Advertisement
Journal of Investigative Dermatology Home

Culprit Drugs Induce Specific IL-36 Overexpression in Acute Generalized Exanthematous Pustulosis

  • Barbara Meier-Schiesser
    Affiliations
    Department of Dermatology, University Hospital, Zürich, Switzerland

    Faculty of Medicine, University of Zürich, Zürich, Switzerland
    Search for articles by this author
  • Laurence Feldmeyer
    Affiliations
    Department of Dermatology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
    Search for articles by this author
  • Author Footnotes
    9 Current affiliation: Novartis Pharma AG, Basel, Switzerland.
    Dragana Jankovic
    Footnotes
    9 Current affiliation: Novartis Pharma AG, Basel, Switzerland.
    Affiliations
    Department of Dermatology, University Hospital, Zürich, Switzerland
    Search for articles by this author
  • Mark Mellett
    Affiliations
    Department of Dermatology, University Hospital, Zürich, Switzerland
    Search for articles by this author
  • Takashi K. Satoh
    Affiliations
    Department of Dermatology, University Hospital, Zürich, Switzerland
    Search for articles by this author
  • Daniel Yerly
    Affiliations
    Department of Rheumatology, Clinical Immunology and Allergology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
    Search for articles by this author
  • Author Footnotes
    10 Current affiliation: Department of Dermatology, University Hospital Basel, Basel, Switzerland.
    Alexander Navarini
    Footnotes
    10 Current affiliation: Department of Dermatology, University Hospital Basel, Basel, Switzerland.
    Affiliations
    Department of Dermatology, University Hospital, Zürich, Switzerland

    Faculty of Medicine, University of Zürich, Zürich, Switzerland
    Search for articles by this author
  • Riichiro Abe
    Affiliations
    Division of Dermatology, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan
    Search for articles by this author
  • Nikhil Yawalkar
    Affiliations
    Department of Dermatology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland
    Search for articles by this author
  • Wen-Hung Chung
    Affiliations
    Department of Dermatology, Drug Hypersensitivity Clinical and Research Center, Chang Gung Memorial Hospital, Taipei and Linkou, Taiwan
    Search for articles by this author
  • Author Footnotes
    7 Current affiliation: Department of Dermatology and Allergology, University Hospital Munich, Munich, Germany.
    ,
    Author Footnotes
    8 These authors contributed equally to this work.
    Lars E. French
    Footnotes
    7 Current affiliation: Department of Dermatology and Allergology, University Hospital Munich, Munich, Germany.
    8 These authors contributed equally to this work.
    Affiliations
    Department of Dermatology, University Hospital, Zürich, Switzerland

    Faculty of Medicine, University of Zürich, Zürich, Switzerland
    Search for articles by this author
  • Author Footnotes
    8 These authors contributed equally to this work.
    Emmanuel Contassot
    Correspondence
    Correspondence: Emmanuel Contassot, Dermatology Department, University Hospital, Gloriastrasse 31, Zürich, Switzerland.
    Footnotes
    8 These authors contributed equally to this work.
    Affiliations
    Department of Dermatology, University Hospital, Zürich, Switzerland

    Faculty of Medicine, University of Zürich, Zürich, Switzerland
    Search for articles by this author
  • Author Footnotes
    7 Current affiliation: Department of Dermatology and Allergology, University Hospital Munich, Munich, Germany.
    8 These authors contributed equally to this work.
    9 Current affiliation: Novartis Pharma AG, Basel, Switzerland.
    10 Current affiliation: Department of Dermatology, University Hospital Basel, Basel, Switzerland.
Open ArchivePublished:November 02, 2018DOI:https://doi.org/10.1016/j.jid.2018.10.023
      Acute generalized exanthematous pustulosis (AGEP) is a severe adverse cutaneous drug reaction. Although an involvement of drug-specific T cells has been reported, the physiopathology of AGEP and mechanism of neutrophilic skin inflammation remain incompletely understood. Recently, mutations in IL-36RN, the gene encoding the IL-36 receptor antagonist, have been reported to be more frequent in AGEP patients and pustular psoriasis. Here, we show that IL-36 cytokines, in particular IL-36γ, are highly expressed in lesional skin of AGEP patients, keratinocytes and macrophages being a major source of IL-36γ. Such an IL-36γ overexpression was not observed in patients with drug-induced maculopapular rash. In vitro, the causative drug specifically induced IL-36γ release either directly by the patient’s peripheral blood monocytes or indirectly by keratinocytes in the presence of autologous peripheral blood mononuclear cells. Such culprit drug induction of IL-36γ secretion in vitro was specific for AGEP and involved toll-like receptor 4 sensing the drug/albumin complex as a danger signal. Our results suggest that IL-36γ secretion by monocytes/macrophages and keratinocytes in response to culprit drug exposure likely plays a key role in the pathogenesis of AGEP.

      Abbreviations:

      ADR (adverse drug reaction), AGEP (acute generalized exanthematous pustulosis), ELISPOT (enzyme-linked immunospot), MPR (maculopapular rash), PBMC (peripheral blood mononuclear cell), Th (T helper), TLR (toll-like receptor)

      Introduction

      Acute generalized exanthematous pustulosis (AGEP) is a severe cutaneous adverse drug reaction (ADR) caused mainly by antibiotics, antimalarials, and antifungals (
      • Sidoroff A.
      • Halevy S.
      • Bavinck J.N.
      • Vaillant L.
      • Roujeau J.C.
      Acute generalized exanthematous pustulosis (AGEP)—a clinical reaction pattern.
      ). Clinically, AGEP is characterized by a disseminated eruption of sterile nonfollicular pustules on the background of a widespread erythematous skin eruption, accompanied by fever and peripheral blood neutrophilia. To date, the pathophysiology of AGEP is not fully elucidated. However, AGEP is currently considered to be a T cell-mediated disease (
      • Pichler W.J.
      Delayed drug hypersensitivity reactions.
      ) and is currently seen as a delayed hypersensitivity reaction. Indeed, drug-specific T cells are suspected to play a central role in AGEP, as evidenced by the high levels of T-cell stimulation induced by causative (culprit) drugs as measured by the lymphocyte transformation test (
      • Anliker M.D.
      • Wuthrich B.
      Acute generalized exanthematous pustulosis due to sulfamethoxazol with positive lymphocyte transformation test (LTT).
      ). Furthermore, drug-specific CD4+ and CD8+ T cells have been derived in vitro from AGEP patients’ peripheral blood. Most of these drug-specific T cells (
      • Nishio D.
      • Izu K.
      • Kabashima K.
      • Tokura Y.
      T cell populations propagating in the peripheral blood of patients with drug eruptions.
      ) produce IL-8 (
      • Britschgi M.
      • Pichler W.J.
      Acute generalized exanthematous pustulosis, a clue to neutrophil-mediated inflammatory processes orchestrated by T cells.
      ,
      • Britschgi M.
      • Steiner U.C.
      • Schmid S.
      • Depta J.P.
      • Senti G.
      • Bircher A.
      • et al.
      T-cell involvement in drug-induced acute generalized exanthematous pustulosis.
      ), a powerful neutrophil chemoattractant. IL-8–producing T cells are, therefore, considered to be the cause of neutrophil survival and recruitment to the skin during the course of the disease (
      • Schaerli P.
      • Britschgi M.
      • Keller M.
      • Steiner U.C.
      • Steinmann L.S.
      • Moser B.
      • et al.
      Characterization of human T cells that regulate neutrophilic skin inflammation.
      ). It has also been suggested that T helper (Th) type 17 effector cytokines, namely IL-17 and IL-22, stimulate keratinocytes to produce IL-8, and that keratinocyte-derived IL-8 also contributes to neutrophil accumulation in AGEP patients’ skin (
      • Kabashima R.
      • Sugita K.
      • Sawada Y.
      • Hino R.
      • Nakamura M.
      • Tokura Y.
      Increased circulating Th17 frequencies and serum IL-22 levels in patients with acute generalized exanthematous pustulosis.
      ). Hence, T-cell– and keratinocyte-derived IL-8 has been proposed to be responsible for the recruitment of neutrophils to the intraepithelial pustules.
      AGEP shares certain clinical and histological features with pustular psoriasis. Recently, genetic studies identified mutations in the IL-36 receptor antagonist gene (IL-36RN) in patients with general pustular psoriasis (
      • Kanazawa N.
      • Nakamura T.
      • Mikita N.
      • Furukawa F.
      Novel IL36RN mutation in a Japanese case of early onset generalized pustular psoriasis.
      ,
      • Marrakchi S.
      • Guigue P.
      • Renshaw B.R.
      • Puel A.
      • Pei X.Y.
      • Fraitag S.
      • et al.
      Interleukin-36-receptor antagonist deficiency and generalized pustular psoriasis.
      ,
      • Onoufriadis A.
      • Simpson M.A.
      • Pink A.E.
      • Di Meglio P.
      • Smith C.H.
      • Pullabhatla V.
      • et al.
      Mutations in IL36RN/IL1F5 are associated with the severe episodic inflammatory skin disease known as generalized pustular psoriasis.
      ,
      • Sugiura K.
      • Takeichi T.
      • Kono M.
      • Ogawa Y.
      • Shimoyama Y.
      • Muro Y.
      • et al.
      A novel IL36RN/IL1F5 homozygous nonsense mutation, p.Arg10X, in a Japanese patient with adult-onset generalized pustular psoriasis.
      ). Mutations in IL36RN have also been reported in AGEP (
      • Navarini A.A.
      • Valeyrie-Allanore L.
      • Setta-Kaffetzi N.
      • Barker J.N.
      • Capon F.
      • Creamer D.
      • et al.
      Rare variations in IL36RN in severe adverse drug reactions manifesting as acute generalized exanthematous pustulosis.
      ), suggesting that IL-36 signaling dysregulation is involved in the physiopathology of both pustular diseases. IL-36 receptor antagonist (IL-36Ra) inhibits the binding of IL-36α, β, and γ to their receptor. A mutation in IL36RN may therefore result in exacerbated IL-36 signaling, leading to the production of IL-1, IL-6, and IL-8 and subsequent neutrophilic skin infiltration with pustule formation.
      IL-36 cytokines have emerged as important cytokines mediating inflammatory responses in the skin. Indeed, several reports have shown that all IL-36 isoforms are overexpressed in psoriatic skin (
      • Debets R.
      • Timans J.C.
      • Homey B.
      • Zurawski S.
      • Sana T.R.
      • Lo S.
      • et al.
      Two novel IL-1 family members, IL-1 delta and IL-1 epsilon, function as an antagonist and agonist of NF-kappa B activation through the orphan IL-1 receptor-related protein 2.
      ,
      • Gresnigt M.S.
      • van de Veerdonk F.L.
      Biology of IL-36 cytokines and their role in disease.
      ,
      • He Q.
      • Chen H.X.
      • Li W.
      • Wu Y.
      • Chen S.J.
      • Yue Q.
      • et al.
      IL-36 cytokine expression and its relationship with p38 MAPK and NF-kappaB pathways in psoriasis vulgaris skin lesions.
      ,
      • Johnston A.
      • Xing X.
      • Guzman A.M.
      • Riblett M.
      • Loyd C.M.
      • Ward N.L.
      • et al.
      IL-1F5, -F6, -F8, and -F9: a novel IL-1 family signaling system that is active in psoriasis and promotes keratinocyte antimicrobial peptide expression.
      ). In keratinocytes, production of IL-36 family members can be induced by tumor necrosis factor, IL-17, and IL-22, cytokines known to be involved in psoriasis, which further supports an important role for IL-36 in this disease (
      • Carrier Y.
      • Ma H.L.
      • Ramon H.E.
      • Napierata L.
      • Small C.
      • O’Toole M.
      • et al.
      Inter-regulation of Th17 cytokines and the IL-36 cytokines in vitro and in vivo: implications in psoriasis pathogenesis.
      ). Recently, it has been reported that antigen-presenting cells located in human skin express high levels of IL-36R and are responsive to IL-36 cytokines (
      • Dietrich D.
      • Martin P.
      • Flacher V.
      • Sun Y.
      • Jarrossay D.
      • Brembilla N.
      • et al.
      Interleukin-36 potently stimulates human M2 macrophages, Langerhans cells and keratinocytes to produce pro-inflammatory cytokines.
      ), an observation that further supports an important role of IL-36 in skin biology.
      Here, we provide evidence that drugs causing AGEP can specifically trigger IL-36 cytokine production by peripheral blood monocytes via toll-like receptor 4 and by keratinocytes from AGEP patients and subsequently induce IL-8 by peripheral blood mononuclear cells in an IL-36–dependent manner. These observations are supportive of a drug-specific dysregulation of IL-36 signaling as a possible driver of AGEP pathogenesis and identify myeloid cells and keratinocytes as important players in AGEP.

      Results

       Enhanced IL-36 expression in lesional skin of AGEP patients

      First, we performed gene expression profiling (Gene Expression Omnibus accession number GSE121421) using Affymetrix (Waltham, MA) Human Exon 1.0 ST chips (see the Supplementary Materials online for description). Total RNA was extracted from lesional skin from patients with AGEP (n = 8) or maculopapular rash (MPR) (n = 6), and from skin from healthy individuals (n = 7). The hierarchical clustering of genes differentially expressed in AGEP, MPR, and healthy skin showed a perfect segregation of these three conditions based on the gene expression profile of individual samples and the magnitude of change in gene expression (Figure 1a, and see Supplementary Table S1 online). IL36G (IL1F9), the gene encoding IL-36γ, was found to be overexpressed in AGEP skin compared with healthy skin (ratio = 3.06, P = 0.0006) (see Supplementary Table S2 online) and MPR (ratio = 2.54, P = 0.005) (see Supplementary Table S3 online). Using quantitative PCR, we then analyzed the expression of IL-36 in skin biopsy samples from patients with AGEP (n = 16) and MPR, a milder form of cutaneous reaction to drugs (n = 16). The expressions of IL36A, IL36B and IL36G, encoding IL-36α, β, and γ, respectively, were found to be up-regulated in AGEP skin compared with MPR (IL36A: 15-fold, P < 0.001; IL36B: 28-fold, P < 0.001; IL36G: 32-fold, P < 0.001) and normal skin (IL36A: 7-fold, P < 0.001; IL36B: 1.35-fold, P < 0.05; IL36G: 3-fold, P < 0.001) (Figure 1b). The expression of IL36RN was found to be higher (4.5-fold, P < 0.001) in AGEP than MPR skin, but it was not different from that found in normal skin. IL1RL2, encoding IL-36 receptor, was similarly expressed in both diseases.
      Figure thumbnail gr1
      Figure 1IL-36α and IL-36β are overexpressed in pustular regions of AGEP lesional skin. (a) Gene expression hierarchical clustering in lesional skin biopsy samples from patients suffering from AGEP (n = 8) and MPR (n = 6) analyzed using Affymetrix (Santa Clara, CA) Human Exon 1.0 ST chips and (b) quantitative real-time reverse transcription–PCR analysis of genes encoding IL-36α, IL-36β, IL-36γ, IL-36R, and IL-36RN in skin biopsy samples from AGEP patients (n = 16), MPR (n = 16), and normal skin (n = 5). (c) Immunohistochemical analysis of lesional skin biopsy samples showed that IL-36α and IL-36γ are overexpressed at the site of pustules in AGEP and PP controls. Representative pictures of 18 AGEP, 10 PP, and 18 MPR cases are shown. Scale bars in the large panels = 500 μm; scale bars in the small panels = 100 μm). Neither IL-36α nor IL-36γ were detectable in normal skin samples from healthy donors (see a). Bottom panels show the semiquantitative evaluation of IL-36α and IL-36γ labeling by immunohistochemistry in the pustular epidermis, nonpustular epidermis, and dermis (see b) of lesional skin biopsy samples from patients with AGEP (n = 18), PP (n = 10), MPR (n = 18), and normal skin (NS) from healthy donors (n=5). Expression levels were qualified as very strong (++++), strong (+++), moderate (++), weak (+) or absent (0). P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. AGEP, acute generalized exanthematous pustulosis; MPR, maculopapular rash; NS, normal skin; PP, pustular psoriasis.
      The relative expression of IL36β, however, was low compared with IL36α and IL36γ. Therefore, we further analyzed the expression and tissue distribution of IL-36α and γ by labeling AGEP and MPR lesional skin biopsy samples with antibodies to both cytokines. IL-36α and γ were found to be highly expressed in pustular and peri-pustular regions of AGEP skin (Figure 1c). Both keratinocytes and dermal immune cells infiltrating pustular areas were found to produce IL-36α and γ. In contrast, IL-36γ was weakly expressed in MPR skin biopsy samples, and IL-36α was barely detectable. In AGEP lesional skin, IL-36γ was predominantly expressed by keratinocytes in the epidermis and by immune cells infiltrating the dermis in pustular areas. Also, IL-36γ staining was particularly intense in nonpustular epidermis surrounding the pustules in AGEP. As previously reported (
      • Song H.S.
      • Kim S.J.
      • Park T.I.
      • Jang Y.H.
      • Lee E.S.
      Immunohistochemical comparison of IL-36 and the IL-23/Th17 axis of generalized pustular psoriasis and acute generalized exanthematous pustulosis.
      ), IL-36α was also expressed in pustules from patients suffering from pustular psoriasis and, as in AGEP biopsy samples, IL-36γ was the predominant form in the epidermis (Figure 1c). Neither IL-36α nor IL-36γ could be detected in skin specimens from healthy donors (see Supplementary Figure S1 online).

       IL-36 production by keratinocytes and macrophages in lesional skin of AGEP patients

      To identify which cell types release IL-36α and γ in AGEP skin, we co-labeled AGEP biopsy sections with antibodies to IL-36α and γ and antibodies to CD68 to identify macrophages and with antibodies to CD3 to identify T cells (Figure 2). IL-36α was predominantly expressed by CD68+ macrophages (41.17% ± 4.71% of positively labeled cells) and, to a lesser extent, by keratinocytes and CD3+ T cells (27.33% ± 2.17% and 22.33% ± 2.39% of positively labeled cells, respectively). IL-36γ was predominantly expressed by keratinocytes and CD68+ macrophages (44.08% ± 4.47% and 43.17% ± 3.31% of positively labeled cells, respectively) and, to a lesser extent, by CD3+ T cells (15.17% ± 1.26% of positively labeled cells). In contrast, IL-36α and IL-36γ were not expressed by neutrophils present in pustular areas (see Supplementary Figure S2 online).
      Figure thumbnail gr2
      Figure 2IL-36γ is expressed by keratinocytes and immune cells in AGEP skin. (a, b) Pustule-containing sections of AGEP skin samples were co-labeled with antibodies to (a) CD68 (green), CD3 (green), and IL-36α (red) (b) or IL-36γ (red). Scale bars in the large panels = 500 μm; scale bars in the small panels = 20 μm). (c) The percentages of epidermal cells (keratinocytes), CD68, and CD3-labeled cells among the IL-36α– or IL-36γ–labeled cells were determined. Cell nuclei appear in blue (DAPI). The mean ± standard deviation of 12 examined patients is shown (patients 8–10 and 20–28). P < 0.05, ∗∗P < 0.01; ∗∗∗P < 0.001.

       Rapid culprit drug-specific induction of IL-36γ in monocytes in AGEP

      Immunofluorescence analyses of AGEP biopsy samples suggested that, in addition to keratinocytes, immune cells may also represent an important source of IL-36γ in the skin. Therefore, we assessed whether peripheral blood mononuclear cells (PBMCs) taken from AGEP patients 8, 9, and 20 more than 6 months after the ADR—when in complete remission—were still able to respond directly to the causative drug. To this end, PBMCs from patients who had experienced AGEP were exposed to their respective causative drug or an irrelevant control drug (i.e., one that did not cause AGEP in the tested patients) in enzyme-linked immunospot (ELISPOT) plates. The number of IL-36γ+ spots was counted after 1, 2, 4, 6, and 8 hours. IL-36γ release was already detected in PBMCs from AGEP patients after 1 hour of exposure to their respective culprit drugs (1.7-fold more spots than control drug, P < 0.001) and reached a plateau after 4 hours (2.7-fold more spots than control drug, P < 0.001) (Figure 3a, left panel), whereas no increase in IL-36γ production was observed with control drugs that did not cause AGEP in the tested patients (Figure 3a, right panel, and see Supplementary Figure S3 online). In contrast, none of the culprit drugs involved in AGEP induced IL-36γ secretion in PBMCs from MPR patients with proven sensitization to the same drug. Similarly, none of the drugs used was able to induce IL-36γ secretion in PBMCs from healthy blood donors. IL-36γ production was concentration dependent and was exclusively observed in PBMCs from patients exposed to the drug that caused their AGEP (see Supplementary Figure S3), whereas none of the drugs used in this study was able to induce IL-36γ in drug-sensitized MPR patients’ or healthy donors’ PBMCs, even at the highest concentration.
      Figure thumbnail gr3
      Figure 3PBMCs and monocytes taken from AGEP patients more than 6 months after the ADR selectively secrete IL-36γ in response to culprit drug exposure. (a) PBMCs from AGEP or MPR patients and from healthy donors were cultured in IL-36γ ELISPOT plates for 8 hours in the presence of the culprit drug having caused AGEP or MPR, respectively (left panel) or a control drug (right panel). The number of spots was counted 1, 2, 4, 6, and 8 hours after drug exposure. The means ± standard deviation of three different AGEP patients (nos. 8, 10, and 21), three MPR patients, and three healthy blood donors are shown. (b) CD14+ monocytes and CD3+ T cells were isolated from AGEP blood with a purity of greater than 96% (right panel) and were cultured in IL-36γ ELISPOT plates (left panel) in the presence of the culprit drug or a control drug and compared with total PBMCs. The number of spots was counted 8 hours after drug exposure (lower panel). The means ± standard deviation of three different patients (patient nos. 8, 9, and 10) are shown. ∗∗P < 0.01, ∗∗∗P < 0.001. ADR, adverse drug reaction; AGEP, acute generalized exanthematous pustulosis; ELISPOT, enzyme-linked immunospot; MPR, maculopapular rash; n.s., not significant; PBMC, peripheral blood mononuclear cell.
      Because immunofluorescence analyses of AGEP skin lesions showed that macrophages and, to a lesser extent, T cells are possible sources of IL-36, we sorted CD14+ (monocytes) and CD3+ cells (T cells) from patients’ PBMCs and exposed them to culprit or control drug to evaluate IL-36γ secretion by ELISPOT (Figure 3b). Whereas the control drug did not induce IL-36γ secretion in PBMCs or in CD14+ and CD3+ cells, the culprit drug was able to induce IL-36γ secretion in CD14+ cells at levels similar to those observed with total PBMCs. In contrast, the culprit drug was not able to induce IL-36γ secretion in CD3+ PBMCs. These results suggest that PBMCs from patients who experienced AGEP are able to respond directly and specifically to the causative drug and that monocytes/macrophages are an important source of IL-36 production in response to culprit drug exposure.

       IL-36 induction by culprit drugs in keratinocytes and PBMCs from patients who experienced AGEP

      By immunolabeling, both keratinocytes and macrophages were positively labeled for IL-36α and γ in pustular areas of AGEP biopsy samples (Figure 2). To assess a possible cross-talk between macrophages and keratinocytes and the relative contribution of each cell type to drug-induced IL-36 secretion, we performed co-culture experiments using autologous hair follicle-derived keratinocytes and PBMCs from three patients (patients 8, 9, and 20) who experienced AGEP but were in remission (i.e., at least 6 months after disappearance of symptoms) at the time of blood and hair collection. Keratinocytes and PBMCs were cultured either alone or together using a Transwell culture system (Corning, Corning, NY). To formally exclude the contamination of one cell type by the other, we confirmed the purity of each cell type by quantitative PCR using primers to CD45 (PBMC marker) and ITGA6 (keratinocyte marker) on both PBMCs and keratinocytes after a 6-hour co-culture (see Supplementary Figure S4 online). When keratinocytes and PBMCs were each cultured alone, only PBMCs showed elevated IL-36γ mRNA levels after exposure to the culprit drug. Both keratinocytes and PBMCs showed culprit drug-specific IL-36γ gene expression when cultured together (Figure 4). The addition of soluble IL-36Ra did not reduce IL-36 expression, suggesting that, at the analyzed time point, there was no auto- or paracrine amplification and/or induction of IL-36 gene expression by IL-36 released by PBMCs or keratinocytes. Taken together, these results indicate that, in culprit drug-induced responses of patient-derived keratinocytes and monocytes, IL-36γ is the predominant form of IL-36 induced. Furthermore, in response to culprit drug exposure, IL-36 is expressed directly by PBMCs and indirectly by keratinocytes.
      Figure thumbnail gr4
      Figure 4IL-36γ is specifically induced by culprit drugs in PBMCs and keratinocytes and is dependent on TLR4 and albumin. (a) Autologous PBMCs and keratinocytes from patients who experienced AGEP but were in remission at the time of blood and hair collection were cultured either alone or together in a Transwell system (Corning, Corning, NY), allowing for soluble factor-mediated interactions. Patients’ cells were exposed to the culprit drug (amoxicillin 1 mg/ml, patient 8; letrozole 100 nmol/L, patient 9; vancomycin 500 μg/ml, patient 20) or a control drug (metamizole 100 μg/ml, patient 8; carbamazepine 10 μg/ml, patient 9; amoxicillin 1 mg/ml, patient 20) in the presence or absence of IL-36RA (1 μg/ml). After 6 hours of culture, RNA was extracted from KCs (left panels) and PBMCs (right panels) to measure IL-36γ mRNA levels. Relative IL-36γ expression in KCs and PBMCs cultured either alone or in co-culture as indicated and exposed to the culprit drug or a control drug in the presence or absence of IL-36Ra are shown for each tested patient. Means ± standard deviation of three replicates are shown. (b) PBMCs from AGEP patients were pretreated with vehicle, MyD88 inhibitory peptide, neutralizing antibodies to TLR2 and TLR4, or recombinant IL-1 receptor antagonist (IL-1RA) and exposed to the culprit drug for 6 hours. An ELISPOT assay was performed, and the number of IL-36γ+ spots were counted using an ELISPOT reader. Untreated cells or cells exposed to an irrelevant control drug were used as basal IL-36γ secretion controls. Means ± standard deviation of the five tested patients are shown. ∗∗P < 0.01, ∗∗∗P < 0.001. (c) PBMCs from AGEP patients were exposed to culprit drug for 6 hours in medium containing 10% FBS, in serum-free medium, or in serum-free medium supplemented with 2.5 mg/ml human serum albumin as indicated. An ELISPOT assay was performed, and the numbers of IL-36γ+ spots were counted using an ELISPOT reader. Untreated cells were used as basal IL-36γ secretion controls. Means ± standard deviation of the five tested patients is shown. P < 0.05, ∗∗∗P < 0.001. Representative IL-36γ ELISPOT pictures are shown below the graph. AGEP, acute generalized exanthematous pustulosis; ELISPOT, enzyme-linked immunospot; FBS, fetal bovine serum; HSA, human serum albumin; inh., inhibitory peptide; KC, keratinocyte; n.d., not determined; PBMC, peripheral blood mononuclear cell; TLR4, toll-like receptor 4; w/o, without.

       IL-36 secretion by AGEP patients’ PBMCs is induced by toll-like receptor 4 recognition and requires albumin

      Toll-like receptors (TLRs) are innate immunity receptors signaling through NF-κB and lead to the transcription of inflammatory cytokines, including IL-36γ (
      • Aschan J.
      • Ringden O.
      • Tollemar J.
      • Ljungman P.
      • Lonnqvist B.
      • Pihlstedt P.
      • et al.
      Improved survival in marrow recipients above 30 years of age with better prevention of graft-versus-host disease.
      ,
      • Lian L.H.
      • Milora K.A.
      • Manupipatpong K.K.
      • Jensen L.E.
      The double-stranded RNA analogue polyinosinic-polycytidylic acid induces keratinocyte pyroptosis and release of IL-36gamma.
      ). A treatment of AGEP patients’ PBMCs exposed to the culprit drug with an inhibitory peptide to the TLR adapter protein MYD88 (Pepinh-MYD) led to a decreased IL-36γ secretion compared with vehicle. Moreover, the inhibition of TLR4 signaling with a neutralizing antibody also resulted in decreased IL-36γ secretion by PBMCs exposed to the culprit drug, whereas a neutralizing antibody to TLR2 had no effect on IL-36γ expression (Figure 4b). Because MyD88 is also involved in IL-1RI/IL-1 signaling, we also treated drug-exposed PBMCs from AGEP patients with a recombinant IL-1 receptor antagonist (IL-1Ra). IL-1Ra was also able to partially reduce IL-36 production in response to the culprit drug, although to a significantly lesser extent than the TLR4-neutralizing antibody.
      Albumin is the most abundant serum protein and is known to have high ligand-binding capacities (
      • Fasano M.
      • Curry S.
      • Terreno E.
      • Galliano M.
      • Fanali G.
      • Narciso P.
      • et al.
      The extraordinary ligand binding properties of human serum albumin.
      ). It has also been reported to be a major target for drugs such as amoxicillin and to be required for the induction of immediate hypersensitivity reactions to drugs (
      • Ariza A.
      • Mayorga C.
      • Salas M.
      • Dona I.
      • Martin-Serrano A.
      • Perez-Inestrosa E.
      • et al.
      The influence of the carrier molecule on amoxicillin recognition by specific IgE in patients with immediate hypersensitivity reactions to betalactams.
      ). To assess whether albumin was required for IL-36γ secretion by AGEP patients’ cells exposed to culprit drugs, cells were first exposed to culprit and control drugs in the absence of proteins in the culture medium (serum-free). In the latter condition, culprit drug-induced IL-36γ secretion was dramatically decreased (Figure 4c). The medium supplementation with albumin restored the ability of PBMCs to secrete IL-36γ in response to drugs. Together, these results suggest that the drug-albumin complex is sensed by TLR4 as a danger signal in patients who experienced AGEP.

       Selective, IL-36–induced IL-8 production by patients’ PBMCs

      IL-8 has been previously reported to be produced by drug-specific T cells in AGEP patients (
      • Britschgi M.
      • Steiner U.C.
      • Schmid S.
      • Depta J.P.
      • Senti G.
      • Bircher A.
      • et al.
      T-cell involvement in drug-induced acute generalized exanthematous pustulosis.
      ), and in accordance with this, IL-8 expression was significantly up-regulated in AGEP skin compared with healthy skin (ratio = 2.25, P = 0.01) (see Supplementary Table S2) but also when compared to MPR (ratio = 2.24, P = 0.008) (see Supplementary Table S3), as shown by Affymetrix gene expression array (Figure 1a). We confirmed this by quantitative PCR (Figure 5a, left panel). Immunohistochemistry analyses further confirmed that IL-8 was overexpressed in AGEP compared with MPR and healthy skin and that cells of the infiltrate in pustular regions of AGEP skin were the main source of IL-8 (Figure 5a, right panels). As previously reported, IL-36 is an inducer of IL-8 (
      • Marrakchi S.
      • Guigue P.
      • Renshaw B.R.
      • Puel A.
      • Pei X.Y.
      • Fraitag S.
      • et al.
      Interleukin-36-receptor antagonist deficiency and generalized pustular psoriasis.
      ,
      • Towne J.E.
      • Renshaw B.R.
      • Douangpanya J.
      • Lipsky B.P.
      • Shen M.
      • Gabel C.A.
      • et al.
      Interleukin-36 (IL-36) ligands require processing for full agonist (IL-36alpha, IL-36beta, and IL-36gamma) or antagonist (IL-36Ra) activity.
      ,
      • Zhang J.
      • Yin Y.
      • Lin X.
      • Yan X.
      • Xia Y.
      • Zhang L.
      • et al.
      IL-36 induces cytokine IL-6 and chemokine CXCL8 expression in human lung tissue cells: implications for pulmonary inflammatory responses.
      ). Consistently, the exposure of PBMCs from AGEP patients to their respective culprit drugs resulted in IL-8 production that was inhibited by recombinant IL-36Ra in a dose-dependent manner (see Supplementary Figure S5 online). Also, using the keratinocyte/PBMC co-culture model described, we assessed whether IL-36 could promote IL8 expression, and if so, which cell type could produce IL-8 in response to IL-36. IL8 expression was induced by culprit drugs in PBMC cultured alone—confirming our ELISPOT data—whereas it was barely detectable in isolated keratinocyte cultures. Levels of IL8 expression in patients’ PBMCs were further increased (4.6-fold increase over control-drug exposure, P < 0.05) when co-cultured with autologous keratinocytes, and this was selectively in response to the culprit drug (Figure 5b). In contrast, keratinocytes co-cultured with PBMCs produced only marginal amounts of IL-8. In this autologous co-culture model, the up-regulation of IL8 expression in AGEP patients’ PBMCs was abrogated by IL-36Ra. Altogether, these results indicate that PBMCs are the main source of drug-induced IL-8 production and that IL-8 production is dependent on PBMC- and/or keratinocyte-derived IL-36.
      Figure thumbnail gr5
      Figure 5IL-8 is expressed in lesional skin biopsy samples from patients suffering from AGEP and is induced in patients’ PBMCs by IL-36. (a) Quantitative real-time reverse transcription–PCR analysis of IL8 in lesional skin biopsy samples of patients with AGEP (n = 16), patients with MPR (n = 16), and normal skin (n = 5). Immunohistochemical analysis of lesional skin biopsy samples showed that IL-8 is expressed at the site of pustules in AGEP in dermal immune cells but not by keratinocytes. Scale bars in the left panels = 300 μm; scale bars in the right panels = 50 μm. Representative picture of 18 AGEP cases is shown (upper panels). IL-8 was not detectable in normal skin from healthy donors (lower panels). (b) Quantitative PCR analysis of IL8 expression in PBMCs (upper panels) and KCs (lower panels) from patient 8 (amoxicillin 1 mg/ml), patient 9 (letrozole 100 nmol/L), and patient 20 (vancomycin 500 μg/ml), cultured either alone or in co-culture as indicated and exposed to the drug that caused the AGEP or a control drug in the presence or absence of IL-36Ra. Gene expression reported as 2–ΔCT, which represents the target gene expression relative to the reference gene (RPL27). Means ± standard deviation of the three tested patients is shown. The figure illustrates a representative experiment that was repeated three times for each patient. P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. AGEP, acute generalized exanthematous pustulosis; KC, keratinocyte; PBMC, peripheral blood mononuclear cell; NS, normal skin.

      Discussion

      T cells and neutrophils are considered to be major players in AGEP (
      • Britschgi M.
      • Steiner U.C.
      • Schmid S.
      • Depta J.P.
      • Senti G.
      • Bircher A.
      • et al.
      T-cell involvement in drug-induced acute generalized exanthematous pustulosis.
      ,
      • Feldmeyer L.
      • Heidemeyer K.
      • Yawalkar N.
      Acute generalized exanthematous pustulosis: pathogenesis, genetic background, clinical variants and therapy.
      ). It has been proposed that drugs causing AGEP elicit drug-specific T-cell responses and that these T cells secrete IL-8 (
      • Britschgi M.
      • Steiner U.C.
      • Schmid S.
      • Depta J.P.
      • Senti G.
      • Bircher A.
      • et al.
      T-cell involvement in drug-induced acute generalized exanthematous pustulosis.
      ). Recently, mutations in IL36RN have been reported in AGEP (
      • Navarini A.A.
      • Valeyrie-Allanore L.
      • Setta-Kaffetzi N.
      • Barker J.N.
      • Capon F.
      • Creamer D.
      • et al.
      Rare variations in IL36RN in severe adverse drug reactions manifesting as acute generalized exanthematous pustulosis.
      ), suggesting that a deregulation in IL-36 signaling may contribute to the physiopathology of AGEP. Our data support this hypothesis, showing that culprit drugs can directly induce IL-36γ secretion in AGEP patients’ cells in vitro. Analysis of IL-36 gene and protein expressions ex vivo also supports this and, furthermore, shows that macrophages and keratinocytes are the main producers of IL-36 in the pustular areas of AGEP skin. IL-36γ was found to be the predominant form of IL-36 expressed in AGEP skin biopsy samples. Causative drugs were able to induce IL-36γ and subsequent IL-8 production in PBMCs taken from patients months to years after AGEP resolution, whereas these cytokines could not be induced by drugs in healthy control donors who had never experienced an adverse drug reaction. This is, to our knowledge, the first experimental and translational evidence that the innate immune system is able to mount a fast proinflammatory response selectively to the culprit drug and exclusively in AGEP patients via the production of IL-36γ. Indeed, such a production by monocytes and keratinocytes from patients could be observed as early as 1 hour after culprit drug exposure in vitro. To further support a role for innate immunity in AGEP, we also provide evidence that TLR4 is involved in the sensing of culprit-drug/albumin complex as a danger signal, inducing IL-36γ exclusively in PBMCs from patients having experienced AGEP.
      Keratinocytes are a major source of IL-36 cytokines in cutaneous inflammation, particularly in psoriasis (
      • Carrier Y.
      • Ma H.L.
      • Ramon H.E.
      • Napierata L.
      • Small C.
      • O’Toole M.
      • et al.
      Inter-regulation of Th17 cytokines and the IL-36 cytokines in vitro and in vivo: implications in psoriasis pathogenesis.
      ). Our data suggest that keratinocytes are also an important source of IL-36 in the skin in AGEP. However, co-culture experiments showed that keratinocytes exposed to culprit drugs were able to produce high amounts of IL-36γ only when cultured in the presence of autologous monocytes and culprit drug. This observation suggests that, when exposed to the causative drug, patients’ monocytes release a soluble factor that further stimulates IL-36γ production by keratinocytes. This factor remains to be identified, but our results using IL-36Ra suggest that it is very likely not IL-36 itself at an early time point. Unlike what has been reported in psoriasis (
      • Carrier Y.
      • Ma H.L.
      • Ramon H.E.
      • Napierata L.
      • Small C.
      • O’Toole M.
      • et al.
      Inter-regulation of Th17 cytokines and the IL-36 cytokines in vitro and in vivo: implications in psoriasis pathogenesis.
      ), we did not observe any IL-36 induction by IL-36 itself in AGEP patients’ keratinocytes at the early time point of 6 hours after drug stimulation, as evidenced by the lack of effect of IL-36Ra.
      We also observed that the use of IL-1Ra led to a reduced IL-36γ production by AGEP patients’ PBMCs exposed to culprit drugs. A great overlap in IL-1 family gene up-regulation by IL-1β and IL-36γ has been recently reported (
      • Swindell W.R.
      • Beamer M.A.
      • Sarkar M.K.
      • Loftus S.
      • Fullmer J.
      • Xing X.
      • et al.
      RNA-seq analysis of IL-1B and IL-36 responses in epidermal keratinocytes identifies a shared MyD88-dependent gene signature.
      ). This overlap includes genes encoding IL-1β and IL-36 themselves, which can be induced by either cytokine, suggesting an IL-1β–IL-36γ amplifying loop in PBMCs stimulated by culprit drugs. The extent of IL-36γ inhibition by IL-1Ra was, however, significantly lower than that seen with a blocking antibody to TLR4. The important role of IL-36 in response to certain drugs in AGEP patients is further supported by the inhibition of IL-8 by IL-36Ra, the latter cytokine already suspected to cause neutrophil recruitment and survival in lesional skin during the course of AGEP (
      • Schaerli P.
      • Britschgi M.
      • Keller M.
      • Steiner U.C.
      • Steinmann L.S.
      • Moser B.
      • et al.
      Characterization of human T cells that regulate neutrophilic skin inflammation.
      ).
      As in other inflammatory diseases such as psoriasis (
      • Nickoloff B.J.
      Keratinocytes regain momentum as instigators of cutaneous inflammation.
      ), contact hypersensitivity (
      • Watanabe H.
      • Gaide O.
      • Petrilli V.
      • Martinon F.
      • Contassot E.
      • Roques S.
      • et al.
      Activation of the IL-1beta-processing inflammasome is involved in contact hypersensitivity.
      ), or sunburn (
      • Feldmeyer L.
      • Keller M.
      • Niklaus G.
      • Hohl D.
      • Werner S.
      • Beer H.D.
      The inflammasome mediates UVB-induced activation and secretion of interleukin-1beta by keratinocytes.
      ), keratinocytes may actively participate in the pathogenesis of AGEP by secreting IL-36, which could contribute to the herein observed and previously reported induction of IL-8 gene expression by macrophages. IL-8 may also originate from drug-specific T cells in AGEP (
      • Britschgi M.
      • Steiner U.C.
      • Schmid S.
      • Depta J.P.
      • Senti G.
      • Bircher A.
      • et al.
      T-cell involvement in drug-induced acute generalized exanthematous pustulosis.
      ). However, the mechanisms leading to drug-specific T cell priming and recruitment to the skin in AGEP have not been elucidated to date. Unlike murine T cells, human T cells do not express IL-36R (
      • Foster A.M.
      • Baliwag J.
      • Chen C.S.
      • Guzman A.M.
      • Stoll S.W.
      • Gudjonsson J.E.
      • et al.
      IL-36 promotes myeloid cell infiltration, activation, and inflammatory activity in skin.
      ) but, in AGEP, IL-36 may nevertheless contribute to T cell-mediated immune responses through its strong stimulatory effects on antigen-presenting cells (
      • Vigne S.
      • Palmer G.
      • Lamacchia C.
      • Martin P.
      • Talabot-Ayer D.
      • Rodriguez E.
      • et al.
      IL-36R ligands are potent regulators of dendritic and T cells.
      ). Our findings that IL-36γ can be induced by a drug very early in innate immune cells does not exclude a delayed drug-specific T-cell response, but whether the induction of IL-36γ in monocytes by culprit drugs and the activation of drug-specific T cells are dependent events in AGEP pathogenesis or not remains to be investigated.
      Th17 cells may also be involved in AGEP, as shown by their previously reported increased frequencies and elevated IL-22 blood levels in AGEP patients (
      • Kabashima R.
      • Sugita K.
      • Sawada Y.
      • Hino R.
      • Nakamura M.
      • Tokura Y.
      Increased circulating Th17 frequencies and serum IL-22 levels in patients with acute generalized exanthematous pustulosis.
      ). IL-17 is known to increase IL-8 secretion by keratinocytes (
      • Pennino D.
      • Eyerich K.
      • Scarponi C.
      • Carbone T.
      • Eyerich S.
      • Nasorri F.
      • et al.
      IL-17 amplifies human contact hypersensitivity by licensing hapten nonspecific Th1 cells to kill autologous keratinocytes.
      ) in allergic contact dermatitis (
      • Albanesi C.
      • Cavani A.
      • Girolomoni G.
      IL-17 is produced by nickel-specific T lymphocytes and regulates ICAM-1 expression and chemokine production in human keratinocytes: synergistic or antagonist effects with IFN-gamma and TNF-alpha.
      ) and in psoriasis (
      • Chiricozzi A.
      • Guttman-Yassky E.
      • Suarez-Farinas M.
      • Nograles K.E.
      • Tian S.
      • Cardinale I.
      • et al.
      Integrative responses to IL-17 and TNF-alpha in human keratinocytes account for key inflammatory pathogenic circuits in psoriasis.
      ). Our in vitro data showed, however, that in response to IL-36, the major source of IL-8 was patients’ monocytes, consistently with the previously reported elevated expression of IL-36R and potential strong responsiveness to IL-36 by dermal myeloid cells (
      • Dietrich D.
      • Martin P.
      • Flacher V.
      • Sun Y.
      • Jarrossay D.
      • Brembilla N.
      • et al.
      Interleukin-36 potently stimulates human M2 macrophages, Langerhans cells and keratinocytes to produce pro-inflammatory cytokines.
      ). Although keratinocytes also express high levels of IL-36R (
      • Dietrich D.
      • Martin P.
      • Flacher V.
      • Sun Y.
      • Jarrossay D.
      • Brembilla N.
      • et al.
      Interleukin-36 potently stimulates human M2 macrophages, Langerhans cells and keratinocytes to produce pro-inflammatory cytokines.
      ), IL-8 production by keratinocytes was marginal, as confirmed by quantitative PCR in co-culture experiments and immunohistochemistry where IL-8 was detected in only the peri-pustular immune infiltrate (Figure 5).
      Although mutations in IL36RN have been identified in AGEP patients, no other marker(s) of genetic predisposition have been formally identified for AGEP to date. Our data show that a deregulation of the IL-36–IL-8 axis can occur independently of IL36RN mutations in AGEP, suggesting that other factors influencing the susceptibility to an IL-36–mediated proinflammatory response to culprit drugs are at play in AGEP patients.
      Overall, we provide evidence that IL-36 is involved in AGEP pathogenesis and identify monocytes and keratinocytes as potential key producers of IL-36 in this ADR. We also identify TLR4 as an essential pattern recognition receptor involved in the sensing of the drug/albumin complex as a danger signal in AGEP patients. Further supporting our findings is the very rapid onset of the disease after the first intake of a drug, suggesting an early response mediated by the innate immune system. However, the mechanisms or factors conferring such a responsiveness to certain drugs on monocytes and keratinocytes in AGEP patients remain to be identified and require extensive investigations. Given that monocytes/PBMCs of all AGEP patients tested have an enhanced propensity to produce IL-36γ upon culprit drug exposure, the identification of predictive markers that would help identify patients prone to pathological IL-36γ production upon exposure to certain medications may be achievable.

      Materials and Methods

       Patients and biological material collection

      The characteristics of all studied AGEP patients from the dermatology department of the University Hospital of Zürich are presented in Table 1 and Supplementary Table S4 online. All human samples were collected after informed written patient consent with approval of the local ethics committees and according to the Declaration of Helsinki Principles. First, 4- to 6-mm punch skin biopsy samples were taken from the lesional skin of patients in the acute phase of the ADR. Healthy skin samples were obtained from the Department of Plastic Surgery, University Hospital of Zürich. For paraffinization, skin samples were fixed in formalin 4% overnight. For RNA isolation, samples were snap-frozen in liquid nitrogen and stored at –80 °C. Peripheral blood from patients was taken both in the acute phase of the disease and under healthy conditions (at least 6 months after the ADR). Healthy control samples were obtained from the Blood Donation Center (Schlieren, Switzerland). PBMCs were isolated using a density gradient (Ficoll-Paque; Pharmacia, Glattbrugg, Switzerland).
      Table 1AGEP Patients: Patch and Prick Tests, LTT, Culprit Drugs, and Presence of IL36RN Mutations
      Boldface indicates that cells from these patients were used in the experiments shown in Figures 3–5.
      Patient NumberPatch TestPrickLTTDrugIL36RN
      1NAnegativenegativeamoxicillinND
      2NANANAterbinafinND
      3negativenegativenegativemetforminno mutation
      4negativenegativeNAunknownno mutation
      5positivepositivenegativeunknownND
      6negativenegativenegativelonsoprazol/clarithromycinND
      7NANANAunknownND
      8NANApositiveamoxicillinno mutation
      9NANApositiveletrozoleND
      10positivepositivepositiveamoxicillinno mutation
      11NANANAamoxicillinND
      12NANANAflucoxacillinND
      13negativenegativeNAceftriaxonND
      14NANANAciprofloxacin/metamizol/paracetamolND
      15NANANAamoxicillinND
      16positivepositivenegativemetamizoleND
      17negativenegativeNAibuprofenno mutation
      18NANANAterbinafineND
      19positivepositivepositiveunknownND
      20NANANAvancomycinND
      21NANApositivefluconazoleND
      22NANANAunknownND
      23NANANAunknownND
      24NANANAamoxicillinND
      25NANANAamoxicillinND
      26NANApositiveacetazolamideND
      27NANANAcotrimoxazoleND
      28positivepositivenegativecotrimoxazole/codeine/paracetamolND
      29NANApositivelisinoprilND
      Abbreviations: LTT, lymphocyte transformation test; NA, not applicable; ND, not determined.
      1 Boldface indicates that cells from these patients were used in the experiments shown in Figure 3, Figure 4, Figure 5.

       Monocyte and T-cell purification

      Where indicated, CD14+ monocytes were sorted from PBMCs by positive selection by using magnetic beads (Miltenyi Biotech, Bergisch Gladbach, Germany). CD3+ cells were sorted from the CD14 fraction by negative selection using a Pan T Cell Isolation Kit (Miltenyi Biotech) according to the manufacturer’s instructions. Purity was assessed by flow cytometry (Facscanto A; Becton Dickinson, Franklin Lakes, NJ) using mouse anti-human CD45-PeCy7 (BioLegend, San Diego, CA), mouse anti-human CD3-APC (BioLegend), and mouse anti-human CD14-FITC (Becton Dickinson,).

       Primary keratinocyte culture

      Hair follicles were collected from patients at least 6 months after resolution of the ADR. Hair follicle extremities containing outer root sheath were cut, washed in complete keratinocyte medium (
      • Strittmatter G.E.
      • Garstkiewicz M.
      • Sand J.
      • Grossi S.
      • Beer H.D.
      Human primary keratinocytes as a tool for the analysis of caspase-1-dependent unconventional protein secretion.
      ), and incubated for 3 minutes in 1 mg/ml Dispase II (Roche, Rotkreuz, Switzerland), followed by three washing steps in medium. The hair was then plated on J2 feeder cells in complete medium and cultured until small colonies of keratinocytes became visible. Cells were split at 80% confluence and passaged three times before being used for experiments.

       Co-culture experiments

      PBMCs and keratinocytes used in these experiments were collected at least 6 months after resolution of the ADR. A total of 70,000 AGEP hair follicle keratinocytes were plated on a 12-mm Transwell plate with 0.4-μm pore polycarbonate membrane inserts (Corning, Corning, NY) overnight to allow them to adhere. The following day, keratinocyte medium was replaced by Gibco Opti-MEM medium (Thermo Fisher Scientific, Waltham, MA), and 1 × 106 PBMCs were added inside the inserts, followed by addition of the culprit drug (amoxicillin 1 mg/ml, letrozole 100 nmol/L, or vancomycin 500 μg/ml) or control drug (metamizole 100 μg/ml, carbamazepine 10 μg/ml, or amoxicillin, 1 mg/ml) or vehicle (RPMI medium containing 10% fetal bovine serum) in both compartments at the same concentration with or without IL-36 inhibition by a IL-36RA 1 μg/ml (R&D Systems, Minneapolis, MN). After 6 hours, cells were harvested and RNA was isolated for quantitative real-time reverse transcription–PCR analysis.

       TLR and IL-1 inhibition

      PBMCs were isolated from AGEP patients and cultured in ELISPOT plates. Cells were stimulated with the culprit drug in the presence or absence of the MyD88 inhibitory peptide Pepinh-MYD (25 mmol/L) (Invivogen, San Diego, CA), anti-TLR2 antibody (5 μg/ml) (R&D Systems, clone no. 383936,), anti-TLR4 antibody (5 μg/ml) (R&D Systems, polyclonal goat IgG), control IgG antibody (5 μg/ml), and IL-1RA (1 μg/ml) (Anakinra; BioVitrum, Stockholm, Sweden). Untreated cells and irrelevant control drug served as controls.

       Statistical analyses

      Differences between groups were assessed using one-way analysis of variance followed by Tukey posttest.

      Conflict of Interest

      The authors state no conflict of interest.

      Ackowledgments

      This work was supported by grants from the Zurich Center for Integrative Human Physiology , the Swiss National Science Foundation (grant 310030-176035 ) to LEF; the Promedica Stiftung to BM-S/LEF; the OPO-Stiftung, EMDO-Stiftung, and Forschungskredit der Universität Zürich to LF; and the Hartmann Müller Stiftung to EC.

      Supplementary Material

      References

        • Albanesi C.
        • Cavani A.
        • Girolomoni G.
        IL-17 is produced by nickel-specific T lymphocytes and regulates ICAM-1 expression and chemokine production in human keratinocytes: synergistic or antagonist effects with IFN-gamma and TNF-alpha.
        J Immunol. 1999; 162: 494-502
        • Anliker M.D.
        • Wuthrich B.
        Acute generalized exanthematous pustulosis due to sulfamethoxazol with positive lymphocyte transformation test (LTT).
        J Investig Allergol Clin Immunol. 2003; 13: 66-68
        • Ariza A.
        • Mayorga C.
        • Salas M.
        • Dona I.
        • Martin-Serrano A.
        • Perez-Inestrosa E.
        • et al.
        The influence of the carrier molecule on amoxicillin recognition by specific IgE in patients with immediate hypersensitivity reactions to betalactams.
        Sci Rep. 2016; 6: 35113
        • Aschan J.
        • Ringden O.
        • Tollemar J.
        • Ljungman P.
        • Lonnqvist B.
        • Pihlstedt P.
        • et al.
        Improved survival in marrow recipients above 30 years of age with better prevention of graft-versus-host disease.
        Transplant Proc. 1990; 22: 195-197
        • Britschgi M.
        • Pichler W.J.
        Acute generalized exanthematous pustulosis, a clue to neutrophil-mediated inflammatory processes orchestrated by T cells.
        Curr Opin Allergy Clin Immunol. 2002; 2: 325-331
        • Britschgi M.
        • Steiner U.C.
        • Schmid S.
        • Depta J.P.
        • Senti G.
        • Bircher A.
        • et al.
        T-cell involvement in drug-induced acute generalized exanthematous pustulosis.
        J Clin Invest. 2001; 107: 1433-1441
        • Carrier Y.
        • Ma H.L.
        • Ramon H.E.
        • Napierata L.
        • Small C.
        • O’Toole M.
        • et al.
        Inter-regulation of Th17 cytokines and the IL-36 cytokines in vitro and in vivo: implications in psoriasis pathogenesis.
        J Invest Dermatol. 2011; 131: 2428-2437
        • Chiricozzi A.
        • Guttman-Yassky E.
        • Suarez-Farinas M.
        • Nograles K.E.
        • Tian S.
        • Cardinale I.
        • et al.
        Integrative responses to IL-17 and TNF-alpha in human keratinocytes account for key inflammatory pathogenic circuits in psoriasis.
        J Invest Dermatol. 2011; 131: 677-687
        • Debets R.
        • Timans J.C.
        • Homey B.
        • Zurawski S.
        • Sana T.R.
        • Lo S.
        • et al.
        Two novel IL-1 family members, IL-1 delta and IL-1 epsilon, function as an antagonist and agonist of NF-kappa B activation through the orphan IL-1 receptor-related protein 2.
        J Immunol. 2001; 167: 1440-1446
        • Dietrich D.
        • Martin P.
        • Flacher V.
        • Sun Y.
        • Jarrossay D.
        • Brembilla N.
        • et al.
        Interleukin-36 potently stimulates human M2 macrophages, Langerhans cells and keratinocytes to produce pro-inflammatory cytokines.
        Cytokine. 2016; 84: 88-98
        • Fasano M.
        • Curry S.
        • Terreno E.
        • Galliano M.
        • Fanali G.
        • Narciso P.
        • et al.
        The extraordinary ligand binding properties of human serum albumin.
        IUBMB Life. 2005; 57: 787-796
        • Feldmeyer L.
        • Heidemeyer K.
        • Yawalkar N.
        Acute generalized exanthematous pustulosis: pathogenesis, genetic background, clinical variants and therapy.
        Int J Mol Sci. 2016; 17: 1214
        • Feldmeyer L.
        • Keller M.
        • Niklaus G.
        • Hohl D.
        • Werner S.
        • Beer H.D.
        The inflammasome mediates UVB-induced activation and secretion of interleukin-1beta by keratinocytes.
        Curr Biol. 2007; 17: 1140-1145
        • Foster A.M.
        • Baliwag J.
        • Chen C.S.
        • Guzman A.M.
        • Stoll S.W.
        • Gudjonsson J.E.
        • et al.
        IL-36 promotes myeloid cell infiltration, activation, and inflammatory activity in skin.
        J Immunol. 2014; 192: 6053-6061
        • Gresnigt M.S.
        • van de Veerdonk F.L.
        Biology of IL-36 cytokines and their role in disease.
        Semin Immunol. 2013; 25: 458-465
        • He Q.
        • Chen H.X.
        • Li W.
        • Wu Y.
        • Chen S.J.
        • Yue Q.
        • et al.
        IL-36 cytokine expression and its relationship with p38 MAPK and NF-kappaB pathways in psoriasis vulgaris skin lesions.
        J Huazhong Univ Sci Technolog Med Sci. 2013; 33: 594-599
        • Johnston A.
        • Xing X.
        • Guzman A.M.
        • Riblett M.
        • Loyd C.M.
        • Ward N.L.
        • et al.
        IL-1F5, -F6, -F8, and -F9: a novel IL-1 family signaling system that is active in psoriasis and promotes keratinocyte antimicrobial peptide expression.
        J Immunol. 2011; 186: 2613-2622
        • Kabashima R.
        • Sugita K.
        • Sawada Y.
        • Hino R.
        • Nakamura M.
        • Tokura Y.
        Increased circulating Th17 frequencies and serum IL-22 levels in patients with acute generalized exanthematous pustulosis.
        J Eur Acad Dermatol Venereol. 2011; 25: 485-488
        • Kanazawa N.
        • Nakamura T.
        • Mikita N.
        • Furukawa F.
        Novel IL36RN mutation in a Japanese case of early onset generalized pustular psoriasis.
        J Dermatol. 2013; 40: 749-751
        • Lian L.H.
        • Milora K.A.
        • Manupipatpong K.K.
        • Jensen L.E.
        The double-stranded RNA analogue polyinosinic-polycytidylic acid induces keratinocyte pyroptosis and release of IL-36gamma.
        J Invest Dermatol. 2012; 132: 1346-1353
        • Marrakchi S.
        • Guigue P.
        • Renshaw B.R.
        • Puel A.
        • Pei X.Y.
        • Fraitag S.
        • et al.
        Interleukin-36-receptor antagonist deficiency and generalized pustular psoriasis.
        N Engl J Med. 2011; 365: 620-628
        • Navarini A.A.
        • Valeyrie-Allanore L.
        • Setta-Kaffetzi N.
        • Barker J.N.
        • Capon F.
        • Creamer D.
        • et al.
        Rare variations in IL36RN in severe adverse drug reactions manifesting as acute generalized exanthematous pustulosis.
        J Invest Dermatol. 2013; 133: 1904-1907
        • Nickoloff B.J.
        Keratinocytes regain momentum as instigators of cutaneous inflammation.
        Trends Mol Med. 2006; 12: 102-106
        • Nishio D.
        • Izu K.
        • Kabashima K.
        • Tokura Y.
        T cell populations propagating in the peripheral blood of patients with drug eruptions.
        J Dermatol Sci. 2007; 48: 25-33
        • Onoufriadis A.
        • Simpson M.A.
        • Pink A.E.
        • Di Meglio P.
        • Smith C.H.
        • Pullabhatla V.
        • et al.
        Mutations in IL36RN/IL1F5 are associated with the severe episodic inflammatory skin disease known as generalized pustular psoriasis.
        Am J Hum Genet. 2011; 89: 432-437
        • Pennino D.
        • Eyerich K.
        • Scarponi C.
        • Carbone T.
        • Eyerich S.
        • Nasorri F.
        • et al.
        IL-17 amplifies human contact hypersensitivity by licensing hapten nonspecific Th1 cells to kill autologous keratinocytes.
        J Immunol. 2010; 184: 4880-4888
        • Pichler W.J.
        Delayed drug hypersensitivity reactions.
        Ann Intern Med. 2003; 139: 683-693
        • Schaerli P.
        • Britschgi M.
        • Keller M.
        • Steiner U.C.
        • Steinmann L.S.
        • Moser B.
        • et al.
        Characterization of human T cells that regulate neutrophilic skin inflammation.
        J Immunol. 2004; 173: 2151-2158
        • Sidoroff A.
        • Halevy S.
        • Bavinck J.N.
        • Vaillant L.
        • Roujeau J.C.
        Acute generalized exanthematous pustulosis (AGEP)—a clinical reaction pattern.
        J Cutan Pathol. 2001; 28: 113-119
        • Song H.S.
        • Kim S.J.
        • Park T.I.
        • Jang Y.H.
        • Lee E.S.
        Immunohistochemical comparison of IL-36 and the IL-23/Th17 axis of generalized pustular psoriasis and acute generalized exanthematous pustulosis.
        Ann Dermatol. 2016; 28: 451-456
        • Strittmatter G.E.
        • Garstkiewicz M.
        • Sand J.
        • Grossi S.
        • Beer H.D.
        Human primary keratinocytes as a tool for the analysis of caspase-1-dependent unconventional protein secretion.
        Methods Mol Biol. 2016; 1459: 135-147
        • Sugiura K.
        • Takeichi T.
        • Kono M.
        • Ogawa Y.
        • Shimoyama Y.
        • Muro Y.
        • et al.
        A novel IL36RN/IL1F5 homozygous nonsense mutation, p.Arg10X, in a Japanese patient with adult-onset generalized pustular psoriasis.
        Br J Dermatol. 2012; 167: 699-701
        • Swindell W.R.
        • Beamer M.A.
        • Sarkar M.K.
        • Loftus S.
        • Fullmer J.
        • Xing X.
        • et al.
        RNA-seq analysis of IL-1B and IL-36 responses in epidermal keratinocytes identifies a shared MyD88-dependent gene signature.
        Front Immunol. 2018; 9: 80
        • Towne J.E.
        • Renshaw B.R.
        • Douangpanya J.
        • Lipsky B.P.
        • Shen M.
        • Gabel C.A.
        • et al.
        Interleukin-36 (IL-36) ligands require processing for full agonist (IL-36alpha, IL-36beta, and IL-36gamma) or antagonist (IL-36Ra) activity.
        J Biol Chem. 2011; 286: 42594-42602
        • Vigne S.
        • Palmer G.
        • Lamacchia C.
        • Martin P.
        • Talabot-Ayer D.
        • Rodriguez E.
        • et al.
        IL-36R ligands are potent regulators of dendritic and T cells.
        Blood. 2011; 118: 5813-5823
        • Watanabe H.
        • Gaide O.
        • Petrilli V.
        • Martinon F.
        • Contassot E.
        • Roques S.
        • et al.
        Activation of the IL-1beta-processing inflammasome is involved in contact hypersensitivity.
        J Invest Dermatol. 2007; 127: 1956-1963
        • Zhang J.
        • Yin Y.
        • Lin X.
        • Yan X.
        • Xia Y.
        • Zhang L.
        • et al.
        IL-36 induces cytokine IL-6 and chemokine CXCL8 expression in human lung tissue cells: implications for pulmonary inflammatory responses.
        Cytokine. 2017; 99: 114-123