Key Points
S100a8- and S100a9-deficient mice have worsened psoriasis.
S100A8 and S100A9 inhibit IL-17 production.
Abstract
High concentrations of the damage-associated molecular patterns S100A8 and S100A9 are found in skin and serum from patients suffering from psoriasis, an IL-17–related disease. Notably, although the expression of these proteins correlates with psoriatic disease severity, the exact function of S100A8 and S100A9 in psoriasis pathogenesis remains unclear. In this study, we investigated the role of S100A8 and S100A9 in psoriasis-associated skin hyperplasia and immune responses using S100a8−/− and S100a9−/− mice in an imiquimod-induced model of psoriasis. We found that S100a8−/− and S100a9−/− psoriatic mice exhibit worsened clinical symptoms relative to wild-type mice and increased expression of S100A9 and S100A8 proteins in keratinocytes, respectively. In addition, the loss of S100A8 enhances proliferation of keratinocytes and disrupts keratinocyte differentiation. We further detected elevated production of IL-17A and -F from CD4+ T cells in the absence of S100A8 and S100A9, as well as increased infiltration of neutrophils in the skin. In addition, treatment with anti–IL-17A and -F was found to reduce psoriasis symptoms and skin hyperplasia in S100a8−/− and S100a9−/− mice. These data suggest that S100A8 and S100A9 regulate psoriasis by inhibiting production of IL-17A and -F, thereby, to our knowledge, providing new insights into their biological functions.
Introduction
Psoriasis is a common immune-mediated skin disease affecting ∼2–4% of the world’s population (1). This multifactorial IL-17–driven condition is influenced by multiple genes, as well as by environmental factors, including trauma and infections. Psoriatic skin is characterized by epidermal hyperplasia, leukocyte infiltration, and more dermal blood vessels (2). Although the exact factors that trigger psoriasis remain unknown, both innate and adaptive immunity are involved in its pathophysiology. In particular, the IL-23–Th17 axis, which promotes keratinocyte and endothelial cell proliferation and the production of antimicrobial peptides, such as S100A8/A9 (3), has been implicated in psoriasis pathogenesis (4, 5).
S100A8 and S100A9 are two Ca2+-binding proteins from the S100 family, which have been characterized as damage-associated molecular patterns (6). Myeloid cells, including neutrophils, monocytes, and dendritic cells (DCs), constitutively express and secrete S100A8 and S100A9 (7, 8), and their expression is inducible in synoviocytes (9), keratinocytes (10), epithelial cells (11), and endothelial cells (12). These proteins occur as noncovalently bonded homodimers and form a noncovalent heterodimer called S100A8/A9, or calprotectin, in the presence of calcium (13). Notably, the S100A8 and S100A9 homodimers are not always coexpressed (12), suggesting they also perform distinct functions in various cellular contexts.
S100A9 has been characterized as a proinflammatory factor, and previous studies have reported that this protein stimulates proinflammatory cytokine secretion from monocytes/macrophages (14, 15), neutrophil phagocytosis (16) and degranulation (17), and phagocyte migration (18). S100a9−/− mice are also resistant to adjuvant-induced arthritis and systemic lupus erythematosus, the latter of which is due, at least in part, to reduced CD8+ T cell activation (19, 20). In addition, S100a9-deleted mice have been shown to exhibit less inflammation, neurodegeneration, and cognitive deficits than wild-type (WT) animals in a mouse model of Alzheimer disease (21). However, deletion of S100a9 in TNF-overexpressing mice worsens joint inflammation and enhances keratinocyte proliferation (22), suggesting distinct functions for S100A9 in different inflammatory diseases.
Similarly, multiple lines of evidence indicate a proinflammatory role for S100A8. This protein is one of the most potent chemotactic factors for neutrophils and monocytes (23), and injection of anti-S100A8 Abs reduces leukocyte recruitment in different models of acute inflammation (24, 25). However, S100A8 is sensitive to oxidation by reactive oxygen species (26–29), and oxidized S100A8 reduces IgE-mediated mast cell degranulation and cytokine secretion (30). In addition, the chemotactic potential of S100A8 is lost when oxidized (26, 29). In mice, S100a8−/− animals have more granulocyte and monocyte precursors, known as granulocyte–monocyte progenitors, in bone marrow, as well as increased numbers of mature neutrophils and monocytes in peripheral blood, relative to WT controls. Further, absence of S100a8 exacerbates the symptoms of collagen-induced arthritis symptoms by inducing differentiation and activation of osteoclasts and enhancing neutrophil infiltration (31), suggesting that S100A8 is anti-inflammatory in this context.
Concentrations of the S100A8/A9 heterodimer are elevated in the serum of psoriasis patients, and the levels are correlated with disease activity (32, 33). This increased expression of both S100A8 and S100A9 is a characteristic of psoriatic skin lesions and is also present in mouse models of psoriasis (34–37). In particular, expression of these proteins is associated with early events in psoriasis models, occurring before histological alterations or cytokine dysregulation (38, 39). Conversely, expression of S100A8/A9 is weak or absent in the skin of healthy individuals and in nonlesional skin from psoriasis patients (40). This increased expression of S100A8/A9 in psoriatic skin has been reported to be driven by high levels of IL-22 (3) and lens epithelium–derived growth factor (LEDGF) (41). Although the precise role of S100A8 and S100A9 in psoriasis remains unclear, these proteins can promote keratinocytes proliferation and have been shown to act as proangiogenic factors in skin by inducing endothelial cell proliferation, migration, and capillary-like tube formation (42, 43). In addition, S100A8/A9 heterodimers that are released by keratinocytes regulate the activation of dermal fibroblasts upon dehydration of human skin (44) and modulate the expression and release of complement C3 by keratinocytes (45). These observations suggest that S100A8 and S100A9 are potential mediators in psoriasis, with possible intracellular and extracellular functions in epidermal hyperplasia, leukocyte infiltration, and angiogenesis.
Therefore, pro- and anti-inflammatory activities have been described for S100A8, S100A9, and S100A8/A9, making it difficult to decipher their respective effects on inflammatory responses. In this study, we investigated the roles of S100A8 and S100A9 in a murine model of imiquimod-induced psoriasis, using both S100a8−/− and S100a9−/− mice. Our data show that in the absence of S100A8 and S100A9, psoriasis symptoms are aggravated because of increased hyperplasia, neutrophil infiltration, and Th17 responses. These results suggest that S100A8 and S100A9 regulate inflammation in Th17-driven diseases.
Materials and Methods
Mice and treatments
All experiments were performed in accordance with the Université Laval Animal Protection Committee (protocol no. 18-005). C57BL/6 (The Jackson Laboratory, Bar Harbor, ME), S100a8Tm1(KOMP)Vclg (S100a8−/−), and S100a9Tm1(KOMP)Vclg (S100a9−/−) (31, 46) mice (8–10 wk old) received daily topical applications of 80 mg commercially available 5% imiquimod cream (Aldara; Bausch Health, Laval, QC, Canada) for 10 d on their shaved back. Erythema, desquamation, and hyperplasia of the skin were scored daily by a blinded observer on a scale from 0 to 3 (0, none; 1, slight; 2, moderate; 3, marked) for a maximum cumulative local Psoriasis Area Severity Index (PASI) score of 9.
Isolation of cells
Mice were sacrificed after 10 d of treatment by isoflurane inhalation, followed by cervical dislocation, and the blood, back skin, femurs, and lymph nodes (LNs) were recovered. Bone marrow cells were collected from the femurs by flushing with RPMI 1640 medium supplemented with 5% FBS. Erythrocytes were lysed by suspending cells in 0.15 M NH4Cl for 5 min, followed by centrifugation at 300 × g, and then washed with PBS. Inguinal LNs were disrupted by passage through a 70-μm filter with a syringe piston.
Dermal cell suspensions were prepared as previously described (47). Briefly, skin from the back was extracted and kept in cold RPMI 1640 containing 5% FBS. Sheets of ear skin were separated using forceps. The skin was then incubated in a solution of HBSS + 2.4 mg/ml Dispase II (Sigma-Aldrich, St. Louis, MO) for 60 min at 37°C to separate epidermis from dermis. The epidermis was discarded using forceps and the dermis cut into small pieces with scissors and then incubated in a solution of HBSS supplemented with 10% FBS and 2 mg/ml Collagenase D (Sigma-Aldrich) for 2 h. Cell suspensions were homogenized by passage through a 70-μm filter with a 19G syringe.
Lymphocyte stimulation
Cells from inguinal LNs and skin were rinsed with sterile PBS + 2% FBS and then seeded at 1 × 106 cells/ml in RPMI 1640 containing 10% FBS, 0.2% Primocin (InvivoGen, San Diego, CA), 1× nonessential amino acids, 50 ng/ml PMA (Sigma-Aldrich), 1 μM ionomycin (Sigma-Aldrich), and 1 μl/ml GolgiStop (BD Biosciences, San Diego, CA) for 5 h at 37°C and 5% CO2. Cells were then centrifuged for 10 min at 300 × g, rinsed with PBS, and analyzed by flow cytometry.
Flow cytometry
Supplemental Table I). Intracellular Ags were detected by labeling with mAbs (listed in Supplemental Table I
Histological analysis
Immunofluorescence staining
ELISAs and simplex assays
Blood was collected and centrifuged, and serum samples were frozen at −20°C. S100A8 and S100A9 homodimers and S100A8/A9 heterodimers were quantified by ELISA, as described previously (46
Skin homogenate preparation
g
Statistical analysis
Statistical analysis was performed using GraphPad Prism v7 (GraphPad, San Diego, CA). Two-way ANOVA with Bonferroni correction was used to assess local PASI scores of imiquimod-induced psoriasis, and the Mann–Whitney U test was used for other experiments; p values <0.05 were considered significant, with *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001.
Results
Deletion of S100a8 and S100a9 aggravates imiquimod-induced psoriasis
To investigate the roles of S100A8 and S100A9 in psoriasis, we first analyzed the expression and localization of these proteins by immunofluorescence in imiquimod-treated skin from WT mice. Whereas keratinocytes from naive mice do not express detectable S100A8 and S100A9, we observed a strong coexpression of S100A8 and S100A9 in the epidermal layer of imiquimod-treated mice as well as in infiltrating dermal leukocytes (Fig. 1A). Interestingly, S100A8 and S100A9 were detected in the cytosol and nuclei of apical keratinocytes. It was previously shown that S100A8 and S100A9 homodimers, as well as S100A8/A9 heterodimers (calprotectin), are secreted by keratinocytes and myeloid cells (48, 49). We therefore quantified these species in serum from imiquimod-treated and control mice. Notably, we detected a >2-fold increase in calprotectin concentrations in serum from imiquimod-treated mice, as compared with naive mice (Fig. 1B). S100A9 homodimers were also found to be slightly increased in imiquimod-treated mice relative to controls (p = 0.5476; Fig. 1C), whereas S100A8 homodimers were undetectable in serum (data not shown).
Extracellular S100A8 and S100A9 alleviate symptoms of imiquimod-induced psoriasis. (A) Skin sections from WT mice treated with imiquimod for 10 d stained with anti-S100A8 (green) or anti-S100A9 (red) and DAPI (blue). Scale bar, 25 μm. (B and C) Levels of S100A8/A9 heterodimer and S100A9 homodimer in serum from mice with imiquimod-induced psoriasis on day 10. Data represent the mean ± SEM; control, n = 5 mice; imiquimod, n = 6 mice. (D) Cumulative local PASI scores for WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis. Data represent the mean ± SEM; n = 6 mice per group. (E) Skin sections from S100a8−/− and S100a9−/− mice treated with imiquimod for 10 d stained with anti-S100A8 (green) or anti-S100A9 (red) and DAPI (blue). Scale bar, 25 μm. (F) Local PASI scores for WT mice with imiquimod-induced psoriasis treated with anti-S100A8 or control Abs. Data represent the mean ± SEM; n = 6 mice per group. (G) Local PASI scores for WT mice with imiquimod-induced psoriasis treated with anti-S100A9 or control Ab. Data represent the mean ± SEM; n = 6 mice per group. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 calculated using the Mann–Whitney U test (B and C) or two-way ANOVA with Bonferroni correction (D, F, and G).
We next induced psoriasis in S100a8- and S100a9-deficient mice and measured imiquimod-induced erythema, skin hypertrophy, and desquamation in knockout and WT control animals. We found that although clinical scores decreased in WT mice by day 6, they remained elevated in both S100a8- and S100a9-deficient mice from day 8 to 10 (Fig. 1D). As expected, expression of S100A8 and S100A9 is absent in skin from S100a8−/− and S100a9−/− mice, respectively. However, de novo expression of S100A9 and S100A8 was observed in skin from imiquimod-treated S100a8−/− and S100a9−/− mice, respectively, with localization primarily detected in the cytosol and nucleus of keratinocytes in the apical side of the epidermis (Fig. 1E). To further decipher the contributions of extracellular and intracellular S100A8 and S100A9 in psoriasis, we then neutralized the activities of extracellular S100A8 and S100A9 proteins using Abs in WT mice (Fig. 1F, 1G). We observed a significant increase in local PASI scores in psoriatic mice treated with anti-S100A8 and anti-S100A9, indicating that extracellular S100A8 and S100A9 function to mitigate the symptoms of psoriasis in this model. Thus, although intracellular functions for S100A8 and S100A9 have been described in keratinocytes (45, 50), these observations indicate that extracellular S100A8 and S100A9 function to reduce psoriasis symptoms.
S100A8 modulates proliferation and differentiation of mouse keratinocytes
Because imiquimod-induced psoriasis is characterized by skin hyperplasia, we measured the dermis and epidermis from skins of naive and imiquimod-treated mice. As expected, for all groups, we observed significant thickening of the dermis and epidermis in imiquimod-treated compared with naive mice (Fig. 2A, 2B). Additionally, no differences in dermal thickness were observed in imiquimod-treated WT, S100a8−/−, and S100a9−/− mice. However, we detected significantly enhanced epidermal hyperplasia in S100a8−/− and S100a9−/− mice relative to WT mice (with a moderate increase for S100a9−/− mice; Fig. 2C). Assessment of keratinocyte proliferation using Ki-67 staining revealed a higher number of Ki-67+ cells in S100a8−/− epidermis compared with WT epidermis (Fig. 2D, 2E), suggesting that S100A8 functions to negatively regulate keratinocyte proliferation. In contrast, no significant differences in epidermal proliferation were observed between S100a9−/− and WT, suggesting that keratinocyte proliferation is independent of S100A9.
S100A8 regulates keratinocyte proliferation and differentiation. (A) Skin sections from WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d stained with H&E. Scale bar, 200 μm. (B) Dermal and (C) epidermal thickness for WT, S100a8−/−, and S100a9−/− mice at treatment day 10. Data represent the mean ± SEM; n = 6 mice per group. (D) Skin sections from WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d stained with anti–Ki-67 (green) and DAPI (blue). Scale bar, 25 μm. (E) Numbers of Ki-67+ nuclei per high-power field (HPF) on treatment day 10 in skin samples from WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis. Data represent the mean ± SEM of at least 18 HPF from slides of n = 6 mice per group. (F) Skin sections from WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d stained with anti-K10 (green) and DAPI (blue). Scale bar, 25 μm. (G) Mean fluorescence intensity of K10 signal in the entire epidermis area from WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d. Data represent the mean ± SEM; n = 2 (WT) or 5 (S100a8−/− and S100a9−/−) mice per group. *p < 0.05, **p < 0.01, calculated using the Mann–Whitney U test (B, C, and E).
In naive skin, keratinocytes proliferate in the basal side of epidermis, then differentiate and express terminal differentiation markers, such as K10 (51). Disruption of this controlled proliferation is accompanied by modified differentiation. We further found that S100A8 and S100A9 are mostly expressed by nonproliferating (Ki-67–negative) cells localized in the apical side of the epidermis in imiquimod-treated S100a9−/− and S100a8−/− mice, respectively (Supplemental Fig. 1A–E). To determine whether keratinocyte differentiation is affected by S100A8 and S100A9 deletion, we measured expression of the early keratinocyte maturation marker K10 by microscopy in WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis. Homogeneous K10 expression was detected along epidermal cells in psoriatic WT and S100a9−/− mice. Moreover, K10 expression was heterogeneously distributed in psoriatic S100a8−/− epidermis, indicating that keratinocyte differentiation is disrupted in the absence of S100A8 (Fig. 2F, Supplemental Fig. 1F). Overall, expression of K10 was found to be decreased in psoriatic S100a8−/− and S100a9−/− epidermis (Fig. 2F, 2G). Collectively, these results suggest that S100A8 both inhibits proliferation and regulates differentiation of keratinocytes in response to imiquimod treatment.
Deletion of S100A8 and S100A9 enhances IL-17 responses in draining LNs and the dermis
We next investigated the effects of S100A8 and S100A9 on the immune response in our model of imiquimod-induced psoriasis. To this end, we first measured expression of S100A8 and S100A9 in CD11c-expressing cells from skin-draining LNs in naive and imiquimod-treated WT mice. We found that whereas fewer than 1% of CD11c+ cells from LNs of naive WT mice coexpress S100A8 and S100A9, we detect a small population (∼4%, p = 0.1) of CD11c+ cells that strongly coexpress S100A8 and S100A9 in imiquimod-treated WT mice (Fig. 3A, 3B, Supplemental Fig. 2A). Among the cells that are recruited to inflammatory sites, monocytes can give rise to DCs, including CD11b+ monocyte-derived DCs (moDCs), which are essential for inducing psoriasis-like skin inflammation (52). Similar numbers of CD11b+ CD11c+ MHC class II (MHCII)+ moDCs were detected in the dermis from naive WT, S100a8-deficient, and S100a9-deficient mice (Supplemental Fig. 2B, 2C). However, we observed increased migration of CD11b+ CD11c+ MHCII+ moDCs to the inguinal LNs in psoriatic S100a8−/− mice, but not S100a9−/− mice, relative to WT animals (Fig. 3C, Supplemental Fig. 2D). These results suggest that in the absence of S100A8, there is increased migration of myeloid DCs from the skin to the draining LNs.
The IL-17A/F response is increased in S100a8−/− and S100a9−/− mice. (A) Expression of S100A8 and S100A9 in CD45+ CD11c+ cells from LNs of naive and imiquimod-treated WT mice on day 10. (B) Percentage of CD45+ CD11c+ cells coexpressing S100A8 and S100A9. Data represent the mean ± SEM; n = 3 mice per group. (C) CD11bhigh CD11c+ MHCII+ cells in inguinal LNs (iLNs) of naive and imiquimod-treated WT, S100a8−/−, and S100a9−/− mice at day 10. Data represent the mean ± SEM; n = 6 mice per group. (D and E) CD4+ TCRγ/δ− IL-17A/F+ and CD4+ TCRγ/δ+ IL-17A/F+ cells in iLN from WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d. Data represent the mean ± SEM; n = 5 (WT) or 6 (S100a8−/− and S100a9−/−) mice per group. (F and G) CD4+ TCRγ/δ− IL-17A/F+ and CD4+ TCRγ/δ+ IL-17A/F+ T cells in dermis of WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d. Data represent the mean ± SEM; n = 4 (WT) or 6 (S100a8−/− and S100a9−/−) mice per group. *p < 0.05, **p < 0.01, calculated using the Mann–Whitney U test (C–F).
We next measured the IL-17A/F–expressing T cells in inguinal LNs and dermis of psoriatic mice by flow cytometry. Imiquimod treatment led to a significant increase in the number of CD4+ TCRγ/δ− IL-17A/F+ in the draining LNs of S100a8−/− and S100a9−/− mice at day 10 compared with WT mice. CD4+ TCRγ/δ+ IL-17A/F+ T cell numbers were also slightly increased in the draining LNs of S100a8−/− and significantly increased in S100a9−/− mice (Fig. 3D, 3E, Supplemental Fig. 3E). In the dermis, CD4+ TCRγ/δ− IL-17A/F+ T cells were found to be significantly increased in S100a8−/− and S100a9−/− mice compared with WT mice, whereas CD4+ TCRγ/δ+ IL-17A/F+ cells were diminished in S100a9−/− mice (Fig. 3F, 3G, Supplemental Fig. 3A, 3B).
Increased IL-17A/F production amplifies keratinocyte proliferation and hyperplasia in S100a8−/− and S100a9−/− mice
To confirm that the increased inflammation observed in the absence of S100A8 and S100A9 is dependent on IL-17, we treated psoriatic WT, S100a8−/−, and S100a9−/− mice with neutralizing anti–IL-17A, anti–IL-17F, or control Abs. We observed a significant reduction in clinical scores for S100a8−/− and S100a9−/− mice treated with anti–IL-17A Abs compared with control Abs, but not for WT mice (Fig. 4A). In addition, injection of neutralizing anti–IL-17F Abs resulted in a significant reduction in psoriasis in WT, S100a8−/−, and S100a9−/− mice (Fig. 4B, Supplemental Fig. 3C, 3D). This was also associated with a significant decrease in epidermal hyperplasia as well as with a reduction in proliferating keratinocytes (Ki-67+ cells) in S100a8−/− and S100a9−/− mice (Fig. 4C, 4D). Together, these results suggest that the Th17 response is enhanced in the absence of S100A8 and S100A9, resulting in increased keratinocyte proliferation and skin hyperplasia.
Treatment with anti–IL-17A and anti–IL-17F inhibits hyperplasia and keratinocyte proliferation in S100a8−/− and S100a9−/− mice. Cumulative local PASI scores for WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis treated with (A) anti–IL-17A or (B) anti–IL-17F and control Ab. Data represent the mean ± SEM; n = 10–12 (control Ab) or 6 (anti–IL-17A and anti–IL-17F) mice per group. (C) Epidermal thickness of WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis treated with anti–IL-17A, anti–IL-17F, and control Ab on day 10. Data represent the mean ± SEM; n = 10–12 (control Ab) or 6 (anti–IL-17A and anti–IL-17F) mice per group. (D) Numbers of Ki-67+ nuclei per high-power field (HPF) at day 10 for WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis treated with anti–IL-17A, anti–IL-17F, or control Ab. Data represent the mean ± SEM of 12 HPF from sections of four individual animals. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, calculated using the Mann–Whitney U test (C and D) or two-way ANOVA with Bonferroni correction (A and B).
Infiltration of neutrophils is increased in imiquimod-treated S100a8−/− and S100a9−/− mice
Finally, we examined skin inflammation in the absence of S100a8 and S100a9 and detected a significant increase in the concentration of the neutrophil chemoattractant CXCL1/GRO-α in the skin of both S100a8−/− and S100a9−/− mice treated with imiquimod relative to their WT littermates (Fig. 5A). We then measured infiltration of neutrophils into the skin and detected a 3-fold increase in the number of Ly-6G+ cells in the dermis of psoriatic S100a8- and S100a9-deficient mice at day 10, as compared with psoriatic WT mice (Fig. 5B, 5C). Because the Th17 response is augmented in S100a8−/− and S100a9−/− mice, we next quantified dermal infiltrating Ly-6G+ cells after treatment with anti–IL-17A and anti–IL-17F. We found that anti–IL-17A and IL-17F had no effect on neutrophil accumulation in the dermis of WT mice. However, we detected significantly fewer Ly-6G+ cells in both psoriatic S100a8−/− and S100a9−/− mice treated with anti–IL-17A or IL-17F, as compared with imiquimod-treated WT mice (Fig. 5D). Thus, our results indicate that deletion of S100a8 and S100a9 results in enhanced expression of IL-17A/F and CXCL1 in an imiquimod-induced model of psoriasis, leading to the accumulation of neutrophils in the dermis.
Enhanced neutrophil infiltration is detected in the skin of S100a8−/− and S100a9−/− mice. (A) CXCL1 levels in skin homogenates from WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d. Data represent the mean ± SEM; n = 4–6 mice per group. (B) Skin sections from WT, S100a8−/−, and S100a9−/− mice treated with imiquimod for 10 d stained with anti–Ly-6G (red) and DAPI (blue). Scale bar, 50 μm. (C) Ly-6G+ cells per high-power field (HPF) of sections from WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis. Data represent the mean ± SEM (n = 6 mice). (D) Ly-6G+ cells per HPF of skin sections of WT, S100a8−/−, and S100a9−/− mice with imiquimod-induced psoriasis treated with anti–IL-17A, anti–IL-17F, or control Ab. Data represent the mean ± SEM; n = 4 mice per group. *p < 0.05, calculated using the Mann–Whitney U test (A, C, and D).
Discussion
Previous studies have reported that the serum and skin from psoriasis patients contain elevated levels of S100A8 and S100A9, and their levels correlate with disease activity (32, 53). In addition, S100A8/A9 is a biomarker of subclinical inflammation (54), and serum S100A8/A9 predicts treatment escalation in chronic inflammatory syndromes, such as irritable bowel disease (55), and nonresponsiveness in rheumatoid arthritis patients (56). Thus, there is a strong correlation between S100A8/A9 levels and disease intensity in chronic inflammation. In this study, we investigated the role of these proteins in psoriasis using a mouse model of disease and identify S100A8 and S100A9 as negative regulators of the IL-17 immune axis in psoriasis. We show that S100A8 and S100A9 are expressed by keratinocytes and myeloid cells in psoriatic skin and serum. In addition, psoriatic S100a8−/− and S100a9−/− mice display aggravated disease relative to WT controls, with elevated production of IL-17A and -F from CD4+ T cells. We further show that the increased levels of IL-17A and -F in S100a8−/− and S100a9−/− mice promote neutrophil migration to the skin as well as keratinocyte proliferation. Taken together, these results suggest that S100A8 and S100A9 negatively regulate psoriatic skin inflammation by inhibiting production of IL-17A and -F.
To our knowledge, this study is the first to compare inflammatory responses in both S100a8−/− and S100a9−/− mice. S100A8 and S100A9 are the most abundant alarmins found in the serum of psoriatic patients (32), and elevated expression of these proteins has also been reported in several mouse models of psoriasis (37, 49, 57). They have therefore been proposed as biomarkers for psoriasis (33) and psoriatic arthritis (58); however, their exact functions in the pathogenesis of psoriasis remain controversial. In this study, we observed worsened psoriasis symptoms in S100a8−/− and S100a9−/− mice, as compared with WT mice, with less-severe clinical scores noted for S100a9−/− compared with S100a8−/− animals. This suggests that the S100A8 and S100A9 homodimers regulate inflammation differently and have distinct functions. However, the possibility still exists that S100A8/A9 heterodimers repress inflammation, with S100A8 homodimers and S100A9 homodimers having anti-inflammatory and proinflammatory functions, respectively. Indeed, a recent study suggests that the S100A8/A9 heterodimer is involved in an autoinhibitory mechanism through the tetramerization of two S100A8/A9 heterodimers in vitro that blocks the cascade of activation linked to TLR4/myeloid differentiation factor-2 (MD-2) binding (22).
Naive S100a8−/− and S100a9−/− mice lack both the respective S100A8 and S100A9 proteins and the S100A8/A9 heterodimer. Although the S100A8/A9 heterodimer is the secreted form predominantly measured in plasma of psoriatic WT mice, the S100A8 and S100A9 homodimers are de novo expressed in inflamed skin and neutrophils from S100a9−/− and S100a8−/− mice, respectively, and can exhibit both intracellular and extracellular functions. In this study, we found that although deletion of S100a8 or S100a9 had similar effects on Th17 activation and neutrophil migration to the skin, increased proliferation and reduced differentiation of keratinocytes were also observed S100a8−/− mice. Because the S100A9 protein is still expressed in psoriatic S100a8−/− keratinocytes, these results suggest that either S100A8 downregulates keratinocyte proliferation or the S100A9 homodimers induce keratinocyte proliferation. Therefore, the phenotypic differences observed in S100a8−/− and S100a9−/− animals, particularly those associated with keratinocyte proliferation and differentiation and DC recruitment, cannot be explained only by the absence of the heterodimer S100A8/A9 but are more likely because of the distinct activities of the S100A8 and S100A9 homodimers. The development of a double mutant for S100a8 and S100a9 and the comparison between both single mutants and the double mutant would be helpful to determine the contribution of both homodimers in the absence of the heterodimer. We further note that no notable differences were observed in keratinocyte proliferation and differentiation in the absence of S100A9 in this model of psoriasis. However, S100A9 appears to control keratinocyte proliferation in some contexts, as epidermal tissue from S100a9−/− mice displays increased numbers of Ki-67–positive keratinocytes in a model of papillomas (59). Given the high expression of S100A8 and S100A9 observed in keratinocytes from hypertrophic scars and keloid tissues (44), it is tempting to speculate that these proteins participate in the hyperproliferation of keratinocytes in pathological conditions.
In this study, we further found that loss of S100A8 and S100A9 resulted in an increased IL-17A/F response in both skin and LNs. In S100a8−/− and S100a9−/− psoriatic mice, IL-17A/F production from CD4+ T cells originated mainly from TCRγ/δ− cells. Moreover, we showed that the amplified hyperplasia that occurs in S100a8−/− and S100a9−/− mice is highly dependent on the secretion of IL-17A and IL-17F. IL-17A and IL-17F have been shown to contribute equally to psoriatic inflammation induced by imiquimod (60). In addition, IL-17A induces increased proliferation and aberrant differentiation of keratinocytes in psoriatic skin (61) and functions to amplify the inflammatory network by promoting release of antimicrobial peptides and proinflammatory cytokines and chemokines (62). We detected an increased migration of DCs to the LNs in psoriatic mice, which might explain the increased presence of Th17 cells. In turn, the enhanced secretion of IL-17A and IL-17F would promote keratinocyte proliferation and the migration of inflammatory cells to the skin, leading to aggravation of psoriasis symptoms. We also found that the IL-17F–neutralizing Ab was more effective in reducing psoriasis symptoms than the anti–IL-17A Ab, which is consistent with the finding that Il17f−/− mice show greater resistance to imiquimod-induced psoriasis than Il17a−/− mice (60). Based on these findings, we hypothesize that myeloid-derived DCs from S100a8−/− and S100a9−/− mice are more efficient in activating Th17 responses via cytokine secretion. However, further investigations will be needed to verify this hypothesis.
Manitz et al. (63) reported that neutrophil migration is weakened in S100a9−/− mice. However, in a previous study, we observed increased neutrophil infiltration into arthritic paws in the absence of S100A8 (31). In addition, in this study, we found that neutrophil recruitment is not diminished in psoriatic skin of S100a8−/− and S100a9−/− mice, suggesting that migration of neutrophils is not impaired in the absence of S100A8, S100A9, and/or the S100A8/A9 heterodimer. In contrast, we observed a greater infiltration of neutrophils through the dermis of both S100a8−/− and S100a9−/− mice, which is associated with an increased concentration of CXCL1 in the skin, and this infiltration was found to be dependent on IL-17A and IL-17F. However, as circulating neutrophils are more numerous in S100a8-deficient mice (31), we cannot rule out the possibility that enhanced migration of neutrophils to the sites of inflammation observed in S100a8−/− and S100a9−/− mice might be due to their increased availability from the circulation.
S100A8 was first identified as a potent neutrophil chemotactic factor (64). In addition, anti-S100A8 was reported to reduce myeloid cell accumulation in models of gout, streptococcal pneumonia, and LPS-induced inflammation (18, 65, 66), indicating that this protein exerts proinflammatory functions. However, glucocorticoids and IL-10 induce S100A8 expression (67–69), suggesting an anti-inflammatory activity for this protein as well. In addition, S100A8 induces expression of IL-10, and S100A8 inhibition aggravates chronic inflammation, including collagen-induced arthritis (31), further indicating that it functions to dampen the inflammatory immune response. In this study, our results support an anti-inflammatory role for S100A8 in psoriasis. Extracellular activity of S100A8 is controlled by oxidation and nitrosylation on its cysteine residue, leading to the formation of covalent bonds between monomers (27, 28). A number of reports demonstrate that oxidized S100A8 exerts anti-inflammatory effects by inhibiting mast cell degranulation and cytokine secretion induced by FcεR cross-linking (28, 30, 70, 71). Thus, we propose a model in which S100A8 is rapidly secreted during inflammatory reactions, thereby enhancing inflammatory responses until its oxidization by reactive oxygen species is produced at the site of inflammation. Oxidized S100A8 would then function to dampen the immune response in chronic inflammation.
S100A9, which is widely viewed as a proinflammatory factor, can induce phagocytosis, degranulation, and reactive oxygen species production by neutrophils and monocytes (16). This protein also activates T cells and can promote cytokine secretion from monocytes via activation of NF-κB signaling and the inflammasome. In addition, studies have shown that blocking S100A9 reduces chronic inflammation (14, 72), and S100a9−/− mice display reduced inflammation in a model of Ag-induced arthritis (73). Surprisingly, we found that in our model of imiquimod-induced psoriasis, inflammation is increased in S100a9−/− mice. This result is in contrast with the protective effect provided by deletion of S100a9 reported in the Jun and JunB double-knockout model of psoriasis (50). However, because these two psoriasis models differ in their requirement for IL-17, we speculate that this discrepancy might be due to the essential role played by IL-17 in the imiquimod model of disease. As in human psoriasis, the IL-23/IL-17 inflammation axis is essential for the mouse model of imiquimod-induced psoriasis (60). However, because the AP-1 transcription factor JunB is required for Th17 cell differentiation (74), this axis cannot be involved in the Jun/JunB double-knockout model of psoriasis. In this study, we found that S100A9 controls both the number of Th17 cells and IL-17 expression. In addition, anti–IL-17 Abs have been shown to inhibit S100A8 and S100A9 expression and enhance IL-17 expression in the DBA-1 model of psoriasis-like dermatitis (75). This suggests the presence of an IL-17–S100A8/A9 regulatory loop in psoriasis, and we therefore speculate that the anti-inflammatory role of S100A9 in the imiquimod model of psoriasis could result from its inhibitory effect on IL-17–induced inflammation.
In conclusion, this study, to our knowledge, uncovers some new and unexpected anti-inflammatory functions for the extracellular damage-associated molecular patterns S100A8 and S100A9 in the context of psoriatic skin inflammation in a mouse model of the disease. When combined with previous findings, our results therefore suggest that these proteins can exert both pro- and anti-inflammatory activities, depending on the disease model and the affected tissues.
Disclosures
The authors have no financial conflicts of interest.
Acknowledgments
We thank the technicians from the Université Laval animal facilities and histology services for technical assistance.
Footnotes
This work was supported by Canadian Institutes of Health Research, Gouvernement du Canada grant to P.A.T. and F.A. (136975).
The online version of this article contains supplemental material.
Abbreviations used in this article:
- DC
- dendritic cell
- K10
- cytokeratin 10
- LN
- lymph node
- MHCII
- MHC class II
- moDC
- monocyte-derived DC
- PASI
- Psoriasis Area Severity Index
- WT
- wild-type.
- Received January 24, 2020.
- Accepted November 26, 2020.
- Copyright © 2021 by The American Association of Immunologists, Inc.