Genome-wide DNA Methylation Study Of Severe Acne

Objective:Acne is the most common inflammatory follicular sebaceous gland disease.Severe acne is the most severe form of acne,which has a wide range of lesions and may produce scars.In sickest cases,it may lead to mental illness or even suicidal tendency.DNA methylation plays an important role in some immune skin diseases such as psoriasis and systemic lupus erythematosus(SLE).But the relationship between DNA methylation and severe acne has not been reported.This study detected the methylation sites of patients with severe acne through whole-genome DNA methylation chip.Finally,differentially methylated sites(DMSs)related to the pathological process of severe acne were found.This is helpful for the in-depth understanding of the molecular mechanism of severe acne and provides the scientific basis for early diagnosis and treatment of severe acne.Methods:Blood samples from patients with severe acne selected from the Acne Genetic Resource Bank.The healthy control samples included in this study strictly screened according to the requirements.The experiment is divided into two phases.The number of severe acne patients and healthy control samples collected in the two phases of the experiment is 24 pairs,and 20 pairs,respectively.Extracted the whole genome DNA of all blood samples by DNA extraction kit.DNA integrity was examined using the agarose gel electrophoresis experiment.DNA concentration was detected using a micro nucleic acid protein concentration analyzer.Performed whole-genome amplification and hybridization after processing the DNA sample with bisulfite.Finally,detected the methylation of the whole genome through the Illumina Methylation EPIC Bead Chip.Cleaned and sorted raw data were obtained from chip,then did the following processing: constructed the methylation profile of normal skin by methylation data from 44 healthy control samples;analyzed the DMSs and differentially methylated regions(DMRs)between patients with severe acne and healthy controls by package Ch AMP in R software;performed GO and KEGG pathway enrichment analysis by KOBAS’s online tool;screened hub genes by the cyto Hubba plugin of Cytoscape software;performed methylation quantitative trait loci(me QTL)analysis by package Matrixe QTL in R software;built six machine learning models for the diagnosis of severe acne by R software including logistic regression,support vector machine,k-nearest neighbor,random forest,gradient-enhanced decision tree,and neural network.Results : The results of agarose gel electrophoresis and micro nucleic acid protein concentration analyzer indicated that the DNA samples were qualified and could be used for subsequent experiments.767,146 methylation sites were identified by 44 healthy control samples.Among these loci: 25.73% were hypomethylated and 49.05% were hypermethylated;19.02% were located in Cp G islands and had lower methylation levels than those located in non-Cp G islands.A total of 904 DMSs were screened out in the first phase of the experiment,which in the second phase were 646.A total of 275 differentially methylated sites were selected as the final results,which were overlapped in the results of the two phases and had the same direction of β-value change.They were located in 194 unique genes.A total of 5 DMRs were obtained,of which 4 regions were located in genes ARG1,HTRA4,RFX3,and SPTLC2,and1 region was located in the intergenic region.The results of GO and KEGG enrichment analysis showed that the differential methylated genes involved many biological processes and pathways such as protein binding,cytokine regulation,and cancer pathways.The top 10 hub genes screened were TCEB1,CUL2,ANAPC7,FBXO9,UBE3 C,ZNRF1,PLCB1,PSMB7,PPP3 CA,and ITPR3,all of which are involved in immune-related signaling pathways,suggesting that they may regulate the development of severe acne through the immune response process.The me QTL analysis revealed a strong association between rs6727948 and cg08977295,suggesting that rs6727948 may regulate the expression of the WDR43 gene by regulating the methylation status of cg08977295.Six models for diagnosing severe acne were constructed through machine learning,of which RF has the best overall performance.Conclusions:Differences existed in genome-wide DNA methylation between patients with severe acne and healthy controls.DNA methylation is likely to be a potential factor in the pathogenesis of severe acne and is one of the epigenetic regulatory mechanisms of severe acne.DMSs and differential genes related to severe acne were screen out by methylation chip technology,which is laying the foundation for epigenetic research,early diagnosis,and treatment of severe acne.Machine learning models can be used to diagnose severe acne,but still,need to be optimized.