Advertisement

Nuclear IL-33 Plays an Important Role in the Suppression of FLG, LOR, Keratin 1, and Keratin 10 by IL-4 and IL-13 in Human Keratinocytes

Open ArchivePublished:April 15, 2021DOI:https://doi.org/10.1016/j.jid.2021.04.002
      IL-33 is a chromatin-associated multifunctional cytokine implicated in the pathogenesis of atopic dermatitis (AD), an inflammatory skin disorder characterized by skin barrier dysfunction. The previous reports show that IL-33 is highly detected in the nucleus of epidermal keratinocytes in AD lesions compared with that in unaffected or normal skin. However, it is unclear whether intracellular IL-33 directly contributes to the pathogenesis of AD. T helper type 2 cytokines IL-4 and IL-13 that are elevated in AD lesions suppress keratinocyte differentiation to impair skin barrier function. We investigated whether intracellular IL-33 is involved in IL-4 and IL-13 function. In monolayer culture and living skin equivalent analyses, IL-4 and IL-13 increased the expression of full-length IL-33 in the nucleus of keratinocytes by activating the MAPK/extracellular signal‒regulated kinase kinase/extracellular signal‒regulated kinase signaling pathway, which is necessary for the inhibition of differentiation markers FLG, LOR, keratin 1, and keratin 10. The nuclear IL-33 functions as a transcription cofactor of signal transducer and activator of transcription 3, increasing the binding of phosphorylated signal transducer and activator of transcription 3 to FLG promoter, thereby inhibiting its transcription, and it inhibits the expression of transcription factor RUNX1 by signal transducer and activator of transcription 3 and signal transducer and activator of transcription 6, thereby downregulating LOR, keratin 1, and keratin 10. Thus, the elevated nuclear IL-33 in the epidermis of AD lesions may be involved in the pathogenesis of AD by inhibiting keratinocyte differentiation and skin barrier function.

      Graphical abstract

      Abbreviations:

      AD (atopic dermatitis), ChIP (chromatin immunoprecipitation), ERK (extracellular signal‒regulated kinase), K (keratin), KC (keratinocyte), LSE (living skin equivalent), NHEK (normal human epidermal keratinocyte), p-STAT (phosphorylated signal transducer and activator of transcription), siRNA (small interfering RNA), STAT (signal transducer and activator of transcription), Th2 (T helper type 2)

      Introduction

      IL-33 belongs to the IL-1 family and was first described as a chromatin-associated nuclear factor constitutively expressed in the nuclei of endothelial and epithelial cells (
      • Baekkevold E.S.
      • Roussigné M.
      • Yamanaka T.
      • Johansen F.E.
      • Jahnsen F.L.
      • Amalric F.
      • et al.
      Molecular characterization of NF-HEV, a nuclear factor preferentially expressed in human high endothelial venules.
      ). Full-length IL-33 (30 kDa) is usually stored in the nucleus as an alarmin, which is rapidly released after cell damage (
      • Cayrol C.
      • Girard J.P.
      Interleukin-33 (IL-33): A nuclear cytokine from the IL-1 family.
      ;
      • Imai Y.
      Interleukin-33 in atopic dermatitis.
      ). IL-33 is susceptible to various stress; it can be cleaved by several intracellular and extracellular proteases, generating different sizes of isoforms (6.5, 18, 20, and 25 and/or 26 kDa) (
      • Cayrol C.
      • Girard J.P.
      Interleukin-33 (IL-33): A nuclear cytokine from the IL-1 family.
      ;
      • Cayrol C.
      • Duval A.
      • Schmitt P.
      • Roga S.
      • Camus M.
      • Stella A.
      • et al.
      Environmental allergens induce allergic inflammation through proteolytic maturation of IL-33.
      ;
      • Dai X.
      • Tohyama M.
      • Murakami M.
      • Shiraishi K.
      • Liu S.
      • Mori H.
      • et al.
      House dust mite allergens induce interleukin 33 (IL-33) synthesis and release from keratinocytes via ATP-mediated extracellular signaling.
      ). Extracellular IL-33 acts as a pro-T helper type 2 (Th2) allergic cytokine, which can activate ST2+ cells, leading to the secretion of Th2 cytokines and eosinophil accumulation, whereas intracellular IL-33 has not been well-characterized (
      • Cayrol C.
      • Girard J.P.
      Interleukin-33 (IL-33): A nuclear cytokine from the IL-1 family.
      ;
      • Imai Y.
      Interleukin-33 in atopic dermatitis.
      ).
      Atopic dermatitis (AD) is a multifactorial, chronic inflammatory skin disorder characterized by impaired skin barrier function, excessive production of Th2 cytokines, and pruritus and/or excoriation. Serum IL-33 levels are elevated in some patients with AD and correlated with disease severity (
      • Dajnoki Z.
      • Béke G.
      • Mócsai G.
      • Kapitány A.
      • Gáspár K.
      • Hajdu K.
      • et al.
      Immune-mediated skin inflammation is similar in severe atopic dermatitis patients with or without filaggrin mutation.
      ;
      • Tamagawa-Mineoka R.
      • Okuzawa Y.
      • Masuda K.
      • Katoh N.
      Increased serum levels of interleukin 33 in patients with atopic dermatitis.
      ). Intracellular IL-33 (mainly located in the nucleus) is detected only in the basal and suprabasal layers of normal skin epidermis but is distributed throughout the epidermis of AD lesions (
      • Balato A.
      • Raimondo A.
      • Balato N.
      • Ayala F.
      • Lembo S.
      Interleukin-33: increasing role in dermatological conditions.
      ;
      • Dajnoki Z.
      • Béke G.
      • Mócsai G.
      • Kapitány A.
      • Gáspár K.
      • Hajdu K.
      • et al.
      Immune-mediated skin inflammation is similar in severe atopic dermatitis patients with or without filaggrin mutation.
      ), possibly owing to the stimulation of certain AD-related cytokines and mite allergens (
      • Dai X.
      • Tohyama M.
      • Murakami M.
      • Shiraishi K.
      • Liu S.
      • Mori H.
      • et al.
      House dust mite allergens induce interleukin 33 (IL-33) synthesis and release from keratinocytes via ATP-mediated extracellular signaling.
      ;
      • Meephansan J.
      • Komine M.
      • Tsuda H.
      • Karakawa M.
      • Tominaga S.
      • Ohtsuki M.
      Expression of IL-33 in the epidermis: the mechanism of induction by IL-17.
      ,
      • Meephansan J.
      • Tsuda H.
      • Komine M.
      • Tominaga S.
      • Ohtsuki M.
      Regulation of IL-33 expression by IFN-γ and tumor necrosis factor-α in normal human epidermal keratinocytes.
      ). In fact, the elevated epidermal nuclear IL-33 (often together with decreased FLG) is more common in patients with AD than increased serum IL-33 (
      • Balato A.
      • Raimondo A.
      • Balato N.
      • Ayala F.
      • Lembo S.
      Interleukin-33: increasing role in dermatological conditions.
      ;
      • Dajnoki Z.
      • Béke G.
      • Mócsai G.
      • Kapitány A.
      • Gáspár K.
      • Hajdu K.
      • et al.
      Immune-mediated skin inflammation is similar in severe atopic dermatitis patients with or without filaggrin mutation.
      ). However, in addition to being the source of cytokine IL-33, it is unclear whether epidermal nuclear IL-33 contributes to the pathogenesis of AD.
      Th2 cytokines IL-4 and IL-13, which play critical roles in AD development (
      • Brandt E.B.
      • Sivaprasad U.
      Th2 cytokines and atopic dermatitis.
      ), activate the Jak‒tyrosine kinase/signal transducer and activator of transcription (STAT) 3, Jak‒STAT6, and MAPK/extracellular signal‒regulated kinase (ERK) kinase/ERK signaling pathways by binding to their receptors, leading to the suppression of differentiation markers FLG, LOR, keratin (K) 1, and K10; barrier dysfunction; and Th2 inflammation (
      • Amano W.
      • Nakajima S.
      • Kunugi H.
      • Numata Y.
      • Kitoh A.
      • Egawa G.
      • et al.
      The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
      ). In this study, we performed monolayer culture and living skin equivalent (LSE) analyses and studied whether intracellular IL-33 is involved in the downregulation of differentiation markers by IL-4 and IL-13 in normal human epidermal keratinocytes (NHEKs). We detected the elevated full-length IL-33 in the nucleus of NHEKs treated with IL-4 and IL-13, which is required for the downregulation of FLG, LOR, K1, and K10. Our study indicates that nuclear IL-33 distributed throughout the epidermis of AD lesions may be involved in the inhibition of keratinocyte (KC) differentiation and barrier function.

      Results

      IL-4 and IL-13 increases IL-33 expression in the nucleus of KCs by activating MAPK/ERK kinase/ERK signaling pathway

      Dupilumab is an IL-4 receptor α-antagonist that inhibits IL-4 and IL-13 signaling through the blockade of the shared IL-4α subunit and is effective against moderate-to-severe AD (
      • Gooderham M.J.
      • Hong H.C.
      • Eshtiaghi P.
      • Papp K.A.
      Dupilumab: a review of its use in the treatment of atopic dermatitis.
      ;
      • Harb H.
      • Chatila T.A.
      Mechanisms of dupilumab.
      ). Treatment with dupilumab resulted in an increase in FLG and a decrease in nuclear IL-33 in the epidermis of AD lesions (Supplementary Figure S1a), suggesting a link between decreased FLG and increased nuclear IL-33 in the signaling activation of IL-4 and IL-13.
      To investigate whether IL-4 and IL-13 regulate IL-33 expression, NHEKs were stimulated with IL-4 and IL-13 for different periods of time. IL-4 and IL-13 increased the IL33 mRNA level by 1.5-fold after 12 hours (Figure 1a) and upregulated IL-33 protein in a time-dependent manner (Figure 1b). We mainly detected 30 kDa IL-33 in NHEKs except for the weak band of approximately 26 kDa (Figure 1b), as previously reported (
      • Dai X.
      • Tohyama M.
      • Murakami M.
      • Shiraishi K.
      • Liu S.
      • Mori H.
      • et al.
      House dust mite allergens induce interleukin 33 (IL-33) synthesis and release from keratinocytes via ATP-mediated extracellular signaling.
      ). The IL-33 protein is located almost in the nucleus, not in the cytoplasm, as verified by specific nuclear and cytoplasmic markers (Figure 1c). To evaluate the role of IL-4 and IL-13 in the skin ex vivo, we treated LSE with IL-4 and IL-13. Figure 1d shows that IL-4 and IL-13 enhanced nuclear IL-33 immunostaining in the epidermis of LSE while suppressing FLG immunostaining in the stratum corneum. Despite the increased nuclear IL-33, IL-4 and IL-13 did not stimulate IL-33 secretion (Supplementary Figure S2a). Next, we investigated whether the signaling pathways activated by IL-4 and IL-13 (
      • Amano W.
      • Nakajima S.
      • Kunugi H.
      • Numata Y.
      • Kitoh A.
      • Egawa G.
      • et al.
      The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
      ;
      • Omori-Miyake M.
      • Yamashita M.
      • Tsunemi Y.
      • Kawashima M.
      • Yagi J.
      In vitro assessment of IL-4- or IL-13-mediated changes in the structural components of keratinocytes in mice and humans [published correction appears in J Invest Dermatol 2018;138:472–3].
      ) contribute to IL-33 upregulation. To prevent ERK activation, NHEKs were pretreated with MAPK/ERK kinase/ERK inhibitor U0126 before exposure to IL-4 and IL-13. Pretreatment with U0126 prevented ERK phosphorylation in IL-4/IL-13‒treated and ‒untreated cells and inhibited the basal and induced expression of IL33 mRNA and IL-33 protein (Figure 1e), suggesting that ERK signal activation is required for IL-33 expression. IL-4 and IL-13 increased the level of phosphorylated STAT (p-STAT) 3, and infection with an AxSTAT3F, that is, an adenovirus vector expressing dominant-negative STAT3 blocked STAT3 activation in IL-4/IL-13‒treated and ‒untreated cells (Figure 1f, upper panel). p-STAT6 was not detected in the control NHEKs, but it was significantly induced by IL-4 and IL-13 (Figure 1g, upper panel). Transfection with a STAT6 small interfering RNA (siRNA) effectively knocked down STAT6 expression and suppressed STAT6 phosphorylation (Figure 1g, upper panel). However, neither STAT3 inactivation nor STAT6 knockdown reduced the IL-33 levels (Figure 1f and g), suggesting that neither STAT3 nor STAT6 is required for IL-33 expression. We detected negative feedback regulation of IL-33 protein on ERK phosphorylation (not STAT3 and STAT6 phosphorylation) (Supplementary Figures S3c and S4a), confirming the unique role of ERK signal in IL-33 expression. In the whole-cell lysates of siRNA-transfected cells (Figure 1g and Supplementary Figure S4a), we detected a small IL-33 band, which appeared to be produced by siRNA transfection (Supplementary Figure S4b).
      Figure thumbnail gr1
      Figure 1IL-4 and/or IL-13 increases the expression of IL-33 in the nucleus of keratinocytes. (a) Effect of IL-4 and IL-13 on IL33 mRNA expression. (b) IL-33 protein levels in whole-cell lysates and (c) in NEs and CEs. (d) Immunofluorescence staining of IL-33 (red) and FLG (green) in IL-4/IL-13‒treated (or ‒untreated) LSE. Keratinocytes pretreated with (e) U0126, (f) infected with adenovirus vector Adex, or (g) transfected with siRNA were stimulated with or without IL-4 and/or IL-13, and the phosphorylation of ERK, STAT3, and STAT6 and the expression of IL33 mRNA and protein were detected. Bar = 50 μM. Data were shown as mean ± SD (n = 3); significance was defined by Student’s t-test. ∗P < 0.05 (compared with the relative control group). CE, cytoplasmic extract; ERK, extracellular signal‒regulated kinase; h, hour; LSE, living skin equivalent; NE, nuclear extract; siRNA, small interfering RNA; p-ERK, phosphorylated extracellular signal‒regulated kinase; p-STAT, phosphorylated signal transducer and activator of transcription; STAT, signal transducer and activator of transcription.
      Notably, the increase in IL33 mRNA level caused by IL-4 and IL-13 was weaker than that in its protein level. This may be because IL-4 and IL-13 can regulate IL-33 expression at the protein level. Deubiquitinating enzymes play an important role in maintaining protein stability (
      • Lim K.H.
      • Ramakrishna S.
      • Baek K.H.
      Molecular mechanisms and functions of cytokine-inducible deubiquitinating enzymes.
      ), among which USP17 and USP21 regulate IL-33 stability (
      • Ni Y.
      • Tao L.
      • Chen C.
      • Song H.
      • Li Z.
      • Gao Y.
      • et al.
      The deubiquitinase USP17 regulates the stability and nuclear function of IL-33.
      ;
      • Tao L.
      • Chen C.
      • Song H.
      • Piccioni M.
      • Shi G.
      • Li B.
      Deubiquitination and stabilization of IL-33 by USP21.
      ). IL-4 and IL-13 induced the expression of USP17 and USP21 (Supplementary Figure S3a), which could increase IL-33 protein stability because inhibition of deubiquitinating enzymes resulted in IL-33 polyubiquitination (Supplementary Figure S3b) and decrease (Supplementary Figure S3c). Thus, IL-4 and IL-13 regulates IL33 expression at the mRNA and protein levels, increasing IL-33 in the nucleus of NHEKs.

      Nuclear IL-33 is involved in the downregulation of FLG, LOR, K1, and K10

      Next, we investigated the role of nuclear IL-33 in IL-4 and IL-13 downregulation of differentiation markers. IL-33 siRNA transfection significantly reduced its mRNA and protein levels in KCs (Figure 2a), suggesting successful IL-33 knockdown. IL-33 knockdown not only prevented IL-4/IL-13‒mediated reduction of FLG, LOR, K1, and K10 mRNA levels but also further induced their transcription (Figure 2b). In IL-4/IL-13‒untreated cells, IL-33 knockdown increased the FLG, LOR, K1, and K10 mRNA levels by 3.5-, 1.6-, 4.1-, and 2.2-fold, respectively (Figure 2b). Western blotting showed that IL-33 knockdown significantly increased the protein levels of FLG (particularly of pro-FLG), LOR, K1, and K10 in IL-4/IL-13‒treated and ‒untreated cells (Figure 2c). Thus, nuclear IL-33 not only participates in the downregulation of FLG, LOR, K1, and K10 by IL-4 and IL-13 but also controls the basal expression of these genes. We detected the increase of IL-33 by IL-4 or IL-13, which was also involved in the downregulation of differentiated markers (Supplementary Figure S1), as detected in IL-4/IL-13‒treated cells. Generally, both IL-4 and IL-13 are highly produced in patients with AD, and they affect KCs differentiation in a similar manner (
      • Brandt E.B.
      • Sivaprasad U.
      Th2 cytokines and atopic dermatitis.
      ), so we stimulated cells with IL-4 and IL-13 in subsequent experiments.
      Figure thumbnail gr2
      Figure 2Nuclear IL-33 is involved in the downregulation of FLG, LOR, K1, and K10. After transfection with siRNA, keratinocytes were treated with or without IL-4 and IL-13, and the expression of (a) IL33 mRNA and protein; (b) FLG, LOR, K1, and K10 mRNA, and (c) FLG, LOR, K1, and K10 protein was examined. (d) LSE was transfected with control or IL-33 siRNA and then treated with IL-4 and/or IL-13 for 3 days. The immunostaining of IL-33 (red), FLG (green), LOR (red), K1 (red), and K10 (green) in LSE was performed by immunofluorescence. Bar = 200 μM. Data are shown as mean ± SD (n = 3); significance was defined by two-way ANOVA. ∗P < 0.05 and #P < 0.05 (compared with the relative control group). h, hour; K, keratin; LSE, living skin equivalent; siRNA, small interfering RNA.
      We further evaluated the role of nuclear IL-33 in LSE. Transfection with IL-33 siRNA decreased IL-33 immunostaining (indicating IL-33 knockdown in LSE) but increased the immunostaining of FLG, LOR, K1, and K10 in the epidermis of IL-4/IL-13‒treated LSE (Figure 2d). Therefore, IL-33 knockdown can restore the KC differentiation suppressed by IL-4 and IL-13, even in LSE.

      Nuclear IL-33 is involved in the downregulation of differentiation markers by regulating the nuclear translocation of p-STAT3 and p-STAT6

      IL-4 and IL-13 inhibits the expression of differentiation markers by activating STAT3 and STAT6 in KCs (
      • Amano W.
      • Nakajima S.
      • Kunugi H.
      • Numata Y.
      • Kitoh A.
      • Egawa G.
      • et al.
      The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
      ;
      • Omori-Miyake M.
      • Yamashita M.
      • Tsunemi Y.
      • Kawashima M.
      • Yagi J.
      In vitro assessment of IL-4- or IL-13-mediated changes in the structural components of keratinocytes in mice and humans [published correction appears in J Invest Dermatol 2018;138:472–3].
      ). Infection with AxSTAT3F, which blocked STAT3 activation (Figure 1f, upper panel), completely restored FLG, LOR, and K10 and partially recovered K1 mRNA levels in IL-4/IL-13‒treated cells and induced excess mRNA expression of these genes in IL-4/IL-13‒untreated cells (Figure 3a). Moreover, STAT3 inactivation restored and increased the protein levels of FLG, LOR, K1, and K10 in IL-4/IL-13‒treated and ‒untreated cells (Figure 3c). Meanwhile, transfection with STAT6 siRNA, which almost eliminated STAT6 expression and phosphorylation (Figure 1g, upper panel), restored the LOR, K1, and K10 but not FLG mRNA and protein levels suppressed by IL-4 and IL-13 (Figure 3b and d). Unlike STAT3, STAT6 did not affect the basal expression of the differentiation markers (Figure 3b), which may be due to the absence of endogenous STAT6 activation in IL-4/IL-13‒untreated cells (Figure 1g, upper panel).
      Figure thumbnail gr3
      Figure 3Nuclear IL-33 contributes to the downregulation of FLG, LOR, K1, and K10 by regulating the nuclear translocation of p-STAT3 and p-STAT6. Keratinocytes infected with AxLacZ or AxSTAT3F were treated with or without IL-4 and/or IL-13 for (a) 24 hours or (c) 30 hours, or keratinocytes transfected with cr or STAT6 siRNA were treated with or without IL-4 and/or IL-13 for (b) 24 hours or (d) 30 hours. The (a, b) mRNA and (c, d) protein levels of FLG, LOR, K1, and K10 were measured. (e) Keratinocytes transfected with cr or IL-33 siRNA were treated with IL-4 and IL-13 for the indicated times, and the levels of p-STAT3, p-STAT6, and IL-33 in the cytoplasmic and nuclear extracts were examined. Data are shown as mean ± SD (n = 3); significance was defined by two-way ANOVA. ∗P < 0.05 and #P < 0.05 (compared with the relative control group). cr, control; h, hour; K, keratin; p-STAT, phosphorylated signal transducer and activator of transcription; siRNA, small interfering RNA; STAT, signal transducer and activator of transcription.
      Next, we performed western blotting of fractionated cellular proteins to investigate whether and how IL-33 acts on STAT3 and STAT6 activation. In whole-cell lysates, IL-33 knockdown significantly inhibited the phosphorylation of STAT3 and STAT6 at 24 hours after stimulation but enhanced their phosphorylation at 8 hours (Supplementary Figure S4a), indicating that intracellular IL-33 is not necessary for their phosphorylation. In control siRNA‒transfected cells, full-length IL-33 was detected in the nucleus rather than in the cytoplasm, and its expression was upregulated by IL-4 and IL-13 but disappeared in IL-33 siRNA‒transfected cells (Figure 3e). After the phosphorylation of STATs, IL-4 and IL-13 trigger the nuclear translocation of p-STAT3 and p-STAT6, which was completely prevented by IL-33 knockdown because both p-STAT3 and p-STAT6 were not detected in the nucleus of NHEKs transfected with IL-33 siRNA (Figure 3e). After IL-4 and IL-13 stimulation, the cytoplasmic p-STAT3 and p-STAT6, which were easily detected in the control cells, were increased by IL-33 knockdown early after stimulation but were significantly inhibited at 24 hours (Figure 3e). In the cytoplasm, we detected a small IL-33 band (Supplementary Figure S4b), which showed no distinct effect on the phosphorylation of STAT3 and STAT6 (Figure 3e). Thus, nuclear IL-33 contributes to the downregulation of FLG, LOR, K1, and K10 by regulating the nuclear translocation and p-STAT3 and p-STAT6.

      Nuclear IL-33 functions as a transcription cofactor of STAT3 and facilitates the DNA binding of p-STAT3, thereby inhibiting FLG transcription

      We next investigated whether nuclear IL-33 regulates FLG promoter activation. A reporter gene assay showed that IL-4 and IL-13 reduced FLG promoter activity, whereas IL-33 siRNA transfection (causing IL-33 knockdown) (Figure 2a) as well as AxSTAT3F infection (causing STAT3 inactivation) (Figure 1f) significantly increased FLG promoter activity in IL-4/IL-13‒treated and ‒untreated cells (Figure 4a).
      Figure thumbnail gr4
      Figure 4Nuclear IL-33 functions as a transcription cofactor of STAT3, facilitating p-STAT3 to the FLG promoter, thereby inhibiting FLG transcription. (a) Keratinocytes were treated with or without IL-4 and IL-13 for 18 hours, and a reporter gene assay was performed. (b) Keratinocytes were treated with IL-4 and IL-13 for 15 hours, and ChIP qPCR was performed. (c) After stimulation with IL-4 and IL-13 for 15 hours, keratinocytes were subjected to ChIP with p-STAT3 or IgG. ChIP samples were separated by SDS-PAGE to perform IB of IL-33 and p-STAT3. (d) After stimulation with IL-4 and/or IL-13 for 15 hours, nuclear proteins were subjected to IP with p-STAT3 or IgG. IP samples were separated by SDS-PAGE to perform IB of IL-33 and p-STAT3. Data are shown as mean ± SD (n = 3); significance was defined by two-way ANOVA. ∗P < 0.05 and #P < 0.05 (compared with the relative control group). ChIP, chromatin immunoprecipitation; IB, immunoblotting; IP, immunoprecipitation; p-STAT, phosphorylated signal transducer and activator of transcription; siRNA, small interfering RNA; STAT, signal transducer and activator of transcription.
      There are two potential STAT3 DNA‒binding sequences in the promoter of FLG (
      • Lee H.
      • Shin J.J.
      • Bae H.C.
      • Ryu W.I.
      • Son S.W.
      Toluene downregulates filaggrin expression via the extracellular signal-regulated kinase and signal transducer and activator of transcription-dependent pathways.
      ) (Supplementary Figure S5a). To detect whether IL-4 and IL-13 stimulate the STAT3 binding to the FLG promoter, we performed a chromatin immunoprecipitation (ChIP) assay using a specific antibody for p-STAT3. As shown by ChIP qPCR (Figure 4b), the binding of p-STAT3 to both FLG promoter elements was detected in IL-4/IL-13‒activated NHEKs, which was significantly inhibited by IL-33 knockdown and by STAT3 inactivation. To study the IL-33‒p-STAT3 interaction, IL-4/IL-13‒activated KCs were subjected to ChIP western blotting or coimmunoprecipitation western blotting. ChIP with a specific antibody against p-STAT3 instead of IgG showed that full-length IL-33 (albeit weaker) is coimmunoprecipitated with p-STAT3 (Figure 4c). The denaturing lysis buffer used in ChIP analysis can denature proteins and break IL-33‒p-STAT3 interaction. To perform coimmunoprecipitation, nuclear proteins were lysed in a nondenaturing lysis buffer. Figure 4d showed that full-length IL-33 were significantly coimmunoprecipitated with p-STAT3 (not with IgG), confirming the IL-33‒p-STAT3 interaction in the nucleus of IL-4/IL-13‒activated NHEKs, which is required for the binding of p-STAT3 to the FLG promoter (Figure 4b). Therefore, nuclear IL-33 interacts with p-STAT3 and facilitates its binding to the FLG promoter, leading to the inhibition of FLG transcription. Nuclear IL-33 can act as a transcription cofactor of STAT3 in inhibiting FLG transcription.

      Nuclear IL-33 participates in the downregulation of LOR, K1, and K10 by reducing the nuclear RUNX1 level

      There is no potential STAT3- (or STAT6)-binding sequence in the promoters of LOR, K1, and K10. However, a potential RUNX1-binding site can be detected in the promoters of LOR, K1, and K10. RUNX1 belongs to the RUNX family of transcription factors, including RUNX1, RUNX2, and RUNX3, which play an essential role in the balance between cell proliferation and differentiation during development (
      • Braun T.
      • Woollard A.
      RUNX factors in development: lessons from invertebrate model systems.
      ). RUNX1 has been detected in epidermal nuclei of normal human skin and is involved in K1 and LOR upregulation during KC differentiation (
      • Masse I.
      • Barbollat-Boutrand L.
      • Molina M.
      • Berthier-Vergnes O.
      • Joly-Tonetti N.
      • Martin M.T.
      • et al.
      Functional interplay between p63 and p53 controls RUNX1 function in the transition from proliferation to differentiation in human keratinocytes.
      ). We also detected ubiquitous RUNX1 in the epidermal nuclei of normal skins; however, RUNX1 staining was much weaker in the epidermis of AD lesions (although clearly detected in dermal infiltrating cells and hair follicle cells) (Figure 5a). In cultured NHEKs, IL-4 and IL-13 reduced the RUNX1 mRNA level in a time-dependent manner but did not reduce the RUNX2 mRNA level (no RUNX3 mRNA was detected, data not shown) (Figure 5b). Moreover, IL-4 and IL-13 inhibited the expression of RUNX1 protein, which was located in KC nuclei, as detected by western blotting of monolayer-cultured NHEKs (Figure 5c) and by immunofluorescence staining of LSE (accompanied with a decrease in K10) (Figure 5d). We further explored the mechanism of IL-4/IL-13‒mediated RUNX1 downregulation. STAT3 inactivation and STAT6 knockdown partially restored the IL-4/IL-13‒inhibited RUNX1 expression, and blocking basal STAT3 activation increased the RUNX1 mRNA level in IL-4/IL-13‒untreated cells (Figure 5e). Therefore, IL-4 and IL-13 reduce the expression of RUNX1 by activating STAT3 and STAT6.
      Figure thumbnail gr5
      Figure 5IL-4 and/or IL-13 decreases the expression of RUNX1 by activating STAT3 and STAT6. (a) RUNX1 immunostaining in NS and AD lesions. Bar = 200 μM. (b) Effect of IL-4 and IL-13 on the mRNA expression of RUNX1 and RUNX2. (c) Effect of IL-4 and IL-13 on RUNX1 protein expression in WCE and in NE. (d) Immunostaining of RUNX1 (red) and K10 (green) in IL-4/IL13‒treated (or ‒untreated) LSE by immunofluorescence. Bar = 50 μM. (e) After infection with adenovirus vector Adex or transfection with siRNA, keratinocytes were treated with or without IL-4 and IL-13 for 20 h, and RUNX1 mRNA was measured. Data are shown as mean ± SD (n = 3); significance was defined by (b) Student’s t-test or (e) two-way ANOVA. ∗P < 0.05 and #P < 0.05 (compared with the relative control group). AD, atopic dermatitis; h, hour; K, keratin; LSE, living skin equivalent; NE, nuclear extract; NS, normal skin; siRNA, small interfering RNA; STAT, signal transducer and activator of transcription; WCE, whole-cell extract.
      Irrespective of IL-4 and IL-13 stimulation, IL-33 knockdown significantly increased the RUNX1 mRNA and nuclear protein levels (Figure 6a and b), indicating that nuclear IL-33 is involved in RUNX1 downregulation. To study the effect of RUNX1, RUNX1 siRNA and IL-33 siRNA were transfected into KCs together, resulting in a knockdown of both, as shown by their mRNA and protein expression (Figure 6a and b). The knockdown of RUNX1 almost blocked IL-33 knockdown‒induced LOR, K1, and K10 but not FLG mRNA expression in IL-4/IL-13‒treated and IL-4/IL-13‒untreated cells (Figure 6c). Thus, nuclear IL-33‒mediated downregulation of LOR, K1, and K10 is due to a decrease in RUNX1 expression in NHEKs.
      Figure thumbnail gr6
      Figure 6Nuclear IL-33 participates in the downregulation of LOR, K1, and K10 by reducing RUNX1 level. (a) After transfection with siRNA, keratinocytes were treated with or without IL-4 and/or IL-13 for 24 hours; the mRNA levels of IL33 and RUNX1 were detected. (b) After transfection with siRNA, keratinocytes were treated with or without IL-4 and/or IL-13 for 20 hours; the protein levels of RUNX1 and IL-33 in nuclear extracts were evaluated. (c) After transfection with siRNA, keratinocytes were treated with or without IL-4 and/or IL-13 for 24 hours; the mRNA levels of LOR, K1, K10, and FLG were detected. Data are shown as mean ± SD (n = 3); significance was defined by two-way ANOVA. ∗P < 0.05 and #P < 0.05 (compared with the relative control group). K, keratin; siRNA, small interfering RNA.

      Discussion

      We show in this study that Th2 cytokines IL-4 and IL-13 increase the expression of IL-33 (full length) in the nucleus of NHEKs, which is essential for inhibiting KC differentiation. Nuclear IL-33 inhibits NF-κB transactivation by sequestering nuclear p65 (
      • Ali S.
      • Mohs A.
      • Thomas M.
      • Klare J.
      • Ross R.
      • Schmitz M.L.
      • et al.
      The dual function cytokine IL-33 interacts with the transcription factor NF-κB to dampen NF-κB-stimulated gene transcription.
      ). We found that nuclear IL-33 is involved in the activation of STAT3 and STAT6. In KCs, nuclear IL-33 appears to act as a transcription cofactor or regulator of STAT3, promoting STAT3 recruitment to the FLG promoter, thereby inhibiting FLG transcription, and it also contributes to the downregulation of LOR, K1, and K10 by reducing the level of the transcription factor RUNX1 by STAT3 and STAT6. This study unraveled that nuclear IL-33 is involved in the suppression of KC differentiation.
      FLG is typically downregulated in patients with AD, irrespective of their FLG genotype (
      • Dajnoki Z.
      • Béke G.
      • Mócsai G.
      • Kapitány A.
      • Gáspár K.
      • Hajdu K.
      • et al.
      Immune-mediated skin inflammation is similar in severe atopic dermatitis patients with or without filaggrin mutation.
      ). The downregulation of FLG in AD skin is mainly due to the production of excessive Th2 cytokines (
      • Amano W.
      • Nakajima S.
      • Kunugi H.
      • Numata Y.
      • Kitoh A.
      • Egawa G.
      • et al.
      The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
      ;
      • Tsuchisaka A.
      • Furumura M.
      • Hashimoto T.
      Cytokine regulation during epidermal differentiation and barrier formation.
      ). STAT3 is a key regulator of KC differentiation, and STAT3 activation is responsible for the downregulation of FLG by various AD-related cytokines (
      • Amano W.
      • Nakajima S.
      • Kunugi H.
      • Numata Y.
      • Kitoh A.
      • Egawa G.
      • et al.
      The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
      ;
      • Cornelissen C.
      • Marquardt Y.
      • Czaja K.
      • Wenzel J.
      • Frank J.
      • Lüscher-Firzlaff J.
      • et al.
      IL-31 regulates differentiation and filaggrin expression in human organotypic skin models.
      ;
      • Kim J.H.
      • Bae H.C.
      • Ko N.Y.
      • Lee S.H.
      • Jeong S.H.
      • Lee H.
      • et al.
      Thymic stromal lymphopoietin downregulates filaggrin expression by signal transducer and activator of transcription 3 (STAT3) and extracellular signal-regulated kinase (ERK) phosphorylation in keratinocytes.
      ;
      • Ryu W.I.
      • Lee H.
      • Bae H.C.
      • Ryu H.J.
      • Son S.W.
      IL-33 down-regulates filaggrin expression by inducing STAT3 and ERK phosphorylation in human keratinocytes.
      ). Jak inhibitor JTE-052 can increase FLG and LOR levels by inhibiting STAT3 activation in AD mouse models (
      • Amano W.
      • Nakajima S.
      • Kunugi H.
      • Numata Y.
      • Kitoh A.
      • Egawa G.
      • et al.
      The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
      ). In IL-4/IL-13‒treated NHEKs, STAT3 but not STAT6 mediated the downregulation of FLG (Figure 2). STAT3 usually activates the transcription of inflammation-, cell proliferation‒, and migration-related genes (
      • Calautti E.
      • Avalle L.
      • Poli V.
      Psoriasis: a STAT3-centric view.
      ;
      • Sano S.
      • Chan K.S.
      • DiGiovanni J.
      Impact of Stat3 activation upon skin biology: a dichotomy of its role between homeostasis and diseases.
      ); however, it can also act as a transcriptional repressor (
      • Lee H.
      • Shin J.J.
      • Bae H.C.
      • Ryu W.I.
      • Son S.W.
      Toluene downregulates filaggrin expression via the extracellular signal-regulated kinase and signal transducer and activator of transcription-dependent pathways.
      ). In this study, we found that STAT3 negatively regulates FLG transcription. In the nucleus of NHEKs activated by IL-4 and IL-13, we identified the IL-33/p-STAT3 complex bound to the FLG promoter. Nuclear IL-33 is likely to act as a cofactor of transcription factor STAT3, facilitating the recruitment of p-STAT3 to the specific STAT3-binding sequences on the FLG promoter, thereby inhibiting FLG transcription. By increasing STAT3 transactivation, IL-4/IL-13‒induced nuclear IL-33 causes FLG downregulation, whereas endogenous nuclear IL-33 in epidermal basal cells can dampen FLG expression in proliferating KCs. Although cytokine IL-33 inhibits FLG expression (
      • Ryu W.I.
      • Lee H.
      • Bae H.C.
      • Ryu H.J.
      • Son S.W.
      IL-33 down-regulates filaggrin expression by inducing STAT3 and ERK phosphorylation in human keratinocytes.
      ), it is not involved in the downregulation of FLG by IL-4 and IL-13 (Supplementary Figure S2).
      IL-4 and IL-13 also inhibited LOR, K1, and K10 expression.
      • Omori-Miyake M.
      • Yamashita M.
      • Tsunemi Y.
      • Kawashima M.
      • Yagi J.
      In vitro assessment of IL-4- or IL-13-mediated changes in the structural components of keratinocytes in mice and humans [published correction appears in J Invest Dermatol 2018;138:472–3].
      reported that STAT6 activation was responsible for IL-4‒ or IL-13‒mediated K1 and K10 inhibition in mouse KCs.
      • Amano W.
      • Nakajima S.
      • Kunugi H.
      • Numata Y.
      • Kitoh A.
      • Egawa G.
      • et al.
      The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
      showed that STAT3 knockdown increased LOR mRNA level in both IL-4‒treated and ‒untreated KCs. In the skin of STAT6-transgenic mice, STAT6 activation significantly inhibits LOR expression (
      • Sehra S.
      • Yao Y.
      • Howell M.D.
      • Nguyen E.T.
      • Kansas G.S.
      • Leung D.Y.
      • et al.
      IL-4 regulates skin homeostasis and the predisposition toward allergic skin inflammation.
      ). In NHEKs, STAT6 activation is mainly responsible for IL-4/IL-13‒mediated reduction of LOR, K1, and K10, whereas STAT3 activation, constitutively detected in NHEKs and slightly increased by IL-4 and IL-13, primarily controls their basal expression (Figure 3). Accordingly, epidermal nuclear IL-33 contributes to the suppression of LOR, K1, and K10 by regulating the activation of STAT3 and STAT6.
      The transcription factor RUNX1 is significantly reduced in the epidermis of AD lesions, which may be caused by excessive Th2 cytokines because IL-4 and IL-13 significantly inhibited RUNX1 expression in KCs through nuclear IL-33‒dependent activation of STAT3 and STAT6 (a finding of this study). In the human epidermis, RUNX1 is a direct target of p63 and plays a role in the initiation of KC differentiation (
      • Masse I.
      • Barbollat-Boutrand L.
      • Molina M.
      • Berthier-Vergnes O.
      • Joly-Tonetti N.
      • Martin M.T.
      • et al.
      Functional interplay between p63 and p53 controls RUNX1 function in the transition from proliferation to differentiation in human keratinocytes.
      ). We show in this study that IL-4/IL-13‒induced RUNX1 decrease is responsible for the downregulation of LOR, K1, and K10 (but not of FLG). Therefore, the RUNX1 decrease detected in the epidermis of AD lesions (probably owing to the elevated nuclear IL-33) can promote AD development by affecting KC differentiation.
      On the basis of the research of skin-expressed IL-33-transgenic mice, cytokine IL-33 produced from epidermis seems to play an important role in AD development (
      • Imai Y.
      • Yasuda K.
      • Nagai M.
      • Kusakabe M.
      • Kubo M.
      • Nakanishi K.
      • et al.
      IL-33-induced atopic dermatitis-like inflammation in mice is mediated by group 2 innate lymphoid cells in concert with basophils.
      ,
      • Imai Y.
      • Yasuda K.
      • Sakaguchi Y.
      • Haneda T.
      • Mizutani H.
      • Yoshimoto T.
      • et al.
      Skin-specific expression of IL-33 activates group 2 innate lymphoid cells and elicits atopic dermatitis-like inflammation in mice.
      ). However, this did not exclude a role for extensive epidermal nuclear IL-33 in mouse skin inflammation. Cytokine IL-33 stimulates the secretion of Th2 cytokines by activating ST2+ cells (
      • Imai Y.
      Interleukin-33 in atopic dermatitis.
      ), whereas Th2 cytokines secreted into the skin can increase KC expression of nuclear IL-33, inhibiting KC differentiation (a finding of this study). Thus, cytokine IL-33 and epidermal nuclear IL-33 may play complementary roles in the development and exacerbation of AD.
      The suppression of differentiation (particularly inhibition of FLG)-related skin barrier disruption is a primary cause of AD (
      • Cabanillas B.
      • Novak N.
      Atopic dermatitis and filaggrin.
      ;
      • Kezic S.
      • Jakasa I.
      IFilaggrin and skin barrier function.
      ). Our findings show that nuclear IL-33 drives the function of IL-4 and IL-13 in the downregulation of FLG, LOR, K1, and K10; it also suppresses their homeostatic expression in KCs. Thus, an increased epidermal nuclear IL-33 level might predispose to the development of AD by promoting skin barrier disruption. Controlling the epidermal nuclear IL-33 level thus shows promise for treating AD lesions and preventing future exacerbations.

      Materials and Methods

      Detailed materials and methods are provided in Supplemental Materials and Methods.

      Human subjects and cell culture

      This study was conducted according to the principles of the Declaration of Helsinki, and all procedures involving human subjects received previous approval from the ethics committee at Ehime University School of Medicine, Japan. Normal human skin biopsies were obtained from plastic surgeries, and lesional skin biopsies were collected from patients with moderate-to-severe AD. All participants provided written informed consent. NHEKs and LSE were prepared and cultured as described previously (
      • Dai X.
      • Tohyama M.
      • Murakami M.
      • Shiraishi K.
      • Liu S.
      • Mori H.
      • et al.
      House dust mite allergens induce interleukin 33 (IL-33) synthesis and release from keratinocytes via ATP-mediated extracellular signaling.
      ).

      Chemical reagents

      Recombinant human IL-4 and IL-13 and U0126 were obtained from R&D Systems (Minneapolis, MN).

      Real-time RT-PCR

      The probes specific for GAPDH, IL-33, FLG, LOR, K1, K10, RUNX1, and RUNX2 (Thermo Fisher Scientific, Yokohama, Japan) were used for real-time RT-PCR.

      Protein isolation and western blotting

      Western blotting of whole-cell extracts, cytoplasmic extracts, and nuclear extracts was performed to evaluate protein expression. The primary antibodies used are listed in Supplementary Table S1.

      siRNA or adenovirus vectors

      The human STAT6 siRNA (Horizon Discovery, Tokyo, Japan), human IL-33 siRNA, and human RUNX1 siRNA (Thermo Fisher Scientific) were respectively transfected into subconfluent NHEKs or LSE to silence STAT6, IL-33, and RUNX1. AxSTAT3F and AxLacZ were infected into subconfluent NHEKs at a multiplicity of infection = 10.

      Immunohistochemistry

      The paraffin-embedded skin sections were reacted with anti-RUNX1 and then stained with ImmPRESS detection system (Vector Laboratories, Burlingame, CA).

      Immunofluorescence

      The paraffin-embedded sections were reacted with primary antibodies, then treated with secondary antibodies labeled with Alexa Fluor 488 (Thermo Fisher Scientific), and photographed using a BZ-X710 All-in-One Fluorescence Microscope (Keyence, Osaka, Japan).

      DNA construction and reporter gene assay

      The pGL3.0/FLG-Luc plasmid was constructed and transfected into KCs as described previously (
      • Lee H.
      • Shin J.J.
      • Bae H.C.
      • Ryu W.I.
      • Son S.W.
      Toluene downregulates filaggrin expression via the extracellular signal-regulated kinase and signal transducer and activator of transcription-dependent pathways.
      ). Luciferase activity was determined using a dual luciferase assay kit (Promega, Madison, WI).

      Coimmunoprecipitation

      The protocol provided by cell signaling technology was followed, with minor modifications. For immunoprecipitation of p-STAT3, the nuclei pellets were lysed in nondenaturing lysis buffer, and Sepharose G beads were used.

      ChIP assay

      NHEKs were harvested and subjected to ChIP assay using a ChIP assay kit (Cell Signaling, Danvers, MA) and anti‒p-STAT3 (Supplementary Table S1). To evaluate the binding of p-STAT3 to the FLG promoter, ChIP qPCR and ChIP western blotting were performed.

      Statistical analysis

      For every data, at least three independent experiments were performed, all of which yielded similar results. The results of real-time PCR, ELISA, and reporter gene assay were expressed as mean ± SD (n = 3) and were analyzed by Student's t-tests for two-group comparison or by ANOVA followed by Bonferroni multiple comparison test. P < 0.05 was statistically significant.

      Data availability statement

      No specific gene or protein database exists. Datasets related to this article can be found in https://doi.org/10.17632/25x7w7hv4y.1, hosted at Mendeley Data.

      ORCIDs

      Conflict of Interest

      KS declares research funding from Kyowa Kirin, Eli Lilly Japan, Torii Pharmaceutical, Maruho, Mitsubishi Tanabe Pharma, Kaken Pharmaceutical, AbbVie Godo, Taiho Pharmaceutical, Sato Pharmaceutical, Nippon Zoki Pharmaceutical, Lydia O'Leary Memorial Pias Dermatological Foundation, and Sun Pharma Japan and honorarium from Kyowa Kirin, Eli Lilly Japan, Torii Pharmaceutical, Maruho, Mitsubishi Tanabe Pharma, Kaken Pharmaceutical, AbbVie Godo, Taiho Pharmaceutical, Janssen Pharma, Eisai, Novel Pharma, Novartis Pharma, Bristol-Myers Squibb, and Sanofi K.K. The remaining authors state no conflict of interest.

      Acknowledgments

      We thank Teruko Tsuda and Eriko Tan for their technical assistance.

      Author Contributions

      Conceptualization: XD, MM, KS, JM, HM, RU, KS; Data Curation: XD, KS; Formal Analysis: XD; Funding Acquisition: KS; Investigation: XD; Methodology: XD, MM, KS; Project Administration: XD, KS; Resources: XD, HM; Visualization: XD, MM, KS; Writing - Original Draft Preparation: XD; Writing - Review and Editing: XD, KS

      Supplementary Materials and Methods

      Keratinocyte culture and treatment

      The human skin samples were obtained after plastic surgery. The epidermis was separated from the dermis, and the keratinocytes (KCs) were collected and were cultured in an MCDB153 medium that was supplemented with insulin (0.02 μM), hydrocortisone (0.5 μΜ), ethanolamine (0.1 mM), phosphoethanolamine (0.1 mM), bovine hypothalamic extract (50 μg/ml), and calcium ion (0.07 mM), as previously described (
      • Shirakata Y.
      • Ueno H.
      • Hanakawa Y.
      • Kameda K.
      • Yamasaki K.
      • Tokumaru S.
      • et al.
      TGF-beta is not involved in early phase growth inhibition of keratinocytes by 1alpha,25(OH)2vitamin D3.
      ). Normal human epidermal KCs (NHEKs) that had been passaged four times were used for the experiments, and the confluent NHEKs were treated with IL-4 (20 ng/ml) and IL-13 (10 ng/ml).

      Preparation and stimulation of living skin equivalent

      To prepare living skin equivalent (LSE), NHEKs and fibroblasts were isolated from human skin samples from plastic surgery and then cultured. A collagen gel containing the fibroblasts was prepared, and the KCs were seeded onto the concave surface of the contracted gel. When the KCs reached confluence, the LSE was raised to the air–liquid interface, and a cornification medium was added. LSE was treated with IL-4 and IL-13 for 3 days after air lifting for 7 days as described previously (
      • Tohyama M.
      • Yang L.
      • Hanakawa Y.
      • Dai X.
      • Shirakata Y.
      • Sayama K.
      IFN-α enhances IL-22 receptor expression in keratinocytes: a possible role in the development of psoriasis.
      ).

      RNA preparation and real-time RT-PCR

      The probes specific for GAPDH, IL-33, FLG, LOR, keratin gene K1 and K10, RUNX1, RUNX2, USP17, and USP21 were obtained from Thermo Fisher Scientific (Yokohama, Japan). Total RNA was isolated and subjected to real-time RT-PCR, and gene expression was analyzed as described previously (
      • Dai X.
      • Sayama K.
      • Yamasaki K.
      • Tohyama M.
      • Shirakata Y.
      • Hanakawa Y.
      • et al.
      SOCS1-negative feedback of STAT1 activation is a key pathway in the dsRNA-induced innate immune response of human keratinocytes.
      ).

      Protein isolation and western blotting

      After stimulation, total, cytoplasmic, and nuclear cell extracts were respectively collected at the indicated times. The whole-cell extracts were collected using RIPA buffer (Wako, Osaka, Japan). NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Fisher Scientific) were used to separate nuclear extracts from cytoplasmic extracts according to the protocol provided, with minor modulation: simply, after removing the cytoplasmic extracts, the insoluble nuclei pellets were washed twice before they were suspended in nuclear extraction reagents. To detect the protein levels, cell lysates were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes. Analyses were performed using a Vistra ECF Kit (Amersham Biosciences, Arlington Heights, IL), and then membranes were scanned using a FluoroImager (Molecular Dynamics, Sunnyvale, CA). The primary antibodies used for western blotting are listed in Supplementary Table S1.

      Small interfering RNA or adenovirus vectors

      The ON-TARGETplus SMART pool of human signal transducer and activator of transcription (STAT) 6 small interfering RNA (siRNA) (Horizon Discovery, Tokyo, Japan) was used for silencing STAT6, Stealth siRNA for human IL-33 and human RUNX1 (Thermo Fisher Scientific) was respectively used for silencing IL-33 and RUNX1, and scrambled siRNA (Thermo Fisher Scientific) was used as a control. Subconfluent NHEKs were transfected with special siRNA or scrambled siRNA using Lipofectamine RNAiMAX Transfection Reagent (Thermo Fisher Scientific), and LSE cultures were transfected with special siRNA or scrambled siRNA using 30% Pluronic F-127 (Sigma-Aldrich, Tokyo, Japan), according to the manufacturer’s instructions. The cells and LSE cultures were allowed to stabilize for at least 24 hours before using them for further experiments.
      Adenovirus vectors expressing dominant-negative STAT3 or AxSTAT3F were prepared and infected into subconfluent KCs at a multiplicity of infection = 10 as described previously (
      • Dai X.
      • Sayama K.
      • Yamasaki K.
      • Tohyama M.
      • Shirakata Y.
      • Hanakawa Y.
      • et al.
      SOCS1-negative feedback of STAT1 activation is a key pathway in the dsRNA-induced innate immune response of human keratinocytes.
      ;
      • Yamasaki K.
      • Hanakawa Y.
      • Tokumaru S.
      • Shirakata Y.
      • Sayama K.
      • Hanada T.
      • et al.
      Suppressor of cytokine signaling 1/JAB and suppressor of cytokine signaling 3/cytokine-inducible SH2 containing protein 3 negatively regulate the signal transducers and activators of transcription signaling pathway in normal human epidermal keratinocytes.
      ), and AxLacZ was used as a control. The cells were allowed to stabilize for 24 hours before being used for further experiments.

      Immunohistochemistry

      To analyze the expression of RUNX1, paraffin-embedded skin sections from three healthy donors and three patients with atopic dermatitis were respectively deparaffinized, adapted for antigen retrieval, blocked with 10% goat serum, and reacted with an anti-RUNX1 antibody (Supplementary Table S1) overnight at 4 °C. Staining was performed using the ImmPRESS detection system (Vector Laboratories, Burlingame, CA), according to the manufacturer’s instructions. A mouse IgG isotype control or the absence of a primary antibody resulted in no specific staining.

      Immunofluorescence

      Paraffin-embedded skin and LSE sections were deparaffinized, adapted for antigen retrieval, and blocked with blocking solution (Thermo Fisher Scientific) and were reacted with the primary antibodies (Supplementary Table S1) overnight at 4 °C. After washing with PBS, the sections were treated with secondary antibodies labeled with Alexa Fluor 488 (Thermo Fisher Scientific) and photographed using a BZ-X710 All-in-One Fluorescence Microscope (Keyence, Osaka, Japan). Control staining with IgG isotype controls or without primary antibodies yielded no positive signal.

      ELISA

      After scratching or not, the culture supernatants were respectively collected and stored at −20 °C before use. Supernatant IL-33 levels were quantified using a commercially available IL-33 ELISA kit (R&D Systems, Minneapolis, MN), according to the manufacturer’s instructions.

      DNA construction and reporter gene assay

      The FLG promoter (−916/+1 base pair) linked to the reporter gene was cloned by performing PCR with human genomic DNA as the template and primers 5′-CTAGCCCGGGCTCGAGGGAGATGCAATCTGCTCAACATAAC-3′ and 5′-GATCGCAGATCTCGAGACATTCAGATTTCCACCTTGGT-3′. PCR fragments were digested with XhoI and BglⅡ and ligated to pGL3.0 luciferase reporter vector (Promega, Madison, WI). KCs were transfected with siRNA or adenovirus vector Adex for 24 hours and cotransfected with the pGL3.0/FLG-Luc plasmid and control vector pRL-TK (Promega) using Lipofectamine 2000 (Thermo Fisher Scientific), followed by stimulation with IL-4 and IL-13. Next, the cells were harvested and subjected to a reporter gene assay using a dual luciferase assay kit (Promega). The relative luciferase activity was normalized to that of Renilla luciferase.

      Chromatin immunoprecipitation qPCR and chromatin immunoprecipitation western blotting

      After stimulation with IL-4 and IL-13 for 15 hours, NHEKs were treated with 1% formaldehyde for 10 minutes at room temperature. The cells were harvested and subjected to chromatin immunoprecipitation (ChIP) using SimpleChIP Enzymatic Chromatin IP Kit (Cell Signaling, Danvers, MA), according to the manufacturer’s instructions. Before ChIP with specific antibodies, we took out 5% of cell lysate (input samples) for immunoblotting of IL-33 and phosphorylated STAT (p-STAT) 3 to confirm the knockdown of IL-33 by IL-33 siRNA (Supplementary Figure S5b) and the inactivation of STAT3 by AxSTAT3F (Supplementary Figure S5c). After ChIP with anti‒p-STAT3 or anti-histone H3 (an assay control) (Supplementary Table S1), the captured genomic fragments were recovered by phenol–chloroform extraction. To exclude the nonspecific binding of antibodies, chromatin complexes were also immunoprecipitated with an anti-IgG as a negative control. Identification of the FLG promoter fragments was performed by qPCR using primers designed on the basis of the STAT3-binding sequences in the FLG promoter (the positions and directions of PCR primers were presented in Supplementary Figure S5a). The sequences of the PCR primers are as follows: for element 1, 5′- GGCATTCATCTCATGGCAAG-3′ (forward) and 5′- GGCTTGACAGGTGCTTGGTA-3′ (reverse) and for element 2, 5′-CAAGCCAAAGTGGGGTTACA-3′ (forward) and 5′-GAAGGCTGAGATAATGGCCC-3′ (reverse), and the primers of RPL30 (used for a control gene) was included in the ChIP assay kit. To perform ChIP western blot (
      • Mahadevan I.A.
      • Kumar S.
      • Rao M.R.S.
      Linker histone variant H1t is closely associated with repressed repeat-element chromatin domains in pachytene spermatocytes.
      ), the p-STAT3 and IgG ChIP samples were eluted, heated in ×2 SDS sample buffer overnight, and separated together with the input sample by SDS-PAGE. The immunoblotting of IL-33 and p-STAT3 was performed as described earlier.

      Coimmunoprecipitation assay

      The protocol provided by cell signaling technology was followed, with minor modifications. For immunoprecipitation of p-STAT3 or IgG, the nuclei pellets (obtained as described earlier) were lysed in a nondenaturing lysis buffer (20 mM Tris hydrogen chloride at pH 8.0, 137 mM sodium chloride, 1% Triton X-100, and 2 mM EDTA), and Sepharose G beads were used. After immunoprecipitation, the immunoprecipitation samples were washed three times with RIPA buffer. After the final wash, the pellets were suspended in ×2 SDS sample buffer, boiled for 3 minutes, and separated together with input samples by SDS-PAGE. The immunoblotting of IL-33 and p-STAT3 was performed.
      Figure thumbnail fx2
      Supplementary Figure S1Both IL-4 and IL-13 increase IL-33 expression, which is required for the downregulation of FLG, LOR, K1, and K10. (a) Detection of FLG (green) and IL-33 (red) by immunofluorescence staining. AD lesion skin biopsies were obtained from the same patient before and after treatment with dupilumab. Bar = 50 μM. (b) Increase of IL33 mRNA and protein expression by IL-4 or by IL-13. (c) Inhibition of IL33 mRNA by IL-33 siRNA transfection. (d) Upregulation of differentiation markers by IL-33 knockdown. Data are shown as mean ± SD (n = 3); significance was defined by (b) Student’s t-test or (c, d) two-way ANOVA. ∗P < 0.05 and #P < 0.05 (compared with the relative control group). AD, atopic dermatitis; h, hour; K, keratin; siRNA, small interfering RNA.
      Figure thumbnail fx3
      Supplementary Figure S2Cytokine IL-33 is not involved in IL-4‒ and IL-13‒mediated FLG reduction. (a) IL-4 and IL-13 did not release IL-33 from keratinocytes. Keratinocytes were treated with IL-4 and IL-13 for the indicated times, then cells were scratched or not, and the supernatants were collected for the detection of IL-33 by ELISA. (b) Blocking IL-33 signaling did not prevent IL-4 and IL-13 from downregulating FLG. NHEKs were pretreated with rhST2 (IL-33R, 100 ng/ml, R&D Systems, Minneapolis, MN) for 1 h and then stimulated with IL-4 and IL-13 or with IL-33 (60 ng/ml, R&D Systems) for 24 h. FLG mRNA expression was detected. Data are shown as mean ± SD (n = 3); significance was defined by Student’s t-test. ∗P < 0.05. h, hour; NHEK, normal human epidermal keratinocyte; rhST2, recombinant human ST2.
      Figure thumbnail fx4
      Supplementary Figure S3IL-4 and IL-13 upregulates IL-33 by increasing its protein stability. (a) Effect of IL-4 and IL-13 on the mRNA expression of USP17 and USP21. Data are shown as mean ± SD (n = 3); significance was defined by Student’s t-test. #P < 0.05 (compared with the relative control group). (b) Keratinocytes were stimulated with IL-4 and IL-13 for 24 h and treated with RP-619 (Sigma-Aldrich, Tokyo, Japan) for 6 h. The immunoblotting of IL-33 was performed. (c) Keratinocytes pretreated with RP-619 were stimulated with IL-4 and IL-13 for 24 h. The effect of RP-619 on IL-33 expression and signal activation was examined. DUB, deubiquitinating enzyme; ERK, extracellular signal‒regulated kinase; h, hour; p-ERK, phosphorylated extracellular signal‒regulated kinase; p-STAT, phosphorylated signal transducer and activator of transcription; STAT, signal transducer and activator of transcription.
      Figure thumbnail fx5
      Supplementary Figure S4IL-33 knockdown regulates the phosphorylation of ERK, STAT3, and STAT6 in NHEKs, and siRNA transfection causes IL-33 cleavage. (a) NHEKs were transfected with siRNA for 48 h and treated with IL-4 and IL-13 for 8 h or 24 h. The expression of IL-33 and the phosphorylation of ERK, STAT3, and STAT6 in whole-cell lysates were evaluated. (b) After transfected with control or IL-33 siRNA, NHEKs were incubated for 48 h in total before subjected to RIPA lysis buffer. At 8 h, 20 h, and 40 h after the addition of siRNA, medium change was performed, that is, the medium contained siRNA transfection reagent was replaced by MCDB 153 medium (without bovine hypothalamic extract). Whole-cell lysates were subjected to western blotting for the detection of IL-33. IL-33 protein is susceptible to various stress and is often cleaved by several proteases (
      • Cayrol C.
      • Girard J.P.
      Interleukin-33 (IL-33): A nuclear cytokine from the IL-1 family.
      ). As shown in b, siRNA transfection causes the cleavage of the stored IL-33, especially when cells were incubated with the siRNA transfection reagent >20 hours, producing a small band (just <30 kDa but >25 and/or 26 kDa), indicating that siRNA transfection can induce some intracellular proteases that target IL-33. This small IL-33 band has not been reported before, and it seems to be relatively stable or resistant to protein degradation (compared with other IL-33 isoforms) because it can be easily detected even after 3 days without distinct change. Western blotting of fractionated cellular proteins showed that this small IL-33 band was detected in the cytoplasm but not in the nucleus, so it appears to be inessential for STAT activation in keratinocytes (e). ERK, extracellular signal‒regulated kinase; h, hour; NHEK, normal human epidermal keratinocyte; p-ERK, phosphorylated extracellular signal‒regulated kinase; p-STAT, phosphorylated signal transducer and activator of transcription; siRNA, small interfering RNA; STAT, signal transducer and activator of transcription.
      Figure thumbnail fx6
      Supplementary Figure S5STAT3-binding elements in the FLG promoter and the knockdown of IL-33 by IL-33 siRNA and the inactivation of STAT3 by AxSTAT3F regarding ChIP samples are presented. (a) The putative STAT3-binding sites in the FLG promoter and the positions and directions of PCR primers used for ChIP qPCR. NHEKs (b) transfected with siRNA or (c) infected with adenovirus vector Adex were treated with IL-4 and IL-13 for 15 h and then cross-linked and harvested. The input samples were used for the immunoblotting of IL-33 and p-STAT3. bp, base pair; ChIP, chromatin immunoprecipitation; h, hour; p-STAT, phosphorylated signal transducer and activator of transcription; siRNA, small interfering RNA; STAT, signal transducer and activator of protein.
      Supplementary Table S1List of Primary Antibodies Used in this Study
      Name of AntibodiesCatalog NumberProviderUsage (Dilution)
      p44/42 MAPK (ERK1/2)9102Cell Signaling (Danvers, MA)Western blotting (1/1,000)
      Phosphorylated p44/42 MAPK (ERK1/2) (Thr202/Tyr204)9101Cell Signaling (Danvers, MA)Western blotting (1/1,000)
      STAT3 (124H6)9139Cell signaling (Danvers, MA)Western blotting (1/1,000)
      Phosphorylated STAT3 (Tyr705) (58E12)9135Cell Signaling (Danvers, MA)Western blotting (1/1,000)
      STAT69362Cell Signaling (Danvers, MA)Western blotting (1/1,000)
      Phosphorylated STAT6 (Tyr641)9361Cell Signaling (Danvers, MA)Western blotting (1/1,000)
      Anti-beta Actin (Ac-15)ab6276Abcam (Tokyo, Japan)Western blotting (1/10,000)
      Ant-IL-33 (Human)PM033MBL (Tokyo, Japan)Western blotting (1/1,000) Immunofluorescence (1/200)
      Filaggrin (M-290)sc-30230Santa Cruz Biotechnology (Dallas, TX)Western blotting (1/500)
      Filaggrin (AKH1)sc-66192Santa Cruz Biotechnology (Dallas, TX)Immunofluorescence (1/100)
      Loricrin (C-13)sc-51130Santa Cruz Biotechnology (Dallas, TX)Western blotting (1/500) Immunofluorescence (1/100)
      Keratin 1-Specific16848-1-APProteintech (Tokyo, Japan)Western blotting (1/1,000) Immunofluorescence (1/100)
      Keratin 10 Ab-2MS-611-P1Thermo Fisher Scientific (Yokohama, Japan)Western blotting (1/500) Immunofluorescence (1/100)
      RUNX1 (A-2)sc-365644Santa Cruz Biotechnology (Dallas, TX)Western blotting (1/500) Immunohistochemistry (1/100) Immunofluorescence (1/100)
      Lamin A613501BioLegend (San Diego, CA)Western blotting (1/200)
      Phosphorylated STAT3 (Tyr705) (D3A7) XP9145Cell Signaling (Danvers, MA)ChIP assay (1/100)
      Histone H32650Cell Signaling (Danvers, MA)ChIP assay (1/100)

      Western blotting (1/5,000)
      Cytochrome C (D18C7)11940Cell Signaling (Danvers, MA)Western blotting (1/1,000)
      Abbreviations: ChIP, chromatin immunoprecipitation; ERK, extracellular signal‒regulated kinase; STAT, signal transducer and activator of transcription.

      References

        • Ali S.
        • Mohs A.
        • Thomas M.
        • Klare J.
        • Ross R.
        • Schmitz M.L.
        • et al.
        The dual function cytokine IL-33 interacts with the transcription factor NF-κB to dampen NF-κB-stimulated gene transcription.
        J Immunol. 2011; 187: 1609-1616
        • Amano W.
        • Nakajima S.
        • Kunugi H.
        • Numata Y.
        • Kitoh A.
        • Egawa G.
        • et al.
        The Janus kinase inhibitor JTE-052 improves skin barrier function through suppressing signal transducer and activator of transcription 3 signaling.
        J Allergy Clin Immunol. 2015; 136: 667-677.e7
        • Baekkevold E.S.
        • Roussigné M.
        • Yamanaka T.
        • Johansen F.E.
        • Jahnsen F.L.
        • Amalric F.
        • et al.
        Molecular characterization of NF-HEV, a nuclear factor preferentially expressed in human high endothelial venules.
        Am J Pathol. 2003; 163: 69-79
        • Balato A.
        • Raimondo A.
        • Balato N.
        • Ayala F.
        • Lembo S.
        Interleukin-33: increasing role in dermatological conditions.
        Arch Dermatol Res. 2016; 308: 287-296
        • Brandt E.B.
        • Sivaprasad U.
        Th2 cytokines and atopic dermatitis.
        J Clin Cell Immunol. 2011; 2: 110
        • Braun T.
        • Woollard A.
        RUNX factors in development: lessons from invertebrate model systems.
        Blood Cells Mol Dis. 2009; 43: 43-48
        • Cabanillas B.
        • Novak N.
        Atopic dermatitis and filaggrin.
        Curr Opin Immunol. 2016; 42: 1-8
        • Calautti E.
        • Avalle L.
        • Poli V.
        Psoriasis: a STAT3-centric view.
        Int J Mol Sci. 2018; 19: 171
        • Cayrol C.
        • Duval A.
        • Schmitt P.
        • Roga S.
        • Camus M.
        • Stella A.
        • et al.
        Environmental allergens induce allergic inflammation through proteolytic maturation of IL-33.
        Nat Immunol. 2018; 19: 375-385
        • Cayrol C.
        • Girard J.P.
        Interleukin-33 (IL-33): A nuclear cytokine from the IL-1 family.
        Immunol Rev. 2018; 281: 154-168
        • Cornelissen C.
        • Marquardt Y.
        • Czaja K.
        • Wenzel J.
        • Frank J.
        • Lüscher-Firzlaff J.
        • et al.
        IL-31 regulates differentiation and filaggrin expression in human organotypic skin models.
        J Allergy Clin Immunol. 2012; 129: 426-433.e4338
        • Dai X.
        • Tohyama M.
        • Murakami M.
        • Shiraishi K.
        • Liu S.
        • Mori H.
        • et al.
        House dust mite allergens induce interleukin 33 (IL-33) synthesis and release from keratinocytes via ATP-mediated extracellular signaling.
        Biochim Biophys Acta Mol Basis Dis. 2020; 1866: 165719
        • Dajnoki Z.
        • Béke G.
        • Mócsai G.
        • Kapitány A.
        • Gáspár K.
        • Hajdu K.
        • et al.
        Immune-mediated skin inflammation is similar in severe atopic dermatitis patients with or without filaggrin mutation.
        Acta Derm Venereol. 2016; 96: 645-650
        • Gooderham M.J.
        • Hong H.C.
        • Eshtiaghi P.
        • Papp K.A.
        Dupilumab: a review of its use in the treatment of atopic dermatitis.
        J Am Acad Dermatol. 2018; 78: S28-S36
        • Harb H.
        • Chatila T.A.
        Mechanisms of dupilumab.
        Clin Exp Allergy. 2020; 50: 5-14
        • Imai Y.
        Interleukin-33 in atopic dermatitis.
        J Dermatol Sci. 2019; 96: 2-7
        • Imai Y.
        • Yasuda K.
        • Nagai M.
        • Kusakabe M.
        • Kubo M.
        • Nakanishi K.
        • et al.
        IL-33-induced atopic dermatitis-like inflammation in mice is mediated by group 2 innate lymphoid cells in concert with basophils.
        J Invest Dermatol. 2019; 139: 2185-2194.e3
        • Imai Y.
        • Yasuda K.
        • Sakaguchi Y.
        • Haneda T.
        • Mizutani H.
        • Yoshimoto T.
        • et al.
        Skin-specific expression of IL-33 activates group 2 innate lymphoid cells and elicits atopic dermatitis-like inflammation in mice.
        Proc Natl Acad Sci USA. 2013; 110: 13921-13926
        • Kezic S.
        • Jakasa I.
        IFilaggrin and skin barrier function.
        Curr Probl Dermatol. 2016; 49: 1-7
        • Kim J.H.
        • Bae H.C.
        • Ko N.Y.
        • Lee S.H.
        • Jeong S.H.
        • Lee H.
        • et al.
        Thymic stromal lymphopoietin downregulates filaggrin expression by signal transducer and activator of transcription 3 (STAT3) and extracellular signal-regulated kinase (ERK) phosphorylation in keratinocytes.
        J Allergy Clin Immunol. 2015; 136: 205-208.e9
        • Lee H.
        • Shin J.J.
        • Bae H.C.
        • Ryu W.I.
        • Son S.W.
        Toluene downregulates filaggrin expression via the extracellular signal-regulated kinase and signal transducer and activator of transcription-dependent pathways.
        J Allergy Clin Immunol. 2017; 139: 355-358.e5
        • Lim K.H.
        • Ramakrishna S.
        • Baek K.H.
        Molecular mechanisms and functions of cytokine-inducible deubiquitinating enzymes.
        Cytokine Growth Factor Rev. 2013; 24: 427-431
        • Masse I.
        • Barbollat-Boutrand L.
        • Molina M.
        • Berthier-Vergnes O.
        • Joly-Tonetti N.
        • Martin M.T.
        • et al.
        Functional interplay between p63 and p53 controls RUNX1 function in the transition from proliferation to differentiation in human keratinocytes.
        Cell Death Dis. 2012; 3: e318
        • Meephansan J.
        • Komine M.
        • Tsuda H.
        • Karakawa M.
        • Tominaga S.
        • Ohtsuki M.
        Expression of IL-33 in the epidermis: the mechanism of induction by IL-17.
        J Dermatol Sci. 2013; 71: 107-114
        • Meephansan J.
        • Tsuda H.
        • Komine M.
        • Tominaga S.
        • Ohtsuki M.
        Regulation of IL-33 expression by IFN-γ and tumor necrosis factor-α in normal human epidermal keratinocytes.
        J Invest Dermatol. 2012; 132: 2593-2600
        • Ni Y.
        • Tao L.
        • Chen C.
        • Song H.
        • Li Z.
        • Gao Y.
        • et al.
        The deubiquitinase USP17 regulates the stability and nuclear function of IL-33.
        Int J Mol Sci. 2015; 16: 27956-27966
        • Omori-Miyake M.
        • Yamashita M.
        • Tsunemi Y.
        • Kawashima M.
        • Yagi J.
        In vitro assessment of IL-4- or IL-13-mediated changes in the structural components of keratinocytes in mice and humans [published correction appears in J Invest Dermatol 2018;138:472–3].
        J Invest Dermatol. 2014; 134: 1342-1350
        • Ryu W.I.
        • Lee H.
        • Bae H.C.
        • Ryu H.J.
        • Son S.W.
        IL-33 down-regulates filaggrin expression by inducing STAT3 and ERK phosphorylation in human keratinocytes.
        J Dermatol Sci. 2016; 82: 131-134
        • Sano S.
        • Chan K.S.
        • DiGiovanni J.
        Impact of Stat3 activation upon skin biology: a dichotomy of its role between homeostasis and diseases.
        J Dermatol Sci. 2008; 50: 1-14
        • Sehra S.
        • Yao Y.
        • Howell M.D.
        • Nguyen E.T.
        • Kansas G.S.
        • Leung D.Y.
        • et al.
        IL-4 regulates skin homeostasis and the predisposition toward allergic skin inflammation.
        J Immunol. 2010; 184: 3186-3190
        • Tamagawa-Mineoka R.
        • Okuzawa Y.
        • Masuda K.
        • Katoh N.
        Increased serum levels of interleukin 33 in patients with atopic dermatitis.
        J Am Acad Dermatol. 2014; 70: 882-888
        • Tao L.
        • Chen C.
        • Song H.
        • Piccioni M.
        • Shi G.
        • Li B.
        Deubiquitination and stabilization of IL-33 by USP21.
        Int J Clin Exp Pathol. 2014; 7: 4930-4937
        • Tsuchisaka A.
        • Furumura M.
        • Hashimoto T.
        Cytokine regulation during epidermal differentiation and barrier formation.
        J Invest Dermatol. 2014; 134: 1194-1196