Ligand-Activated PPARα-Dependent DNA Demethylation Regulates the Fatty Acid β-Oxidation Genes in the Postnatal Liver

Birth is a principal change in nutritional environment, when the major source of energy switches from glucose in the cord blood to lipids in breast milk. The metabolic function of the liver changes sequentially during early life in mammals, so that it becomes matured as a major metabolic organ, thereby adapting to the marked changes in nutritional environment. After birth, metabolic pathways such as gluconeogenesis, fatty acid β-oxidation, and de novo lipogenesis are known to be upregulated in the liver (1–3), whereas hepatic extramedullary hematopoiesis is downregulated (4). Indeed, hepatic fatty acid β-oxidation is activated progressively during the neonatal period to produce energy from breast milk lipids (2,3). It should be critically involved in the regulation of metabolic homeostasis; deficiency of certain fatty acid β-oxidation enzyme(s) results in sudden infant death due to fasting intolerance (5). However, how a particular metabolic pathway is activated during the liver maturation is poorly understood.

The nuclear receptor-type transcription factor peroxisome proliferator–activated receptor α (PPARα) is considered as a key regulator of lipid metabolism in the liver (6,7). Mice with targeted disruption of PPARα develop high-fat diet– and fasting-induced fatty liver (8). PPARα, when activated by various lipid ligands, can stimulate transcription of its target genes in the fatty acid β-oxidation pathway (6,9). During the suckling period, it may be activated in response to breast milk–derived lipid ligand(s) (10). It is, therefore, conceivable that PPARα serves as a master transcriptional regulator of the hepatic fatty acid β-oxidation pathway in early life.

The methylation of cytosine residues in DNA is a major heritable and reversible epigenetic modification. DNA methylation of the promoter region is a well-studied repressive modification for gene expression (11). It is essential for embryonic development; DNA methylation contributes to a variety of critical biological processes such as genome imprinting and X-chromosome inactivation (12–14). However, the functional role of DNA methylation after organogenesis has not been fully addressed except for carcinogenesis. Recent studies showed that gene- and life-stage–specific changes in DNA methylation occur in certain metabolic genes during the fetal period and even after birth (15,16). We have reported that expression of the lipogenic glycerol-3-phosphate acyltransferase 1 gene (Gpam) is induced in the postnatal mouse period via a DNA demethylation mechanism (16). On the other hand, genome-wide analysis revealed marked changes in DNA methylation in numerous genes in the fetal and neonatal livers in both the mouse and human (17,18). However, the detailed molecular mechanism and physiologic implication of the gene- and life-stage–specific changes in DNA methylation has not been fully addressed.

Here we show that the ligand-activated PPARα-dependent DNA demethylation occurs in the promoter regions of its target genes, which are mapped onto the fatty acid β-oxidation pathway in the postnatal mouse liver. We also demonstrate that DNA demethylation occurs specifically in the fatty acid β-oxidation genes from the fetal to adult human liver. The data of this study highlight the role of a lipid-sensing nuclear receptor in the gene- and life-stage–specific DNA demethylation of a particular metabolic pathway.

All animal experiments were approved by the Tokyo Medical and Dental University Committee on Animal Research (0090041). Eight-week-old male and female C57BL6 (Japan SLC Inc., Hamamatsu, Japan) mice were crossbred, and the offspring were used for analysis. We obtained the liver samples at 18.5 days after fertilization (e18.5, before birth), 2 days after birth (D2, just after birth), 16 days after birth (D16, preweaning), and 28 days after birth (D28, postweaning).

Since it was reported that differentiated hepatocytes expressing specific metabolic enzymes are observed on embryonic day 15.5 (e15.5) (19), maternal administration of PPARα ligand Wy14643 (Wy) was performed at 40 mg/kg/day from gestation day 15 to 18 and from day 3 to 16 of the lactation period by intraperitoneal injection. The PPARα-deficient (KO) mice (19) (strain B6.129S4-Ppara^tm1Gonz/J) were purchased from The Jackson Laboratory (Bar Harbor, ME). Homozygous male and female PPARα-KO mice were crossed to obtain the PPARα-KO offspring. The age-matched C57BL6 mice were used as wild-type (WT) control. The detailed protocol of Wy-administration experiments is illustrated in Supplementary Fig. 1.

The mouse liver genomic DNA was extracted by the standard proteinase K method. Human fetal (male 18 and 22 weeks and female 18 weeks after fertilization) and adult (male 20, 44, and 59 years old) liver genomic DNAs were purchased from BioChain Institute Inc.‎ (Newark, CA). The Microarray-Based Integrated Analysis of Methylation by Isoschizomers (MIAMI) analysis, a genome-wide analysis of DNA methylation using a gene promoter array and methylation-sensitive restriction enzyme, was performed as described (20–22). Detailed experimental procedure is described in the Supplementary Data.

RNA was isolated from the livers of sex-matched mice at e18.5, D2, D16, and D28 (five samples for each group are combined). Human fetal (male 20 weeks and female 38 weeks after fertilization) and adult (male 26 and 44 years old) liver RNAs were purchased from BioChain Institute Inc.‎ They were labeled with a cyanine 3-CTP using a Low Input Quick Amp Labeling Kit (Agilent Technologies Inc., Santa Clara, CA) and hybridized to the Agilent whole mouse genome 4 × 44 K microarray or Human Gene Expression 4 × 44 K version 2 MicroArray. Signal detection was performed according to manufacturer’s manual, and fluorescence signals obtained were normalized with the global normalization method. Signals >1.5-fold and <1.5-fold were considered as increase and decrease, respectively. To describe the correlation between changes in DNA methylation and gene expression, scatter plot graphs were depicted with the log-transformed methylation differences from MIAMI data and the log-transformed fold changes from cDNA microarray data.

We used DAVID version 6.7 (http://david.abcc.ncifcrf.gov/) (23) for the gene ontology (GO) analysis. The functional annotation chart and clustering of DAVID were applied to group similar GO terms. Benjamini-Hochberg multiple testing corrections were performed, and the corrected P values were used for judging the candidate genes (P < 0.05 were considered significant). For the GO pie chart (Fig. 1C), we counted the total number of genes annotated to each GO cluster.

We used the MATCH software (24) (BIOBASE GmbH, Worfenbuettel, Germany) to explore transcriptional binding motifs in the promoter regions (from −1,000 to +100 of each gene, taking the first nucleotide of exon 1 as +1) of the genes of interest.

For pathway analysis, we used DAVID 6.7. Pathways for the genes undergoing DNA methylation change were enriched by the corrected P values from Benjamini-Hochberg multiple testing corrections (P < 0.05 were considered significant). R heatmap.2 (function of gplot package) was used to create a heat map from the pathway analysis results (Fig. 4C). The logarithms of the P values to base 10 were converted into the heat map on blue-red scale.

Quantitative RT-PCR analysis was performed according to the previous report (16) using the following primers: Acox1 forward GCCTTTGTTGTCCCTATCCGT, reverse CGATATCCCCAACAGTGATGC; Cpt1a forward CCTGCATTCCTTCCCATTTG, reverse TGCCCATGTCCTTGTAATGTG; Ehhadh forward CCAATGCAAAGGCTCGTGTT, reverse GGTAGAAGCTGCGTTCCTCTTG.

Bisulfite DNA methylation analysis was performed according to the previous report (16) using the following specific primers: Acox1 forward AATTGTGGGAGAGGGTGGGTTA, reverse AAACCATATCTCCAACCCCCTTATAT; Cpt1a forward GGATATTTTTATTTTTTGGTGGGAATAG, reverse ACTCCATAATCATTCTCTCTAACCTCCT; Ehhadh forward GGTTTAAGTTAAATTTGTAGTAGTTTGGTTGG; reverse AAAACAAAATCTAAATAAATCAACAATTCCT. We used a web-based quantification tool for the bisulfite sequencing analysis (http://quma.cdb.riken.jp/) (25).

Western blot analysis was performed as described previously (26). Anti-ACOX1 (sc-98499; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), anti-CPT1A (ab128568; Abcam, Inc., Cambridge, MA), and anti-β-actin (sc-81178; Santa Cruz Biotechnology, Inc.) were used as primary antibodies.

After extraction with 2:1 (vol/vol) chloroform/methanol, hepatic triacylglycerol (TG) contents were measured by the enzymatic colorimetric method using triglyceride E tests (Wako Pure Chemicals, Osaka, Japan) according to the previous report (16).

Hepatic acetyl-CoA levels were measured by Pico-Probe Acetyl-CoA Fluorometric Assay kit (BioVision, Milpitas, CA).

Serum TG, total cholesterol, nonesterified fatty acid (NEFA), and glucose concentrations were measured by Test Wako Kits (Wako). Serum insulin and total ketone body concentrations were measured by Mouse Insulin ELISA Kit (Morinaga Institute of Biological Science Inc., Yokohama, Japan) and EnzyChrom Ketone Body Assay Kit (BioAssay Systems, Hayward, CA), respectively.

Statistical analysis was performed using Student t test and ANOVA followed by Scheffé test. Data are expressed as the mean ± SEM. P < 0.05 was considered significant.

We analyzed the genome-wide methylation changes in the postnatal mouse liver, comparing e18.5 with D2, D2 with D16, and D16 with D28. By the MIAMI analysis, we found marked changes in DNA methylation; 1,721 genes that gained DNA methylation and 568 genes that lost DNA methylation from D2 to D16 (Fig. 1A).

To identify genes with an inverse correlation between DNA methylation and mRNA expression, we combined the DNA methylation status of the genes with their mRNA expression. Figure 1B shows the relationship between the DNA methylation and mRNA expression levels. From D2 to D16, 433 genes gained DNA methylation with decreased mRNA expression (25.2% of 1,721 genes) and 249 genes lost DNA methylation with increased mRNA expression (43.8% of 568 genes). Correlation analysis revealed that negative correlation between methylation and expression is only observed in the genes that lost DNA methylation from D2 to D16 (R = −0.11; P = 0.013) but not in those that gained DNA methylation from D2 to D16 (R = 0.0036; P = 0.89) or those that lost DNA methylation from D16 to D28 (R = 0.0039; P = 0.97). The number of genes that underwent DNA methylation change are summarized in Supplementary Table 1.

We next performed GO analysis. From D2 to D16, 10 clusters were produced for genes that gained DNA methylation with decreased mRNA expression and those that lost DNA methylation with increased mRNA expression (Supplementary Table 2). The GO analysis data suggest that the genes with marked changes in DNA methylation and mRNA expression are related to metabolic and hematocyte functions (Supplementary Table 2). In this study, the genes that lost DNA methylation with increased mRNA expression from D16 to D28 did not produce any significant GO terms.

We illustrated the existence ratio of the genes classified with hematocyte and/or metabolic functions in those with marked difference in DNA methylation between D2 and D16 (Fig. 1C). The genes with metabolic function were predominant among those that lost DNA methylation with increased mRNA expression from D2 to D16 [indicated as (c) in Fig. 1C, pie graph on the right], suggesting that some metabolic processes are activated in parallel with DNA demethylation in the postnatal mouse liver. By contrast, many genes with increased DNA methylation and decreased mRNA expression were related to hematocyte function [indicated as (a) in Fig. 1C, pie graph on the left], which is consistent with reduced hepatic extramedullary hematopoiesis upon birth (4).

We investigated transcriptional factor binding motifs in the promoter regions of the genes shown in Fig. 1C. Eight motifs were significantly enriched in the genes that lost DNA methylation from D2 to D16 (Supplementary Table 3). Several potential binding motifs of nuclear receptor PPARs (V$PPARG, V$PPAR_DR1, V$PPARA) were highlighted. By contrast, analysis of the genes that gained DNA methylation revealed no particular binding motifs. These observations suggest that the genes that lost DNA methylation with increased mRNA expression are potential targets of PPARs.

We next conducted a pathway analysis using the gene set shown in Fig. 1C. Several pathways were enriched, with the “PPAR signaling pathway” and “fatty acid metabolism” being significant (Supplementary Table 4). Because among PPARs PPARα is considered as a key regulator of lipid metabolism in the liver (6,9), we hereafter focused on PPARα. We found that 9 out of 249 genes that lost DNA methylation from D2 to D16 are mapped onto the fatty acid β-oxidation pathway (Fig. 1D).

As the MIAMI analysis evaluates DNA methylation of the HpaII digestion site, we performed the detailed bisulfite sequencing analysis of representative β-oxidation genes, Acox1, Ehhadh, and Cpt1a, at several CpG sites, including the HpaII sites. In all cases, the HpaII sites and neighboring CpG sites showed a concordant decrease in DNA methylation (Fig. 1E). We also examined mRNA expression and protein levels from D2 to D16 (Fig. 1F and Supplementary Fig. 2A and B). There were significant negative correlations between DNA methylation and mRNA expression levels in all the genes highlighted in Fig. 1D. PPARα mRNA expression was also increased from D2 to D16. Serum TG concentrations and hepatic TG contents were significantly decreased in D16 relative to D2 (Fig. 1G), which is consistent with the upregulation of the fatty acid β-oxidation in the liver. As previously reported (27), serum glucose and insulin concentrations were significantly increased and decreased in D16 relative to D2, respectively. Serum total cholesterol concentration was significantly decreased in D16 (Supplementary Fig. 2C).

We next performed genome-wide DNA methylation analysis of the livers from PPARα-KO mice. There was only a slight difference in DNA methylation between WT and PPARα-KO mice on D2 (25 genes with DNA hypermethylation and 14 genes with DNA hypomethylation in the KO liver relative to WT liver) (Fig. 2A, left). By contrast, there were more genes with DNA hypermethylation in PPARα-KO mice on D16 (110 genes with DNA hypermethylation and 72 genes with DNA hypomethylation) (Fig. 2A, right). The PPAR binding motif was enriched in the promoter regions of the genes with DNA hypermethylation (Supplementary Table 3), which are involved in fatty acid metabolism (Supplementary Table 4). Consistently, many of the genes with DNA hypermethylation were mapped onto the fatty acid β-oxidation pathway (Fig. 2B). Decreased DNA methylation of Acox1 and Ehhadh observed in the WT liver were partially but significantly attenuated in the KO liver from D2 to D16 (Fig. 2C). Consistently, mRNA expression of Acox1 and Ehhadh was not different between the WT and PPARα-KO mice on D2, whereas it was significantly decreased in PPARα-KO mice on D16 (Fig. 2D). In addition, serum TG, total cholesterol, and NEFA concentrations and hepatic TG contents were significantly increased in PPARα-KO mice relative to WT on D16 (Fig. 2E and Supplementary Fig. 3). There was no significant difference in serum glucose and insulin concentrations between WT and KO mice on D16 (Supplementary Fig. 3).

We found no significant difference in mRNA expression of Cpt1a between WT and KO mice on both D2 and D16 (Fig. 2D), suggesting that DNA methylation and mRNA expression of Cpt1a do not depend upon PPARα during the neonatal period. We therefore focused on Acox1 and Ehhadh for further analysis.

To explore if DNA demethylation of the fatty acid β-oxidation genes depends upon PPARα activation, we administered a PPARα ligand Wy14643 (Wy) to dams in the late gestation and lactation periods (Supplementary Fig. 1A). DNA demethylation and increased mRNA expression of Acox1 and Ehhadh was observed in the liver of the D16 offspring derived from the Wy-administered dams (Fig. 3A and B). In this study, PPARα mRNA expression was significantly increased (Supplementary Fig. 4A). We also examined parameters related to the fatty acid β-oxidation. (Fig. 3C). Although hepatic TG contents were unchanged, serum TG and NEFA concentrations were significantly decreased, whereas hepatic acetyl-CoA levels were significantly increased, in the Wy-administered offspring on D16 (Fig. 3C and Supplementary Fig. 4B). Moreover, serum total ketone body concentrations were increased, although not significantly, in the Wy-administered offspring.

Notably, the Wy-induced DNA demethylation and increased mRNA expression of Acox1 and Ehhadh also occurred, although slightly, in the liver of D2 offspring (Fig. 3D and E), which may be referred to as premature DNA demethylation. The Wy-induced DNA demethylation was also observed in other genes in the liver of D2 (32 genes with DNA hypomethylation and 19 genes with DNA hypermethylation) (Fig. 3F). Pathway analysis of the genes undergoing DNA demethylation identified the PPARα-related pathways (Supplementary Table 4).

We next examined DNA methylation and mRNA expression in the liver of both D2 and D16 offspring derived from the Wy-administered PPARα-KO dams (Supplementary Fig. 1B). The Wy-induced DNA demethylation and increased mRNA expression of Acox1 and Ehhadh (Fig. 3A, B, D, and E) were not observed in the Wy-administered PPARα-KO offspring (Fig. 3G and H). Consistent with this, hepatic acetyl-CoA levels and serum ketone body concentrations were not increased or serum TG concentrations were not decreased in the Wy-administered PPARα-KO offspring. (Supplementary Fig. 4C).

To investigate if the DNA methylation changes in the postnatal mouse liver (Fig. 1A–D) are conserved in humans, we performed a genome-wide analysis of DNA methylation and cDNA microarray analysis in the fetal and adult human liver samples. The MIAMI analysis identified numerous genes with marked changes in DNA methylation (529 genes with DNA hypomethylation and 299 genes with DNA hypermethylation) (Fig. 4A) from the fetus to adult human liver. Among those with DNA hypomethylation, 122 genes showed increased mRNA expression. Pathway analysis highlighted the PPARα-related pathways in the genes with decreased DNA methylation and increased mRNA expression (Supplementary Table 4 and Fig. 4B). A heat map generated with the pathway analysis data confirmed that the decreased DNA methylation with increased mRNA expression in the PPAR signaling and fatty acid metabolism pathways occur in both the mouse and human (Fig. 4C, compare D2 vs. D16 and fetus vs. adult). Interestingly, PPARα-dependent DNA methylation changes were only observed in the genes mapped onto the two pathways (Fig. 4C, compare D2 vs. D16, WT vs. KO, and vehicle vs. Wy).

Epigenetic modification of the genome such as DNA methylation may play an important role in the organ development and maturation during the fetal and neonatal periods. A previous study has suggested a role for DNA methylation in early postnatal liver development (18). Moreover, hepatic metabolic genes such as Tat (tyrosine amino transferase) and Gpam are upregulated in parallel with DNA demethylation before or even after birth (15,16). Here we demonstrate that many of the fatty acid β-oxidation genes undergo DNA demethylation with increased mRNA expression in the postnatal mouse liver. The DNA demethylation and increased mRNA expression are accompanied by increased protein expression. Notably, the DNA demethylation seems to be neonatal-stage–specific; it does not occur before birth under the physiologic condition. We also find that DNA methylation is reduced in the fatty acid β-oxidation genes in the adult human liver relative to the fetal liver, suggesting that it is functionally conserved across species. This study is the first demonstration of gene- and life-stage–specific DNA demethylation of a particular metabolic pathway. It is conceivable that DNA demethylation of the fatty acid β-oxidation genes leads to their coordinate induction in the neonatal liver, thereby representing an adaptative mechanism to the marked changes in nutritional environment after birth.

Because known factors involved in DNA demethylation possess no DNA sequence specificity, how the genes specific to a particular metabolic pathway are coordinately demethylated is an important issue to be addressed. In this study, analysis of the genes with DNA demethylation reveal PPAR binding motifs, suggesting that they are transcriptional targets of PPARs. Notably, DNA demethylation of the fatty acid β-oxidation genes during the neonatal period is suppressed in PPARα-KO mice. On the other hand, maternal administration of a PPARα ligand accelerates the DNA demethylation in the Wy-administered offspring. These observations suggest that ligand-activated PPARα triggers DNA demethylation of the fatty acid β-oxidation genes in the liver, thereby leading to increased mRNA expression during the neonatal period. In this regard, previous studies showed that transcription factors and their DNA target sequences are critical determinants of the DNA methylation state in vitro (28,29). There is also evidence that thymine DNA glycosylase, which is considered as a critical regulator in the multiple enzymatic reactions of DNA demethylation (30), binds to several nuclear receptors (31,32). It is, therefore, interesting to speculate that the ligand-activated PPARα recruits factors involved in DNA demethylation to the promoter regions of the fatty acid β-oxidation genes via the PPAR-responsive element, thus leading to the gene-specific DNA demethylation. Further studies are required to understand how PPARα triggers DNA demethylation in a gene-specific manner.

The detailed mechanism of life-stage–specific DNA demethylation of the fatty acid β-oxidation genes in the liver is poorly understood. In this study, we find that DNA demethylation of the fatty acid β-oxidation genes does not occur in the fetal liver, where PPARα is expressed abundantly. It is, therefore, unlikely that increased PPARα mRNA expression affects DNA demethylation and mRNA expression of PPARα target genes. On the other hand, maternal administration of Wy during the gestation period results in premature DNA demethylation in the liver of WT but not PPARα-KO offspring just after birth, indicating that DNA demethylation of the fatty acid β-oxidation genes in the liver is ligand-activated PPARα dependent. We speculate that the PPARα-dependent DNA demethylation does not occur in the fetal liver, where PPARα ligands are unavailable under the physiological condition. PPARα is known to be activated by lipid ligands (33), which, during the suckling period, are supplied by breast milk (10). This discussion is supported by our observation that there is no significant difference in the DNA methylation state of the fatty acid β-oxidation genes between PPARα-KO and WT mice on D2. These observations suggest that the neonatal-stage–specific DNA demethylation of the fatty acid β-oxidation genes is not a programmed process of liver maturation; it may be induced in response to the changes in nutritional environment, when PPARα lipid ligands are available. Further studies are required to identify the natural occurring PPARα ligands in breast milk lipids.

Transcriptional activation of the PPARα-target genes depends upon the amount and kinds of fatty acid ligands (33,34). In this study, maternal administration of Wy during the lactation period triggers DNA demethylation of the fatty acid β-oxidation genes, suggesting that DNA demethylation in the neonatal liver can be controlled by the PPARα ligands delivered to the pup via breast milk. Evidence has suggested that the fatty acid composition of breast milk is affected by maternal nutritional condition (35,36). It is reported that PPARα serves as an adaptive mechanism to nutritional changes such as an excess of lipid intake and fasting (8). In this study, analysis of serum and hepatic metabolic parameters revealed that hepatic lipid metabolism is impaired in PPARα-KO mice and that maternal administration of Wy accelerates the fatty acid β-oxidation of WT but not of the PPARα-KO offspring. It is conceivable that PPARα-dependent DNA demethylation of the fatty acid β-oxidation genes in the neonatal liver is an adaptive response to the changes in nutritional environment at the onset of lactation.

Given that the DNA methylation status established in early life remains relatively stable throughout life and thus affects hepatic lipid metabolism and susceptibility to metabolic diseases in later life (37), development of formula milk with the differing amount and kinds of PPARα ligands may offer therapeutic strategies in early life to prevent the development of metabolic diseases in adulthood.

This study is the first demonstration of gene- and life-stage–specific DNA demethylation on the fatty acid β-oxidation pathway, which is triggered by the ligand-activated PPARα during the liver maturation (Supplementary Fig. 5). Before birth, when glucose, a major source of energy during the fetal period, is provided via the cord blood, expression of all the fatty acid β-oxidation genes examined may be suppressed in a DNA methylation-dependent manner, which is partly because PPARα ligands are unavailable. After birth, activation of hepatic PPARα by milk lipid ligands may result in the activation of the fatty acid β-oxidation pathway via a DNA demethylation mechanism. It is, therefore, conceivable that during the neonatal period, milk lipids serve as a nutrient signal as well as nutrients so that they can be oxidized efficiently as an energy source.

Acknowledgments. The authors thank Dr. Kyoichiro Tsuchiya of Tokyo Medical and Dental University for critical reading of the manuscript.

Funding. This work was supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and the Ministry of Health, Labor, and Welfare of Japan. This work was also supported by research grants from Ono Medical Foundation, The Naito Foundation, and Nestlé Nutrition Council, Japan.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. T.E. and Y.K. designed and performed research, analyzed data, and wrote the manuscript. X.Y., M.T., S.K., E.T., K.T., and T.T. performed research. Y.N., H.S., and I.H. contributed new reagents/analytic tools. T.T.-I. contributed new reagents/analytic tools and analyzed data. T.S. and K.H. analyzed data. Y.O. designed research and wrote the manuscript. Y.O. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.