication show defective production and response to IFN-γ^41,46.

Taken together, our results strongly suggest that MC-derived IL-13 suppresses the Th1 response to cutaneously encountered antigens in AD. Local IL-13 blockade may be useful in correcting the Th2:Th1 imbalance in AD and in promoting a protective Th1 immune response to cutaneous viral infections that afflict AD patients.

CLINICAL IMPLICATIONS.

Local IL-13 blockade may be useful in correcting the Th2:Th1 imbalance in lesional AD skin and promoting a protective Th1 immune response to cutaneous viral infections that afflict AD patients.

ABBREVIATIONS

AD

Atopic dermatitis

DCs

Dendritic cells

TSLP

Thymic stromal lymphopoietin

MCs

Mast cells

ILC2s

Type 2 innate lymphoid cells

dLNs

Draining lymph nodes

OVA

Ovalbumin

CHS

Contact hypersensitivity

DNFB

Dinitrofluorobenzene

IFN-γ

Interferon-γ

BMMCs

bone marrow-derived mast cells