C57Bl/6 wild-type (WT) mice were obtained from Charles River Laboratories Sulzfeld, Germany. hK14mIL33tg were generated on a C57Bl/6 background (see below). KRT14-cre (CD1 background) mice were obtained from Jackson Laboratory (Bar Harbor, ME, USA). The mmIL-33-GFP-CRE mice (C57Bl/6 background) were kindly provided by Dr. S. Wirtz (27).

The transgene is expressed in a K14 expression vector (28). It consists of a 2-kb human keratin 14 promoter (Gen-Bank accession no. DQ343282.1), a 0.66-kb rabbit-beta globin intron followed by a 0.8-kb full-coding sequence of mouse IL-33 cDNA (Gen-Bank accession no. NM_133775.2) and a 0.35 kb human keratin 14 poly(A) signal DNA fragment. For generation of the vector encoding a secreted version of IL-33 (mmIL-33) controlled by skin-specific regulatory elements, we inserted mouse mature IL-33 cDNA into the K14 vector (27).

The human keratinocyte cell line HaCaT (#300493, CLS Cell Lines Service, Germany) and the murine keratinocyte cell line PDV were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen, Carlsbad, CA, USA), 100 U/ml penicillin and 100 mg/ml streptomycin. Cells were maintained at 37°C and 5% CO₂. The K14 expression vector was transiently transfected into HaCaT cells using Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Murine keratinocytes were transfected using FuGENE^® HD Transfection Reagent according to the manufacturer’s instruction (Promega, Madison, WI, USA).

After collecting the supernatant, all cells were pelleted, washed with cold phosphate-buffered saline (PBS), and lysed in RIPA buffer. Before denaturation with Laemmli buffer, protein concentration adjustment was conducted. Supernatants were resuspended in 6× Laemmli and boiled at 98°C for 10 min. Cell pellets and supernatants were separated on a 10% SDS-polyacrylamide gel. Membranes were blocked with 5% non-fat dry milk (Roth, Karlsruhe, Germany), probed with mouse anti-human IL-33 antibody (1:1,000; Nessy-1, Enzo Life Sciences, Lörrach, Germany), and peroxidase-labeled anti-mouse as secondary antibody (Dako, Denmark). Membranes were developed using the enhanced chemiluminescence method.

Linear transgene K14-IL-33 DNA fragment was separated from vector backbone by gel electrophoresis. The K14-IL-33 fragment was cut out, transferred in a dialysis bag (Spectra/Por MWCO 3500, Spectrumlabs, DG Breda, The Netherlands), filled with running buffer (1× TAE), and closed with clamps. After electrophoresis for 1 h at 80 V, 90% of the DNA was eluted. The DNA was purified by Elutip-D minicolumns from Schleicher & Schuell (Dassel, Germany).

The generation of transgenic mice by pronuclear injection of K14-IL-33 transgene was performed by the Transgenic Mouse Facility, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany. Of 16 pups born, 2 mice were positive for the transgene. The transgenic founders were identified by PCR for expression of the K14 promoter (IL-33-seq7fw 5′ CAGTTGATCCCAGGAAGAGC 3′ and IL-33-seq6rev GCAGGCTACACTTTCCCATC) and K14-IL-33 transgene (IL-33-seq7fw CAGTTGATCCCAGGAAGAGC and IL-33-seq2rev GTTGCAGCTCTCATCTTTCTCC) and by quantitative real-time PCR for the expression of mouse IL-33 using the primer pair: K14IL-33int/ex fw CTGCAAGTCAATCAGGCGAC and K14IL-33 int/ex rv TGCAGCCAGATGTCTGTGTC. A mouse line that highly overexpressed IL-33 was generated from these mice by breeding with littermates. To establish K14-CreER™/GFPmmIL33 mice, homozygous K14-CRE™ mice were crossed with heterozygous tamoxifen inducible mmIL-33-GFP-CRE mouse. The offspring were tested for mmIL-33 expression using flow cytometric analysis of blood for GFP expression (Figure 3B) thus using negative littermate animals as control mice for all experiments.

All protocols used in these studies were in compliance with federal guidelines and the Amgen Institutional Animal Care and Use Committee. All mice were maintained in a SPF facility. All animal experiments were approved by the animal welfare committee and approved by the “Regierung Unter-/Mittelfranken.”

The right ears of both WT and hK14mIL33tg mice at 8–10 weeks of age were topically treated with an IMQ-containing cream (Aldara, MEDA Pharma) for 7 consecutive days. Control ears were treated similarly with a vehicle cream. Ear swelling was measured daily before treatment. On day 7, after sacrificing the mice, ears were collected for H&E, immunohistochemistry (IHC) and analyzed for epidermal thickness.

Barrier disruption due to removement of corneocytes was induced by skin stripping as reported previously (29). On day 3 after treatment, back skin of WT and hK14mIL33tg (Tg) were collected for immunohistochemical staining for IL-33, H&E and analyzed for epidermal thickness (mm).

500 ng IL-23 (R&D Systems, Minneapolis, MN, USA) in a total volume of 20 µL was intradermally injected in the ears from WT and hK14mIL33tg (Tg) every other day. Sterile PBS was used as a vehicle control. Ear thickness was monitored before each injection. On day 15, ears were analyzed for epidermal thickness and collected for immunostaining.

Dermatitis was induced by locally and fractionated irradiation of the right ear similar to a procedure used for mouse tumor irradiation with slight modifications (30). For irradiation, mice were anesthetized by isoflurane inhalation and placed in a special manufactured plexiglas^® box. During the irradiation procedure, the mice were kept under isoflurane inhalation anesthesia to avoid movement of the mice. The mice were locally irradiated with four single fractions of 12 Gy (cumulative dose of 48 Gy) only at the right ear. The irradiation field of the 6 MV linear accelerator (PRIMART, Siemens, Munich, Germany) was minimized, and irradiation was performed tangentially to preserve the head of the mice. This procedure was performed at days 1, 3, 5, and 7. Starting on day 2 and every second day thereafter, animals were scored in a blinded manner for ear thickness and dermatitis. Dermatitis scoring of all animals was performed analog to the Cancer Therapy Evaluation Program scale (31, 32). Briefly, dermatitis scoring ranged from: 0 = normal, no changes; 1 = mild erythema; 2 = moderate to severe erythema, slight desquamation; 3 = desquamation of 25–50% of irradiated area; 4 = desquamation of >50% of irradiated area; to 5 = frank ulcer.

K14 fl IL-33 transfected murine keratinocytes (PDV) were incubated at 37°C, 5% CO₂ in humidified air in 24 well, flat-bottom plates. Ionizing irradiation was performed with an X-ray generator (120 kV; GE Inspection Technologies, Hürth, Germany). The dose rate was 8 Gy/min. Irradiated and control keratinocytes monolayer cultures were immediately returned to the incubator at 37°C in a humidified environment and cultured for 6–48 h. Supernatant and cells were analyzed by ELISA and western blot to assess the presence of the IL-33 cytokine.

Tamoxifen (Cayman Chemical, Ann Harbor, MI, USA) was dissolved in an ethanol/DMSO mixture (equal parts) solutions at 100 mg/ml. Tamoxifen solution was freshly prepared the day prior to each administration and placed on a rolling device to dissolve overnight at room temperature. Before treatment, excess fur was shaved from the backs of recipient mice, which were topically treated with 20 mg Tamoxifen (200 µl volume).

Mouse skin and ear samples were fixed in 4% formaldehyde and stained with hematoxylin and eosin. Epithelial hyperplasia was assessed using a Zeiss Axio Lab.A1 light microscope (Zeiss, Oberkochen, Germany) and quantified by measuring the mean distance from the stratum basale to the bottom of the stratum corneum in a blinded manner. Mast cells were stained by toluidine blue, numbers counted per full skin section and standardizing for the tissue area of different sections [using image analysis system (OsteoMeasure; OsteoMetrics, Decatur, GA, USA)].

For IHC, 8-µm paraffin sections were incubated with anti-mouse IL-33 antibody (Enzo Life Sciences, Lörrach) overnight at 4°C. Tissues were subsequently labeled with biotinylated goat anti-mouse IgG antibody (Vector Laboratories, Burlingame, CA, USA) at room temperature for 1 h, following streptavidin-HRP (Dako, Denmark) and DAB Chromogen (Dako, Denmark) incubation. Positive staining developed as a brown reaction.

Total RNA was isolated using peqGold TriFast from Peqlab (Erlangen, Germany). RNA was reverse transcribed using the MultiScribe™MuLV reverse transcriptase (Invitrogen, Thermo Fisher Scientific Inc., Waltham, MA, USA). Relative gene expression was measured with Sybr Green RT mix by quantitative real-time PCR using β-actin as endogenous control according to the manufacturer’s manual (Applied Biosystems 7500 fast-real-time-PCR System or Quant StudioTM 6 Flex Real-Time PCR systems). For evaluation of target gene expression, the ΔCt as well as the ΔΔCt method was used (all Applied Biosystems, Carlsbad, CA, USA).

The following primers were used: mIL-33 (fw 5′CTGCAAGTCAATCAGGCGAC 3′, rv 5′ TGCAGCCAGATGTCTGTGTC 3′) β-Actin (fw 5′ TGTCCACCTTCCAGCAGATGT 3′, rv 5′ AGCTCAGTAACAGTCCGCCTAGA 3′) mIL-1β (fw 5′ CAGGCAGGCAGTATCACTCA 3′, rv 5′ AGGTGCTCATGTCCTCATCC 3′) mIL-6 (fw 5′ TCCATCCAGTTGCCTTCTTG 3′, rv 5′ TTCCACGATTTCCCAGAGAAC 3′) mIL-19 (fw 5′ GCCAACTCTTTCCTCTGCGT 3′, rv 5′ GGTGGCTTCCTGACTGCAGT 3′) mIL-5 (fw 5′ AGCACAGTGGTGAAAGAGACCTT 3′, rv 5′ TCCAATGCATAGCTGGTGATTT 3′) mIL-10 (fw 5′ ACTGCACCCACTTCCCAGT 3′, rv 5′ TTGTCCAGCTGGTCCTTTGT 3′) mIL-4 (fw 5′ AGATGGATGTGCCAAACGTCCTCA 3′, rv 5′ AATATGCGAAGCACCTTGGAAGCC 3′) mIL-13 (fw 5′ TGAGGAGCTGAGCAACATCACACA 3′, rv 5′ TGCGGTTACAGAGGCCATGCAATA 3′) mCCR6 (fw 5′ CTGCAGTTCGAAGTCATC 3′, rv 5′ GTCATCACCACCATAATGTTG 3′) mTNFα (fw 5′ GCTGAGCTCAAACCCTGGTA 3′, rv 5′ CGGACTCCGCAAAGTCTAAG 3′) BD4 (fw 5′ GGCTTCAGTCATGAGGATCCAT 3′, rv 5′ TTTGGGTAAAGGCTGCAAGTG 3′) BD14 (fw 5′ GTGGCCGGTGTGCTGTACT 3′, rv 5′ CGCTATTAGAACATCGACCTATTTGT 3′) mLCN2 (fw 5′ TGGAAGAACCAAGGAGCTGT 3′, rv 5′ GGTGGGGACAGAGAAGATGA 3′).

ELISA was performed with ELISA Kits (eBioscience, San Diego, CA, USA), according to the manufacturer’s instructions. Absorbance was measured at λ450/540 nm with SpectraMax 190 ELISA-Reader and analyzed with Softmax Pro Version 3.0 (Molecular Devices) software. IL-33 ELISA detects mature recombinant IL-33 (rIL-33) and processed forms of full length IL-33 (16).

Differences between groups were evaluated by unpaired two-tailed t-test. We applied the Bonferroni correction for multiple testing. For in vivo radiation experiments, differences were evaluated using two 2 × 2 ANOVAs, each incorporating time as a within-subjects characteristic (difference between first measurement at day 2 and the last measurement at day 25) and group assignment as a between-subjects characteristic (control mice vs. K14 IL33 mice). p < 0.05 was considered to be significant.

To test, if mature IL-33 exclusively linked to keratinocytes induce spontaneous inflammation we used an inducible K14-IL-33-CRE™/GFPmmIL33 mouse (iTG). This mouse contains a transgene DNA with loxP-flanked GFP and stop codon, which prevents transcription of the mature form of mouse IL-33 (mmIL-33) cDNA and was crossed with a tamoxifen inducible K14-CreER™ mouse (Figures 1A,B). For the Cre-mediated recombination and K14-specific overexpression of the mmIL-33, tamoxifen (TM) was administrated for 6 days topical on the shaved back skin (Figure 1C). Five days after the last TM administration, adult iTG mice developed thickening and scaling of the skin as well as weight loss (Figures 1D,E). In induced skin, IL-33 expression increased about 160-fold (Figure 1F). Histological analysis of skin biopsies by hematoxylin and eosin staining revealed an increase in epidermis and dermis thickness in iTG compared to WT mouse (Figure 1G). Abundant infiltrates of mast cells were detected by toluidine blue staining (Figure 1H). Furthermore, IL-33 overexpression in skin, lead to high production of the cytokines IL-6, IL-1β, IL-13, IL-10, IL-19, CCR6, and LCN2 (Figure 2).

In order to assess the role of IL-33 in skin inflammation, we developed transgenic mice that expressed the full-length mouse IL-33 under the control of a human K14 promoter via pronuclear injection of K14-IL-33 construct (hK14mIL33tg, Figure 3A). To test the functionality of this construct, HaCaT cells were transfected with the plasmid containing hK14mIL33tg or mock control. Lysates and supernatants were tested for IL-33 expression using Western blot analysis (Figure 3B) and IHC (Figure 3C). IL-33 (30 kDa) was strongly expressed in cell lysates after 48 h and 72 h (Figure 3B). IHC showed nuclear expression in transfected HaCaT cells compared to the mock control (Figure 3C). The transgenic mouse line expressing the hK14mIL33tg was generated as described above. Two founder mice were tested positive for the transgene and used as hK14mIL33tg mouse line, all of which had a similar IL-33 expression in the skin. The studies reported in this work were performed in hK14mIL33tg with at least eightfold higher mRNA levels in skin relative to the endogenous IL-33 mRNA level (Figure 3F). To generate littermates, female heterozygous hK14mIL33tg mice were crossed with male WT C57BL/6 mice. The hK14mIL33tg mice grew normally and did not develop phenotypic abnormalities (Figure 3D), treatment with topic tamoxifen did not result in skin changes (data not shown). Skin IHC staining for IL-33 revealed higher expression in transgenic compared to WT mice; however, no skin pathology was observed (Figures 3D,E).

The obvious lack of a spontaneous skin phenotype in hK14mIL33tg mice showed that a mere increase in intracellular IL-33 levels was not sufficient to induce local inflammation. This observation in turn indicated that additional signals and events such as an inflammatory trigger or tissue damage are necessary to allow secretion and/or action of full-length IL-33. We therefore aimed to elucidate whether a local inflammatory stimulus was sufficient to release and activate full-length IL-33 in the hK14mIL33tg mouse and performed different skin inflammation models including rIL-23 ear injection, topical IMQ, and barrier disruption using skin tape stripping. The morphology of IL-23-injected skin is similar to lesional skin of human psoriasis (33, 34); therefore, we used an IL-23 ear injection model to estimate the effect of IL-33 overexpression during skin inflammation. rIL-23 was intradermally injected in the ear of hK14mIL33tg and WT mice every other day. Ear thickness and histological analysis such as epidermal thickness were used as readout (Figures 4A–C). Measurement of ear and epidermal thickness did not reveal any difference between hK14mIL33tg and WT mice. IHC and HE staining of ear sections showed that transgenic mice express more nuclear IL-33 compared to WT mice (Figure 4C), but showed no difference in epidermal hyperplasia. To assess whether overexpression of IL-33 in mice keratinocytes worsens the development of IMQ-induced psoriasiform dermatitis, we applied IMQ cream on the right ear of hK14mIL33tg and WT mice for 7 consecutive days. Daily measurements of ear thickness but also histological analysis did also not show any difference in inflammation comparing hK14mIL33tg to WT mice (Figures 4D–F). Furthermore, no difference was observed using a barrier disruption method (Figures 4G–I). In comparison to the iTG mice, we also treated hK14mIL33tg mice with tamoxifen; similar to iWT mice (Figure 1D), no effect was observed (data not shown). Thus, despite different immunological provocations no phenotype of full length (fl) IL-33 overexpression could be observed, which showed that skin inflammation per se does not promote action of IL-33.

In the here reported mouse models, only secreted mmIL-33 led to a clinically relevant phenotype. In serum analysis, IL-33 was also only clearly expressed in the induced K14-mmIL-33-CRE-TM (Figure S1 in Supplementary Material). Fl IL-33 has been demonstrated to be released from damaged cells (15). As radiation induces a strong dermatitis due to apoptosis and necrosis, we next determined the effect of fractionated radiation (single fraction 12 Gy, 4 fractions within 7 days) in hK14mIL33tg and WT mice (Figures 5A,B). In comparison to WT mice, an increase in ear thickness in hK14mIL33tg was observed 25 days after irradiation (Figure 5B). Macroscopic examination of irradiated ears indicated that transgenic mice have more severe skin symptoms compared to WT mice (Figure 5A). To confirm that IL-33 is affected in irradiated keratinocytes in vitro, we used the murine keratinocyte cell line PDV and the HaCaT transfected with the K14-flIL-33 vector and irradiated the cells. Radiation with 20 Gy induced an intracellular decrease of IL-33 in both cell lines (Figure 6). In conclusion, IL-33 contributed to a skin radiation-induced phenotype.