Homozygous CD11c-IL-17A^ind/ind mice were obtained from the University Medical Center of the Johannes Gutenberg-University Mainz (19). These mice are characterized by low but sustained production of the transgenic cytokine IL-17A in CD11c⁺ DCs. CD11c⁺ DCs account for an average of 2%–3% of leukocytes in all murine tissue types, which roughly corresponds to the number of IL-17A-producing leukocytes in human psoriatic plaques. The overexpression of IL-17A in CD11c-IL-17A^ind/ind mice leads to the development—starting in homozygous mice at an age of approximately 8 weeks—of gradually progressing plaque psoriasis, which is very similar to human psoriasis vulgaris.

The animal experiment was approved by the regional council (Regierungspräsidium Darmstadt, approval number: DA17/1000) and conducted in accordance with the guidelines of Federation of European Laboratory Animal Science Associations (FELASA). Mice were housed in groups using open cages placed in a ventilated scantainer (Scanbur) maintaining standard conditions [20–22 °C; 55% relative humidity (RH); 12-h light–dark cycle]. Animals had ad libitum access to food and water and were kept in an enriched environment.

CD11c-IL-17A^ind/ind mice were exposed to radon in a chamber especially designed for animal and cell experiments. Using this chamber, radon concentration and temperature/relative humidity could be reproducibly adjusted at 22.7 ± 0.5 °C and 50%–70% RH (20). Mice [8.5 weeks old (±3 days)] were randomly distributed in groups to be exposed 1h either once to 539.0 ± 2.8 kBq/m³ (n=10; 6 females, 4 males) or on 10 consecutive workdays to 39.2 ± 2.0 kBq/m³ (n=22; 9 females, 13 males) radon activity concentration. Sham-exposed animals (single exposure: n=10; 7 females, 3 males; multiple exposures: n=21; 8 females, 13 males) underwent the same procedure under similar conditions except that the radon activity concentration was 0.12 ± 0.04 kBq/m³, which is the normal background concentration in buildings. At the time points of exposure, none of the animals already showed plaques (score 4, see section 2.3). A few animals showed a skin score of 1 or 2 (dry or scaly skin) before the start of radon treatment. These animals were equally distributed among the groups, resulting in a similar overall score at the start of radon exposure.

Considering that commonly used scores like PASI (Psoriasis Area and Severity Index) were not adequate for this purpose, we established a modified severity score from 1 to 6 (see Supplementary Table S1; Supplementary Figure S1). Scoring was conducted three times per week. A score of 4 was counted as plaque. To be considered as “plaque induction”, the plaque had to be observed at least 2× per week and persist for at least 2 weeks.

Mice were euthanized when they reached a score higher than 5.5 or at defined time points after radon exposure, by cervical dislocation under deep isoflurane inhalation anesthesia (5% isoflurane) following the FELASA and AVMA guidelines for the euthanasia of animals. Skin tissue was fixed in 4% PFA solution for histological analysis or stored in RNAlater for RNA sequencing. Spleen and lymph nodes (LNs) were collected for immunophenotyping.

Blood was drawn from the tail vein one or two times before and every week during and after radon treatment period, and after sacrificing the mice. The blood was incubated in clotting activation tubes (Vacuette^® Blood Collection Tubes, Greiner Bio-One GmbH) for 30min and centrifuged (1,500 ×g, 10min, RT), and serum was separated and frozen. IL-17A and IL-19 concentrations in serum were measured via enzyme-linked immunosorbent assay (ELISA) according to the manufacturer’s instructions (Uncoated ELISA-Kits, Invitrogen AG).

Tissue samples were processed and analyzed microscopically as previously described (21) (see Supplementary Materials). Antibody staining was performed for Ki67 (monoclonal, SP6, abcam), FOXP3 (monoclonal, FJK 16s, eBioscience), RORγT (monoclonal, AFKJS-9, Invitrogen AG), or Ly6g (monoclonal, EPR22909-135, abcam).

For staining, 2 × 10⁶ cells per sample of spleen and LN were used. The applied standard procedure is described in the Supplementary Materials. Staining was performed with the following antibodies:

DC panel: CD11c (PerCP, N418) and MHCII (APC, M5/114.15.2) from BioLegend, B220 (Pacific Blue, RA3-6B2) and CD8a (APC-Cy7, 53-6.7) from BD Pharmingen, and CD11b (PE, M1/70) from eBiosciences; T-cell panel: CD3e (PE-Cy7, 145-2C11), CD4 (Pacific Blue, RM4-5), and CD25 (PE, 7D4) from BD Pharmingen and CCR6 (PerCP-Cy5.5, 29-2L17). Gating strategies are shown in Supplementary Figure S3.

Whole RNA sequencing of skin samples was performed by Microsynth Seqlab GmbH. Tissue samples were removed from RNAlater and placed in tubes with 750 µL of RLTplus (Qiagen) with 10 µL/mL β-mercaptoethanol (Thermo Fisher) and 2-mm steel beads, followed by processing on an FP24 at speed 6 for 1min. Lysate (200 µL) was used for RNA isolation with the Qiagen RNeasy plus kit. The RNA was quantified with the RiboGreen (Thermo Fisher) assay, and qualitative analysis was carried out on an Agilent FragmentAnalyzer.

Raw read counts were imported into R prior to differential gene expression analysis using the R package DEseq2 (22). Only genes that had a total sum count of 10 times the number of samples were kept. Genes were considered differentially expressed if the unadjusted p-value was smaller than 0.05.

Gene set enrichment analysis (GSEA) was performed as described in the Supplementary Materials.

Results were plotted and statistically analyzed with GraphPad Prism 9.3.1 software, R, or the online tool venny2.1. Independent Kaplan–Meier plots were compared via the Log-rank (Mantel–Cox) test. Independent samples/groups were compared using unpaired t-test when the Shapiro–Wilk test indicated normal distribution or the Mann–Whitney U test when the Shapiro–Wilk test showed no normal distribution.

We previously reported that homozygous CD11c-IL-17A^ind/ind mice develop psoriatic skin lesions between the age of 8 and 18 weeks (19). To investigate if radon alleviates this phenotype, 8.5-week-old (±3 days) CD11c-IL-17A^ind/ind mice received total radon activity concentrations comparable to those in a radon gallery where patients are treated. Radon was administered either in a single exposure for 1h or—similar to patient treatment—in 10 consecutive 1-h exposures each over a period of 2 weeks by using a specially designed radon chamber (20). Before, during, and for several weeks after treatment, the severity of skin reactions in radon- and sham-exposed mice was scored (Supplementary Table S1). As depicted in Figure 1A, multiple radon exposures significantly reduced the mean severity score of skin reactions by 1–1.6 units as compared to sham-exposed control animals, starting from the second week of treatment and up to 3 weeks after the last session. In addition, the time point of full plaque induction (attaining a score of 4) was significantly delayed by 2–3 weeks (Figure 1B), postponing the median age of plaque development from 11 weeks in the sham-exposed group to 14 weeks in multiply radon-exposed mice (Figure 1C). Single radon exposures with a comparable total radon activity concentration showed a slightly delayed onset of plaque induction, which is not significant (Supplementary Figure S2). In conclusion, these data indicate that repeated exposures to radon result in a delayed development of psoriatic skin lesions in CD11c-IL-17A^ind/ind mice.

To check whether the significant improvement in skin reactions of CD11c-IL-17A^ind/ind mice was facilitated by a systemic change of key cytokines after multiple radon exposures, we measured the serum concentrations of two central cytokines of the pathogenic IL-23/IL-17 axis, namely, IL-17A and IL-19, at different time points (13, 23).

The serum concentration of IL-17A was measured before and up to 8 weeks after starting the multiple radon exposures. No radon-induced changes were observed compared to controls. Furthermore, no changes in IL-17A expression were detected over time (Figure 2A).

The serum concentration of IL-19 was analyzed 3 days and 2 weeks after the last treatment session as the most pronounced difference in the score of skin reactions was detectable during and up to 3 weeks after multiple radon exposures. Consistently with IL-17A, the serum concentration of IL-19 did not differ between radon-exposed and sham-treated mice (Figure 2B).

These data indicate that transgenic overexpression of IL-17A in CD11c-IL-17A^ind/ind mice does not change after radon exposures. This also accounts for its downstream amplifier IL-19.

The onset of psoriatic skin inflammation is promoted by different DC subtypes, with pDCs being the initiators of the inflammatory cascade and cDCs being the activators for Th1 and Th17 cells. The resulting imbalance of regulatory T (Treg) and effector T cells including the associated cytokine profile further promotes the inflammatory response (11, 16).

To investigate if multiple radon exposures can induce systemic changes in the inflammatory status of DCs and T cells, we conducted immunophenotyping in cervical LNs and spleen of radon- and sham-treated CD11c-IL-17A^ind/ind mice 3 days and 2 weeks after the last session. In LNs, we detected a decrease in the percentage of mature DCs (CD11c⁺MHCII⁺), pDCs (CD11c⁺MHCII⁺B220⁺CD11b⁻), and cDC1 (CD11c⁺MHCII⁺CD8a⁺CD11b⁻) 3 days after radon exposures as compared to sham-treated mice, but only with a statistical significance for pDCs (Figures 3A–C). No differences were found in cDC2 (CD11c⁺MHCII⁺CD8a⁻CD11b⁺) (Figure 3D), migratory DCs (CD11c^intMHCII^high), and LN-resident DCs (CD11c^highMHCII^int) at this early time point (Table 1; Supplementary Figures S3C, D). Additionally, the fractions of Th cells (CD3e⁺CD4⁺), cytotoxic T (Tc) cells (CD3e⁺CD4⁻), Treg cells (CD3e⁺CD4⁺CD25⁺), and Th17 cells (CD3e⁺CD4⁺CCR6⁺) showed no differences in LN of radon- and sham-treated animals at the 3-day time point (Table 1, Supplementary Figures S3I–L). Two weeks after the last radon exposure and in line with a transient effect, the identified effects in the immune cell status, namely, for DC, were no longer detectable in LN (Supplementary Figures S4A–D, I–L).

In contrast, in spleen, no differences were detected in radon- and sham-treated mice in the concentrations of the investigated DC and T-cell subtypes, neither for the 3-day (Table 1; Supplementary Figures S3E–H, M–P) nor for the 2-week time point (Supplementary Figures S4E–H, M–P). Taken together, the results show a radon-induced reduction of pDC in LNs but not in spleen of CD11c-IL-17A^ind/ind mice after multiple radon exposures.

As shown above, multiple radon exposures significantly delayed the development of severe skin lesions in CD11c-IL-17A^ind/ind mice. We hypothesized that this observation is related to the immune status in the skin. Therefore, we performed an immunohistological analysis with skin samples. Only a small percentage of mice showed psoriatic plaques at the time points of analysis. Thus, mainly non-lesional skin samples were analyzed to generate comparable datasets of radon- and sham-exposed mice.

To investigate if radon exposure leads to local immunosuppressive effects in the skin, we stained for FOXP3, a transcription factor specific for Treg cells that counteract psoriasiform inflammation (24). Additionally, we stained for the proliferation marker Ki67 to analyze if the hyperproliferation of keratinocytes, which is one hallmark of psoriasis (17, 25), is affected by radon treatment. Neither the fraction of FOXP3⁺ cells nor Ki67⁺ proliferating cells detected 3 days and 2 weeks after the last of multiple radon exposures were significantly different from controls (Figure 4).

We aimed at identifying neutrophils (via Ly6G staining), accumulating in psoriatic skin of CD11c-IL-17A^ind/ind mice (19), and Th17 cells (via RORγT staining), releasing IL-17A, both key events in the pathomechanism of human plaque psoriasis (13). As expected, in the lesional back skin of CD11c-IL-17A^ind/ind mice, both markers were detected (Supplementary Figure S5), but not in non-lesional skin. As only a few plaques were present at this early time point, statistical analysis was not possible. However, if a plaque could be analyzed, the number of neutrophils and Th17 cells in lesional skin was comparable in radon-exposed and sham-treated mice. Taken together, Th17 cells and neutrophils, two key immune cell types in psoriatic lesions, as well as keratinocyte proliferation in the skin of CD11c-IL-17A^ind/ind mice were not affected after radon exposures.

As flow cytometric and histological analysis did not reveal a clear immune cell signature supporting the significant clinical improvement in skin reactions of CD11c-IL-17A^ind/ind mice after multiple radon exposures, we next used an unbiased approach. We performed a comprehensive transcriptome profiling by RNA sequencing of non-lesional skin samples to quantify gene expression in radon-exposed CD11c-IL-17A^ind/ind mice and sham-treated animals.

An overview of the significant changes in the entirety of the analyzed genes is depicted in volcano plots for 3 days and 2 weeks after radon treatment. As shown in Figures 5A, B, radon exposure affected the transcription of numerous genes even in non-lesional skin. These changes were very different for the time points investigated, i.e., 3 days (Figure 5A) and 2 weeks (Figure 5B) after treatment, pointing to transient effects that evolved over time after the end of multiple radon exposures.

Additionally, we interrogated differences in the expression levels of distinct genes that are related to psoriasis and radiation response.

The expression of the transcription factor Gata3, specific for T cells counteracting the psoriatic inflammation by association with Th2 cells (26), was upregulated 3 days after radon exposures (Figure 6A), while this was not detectable anymore at 2 weeks after exposures (Figure 6B). The expression of the anti-inflammatory factors Il10 and Foxp3 was not modified.

Regarding the initiation phase of psoriasis, the expression of the transcription factor Tbx21 associated with Th1 cells was significantly downregulated at 3 days after exposure (27) (Figure 6A), while this was no longer detectable at the later time point (2 weeks, Figure 6B).

Not at 3 days, but at 2 weeks after radon exposure, a significant downregulation of the genes Tlr7, Ccr6, and Il23r was detected. These genes encode for relevant receptors involved in the initiation of the immune response promoting the progression of psoriasis (28–31) (Figure 6B).

Genes encoding autoantigens involved in psoriasis and related to the initiation phase were not significantly changed, i.e., Adamtsl5, encoding a psoriasis-related melanocyte-derived autoantigen (15), and Camp, encoding the autoantigen LL-37 (14). Also not affected by radon exposure were genes involved in the immune response, Itgax (T-cell marker, elevated in psoriasis), Ciita (marker for DC), and Rorc (marker for Th17 cells) (32–34).

While we have observed attenuated development of skin lesions in CD11c-IL-17A^ind/ind mice after radon exposures, the gene expression of soluble inflammatory or differentiation factors produced by immune cells or keratinocytes that are related to the chronic state of psoriasis was either not affected or significantly increased as in the case of Il20 at the early time point or of Il1f9 and Defb3 at the late time point (35, 36). One exception at the late time point was Tgfbr1, part of the disease promoting signaling in psoriasis, which was significantly downregulated (37).

Additionally, genes that encode keratins 10, 14, and 16 were significantly upregulated at the 2-week time point, indicating a complex relation to keratinocyte differentiation (38). Furthermore, the proliferation marker Mki67 as well as typical radiation-induced genes, i.e., Trp53, Chek2, and Mdm2 (39, 40), were not differentially expressed after radon exposures, except Cdkn1a (target gene of Trp53) that was significantly upregulated 2 weeks after exposure.

Since the identified differences in the expression levels of distinct psoriasis-related genes indicated transient immunosuppressive effects after multiple radon exposures in CD11c-IL-17A^ind/ind mice, we performed a pathway enrichment analysis as described in detail in the Supplementary Materials (Supplementary Figures S6, S7). This revealed that radon exposure mainly affects the organization and maturation of different skin compartments, ion transport and regulation, kinases and T cells, biosynthesis, as well as processes associated with ultraviolet (UV) response. In conclusion, the transcriptome profiling in the skin of CD11c-IL-17A^ind/ind mice revealed a transient upregulation of the expression of psoriasis-suppressive genes, whereas genes related to the initiation of psoriasis were downregulated after multiple radon exposures. Genes related to keratinocyte proliferation were not differentially expressed after radon treatment or showed a complex pattern for keratinocyte differentiation, and the cell cycle inhibitor Cdkn1a was significantly upregulated.

Radon exposure is a radiation-based and non-invasive treatment, similar to UV irradiation, which is one of the conventional therapy options for psoriatic skin lesions. As described in more detail in the Supplementary Materials, we compared the transcriptome profile of non-lesional skin from CD11c-IL-17A^ind/ind mice after radon exposure with three GEO-available gene expression analyses of human psoriatic skin lesions after UV treatment to assess the relevance of our results for patients. This comparison revealed 33 differentially expressed genes (DEGs) (P<0.05) that are identical after radon and UV treatment (Supplementary Figure S8). According to UniProt and GeneCards, the DEGs that are identical after radon and UV therapy are involved in the regulation of cell proliferation and differentiation, calcium concentration, T-cell activation, innate immunity, class I MHC antigen presentation, and TNF-alpha signaling. In addition, these DEGs modulate the MAPK cascade and ERK signaling, with ERK1/2 being of high interest in the pathogenesis of psoriasis because it serves as a target for glucocorticoids, which are used as a standard treatment (41, 42). In summary, this comparative gene expression analysis of radon- and UV-exposed psoriasiform skin lesions shows partially overlapping changes in psoriasis-relevant pathways.