ADAM17-Deficient Mice Model the Transcriptional Signature of Human Atopic Dermatitis

Atopic dermatitis (AD) is an inflammatory skin disease characterized by barrier defect, immune dysregulation, and microbial dysbiosis. AD frequently precedes atopic disease at other barrier tissues, such as food allergies and asthma (Dharmage et al., 2014Dharmage S.C. Lowe A.J. Matheson M.C. Burgess J.A. Allen K.J. Abramson M.J. Atopic dermatitis and the atopic march revisited.Allergy. 2014; 69: 17-27Crossref PubMed Scopus (252) Google Scholar). Despite the prevalence and impact of the disease, there are still few steroid-sparing treatment options and only one targeted therapy approved by the US Food and Drug Administration (Simpson et al., 2016Simpson E.L. Bieber T. Guttman-Yassky E. Beck L.A. Blauvelt A. Cork M.J. et al.Two phase 3 trials of dupilumab versus placebo in atopic dermatitis.N Engl J Med. 2016; 375: 2335-2348Crossref PubMed Scopus (1106) Google Scholar), reflecting the limits of current knowledge and strategies for intervention. To advance understanding of pathogenesis and treatment, a robust preclinical animal model is needed. Mice lacking Adam17 in Sox9-expressing tissue including epidermis (Adam17fl/flSox9-Cre, henceforth A17) have been previously identified as a model for severe human AD based on the eczematous phenotype and spontaneous staphylococcal dysbiosis of A17 skin (Kobayashi et al., 2015Kobayashi T. Glatz M. Horiuchi K. Kawasaki H. Akiyama H. Kaplan D.H. et al.Dysbiosis and Staphylococcus aureus colonization drives inflammation in atopic dermatitis.Immunity. 2015; 42: 756-766Abstract Full Text Full Text PDF PubMed Scopus (322) Google Scholar). Although useful, these endpoints capture only the most grossly aberrant features of AD and cannot validate the model for as-yet-unknown mechanisms of pathogenesis. To address these limitations, we analyzed unbiased changes across the entire skin transcriptome in A17 mouse skin. All experiments were performed under an animal study proposal approved by the National Cancer Institute Animal Care and Use Committee. We compared whole skin and epidermis RNA sequencing of three A17 mice with those of three age-matched wild-type controls. Back skin was enzymatically dissociated to obtain cell suspensions as described previously (Kobayashi et al., 2015Kobayashi T. Glatz M. Horiuchi K. Kawasaki H. Akiyama H. Kaplan D.H. et al.Dysbiosis and Staphylococcus aureus colonization drives inflammation in atopic dermatitis.Immunity. 2015; 42: 756-766Abstract Full Text Full Text PDF PubMed Scopus (322) Google Scholar). Total mRNA was used to generate cDNA libraries that were sequenced, filtered for low-quality reads, aligned with STAR (2.4.1d), and quantified to mm10 (Ensembl Transcripts, release 82) in PartekFlow (Partek Inc., Chesterfield, MO). Read counts were normalized to total count and analyzed for differential expression using gene-specific analysis (GSA) per PartekFlow recommended settings. Sequencing data have been deposited in the National Center for Biotechnology Information Gene Expression Omnibus and are accessible through Gene Expression Omnibus Series accession number GSE107687. Of 13,573 genes analyzed by gene-specific analysis, 2,514 differentially expressed genes (DEGs) were detected in A17 skin using a standard definition including both statistically significant change (P-value < 0.01, false discovery rate < 0.05) and effect size (fold change > 2 or < –2) (Figure 1a). These DEGs included genes important in lipid metabolism and keratinization, consistent with the understanding of barrier dysfunction in atopic inflammation, as well as significant immunologic pathways in adaptive and innate immunity previously identified in the human AD transcriptome (Figure 1b) (Ewald et al., 2015Ewald D.A. Malajian D. Krueger J.G. Workman C.T. Wang T. Tian S. et al.Meta-analysis derived atopic dermatitis (MADAD) transcriptome defines a robust AD signature highlighting the involvement of atherosclerosis and lipid metabolism pathways.BMC Med Genomics. 2015; 8: 60Crossref PubMed Scopus (103) Google Scholar). Multiple up-regulated pathways implied dermal effector cell involvement, including those for T-cell activation, dendritic cell maturation, and cytokine signaling. Of particular interest was the up-regulation of JAK signaling pathways, which are targetable via inhibitors in humans. Nearly identical lists of DEGs were obtained from whole skin and epidermis samples (see Supplementary Table S1 online). Transcriptomic data obtained from whole skin samples were further analyzed to reflect skin biopsies performed in humans. To assess the similarity of A17 mouse skin and human AD skin more directly, we queried human gene symbols corresponding to mouse gene symbols using the bioinformatics package biomaRt and investigated the overlap between these human orthologs of mouse DEGs (hDEGs) in A17 whole skin (n = 2,250) and a previously published, meta-analysis–derived DEG list (“MADAD”) of 595 genes from human AD skin (Ewald et al., 2015Ewald D.A. Malajian D. Krueger J.G. Workman C.T. Wang T. Tian S. et al.Meta-analysis derived atopic dermatitis (MADAD) transcriptome defines a robust AD signature highlighting the involvement of atherosclerosis and lipid metabolism pathways.BMC Med Genomics. 2015; 8: 60Crossref PubMed Scopus (103) Google Scholar). We found that the A17 hDEGs contained 34% of human DEGs in MADAD (Figure 1c), on the upper end of overlap in other murine AD models for which published DEGs were available (Ewald et al., 2017Ewald D.A. Noda S. Oliva M. Litman T. Nakajima S. Li X. et al.Major differences between human atopic dermatitis and murine models, as determined by using global transcriptomic profiling.J Allergy Clin Immunol. 2017; 139: 562-571Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). The overall trend was maintained even when directionality was taken into account (Figure 1c). Among overlapping genes were those important for mitosis, disulfide bond formation, chemotaxis, and innate immunity (Table 1, and see Supplementary Table S2 online). Greatest enrichment was seen in genes related to cell division, suggesting that A17 uniquely captures a proliferative phenotype within MADAD not reported for other models (Ewald et al., 2017Ewald D.A. Noda S. Oliva M. Litman T. Nakajima S. Li X. et al.Major differences between human atopic dermatitis and murine models, as determined by using global transcriptomic profiling.J Allergy Clin Immunol. 2017; 139: 562-571Abstract Full Text Full Text PDF PubMed Scopus (79) Google Scholar). Both A17 and IL-23, the model with the highest MADAD coverage, captured innate immune activation seen in MADAD, with more direct comparison of the two models showing enriched IL-1 activity (Table 1, and see Supplementary Table S2). Most human DEGs captured in MADAD remain unrepresented in A17 or any other mouse model, speaking to the challenges of recapitulating the full range of change seen in polygenic diseases like AD with models that manipulate single variables in a controlled fashion.Table 1Functional annotation clusters for MADAD and human orthologs of A17 DEGsAnnotationsCluster Enrichment ScoreDEGs in both MADAD and human ortholog of A17 (n = 200) Mitosis, cell cycle, cell division7.84 Signal peptide, disulfide bond, glycoprotein3.12 Inflammatory response, chemotaxis, cytokine2.93 Innate immunity2.67DEGs in A17 only (n = 2,050) Chemical carcinogenesis, metabolism of xenobiotics by cytochrome P4505.03 NAD (nicotinamide adenine dinucleotide), nucleotide phosphate-binding region: NAD4.38 Cell cycle, mitosis4.00 Retinol metabolism, retinol dehydrogenase activity3.51DEGs in MADAD only (N = 383) Signal peptide, disulfide bond, glycoprotein9.64 Innate immunity, antiviral defense7.57 Adaptive immune response7.16 Chemotaxis, chemokine signaling pathway3.47DEGs in MADAD, A17, and IL-23 (n = 96) Signal peptide, disulfide bond, glycoprotein4.88 Innate immune response4.35 Leukocyte migration, neutrophil chemotaxis4.10DEGs in A17 and IL-23 (n = 596) Signal peptide, disulfide bond, glycoprotein7.91 Oxidoreductase activity4.52 Cytokine activity, IL-14.34Abbreviations: A17, Adam17fl/flSox9-Cre mice; DEG, differentially expressed gene; MADAD, meta-analysis derived in atopic dermatitis. Open table in a new tab Abbreviations: A17, Adam17fl/flSox9-Cre mice; DEG, differentially expressed gene; MADAD, meta-analysis derived in atopic dermatitis. For all mouse models, the extent of differential expression, or number of hDEGs in each model, predicted the extent of overlap with MADAD (r2 = 0.821, P < 0.001) (Figure 1d). This raised the question of whether an increasingly extreme but nonspecific model of skin disease might, by sheer number of DEGs alone, capture more human AD hDEGs. To address this concern, we analyzed the A17 hDEGs that did not overlap with MADAD. We found that these genes played a role in cytochrome P450 metabolism, nicotinamide adenine dinucleotide, cell division, and retinol metabolism (Table 1), all of which have been associated with AD previously (Drucker et al., 2017Drucker A.M. Li W.-Q. Park M.K. Li T. Qureshi A.A. Cho E. Niacin intake and incident adult-onset atopic dermatitis in women.J Allergy Clin Immunol. 2017; 139: 2020-2022Abstract Full Text Full Text PDF PubMed Scopus (8) Google Scholar, Ohno et al., 2015Ohno Y. Nakamichi S. Ohkuni A. Kamiyama N. Naoe A. Tsujimura H. et al.Essential role of the cytochrome P450 CYP4F22 in the production of acylceramide, the key lipid for skin permeability barrier formation.Proc Natl Acad Sci USA. 2015; 112: 7707-7712Crossref PubMed Scopus (108) Google Scholar, Rollman and Vahlquist, 1985Rollman O.L.A. Vahlquist A. Vitamin A in skin and serum—studies of acne vulgaris, atopic dermatitis, ichthyosis vulgaris and lichen planus.Br J Dermatol. 1985; 113: 405-413Crossref PubMed Scopus (31) Google Scholar, Ungar et al., 2017Ungar B. Garcet S. Gonzalez J. Dhingra N. da Rosa J. Shemer A. et al.An integrated model of atopic dermatitis biomarkers highlights the systemic nature of the disease.J Invest Dermatol. 2017; 137: 603-613Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar). Although these associations cannot prove the absence of off-target, non-AD–like inflammation, they also do not provide evidence of an additional, wholly unrelated skin pathology driving differential gene expression in A17 skin. Critically, the absence of any A17 hDEG in MADAD may also reflect low sensitivity of MADAD itself to detect subtle transcriptional change in human AD, because failure to detect low-expression genes on whole-tissue microarrays would carry through the meta-analysis approach. Indeed, the human orthologs of several nonoverlapping A17 hDEGs, including filaggrin (FLG), inflammasome regulator NLRP10 (NLRP10), putative antioxidant CCDC80 (CCDC80), and IL-2 receptor alpha (IL2RA), were missing from MADAD but have been implicated in human AD genome-wide association studies (Paternoster et al., 2015Paternoster L. Standl M. Waage J. Baurecht H. Hotze M. Strachan D.P. et al.Multi-ancestry genome-wide association study of 21,000 cases and 95,000 controls identifies new risk loci for atopic dermatitis.Nat Genet. 2015; 47: 1449-1456Crossref PubMed Scopus (362) Google Scholar) and other work (Elias et al., 2017Elias M.S. Long H.A. Newman C.F. Wilson P.A. West A. McGill P.J. et al.Proteomic analysis of filaggrin deficiency identifies molecular signatures characteristic of atopic eczema.J Allergy Clin Immunol. 2017; 140: 1299-1309Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar, Furue et al., 1994Furue M. Sugiyama H. Tsukamoto K. Ohtake N. Tamaki K. Serum soluble IL-2 receptor (sIL-2R) and eosinophil cationic protein (ECP) levels in atopic dermatitis.J Dermatol Sci. 1994; 7: 89-95Abstract Full Text PDF PubMed Scopus (33) Google Scholar, Miyai et al., 2016Miyai M. Yamamoto-Tanaka M. Inoue K. Tsuboi R. Hibino T. Atopic dermatitis susceptible gene NLRP10 suppresses inflammatory reaction and NLRP10 SNP mutation down-regulates NLRP10 expression.J Dermatol Sci. 2016; 84: e69Abstract Full Text Full Text PDF Google Scholar). Our results suggest that A17 shows among the greatest degree of transcriptional similarity with human AD and are consistent with eczematous phenotypes seen previously (Kobayashi et al., 2015Kobayashi T. Glatz M. Horiuchi K. Kawasaki H. Akiyama H. Kaplan D.H. et al.Dysbiosis and Staphylococcus aureus colonization drives inflammation in atopic dermatitis.Immunity. 2015; 42: 756-766Abstract Full Text Full Text PDF PubMed Scopus (322) Google Scholar). Similarities to other inflammatory processes as seen in psoriasis remain to be explored and may elucidate common features of these diseases. The authors state no conflict of interest. This work was supported by the Intramural Research Program, Center for Cancer Research, National Cancer Institute, National Institutes of Health. TK is a JSPS-NIH fellow. TW was funded by the National Institutes of Health Medical Research Scholars Program. Download .xlsx (5.45 MB) Help with xlsx files Supplementary Table S1 Download .xlsx (.07 MB) Help with xlsx files Supplementary Table S2