Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Feb 12;9(1):636.
doi: 10.1038/s41467-018-03038-w.

Epigenetic Modulation of Fgf21 in the Perinatal Mouse Liver Ameliorates Diet-Induced Obesity in Adulthood

Affiliations
Free PMC article

Epigenetic Modulation of Fgf21 in the Perinatal Mouse Liver Ameliorates Diet-Induced Obesity in Adulthood

Xunmei Yuan et al. Nat Commun. .
Free PMC article

Abstract

The nutritional environment to which animals are exposed in early life can lead to epigenetic changes in the genome that influence the risk of obesity in later life. Here, we demonstrate that the fibroblast growth factor-21 gene (Fgf21) is subject to peroxisome proliferator-activated receptor (PPAR) α-dependent DNA demethylation in the liver during the postnatal period. Reductions in Fgf21 methylation can be enhanced via pharmacologic activation of PPARα during the suckling period. We also reveal that the DNA methylation status of Fgf21, once established in early life, is relatively stable and persists into adulthood. Reduced DNA methylation is associated with enhanced induction of hepatic FGF21 expression after PPARα activation, which may partly explain the attenuation of diet-induced obesity in adulthood. We propose that Fgf21 methylation represents a form of epigenetic memory that persists into adulthood, and it may have a role in the developmental programming of obesity.

Conflict of interest statement

The authors declare no competing financial interests.

Figures

Fig. 1
Fig. 1
Genome-wide DNA methylation analysis of Wy-offspring. a Experimental protocol of genome-wide DNA methylation analysis in the liver of offspring derived from dams-administered Wy (Wy-offspring) or DMSO (Veh-offspring) during the late gestation (e14–e18) and lactation periods (D2–D16). b, c MIAMI analysis comparing vehicle (labeled with Cy3) with Wy (Cy5) at D16 (b) and 14W (c). Plots of log transformed values of HpaII (methylation-sensitive, horizontal axis) and MspI signal difference (methylation-insensitive, vertical axis) between samples. The values of HpaII signal difference/MspI signal difference judged as increased and decreased DNA methylation are above 1.4 and below 0.65, respectively. The regression line is in yellow and red lines are located log10 1.4 and log10 0.65 of the horizontal distance from the regression line. Orange and blue circles are judged as hypermethylated and hypomethylated, respectively. d A correlation plot showing the differences at 14W (y-axis) vs. D16 (x-axis). Genes, which are hypomethylated both at D16 and 14W are highlighted as blue circles. e Venn-diagram showing the result of MIAMI analysis. Four-hundred and twenty-four genes were DNA hypomethylated in Wy-offspring relative to that in Veh-offspring at D16 and 33 genes at 14W, respectively. Consequently, we identified 25 genes that were DNA hypomethylated in Wy-offspring relative to that in Veh-offspring both at D16 and 14W
Fig. 2
Fig. 2
DNA methylation analysis of Fgf21. a Schematic representation of the promoter region of Fgf21. Open circles and gray boxes indicate CpG sites and PPAR response elements (PPREs), respectively. Bisulfite-sequencing (BS) analysis region encompassing the transcription start site is indicated. b Bisulfite-sequencing analysis of Fgf21 in Wy- and DMSO (Veh)-treated offspring and PPARα-KO mice. Closed and open circles indicate methylated and unmethylated CpGs, respectively. Representative data of three independent experiments are shown. c Graphic presentation of statistical analysis of the bisulfite-sequencing data. The gray-shaded box indicates the period of maternal administration of Wy or Veh, (n = 3–6 at each time point). Statistics by one-way ANOVA with Tukey’s multiple comparison test. Data are expressed as mean ± SEM. *P < 0.05; **P < 0.01; ###P < 0.001 vs. Veh-offspring
Fig. 3
Fig. 3
Life stage-specific DNA demethylation of Fgf21. a Experimental protocol of maternal administration of Wy or DMSO (Veh) during the late gestation (e14–e18) and lactation periods (D2–D16) (left). Bisulfite-sequencing analysis of the offspring on D16 (right) (n = 3–4 for each group). Statistics by unpaired Student’s t-test. Data (% DNA methylation) are expressed as mean ± SEM. **P < 0.01; N.S.: not significant vs. Veh-offspring. b Experimental protocol (top) and bisulfite-sequencing analysis (bottom) in mice treated with Wy or Vehicle from 4W to 6W (Wy- and Veh-mice, respectively). Bisulfite-sequencing analysis of Fgf21 in Wy- and Veh-mice (bottom left). Representative data of three independent experiments are shown. Graphic presentation of statistical analysis of the bisulfite-sequencing data (bottom right). The gray-shaded box indicates the period of Wy or Veh administration. (n = 4–5 per group, statistics by one-way ANOVA with Tukey’s multiple comparison test). Data are expressed as mean ± SEM. **P < 0.01; N.S., not significant vs. Veh-mice
Fig. 4
Fig. 4
Histone modifications of Fgf21 in Wy-offspring. a Schematic representation of the promoter region of Fgf21. Open circles and gray boxes indicate CpG sites and PPAR response elements (PPREs), respectively. BS, bisulfite sequencing. b, c ChIP assays of histone marks in Wy- and Veh-offspring at D16 (b) and 14W (c), with the indicated antibodies. Primers amplifying the region of −106 to +21 bp were used for ChIP-qPCR analysis (n = 5–6 per group). d ChIP assays of the recruitment of PPARα to PPRE1. Primers amplifying the region of −997 to −923 bp were used for ChIP-qPCR analysis (n = 4–8 per group). Statistics by unpaired Student’s t-test. Data are expressed as mean ± SEM. **P < 0.01; ***P < 0.001; N.S., not significant vs. Veh-offspring
Fig. 5
Fig. 5
Enhanced recruitment of TET2 to Fgf21 promoter in Wy-offspring at D16. a Hepatic Tet1, Tet2, Tet3, Dnmt3a, and Dnmt3b mRNA expression in mice from D2 to 14W (n = 4–8 per group, statistics by one-way ANOVA with Tukey’s multiple comparison test). Data are expressed as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; N.S., not significant vs. D16. b ChIP assays at D16 with the indicated antibodies. Primers amplifying the region of −106 to +21 bp were used for ChIP-qPCR analysis (Fig. 4a) (n = 4–8 per group, statistics by unpaired Student’s t-test). Data are expressed as mean ± SEM. *P < 0.05; N.S., not significant vs. Veh-offspring. c ChIP assays at 6W for TET2 in mice treated with Wy (Wy-mice) or DMSO (Veh-mice) from 4W to 6W (Fig. 3b) (n = 4 per group, statistics by unpaired Student’s t-test). Data are expressed as mean ± SEM. N.S., not significant vs. Veh-mice
Fig. 6
Fig. 6
Correlation between DNA methylation status of Fgf21 and FGF21 expression. a, b Hepatic Fgf21 mRNA expression and serum FGF21 concentrations in Wy- and DMSO (Veh)-treated offspring from D2 to 14W. The gray-shaded box indicates the period of maternal administration of Wy or Veh (n = 8 per group). c Hepatic Fgf21 mRNA expression after a single Wy injection in Wy- and Veh-offspring at 14W (left, n = 5–8 per group). Correlation between Fgf21 mRNA expression and DNA methylation (right, n = 13). d Serum FGF21 concentrations after a single Wy injection in Wy- and Veh-offspring at 14W (left, n = 4–7 per group). Correlation between serum FGF21 concentrations and DNA methylation (right, n = 11). e Serum FGF21 concentrations during fasting and refeeding (left, n = 7–9 per group). Correlation between serum FGF21 concentrations after a 24-h fast and DNA methylation (right, n = 16). f Serum NEFA concentrations during fasting and refeeding (n = 7–9 per group). g, h ChIP assays for Pol II in Wy- and Veh-offspring at D16 (g) and with a single DMSO (Veh) or Wy injection at 14W (h). Primers amplifying the region of −106 to +21 bp were used for Pol II (Fig. 4a) (n = 7–8 per group). Statistics by unpaired Student’s t-test (a, b, ce left, f, g, h) or Spearman’s rank correlation coefficient (ce right). The r- and p-values are indicated on the graph. Data are expressed as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; N.S., not significant vs. Veh-offspring
Fig. 7
Fig. 7
Metabolic phenotypes of Wy-offspring during high-fat diet (HFD) feeding. a Experimental protocol of Veh-offspring and Wy-offspring fed HFD diet, which are referred to as Veh-HFD and Wy-HFD, respectively. The gray-shaded box indicates the period of maternal administration of Wy or Veh. b, c Body weight changes (b) and total food intake during HFD feeding (c) (n = 11 per group, statistics by two-way ANOVA with repeated measures). d Tissue weight of Wy- and DMSO (Veh)-treated HFD mice at 14W (n = 11 per group). e Hematoxylin and eosin (HE) staining (top, representative image of ten individuals per group) and quantification of adipocyte diameter (bottom) of eWAT. Histograms of adipocyte diameter (bottom left). Horizontal lines with bilateral squares indicate interquartile range (IQR). Arrows indicate the median values (numbers above the horizontal lines) of Veh-HFD and Wy-HFD. Statistical analysis (bottom right) of mean adipocyte diameters are shown. Scale bar = 100 µm (n = 10 per group). f Bisulfite-sequencing analysis (left, representative data of three independent experiments) and graphical presentation of statistical analysis (right, n = 4–5 per group) of Fgf21 in Wy-HFD and Veh-HFD at 4W and 14W. g Hepatic Fgf21 mRNA expression in Wy-HFD and Veh-HFD at 14W. (n = 10 per group). h Circadian variation of serum FGF21 concentrations. ZT, zeitgeber time (n = 11 per group). i Relative mRNA expression of Egr1, c-fos, Hsl, and Atgl in eWAT (n = 10 per group). Statistics by unpaired Student’s t-test otherwise indicated. Data are expressed as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; N.S., not significant vs. Veh-HFD
Fig. 8
Fig. 8
Analysis of FGF21-deficient (KO) mice. a Experimental protocol of maternal administration of Wy or DMSO (Veh) to FGF21-KO dams during the late gestation (e14–e18) and lactation periods (D2–D16). Offspring derived from Wy- or Veh-administered dams are referred to as Wy-FGF21-KO and Veh-FGF21-KO, respectively. These offspring were treated with HFD diet for 10 weeks in adulthood (from 4W to 14W). b Body weight changes during HFD feeding (n = 7–9 per group, statistics by two-way ANOVA with repeated measures). c Tissue weight of Wy- and Veh-FGF21-KO at 14W. d Serum FGF21 levels of Wy- and Veh-FGF21-KO at 14W. e Hematoxylin and eosin (HE) staining (top, representative image of 7–9 individuals per group) and quantification of adipocyte diameter (bottom) of eWAT. Histograms of adipocyte diameter (bottom left). Horizontal lines with bilateral squares indicate interquartile range (IQR). Arrows indicate the median values (numbers above the horizontal lines) of Veh-FGF21-KO and Wy-FGF21-KO. Statistical analysis (bottom right) of mean adipocyte diameters ± SEM are shown. Scale bar = 100 µm. f Relative mRNA expression of Egr1, c-fos, Hsl, and Atgl in eWAT, n = 7–9 per group. Statistics by unpaired Student’s t-test otherwise indicated. Data are expressed as mean ± SEM. N.S., not significant vs. Veh-FGF21-KO
Fig. 9
Fig. 9
Schematic representation of epigenetic memory of Fgf21. Ligand-activated PPARα induces DNA demethylation of Fgf21 in the postnatal mouse liver; DNA methylation status of Fgf21, once established in early life, persists into adulthood, as an epigenetic memory. In adulthood, DNA methylation status determines the magnitude of Fgf21 expression response to environmental cues. This may lead to the reduction of body weight and adipose tissue mass during HFD feeding in adulthood

Similar articles

See all similar articles

Cited by 12 articles

See all "Cited by" articles

References

    1. Hanson MA, Gluckman PD. Early developmental conditioning of later health and disease: physiology or pathophysiology? Physiol. Rev. 2014;94:1027–1076. doi: 10.1152/physrev.00029.2013. - DOI - PMC - PubMed
    1. Waterland RA. Epigenetic mechanisms affecting regulation of energy balance: many questions, few answers. Annu. Rev. Nutr. 2014;34:337–355. doi: 10.1146/annurev-nutr-071813-105315. - DOI - PubMed
    1. Lillycrop KA, Burdge GC. Epigenetic changes in early life and future risk of obesity. Int. J. Obes. (Lond.). 2011;35:72–83. doi: 10.1038/ijo.2010.122. - DOI - PubMed
    1. D’Urso A, Brickner JH. Mechanisms of epigenetic memory. Trends Genet. 2014;30:230–236. doi: 10.1016/j.tig.2014.04.004. - DOI - PMC - PubMed
    1. Waterland RA, Garza C. Potential mechanisms of metabolic imprinting that lead to chronic disease. Am. J. Clin. Nutr. 1999;69:179–197. - PubMed

Publication types

Feedback