Early-onset childhood atopic dermatitis is related to NLRP2 repression
2017; Elsevier BV; Volume: 141; Issue: 4 Linguagem: Inglês
10.1016/j.jaci.2017.11.018
ISSN1097-6825
AutoresLoreen Thürmann, Konrad Grützmann, Matthias Klös, Matthias Bieg, Marcus Winter, Tobias Polte, Tobias Bauer, Matthias Schick, Melanie Bewerunge‐Hudler, Stefan Röder, Mario Bauer, Dirk K. Wissenbach, Ulrich Sack, Dieter Weichenhan, Oliver Mücke, Christoph Plass, Michael Borte, Martin von Bergen�, Irina Lehmann, Roland Eils, Saskia Trump,
Tópico(s)Food Allergy and Anaphylaxis Research
ResumoAtopic dermatitis (AD) is a common skin disease in children. While genetic predisposition and environmental factors are accepted mediators of AD development, the role of the innate immune system is not yet completely elucidated. Likely mediators of AD are nucleotide-binding oligomerization and pyrin domain containing-receptors (NLRPs), which are involved in the innate immune response.1Fontalba A. Gutierrez O. Fernandez-Luna J.L. NLRP2, an inhibitor of the NF- B pathway, is transcriptionally activated by NF- B and exhibits a nonfunctional allelic variant.J Immunol. 2007; 179: 8519-8524Crossref PubMed Scopus (52) Google Scholar So far, few studies investigated the role of NLRPs in adult AD, primarily focusing on single nucleotide polymorphisms (SNPs), while only reporting transcriptional changes for NLRP2 and NLRP3.2Niebuhr M. Baumert K. Heratizadeh A. Satzger I. Werfel T. Impaired NLRP3 inflammasome expression and function in atopic dermatitis due to Th2 milieu.Allergy. 2014; 69: 1058-1067Crossref PubMed Scopus (45) Google Scholar, 3Yoshikawa Y. Sasahara Y. Takeuchi K. Tsujimoto Y. Hashida-Okado T. Kitano Y. et al.Transcriptional analysis of hair follicle-derived keratinocytes from donors with atopic dermatitis reveals enhanced induction of IL32 gene by IFN-gamma.Int J Mol Sci. 2013; 14: 3215-3227Crossref PubMed Scopus (7) Google Scholar To elucidate the relevance of NLRPs in childhood AD, this study focused on these 2 NLRP family members. We started by evaluating the transcription of NLRP2 and NLRP3 in whole blood of children of the Lifestyle and Environmental Factors and Their Influence on Newborns Allergy Risk (LINA) cohort (see Fig E1 in this article's Online Repository at www.jacionline.org), showing AD symptoms or were diagnosed with AD by a physician until the age of 6 years ("AD ever") (Fig 1; see also Table E1 and Fig E1 in this article's Online Repository at www.jacionline.org). While NLRP3 expression was unaltered, NLRP2 was decreased in 1-year-old children with AD compared with healthy controls (Fig 1, A; see also Tables E2 and E3 in this article's Online Repository at www.jacionline.org). Since there are different pheno- and endotypes of AD, supposedly based on different molecular and cellular characteristics, we further separated the occurrence of AD into early- and late-onset (see Fig E2 in this article's Online Repository at www.jacionline.org). Only for children with early-onset AD was NLRP2 differentially expressed in comparison to healthy controls (Fig 1, B, Tables E2 and E3). NLRP3 was neither differentially transcribed in early- nor late-onset AD compared with controls (Tables E2 and E3). As DNA-methylation is as a potent epigenetic regulator of gene expression, we next investigated NLRP2 promoter methylation. Using whole genome bisulfite sequencing data from cord blood samples of a subset of children from our cohort, we observed NLRP2 promoter hypermethylation in children with early-onset AD compared with controls (Fig 2, A; see also Table E4 in this article's Online Repository at www.jacionline.org). Validation of DNA-methylation (see Fig E3 and Table E5 in this article's Online Repository at www.jacionline.org) in the entire cohort confirmed an AD-related promoter hypermethylation at birth and at age 1 year, which was associated with an increased risk for the children to develop early-onset AD (Fig 2, B). At both time points this hypermethylation was inversely correlated with NLRP2 transcription (see Fig E4 in this article's Online Repository at www.jacionline.org), suggesting a functional translation of the methylation change to the transcriptional level. Subsequently, we investigated whether the alterations in NLRP2 expression and methylation arose from a specific blood cell subpopulation. Although monocytes, B cells, CD8+, and CD4+ T cells of children with AD showed a significant difference in NLRP2 expression and methylation compared with controls, the pattern was similar in all subpopulations, suggesting that AD-associated differences in NLRP2 are not related to changes in these specific cell types (see Fig E5 in this article's Online Repository at www.jacionline.org). Another level, by which variation in gene expression can be explained are expression quantitative trait loci (eQTLs). Because the Genotype-Tissue-Expression portal describes rs514148 as the most significant eQTL for NLRP2 in blood, we analyzed this SNP in our cohort. Both NLRP2 transcription and promoter methylation were different between rs514148 genotypes (see Fig E6 in this article's Online Repository at www.jacionline.org), characterizing rs514148 also as a methylation QTL for NLRP2. Reevaluating our AD-related regression analyses by including the rs514148 variant as a confounder retained the significant risk increase for AD development in relation to NLRP2 repression and promoter hypermethylation (age 1 year: transcription, adjusted odds ratio [adj OR] = 0.75, 95% CI: 0.62-0.91; methylation, adj OR = 3.75, 95% CI: 1.26-11.18). Only recently released genotyping platforms cover rs514148 or other SNPs in linkage disequilibrium (r2 > 0.5). As per our knowledge, none of these SNPs were included in prior AD–genome-wide association studies, so we hope future studies will show whether rs514148 will be confirmed as an AD risk locus. Since the genetic background alone could not explain the link between NLRP2 and AD, further factors, such as environmental exposure might have an additional impact on NLRP2 regulation in AD. We assessed whether prenatal tobacco smoke exposure as a known risk factors for AD4Winter M. Thürmann L. Gu Z. Schüürmann G. Herberth G. Hinz D. et al.The benzene metabolite 1,4-benzoquinone reduces regulatory T-cell function: a potential mechanism for tobacco smoke–associated atopic dermatitis.J Allergy Clin Immunol. 2017; 140: 603-605Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar, 5Saulyte J. Regueira C. Montes-Martinez A. Khudyakov P. Takkouche B. Active or passive exposure to tobacco smoking and allergic rhinitis, allergic dermatitis, and food allergy in adults and children: a systematic review and meta-analysis.PLoS Med. 2014; 11: e1001611Crossref PubMed Scopus (123) Google Scholar might play a role. Indeed, mediation analysis revealed that prenatal benzene exposure—a valid proxy for tobacco smoke exposure (see Fig E7 in this article's Online Repository at www.jacionline.org)—influenced AD development via changes in NLRP2 promoter methylation and transcription (see Fig E8 in this article's Online Repository at www.jacionline.org). Previously, in IFN-γ stimulated hair follicle-derived keratinocytes from AD patients—mimicking an inflammatory response—an increased NLRP2 expression compared with that of healthy adults was observed.3Yoshikawa Y. Sasahara Y. Takeuchi K. Tsujimoto Y. Hashida-Okado T. Kitano Y. et al.Transcriptional analysis of hair follicle-derived keratinocytes from donors with atopic dermatitis reveals enhanced induction of IL32 gene by IFN-gamma.Int J Mol Sci. 2013; 14: 3215-3227Crossref PubMed Scopus (7) Google Scholar This is in line with the immunosuppressive function of NLRP2 to blunt excessive nuclear factor-κB activation, as NLRP2 is a known nuclear factor-κB inhibitor.1Fontalba A. Gutierrez O. Fernandez-Luna J.L. NLRP2, an inhibitor of the NF- B pathway, is transcriptionally activated by NF- B and exhibits a nonfunctional allelic variant.J Immunol. 2007; 179: 8519-8524Crossref PubMed Scopus (52) Google Scholar Although, these results seem to contradict our findings, our observations were made in different cell types and at different time points (adulthood vs early childhood). More importantly, the small number of studied hair follicle-derived keratinocytes increases the likelihood that the genotype distribution determined the increased NLRP2 expression in AD patients. The results presented here rather suggest that early dysregulation of NLRP2 in unstimulated circulating immune cells constitutes a risk factor for the child to develop AD by diminishing the immunosuppressive potential of NLRP2. As NLRP2 inhibits nuclear factor-κB signaling, concomitant cytokine alterations might have an impact on AD. In fact, children with early-onset AD showed elevated IL-10 blood protein concentrations (see Table E6 in this article's Online Repository at www.jacionline.org) that were associated with both, NLRP2 hypermethylation (adj mean ratio [MR] = 1.15, 95% CI: 1.03-1.27) and transcriptional repression (adj MR = 0.87, 95% CI: 0.78-0.9). AD-related increase of IL-10 has been observed in skin lesions, but also in various blood cell types.6Ohmen J.D. Hanifin J.M. Nickoloff B.J. Rea T.H. Wyzykowski R. Kim J. et al.Overexpression of IL-10 in atopic dermatitis: contrasting cytokine patterns with delayed-type hypersensitivity reactions.J Immunol. 1995; 154: 1956-1963PubMed Google Scholar Ricci et al7Ricci G. Patrizi A. Federica B. Calamelli E. Dell'Omo V. Bendani B. et al.Cytokines levels in children affected by atopic and nonatopic eczema.Open Dermatol J. 2008; 2: 18-21Crossref Google Scholar determined higher IL-10 serum concentrations in children with eczema, in line with our observation that elevated IL-10 increased the children's risk to develop early-onset AD (adj OR = 2.81, 95% CI: 1.20-6.56). High concentrations of IL-10 have been implicated in the downregulation of the antimicrobial peptide human beta-defensin 2 contributing to recurrent skin infections in AD.8Howell M.D. Novak N. Bieber T. Pastore S. Girolomoni G. Boguniewicz M. et al.Interleukin-10 downregulates anti-microbial peptide expression in atopic dermatitis.J Invest Dermatol. 2005; 125: 738-745Abstract Full Text Full Text PDF PubMed Scopus (172) Google Scholar Notably, NLRP2 seems to be a mediator of human beta-defensin expression as Ji et al9Ji S. Shin J.E. Kim Y.S. Oh J.E. Min B.M. Choi Y. Toll-like receptor 2 and NALP2 mediate induction of human beta-defensins by fusobacterium nucleatum in gingival epithelial cells.Infect Immun. 2009; 77: 1044-1052Crossref PubMed Scopus (54) Google Scholar observed a decrease in the bacterial-activation of human beta-defensins by NLRP2 silencing, suggesting an impaired host defense due to NLRP2 repression. Susceptibility of the skin to adverse environmental factors might increase as a result of such underlying immune deviations. Therefore, infiltrating immune cells with an impaired NLRP2 expression, as we have observed in children with early-onset AD, might not only contribute to IL-10 elevation in skin lesions of AD patients, but it might also impair pathogen defense due to its negative effect on defensin expression. This study broadens the view on how genetic variations in NLR genes might contribute to AD development. It describes a functional translation of the rs514148 genotype to NLRP2 alterations, which increase the risk for early-onset AD. These early life changes in NLRP2 might provide an opportunity to mitigate AD development by modulating innate immunity at a young age. We would like to thank Marion Bähr and Monika Helf who provided support in MassARRAY validation and Kathrin Jäger from the Core Unit Fluorescence Technologies at the University of Leipzig for cell sorting of PBMC subpopulations. We are grateful to Dr Gunda Herberth who provided information on cytokine concentrations in the LINA cohort. We cordially thank the participants of the LINA study as well as Susanne Arnold, Beate Fink, Anne Hain, Anna Regis, and Melanie Bänsch for their excellent technical assistance and fieldwork. This study is based on the prospective birth cohort LINA, which recruited 629 mother-child pairs from March 2006 until December 2008 in Leipzig, Germany. Standardized questionnaires (self-administered by the parents) on lifestyle/environmental exposures, and atopic outcomes of the children were recorded annually. Blood samples of children were obtained at the time of birth and each following year at scheduled clinical visits. Early-onset AD was defined as disease occurrence either reported as physician-diagnosed or as symptoms according to Hanifin and RajkaE1Hanifin J.M. Rajka G. Diagnostic features of atopic eczema.Acta Derm Venereol. 1980; 92: 44-47Google Scholar criteria until the age of 2 yearsE2Carlsten C. Dimich-Ward H. Ferguson A. Watson W. Rousseau R. Dybuncio A. et al.Atopic dermatitis in a high-risk cohort: natural history, associated allergic outcomes, and risk factors.Ann Allergy Asthma Immunol. 2013; 110: 24-28Abstract Full Text Full Text PDF PubMed Scopus (82) Google Scholar and late-onset AD as diagnosis after age 2 years until the age of 6 years. Controls were neither diagnosed with AD nor did they ever show any symptoms of AD. Participation in LINA was voluntary and approved by the Institutional Review Board of the University of Leipzig (046-2006, 160-2008, 160-2008, 160b/2008, 144-10-31052010, 113-11-18042011). Informed written consent was given by the parents. This study is based on a subcohort of the entire LINA cohort considering all children, for whom blood samples at birth and 1 year after were available together with AD information up to the age of 6 years. For cohort characteristics see Table E1. An overview of biosamples is given in Fig E1. RNA isolation and cDNA generation was performed as previously described.E3Bauer T. Trump S. Ishaque N. Thurmann L. Gu L. Bauer M. et al.Environment-induced epigenetic reprogramming in genomic regulatory elements in smoking mothers and their children.Mol Syst Biol. 2016; 12: 861Crossref PubMed Scopus (78) Google Scholar Briefly, cDNA was used as templates for gene expression measurements on the 96.96 Dynamic Array (Fluidigm, San Francisco, Calif) of NLRP2 (forward primer: 5′-ACGGAAAGAACGACCACCT-3′, reverse primer: 5′-GATGACATTTCCTTTCGTTTCTG-3′, Roche Universal Probe Library [UPL] #67) and NLRP3 (forward primer: 5′-GTGTCCTCCCAAGCTCCTCT-3′, reverse primer: 5′-AAGTGAGGTGGCTGTTCACC-3′, UPL #27) using GAPD (forward primer: 5′-GCTCTCTGCTCCTCCTGTTC-3′, reverse primer: 5′-ACGACCAAATCCGTTGACTC-3′, UPL #60) and GUSB (forward primer: 5′-CGCCCTGCCTATCTGTATTC-3′, reverse primer: 5′-TCCCCACAGGGAGTGTGTAG-3′, UPL #57) as reference genes. All reactions were performed in triplicates. Gene expression values were determined by 2−ΔΔCT method and normalized to the lowest measured value. Available cord blood raw DNA-methylation values of a subset of LINA samples with available WGBS data (European Genome-Phenome Archive under accession number EGAS00001000455, n = 13, bisulfite conversion efficiency per sample > 99.8%)E3Bauer T. Trump S. Ishaque N. Thurmann L. Gu L. Bauer M. et al.Environment-induced epigenetic reprogramming in genomic regulatory elements in smoking mothers and their children.Mol Syst Biol. 2016; 12: 861Crossref PubMed Scopus (78) Google Scholar, E4Trump S. Bieg M. Gu Z. Thurmann L. Bauer T. Bauer M. et al.Prenatal maternal stress and wheeze in children: novel insights into epigenetic regulation.Sci Rep. 2016; 6: 28616Crossref PubMed Scopus (48) Google Scholar were used for a preliminary assessment of NLRP2 promoter DNA-methylation differences in relation to AD development. Based on ENCODE ChromHMM definitions 57 cytosine guanine dinucleotides (CpGs) (with a WGBS coverage >5) in the promoter were analyzed. Genomic DNA of whole blood samples of time of birth, age 1 and 4 years was isolated using the QIAmp DNA Blood Mini Kit (Qiagen, Hilden, Germany), followed by bisulfite conversion using the EZ methylation kit (Zymo Research, Freiburg, Germany) according to the manufacturer's instructions (refer to Fig E1 for sample overview). As recommended by Suchiman et alE5Suchiman H.E. Slieker R.C. Kremer D. Slagboom P.E. Heijmans B.T. Tobi E.W. Design, measurement and processing of region-specific DNA methylation assays: the mass spectrometry-based method EpiTYPER.Front Genet. 2015; 6: 287Crossref PubMed Scopus (59) Google Scholar and Bock et al,E6BLUEPRINT ConsortiumQuantitative comparison of DNA methylation assays for biomarker development and clinical applications.Nat Biotechnol. 2016; 34: 726-737Crossref PubMed Scopus (215) Google Scholar the quality of the bisulfite-treated DNA was controlled by test PCRs yielding amplicons with increasing length. The bisulfite conversion efficiency was assed based on the MassARRAY R package version 1.28.0.E7Thompson R.F. Suzuki M. Lau K.W. Greally J.M. A pipeline for the quantitative analysis of CG dinucleotide methylation using mass spectrometry.Bioinformatics. 2009; 25: 2164-2170Crossref PubMed Scopus (67) Google Scholar Using the amplicon "TMEM241" (refer to Bauer et alE3Bauer T. Trump S. Ishaque N. Thurmann L. Gu L. Bauer M. et al.Environment-induced epigenetic reprogramming in genomic regulatory elements in smoking mothers and their children.Mol Syst Biol. 2016; 12: 861Crossref PubMed Scopus (78) Google Scholar and Table E5) quality scores of conversion efficiency were determined to be <2, as recommended by Suchiman et al.E5Suchiman H.E. Slieker R.C. Kremer D. Slagboom P.E. Heijmans B.T. Tobi E.W. Design, measurement and processing of region-specific DNA methylation assays: the mass spectrometry-based method EpiTYPER.Front Genet. 2015; 6: 287Crossref PubMed Scopus (59) Google Scholar PCR primers (Table E5) for the targeted DNA-methylation analysis of the NLRP2 promoter by MassARRAY in the entire LINA cohort—covering 12 CpGs—were designed according to the DNA-methylation changes observed in the WGBS data (Table E4). Bisulfite-treated DNA was PCR amplified using HotStarTaq DNA Polymerase (Qiagen) with the following cycling program: 95°C for 15 minutes, followed by 45 cycles of 94°C for 30 seconds, 52°C for 30 seconds, 72°C for 1 minute and a final elongation step at 72°C for 5 minutes on a T1 Thermocycler (Biometra, Göttingen, Germany). The PCR product was in vitro transcribed and enzymatically cleaved using the EpiTyper T Complete Reagent Set (Agena Bioscience, Hamburg, Germany) and subjected to mass spectrometry analysis. The validity of the derived DNA-methylation values was confirmed by correlation of expected and measured DNA-methylation values (NLRP2: R2 = 0.96) (Fig E3) using DNA-methylation standards (0%, 20%, 40%, 60%, 80%, and 100% methylated genomic DNA). For analysis of NLRP2/3 transcription and NLRP2 promoter DNA-methylation in monocytes, B cells, CD4+ and CD8+ T cells, PBMCs of 7 1-year-old healthy children and 7 age-matched children diagnosed with early-onset AD and showing AD symptoms according to Hanifin and RajkaE1Hanifin J.M. Rajka G. Diagnostic features of atopic eczema.Acta Derm Venereol. 1980; 92: 44-47Google Scholar criteria were used. The genotype distribution of rs514148 was not different between the control (n[AA] = 2, n[AG] = 3, n[GG] = 2) and the early-onset AD group (n[AA] = 3, n[AG] = 2, n[GG] = 2; Fisher exact test for cross relationship: P = 1.000). First, cells were stained with Fixable Viability Dye eFluor 450 (#65-0863-14, eBioscience, Frankfurt, Germany) for 30 minutes at 4°C. Subsequently, cells were washed with PBS containing 1% FCS and incubated with Cyto-Stat Coulter Clone CD3-PC5 liquid murine monoclonal antibody (PN 6607010, Beckman-Coulter, Krefeld, Germany), CD4 Antibody Qdot 605 conjugate (#Q10008, Molecular Probes, Thermo Fisher Scientific, Darmstadt, Germany), CD8-PE (#0452, Beckman-Coulter), CD14-FITC (PN IM0645, Beckman-Coulter), and allophycocyanin Mouse Anti-Human CD19 (Clone: HIB19: #561742, BD Bioscience, Heidelberg, Germany) for 20 minutes at 4°C. Washed and stained cells were sorted on a BD FACS Aria III, SORP (BD Bioscience). Total RNA was extracted using peqGold RNA Pure (peqlab, Erlangen, Germany) following manufacturer's instructions and, except for monocyte cell fraction, genomic DNA was isolated QIAmp DNA Blood Mini Kit (Qiagen, Hilden, Germany). Gene expression in PBMC subpopulations was assessed on a LightCycler 480 (Roche Applied Science, Mannheim, Germany) with the following settings: 5 minutes at 95°C, 45 cycles at 95°C for 15 seconds, followed by 1 minute at 60°C and 30 seconds at 72°C using the FastStart Universal Probe Master Mix (Rox) (Roche Applied Science). A semiquantitative approach based on external standard curves and in-run calibrator samples was used to determine gene expression normalized to the calibrator and GAPD as reference gene. To determine the genotype of the SNP rs514148 (chr19:55,474,502, GRCh37) in the subcohort with available DNA-methylation information at birth and age 1 year (n = 365), PCR products (forward primer: 5′-ttcaagacctcagacaggtt-3′, reverse primer: 5′-cttactgaaggctacaggc-3′, sequencing primer: 5′-aaatgggtaggatcatggc-3′) obtained from whole blood genomic DNA were analyzed by Sanger sequencing (GATC Biotech AG, Köln, Germany). Genotype distribution and allele frequency were tested for Hardy-Weinberg equilibrium (goodness-of-fit chi-square, P = .12, n[AA] = 118, n[AG] = 166, n[GG] = 81). Concentrations of 10 different cytokines (see Table E6) were determined in the supernatants of whole blood samples of 1-year-old children as previously described.E8Englich B. Herberth G. Rolle-Kampczyk U. Trump S. Röder S. Borte M. et al.Maternal cytokine status may prime the metabolic profile and increase risk for obesity in children.Int J Obes (Lond). 2017; 41: 1440-1446Crossref PubMed Scopus (16) Google Scholar Parental smoking behavior including the number of cigarettes smoked in the homes of study participants during pregnancy was assessed by questionnaire. Indoor benzene concentrations were measured between the 34th and 36th week of gestation using passive samplers.E9Winter M. Thürmann L. Gu Z. Schüürmann G. Herberth G. Hinz D. et al.The benzene metabolite 1,4-benzoquinone reduces Treg function: a potential mechanism for tobacco smoke-associated atopic dermatitis.J Allergy Clin Immunol. 2017; 140: 603-605Abstract Full Text Full Text PDF PubMed Scopus (2) Google Scholar Chi-square was used to test for equal parameter distribution in the subcohort and the entire LINA cohort. As nonparametric tests, either rank sum analysis by Mann–Whitney U or Spearman correlations were applied. Logistic regression models were applied to determine OR and 95% CI adjusting for possible confounders of AD (sex, maternal AD, parental educational level, smoking during pregnancy, breast-feeding, and cat keeping). Transcription and DNA-methylation models were additionally adjusted for gestational age and mode of delivery. Multiple regression models were applied for the association analysis of continuous variables, results are given as mean ratios (MRs) with 95% CIs. All continuous variables were log-transformed. Two-way ANOVA was applied to assess differential methylated CpGs in the NLRP2 promoter. To evaluate the relationship of indoor benzene concentrations and tobacco smoke exposure 1-way ANOVA was applied. Genotype differences in NLRP2 transcription and promoter DNA-methylation were assessed by Kruskal-Wallis test. To evaluate the relationship of prenatal benzene exposure, NLRP2 DNA-methylation, DNA-transcription, and AD development, a mediation analysis was conducted (PROCESS macro v2.16 for SPSS, www.processmacro.org) with a bootstrap approach using 10,000 samples to estimate the statistical significance of the indirect effect as bias-corrected CIs.E10Hayes A.F. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. Guilford Press, New York, NY2013Google Scholar P values ≤ .05 were considered significant. Analyses were performed in either STATISTICA 13.0 for Windows (Dell Inc, Round Rock, Tex), SigmaPlot version 12.0 (Systat Software GmbH, Erkrath, Germany), IBM SPSS Statistics version 22 (IBM Corp, Armonk, NY) or R 3.0.0.E11R Core TeamR: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria2013http://www.R-project.org/Google ScholarFig E2AD incidence. Incidences of AD in the LINA cohort up to the age of 6 years are shown separately for children with early- or late-onset AD. The percentage was calculated as the number of new AD cases related to the number of valid questionnaires per year.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E3Quality control of the MassARRAY amplicon used for NLRP2 promoter DNA-methylation assessment. The graph shows DNA-methylation values derived by MassARRAY measurements of standard samples (0%, 20%, 40%, 60%, 80%, and 100% methylated genomic DNA) representing the mean DNA-methylation of all 12 CpGs within the MassARRAY amplicon (given are mean ± SD DNA-methylation values of 4 independent measurements).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E4Relationship between NLRP2 DNA-methylation and transcription. A, NLRP2 promoter DNA-methylation is significantly negatively correlated with NLRP2 transcription at the time of birth (n = 326) and age 1 year (n = 386, P value from Spearman correlation). B, This relationship is sustained in the adjusted regression model at the time of birth and age 1 year (birth: adj MR = 0.57, 95% CI: 0.63-0.71; age 1 year: adj MR = 0.57, 95% CI: 0.63-0.69; MRs calculated by multiple regression, adjusted for AD confounders, mode of delivery, and gestational age).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E5NLRP2 and NLRP3 transcription in PBMC subpopulations of children with early-onset AD. A, NLRP2 transcription is decreased in children with early-onset AD in comparison to healthy controls in the different PBMC cell subtypes investigated (2-way ANOVA: P = .0003). B, NLRP2 promoter hypermethylation observed in children with early-onset AD compared with healthy controls in the different PBMC subpopulations. DNA-methylation was determined by MassARRAY. Numbers in the black bars indicate the DNA-methylation difference between healthy controls and children suffering from early-onset AD (2-way ANOVA: P = .009). C, Normalized NLRP3 expression in the different PBMC subtypes showed no difference comparing children with early-onset AD to healthy controls. (*P < .05 and ***P < .0005 from 2-way ANOVA, mean ± SEM, n = 7 AD, n = 7 controls; n = 6 AD and n = 5 controls for NLRP2 in monocytes; n = 6 AD and n = 5 controls for NLRP2 methylation in CD8+ T cells.)View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E6NLRP2 regulation is influenced by the genotype. NLRP2 transcription and promoter methylation is dependent on the genotype of rs514148 (chr19:55474502, human genome assembly GRCh37) located upstream of the NLRP2 promoter line represents the median, P value from Kruskal-Wallis (KW) test, expression quantitative trait loci (eQTL) n= 111 AA, 158 AG, 78 GG and methylation QTL (meQTL) n= 118 AA, 166 AG, 81GG.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E7Indoor benzene concentrations as a proxy for smoking behavior. Shown are ln-transformed indoor benzene concentrations measured in the homes of study participants during pregnancy and their relationship to the frequency of indoor smoking behavior (n = 378 never-, n = 27 occasional-, n = 25 almost daily smokers) (A) and the total number of cigarettes smoked in the homes of the study participants (no cigarettes: n = 374, 1-10 cigarettes per day: n = 31, >10 cigarettes per day: n = 14) (B). Graphs depict mean ± SEM.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E8Effect of prenatal tobacco smoke exposure on early-onset AD is mediated by NLRP2. The mediation analysis suggests that prenatal exposure to high benzene concentrations influences the development of early-onset AD via changes in NLRP2 promoter methylation and transcription. The table summarizes unstandardized effect sizes (n = 292, adjusted for AD confounders and rs514148 genotype, significance determined by bias-corrected 90% CI of 10,000 bootstrapped samples, n.a.: no P-values are calculated for this model type by the PROCESS macro, effects are significant if confidence intervals do not contain 0).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table E1Characteristics of the entire LINA cohort and the investigated AD subcohortEntire cohortn = 629n (%)∗Numbers may be different from total sum due to missing data.Subcohortn = 430n (%)∗Numbers may be different from total sum due to missing data.P value†P value from chi-squared test for cross relationship.Sex of the child Male330 (52.5)234 (54.4).573 Female299 (47.5)196 (45.6)Mode of delivery Spontaneous469 (74.6)339 (78.9).665 Caesarean section126 (20.0)84 (19.5)Gestational age 33rd-38th week124 (19.7)83 (19.3).723 39th-41st week427 (67.9)287 (66.7) 42nd-43rd week77 (12.2)60 (14.0)Tobacco smoke exposure during pregnancy‡Answers based on question "Did you or anybody else smoke inside your home during the last 12 months?" (Almost) daily48 (7.6)25 (5.8).366 Occasionally47 (7.5)27 (6.3) Never534 (84.9)378 (87.9)Breast-feeding (until 6th month) Yes442 (70.3)313 (72.8).719 No139 (22.1)105 (24.4)Parental education level§Low, 9 years of schooling or less "Hauptschulabschluss"; intermediate, 10 years of schooling "Mittlere Reife"; high, 12 years of schooling or more "(Fach-)hochschulreife." Low16 (2.5)9 (2.1).736 Intermediate144 (22.9)106 (24.7) High469 (74.6)315 (73.3)Cat keeping Yes95 (15.1)66 (15.3).955 No511 (81.2)364 (84.7)Maternal history of AD Positive112 (17.8)74 (17.2).866 Negative517 (82.2)356 (82.8)AD Never378 (60.1)264 (61.4).929 Ever (1st-6th year)228 (36.2)166 (38.6) Early-onset (1st or 2nd year)138 (21.9)105 (24.4) Late-onset (>2nd year)90 (14.3)61 (14.2)∗ Numbers may be different from total sum due to missing data.† P value from chi-squared test for cross relationship.‡ Answers based on question "Did you or anybody else smoke inside your home during the last 12 months?"§ Low, 9 years of schooling or less "Hauptschulabschluss"; intermediate, 10 years of schooling "Mittlere Reife"; high, 12 years of schooling or more "(Fach-)hochschulreife." Open table in a new tab Table E2ORs for risk increase of AD development in association to transcription of NLRP2 and NLRP3GeneAgeRaw modelAdjusted modeln (control/AD)OR (CI 95%)P valuen (control/AD)Adj OR (CI 95%)∗Logistic regression adjusted for sex, breast-feeding, maternal history of atopic eczema, parental educational level, cat keeping, and prenatal tobacco smoke exposure.P valueAD ever NLRP21253/1590.82 (0.72-0.94).004247/1530.83 (0.72-0.95).009 NLRP2499/870.81 (0.65-1.01).06497/830.79 (0.62-1.02).064 NLRP31252/1560.91 (0.76-1.10).319246/1500.93 (0.77-1.13).456 NLRP3499/881.13 (0.69-1.86).61997/841.20 (0.71-2.04).490Early-onset AD NLRP21253/1020.76 (0.65-0.88).0002247/1000.76 (0.65-0.89).0005 NLRP2499/520.71 (0.55-0.93).01197/500.71 (0.52-0.95).021 NLRP31252/990.89 (0.72-1.11).311246/970.49 (0.75-1.18).610 NLRP3499/531.15 (0.65-2.03).62097/511.24 (0.66-2.34).497Late-onset AD NLRP21253/571.01 (0.82-1.25).897247/531.02 (0.81-1.29).856 NLRP2499/350.98 (0.74-1.82).87497/330.95 (0.70-1.29).720 NLRP31252/570.94 (0.72-1.22).633246/530.90 (0.68-1.02).481 NLRP3499/351.10 (0.55-2.22).78097/331.15 (0.53-2.50).717∗ Logistic regression adjusted for sex, breast-feeding, maternal history of atopic eczema, parental educational level, cat keeping, and prenatal tobacco smoke exposure. Open table in a new tab Table E3Mean and median of normalized expression values of NLRP2 for controls and children who develop AD ever, early-onset, or late-onset AD up to the age of 6 yearsGene (age)ControlAD everEarly-onset ADLate-onset ADMean ± SEMMedian (IQR)Mean ± SEMMedian (IQR)Mean ± SEMMedian (IQR)Mean ± SEMMedian (IQR)NLRP2 (1)6.62 ± 0.096.83 (5.82-7.57)6.17 ± 0.146.46 (5.47-7.31)5.89 ± 0.196.38 (5.01-7.18)6.68 ± 0.176.73 (5.82-7.47)NLRP2 (4)5.93 ± 0.136.18 (5.26-6.89)5.51 ± 0.145.59 (4.73-6.37)5.26 ± 0.165.29 (4.54-6.29)5.89 ± 0.266.13 (5.18-6.82)Controls: n = 247, age 1: 97, age 4: 150; children who develop AD ever: n = 153, age 1 year: 83, age 4 years: 70; early-onset: n = 100, age 1 year: 50, age 4 years; or late-onset AD: n = 53, age 1 year: 33, age 4 years: 20.IQR, Interquartile range. Open table in a new tab Table E4NLRP2 promoter DNA-methylation differences between children with early-onset AD (n = 7) and controls (n = 6) based on WGBS dataGenomic position of CpG∗Human genome assembly GRCh37.Mean DNA-methylation control in % (SE)Mean DNA-methylation AD in % (SE)ΔDNA-methylation %Bonferroni corrected P valuechr19:5547745471.1 (7.6)84.5 (3.5)13.4.367chr19:5547747057.4 (5.7)80.3 (3)22.9.099chr19:5547747466.5 (7.9)82.2 (2.5)15.7.261chr19:5547749947.8 (5.9)78.9 (4.2)31.1.014chr19:5547755747.2 (7.7)76 (2.9)28.8.015chr19:5547756148 (8.5)71.4 (4.8)23.4.031chr19:5547763332.6 (9.4)59.3 (6)26.7<.001chr19:5547765339.1 (10.2)67.7 (5.1)28.6.002chr19:5547767034.9 (11.1)65.1 (4.7)30.2.017chr19:5547769732 (7.2)61.8 (4.6)29.8<.001chr19:5547772334.1 (7.5)59 (6.4)24.9.007chr19:5547772943.2 (9.3)74.1 (7.2)30.9.004chr19:5547773436.9 (7.6)67.6 (5.6)30.8<.001chr19:5547775539.5 (7.1)72.2 (5.7)32.7.002chr19:5547775941.5 (7.2)67.1 (5.8)25.6.011chr19:5547780745.8 (8.4)70.2 (6.5)24.5.021chr19:5547781044.8 (9.4)72.9 (4.9)28.1.003chr19:5547781343.4 (7.8)70.6 (5.8)27.2.01chr19:5547784939.2 (9.4)72.4 (6.9)33.2<.001chr19:5547793638.7 (9.5)66.6 (8)27.8<.001chr19:5547794436.6 (9.9)69.8 (8.6)33.2<.001chr19:5547795233.5 (7.6)69.8 (7.6)36.3<.001chr19:5547796842.1 (10.9)65.6 (6.3)23.5.202chr19:5547797440 (9.5)68.9 (6.3)28.9.107chr19:5547797636.4 (9.7)62.3 (7.7)25.8.107chr19:5547798237.4 (9.4)71.3 (6.2)34.0<.001chr19:5547799946.8 (11.7)71.7 (6.4)25.0.002chr19:5547800633.3 (9.6)64.8 (6.6)31.6<.001chr19:5547803835.5 (9.3)73.9 (5.4)38.4<.001chr19:5547810448.1 (10.9)66.3 (6.4)18.2.027chr19:5547814057.3 (6.3)75 (6.9)17.7.224chr19:5547814548.2 (9)69.1 (6.3)20.8.058chr19:5547816147.8 (8.7)63.7 (7.8)15.9.141chr19:5547817348.2 (9.6)74.7 (7.1)26.5.017chr19:5547818343.6 (10.6)72.2 (11.6)28.5.007chr19:5547818541.6 (9)75.6 (9.4)34.0.002chr19:5547826246.1 (13)79.8 (8.6)33.7.063chr19:5547827841.9 (9.5)79.8 (8.8)37.9<.001chr19:5547832166.8 (5)86.5 (5)19.7.222chr19:5547834875.2 (6.2)83.7 (5.3)8.5.609chr19:5547844283.5 (2.8)92.7 (2.6)9.2.626chr19:5547847286.9 (2.3)92.6 (2.1)5.6.772chr19:5547852491.2 (2.2)93.9 (2.1)2.8.89chr19:5547852697.2 (1.4)94.6 (1.9)−2.6.897chr19:5547853095.8 (1.4)92.5 (2.8)−3.2.865chr19:5547858891.4 (3.1)94.8 (2.7)3.4.867chr19:5547859895.8 (1.4)93.1 (1.4)−2.7.893chr19:5547860384.6 (4.7)89.1 (3.4)4.6.793chr19:5547861691.7 (2.1)96.3 (1.8)4.6.82chr19:5547862198.6 (1.4)94.6 (4.2)−4.0.826chr19:5547864090.3 (2.6)91.6 (1.7)1.2.946chr19:5547864895.9 (1.5)97.8 (1.1)1.9.927chr19:5547871096.6 (2.4)96.4 (1.8)−0.2.995chr19:5547871898.2 (1.3)96.5 (1.5)−1.7.935chr19:5547877477.7 (6.9)91.8 (3.9)14.1.401chr19:5547878197.1 (1)97.7 (1.5)0.7.976chr19:5547879098.8 (0.8)94.3 (1.7)−4.5.827CpGs considered in the MassARRAY analysis are in boldface. Shown are Bonferroni-corrected P values from 2-way ANOVA.∗ Human genome assembly GRCh37. Open table in a new tab Table E5MassARRAY primerAssay nameAmplicon coordinate∗Human genome assembly GRCh37.Forward primer (5′→3′)Reverse primer(5′→3′)TMEM241chr18:21002501-21002763x-TTATGTTTGGTTTATAGTATTAGy-TATACTACCCCTACACTATACNLRP2chr19:55477611-55477835x-TGGAGGATTATAGGTGGAGy-TCCCAACACCTAACCCTACx = AGGAAGAGAG; y = CAGTAATACGACTCACTATAGGGAGAAGGCT.∗ Human genome assembly GRCh37. Open table in a new tab Table E6Cytokine protein concentrations in sera of 1-year-old children: Controls (n = 262), diagnosed with early-onset AD (n = 62)CytokineControl mean ± SE(pg/mL)Early-onset AD mean ± SE (pg/mL)Control median (IQR) (pg/mL)Early-onset AD median (IQR) (pg/mL)P value∗P value from Mann-Whitney U test.Bonferroni corrected P < .05IL-48.4 ± 0.38.0 ± 0.57.3 (5.8-9.3)7.1 (5.8-9).7056NoIL-54 ± 0.14.2 ± 0.33.8 (3.3-4.4)3.8 (3.5-4.3).5887NoIL-671.2 ± 18.294.4 ± 32.124.1 (13.1-46.5)32.4 (15.3-71.6).1295NoIL-8442.7 ± 33.4751.3 ± 139.7223.2 (106.2-557)402.1 (105.3-745.7).1506NoIL-109.4 ± 0.211.0 ± 0.88.5 (7.1-10.5)9.8 (8.2-11.8).0001.001IL-126.4 ± 0.36.5 ± 0.45.6 (4.6-7.1)6.0 (4.6-7.9).2962NoIL-138.1 ± 0.38.4 ± 0.57.5 (4.7-10.2)8.1 (6.2-10.5).2667NoIFN-γ10.4 ± 0.410.8 ± 1.710.2 (5.1-14.1)9.7 (2.8-13.4).2525NoMCP352.1 ± 22.9465 ± 64.7232.9 (154.9-416.4)275.5 (154.5-549).4357NoTNF-α44.8 ± 12.562.0 ± 22.419.2 (12.1-34.9)21.6 (13.6-34.3).3741NoMCP, Monocyte chemoattractant protein.∗ P value from Mann-Whitney U test. Open table in a new tab Controls: n = 247, age 1: 97, age 4: 150; children who develop AD ever: n = 153, age 1 year: 83, age 4 years: 70; early-onset: n = 100, age 1 year: 50, age 4 years; or late-onset AD: n = 53, age 1 year: 33, age 4 years: 20. IQR, Interquartile range. CpGs considered in the MassARRAY analysis are in boldface. Shown are Bonferroni-corrected P values from 2-way ANOVA. x = AGGAAGAGAG; y = CAGTAATACGACTCACTATAGGGAGAAGGCT. MCP, Monocyte chemoattractant protein.
Referência(s)