When adult cells are reprogrammed into induced pluripotent stem cells (iPSCs), their DNA methylation landscape is extensively rewritten. Yet some iPSC lines retain somatic-like marks or acquire patterns that differ from embryonic stem cells. A new study published in Epigenetics & Chromatin has mapped these changes genome-wide in human gingival fibroblast-derived iPSCs, reporting broad convergence towards a pluripotent reference state alongside smaller sets of residual and culture-specific methylation patterns.
The research team compared three cell populations: human gingival fibroblasts (HGF), iPSCs derived from them (GP), and an H9 embryonic stem-cell reference line (H9CL). The analysis covered 693,271 CpG sites across nine culture-and-assay replicates, with three replicates per cell population.
Overall, the gingival fibroblast-derived iPSCs showed greater genome-wide methylation similarity to the embryonic stem-cell reference than to the fibroblasts they were derived from, indicating substantial epigenetic resetting during reprogramming.
The scale of the differences varied sharply between comparisons. Using a Benjamini–Hochberg false-discovery rate below 0.05 and an absolute methylation difference of at least 0.20, the researchers identified 113,520 differentially methylated positions between fibroblasts and iPSCs, 115,056 between fibroblasts and the embryonic stem-cell reference, but only 10,926 between the iPSCs and the reference.
To classify how reprogramming resolved each methylation difference, the team examined the positions that distinguished fibroblasts from the embryonic reference. Of these, 83.04 percent had shifted to the embryonic-like state in the iPSCs. A further 4.91 percent remained fibroblast-like, 7.27 percent showed intermediate methylation, and 4.78 percent deviated beyond both reference states, representing patterns specific to the derived iPSC cultures.
The authors also ran an independent raw-IDAT processing workflow as a check on their findings. Within that workflow, most differentially methylated positions remained significant after adjustment for slide and run: 99.72 percent, more than 99.99 percent, and 93.79 percent for the three respective comparisons. The researchers emphasise that these percentages are within-workflow comparisons and should not be read as replication rates for the primary DMP sets.
The study’s conclusions are deliberately framed around its own scope. The gingival fibroblast-derived iPSC cultures showed broad methylation convergence towards the embryonic stem-cell reference, together with smaller sets of fibroblast-like, intermediate, and iPSC-specific patterns.
Because the analysis used a single fibroblast-derived iPSC population compared against one reference line, the authors note that the findings should be pursued in additional donor-matched cultures and independently derived iPSC lines. The residual somatic-like and deviant methylation patterns documented here provide a concrete starting point for that next stage of investigation.
Subject of Research: Genome-wide DNA methylation patterns in human gingival fibroblast-derived induced pluripotent stem cells
Article Title: Genome-wide DNA methylation profiles of human gingival fibroblast-derived induced pluripotent stem cells show broad pluripotent-state similarity and residual somatic-like patterns
Article References: Li, J. (2026). Genome-wide DNA methylation profiles of human gingival fibroblast-derived induced pluripotent stem cells show broad pluripotent-state similarity and residual somatic-like patterns. Epigenetics & Chromatin. https://doi.org/10.1186/s13072-026-00709-7
Image Credits: AI Generated
DOI: 10.1186/s13072-026-00709-7
Keywords: DNA methylation, induced pluripotent stem cells, gingival fibroblasts, epigenetics, reprogramming, differentially methylated positions, embryonic stem cells, CpG methylation, pluripotency, Epigenetics & Chromatin, Genome-wide, methylation
News Source: Drew Townsend. (October 27, 2026). Gingival fibroblast-derived iPSCs show broad methylation convergence with residual somatic patterns. Scienmag.
