Limited Presence of IL-22 Binding Protein, a Natural IL-22 Inhibitor, Strengthens Psoriatic Skin Inflammation

Psoriasis is a chronic inflammatory skin disease that affects ∼2% of the Caucasian population (1). Typical macroscopic skin alterations present as sharply demarcated, red, and slightly raised lesions with silver-whitish scales. It is generally acknowledged that these skin changes result from chronic dysregulated activation of the cutaneous immune system that—by secreted cytokines—alters the biology of local tissue cells (2, 3). Epidermal keratinocytes respond by hyperproliferation and impairment of their terminal differentiation, leading to epidermal thickening, hypogranularity, and hyperkeratosis. At the same time, these cells secrete high amounts of chemokines enabling the recruitment of further immune cells into the skin to create a self-sustained inflammatory milieu. Moreover, the high keratinocyte production of antimicrobial peptides prevents infections of the highly disturbed psoriatic epidermis. Among a range of immune mediators present in the psoriatic skin, an essential role for IL-23 and its downstream cytokine IL-17A in the maintenance of psoriatic lesions has been recently confirmed by the dramatic clinical success of their therapeutic blockade in these patients (4–7).

Another major IL-23 downstream cytokine is the IL-10 family member IL-22. IL-22 is also present in large quantities in psoriatic lesions (8). Main producers of IL-22 include different CD4⁺ T cell populations and group 3 innate lymphoid cells (9). Because the expression of the membrane IL-22R is restricted to epithelial/epithelioid cells (8), IL-22 assumes major cross-talk functions between immune and epithelial cells, especially at body barriers (10, 11). By inducing antimicrobial peptides and proinflammatory chemokines, as well as by inhibiting the terminal differentiation of keratinocytes, IL-22 has been shown to largely contribute to inflammation and epithelial remodeling of the psoriatic skin (12–16). Interestingly, the IL-22 activity is regulated by IL-22 binding protein (IL-22BP). IL-22BP is a soluble single chain receptor encoded by the gene IL22RA2 that specifically prevents the binding of IL-22 to IL-22R (17–20). Recently, we and others showed that conventional dendritic cell–derived IL-22BP exerts important inhibitory functions that impair the IL-22–mediated protection during acute colitis in rodents, although preventing potentially dangerous long-lasting proliferative effects on intestinal epithelial cells (21–23). Whether IL-22BP is expressed and plays a role in psoriatic skin inflammation is, however, unknown so far. Our study answers these questions in a translational approach using skin and blood samples from healthy control donors and psoriasis patients, in vitro experiments with human keratinocytes and reconstituted epidermis, and genetic deletion and pharmacological blockade approaches in rodents.

For analyses of skin expression, biopsies were obtained from adult healthy participants (22–63 y old [mean ± SD: 44.0 ± 10.9], 28.6% female), patients with plaque psoriasis (25–67 y old [mean ± SD: 46.4 ± 12.8], 22.7% female, 68.2% moderate to severe disease [psoriasis area and severity index (PASI) ≥10]), and patients with atopic dermatitis (22–40 y old [mean ± SD: 30.0 ± 7.7], 50.0% female, 100% moderate to severe disease). For analyses of blood mediators, blood samples were obtained from adult control participants (26–57 y old [mean ± SD: 38.3 ± 9.3], 72.0% female) and patients with plaque psoriasis (19–64 y old [mean ± SD: 41.2 ± 13.7], 61.1% female, 27.8% moderate to severe disease). The skin and blood samples were approved by the clinical institutional review board of the Charité University Medicine, Berlin. Written consent was obtained from all participants. The study was conducted according to the principles of the Declaration of Helsinki.

Underdeveloped EpiDerm-201 human epidermis models (MatTek) were cultured in inserts at the air–liquid interface as described previously (24) and stimulated or not (control) by supplementing the culture medium with either 20 ng/ml IL-22, 10 ng/ml IL-17A, a mixture of both, or 20 ng/ml IL-24 (all from R&D Systems, Wiesbaden-Nordenstadt, Germany). After 72 h, samples were taken for quantitative RT-PCR (qRT-PCR) analysis.

Il22ra2^−/− and Il22ra2^+/+ control littermate rats were generated on the Sprague-Dawley background using zinc-finger nucleases (Sigma-Aldrich, St. Louis, MO) at our local Rats Transgenesis Platform facility IBISA-CNRS as described previously (22). C57BL/6J mice were purchased from Centre d’Elevage Janvier (Le Genest-Saint-Isle, France).

All animals were kept under specific pathogen-free conditions. All animal studies were conducted in accordance with the EU Directive 2010/63/EU for animal experiments, and the guidelines of the French Agriculture Ministry. These studies were approved by the Veterinary Departmental Services committee (E.44011).

Snap frozen skin samples or epidermis models were mechanically homogenized (25). Total cellular RNA was isolated using Invisorb RNA kit II (Invitek/Stratec Molecular) (human samples) or TRIzol reagent (Invitrogen) (rodent samples) according to the manufacturers’ instructions. Reverse transcription was performed using murine moloney leukemia virus reverse transcriptase (Invitrogen) following the manufacturer’s instructions. Quantitative PCR on reverse transcribed mRNA was performed using the Maxima Probe/ROX qPCR Master Mix (Thermo Fisher Scientific/Fermentas) (human samples) or the TaqMan Fast Advanced Master Mix 2× reagent (rodent samples) (both from Applied Biosystems) and the StepOne Plus device (Applied Biosystems). Primers and double-labeled fluorescent probes were either self-designed (human IL-20, IL-22, IL-24, IL-22BP, and BD2) (26) or purchased from Applied Biosystems (all others). Expressions were normalized to HPRT using the 2-ΔΔ cycle threshold method, and results were expressed in arbitrary units.

Detection of blood serum levels of IL-22 and IL-22BP was performed using ELISA kits from Bio-Techne (Quantikine system) and Biozol (Cloud-Clone), respectively.

Psoriasis-like skin inflammation was induced in the right ear of Il22ra2^+/+ and Il22ra2^−/− rats, and in the back skin of C57BL/6J mice as previously described (27). Briefly, topical application of Aldara cream (3M Pharmaceuticals) was performed daily for 5 d. Skin thickness was measured daily with a Digimatic Caliper and the percentage of skin thickness increase relative to day 0 was calculated every day. Animals were sacrificed at day 5 for qRT-PCR and histopathological analyses on skin tissues. Skin sections were stained with H&E.

Anti–IL-22BP neutralizing Ab (28) or control isotype was injected i.p. at a dose of 10 mg/kg at days −1, 0, 2, and 4 in Aldara-treated mice.

Statistical analysis was performed with GraphPad Prism Software (GraphPad Software, San Diego, CA) or IBM SPSS Statistics 23.0 software (IBM, New York). Mean comparisons of unpaired samples were performed using the Mann–Whitney U test. The Wilcoxon matched-pairs signed-rank test (two-tailed) was used for paired samples. Correlations were calculated using the Spearman rank correlation test. The p values <0.05 were considered statistically significant.

To assess whether IL-22BP could exert a regulatory role in psoriasis, we first analyzed IL22RA2 expression in the skin from psoriasis patients as well as from atopic dermatitis patients and healthy donors as comparison groups. Confirming our previous works (17, 20), constitutive expression of IL-22BP was found in healthy donors’ skin. IL-22BP was also detected in the nonaffected skin of psoriasis patients. However, although IL-22BP expression in atopic dermatitis appeared rather increased as compared with healthy donors, levels in the nonaffected skin of psoriasis patients were in fact lower (Fig. 1A). We next analyzed IL-22 and IL-22BP expression in paired biopsies from nonaffected, perilesional, and lesional skin of psoriasis patients. In line with previous studies including ours (8, 15, 29), IL-22 expression was almost absent in nonlesional skin but showed a strong upregulation in perilesional (∼1200-fold) and lesional (2500-fold) skins (Fig. 1B). Importantly, in contrast to IL-22, the increase of IL-22BP expression was minimal in perilesional and lesional psoriatic skin (∼2-fold induction) (Fig. 1B). Thus, the very small IL-22/IL-22BP ratio in nonlesional skin (0.01) was largely increased in perilesional and lesional psoriatic skin to ∼3 and 7, respectively.

Taken together, these data suggest that constitutive low levels of IL-22BP in nonlesional psoriatic skin and their limited upregulation in lesional skin could enable broadly unregulated IL-22 to be effective and therefore might contribute to the development and maintenance of epidermal alterations in these patients.

In order to demonstrate that IL-22BP deficiency indeed strengthens IL-22–mediated pathogenicity in vivo during skin inflammation, we first took advantage of IL-22BP–deficient rats we generated recently (22) and characterized the impact of IL-22BP deficiency on the severity of skin lesions in the model of skin inflammation induced by imiquimod (30). Following daily topical application, TLR7 activation by imiquimod and inflammasome activation by the vehicle lead to an acute skin inflammatory process that critically depends on IL-23 and downstream mediators IL-17 and IL-22 (30–32). As expected, imiquimod application on the right ear of Il22ra2^+/+ rats led to skin erythema and thickening within 5 d (Fig. 2A, 2B). Histological analyses confirmed the induction of acanthosis, parakeratosis, Munro’s microabscesses, and dermal immune infiltration (Fig. 2C). Importantly, skin alterations were characterized by a strong induction of IL-22 and a moderate increase of IL-22BP, somehow mimicking the regulation of these two molecules in human psoriatic skin (Fig. 2D). Confirming our expectation that IL-22BP should negatively regulate IL-22 pathogenicity during skin inflammation, clinical and histological skin alterations were severely worsened in Il22ra2^−/− versus Il22ra2^+/+ littermate controls (Fig. 2A–C). In agreement with their more severe phenotype, enhanced expression of inflammatory cytokines, including IL-17A, TNF-α, and IFN-γ, was detected in lesional skin of Il22ra2^−/− rats (Fig. 2E). Moreover, concordant with exacerbated actions of IL-22 in the absence of IL-22BP, β-defensin-2 (BD2) and lipocalin-2 (LCN2), two antimicrobial peptides known to be induced by IL-22 in rodent epithelial cells (33, 34), showed significantly higher expression in the lesional skin of Il22ra2^−/− versus Il22ra2^+/+ littermate controls (Fig. 2F). Of note, there was no significant difference in the expression of these inflammatory cytokines and antimicrobial peptides in the skin of Il22ra2^−/− rat versus Il22ra2^+/+ littermate controls at steady state (data not shown).

To further demonstrate the protective role of IL-22BP against IL-22 in skin inflammation and avoid any potential bias that might arise from studies in genetically engineered animals, we sought to recapitulate the data obtained with IL-22BP–deficient rats through a pharmacological blockade of endogenous IL-22BP during imiquimod-induced skin disease in wild-type animals. As no neutralizing Ab against rat IL-22BP is currently available but does exist against mouse IL-22BP (28), we launched a new set of experiments in wild-type mice treated with imiquimod and injected i.p. with anti–IL-22BP or isotypic control Ab (Fig. 3A). Back skin lesions, including erythema, thickening, and scaling, were dramatically worsened in mice treated with the anti–IL-22BP Ab as compared with those injected with the control Ab (Fig. 3B, 3C), as confirmed by histological analyses (Fig. 3D).

Collectively, data obtained from rodent experiments thus argued in favor of a protective role of IL-22BP during skin inflammation through its ability to neutralize IL-22 pathogenic actions on keratinocytes.

To extend our findings in rodents and provide evidence that IL-22BP controls IL-22 in human psoriatic skin, we hypothesized that the IL-22/IL-22BP ratio should reflect the level of free bioactive IL-22 (i.e., not bound to IL-22BP) and should thus correlate with the expression of several molecules induced by IL-22 in the lesional psoriatic skin. As demonstrated in Fig. 4A, IL-22/IL-22BP ratios effectively positively correlated with IL-20 and IL-36γ, two cytokines coinduced in keratinocytes by IL-22 (35, 36) and playing a key role in psoriatic skin alterations (37, 38). Positive correlations were also measured between IL-22/IL-22BP ratios and the neutrophil-recruiting chemokine CXCL1, as well as the antimicrobial peptide BD2 (Fig. 4A), both genes also known to be strongly induced in keratinocytes by IL-22 (8, 14). As expected, the expression of CXCL1 and BD2 also positively correlated with that of IL-17A, although without reaching statistical significance for BD2 (p = 0.057) (Fig. 4B). This clearly matched the known facts about IL-17A and IL-22 synergistic actions in inducing these two molecules in keratinocytes (39). Finally, we observed a positive correlation between the IL-22/IL-22BP ratio and IL-24 (Fig. 4A), another member of the IL-10 family (9) known to participate in psoriasis skin lesions (40–42), but whose relationship with IL-22 has remained poorly established so far.

To get more insight into the relationship between IL-22 and IL-24, we moved to in vitro experiments using reconstituted human epidermis. IL-24 has been shown to be produced by keratinocytes (41), and we observed that its expression was in fact directly induced by IL-22 stimulation (Fig. 5A). Interestingly, a similar effect was observed after IL-17A stimulation, and the presence of both cytokines in the culture medium further increased the level of induction (Fig. 5B). This suggested that, as for BD2 and CXCL1 (39, 43), synergistic actions existed between IL-22 and IL-17A in regards to IL-24 induction in keratinocytes. Importantly, further characterization of the IL-24 action showed its autocrine action to inhibit the expression of genes, including CALML5, involved in the terminal differentiation of keratinocytes (Fig. 5C).

Together, these data indicate that, as previously described for IL-20 (35, 44), IL-22 together with IL-17, induces the expression of IL-24, which, by inhibiting keratinocytes differentiation, supports the installation of a self-sustained amplification loop responsible for skin alterations in psoriasis.

In the last part of our study we quantified IL-22 and IL-22BP protein concentrations in sera from psoriasis patients and healthy controls. Confirming previous reports including ours (13, 29, 45), levels of IL-22 were significantly elevated in psoriasis patients (Fig. 6A). We also found a slight but significant elevation of IL-22BP serum concentration in psoriasis patients as compared with healthy controls (Fig. 6A). These data thus reflected the extent of the expressional increase of IL-22 and IL-22BP in psoriatic lesions (Fig. 1A). Moreover, the IL-22/IL-22BP serum protein ratio showed excellent positive correlation with disease severity, as assessed by the PASI, but not with disease duration (Fig. 6B). These data thus further substantiated a pathophysiological role for IL-22BP in controlling IL-22–dependent skin inflammation in psoriasis by suggesting that the higher the level of free IL-22, the more severe psoriasis is.

Psoriasis is a common chronic skin disease whose pathogenesis is now admitted to largely consist of deregulated responses of the immune system promoting the production of high amounts of cytokines that are directly responsible for skin alterations (2). Due to the role of the immune system, the high prevalence of the disease, and the easily visible and quantifiable skin lesions, psoriasis can even be seen as an easy-to-access model of immune-mediated inflammatory diseases (IMIDs). This is especially true for IMIDs, which, like psoriasis, depend on the IL-23/IL-17/IL-22 pathway, including ankylosing spondylitis or inflammatory bowel disease for instance (46). Concordantly, the efficacy of several novel immune therapies targeting components of this pathway have often been successfully tried in psoriasis, firstly, as illustrated, with anti–IL-17 or anti–IL-23 Abs (4–7). Beyond this, however, there is currently very limited knowledge about the in situ regulation of the activity of these cytokines in psoriasis.

This question is particularly relevant for IL-22, which is part of the rare cytokines that possess a specific endogenous, secreted inhibitor that is IL-22BP (17–20). In a previous work, we showed that IL-22BP impaired the protective functions of IL-22 on the gut epithelium during dextran sulfate sodium–induced colitis, and suggested it may thus have a pathogenic role in inflammatory bowel disease (22). In this study, we provide evidence suggesting that insufficient production of IL-22BP in the psoriatic skin contributes to the development and maintenance of epidermal alterations. We indeed showed that, concomitant to the strong induction of IL-22 in the inflamed psoriatic skin, the IL-22BP increase was very moderate. Based on our results obtained in animal models, it is tempting to speculate that lower levels of IL-22BP expression we measured in nonlesional skin of psoriasis patients, as compared with healthy donors and atopic dermatitis patients, may predispose to emergence and maintenance of psoriatic lesions. The cellular and molecular mechanisms of this decrease nevertheless still remain to be understood. When preparing this manuscript, a report appeared in which IL-22BP expression was analyzed specifically in epidermis and dermis of skin biopsies from psoriasis patients (47). In contrast to our study, the authors did not observe any difference in both compartments between healthy donors and nonlesional skin of psoriasis patients (47). The reason for this discrepancy is unclear but might be due to the long-lasting and stressing procedure the authors used in their study to separate the epidermis from the dermis, which may have altered mRNA expressions and masked any possible preexisting differences.

To demonstrate that IL-22BP is able to dampen the extent and severity of IL-22–mediated skin inflammation in psoriasis, we chose to use the highly IL-22–dependent model of skin inflammation induced by the topical application of imiquimod (30, 32) through two complementary approaches. We first showed that, in comparison with littermate controls, a constitutive deficiency of IL-22BP in Il22ra2^−/− rats was associated with increased severity of the disease because of unleashed actions of IL-22, as further confirmed by enhanced expression of known IL-22–inducible genes in the lesional skin. These data confirm a recently accepted publication showing the same model of skin inflammation in Il22ra2^−/− mice (47). Importantly, we further corroborated these data by performing for the first time, to our knowledge, a pharmacological blockade of endogenous IL-22BP through injecting a neutralizing Ab (28) to imiquimod-treated mice and thus recapitulating our observations made in IL-22BP–deficient rats. Results from these two strategies thus allowed us to formally confirm that IL-22BP is able to play protective functions during skin inflammation by actively inhibiting the extent of IL-22 action. Moreover, this data obtained with an IL-22BP neutralizing Ab is of particular relevance because it highlights the possibility to target IL-22BP in conditions where the IL-22BP–mediated control of IL-22 is expected to be detrimental, like inflammatory bowel disease, as we previously suggested (22). In this context, it is important to note that not only the structure of respective genes, mRNA transcripts, and proteins (48), but also the IL-22–neutralizing capacity are virtually identical between human and rodent IL-22BP (23).

No correlation existed between IL-22BP and IL-22 expression in the lesional psoriatic skin (data not shown), suggesting that different inducer signals are involved in their respective upregulation. This different regulation of IL-22 and IL-22BP indicates that accurate interpretation of high levels of IL-22 production in the psoriatic skin must take concomitant levels of IL-22BP into account. In this sense, we observed a close positive association between the IL-22/IL-22BP ratio and the expression of several IL-22–inducible molecules known for their pathogenic roles in psoriasis (i.e., IL-20, IL-36γ, and CXCL1). Furthermore, the IL-22/IL-22BP ratio also positively correlated with IL-24, another member of the IL-20 subfamily of cytokines sharing pathogenic functions with IL-22 in psoriasis (9, 40–42). As previously suggested by in vitro experiments in cultured keratinocytes transfected with IL-22R cDNA (44), we confirmed IL-24 as an additional target gene of IL-22 in psoriasis. Interestingly, similar to other genes including those encoding CXCL1 and BD2 (39), IL-24 expression was further increased when IL-17A was added to IL-22 in the culture medium. This suggested that a number of pathogenic effects attributed to either IL-22 or IL-17A in the psoriatic skin are in fact dependent on synergic actions of both cytokines on keratinocytes, and are thus subjected to IL-22BP negative regulation. For psoriasis pathogenesis, this also indicates the existence of a self-sustained amplification loop in which the T cell–derived cytokines not only directly induce the visible epidermal alterations in psoriasis but also provoke the expression of keratinocyte mediators further worsening these alterations.

Finally, we substantiated the importance of the protective functions assumed by IL-22BP through addressing its serum levels in relation to those of IL-22 in psoriasis patients. We indeed showed that the IL-22/IL-22BP serum protein ratio strongly correlated with the disease severity as assessed by the PASI, which suggests that the higher the control of IL-22 by IL-22BP, the less severe psoriasis is. In future prospective studies it will be interesting to address whether the IL-22/IL-22BP blood protein ratio may predict therapy success (e.g., anti–IL-17 or anti-p40 therapy), proofing the value of this ratio as a potential new biomarker in psoriasis.

In conclusion, we showed that IL-22BP critically controls deleterious actions of IL-22 during skin inflammation in rodents and suggest it is insufficiently present to limit the extent of IL-22 pathogenicity in human psoriasis. Together with our recent report on the role of IL-22BP during inflammatory bowel diseases, our results identify IL-22BP as a critical regulator of IL-22 actions on epithelial cells during IMIDs.

We are grateful to La Plate-Forme MicroPICell (Structure Fédérative de Recherche François Bonamy, Nantes University, INSERM) for help with immunohistology experiments.

The authors have no financial conflicts of interest.