Fig. 1: Analysis of global DNA methylome in human alpha and beta cells.
a, Workflow for the current study. Human pancreatic islets from 24 donors with or without T2D were used for sorting of alpha and beta cells after staining for glucagon and insulin. WGBS and RNA-seq were used to analyse the methylome and transcriptome in sorted alpha and beta cells. Bioinformatic analyses were used to identify differential DNA methylation and gene expression. Follow-up experiments were done in human islets, EndoC-βH1 cells, INS-1 beta cells and GK rats, by epigenetic editing, analysing gene expression, insulin secretion and metabolism. b, Filtering islet samples for WGBS and RNA-seq. The Lund University Diabetes Centre (LUDC) islet sorting cohort consists of alpha and beta cells sorted from 24 human islet donors. Based on samples with enough DNA, WGBS was performed on sorted alpha and beta cells from 11 and 14 islet preparations, respectively. Among WGBS samples passing quality control (one alpha cell sample did not), 3 donors had pre-T2D/T2D (both alpha and beta cell), and 7 (alpha cell) and 11 (beta cell) were controls. RNA-seq was performed on sorted alpha and beta cells from 22 and 24 islet preparations, respectively. Among RNA-seq samples passing quality control (two beta cell samples did not), 7 donors had pre-T2D/T2D and 15 were controls (both alpha cells and beta cells). c, Density plot showing the global methylome of human alpha (mean 76.1%) and beta (mean 77.2%; n = 7 donors) cells. Peaks near 0% and 100% methylation. d,e, Density plots showing the degree of DNA methylation in alpha and beta cells in different genomic (d) and CpG island (e) regions. d, The region 1 to 5 kilobase pairs (kb) upstream of the transcription start site (TSS) had mean methylation levels of 66.7% (alpha cells) and 67.0% (beta cells). Promoters (1,000 bp upstream of TSS) had mean methylation levels of 25.5% (alpha cells) and 25.4% (beta cells). 5’ UTRs had mean methylation levels of 13.5% (alpha cells) and 13.1% (beta cells). Exons had mean methylation levels of 62.4% (alpha cells) and 62.2% (beta cells). Introns had mean methylation levels of 79.8% (alpha cells) and 79.9% (beta cells). 3’ UTRs had mean methylation levels of 80.6% (alpha cells) and 80.0% (beta cells). Intergenic regions had mean methylation levels of 78.1% (alpha cells) and 80.2% (beta cells). e, The mean degree of DNA methylation of Shelves (81.9% in alpha and 82.6% in beta cells), Shores (62.2% in alpha and 62.4% in beta cells), CpG islands (15.8% in alpha and in beta cells) and Open sea (82.4% in alpha and 83.7% in beta cells). A CpG island is defined as a ≥200-bp-long stretch of DNA with a CG content of ≥50% and an observed CpG/expected CpG in excess of 0.6. The Shores are the 2-kb flanking regions of CpG islands, the Shelves are the 2-kb regions outside island shores, and the Open Sea is everything else. f,g, Average DNA methylation levels, using the WGBS data, in different gene regions of genes divided according to expression level (no, low, medium and high expression) in alpha (f; n = 7 donors) and beta cells (g; n = 11 donors), respectively. The WGBS data were analysed by dmrseq25, and the Friedman rank-sum test was used to test differences in the average DNA methylation levels between the groups. Data are presented as the mean ± s.e.m.; for f, ***P1-5kb = 0.000105, Ppromoters = 0.000105, P5’ UTR = 0.000105, Pexons = 0.000105, Pintrons = 0.000172, P3’ UTR = 0.000105; for g, ***P1–5 kb = 0.000000835, Ppromoters = 0.000000322, P5’ UTR = 0.000000322, Pexons = 0.000000835, Pintrons = 0.000000322, P3’ UTR = 0.000000322, as analysed by two-sided Friedman rank-sum tests, not corrected for multiple testing. The vertical lines in c–e represent the mean DNA methylation. Schematic in a created in BioRender; Ofori, J. https://biorender.com/h6zumch (2026).
