Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 Sep;9(9):1195-206.
doi: 10.4161/epi.29856. Epub 2014 Jul 10.

Methylome, Transcriptome, and PPAR(γ) Cistrome Analyses Reveal Two Epigenetic Transitions in Fat Cells

Free PMC article

Methylome, Transcriptome, and PPAR(γ) Cistrome Analyses Reveal Two Epigenetic Transitions in Fat Cells

Hitomi Takada et al. Epigenetics. .
Free PMC article


Although DNA modification is adaptive to extrinsic demands, little is known about epigenetic alterations associated with adipose differentiation and reprogramming. We systematically characterized the global trends of our methylome and transcriptome data with reported PPAR(γ) cistrome data. Our analysis revealed that DNA methylation was altered between induced pluripotent stem cells (iPSCs) and adipose derived stem cells (ADSCs). Surprisingly, DNA methylation was not obviously changed in differentiation from ADSCs to mature fat cells (FatCs). This indicates that epigenetic predetermination of the adipogenic fate is almost established prior to substantial expression of the lineage. Furthermore, the majority of the PPAR(γ) cistrome corresponded to the pre-set methylation profile between ADSCs and FatCs. In contrast to the pre-set model, we found that a subset of PPAR(γ)-binding sites for late-expressing genes such as Adiponectin and Adiponectin receptor2 were differentially methylated independently of the early program. Thus, these analyses identify two types of epigenetic mechanisms that distinguish the pre-set cell fate and later stages of adipose differentiation.

Keywords: DNA methylation; adipose derived stem cells; epigenetics; fat cells; fat differentiation; induced pluripotent stem cells; reprogramming.


Figure 1. Relationship between promoter or gene-body methylation and gene expression in ADSCs. (A) Promoter CpG content in the human genome. Low- and high-CpG promoters were divided by the threshold of 0.03. (B) The FPKM distributions are shown for all genes, as well as those in the lowest 10% with respect to mC ratio, and those in the highest 10%. P values were calculated by the Wilcoxon rank sum test. (C) Gene-body CpG content in the human genome. (D) Analysis similar to that described for (B) was performed for the mC ratios calculated from gene bodies. Symbols *: P < 0.05; **: P < 10−3. Box: 25–75th percentile. The number of plotted genes is shown above each whisker. Genes with no expression in all the three cell types were excluded from the analysis.
Figure 2. Differential methylation between ADSCs, iPSCs, and FatCs. (A,B) Differential promoter methylation between ADSCs and iPSCs (A) and between ADSCs and FatCs (B). (C,D) Differential gene-body methylation between ADSCs and iPSCs (C) and between ADSCs and FatCs (D). (E, F) Differential expression between ADSCs and iPSCs (E) and between ADSCs and FatCs (F). The FPKM variation was evaluated by the Pearson correlation coefficient (PCC). Despite a lack of change in promoter and gene-body methylation, the differentiation to FatCs involves a large variation in expression, as does reprogramming to iPSCs.
Figure 3. Differential methylation accompanied by differential expression in ADSCs and iPSCs. The distributions of activated and repressed genes are presented as enrichment relative to the background distribution of all genes. (A) All genes are plotted. Hypomethylation and hypermethylation were correlated with activation and repression, respectively. (B) Low-CpG promoters are plotted. Low-CpG promoters are responsible for the correlation of hypomethylation with activation and hypermethylation with repression. (C) High-CpG promoters are plotted. High-CpG promoters do not contribute to the correlation of hypermethylation with repression.
Figure 4. Methylation of PPARγ binding sites. (A,C) Shown are mC values in the C/EBPα (A) and PPARγ (C) gene loci. Black bars; PPARγ binding regions based on ChIP-Seq data, red bars; differentially methylated regions between ADSCs and iPSCs, blue bars; differentially methylated regions between ADSCs and FatCs. (B, D) The expression level of C/EBPα (B) and PPARγ (D) in each cell type. (E-G) The histogram of mC ratio at 52040 PPARγ sites compared with those for random genomic loci in each cell type. (H-J) The histogram for 33402 ChIP signal sites of H3K4me3 (activation mark) in each cell type. (K-M) The histogram for 54130 ChIP signal sites of H3K27me3 (repression mark) in each cell type.
Figure 5. Differential methylation at PPARγ binding regions between ADSCs and FatCs. (A-C) Differential methylation of PPARγ binding sites (A), H3K4me3 (B), and H3K27me3 (C) between ADSCs and FatCs. Hypomethylated signal sites are over-represented specifically for PPARγ, but not for H3K4me3 or H3K27me3. The numbers of hypomethylated signal sites are 7826 out of 52040 for PPARγ, 971 out of 33402 for H3K4me3, and 1878 out of 54130 for H3K27me3. (D) Over-representation of differentially methylated signal sites in ADSCs and FatCs. Enrichment and P values were calculated by the binomial test. Over- and under-representation with P < 0.05 are colored in yellow and blue, respectively. (E) Enrichment of differentially methylated PPARγ binding sites among differentially expressed genes in ADSCs and FatCs. P values were calculated by the Fisher exact test. Enrichment and depletion (dis-enrichment) with P < 0.05 are colored in yellow and blue, respectively.
Figure 6. Epigenetic markers for adipogenesis. (A,C) mC values in the ADIPOQ (A) and ADIPOR2 (C) gene loci for each cell type. (B,D) The expression levels of ADIPOQ (B) and ADIPOR2 (D) in each cell type.

Similar articles

See all similar articles

Cited by 1 article


    1. Scherer PE. Adipose tissue: from lipid storage compartment to endocrine organ. Diabetes. 2006;55:1537–45. doi: 10.2337/db06-0263. - DOI - PubMed
    1. Townsend K, Tseng YH. Brown adipose tissue: Recent insights into development, metabolic function and therapeutic potential. Adipocyte. 2012;1:13–24. doi: 10.4161/adip.18951. - DOI - PMC - PubMed
    1. Nishimura S, Manabe I, Nagasaki M, Hosoya Y, Yamashita H, Fujita H, Ohsugi M, Tobe K, Kadowaki T, Nagai R, et al. Adipogenesis in obesity requires close interplay between differentiating adipocytes, stromal cells, and blood vessels. Diabetes. 2007;56:1517–26. doi: 10.2337/db06-1749. - DOI - PubMed
    1. Balistreri CR, Caruso C, Candore G. The role of adipose tissue and adipokines in obesity-related inflammatory diseases. Mediators Inflamm. 2010;2010:802078. doi: 10.1155/2010/802078. - DOI - PMC - PubMed
    1. Lefterova MI, Lazar MA. New developments in adipogenesis. Trends Endocrinol Metab. 2009;20:107–14. doi: 10.1016/j.tem.2008.11.005. - DOI - PubMed

Publication types

MeSH terms