Vitamin C promotes epidermal proliferation by enhancing the DNA demethylation of proliferation-related genes in human epidermal equivalents (WGBS)

The stratified structure of the epidermis, the outermost layer of the skin, is highly regulated through the differentiation of keratinocytes. Vitamin C act as a cofactor for DNA demethylation enzyme ten-eleven translocation (TET) in epigenetic mechanism. Previous study revealed that VC deficiency causes epidermal atrophy in hairless mice unable to synthesize VC. Genomic DNA was extracted from a half portion of human epidermal equivalents using the Monarch Genomic DNA Purification Kit (New England Biolabs, Ipswich, MA, US). The integrity and quality of the DNA were evaluated using a NanoDrop One and TapeStation (Agilent Technologies) prior to analysis. DNA methylation sequencing was performed and Chithe data analysed using a previously described WGBS methodology63. Bisulfite treatment was performed by using a Swift BiosciencesTM Accel-NGS® Methyl-Seq DNA Library Kit (Swift Biosciences, Ann Arbor, Michigan, US) to generate WGBS libraries. The amplified libraries were sequenced using an Illumina Nova Seq 6000 with 150 bp paired-end reads following standard Illumina sequencing protocols. To analyse the bisulfite sequencing results, the quality of the raw paired-end sequence reads was assessed with FastQC (version 0.11.7). Low-quality (< 20 bases) or 5'-end 10-base sequences and adapter sequences were trimmed by Trim Galore (version 0.5.0) with the options "-q 20 --phred33 --clip_R1 10 --clip_R2 10 --paired". The trimmed reads were aligned to the reference genome using methylpy (version 1.4.6) with the options "--remove-clonal True --trim-reads FALSE". Annotation was performed with in-house software. The number of methylated cytosines in the 1,000 bases of nonoverlapping windows across the whole genome was estimated by the tileMethylCounts function from the methylKit (version 1.10.0) package. Normalization was performed using the normalizeCoverage function in methylKit. Bases with less than 10× read coverage and bases with more than 99.9th percentile coverage in each sample (likely PCR artifacts with abnormally high coverage) were removed by the filterByCoverage function of methylKit. All the samples were then merged by using the unite function in methylKit with the destrand = FALSE option. To identify DMRs, differential methylation was calculated using the calculate DiffMeth function of methylKit. Regions with a q value less than 0.01 and a methylation change greater than 25% between the groups were considered significant DMRs. A cluster analysis heatmap of the DMRs and DEGs was generated with ComplexHeatmap (version 2.12.1). To perform an integrated analysis of the gene expression and DNA methylation profiles, normalization of the gene expression levels was performed using the robust multiarray average (RMA) method with oligo software, and the expression levels were corrected among the samples. The program limma was used for comparisons between the groups.