| Background Psoriasis(PsO)is a common chronic,inflammatory,and recurrent skin disease,and its pathogenesis is not completely clear.Epigenetic studies have shown that DNA methylation(DNAm)is associated with PsO,and DNAm regulates the expression of PsO susceptible genes,thus affects the risk of PsO in populations.Epidemiological studies have shown that some autoimmune diseases coexist with PsO,and PsO is strongly associated cardiovascular diseases(CVD)and cardiovascular risk factors,as well as possible mutual effect relationships between PsO and those diseases or risk factors.In the past ten years,genome-wide association study(GWAS)has been widely used in many diseases and traits,and a large amount of datasets has been accumulated.Researchers have performed GWAS Imputation and obtained many more genetic loci.Meanwhile,many new statistical methods have also been developed,making it possible for a series of subsequent data analysis and in-depth mining.Objective The aims of this study are:To explore whether PsO has an effect on DNAm age of psoriatic lesional skin tissue.To explore the pleiotropic loci and possible shared pathways between PsO and five kinds of autoimmune diseases(systemic lupus erythematosus(SLE),rheumatoid arthritis(RA),multiple sclerosis(MS),Crohn's disease(CD),and ulcerative colitis(UC)).To explore whether there are causal associations between PsO risk and genetic variants of CVD(coronary heart disease(CHD),stroke,heart failure(HF),and atrial fibrillation(AF))or cardiovascular risk factors(high density lipoprotein cholesterol(HDL),low density lipoprotein cholesterol(LDL),total cholesterol(TC),triglyceride(TG),diastolic blood pressure(DBP),and systolic blood pressure(SBP)).Data sources and research methods The DNAm dataset of the PsO(Including 114 psoriastic lesional skin tissue(PP)from PsO patients,41 psoriastic non-lesional skin tissue(PN)from PsO patients,and 62 normal skin tissue(NN)from healthy controls),PsO and SLE GWASs Imputation datasets of the Chinese Han population were derived from the research projects carried out by Institute of Dermatology,Anhui Medical University.We got PsO GWAS Imputation dataset of European population from Professor James T Elder's research group at The University of Michigan,United States;the European population SLE GWAS Imputation dataset of European population from the Professor Timothy J Vyse's research group at King's College London,United Kingdom;MS GWAS Imputation dataset from International Multiple Sclerosis Genetics Consortium(IMSGC);CD and UC GWASs Imputation datasets from International Inflammatory Bowel Disease Genetics Consortium(IIBDGC).The UK Biobank(UKBB)GWASs datasets for five kinds of autoimmune diseases(SLE,RA,MS,CD,and UC)in European populations were from Benjamin Neale's laboratory(http://www.nealelab.is/uk-biobank)at Broad Institute of MIT and Harvard.Instrumental variables(IVs)were extracted from previous published literatures based on inclusion criteria.Data analysis was performed by using methods,such as multiple-tissue methylation clock model,empirical weighted Linear Combination test statistics(eLC),the Database for Annotation,Visualization and Integrated Discovery(DAVID),Gene Set Enrichment Analysis(GSEA),Mendelian Randomization Analysis(MRA),and Robust Adjusted Profile Score(RAPS).Results(Part I)DNAm age analysis results:This study indicated that the applicability of the multi-tissue methylation clock model in skin tissue.For PN and N-N,the calculated DNAm age is significantly positively correlated with chronological age(Spearman's ??0.86,P<2.2E-16),and the age-accelerated residual is close to 0(P>0.15).For PP,there is also a strong correlation between DNAm age and chronological age(Spearman's p=0.78,P<2.2E-16),and the age-accelerated residual is close to NN or PN(P>0.3).In addition,in PP,there is no correlation between DNAm age-accelerated residuals and PsO area and severity index(PASI)score,related features(age,gender,body mass index(BMI),and smoking status).This study also found that there are 100,588,and 8,032 CpG sites related to DNAm age in NN,PN,and PP,respectively(False discovery rate(FDR)<0.05).These DNAm age-related CpG sites have overlaps and have consistent correlation coefficients among these three groups(NN,PN,and PP)and with external dataset.It is worth noting that 39 DNAm age-related CpGs in PP and 2 clock CpGs also differentially methylated in PsO.Comparison between PP vs.PN and PP vs.NN identified 301 and 881 PsO-dependent DNAm age-related CpGs(P<0.01),respectively,but none of these sites reached the significant threshold of FDR<0.05.(Part II)Results of pleiotropic loci and shared pathways between PsO and five kinds of autoimmune diseases:In the European population,LD-Hub calculation results show that PsO is significantly genetically related to autoimmune diseases,metabolic diseases,and basic human characteristics.The results of the stage I showed that the common significant loci/genomic regions/independent loci of PsO and five kinds of autoimmune diseases were:SLE(1244/14/35),RA(11535/16/114),MS(10426/23/102),CD(1175/22/73),and UC(97/1/11).The significant pleiotropic loci from stage I were further verified in corresponding UKBB datasets,the number of verified independent loci of five autoimmune diseases are:SLE(8),RA(16),MS(11),CD(12),and UC(7).Meanwhile,bivariate GWAS analysis of GWASs Imputation datasets of PsO and SLE in Chinese Han population showed that only SNP rs 1887428(chromosome 9,the nearest gene is JAK2,P_dLC=6.20E-09)was statistically significant in outside the MHC region,and all other significant loci are in MHC region and were not significant in previous GWASs.Pathway analysis based on the significant poetential pleiotropic loci from stage I by using DAVID and GSEA,results showed that the shared pathways of PsO and four kinds of autoimmune diseases(SLE,RA,MS,and CD)mainly involving autoimmune diseases,infectious diseases,inflammatory diseases related pathways and other pathways,and there are some overlaps.(Part III)MRA results of causal associations between PsO and CVD or cardiovascular risk factors:After corresponding selection criteria screening and data harmonisation,29,26,10,131,83,63,75,50,121,and 105 SNPs of CHD,Stroke,HF,AF,HDL,LDL,TC,TG,DBP,and SBP were included and used as IVs for Two-Sample Mendelian Randomization(TSMR)analysis.The horizontal pleiotropic test shows that the genetic variation related to stroke has horizontal pleiotropy.TSMR analysis results showed that genetically predicted CHD and PsO risk was statistically significant(OR=1.20,95%CI:1.06-1.45,P=3.05E-03),SBP and PsO risk with suggestive statistical significance(OR=1.51,95%CI:1.08-2.11,P=1.53E-02),and there was no significant genetic causal association between other exposures and PsO risk.For stroke,Mendelian Randomization Pleiotropy RESidual Sum and Outlier(MR-PRESSO)analysis result also showed no statistical significance between stroke and risk of PsO.Further RAPS analysis results showed that CHD(OR=1.21,95%CI:1.06-1.37,P=3.44E-03)and HF(OR=1.72,95%CI:1.22-2.43,P=2.08E-03)have significant associations with PsO risk,SBP has suggestive association with PsO risk(OR=1.53,95%CI:1.09-2.14,P=1.30E-02).There was no significant heterogeneity in PsO with CHD,HF,and SBP.Leave-one-out analysis showed the continuous and significant causal effects of CHD,HF,and SBP on PsO risk,supporting the robustness of the results of TSMR and RAPS analysis.Conclusions This study comfirms that the multi-tissue methylation clock model has good applicability in the skin tissues of Chinese Han population.No significant change in DNAm age was observed in PN and PP,and there was a lack of correlation between related features and DNAm age accelerated residuals.PsO-induced keratinocyte proliferation and DNAm changes may not affect the biological age of PsO skin tissue.Some pleiotropic loci and shared pathways were identified,these results can partially explain the coexistence between PsO and autoimmune diseases,which will help us to better understand the genetic mechanism of PsO and autoimmune diseases.Causal associations between CHD,HF,and SBP and PsO risk were identified by using MR methods,which will help us to better understand the genetic relationships between PsO risk and CVD or cardiovascular risk factors,as well as further treat and manage PsO patients with coexisting diseases reasonably. |
PDF Full Text Request