Dysregulated skin barrier function in Tmem79 mutant mice promotes IL‐17A‐dependent spontaneous skin and lung inflammation

Atopic dermatitis (AD) is associated with a dysregulation of the skin barrier and may predispose to the development of secondary allergic conditions, such as asthma. Tmem79ma/ma mice harbor a mutation in the gene encoding Transmembrane Protein 79 (or Mattrin), which has previously been associated with AD. As a result of the Tmem79 gene mutation, these mice have a defective skin barrier and develop spontaneous skin inflammation. In this study, Tmem79ma/ma mice were assessed for the underlying immunological response in the development of spontaneous skin and lung inflammation.


| INTRODUC TI ON
Atopic dermatitis (AD) is a common inflammatory skin disorder affecting 15%-30% of children in the western world. 1 Individuals with a history of AD are predisposed to develop secondary allergic conditions in distal tissue sites, such as the lung, in what is known as the atopic march. 2,3 The etiology of AD is multifactorial, with roles for genetic predisposition and environmental factors in the development of aberrant skin inflammation. 4 Disease progression is complex and significantly influenced by the interplay between the functional integrity of the skin barrier and the immune system. 5 Recent studies have illustrated the complexity and heterogeneity of inflammatory gene expression in the skin of patients with AD. Therefore, the spectrum of underlying immune responses associated with AD now includes CD4 + Thelper (h)2, IL-22 (Th22), IL-17 (Th17) as well as IL-23 mediated cellular responses in different patient cohorts, and in mouse models that develop AD-like inflammation. [6][7][8] There is increased expression of IL17A in both acute and chronic lesions of patients with AD. 9,10 While studies have indicated a potential role for IL17A in disease pathogenesis, the function and source of IL-17A in AD patients are not well defined. In mouse models of skin inflammation, IL-17A has been shown to be involved in the generation of AD-like skin inflammation, 11 with a critical role for microbiota-induced IL-17A secretion. [12][13][14] The flaky tail double-mutant mouse, with mutations in Filaggrin (Flg) and Tmem79, has been extensively used for studies of skin barrier defects and development of AD-like skin inflammation. 15 The skin inflammation in flaky tail mice on C57BL/6 background is due to mutations in Tmem79, and not Flg. 16,17 Tmem79 ma mutant mice do not express the transmembrane transport protein mattrin leading to a defective skin barrier and the development of skin inflammation.
The identification of the Tmem79 gene in mice led to detection of a missense SNP in the human homologue, TMEM79/MATT, that significantly increases the risk for AD. 16 In this study, we now demonstrate that the Tmem79 ma/ma mutant mice develop spontaneous IL-17A-dependent skin inflammation with secondary progression to lung pathology. Indeed, cutaneous Conclusion: Mice lacking Tmem79/Mattrin expression have a defective skin barrier.
In adulthood, these mice develop dermatitis with secondary progression to lung inflammation. The development of skin and lung inflammation is IL-17A-dependent and mediated by TCRγδ T cells.

K E Y W O R D S
cutaneous inflammation, interleukin-17A, lung inflammation, skin barrier, Tmem79

G R A P H I C A L A B S T R A C T
A single mutation in Tmem79 causes a dysregulation of skin barrier integrity, which -with increasing age -lead to the development of spontaneous dermatitis and subsequent lung inflammation. In Tmem79 ma/ma mice, IL-17A-producing γδ T cells and Th17 cells infiltrate the skin prior to the occurrence of pulmonary disease. Progression from skin to lung inflammation is dependent on the presence of γδ T cells and IL-17A. Abbreviations: Tmem79, transmembrane protein 79 expansion of IL-17A-expressing Th17 and TCRγδ cells facilitated progression to pulmonary inflammation, which provides important insights into the pathogenesis of the atopic march.

| Mice
The Tmem79 ma mutation was separated from double-mutant flaky tail (Matt ma/ma Flg ft/ft ) mice, and Tmem79 ma mice were backcrossed to congenic C57BL/6J background as previously outlined. 16 Homozygous female Tmem79 ma/ma mice were used in all experiments. TCRδ-deficient (Strain No: 002120, Jackson Laboratories),

| Clinical scoring
The severity of inflammation and AD-like pathology was scored using the macroscopic diagnostic criteria as described for the skin inflammation in the Nc/Nga mouse model of AD. 19 A scoring system (score 0, none; score 1, mild; score 2, moderate; score 3, severe) was applied to the symptoms of edema, erosion, scaling, and erythema.
Total scores for each mouse, to a maximum of 12, were calculated from the sum of individual parameters.

| Scratching behavior
Individual mice were videoed and the spontaneous scratching of mice was quantified as described. 20 Scratches per minutes were determined based on an individual animal lifting a hind paw from the cage floor, scratching with the paw, and then returning the paw to the cage floor.

| Trans-epidermal water loss (TEWL)
The measurement of TEWL was performed on ear skin using a Courage and Khazaka Tewameter TM210 (Enviroderm), as described previously. 15

| Histology
Skin or lungs were fixed in 10% formaldehyde saline. Skin sections were stained with hematoxylin and eosin (H&E). Lung sections were stained with H&E and Masson's trichrome for collagen quantification, as described. 21

| RNA isolation and real-time PCR
RNA was isolated from skin using the RNeasy kit and reverse transcribed with the Quantitect reverse transcription kit (Qiagen).

Real-time quantitative PCR was performed on an AB StepOnePLus
Real-time PCR system (Life Technologies), using TaqMan gene expression assays specific for murine Ifnγ (Mm01168134_m1), Il4
Mice were tracheostomized and ventilated using a whole-body plethysmography, with a pneumotachograph linked to a transducer (EMMS).
Changes in lung resistance (R L ) and compliance (C dyn ) in response to increasing doses of nebulized and inhaled methacholine (10, 30, 60, and 100 mg/mL; Sigma-Aldrich), were recorded as described. 15,21

| Bronchoalveolar lavage
Bronchoalveolar lavage (BAL) fluid was collected following AHR analyses. 22 Differential cell counts were performed on BAL cells on cytospins.

| Collagen assay
Briefly, lung tissue was collected and homogenized, with subsequent overnight acid-pepsin collagen extraction at 4°C. Total soluble collagen content was determined with a Sircol Collagen Assay kit (Biocolor). Pulmonary collagen was expressed as μg collagen per mg lung protein. 21

| Protein extraction from lung and cytokine measurement
Lungs were homogenized in a buffer containing PBS, 2% fetal bovine serum and 0.5% cetyltrimethylammonium bromide using an IKA T10 Basic Ultra-Turrax homogenizer. The homogenates were centrifuged at 13 000 g for 15 minutes. Cytokines were determined using commercial ELISA kits (R&D systems) and expressed as μg cytokine per mg lung protein.  analyses were performed using FlowJo software (Tree Star Inc).

| IgE and cytokine ELISA
Total serum IgE was measured in serum from mice according to the manufacturer's instructions (BD Pharmingen). Cytokines in cell culture supernatants were analyzed by ELISA as per the manufacturer's instructions (R&D Systems).

| Statistical analyses
Data are expressed as mean ± SEM and were analyzed by two-way analysis of variance (ANOVA) test or unpaired Student's t tests (Prism 7; GraphPad Software). Significance for was shown in figures as: NS = not significant, P < .05 (*), P < .01 (**), P < .001 (***) and P < .0001 (****).  Figure 2F). While production of IFN-γ and IL-4 was also increased by LN cells from Tmem79 ma/ ma mice, the increase was not statistically compared to cells from infected WT animals ( Figure 2F). These results demonstrate that spontaneous skin inflammation in Tmem79 ma/ma mice renders them refractory to S aureus infection and was characterized by an IL-17Adominant cutaneous immune response.

| Tmem79 ma/ma mice develop spontaneous pulmonary inflammation
As AD can initiate the progression to allergic pulmonary inflammation, we analyzed the lungs of Tmem79 ma/ma mice. At 16 weeks of age, mutant mice had developed skin inflammation ( Figure 1A) but there was no overt lung inflammation (Figure. S1C). However, subsequent analysis at 32 weeks of age revealed that Tmem79 ma/ma mice had alterations in AHR, with significant changes in dynamic lung compliance (C dyn ) in mutant mice compared to WT controls but not in lung resistance (R L ) ( Figure 3A).
In support of the altered lung function and presence of pulmonary inflammation, there were significantly (P < .05) more cells in BAL of Tmem79 ma/ma mice relative to WT mice with an increase in total cell numbers observed across all cell populations ( Figure 3B). Notably, there was a significant increase in the frequency of both neutrophils F I G U R E 1 Tmem79-deficiency leads to spontaneous and progressive skin inflammation. A, Macroscopic clinical scoring of Tmem79 ma/ma vs wild-type (WT) mice. Data are from 10 to 20 mice per strain, scored longitudinally. ****P < .0001, two-way ANOVA. B, Representative photograph depicting the gross phenotype of Tmem79 ma/ma mouse (Skin Score 11) in comparison to WT control at 32 wk of age. C, Representative H&E-stained biopsies of lesional skin from 32-wk-old Tmem79 ma/ma and WT mice. Scale bars, 400 μm. D, Scratches per minute of individual 32-wk-old Tmem79 ma/ma (red circles) and WT (black circles) mice. E, Fold change in Ifng, Il4, Il13, Il17a, Il1b, Il25, Il33, and Tslp mRNA expression (n = 6-12 per group) in skin relative to Gapdh. Data are representative of three experiments. Relative quantification and statistics in Figure S1B (P < .001) and eosinophils (P < .05) ( Figure 3B). The compromised AHR in Tmem79 ma/ma mice was reflected in marked lung pathology with mixed peribronchial cellular infiltrates ( Figure 3C). The lungs of Tmem79 ma/ma mice did not develop bronchial goblet cell hyperplasia, marked peribronchial eosinophilia, or airway occlusion (data not shown). In lung homogenates from Tmem79 ma/ma mice, IL-17A was significantly (P < .0001) increased, as was IL-1β (P < .0001; Figure 3D).
The alarmins IL-25 (P < .01) and TSLP (P < .05) were also significantly elevated in the lungs of Tmem79 ma/ma mice ( Figure 3D Collectively, Tmem79 ma/ma mice spontaneously develop lung pathology, that was secondary to skin inflammation, with decreased lung compliance, increased parenchymal collagen deposition, eosinophil and neutrophil infiltration, and pulmonary inflammation associated with IL-17A.

| Tmem79 ma/ma mice have an expansion of IL-17A producing T cells
Given the pronounced increase in IL-17A levels in both lesional and nonlesional skin of Tmem79 ma/ma mice, we investigated the cellular source of IL-17A by generating dual Il17eGFP reporter Tmem79 ma/ma mice. Assessing the levels of IL-17A-producing T cells in the skin-draining LN of Tmem79 ma/ma mice, prior to the development of eczematous skin inflammation, revealed that IL-17A was expressed by both γδ T cells and CD4 + Th cells ( Figure 4A,B).
The lung inflammation that develops in Tmem79 ma/ma mice was associated with increased IL-17A levels in lung homogenates ( Figure 3D), yet Il17eGFP Tmem79 ma/ma mice experienced no F I G U R E 3 Spontaneous lung inflammation in Tmem79 ma/ma mice. A, Airway hyperresponsiveness in WT and Tmem79 ma/ma mice. Dynamic compliance (C dyn ) and resistance (R L ) in response to increasing doses of methacholine (Mch) were recorded. Data are representative of three experiments, n = 6-8 per group. Statistical significance was determined with two-way ANOVA test. B, Total and differential cell counts from BAL of WT (black bars) and  Figure S3C; data not shown). Taken together these data indicate that skin inflammation in Tmem79 ma/ma mice is associated with a type 17 cellular response.

| Development of spontaneous skin and lung inflammation in Tmem79 ma/ma is mediated by adaptive immunity
We sought to address if inflammation in Tmem79 ma/ma mice was mediated by adaptive immunity by generating T and B cell deficient Rag1 -/-Tmem79 ma/ma mice and, since γδ T cells are increased in skin of mutant mice, dual Tcrδ -/-Tmem79 ma/ma mice ( Figure 5A). Both Rag1 -/-Tmem79 ma/ ma and Tcrδ -/-Tmem79 ma/ma mice were protected from overt skin inflammation ( Figure 5A). In contrast, dual Tcrα -/--Tmem79 ma/ma mice developed skin inflammation (data not shown). The defect in the skin barrier of Tmem79 ma/ma mice, evident by significantly elevated (P < .005) TEWL in nonlesional skin of 8-week-old mutant mice compared to WT mice, was also present in the dual Rag1 -/--or Tcrδ -/--Tmem79 ma/ma mice ( Figure 5B). Importantly, both strains of mice did not progress to develop lung pathology ( Figure 5C,D), indicating a critical role for T and B cells, and among these specifically γδ T cells, in the pathogenesis of dermatitis and progression to lung pathology in Tmem79 ma/ma mice. Our data suggest that development of spontaneous skin and lung inflammation in Tmem79 ma/ma mice are dependent on TCRγδ T cells.

| Spontaneous skin and lung inflammation in Tmem79 ma/ma mice is dependent on IL-17A
In order to define the contribution of IL-17 to inflammation, Il17a -/-Tmem79 ma/ma mice were generated. Il17a -/-Tmem79 ma/ma mice had a profound amelioration in the magnitude of skin inflammation ( Figure 6A,B), with significantly (P < .0001) reduced clinical skin inflammation relative to Tmem79 ma/ma controls ( Figure 6A).
Consequently, scratching was significantly reduced in IL-17Adeficient Tmem79 ma/ma mice ( Figure 6B). Overt inflammatory lesions did not develop and dermal inflammatory cell infiltration and epidermal hyperplasia were decreased in Il17a -/-Tmem79 ma/ma mice F I G U R E 4 Frequency of IL-17producing cells in the skin and lung of Tmem79 ma/ma mice. (A, B), Frequency of IL-17 + γδ T cells (A) and frequency of Th17 (B) cells in skin-draining lymph nodes of WT and Tmem79 ma/ma crossed to Il17eGFP reporter mice. (C, D), Frequency of IL-17 + γδ T cells (C) and IL-17 + CD4 T cells (D) in lesional skin of WT(Il17eGFP) and Tmem79 ma/ma (Il17eGFP) mice. E, F: Frequency of IL-17 + γδ T cells (E) and Th17 cells (F) in whole lungs of WT(Il17eGFP) and Tmem79 ma/ma (Il17eGFP)mice. Data is representative of two experiments, n = 4-6 per group. **P < .01; ***P < .001; Student's t test ( Figure 6C). At 32 weeks of age, IL-17A-deficient mutant mice had no lung inflammation, while Tmem79 ma/ma mice had developed pronounced lung pathology ( Figure 6D,E). These results demonstrate that pathogenesis of overt skin inflammation as well as the progression to lung pathology in mice with an impaired skin barrier due to Tmem79-deficiency is dependent on IL-17A.

| D ISCUSS I ON
Increasing evidence from genome-wide association studies (GWAS), immunochip and transcriptome analyses firmly implicate genetic predisposition associated with skin barrier defects, as a causal factor in AD pathogenesis. 4 Previous studies on a mouse model with a mutation in Tmem79, identified in the Flaky tail dual Flg-Tmem79 mutant strain, with an impaired skin barrier, 17 led to the demonstration of a missense SNP in TMEM79 that was associated with human AD. 16 In contrast, mice with a mutation in the filaggrin gene (Flg ft/ ft ) develop no overt skin inflammation on the C57BL/6 background (Figure. S1A), 16 while they develop AD-like inflammation on the BALB/c background in an IL-1β-and mast cell-dependent manner. 28 We now show that Tmem79 ma/ma mice develop spontaneous skin inflammation that progresses to lung inflammation, through an IL-17A dependent mechanism that is mediated by TCRγδ T cells.
The AD-like skin inflammation in Tmem79 ma/ma mice is characterized by a ~40-fold increase in Il17a expression, and a marked infiltration of both Th17 and IL-17 + γδ T cells. TCRγδ T cells are major producers of IL-17A and rapidly respond to dysregulation of the skin barrier. 29 Importantly, we observe a specific ~7-fold increase in Il17a in nonlesional skin, which is associated with an increase in the frequency of skin IL-17 + γδ T cells in mutant mice prior to overt skin inflammation. These high basal levels of IL-17A in the skin of Tmem79 ma/ma mice mediate refractoriness to S auerus challenge. Indeed, specific alterations in responses toward S aureus have been reported in a number of other mouse strains with defective skin barrier integrity. 14,28,30 The upregulation in Il1β expression in the nonlesional skin of both Tmem79 ma/ ma mice is consistent with previous work, in which epidermal IL-1β was shown to be important in AD patients and murine AD models. [31][32][33] As IL-1β can activate IL-17 + γδ T cells in the skin, 34 the increased IL-1β in skin of mutant mice may increase the frequency of cutaneous IL-17 + γδ T cells to amplify skin inflammation suggesting that IL-1β acts upstream of IL-17A-mediated pathology.
Indeed, in other mouse models of disrupted skin barrier, there is increased infiltration of IL-17 + γδ T cells in the skin. 11,14,30,32,35 In two distinct mouse models of defective epidermal barrier leading to cutaneous inflammation, γδ T cells were increased in the skin but were redundant for the genesis of skin inflammation. 14,35 In Tmem79 ma/ma mice the development of spontaneous IL-17A- We generated dual IL-17A-deficient and Tmem79 ma/ma mice that demonstrated that the development of skin and lung inflammation in mutant mice was IL-17A-dependent. Previously, roles for IL-17A in skin inflammation have been reported in the Flaky tail mice that the Tmem79 mutant mice were derived from. 11,36 While one study suggested that IL-17A-mediated inflammation was a consequence of the filaggrin mutation present in flaky tail mice, 37 we could now show that in fact the Tmem79 ma/ma mutation was responsible for increased IL-17A production and neutrophil infiltration. In mice with defective IL-17RA signaling the skin barrier is compromised, with IL-17A treatment of primary murine keratinocytes altering filaggrin monomer formation. 14 IL-17A treatment of human keratinocytes has been shown to downregulate monomeric filaggrin protein, and the expression of genes implicated in keratinocyte adhesion. 38 Therefore, mutations in Tmem79 may further decrease skin barrier integrity through modulating filaggrin expression by increased IL-17A production. These studies highlight that IL-17A has a central role in skin barrier maintenance and the development of cutaneous inflammation in Tmem79 ma/ma mice. Future studies are needed to address the interplay of mutations affecting skin barrier integrity and the ensuing AD to personalize therapy in affected AD patients.
The pulmonary inflammation in Tmem79 ma/ma mice was characterized by increased collagen deposition, eosinophil and neutrophil infiltration in the airways, and increased IL-17A and IL-1β, leading to compromised lung function. While Tmem79 ma/ma mice have altered dynamic lung compliance, there is no change in airway resistance, a surrogate marker for central airway allergic inflammation.
These alterations in lung compliance in Tmem79 ma/ma mice suggest that the aberrant lung function is predominately due to peripheral changes, such as lung parenchyma elasticity, with lesser effects on central airway function. 39 In mouse models, it has been previously shown that Th17 cells mediate steroid-resistant airway inflammation and are capable of inducing AHR and neutrophil recruitment in response to Ovalbumin-challenge. 40 Indeed, an IL-17A(high) immunophenotype has been described in patients with steroid-resistant asthma. 41 We herein present the first example of a mouse model of IL-17A-dependent spontaneous lung inflammation that occurs subsequent to spontaneous skin inflammation. Currently, the main source of the increased IL-17A in the lung remains unclear.
While we observed a minor increase in the frequency of pulmonary TCRγδ and Th17 cells, their increased cell numbers most likely increase the overall abundance of IL-17A. Furthermore, several other cells have been reported to produce IL-17A, including neutrophils and eosinophils. 42,43 It will be interesting to further explore the sensitivity of Tmem79 ma/ma mice to steroids, or other interventions, to treat pulmonary inflammation. The exact factors that drive the atopy march are still not entirely clear. Thus, both Tmem79 ma/ma and Flg ft/ft mice, which develop pulmonary inflammation following overt AD-like inflammation, are valuable tools to investigate the factors promoting allergic escalation. 32 We generated Tmem79 fl/fl mice to conditionally delete mattrin in a cell-or site-specific manner. Total deletion of Tmem79, using ACTB-Cre mice, validated the Cre-lox system with mice developing comparable phenotypes as the mutant strain. Conditional targeting uisng K14-cre mice, demonstrated that keratinocyte-derived Tmem79 was responsible for the spontaneous skin and lung inflamamtion observed in Tmem79 ma/ma mice. Indeed, others have generated a similar mouse model and could show that deletion of Tmem79 in keratinocytes increased expansion of IL-17A + γδ-T cells and elicited mast cell-mediated histaminergic itch. 44 While they did not observe overt lung inflammation, we only observed pulmonary inflammation starting after 16 weeks of age ( Figure.

CO N FLI C T O F I NTE R E S T
Dr Saunders has nothing to disclose. Dr Floudas has nothing to disclose. Dr Moran has nothing to disclose. Dr Byrne has nothing to disclose. Dr Rooney has nothing to disclose. Dr Fahy has nothing to disclose. Dr Geoghegan has nothing to disclose. Dr Iwakura has nothing to disclose. Dr Fallon has nothing to disclose. Dr Schwartz has nothing to disclose.