Gestational high‐fat diet impaired demethylation of Ppar α and induced obesity of offspring

Abstract Gestational and postpartum high‐fat diets (HFDs) have been implicated as causes of obesity in offspring in later life. The present study aimed to investigate the effects of gestational and/or postpartum HFD on obesity in offspring. We established a mouse model of HFD exposure that included gestation, lactation and post‐weaning periods. We found that gestation was the most sensitive period, as the administration of a HFD impaired lipid metabolism, especially fatty acid oxidation in both foetal and adult mice, and caused obesity in offspring. Mechanistically, the DNA hypermethylation level of the nuclear receptor, peroxisome proliferator‐activated receptor‐α (Pparα), and the decreased mRNA levels of ten‐eleven translocation 1 (Tet1) and/or ten‐eleven translocation 2 (Tet2) were detected in the livers of foetal and adult offspring from mothers given a HFD during gestation, which was also associated with low Pparα expression in hepatic cells. We speculated that the hypermethylation of Pparα resulted from the decreased Tet1/2 expression in mothers given a HFD during gestation, thereby causing lipid metabolism disorders and obesity. In conclusion, this study demonstrates that a HFD during gestation exerts long‐term effects on the health of offspring via the DNA demethylation of Pparα, thereby highlighting the importance of the gestational period in regulating epigenetic mechanisms involved in metabolism.


| INTRODUC TI ON
Obesity is a major health issue worldwide, with excessive fat intake contributing to disease progression. Increasing evidence shows that obesity increases the risk of developing type 2 diabetes, nonalcoholic fatty liver disease and cardiovascular disease, thereby reducing the quality of life and life expectancy of patients. 1,2 However, it remains unknown whether there is a critical period during development from the gamete to the adult, and whether there is a relationship between high-fat diet (HFD)-induced obesity and gender. Nonetheless, maternal pre-pregnancy and gestational obesity have adverse health outcomes in offspring, and this positive relationship has been confirmed in several studies. 3,4 The effects of a HFD given to mothers during lactation on metabolism in offspring are still controversial. 5,6 Presently, a wealth of food choices has resulted in bad eating habits, suggesting that changing environments may lead to adverse health outcomes in adults, which is a major driver of the obesity epidemic. However, it is unknown which period is most critical on the metabolism of offspring.
The pathogenesis of obesity not only includes various genetic factors, but also social and obesogenic environmental factors. 5,7 One study has investigated the epigenetic signature related to obesity in later life. 8 However, there is little information on the specific mechanisms of metabolic programming, including when and how epigenetic changes occur. DNA methylation is an important epigenetic hallmark, and DNA methylation patterns of genes are often altered by deleterious environmental factors.
Furthermore, DNA methylation levels are regulated by DNAmodifying enzymes. The ten-eleven translocation (TET) family is comprised of important DNA-modifying enzymes with roles in the epigenetic regulation of genes. 9 TETs promote DNA demethylation and increase gene expression. DNA demethylation occurs in foetal and adult livers and associates with the fatty acid β-oxidation genes. 10 However, the major challenge is whether we can gauge the critical timing and the effects of HFD that can contribute to obesity.
In the present study, we established a mouse model to identify the most sensitive stage of obesity development in response to HFD. Furthermore, we confirmed its possible mechanism of action in obese offspring.

| Animal care
All animal protocols were approved by the Zhejiang University Animal Care and Use Committee. All mice were purchased from Beijing Weitonglihua Laboratory Animal (Beijing, China). The mouse HFD model was established as shown in Figure S1. Virgin C57BL/6J females (age, 6 weeks; weight, 12 g) were purchased and provided either a HFD (#D12492' Research Diets, containing 20% protein, 60% fat) or a normal chow diet (NCD) (#D12450B; Research Diets, containing 20% protein, 10% fat) for 8 weeks before mating. Mice were provided either a NCD or a HFD during pre-pregnancy and pregnancy. C57BL/6J male mice (age, 10 weeks) were used for mating. After birth, each group was divided into two groups: NCD and HFD groups based on the food provided to the mothers. After weaning, offspring in each group (four groups in total) was further divided into two groups: NCD and HFD groups based on the food provided to the offspring. This resulted in eight groups of offspring of two sexes, and there is only one pup from the same litter in each group ( Figure S1).

| Serum biochemical measurements
Blood specimens were collected from 16-week-old mice after overnight fasting. The serum levels of fasting triglyceride (TG), total cholesterol (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL) and non-esterified fatty acids (NEFA) were measured using a biochemical analyser (TBA120FR; Toshiba, Tokyo, Japan). The serum level of fasting insulin was quantified using a mouse insulin enzyme-linked immunosorbent assay (Crystal Chem, Downers Grove, IL, USA). The fasting glucose level was measured using a biochemical analyser (TBA120FR; Toshiba). After euthanization, livers and free gonadal adipose tissues were harvested and measured to calculate the liver-to-weight ratio and gonadal fat-toweight ratio.

| Tissue collection, histology and analysis of gene expression
Livers were harvested after collecting blood. For lipid accumulation analysis, left liver lobes were fixed, embedded in paraffin and sectioned for staining with Oil Red O. Total RNA from livers was extracted using TRIzol reagent (Life Technologies, Grand Island, NY, USA). cDNA was synthesized using oligo-dT and random primers (Takara, Tokyo, Japan). The levels of target genes were measured using quantitative real-time polymerase chain reaction (PCR) (LightCycler 480 II; Roche, Switzerland) with SYBR green detection (Takara). The level of each target gene was normalized to the β-actin level. All primer (Sangon Biotech, Shanghai, China) sequences are shown in Table S1.  Table S2. . All data are presented as the mean ± standard error of the mean. Unpaired Student's t tests were used to analyse statistical significance between two independent groups. One-way analysis of variance followed by the least significance difference post hoc test was used to analyse statistical significance among three or more groups. All reported P values are two-sided. P < .05 was considered statistically significant.

F I G U R E 2 Oil
Red O-stained liver cross-sections showing fat accumulation in F1 offspring mice at 16 weeks of age. A, administration of a NCD after weaning and C, a HFD after weaning in male mice; B, administration of a NCD after weaning and D, a HFD after weaning in female mice

| Administration of a HFD during gestation affects lipid metabolism and causes obesity in offspring
We analysed the offspring that were given a HFD during gestation, during weaning and after weaning. Irrespective of whether the offspring were given a NCD or HFD after weaning, the offspring that were given a HFD during gestation had a significantly

| Administration of HFD altered serum TG, LDL, NEFA, fasting glucose and insulin levels in offspring
Given that the important hallmarks of obesity are metabolic disorders and insulin resistance, we measured serum TG, TC, LDL, HDL, NEFA, glucose and insulin levels in offspring. Compared with the control group (N-N-N), offspring exposed to a HFD during gestation or lactation showed increased TG, LDL, NEFA, fasting glucose and insulin levels ( These results indicate that mice exposed to a HFD during early life can develop lipid and glucose metabolism disorders that may increase the risk of obesity in later life. These results suggest that gestational HFD exposure alters lipid metabolism-related genes in the liver, and gestation might be a sensitive period of HFD exposure.

| Administration of a HFD impaired fatty acid oxidation earlier than lipogenesis at E18.5 day
We performed qPCR analysis to examine the levels of the genes associated with lipid metabolism in the foetal liver at E18.5 day. The genes related to lipid metabolism included those involved in lipid transport, cholesterol metabolism, fatty acid transport and lipogenesis. Compared with the foetuses from mothers given a NCD TA B L E 2 Metabolic parameters in offspring mice at 16 weeks of age during gestation, Cpt-1a and Pparα mRNA levels were significantly decreased in the livers of both males ( Figure 4A[e, f], male) and females ( Figure 4B[e, f], female) at E18.5 day from the mothers given a HFD during gestation. These results are consistent with those in the livers of adult offspring ( Figure 3) and provide further evidence that gestational HFD exposure alters genes related to lipid metabolism in the liver and gestation is a sensitive period of HFD exposure.

| Administration of a HFD reprogrammes DNA methylation patterns by decreasing Tets expression instead of increasing Dnmts expression at E18.5 day
Based on our finding of decreased Pparα expression in mice at E18.5 day and 16 weeks of age, we further examined the DNA methylation pattern of Pparα and the relative mRNA levels of DNA-modifying enzymes. We found increased methylation in CpG sites of the Pparα promoter in the livers of foetuses at E18.5 day from mothers given a HFD during gestation compared to mothers given a NCD during gestation ( Figure 5A [male], B [female]). The mean methylation level of the Pparα promoter was significantly higher in offspring of the gestational HFD group than that in offspring of the NCD group ( Figure 5E [male], F [female]). These hypermethylation patterns may have been caused by the decreased expression levels of Tet1 and Tet2, because we found that the expression levels Tet1 and Tet2 were significantly lower in the liver of foetuses at E18.5 day from mothers given a HFD during gestation than those given a NCD ( Figure 5C

| Administration of HFD induces hypermethylation of Pparα by decreasing Tets expression in offspring at 16 weeks of age
We examined the DNA methylation patterns of Pparα in male and female offspring at 16 weeks of age. Similar to the findings in foetuses, we found increased methylation in CpG sites of the Pparα promoter in the livers of offspring at 16 weeks of age from mothers given a HFD during gestation than that from mothers given a NCD ( Figure 6A [male], B [female]). The mean methylation level of the Pparα promoter was significantly higher in offspring of the gestational HFD group than those of the NCD group ( Figure 6E [male], F [female]). We also detected a lower expression level of Tet2 in the livers of male offspring ( Figure 6C [male]) and lower expression levels of Tet1 and Tet2 in the livers of female offspring ( Figure 6D [female]) at 16 weeks of age from mothers given a HFD during gestation than those given a NCD, whereas there were no significant differences in the expression levels of Dnmt1 and Dnmt3a between these two groups ( Figure 6C [male], D [female]). These results provide further evidence that a HFD given during gestation induces hypermethylation of Pparα by decreasing the expression of Tet1 and Tet2.

| D ISCUSS I ON
'Mismatch'nutritional changes and/or a sedentary lifestyle can increase the risk of obesity. 11 Previous studies have shown that a maternal HFD-or self-HFD-induced obesity leads to different metabolism disorders. [12][13][14] Here, we provide evidence to support two fundamental principles of programming: sensitivity and duration.
We found that gestation was the most sensitive period to induce obesity in late life, and there was no difference between the sexes The administration of a HFD during lactation may induce obesity in mice, consistent with the results of previous studies. 18,19 Sexual dimorphism is a common research area in clinical trials and basic medical research. 20 In this study, we did not observe sexual dimorphism in lipid metabolism and obesity. One possible reason is that diet-induced obesity in mice varies by age of onset, 21 and oestrogen in adult females may have caused a difference between the sexes.
Another possibility is that there are significant differences in adipose tissue distribution between males and females, 22 was not only observed in livers of offspring but also in livers of foetuses from mothers given a HFD during gestation. PPARα is a key regulator of fatty acid oxidation in mice. 27 Hepatocyte-specific PPARα deficiency can lead to hepatic lipid accumulation. 28 PPARα can also stimulate the transcription of fatty acid β-oxidation relative genes. 29 In the present study, the administration of a HFD during pregnancy led to decreased expression of Pparα in foetal livers, possibly by inhibiting the demethylation of Pparα-dependent fatty acid oxidation-related genes and ultimately inhibiting lipid catabolism and inducing obesity.
There is increasing evidence that environmental stress can affect gene transcription, and gene transcription rates can be regulated through DNA methylation levels. Several studies have hypothesized that DNA methylation is fully established by the time of birth. [30][31][32] However, many fatty acid β-oxidation-related genes undergo DNA demethylation with increased mRNA expression in the postnatal mouse liver. 33 DNA methylation is a gene transcription silencing mechanism. In this study, we speculated that the relationship between the reduced expression of Tets and the increased methylation level of Pparα may have started during gestation, thereby leading to lipid metabolism disorders and obesity in offspring later in life.
Furthermore, it has been reported that DNA methylation patterns of metabolism-related genes in the liver change dynamically in early life, thereby activating hepatic metabolic processes to adapt to the nutritional environment. [34][35][36] The administration of a HFD during gestation down-regulated the levels of Tet1 and Tet2 in offspring at E18.5 days, whereas the Dnmts level was not altered. TET2 is necessary for cell development, 37 and postnatal demethylation in the liver was mediated by TET2. 38  and PPARα mediates DNA demethylation of fatty acid β-oxidation genes. 34 In the present study, a gestational HFD reduced Tet1 and Tet2 expression, thereby resulting in hypermethylation of hepatic Pparα and decreased PPARα expression. Finally, reduced PPARα expression increased the risk of obesity in offspring from mothers given a HFD.
In summary, we found that a HFD given during gestation alters F I G U R E 6 Methylation patterns of Pparα and the relative expression of DNA-modifying enzymes in livers of mice at 16 weeks of age. DNA methylation levels of the Pparα promoter in livers of A, male and B, female mice given a normal chow diet after weaning (male, n = 5 mice per group; female, n = 6 mice per group), relative expression of DNA-modifying enzymes in livers of C, male and D, female mice, E, mean DNA methylation in male mice; F, mean DNA methylation in female mice (male, n = 5 mice per group; female, n = 6 mice per group). Data are presented as the mean ± SEM, significance determined by ANOVA. *P < .05, **P < .01