Assay for Transposase-Accessible Chromatin Using Sequencing Analysis Reveals a Widespread Increase in Chromatin Accessibility in Psoriasis

Psoriasis is a complex, chronic inflammatory skin disease characterized by keratinocyte hyperproliferation and disordered immune response; however, its exact etiology remains unknown. To better understand the regulatory network underlying psoriasis, we explored landscape of chromatin accessibility using an assay for transposase-accessible sequencing analysis 15 psoriatic, 9 nonpsoriatic, 19 normal tissue samples, data were integrated with genomic, epigenomic, transcriptomic datasets. We identified 4,915 genomic regions that displayed differential in psoriatic samples compared both nonpsoriatic nearly all which exhibited increased tissue. These differentially accessible tended to be more hypomethylated correlated expression their linked genes, comprised several psoriasis susceptibility loci. Analyses region sequences showed they most highly enriched FRA1 and/or activator protein-1 transcription factor DNA-binding motifs. also found AIM2, encodes important inflammasome component triggers inflammation, direct target protein-1. Our study provided clear insights resources improved understanding pathogenesis psoriasis. disease-associated might serve as therapeutic targets treatment future. common complex global prevalence varies from 0.09% 11.4% (Michalek et al., 2017Michalek I.M. Loring B. John S.M. A systematic review worldwide epidemiology psoriasis.J Eur Acad Dermatol Venereol. 2017; 31: 205-212Crossref PubMed Scopus (419) Google Scholar; Parisi 2013Parisi R. Symmons D.P. Griffiths C.E. Ashcroft D.M. Identification Management Associated ComorbidiTy (IMPACT) project team, al. Global psoriasis: incidence prevalence.J Invest Dermatol. 2013; 133: 377-385Abstract Full Text PDF (1513) Scholar). Psoriatic lesions are generally parakeratosis, acanthosis, lymphocytic infiltration. histopathological features caused synergy between aberrant activation cells impaired proliferation keratinocytes (KCs) (Griffiths Barker, 2007Griffiths Barker J.N. Pathogenesis clinical psoriasis.Lancet. 2007; 370: 263-271Abstract (1319) Schön Boehncke, 2005Schön M.P. Boehncke W.H. Psoriasis.N Engl J Med. 2005; 352: 1899-1912Crossref (998) GWASs have than 100 psoriasis-associated variants (Harden 2015Harden J.L. Krueger J.G. Bowcock A.M. The immunogenetics Psoriasis: comprehensive review.J Autoimmun. 2015; 64: 66-73Crossref (323) Mahil 2015Mahil S.K. Capon F. Genetics psoriasis.Dermatol Clin. 33: 1-11Abstract (57) Nair 2000Nair R.P. Stuart P. Henseler T. Jenisch S. Chia N.V. Westphal E. al.Localization psoriasis-susceptibility locus PSORS1 60-kb interval telomeric HLA-C.Am Hum Genet. 2000; 66: 1833-1844Abstract (226) Tsoi 2017Tsoi L.C. P.E. Tian C. Gudjonsson J.E. Das Zawistowski M. al.Large scale meta-analysis characterizes genetic architecture associated variants.Nat Commun. 8: 15382Crossref (142) Zhou 2016aZhou Cao H. Zuo X. Zhang Liu al.Deep MHC Chinese population contributes studies disease.Nat 2016; 48: 740-746Crossref (134) Many these located near genes encoding subunits epidermal differentiation major histocompatibility well involved regulation immune-related signaling pathways, indicating system epithelial barrier critical roles (Mahil Although GWAS many loci only explain small fraction cases heritability. Moreover, biological function reside noncoding intergenic explored. maturation KCs modulated layers epigenetic modifications (Cavazza 2016Cavazza A. Miccio Romano O. Petiti L. Malagoli Tagliazucchi G. Peano al.Dynamic transcriptional human differentiation.Stem Cell Reports. 6: 618-632Abstract (38) Scholar), such DNA methylation (DNAm), dysregulation can contribute through changes gene (Chandra 2015Chandra Ray Senapati Chatterjee Genetic basis pathogenesis.Mol Immunol. 313-323Crossref (102) Several hundreds CpG DNAm (Gervin 2012Gervin K. Vigeland M.D. Mattingsdal Hammerø Nygård Olsen A.O. al.DNA monozygotic twins discordant identification epigenetically dysregulated genes.PLoS 2012; 8e1002454Crossref (121) Roberson 2012Roberson E.D. Y. Ryan Joyce Duan al.A subset methylated sites differentiate skin.J 132: 583-592Abstract (108) 2013Zhang Zhao Liang Yin Huang D. Su al.Whole-genome patients vulgaris.J 41: 17-24Crossref (104) 2016bZhou Wang W. Shen Li Zheng al.Epigenome-wide association nine 136: 779-787Abstract (51) Besides DNAm, at elements promoters enhancers seem (Klemm 2019Klemm S.L. Shipony Z. Greenleaf W.J. Chromatin epigenome.Nat Rev 2019; 20: 207-220Crossref (427) Given transcriptome-level lesions, it conceivable state may altered significant proportion regions. However, best our knowledge, no has comprehensively evaluated skin. address this, profiled genome-wide pattern performing (ATAC-seq) (Buenrostro 2013Buenrostro J.D. Giresi P.G. Zaba Chang H.Y. Transposition native fast sensitive epigenomic profiling open chromatin, proteins nucleosome position.Nat Methods. 10: 1213-1218Crossref (2895) Scholar) (PP) (PN) tissues (NN) healthy individuals. first ATAC signals could largely separate PP NN skin, peaks was specifically Thereafter, integrating previous methylome transcriptome individuals same population, observed decreased transcription. Overall, analyses provide into molecular mechanisms onset development conducted ATAC-seq PP, PN, samples. Each dataset generated 32.80 54.50 million reads (median, 50.15 million) (Supplementary Table S1). After alignment mapping raw data, called model-based immunoprecipitation-sequencing 2 992,693 consensus peaks, each being supported one or this represent largest datasets date measuring Assessment 305,828 recurrent (supported least two samples) group indicated strong enrichment around start sites, mostly those distal intronic (Figure 1a Supplementary Figure S1a). then derived DNase I hypersensitive multiple adult KC cell line Encyclopedia Elements portal (Sloan 2016Sloan C.A. Chan E.T. Davidson J.M. Malladi V.S. Strattan J.S. Hitz B.C. al.ENCODE ENCODE portal.Nucleic Acids Res. 44: D726-D732Crossref (254) More 40% accessibility-related overlapped line, showing highest degree consistency among interrogated 1b). furthermore large, consistent coordinates housekeeping CTCF, RPS24, SF1, UBE4B across replicates S1b). overall differences three groups principal normalized counts per (c.p.m.) mapped read values. clearly separated other 1c). difference mainly captured 1 S2a). On further exploration top components, interestingly PN seemed distinct on level 4 S2b). comparable variance relative S2c–d) possibly variability technical issues. results substantially different tissues. aimed identify displaying By applying custom normalization methods, 22,839 50,845 respectively (false discovery rate <0.01). Of these, 94.3% (21,543 22,839) 84.2% (42,790 50,845), respectively, S3a–b). met above false cutoff threshold <0.01 comparisons fold (Pearson’s correlation, r = 0.8; P < 2.2e−16) 2a). obtain unique set high-confidence applied stringent (|log2 change| >1 comparisons) (hereafter denoted [DARs]) 2b). Notably, but DAR examined markers dermal types (e.g., fibroblasts macrophages) reportedly epidermis Langerhans T helper 17 cells), did not observe overrepresentation DARs S2), suggesting significantly dynamic types. Analysis 10 presence previously been (Table 1). For example, 1q21 (PSORS4) containing S100A family harbored 2c S4). Fitting multilinear regression model median intensity against features, former smokers (three patients) tend (P 0.0107) S3). because limited sample size, needed verify this. Interestingly, representing early replication domains KCs, whereas related moderately late 2d).Table 1The Most Significant Psoriasis-Associated Differential Accessibility PeaksPeakDistance TSSSymbol2Symbol: name, according HUGO Gene Nomenclature Committee.Mean_PP3Mean_PP: Average peak scores groups.Mean_PNMean_NNPP Versus PNPP NNChrStart1Start end: Genomic coordinate_GRCh37 end peaks.EndlogFCFDRlogFCFDRchr1153,329,982153,330,4670S100A98.872.432.331.863.71E-141.951.51E-30chr2250,980,02150,980,522−5,940KLHDC7B7.292.521.821.541.86E-102.021.51E-30chr1812,890,00712,890,531−5,673PTPN26.341.671.471.924.83E-132.141.73E-30chr2237,678,552237,679,351200,172ACKR315.454.013.571.941.64E-132.142.60E-29chr1321,005,13721,005,7552,230MIR449910.262.452.382.055.42E-142.127.04E-29chr1590,057,96990,058,7088,557LINC0092811.562.752.762.073.47E-132.076.45E-28chr5149,011,842149,012,93950,707ARHGEF3723.437.466.841.653.52E-121.795.25E-27chr193,154,9003,155,54718,709GNA1512.944.473.861.531.63E-111.776.92E-27chr1117,079,958117,080,4846,742CD588.491.541.772.441.68E-142.271.75E-26chr647,135,78947,136,389−125,707ADGRF110.982.142.542.148.02E-152.128.13E-26Abbreviations: Chr, chromosome number; FC, change; FDR, rate; HUGO, Human Genome Organization; NN, tissues; psoriasis; lesions; TSS, site.FDR ordered decreasing statistical significance versus NN.1 Start peaks.2 Symbol: Committee.3 Mean_PP: groups. Open table new tab Abbreviations: site. FDR NN. them available anatomical locations S3c). recent histone (Klein 2017Klein R.H. Lin Hopkin A.S. Gordon al.GRHL3 binding rearrange transition functional states.PLoS 13e1006745Crossref (20) When comparing dataset, estimated 27.9% (1,372 4,915) likely active enhancers, marked H3K27ac H3K4me3 S3d). ontology pathway overlap processes development, migration, TNF pathway. There some unreported pathways axon guidance Rap1 S5a–d). Taken together, DAR-related key performed Illumina 450K array (San Diego, CA) (Zhou RNA-sequencing (Dou 2017Dou J. Xie Ye Yang Wen al.Integrative reveal psoriasis.Br 177: 1349-1357Crossref (24) lesional nonlesional partially included (see S6 details), correspond biopsy. This allowed us explore connections variation expression, 1,676 CpGs (here designated CpGs) when either 90 resided inside S4), corresponded 7.28% (90 1,236) assayed DARs. 0.61% (813 133,965) peak-residing there 10-fold regions, (hypergeometric test, 2.2e−16). levels 95.56% (86 90) groups, including promoter-localized 3a). reports (Corces 2018Corces M.R. Granja Shams Louie B.H. Seoane J.A. al.The primary cancers.Science. 2018; 362eaav1898Crossref (370) Kim 2018Kim H.S. Tan Ma Merkurjev Destici Q. al.Pluripotency factors functionally premark cell-type-restricted ES cells.Nature. 556: 510-514Crossref (25) Northcott 2017Northcott P.A. Buchhalter I. Morrissy Hovestadt V. Weischenfeldt Ehrenberger whole-genome medulloblastoma subtypes.Nature. 547: 311-317Crossref (491) Thurman 2012Thurman R.E. Rynes Humbert Vierstra Maurano M.T. Haugen genome.Nature. 489: 75-82Crossref (1735) suggest correlation 1,886 2,616 consistently upregulated downregulated, 0.01 comparisons). DAR-linked 632 274 3b). represented revealed death 2.58e-8) effector 9.4e-6), others. Taking AIM2 promoter contained cg07195224, site shown (Dombrowski 2011Dombrowski Peric Koglin Kammerbauer Göss Anz al.Cytosolic lesions.Sci Transl 2011; 3: 82ra38Crossref (297) As expected, promoter-associated strongly negatively cg07195224 (Spearman’s ρ −0.49; 7.9e-3) 3c), positively mRNA (ρ 0.89; 1.4e-3) 3c); −0.45; 3.7e-4) 3c). analyzed 531 SNPs catalog (0.38%) list DARs, namely rs2145623 PSMA6 NFKBIA, rs2779255 NOS2. S7a–b) respective S7c–d). close linkage (r2 > 0.6) additionally 36 28 S5a). imputation East Asian haplotypes 1000 Genomes Project reference panel, calculated (Zhang 2009Zhang X.J. Sun L.D. F.Y. Zhu Q.X. al.Psoriasis identifies within LCE cluster 1q21.Nat 2009; 205-210Crossref (355) 3,602 non-HLA 10e-4 potential risk variants. 18 (0.49%) S5b). findings explained next determined motifs Homer toolkit motif analysis. (AP-1) complexes 4a), consist structurally similar heterodimers belonging c-Fos (c-JUN, JUN-B, JUN-D), c-Jun (c-FOS, FOS-B, FRA1, FRA2), ATF, JDP families. FOSL1, ranked list. Examination FOSL1 eight-fold (t-test; 0.01), public AP-1 S8a–b). validated immunohistochemical immunofluorescence analyses. expressed CD45, number much lower 4b). Based =1.03e-9) FRA1-binding motif. inflammation processing proinflammatory cytokine, pro-IL-1β, form 363 base pair 57 upstream AIM2. suspected directly targeted FRA1. luciferase reporter activity 293T cotransfected pGL3-AIM2 vector (pGL3 wild type mutant fused mRNA) 4c). exogenous signal transfected promoter, promoter. suggested binds myeloid leukemia breast cancer 4d S9). In summary, upregulation inflammasome. extent use various lines (Cusanovich 2018Cusanovich D.A. Hill A.J. Aghamirzaie Daza R.M. Pliner H.A. Berletch J.B. single-cell atlas vivo mammalian accessibility.Cell. 174: 1309-1324.e18Abstract (273) Wu 2016Wu Chen Xiang preimplantation embryos.Nature. 534: 652-657Crossref (327) Here, 300,000 chromatin-linked Approximately measured DNase-seq. resource genome. openness study, evidence First, nonstochastic near-unidirectional increase reported increases hyperproliferation-related diseases, particularly cancers (Denny 2016Denny Chuang C.H. Brady J.J. Lim Grüner B.M. al.Nfib promotes metastasis widespread 166: 328-342Abstract (199) reflect rates require ener