Advertisement

Suppressive effects of antimycotics on thymic stromal lymphopoietin production in human keratinocytes

      Abstract

      Background

      Thymic stromal lymphopoietin (TSLP) is produced by epidermal keratinocytes, and it induces Th2-mediated inflammation. TSLP expression is enhanced in lesions with atopic dermatitis, and is a therapeutic target. Antimycotic agents improve the symptoms of atopic dermatitis.

      Objective

      The objective of this study was to examine whether antimycotics suppress TSLP expression in human keratinocytes.

      Methods

      Normal human keratinocytes were incubated with polyinosinic–polycytidylic acid (poly I:C) plus IL-4 in the presence of antimycotics. TSLP expression was analyzed by ELISA and real time PCR. Luciferase assays were performed to analyze NF-κB activity. IκBα degradation was analyzed by Western blot analysis.

      Results

      Poly I:C plus IL-4 increased the secretion and mRNA levels of TSLP, which was suppressed by an NF-κB inhibitor, and also enhanced NF-κB transcriptional activities and induced the degradation of IκBα in keratinocytes. The antimycotics itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine suppressed the secretion and mRNA expression of TSLP, NF-κB activity, and IκBα degradation induced by poly I:C plus IL-4. These suppressive effects were similarly manifested by 15-deoxy-Δ-12,14-PGJ2 (15d-PGJ2), a prostaglandin D2 metabolite. Antimycotics increased the release of 15d-PGJ2 from keratinocytes and decreased the release of thromboxane B2, a thromboxane A2 metabolite. Antimycotic-induced suppression of TSLP production and NF-κB activity was counteracted by an inhibitor of lipocalin type-prostaglandin D synthase.

      Conclusions

      Antimycotics itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine may suppress poly I:C plus IL-4-induced production of TSLP by inhibiting NF-κB via increasing 15d-PGJ2 production in keratinocytes. These antimycotics may block the overexpression of TSLP in lesions with atopic dermatitis.

      Abbreviations:

      AD (atopic dermatitis), ANOVA (analysis of variance), CRTH2 (chemoattractant receptor-homologous molecule expressed on Th2 cells), DMSO (dimethylsulfoxide), DP (D prostanoid receptor), 15d-PGJ2 (15-deoxy-Δ-12,14-prostaglandin J2), GAPDH (glyceraldehyde-3-phosphate dehydrogenase), HPGDS (hematopoietic prostaglandin D synthase), IκB (inhibitory κB), IKK (IκB kinase), KBM (keratinocyte basal medium), LPGDS (lipocalin-type prostaglandin D synthase), NEMO (NF-κB essential modifier), PG (prostaglandin), poly I:C (polyinosinic–polycytidylic acid), PPARγ (peroxisome proliferator-activated receptor γ), RIP1 (receptor interacting protein 1), TAK1 (TGF-β activated kinase 1), TNFR (TNF receptor), TRADD (TNF receptor-associated death domain), TRAF (TNF receptor-associated factor), TRIF (Toll/IL-1 receptor domain-containing adaptor inducing IFN-β), TSLP (thymic stromal lymphopoietin), TXA2 (thromboxane A2), TXAS (thromboxane A2 synthase)

      Keywords

      1. Introduction

      A cytokine thymic stromal lymphopoietin (TSLP) activates dendritic cells to induce proallergic Th2 cell responses [
      • Ziegler S.F.
      • Liu Y.J.
      Thymic stromal lymphopoietin in normal and pathogenic T cell development and function.
      ]. TSLP activates dendritic cells to produce Th2-attracting chemokines CCL17/CCL22 and to express OX40 ligand, a key Th2 costimulatory factor [
      • Soumelis V.
      • Reche P.A.
      • Kanzler H.
      • Yuan W.
      • Edward G.
      • Homey B.
      • et al.
      Human epithelial cells trigger dendritic cell mediated allergic inflammation by producing TSLP.
      ]. TSLP-activated dendritic cells prime naïve CD4+ T cells to differentiate into inflammatory Th2 cells producing IL-4, IL-5, IL-13, and TNF-α through OX40 ligand [
      • Soumelis V.
      • Reche P.A.
      • Kanzler H.
      • Yuan W.
      • Edward G.
      • Homey B.
      • et al.
      Human epithelial cells trigger dendritic cell mediated allergic inflammation by producing TSLP.
      ,
      • Wang Y.H.
      • Ito T.
      • Wang Y.H.
      • Homey B.
      • Watanabe N.
      • Martin R.
      • et al.
      Maintenance and polarization of human TH2 central memory T cells by thymic stromal lymphopoietin-activated dendritic cells.
      ]. In skin lesions with atopic dermatitis (AD), the expression of TSLP in epidermal keratinocytes is enhanced [
      • Soumelis V.
      • Reche P.A.
      • Kanzler H.
      • Yuan W.
      • Edward G.
      • Homey B.
      • et al.
      Human epithelial cells trigger dendritic cell mediated allergic inflammation by producing TSLP.
      ], which may trigger Th2-mediated inflammation in the skin. Compared with healthy subjects, TSLP levels in the stratum corneum, reflecting their expression levels in keratinocytes, are increased in AD patients, and correlate with severity scoring of AD and dry skin score. The stratum corneum TSLP levels as well as severity scoring of AD are reduced by topical application of moisturizer [
      • Sano Y.
      • Masuda K.
      • Tamagawa-Mineoka R.
      • Matsunaka H.
      • Murakami Y.
      • Yamashita R.
      • et al.
      Thymic stromal lymphopoietin expression is increased in the horny layer of patients with atopic dermatitis.
      ]. It is known that TSLP production is induced by the stimulation of Toll-like receptors (TLRs), such as TLR3 or TLR2, or several cytokines like TNF-α, IL-1, IFN-α, TGF-α, or IL-4 [
      • Vu A.T.
      • Chen X.
      • Xie Y.
      • Kamijo S.
      • Ushio H.
      • Kawasaki J.
      • et al.
      Extracellular double-stranded RNA induces TSLP via an endosomal acidification- and NF-κB-dependent pathway in human keratinocytes.
      ,
      • Vu A.T.
      • Baba T.
      • Chen X.
      • Le T.A.
      • Kinoshita H.
      • Xie Y.
      • et al.
      Staphylococcus aureus membrane and diacylated lipopeptide induce thymic stromal lymphopoietin in keratinocytes through the Toll-like receptor 2-Toll-like receptor 6 pathway.
      ,
      • Takai T.
      TSLP expression: cellular sources, triggers, and regulatory mechanisms.
      ], and that the activation of NF-κB is required for the TSLP production [
      • Vu A.T.
      • Chen X.
      • Xie Y.
      • Kamijo S.
      • Ushio H.
      • Kawasaki J.
      • et al.
      Extracellular double-stranded RNA induces TSLP via an endosomal acidification- and NF-κB-dependent pathway in human keratinocytes.
      ].
      Antimycotic agents, azole or non-azole, are reported to be therapeutically effective in treating patients with AD [
      • Ikezawa Z.
      • Kondo M.
      • Okajima M.
      • Nishimura Y.
      • Kono M.
      Clinical usefulness of oral itraconazole, an antimycotic drug, for refractory atopic dermatitis.
      ,
      • Nikkels A.F.
      • Piérard G.E.
      Framing the future of antifungals in atopic dermatitis.
      ]. One possible mechanism is that antimycotics suppress the colonization of fungi like Malassezia, acting as allergens for AD patients [
      • Darabi K.
      • Hostetler S.G.
      • Bechtel M.A.
      • Zirwas M.
      The role of Malassezia in atopic dermatitis affecting the head and neck of adults.
      ], by exerting fungicidal or fungistatic effects [
      • Nikkels A.F.
      • Piérard G.E.
      Framing the future of antifungals in atopic dermatitis.
      ]. The effectiveness of antimycotics is also reported for patients with seborrheic dermatitis whose pathogenesis may involve Malassezia [
      • Squire R.A.
      • Goode K.
      A randomised, single-blind, single-centre clinical trial to evaluate comparative clinical efficacy of shampoos containing ciclopirox olamine (1.5%) and salicylic acid (3%), or ketoconazole (2%, Nizoral) for the treatment of dandruff/seborrhoeic dermatitis.
      ]. Another proposed mechanism is that antimycotics may directly act on keratinocytes or lymphocytes of AD patients and manifest anti-inflammatory effects [
      • Kanda N.
      • Watanabe S.
      Suppressive effects of antimycotics on tumor necrosis factor-α-induced CCL27, CCL2, and CCL5 production in human keratinocytes.
      ,
      • Kanda N.
      • Watanabe S.
      Ketoconazole suppresses interleukin-4 plus anti-CD40-induced IgE class switching in surface IgE negative B cells from patients with atopic dermatitis.
      ]. We have found that an antimycotic ketoconazole reduces IL-4 and IL-13 production in human T cells [
      • Kanda N.
      • Watanabe S.
      Ketoconazole suppresses interleukin-4 plus anti-CD40-induced IgE class switching in surface IgE negative B cells from patients with atopic dermatitis.
      ], and that ketoconazole and another antimycotic terbinafine suppress CCL27 production in human keratinocytes [
      • Kanda N.
      • Watanabe S.
      Suppressive effects of antimycotics on tumor necrosis factor-α-induced CCL27, CCL2, and CCL5 production in human keratinocytes.
      ]. However, the effects of antimycotics on TSLP production in keratinocytes are unknown.
      We previously found that the antimycotics itraconazole and terbinafine increase the secretion of prostaglandin D2 (PGD2) from human keratinocytes [
      • Kanda N.
      • Kano R.
      • Ishikawa T.
      • Watanabe S.
      The antimycotic drugs itraconazole and terbinafine hydrochloride induce the production of human β-defensin-3 in human keratinocytes.
      ]. PGD2 is intracellularly converted from PGH2 by lipocalin-type PGD synthase (LPGDS) or hematopoietic PGD synthase (HPGDS) [
      • Scher J.U.
      • Pillinger M.H.
      15d-PGJ2: the anti-inflammatory prostaglandin?.
      ,
      • Herlong J.L.
      • Scott T.R.
      Positioning prostanoids of the D and J series in the immunopathogenic scheme.
      ]. PGH2 is also converted to thromboxane A2 (TXA2) by thromboxane A2 synthase (TXAS) [
      • Kanda N.
      • Kano R.
      • Ishikawa T.
      • Watanabe S.
      The antimycotic drugs itraconazole and terbinafine hydrochloride induce the production of human β-defensin-3 in human keratinocytes.
      ]. Itraconazole and terbinafine appear to suppress the activity of TXAS, and thus, redirect the conversion of PGH2 from TXA2 to PGD2 [
      • Kanda N.
      • Kano R.
      • Ishikawa T.
      • Watanabe S.
      The antimycotic drugs itraconazole and terbinafine hydrochloride induce the production of human β-defensin-3 in human keratinocytes.
      ]. PGD2 is intracellularly dehydrated into 15-deoxy-Δ-12,14-PGJ2 (15d-PGJ2) non-enzymatically [
      • Scher J.U.
      • Pillinger M.H.
      15d-PGJ2: the anti-inflammatory prostaglandin?.
      ,
      • Herlong J.L.
      • Scott T.R.
      Positioning prostanoids of the D and J series in the immunopathogenic scheme.
      ]. 15d-PGJ2 is known to suppress the activity of NF-κB and NF-κB-dependent gene expression, dependent or independent of intracellular receptor peroxisome proliferator-activated receptor γ (PPARγ) [
      • Scher J.U.
      • Pillinger M.H.
      15d-PGJ2: the anti-inflammatory prostaglandin?.
      ,
      • Straus D.S.
      • Pascual G.
      • Li M.
      • Welch J.S.
      • Ricote M.
      • Hsiang C.H.
      • et al.
      15-Deoxy-delta 12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.
      ]. 15d-PGJ2 binds PPARγ, and the liganded PPARγ (i) induces the expression of NF-κB inhibitory subunit, inhibitory κB (IκB) or (ii) competes with NF-κB for the transcriptional coactivators like p300/CBP [
      • Scher J.U.
      • Pillinger M.H.
      15d-PGJ2: the anti-inflammatory prostaglandin?.
      ]. Independent of PPARγ, 15d-PGJ2 covalently binds to IκB kinase (IKK) and inhibits its function to phosphorylate and inactivate IκB [
      • Straus D.S.
      • Pascual G.
      • Li M.
      • Welch J.S.
      • Ricote M.
      • Hsiang C.H.
      • et al.
      15-Deoxy-delta 12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.
      ].
      In this study, we aimed to examine if antimycotics suppress TSLP production in human keratinocytes and to elucidate the mechanism for the effects related to the induction of 15d-PGJ2. The antimycotics used were itraconazole, ketoconazole, fluconazole, voriconazole and luliconazole (azole class), terbinafine (allylamine class), butenafine (benzylamine class), and amorolfine (morpholine class) (Supplementary Fig. S1) [
      • Koga H.
      • Nanjoh Y.
      • Makimura K.
      • Tsuboi R.
      In vitro antifungal activities of luliconazole, a new topical imidazole.
      ]. The azole class is potent against both Trichophyton spp. and Candida albicans and inhibits C14α-lanosterol demethylase in the biosynthetic pathway of ergosterol, an essential component of fungal cell membranes. In contrast, the allylamine (terbinafine) and benzylamine (butenafine) classes inhibit squalene epoxidase, an early step in the ergosterol biosynthesis, show stronger anti-fungal activity against Trichophyton spp., but less effective against C. albicans. On the other hand, the morpholine class (amorolfine) inhibits both C14-reductase and C7–C8 isomerase, later stages in the ergosterol biosynthesis, shows comparable potency against Trichophyton spp., and C. albicans. The azole class also shows potent antifungal activity against Malassezia spp. [
      • Koga H.
      • Nanjoh Y.
      • Makimura K.
      • Tsuboi R.
      In vitro antifungal activities of luliconazole, a new topical imidazole.
      ]. Among azole antimycotics, itraconazole and ketoconazole, fluconazole and voriconazole are structurally alike, respectively, while luliconazole is distinct; among non-azole antimycotics, terbinafine and butenafine are structurally similar while amorolfine is dissimilar (Supplementary Fig. S1). The differential effects of these drugs on TSLP production were compared.

      2. Materials and methods

      2.1 Reagents

      Helenalin was purchased from Calbiochem (La Jolla, CA, USA). Recombinant human IL-4, TNF-α, IFN-α, and TGF-α were purchased from R&D Systems (Minneapolis, MN, USA). 15d-PGJ2, PGE2, PGF, rosiglitazone, BW245C, 13,14-dihydro-15-keto-PGD2, AT-56, and HQL-79 were purchased from Cayman Chemical (Ann Arbor, MI, USA). Pam3Cys-Ser-(Lys)4·3HCl was purchased from IMGENEX (San Diego, CA, USA). β-Glucan and zymosan from Baker's yeast (Saccharomyces cerevisiae) and polyinosinic–polycytidylic acid (poly I:C) potassium salt were purchased from Sigma–Aldrich (St. Louis, MO, USA). Mite extract from Dermatophagoides pteronyssinus was purchased from Cosmo Bio (Tokyo, Japan). Lysophosphatidic acid sodium salt was from Enzo Life Sciences (Farmingdale, NY, USA). Itraconazole, ketoconazole, fluconazole, voriconazole, and terbinafine hydrochloride were from Wako Pure Chemical (Osaka, Japan). Luliconazole, butenafine hydrochloride, and amorolfine hydrochloride were donated by POLA PHARMA (Tokyo, Japan), Kaken Pharmaceutical Co. (Tokyo, Japan), and Kyorin Pharmaceutical Co. (Tokyo, Japan), respectively. The antimycotics were dissolved in dimethylsulfoxide (DMSO) at 100 mM to create solutions and subsequently diluted in the experimental media to yield the final concentrations. The DMSO concentration as a vehicle control was 0.1% (v/v).

      2.2 Keratinocyte culture

      Human neonatal foreskin keratinocytes were purchased from Clonetics (Walkersville, MD, USA). The keratinocytes were cultured in serum-free keratinocyte growth medium (Clonetics) containing keratinocyte basal medium (KBM) supplemented with 0.5 μg/ml hydrocortisone, 5 ng/ml epidermal growth factor, 5 μg/ml insulin, and 0.5% bovine pituitary extract. The cells in the third passage were used. Each experiment was performed four times using the same lot of keratinocytes from a single donor.

      2.3 ELISA

      Keratinocytes (5 × 104 cells/well) were seeded in triplicate in 24-well plates in 0.4 ml keratinocyte growth medium, and allowed to adhere overnight. The cells were washed and incubated with supplement-free KBM for 24 h. The cells were washed and preincubated with the vehicle (DMSO) or the indicated concentrations of antimycotics or prostanoids, or 1 μM helenalin for 30 min, and subsequently incubated with IL-4 (10 ng/ml) and/or poly I:C (10 μg/ml), TNF-α (1, 10, or 100 ng/ml), IL-1β (1, 10, or 100 ng/ml), TGF-α (1, 10, or 100 ng/ml), IFN-α (1, 10, or 100 ng/ml), Pam3Cys-Ser-(Lys)4, β-glucan, zymosan, or lysophosphatidic acid (each 1, 10 or 100 μg/ml) in KBM for 48 h. The culture supernatants were analyzed by performing ELISA for TSLP (R&D Systems), 15d-PGJ2 (Abnova, Taipei, Taiwan), PGD2, TXB2, PGE2, or PGF (Cayman Chemical), according to the manufacturers’ instructions. In some experiments, the cells were pretreated with HQL-79 or AT-56 (each 10 μM) for 30 min before the addition of antimycotics.

      2.4 Real-time PCR

      The keratinocytes were incubated for 24 h as described above, and the total cellular RNA was extracted using RNeasy Mini kits (SABiosciences, Frederick, MD, USA). cDNA was synthesized using Superscript III First Strand synthesis kits (Invitrogen, Carlsbad, CA, USA). TSLP mRNA levels were quantified by performing TaqMan gene expression assay, using a probe set specific for the long form of the TSLP transcript (Hs01572933 m1; Applied Biosystems, Warrington, UK) [
      • Xie Y.
      • Takai T.
      • Chen X.
      • Okumura K.
      • Ogawa H.
      Long TSLP transcript expression and release of TSLP induced by TLR ligands and cytokines in human keratinocytes.
      ]. CCL5 and CXCL8 mRNA levels were quantified by performing SYBR Green I® gene expression assay (Takara Bio Inc., Tokyo, Japan). The mRNA levels of TSLP, CCL5 and CXCL8, normalized to those of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) are represented as the fold change relative to the control keratinocytes treated with KBM alone.

      2.5 Plasmids and transfection

      NF-κB-responsive firefly luciferase vector and constitutive Renilla luciferase vector (SABiosciences) were mixed with Fugene HD (Roche, Indianapolis, IN, USA) and added to the keratinocytes (3.0 × 104 cells/well) in the 24-well plates. After 24 h, the cells were washed and incubated with KBM for 24 h, and preincubated with the vehicle (DMSO) or the indicated concentrations of antimycotics or prostanoids, or 1 μM helenalin for 30 min and subsequently incubated with 10 ng/ml IL-4 and/or 10 μg/ml poly I:C. After 18 h, the firefly and Renilla luciferase activities in the cell extracts were quantified using the Dual-Luciferase Assay System (Promega, Madison, WI, USA). The transcriptional activities of NF-κB are expressed as the ratio of firefly and Renilla luciferase activities.

      2.6 Western blotting

      Western blotting was performed by a previously described method [
      • Kanda N.
      • Kano R.
      • Ishikawa T.
      • Watanabe S.
      The antimycotic drugs itraconazole and terbinafine hydrochloride induce the production of human β-defensin-3 in human keratinocytes.
      ]. In brief, keratinocytes were preincubated with the vehicle (DMSO) or 10 μM antimycotics or prostanoids for 30 min and incubated with 10 ng/ml IL-4 and/or 10 μg/ml poly I:C for 15 min, and lysed in a lysis buffer; the resultant lysates were separated by performing SDS-PAGE, and transferred to polyvinylidene difluoride membranes. The membranes were blocked and probed using the primary antibodies anti-IκBα and anti-β-actin (Cell Signaling Technology Inc., Boston, MA, USA). Appropriate secondary antibodies conjugated to horseradish peroxidase (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were added, and the immunoreactivities were visualized using LumiGLO® (Cell Signaling Technology), according to the manufacturer's instructions. Protein bands were analyzed by performing densitometry, and the levels of IκBα normalized against those of β-actin are represented as fold changes.

      2.7 Statistical analyses

      All the statistical analyses were performed using Ekuseru-Toukei 2012 (Social Survey Research Information Co. Ltd., Tokyo, Japan). One-way analysis of variance (ANOVA) with Dunnett's test was used for analyzing the data represented in Fig. 1, Fig. 2, Fig. 3, and Supplementary Figs. S2 and S4, and one-way ANOVA with Scheffe's test was used for analyzing the data represented in Fig. 1, Fig. 2, Fig. 4, Fig. 5, Fig. 6 and Supplementary Fig. S3. A P value of <0.05 was considered significant.
      Figure thumbnail gr1
      Fig. 1Antimycotics suppress poly I:C plus IL-4-induced TSLP secretion and mRNA expression. (A and C) Keratinocytes were preincubated with the vehicle (DMSO) or 1 μM helenalin for 30 min, and incubated with 10 μg/ml poly I:C and/or 10 ng/ml IL-4. TSLP secretion or mRNA levels were measured at 48 h or at 24 h, respectively. *P < 0.05, by one-way ANOVA with Scheffe's test. Means ± SD (n = 4) are shown. (B and D) Keratinocytes were preincubated with the vehicle (DMSO) or the indicated concentrations of antimycotics for 30 min, and incubated with poly I:C and/or IL-4. TSLP secretion or mRNA levels were measured at 48 h or at 24 h, respectively. *P < 0.05 vs. poly I:C plus IL-4, by one-way ANOVA with Dunnett's test. Means ± SD (n = 4) are shown. pIC, poly I:C; ITC, itraconazole; KTC, ketoconazole; LLC, luliconazole; TRB, terbinafine; BTN, butenafine; AMR, amorolfine.
      Figure thumbnail gr2
      Fig. 2Antimycotics suppress NF-κB activity and IκBα degradation induced by poly I:C plus IL-4. (A) Keratinocytes transfected with NF-κB-responsive firefly luciferase vector and Renilla luciferase vector were preincubated with the vehicle (DMSO) or 1 μM helenalin for 30 min, and incubated with 10 μg/ml poly I:C and/or 10 ng/ml IL-4. At 18 h, the transcriptional activities were analyzed. The transcriptional activities of NF-κB are expressed as the ratio of firefly and Renilla luciferase activities. *P < 0.05, by one-way ANOVA with Scheffe's test. (B) The transfected keratinocytes were preincubated with the vehicle (DMSO) or the indicated concentrations of antimycotics for 30 min, and incubated with poly I:C and/or IL-4. At 18 h, the transcriptional activities were analyzed. *P < 0.05 vs. poly I:C plus IL-4, by one-way ANOVA with Dunnett's test. Means ± SD (n = 4) are shown. (C) The untransfected keratinocytes were preincubated with the vehicle (DMSO) or 10 μM of antimycotics for 30 min, and incubated with 10 μg/ml poly I:C and/or 10 ng/ml IL-4. At 15 min, the levels of IκBα were analyzed by Western blotting. The IκBα levels normalized against those of β-actin are represented as the fold change. Results represent 4 separate experiments. pIC, poly I:C; ITC, itraconazole; KTC, ketoconazole; FLC, fluconazole; VRC, voriconazole; LLC, luliconazole; TRB, terbinafine; BTN, butenafine; AMR, amorolfine.
      Figure thumbnail gr3
      Fig. 315d-PGJ2 suppresses TSLP secretion (A) and mRNA expression (B), NF-κB activity (C) and IκBα degradation (D) induced by poly I:C plus IL-4. (A–C) The keratinocytes, untransfected or transfected with NF-κB-responsive firefly luciferase vector and Renilla luciferase vector, were preincubated with the vehicle (DMSO) or indicated concentrations of BW245C, 13,14-dihydro-15-keto-PGD2 (DK-PGD2), rosiglitazone (RSG) or 15d-PGJ2 for 30 min, and incubated with 10 μg/ml poly I:C (pIC) and/or 10 ng/ml IL-4. The TSLP secretion, mRNA levels, or NF-κB activities were analyzed at 48 h, 24 h, or 18 h, respectively. *P < 0.05 vs. poly I:C plus IL-4, by one-way ANOVA with Dunnett's test. Means ± SD (n = 4) are shown. (D) The keratinocytes were preincubated with the vehicle (DMSO) or 10 μM of BW245C (BW), 13,14-dihydro-15-keto-PGD2 (DK), rosiglitazone (RSG), PGE2, PGF, or 15d-PGJ2 for 30 min, and incubated with 10 μg/ml poly I:C (pIC) and/or 10 ng/ml IL-4 for 15 min. The IκBα levels normalized against those of β-actin are represented as the fold change. Results represent 4 separate experiments.

      3. Results

      3.1 Antimycotics itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine suppress poly I:C plus IL-4-induced TSLP production in keratinocytes

      TLR3 agonist poly I:C alone moderately increased TSLP secretion; the secretion was further enhanced by IL-4, although IL-4 alone was ineffective (Fig. 1A). TNF-α, IL-1β, TGF-α, IFN-α, Pam3Cys-Ser-(Lys)4, mite extract, lysophosphatidic acid, β-glucan or zymosan did not effectively increase TSLP secretion either in the presence or absence of IL-4 (data not shown). Thus poly I:C plus IL-4 were used as the TSLP-inducing stimuli in further examination. The poly I:C plus IL-4-induced TSLP secretion was suppressed by the NF-κB inhibitor helenalin (Fig. 1A), indicating NF-κB-dependent production. The poly I:C plus IL-4-induced TSLP secretion was suppressed by the azole antimycotics itraconazole, ketoconazole, luliconazole, and by the non-azole antimycotics terbinafine, butenafine, and amorolfine (Fig. 1B) in a dose-dependent manner, while other azole antimycotics (fluconazole and voriconazole) were ineffective (Supplementary Fig. S2A). These antimycotics did not reduce the cell viability at less than 100 μM (>95% of controls) as examined by a trypan blue dye exclusion test.
      Poly I:C alone increased the TSLP mRNA level, and the level was further enhanced by IL-4, but IL-4 alone was ineffective (Fig. 1C). The increase in TSLP mRNA level was suppressed by helenalin (Fig. 1C), and also by antimycotics except for fluconazole or voriconazole (Fig. 1D and Supplementary Fig. S2B) in manners similar to those observed with secretion, indicating transcriptional repression by these antimycotics. Poly I:C also increased mRNA levels of chemokines, CCL5 and CXCL8, and the levels were further enhanced by IL-4, though IL-4 alone was ineffective (Supplementary Fig. S3A and B). The increases in CCL5 and CXCL8 mRNA levels were suppressed by helenalin, indicating NF-κB-dependent expression, and were suppressed by itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine but not by fluconazole or voriconazole (Supplementary Fig. S3A and B). Thus antimycotics suppressed NF-κB-dependent CCL5 and CXCL8 expression induced by poly I:C plus IL-4 in manners similar to that in TSLP.

      3.2 Antimycotics itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine suppress the activation of NF-κB induced by poly I:C plus IL-4

      We then examined if these antimycotics may suppress NF-κB activities that are possibly required for TSLP production. Poly I:C alone increased the transcriptional activity of NF-κB, and the activity was further enhanced by IL-4, although IL-4 alone was ineffective (Fig. 2A). The poly I:C plus IL-4-induced NF-κB activity was suppressed by helenalin (Fig. 2A), and was reduced in a dose-dependent manner by itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine (Fig. 2B), while NF-κB activity was not altered by fluconazole or voriconazole (Supplementary Fig. S2C). These results suggest that antimycotics suppress NF-κB activity induced by poly I:C plus IL-4 in parallel with the suppression of TSLP production, indicating that the suppression of TSLP production may be mediated by that of NF-κB activity.
      The NF-κB p50/p65 complex is constitutively combined with IκBα and sequestered in the cytosol. Certain stimuli activate IKK, which phosphorylates IκBα and induces its degradation, allowing the release and nuclear translocation of the active NF-κB p50/p65 complex [
      • Scher J.U.
      • Pillinger M.H.
      15d-PGJ2: the anti-inflammatory prostaglandin?.
      ]. We thus examined if antimycotics suppress the degradation of IκBα. Poly I:C plus IL-4 reduced the amount of IκBα in the keratinocytes (Fig. 2C), indicating the degradation of IκBα via phosphorylation by IKK. The poly I:C plus IL-4-induced degradation of IκBα was suppressed by itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine, while fluconazole or voriconazole were ineffective (Fig. 2C). Our results indicate that these antimycotics may suppress the poly I:C plus IL-4-induced activation of IKK. When the keratinocytes pretreated with antimycotics were washed before the addition of poly I:C plus IL-4, the suppression of IκBα degradation by antimycotics was still revealed (data not shown), indicating that the extracellular physical interaction of antimycotics with poly I:C may not be essential if any.

      3.3 15d-PGJ2 suppresses the poly I:C plus IL-4-induced NF-κB activity and TSLP production

      It is known that cyclopentenone prostanoid 15d-PGJ2 suppresses NF-κB activity and NF-κB-dependent gene expression [
      • Scher J.U.
      • Pillinger M.H.
      15d-PGJ2: the anti-inflammatory prostaglandin?.
      ,
      • Straus D.S.
      • Pascual G.
      • Li M.
      • Welch J.S.
      • Ricote M.
      • Hsiang C.H.
      • et al.
      15-Deoxy-delta 12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.
      ]. Exogenous 15d-PGJ2 (i) binds and stimulates cell surface PGD2 receptors, D prostanoid receptor (DP) or chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2), or (ii) enters cells and binds to the intracellular receptor PPARγ, or (iii) covalently interacts with IKK independent of these receptors [
      • Scher J.U.
      • Pillinger M.H.
      15d-PGJ2: the anti-inflammatory prostaglandin?.
      ]. We thus examined if 15d-PGJ2 suppresses the poly I:C plus IL-4-induced NF-κB activity and TSLP production dependent on DP, CRTH2, or PPARγ. 15d-PGJ2 was observed to suppress TSLP secretion (Fig. 3A) and mRNA expression (Fig. 3B), and NF-κB activity (Fig. 3C) and IκBα degradation (Fig. 3D) induced by poly I:C plus IL-4. In parallel, 15d-PGJ2 suppressed CCL5 and CXCL8 mRNA expression induced by poly I:C plus IL-4 (Supplementary Fig. S3A and B). In contrast, the poly I:C plus IL-4-induced TSLP secretion and mRNA expression, NF-κB activity, and IκBα degradation were not suppressed by DP, CRTH2, or PPARγ agonists, BW245C, 13,14-dihydro-15-keto-PGD2, or rosiglitazone, respectively (Fig. 3A–D), or by the other prostanoids PGE2 or PGF (Supplementary Fig. S4A–C and Fig. 3D). These results indicate that DP, CRTH2, and PPARγ are not involved in the 15d-PGJ2-induced suppression of TSLP production and NF-κB activity and that the other prostanoids, PGE2 or PGF may not suppress the poly I:C plus IL-4-induced TSLP production or NF-κB activity.

      3.4 Antimycotics itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine increase the release of 15d-PGJ2 and its precursor PGD2 and decrease the release of TXB2

      Since itraconazole and terbinafine increase the release of PGD2, a precursor of 15d-PGJ2 from keratinocytes [
      • Kanda N.
      • Kano R.
      • Ishikawa T.
      • Watanabe S.
      The antimycotic drugs itraconazole and terbinafine hydrochloride induce the production of human β-defensin-3 in human keratinocytes.
      ], we examined if antimycotics also induce the release of 15d-PGJ2 from keratinocytes. Poly I:C plus IL-4 reduced the release of 15d-PGJ2 (Fig. 4A), and the release of 15d-PGJ2 was increased by the addition of itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine, while the release was not altered by fluconazole or voriconazole (Fig. 4A). In parallel with 15d-PGJ2, the release of its precursor PGD2 was reduced by poly I:C plus IL-4 (Fig. 4B), and the release of PGD2 was increased by the addition of itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine, while fluconazole and voriconazole were ineffective (Fig. 4B). In contrast, poly I:C plus IL-4 increased the release of TXB2, a metabolite of TXA2 (Fig. 4C), and the release of TXB2 was reduced by itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine, while fluconazole and voriconazole were ineffective (Fig. 4C). We also analyzed if antimycotics may alter the release of the other prostanoids, PGE2 and PGF in the presence of poly I:C plus IL-4. Poly I:C or IL-4, either alone or together increased the release of PGE2, and the poly I:C plus IL-4-induced release of PGE2 was not significantly altered by any of the antimycotics (Fig. 4D). IL-4 alone or together with poly I:C increased the release of PGF, and the release of PGF was further increased by itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine, while fluconazole and voriconazole were ineffective (Fig. 4E). These results suggest that antimycotics may enhance the release of 15d-PGJ2, PGD2, and PGF while reducing the release of TXB2 in poly I:C plus IL-4-stimulated keratinocytes.
      Figure thumbnail gr4
      Fig. 4Effects of antimycotics on the release of 15d-PGJ2 (A), PGD2 (B), TXB2 (C), PGE2 (D) and PGF (E) in poly I:C plus IL-4-stimulated keratinocytes. Keratinocytes were preincubated with the vehicle (DMSO) or 10 μM of antimycotics for 30 min, and incubated with 10 μg/ml poly I:C and/or 10 ng/ml IL-4 for 48 h. The culture supernatants were analyzed for prostanoids. *P < 0.05 vs. controls and P < 0.05 vs. poly I:C plus IL-4, by one-way ANOVA with Scheffe's test. Means ± SD (n = 4) are shown. pIC, poly I:C; ITC, itraconazole; KTC, ketoconazole; FLC, fluconazole; VRC, voriconazole; LLC, luliconazole; TRB, terbinafine; BTN, butenafine; AMR, amorolfine.
      We then examined if antimycotics may alter basal release of these prostanoids in the absence of poly I:C plus IL-4. Itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine increased the basal release of 15d-PGJ2, PGD2, PGE2, and PGF (Fig. 5A, B, D and E) while reduced that of TXB2 (Fig. 5C). Fluconazole or voriconazole did not alter the basal release of these prostanoids.
      Figure thumbnail gr5
      Fig. 5Effects of antimycotics on the basal release of 15d-PGJ2 (A), PGD2 (B), TXB2 (C), PGE2 (D) and PGF (E). Keratinocytes were incubated with the vehicle (DMSO) or 10 μM of antimycotics for 48 h. *P < 0.05 vs. controls by one-way ANOVA with Scheffe's test. Means ± SD (n = 4) are shown. ITC, itraconazole; KTC, ketoconazole; FLC, fluconazole; VRC, voriconazole; LLC, luliconazole; TRB, terbinafine; BTN, butenafine; AMR, amorolfine.

      3.5 15d-PGJ2 is involved in the antimycotic-induced suppression of TSLP production and NF-κB activity

      We then analyzed if 15d-PGJ2 may be involved in the antimycotic-induced suppression of TSLP production and NF-κB activity, using the LPGDS or HPGDS inhibitors, AT-56 or HQL-79, respectively. AT-56 counteracted the butenafine-induced suppression of TSLP secretion (Fig. 6A), mRNA expression (data not shown), and NF-κB activity (Fig. 6B), while HQL-79 did not. Similar results were obtained for the suppression of TSLP secretion, mRNA expression, and NF-κB activity induced by the other antimycotics, itraconazole, ketoconazole, luliconazole, terbinafine, and amorolfine (data not shown). In parallel, AT-56 suppressed the butenafine-induced release of 15d-PGJ2 and PGD2, while HQL-79 did not significantly alter the release (Fig. 6C and D), indicating the LPGDS-dependent production of 15d-PGJ2 and PGD2. Similar results were obtained for the increase in 15d-PGJ2 and PGD2 release induced by the other antimycotics, itraconazole, ketoconazole, luliconazole, terbinafine, and amorolfine (data not shown). These results clearly suggest that these antimycotics may suppress NF-κB activity and NF-κB-dependent TSLP production via the production of 15dPGJ2, dependent on LPGDS.
      Figure thumbnail gr6
      Fig. 6LPGDS inhibitor AT-56 counteracts the suppression of TSLP secretion (A) and NF-κB activity (B), and the release of 15d-PGJ2 (C) and PGD2 (D) induced by butenafine. The keratinocytes, untransfected or transfected with NF-κB-responsive firefly luciferase vector and Renilla luciferase vector, were preincubated with the vehicle (DMSO) or 10 μM AT-56 or HQL-79 for 30 min, and incubated with 10 μM butenafine (BTN) for 30 min, and then treated with 10 μg/ml poly I:C (pIC) and/or 10 ng/ml IL-4. The secretion of TSLP or prostanoids was analyzed at 48 h while NF-κB activities were analyzed at 18 h. *P < 0.05 vs. controls, P < 0.05 vs. poly I:C plus IL-4, and P < 0.05 vs. poly I:C plus IL-4 plus butenafine, by one-way ANOVA with Scheffe's test. Means ± SD (n = 4) are shown.

      4. Discussion

      The antimycotics itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine suppressed the poly I:C plus IL-4-induced NF-κB activity and NF-κB-dependent production of TSLP via increasing 15d-PGJ2 production. It is reported that the cyclopentenone ring of 15d-PGJ2 avidly reacts with intracellular signaling molecules [
      • Straus D.S.
      • Pascual G.
      • Li M.
      • Welch J.S.
      • Ricote M.
      • Hsiang C.H.
      • et al.
      15-Deoxy-delta 12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.
      ,
      • Rajakariar R.
      • Hilliard M.
      • Lawrence T.
      • Trivedi S.
      • Colville-Nash P.
      • Bellingan G.
      • et al.
      Hematopoietic prostaglandin D2 synthase controls the onset and resolution of acute inflammation through PGD2 and 15-deoxy-delta 12,14-prostaglandin J2.
      ,
      • Trivedi S.G.
      • Newson J.
      • Rajakariar R.
      • Jacques T.S.
      • Hannon R.
      • Kanaoka Y.
      • et al.
      Essential role for hematopoietic prostaglandin D2 synthase in the control of delayed type hypersensitivity.
      ], especially covalently binds to specific residues in IKKβ, and thus inhibits its ability to phosphorylate and degrade IκBα and to allow the release and nuclear entry of the active NF-κB complex p50/p65 [
      • Straus D.S.
      • Pascual G.
      • Li M.
      • Welch J.S.
      • Ricote M.
      • Hsiang C.H.
      • et al.
      15-Deoxy-delta 12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.
      ,
      • Rajakariar R.
      • Hilliard M.
      • Lawrence T.
      • Trivedi S.
      • Colville-Nash P.
      • Bellingan G.
      • et al.
      Hematopoietic prostaglandin D2 synthase controls the onset and resolution of acute inflammation through PGD2 and 15-deoxy-delta 12,14-prostaglandin J2.
      ,
      • Trivedi S.G.
      • Newson J.
      • Rajakariar R.
      • Jacques T.S.
      • Hannon R.
      • Kanaoka Y.
      • et al.
      Essential role for hematopoietic prostaglandin D2 synthase in the control of delayed type hypersensitivity.
      ]. It is reported that 15d-PGJ2 inhibits TNF-α-induced activation of IKK and resultant phosphorylation and degradation of IκBα in human synovial fibroblasts [
      • Lin T.H.
      • Tang C.H.
      • Wu K.
      • Fong Y.C.
      • Yang R.S.
      • Fu W.M.
      15-Deoxy-Δ(12,14) -prostaglandin-J2 and ciglitazone inhibit TNF-α-induced matrix metalloproteinase 13 production via the antagonism of NF-κB activation in human synovial fibroblasts.
      ]. Our present results indicate that the possible inhibition of IKK by 15d-PGJ2 may mediate the inhibition of NF-κB activity and TSLP production in poly I:C plus IL-4-stimulated keratinocytes. It is also reported that 15d-PGJ2 inhibits NF-κB p50/p65 binding to DNA via alkylation of conserved cysteine residues in the binding domain, p65/Cys38 or p50/Cys62 [
      • Straus D.S.
      • Pascual G.
      • Li M.
      • Welch J.S.
      • Ricote M.
      • Hsiang C.H.
      • et al.
      15-Deoxy-delta 12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.
      ].
      The 15d-PGJ2-induced inhibition of NF-κB may not involve cell surface DP or CRTH2 receptors but rather depend on its intracellular action at least in poly I:C plus IL-4-stimulated keratinocytes. The intracellularly produced 15d-PGJ2 may directly access IKK and thus more effectively suppress its activity compared with exogenously added 15d-PGJ2. Besides 15d-PGJ2 is unstable since it is likely to be metabolized by glutathione S-transferase [
      • Hayes J.D.
      • Flanagan J.U.
      • Jowsey I.R.
      Glutathione transferases.
      ]. Therefore, the antimycotics inducing intracellular production of 15d-PGJ2 may act as more stable and efficient inhibitors of NF-κB and TSLP production, compared with exogenous 15d-PGJ2.
      The antimycotics, itraconazole, ketoconazole, luliconazole, terbinafine, butenafine, and amorolfine decreased the release of TXB2, a TXA2 metabolite, and simultaneously increased the release of PGD2 and its metabolite 15d-PGJ2 and another prostanoid PGF in keratinocytes either in the presence or absence of poly I:C plus IL-4 (Fig. 4, Fig. 5). Though these antimycotics increased basal release of PGE2 (Fig. 5D), they did not further increase the poly I:C plus IL-4-induced release of PGE2 (Fig. 4D), possibly because the capability of keratinocytes to increase its production may be saturated. Among these antimycotics, ketoconazole is known to bind TXAS and suppress its activity [
      • Tolman E.L.
      • Fuller B.L.
      Inhibition of thromboxane synthesis in guinea pig lung and human platelets by clotrimazole and other imidazole antifungals.
      ]. The inhibition of TXAS redirects the metabolism of PGH2 from TXA2 to the other prostanoids including PGD2 [
      • Tolman E.L.
      • Fuller B.L.
      Inhibition of thromboxane synthesis in guinea pig lung and human platelets by clotrimazole and other imidazole antifungals.
      ], which is further non-enzymatically dehydrated to become 15d-PGJ2, an anti-inflammatory product [
      • Gandhi U.H.
      • Kaushal N.
      • Ravindra K.C.
      • Hegde S.
      • Nelson S.M.
      • Narayan V.
      • et al.
      Selenoprotein-dependent up-regulation of hematopoietic prostaglandin D2 synthase in macrophages is mediated through the activation of peroxisome proliferator-activated receptor (PPAR) γ.
      ]. It is thus hypothesized that the possible suppression of TXAS activity may mediate the decrease of TXB2 and coincident increase of 15d-PGJ2 by antimycotics, itraconazole, luliconazole, terbinafine, butenafine, and amorolfine as well as ketoconazole. Alternatively, some antimycotics may directly activate LPGDS. These possibilities should further be examined.
      Differently from 15d-PGJ2, PGE2 or PGF did not appear to suppress NF-κB activity or TSLP production induced by poly I:C plus IL-4 (Supplementary Fig. S4). The results indicate that these prostanoids may be dispensable for the inhibition of poly I:C plus IL-4-induced NF-κB and TSLP production by antimycotics. On the other hand, ketoconazole and terbinafine suppressed TNF-α-induced NF-κB activity and CCL27, CCL2, and CCL5 production via PGE2 [
      • Kanda N.
      • Watanabe S.
      Suppressive effects of antimycotics on tumor necrosis factor-α-induced CCL27, CCL2, and CCL5 production in human keratinocytes.
      ]. The discrepancy with the present study may be due to the difference in NF-κB-activating stimuli, and PGE2 may suppress NF-κB activation induced by TNF-α but not that by poly I:C. The ligation of TNF receptor 1 (TNFR1) with TNF-α induces the recruitment of adaptor molecules, TNFR-associated death domain (TRADD), receptor interacting protein 1 (RIP1), and TNFR-associated factor 2 (TRAF2) through poly-ubiquitination, and the TNFR1-TRADD-RIP1-TRAF2 complex further recruits and activates TGF-β-activated kinase 1 (TAK1) and NF-κB essential modifier (NEMO) complexed with IKKα/IKKβ, leading to the phosphorylation and activation of IKKβ by TAK1 [
      • Scheidereit C.
      IκB kinase complexes: gateways to NF-κB activation and transcription.
      ]. The ligation of TLR3 with poly I:C recruits adaptor protein Toll/IL-1 receptor domain-containing adaptor inducing IFN-β (TRIF), and the activated TRIF induces NF-κB-activating signaling pathways, TRAF6/TAK1/NEMO/IKKα and IKKβ or TRADD/Pellino-1/RIP1/TAK1/NEMO/IKKα and IKKβ [
      • Kawai T.
      • Akira S.
      The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors.
      ]. PGE2, induced by antimycotics, may generate cyclic AMP or Ca2+ signal via EP2 or EP3 receptors, respectively, and either one or both signals may block TNF-α-dependent but poly I:C-independent points in NF-κB-activating pathway, suppressing the former – but not the latter-induced gene expression. It is reported that cyclic AMP suppresses TRADD and RIP1 binding to TNFR1 via cyclic AMP-dependent protein kinase A in rat hepatocytes [
      • Bhattacharjee R.
      • Xiang W.
      • Wang Y.
      • Zhang X.
      • Billiar T.R.
      cAMP prevents TNF-induced apoptosis through inhibiting DISC complex formation in rat hepatocytes.
      ], and that cyclic AMP reduces the protein levels of TRAF2 in murine CD8+ CD4+ thymocytes [
      • Guevara Patiño J.A.
      • Ivanov V.N.
      • Lacy E.
      • Elkon K.B.
      • Marino M.W.
      • Nikolic-Zugić J.
      TNF-α is the critical mediator of the cyclic AMP-induced apoptosis of CD8+4+ double-positive thymocytes.
      ].
      In contrast to the other antimycotics, fluconazole and voriconazole did not alter the release of 15d-PGJ2, PGD2, TXB2, PGE2, and PGF either in the presence or absence of poly I:C plus IL-4 (Fig. 4, Fig. 5). One possible reason for the difference is that these two antimycotics may not suppress the activity of TXAS. It is reported that the inhibitor-binding site of TXAS contains a hydrophobic domain [
      • Hsu P.Y.
      • Tsai A.L.
      • Kulmacz R.J.
      • Wang L.H.
      Expression, purification, and spectroscopic characterization of human thromboxane synthase.
      ,
      • Wang L.H.
      • Tsai A.L.
      • Hsu P.Y.
      Substrate binding is the rate-limiting step in thromboxane synthase catalysis.
      ]. Since fluconazole and voriconazole have hydrophilic side chains [
      • Matsuura K.
      • Yoshioka S.
      • Tosha T.
      • Hori H.
      • Ishimori K.
      • Kitagawa T.
      • et al.
      Structural diversities of active site in clinical azole-bound forms between sterol 14alpha-demethylases (CYP51s) from human and Mycobacterium tuberculosis.
      ], the binding of fluconazole or voriconazole to TXAS may be less potent than that of other antimycotics like itraconazole or ketoconazole, containing lipophilic phenylpiperazinyl ether side chains [
      • Steel H.C.
      • Tintinger G.R.
      • Anderson R.
      Comparison of the anti-inflammatory activities of imidazole antimycotics in relation to molecular structure.
      ]. We should further examine the relationship between TXAS inhibition and the structure of antimycotics.
      Antimycotics suppress in vitro TSLP production in keratinocytes at 1–10 μM, which is close to the concentrations obtained by oral administration of antimycotics; peak serum concentrations were 14.9 μM, 0.71 μM, or 3.4 μM by oral 200 mg ketoconazole, 200 mg itraconazole, or 250 mg terbinafine, respectively [
      • Faergemann J.
      • Zehender H.
      • Jones T.
      • Maibach I.
      Terbinafine levels in serum, stratum corneum, dermis-epidermis (without stratum corneum), hair, sebum and eccrine sweat.
      ]. It is thus indicated that these antimycotics, administered in vivo, may suppress TSLP production in epidermal keratinocytes at physiological concentrations. TSLP is overexpressed in lesions with AD and initiates Th2 responses in the acute phase of AD, and is thus a key therapeutic target for this disease [
      • Vu A.T.
      • Baba T.
      • Chen X.
      • Le T.A.
      • Kinoshita H.
      • Xie Y.
      • et al.
      Staphylococcus aureus membrane and diacylated lipopeptide induce thymic stromal lymphopoietin in keratinocytes through the Toll-like receptor 2-Toll-like receptor 6 pathway.
      ,
      • Bogiatzi S.I.
      • Guillot-Delost M.
      • Cappuccio A.
      • Bichet J.C.
      • Chouchane-Mlik O.
      • Donnadieu M.H.
      • et al.
      Multiple-checkpoint inhibition of thymic stromal lymphopoietin-induced TH2 response by TH17-related cytokines.
      ]. The treatment with these antimycotics may suppress the overexpression of TSLP via increasing 15dPGJ2 in lesions with AD and thus block the priming of Th2-mediated inflammation. The suppression of TSLP production by antimycotics may thus be one possible mechanism for their therapeutic efficacy for treating AD. Future studies should determine if the topical or systemic administration of these antimycotics may suppress TSLP expression in lesions with AD. Despite the reported effectiveness of antimycotics for patients with AD of head and neck type or seborrheic dermatitis, the results denying their therapeutic efficacy are reported [
      • Wong A.W.
      • Hon E.K.
      • Zee B.
      Is topical antimycotic treatment useful as adjuvant therapy for flexural atopic dermatitis: randomized, double-blind, controlled trial using one side of the elbow or knee as a control.
      ] and overall effectiveness in large scaled study is lacking. This discrepancy is possibly due to the heterogeneity of patients’ population. The therapeutic effects of antimycotics may depend on the patients’ immunological and/or microbial phenotypes. Firstly, the cytokine profiles differ with individual patients or inflammation phases; extrinsic AD patients with high IgE levels or acute phase of AD lesions apt to show Th2-shifted phenotype while intrinsic AD patients with low IgE levels or chronic phase of AD lesions are likely to show Th1-enhanced phenotype [
      • Kabashima-Kubo R.
      • Nakamura M.
      • Sakabe J.
      • Sugita K.
      • Hino R.
      • Mori T.
      • et al.
      A group of atopic dermatitis without IgE elevation or barrier impairment shows a high Th1 frequency: possible immunological state of the intrinsic type.
      ]. Thus antimycotics suppressing TSLP production may be less effective for the latter population. Secondly, the amounts of colonized Malassezia may differ with the patients, and Malassezia may not play a causative role in patients with low levels of colonization; Malassezia globosa or Malassezia restricta induces IL-5, IL-10, IL-13 or IL-4 secretion from keratinocytes, respectively [
      • Ishibashi Y.
      • Sugita T.
      • Nishikawa A.
      Cytokine secretion profile of human keratinocytes exposed to Malassezia yeasts.
      ], and thus antimycotics suppressing Malassezia colonization may be less effective for the patients with lower levels of its colonization. Further study should elucidate the relationship between the patients’ phenotypes and effectiveness of antimycotics.

      Acknowledgments

      We thank POLA PHARMA, Kaken Pharmaceutical Co., and Kyorin Pharmaceutical Co. for the donation of luliconazole, butenafine hydrochloride, and amorolfine hydrochloride, respectively.

      References

        • Ziegler S.F.
        • Liu Y.J.
        Thymic stromal lymphopoietin in normal and pathogenic T cell development and function.
        Nat Immunol. 2006; 7: 709-714
        • Soumelis V.
        • Reche P.A.
        • Kanzler H.
        • Yuan W.
        • Edward G.
        • Homey B.
        • et al.
        Human epithelial cells trigger dendritic cell mediated allergic inflammation by producing TSLP.
        Nat Immunol. 2002; 3: 673-680
        • Wang Y.H.
        • Ito T.
        • Wang Y.H.
        • Homey B.
        • Watanabe N.
        • Martin R.
        • et al.
        Maintenance and polarization of human TH2 central memory T cells by thymic stromal lymphopoietin-activated dendritic cells.
        Immunity. 2006; 24: 827-838
        • Sano Y.
        • Masuda K.
        • Tamagawa-Mineoka R.
        • Matsunaka H.
        • Murakami Y.
        • Yamashita R.
        • et al.
        Thymic stromal lymphopoietin expression is increased in the horny layer of patients with atopic dermatitis.
        Clin Exp Immunol. 2013; 171: 330-337
        • Vu A.T.
        • Chen X.
        • Xie Y.
        • Kamijo S.
        • Ushio H.
        • Kawasaki J.
        • et al.
        Extracellular double-stranded RNA induces TSLP via an endosomal acidification- and NF-κB-dependent pathway in human keratinocytes.
        J Invest Dermatol. 2011; 131: 2205-2212
        • Vu A.T.
        • Baba T.
        • Chen X.
        • Le T.A.
        • Kinoshita H.
        • Xie Y.
        • et al.
        Staphylococcus aureus membrane and diacylated lipopeptide induce thymic stromal lymphopoietin in keratinocytes through the Toll-like receptor 2-Toll-like receptor 6 pathway.
        J Allergy Clin Immunol. 2010; 126: 985-993
        • Takai T.
        TSLP expression: cellular sources, triggers, and regulatory mechanisms.
        Allergol Int. 2012; 61: 3-17
        • Ikezawa Z.
        • Kondo M.
        • Okajima M.
        • Nishimura Y.
        • Kono M.
        Clinical usefulness of oral itraconazole, an antimycotic drug, for refractory atopic dermatitis.
        Eur J Dermatol. 2004; 14: 400-406
        • Nikkels A.F.
        • Piérard G.E.
        Framing the future of antifungals in atopic dermatitis.
        Dermatology. 2003; 206: 398-400
        • Darabi K.
        • Hostetler S.G.
        • Bechtel M.A.
        • Zirwas M.
        The role of Malassezia in atopic dermatitis affecting the head and neck of adults.
        J Am Acad Dermatol. 2009; 60: 125-136
        • Squire R.A.
        • Goode K.
        A randomised, single-blind, single-centre clinical trial to evaluate comparative clinical efficacy of shampoos containing ciclopirox olamine (1.5%) and salicylic acid (3%), or ketoconazole (2%, Nizoral) for the treatment of dandruff/seborrhoeic dermatitis.
        J Dermatol Treat. 2002; 13: 51-60
        • Kanda N.
        • Watanabe S.
        Suppressive effects of antimycotics on tumor necrosis factor-α-induced CCL27, CCL2, and CCL5 production in human keratinocytes.
        Biochem Pharmacol. 2006; 72: 463-473
        • Kanda N.
        • Watanabe S.
        Ketoconazole suppresses interleukin-4 plus anti-CD40-induced IgE class switching in surface IgE negative B cells from patients with atopic dermatitis.
        J Invest Dermatol. 2002; 119: 590-599
        • Kanda N.
        • Kano R.
        • Ishikawa T.
        • Watanabe S.
        The antimycotic drugs itraconazole and terbinafine hydrochloride induce the production of human β-defensin-3 in human keratinocytes.
        Immunobiology. 2011; 216: 497-504
        • Scher J.U.
        • Pillinger M.H.
        15d-PGJ2: the anti-inflammatory prostaglandin?.
        Clin Immunol. 2005; 114: 100-109
        • Herlong J.L.
        • Scott T.R.
        Positioning prostanoids of the D and J series in the immunopathogenic scheme.
        Immunol Lett. 2006; 102: 121-131
        • Straus D.S.
        • Pascual G.
        • Li M.
        • Welch J.S.
        • Ricote M.
        • Hsiang C.H.
        • et al.
        15-Deoxy-delta 12,14-prostaglandin J2 inhibits multiple steps in the NF-κB signaling pathway.
        Proc Natl Acad Sci USA. 2000; 97: 4844-4849
        • Koga H.
        • Nanjoh Y.
        • Makimura K.
        • Tsuboi R.
        In vitro antifungal activities of luliconazole, a new topical imidazole.
        Med Mycol. 2009; 47: 640-647
        • Xie Y.
        • Takai T.
        • Chen X.
        • Okumura K.
        • Ogawa H.
        Long TSLP transcript expression and release of TSLP induced by TLR ligands and cytokines in human keratinocytes.
        J Dermatol Sci. 2012; 66: 233-237
        • Rajakariar R.
        • Hilliard M.
        • Lawrence T.
        • Trivedi S.
        • Colville-Nash P.
        • Bellingan G.
        • et al.
        Hematopoietic prostaglandin D2 synthase controls the onset and resolution of acute inflammation through PGD2 and 15-deoxy-delta 12,14-prostaglandin J2.
        Proc Natl Acad Sci USA. 2007; 104: 20979-20984
        • Trivedi S.G.
        • Newson J.
        • Rajakariar R.
        • Jacques T.S.
        • Hannon R.
        • Kanaoka Y.
        • et al.
        Essential role for hematopoietic prostaglandin D2 synthase in the control of delayed type hypersensitivity.
        Proc Natl Acad Sci USA. 2006; 103: 5179-5184
        • Lin T.H.
        • Tang C.H.
        • Wu K.
        • Fong Y.C.
        • Yang R.S.
        • Fu W.M.
        15-Deoxy-Δ(12,14) -prostaglandin-J2 and ciglitazone inhibit TNF-α-induced matrix metalloproteinase 13 production via the antagonism of NF-κB activation in human synovial fibroblasts.
        J Cell Physiol. 2011; 226: 3242-3250
        • Hayes J.D.
        • Flanagan J.U.
        • Jowsey I.R.
        Glutathione transferases.
        Annu Rev Pharmacol Toxicol. 2005; 45: 51-88
        • Tolman E.L.
        • Fuller B.L.
        Inhibition of thromboxane synthesis in guinea pig lung and human platelets by clotrimazole and other imidazole antifungals.
        Biochem Pharmacol. 1983; 32: 3488-3490
        • Gandhi U.H.
        • Kaushal N.
        • Ravindra K.C.
        • Hegde S.
        • Nelson S.M.
        • Narayan V.
        • et al.
        Selenoprotein-dependent up-regulation of hematopoietic prostaglandin D2 synthase in macrophages is mediated through the activation of peroxisome proliferator-activated receptor (PPAR) γ.
        J Biol Chem. 2011; 286: 27471-27482
        • Scheidereit C.
        IκB kinase complexes: gateways to NF-κB activation and transcription.
        Oncogene. 2006; 25: 6685-6705
        • Kawai T.
        • Akira S.
        The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors.
        Nat Immunol. 2010; 11: 373-384
        • Bhattacharjee R.
        • Xiang W.
        • Wang Y.
        • Zhang X.
        • Billiar T.R.
        cAMP prevents TNF-induced apoptosis through inhibiting DISC complex formation in rat hepatocytes.
        Biochem Biophys Res Commun. 2012; 423: 85-90
        • Guevara Patiño J.A.
        • Ivanov V.N.
        • Lacy E.
        • Elkon K.B.
        • Marino M.W.
        • Nikolic-Zugić J.
        TNF-α is the critical mediator of the cyclic AMP-induced apoptosis of CD8+4+ double-positive thymocytes.
        J Immunol. 2000; 164: 1689-1694
        • Hsu P.Y.
        • Tsai A.L.
        • Kulmacz R.J.
        • Wang L.H.
        Expression, purification, and spectroscopic characterization of human thromboxane synthase.
        J Biol Chem. 1999; 274: 762-769
        • Wang L.H.
        • Tsai A.L.
        • Hsu P.Y.
        Substrate binding is the rate-limiting step in thromboxane synthase catalysis.
        J Biol Chem. 2001; 276: 14737-14743
        • Matsuura K.
        • Yoshioka S.
        • Tosha T.
        • Hori H.
        • Ishimori K.
        • Kitagawa T.
        • et al.
        Structural diversities of active site in clinical azole-bound forms between sterol 14alpha-demethylases (CYP51s) from human and Mycobacterium tuberculosis.
        J Biol Chem. 2005; 280: 9088-9096
        • Steel H.C.
        • Tintinger G.R.
        • Anderson R.
        Comparison of the anti-inflammatory activities of imidazole antimycotics in relation to molecular structure.
        Chem Biol Drug Des. 2008; 72: 225-228
        • Faergemann J.
        • Zehender H.
        • Jones T.
        • Maibach I.
        Terbinafine levels in serum, stratum corneum, dermis-epidermis (without stratum corneum), hair, sebum and eccrine sweat.
        Acta Derm Venereol. 1991; 71: 322-326
        • Bogiatzi S.I.
        • Guillot-Delost M.
        • Cappuccio A.
        • Bichet J.C.
        • Chouchane-Mlik O.
        • Donnadieu M.H.
        • et al.
        Multiple-checkpoint inhibition of thymic stromal lymphopoietin-induced TH2 response by TH17-related cytokines.
        J Allergy Clin Immunol. 2012; 130: 233-240
        • Wong A.W.
        • Hon E.K.
        • Zee B.
        Is topical antimycotic treatment useful as adjuvant therapy for flexural atopic dermatitis: randomized, double-blind, controlled trial using one side of the elbow or knee as a control.
        Int J Dermatol. 2008; 47: 187-191
        • Kabashima-Kubo R.
        • Nakamura M.
        • Sakabe J.
        • Sugita K.
        • Hino R.
        • Mori T.
        • et al.
        A group of atopic dermatitis without IgE elevation or barrier impairment shows a high Th1 frequency: possible immunological state of the intrinsic type.
        J Dermatol Sci. 2012; 67: 37-43
        • Ishibashi Y.
        • Sugita T.
        • Nishikawa A.
        Cytokine secretion profile of human keratinocytes exposed to Malassezia yeasts.
        FEMS Immunol Med Microbiol. 2006; 48: 400-409