Exacerbated obesogenic response in female mice exposed to early life stress is linked to fat depot-specific upregulation of leptin protein expression

Abstract

Early life stress (ELS) is an independent risk factor for increased BMI and cardiometabolic disease risk later in life. We have previously shown that a mouse model of ELS, maternal separation and early weaning (MSEW), exacerbates high-fat diet (HF)-induced obesity only in adult female mice. Therefore, the aim of this study was to investigate 1) whether the short- and long-term effects of HF on leptin expression are influenced by MSEW in a sex-specific manner and 2) the potential epigenetic mechanisms underlying the MSEW-induced changes in leptin expression. After 1 wk of HF, both MSEW male and female mice displayed increased fat mass compared with controls (P < 0.05). However, only MSEW female mice showed elevated leptin mRNA expression in gonadal white adipose tissue (gWAT; P < 0.05). After 12 wk of HF, fat mass remained increased only in female mice (P < 0.05). Moreover, plasma leptin and both leptin mRNA and protein expression in gWAT were augmented in MSEW female mice compered to controls (P < 0.05), but not in MSEW male mice. This association was not present in subcutaneous WAT. Furthermore, among 16 CpG sites in the leptin promoter, we identified three hypomethylated sites in tissue from HF-fed MSEW female mice compared with controls (3, 15, and 16, P < 0.05). These hypomethylated sites showed greater binding of key adipogenic factors such as PPARγ (P < 0.05). Taken together, our study reveals that MSEW superimposed to HF increases leptin protein expression in a sex- and fat depot-specific fashion. Our data suggest that the mechanism by which MSEW increases leptin expression could be epigenetic.

INTRODUCTION

The worldwide prevalence of obesity has nearly tripled in the last four decades, becoming a serious public health concern that has reached epidemic proportions (66). Therefore, it is necessary to uncover risk factors contributing to the uncontrolled development of obesity and related comorbidities. Early life stress (ELS), or adverse childhood experiences, is any form of abuse, neglect, and/or household dysfunction experienced during the critical developmental years of childhood that leads to negative health outcomes in adulthood (44, 57). There is a positive correlation between the cumulative number of adverse childhood experiences and the risk of developing chronic disease in adult life (2, 10, 16, 60). Epidemiological studies have linked ELS with several neuroendocrine, immune, and metabolic abnormalities; in addition, an increasing number of studies are showing that women exposed to ELS may be at a higher risk of developing obesity compared with men (3, 37, 38, 67).

To study the impacts of ELS on obesity, maternal separation and early weaning (MSEW) is a well-established mouse model of neglect that effectively mirrors a wide range of physiological effects of adverse childhood experiences in humans in a mouse (20, 55). Previously, our laboratory has shown that MSEW promotes fat expansion in a sex-specific manner (42). Specifically, when fed a high-fat (HF) diet, both male and female control mice displayed increased fat mass; however, only females exposed to MSEW showed exacerbated fat mass compared with non-stressed mice (42). Interestingly, this elevation in fat mass was due to greater visceral adipose tissue deposition, a “bad” fat associated with increased cardiometabolic risk (4, 12). Importantly, this increase in fat deposition was not linked to differences in food intake despite the existence of hyperleptinemia (42).

Body weight is a process that is tightly regulated by several genes associated with metabolic function. The ob gene transcript leptin is a hormone released from adipocytes into the bloodstream that affects both anabolic and catabolic pathways (34). Epidemiological studies have linked ELS with increased risk of morbidity and mortality as well as identified it as an independent risk factor for obesity, increased waist circumference, metabolic syndrome, and cardiovascular disease associated with elevated levels of plasma leptin (5, 13, 26, 48, 52, 61).

Despite a large amount of evidence suggesting that ELS may exert sex-specific effects on fat deposition and metabolic disease, the mechanisms underlying increases in women’s visceral adiposity are not fully understood. Therefore, the aim of this study was to investigate whether the short- and long-term effects of HF on visceral fat-derived leptin expression are influenced by MSEW in a sex-specific manner. Hence, we determined body composition and metabolic status in control and MSEW mice from both sexes exposed to HF for either 1 or 12 wk. In addition, analysis of plasma leptin and adipose tissue mRNA and protein levels was completed in both subcutaneous and visceral adipose tissue depots. Furthermore, because both ELS and obesity are lifestyle risk factors associated with epigenetic modifications (11, 20, 23, 40, 46), we determined the effects of MSEW superimposed to HF on the site-specific methylation patterns of the leptin promoter expressed in visceral fat.

METHODS

Animal model.

All animal protocols were approved by the Institutional Animal Care and Use Committee at the University of Kentucky. Maternal separation and early weaning (MSEW) was performed using C57BL/6 breeders given ad libitum access to food (2918 Teklad irradiated Global 18% Protein Rodent Diet) and water (Lexington, KY, city tap water treated by reverse osmosis). Animal rooms were maintained at 21 ± 2°C and kept on a 14:10 light-dark cycle. MSEW litter pups were separated from the dam for 4 h/day from postnatal days (PD) 2 to 5 and for 8 h a day from PD 6 to 16. Pups were weaned early on PD 17. Separation occurred at the same time of day, during which the mother was removed from the cage and pups were placed into a clean cage in an incubator (30 ± 1°C, 60% humidity). Normally reared litters served as controls and remained undisturbed with the dam until weaning on PD 21. To standardize maternal care and nutrition, litter size was culled to six to eight pups. After weaning, mice were housed three per cage, divided by sex and treatment. When placed on special diet, ideally, one male and one female from each litter were randomly assigned to either a low-fat (LF) or HF diet for either 1 or 12 wk. For example, in a typical control or MSEW litter of eight, which included four male mice and four female mice, each mouse would be randomly assigned to the following groups: 1 wk LF, 1 wk HF, 12 wk LF, and 12 wk HF for each sex. For body weight and adiposity, biological replicates were averaged to represent n = 1. Thus, for these measurements, the n reflects the total number of litters in each group. Then, mice were randomized for representing the litter in different experiments.

Following, mice were placed on a LF diet (D12450J, 10% Kcal from fat; Research Diets, New Brunswick, NJ) or a HF diet (D12492, 60% kcal from fat; Research Diets).

Experimental design.

After weaning, mice were fed a standard chow diet until 8 wk of age.

To determine early changes in body composition and adipose tissue leptin secretion in response to HF diet, mice were placed on either a LF or HF diet for 1 wk. To establish the trajectory of these changes, a subset of littermates was kept on either a LF or HF diet for a total of 12 wk.

Body weight (BW) and body adiposity (fat mass in %BW) were measured at weeks 8, 9, and 12 by Echo Magnetic resonance imaging (Echo-MRI; Echo Medical System, Houston, TX). An oral glucose tolerance test (OGTT) was performed at 11 wk of special feeding. Then, mice were euthanized using a ketamine-xylazine cocktail (100/10 mg/kg ip) for exsanguination via cardiac puncture and tissue harvesting. Subcutaneous (sc) white adipose tissue (WAT) and gonadal WAT (gWAT; epidydimal fat in males and periovaric fat in females) were collected and snap-frozen in liquid nitrogen for mRNA expression analysis. Plasma was snap-frozen for later adipokine analysis. In a subset of female mice fed a LF or HF diet for 16 wk, in accordance with similar sex-specific differences in adiposity previously reported (42), gWAT was collected for differential methylation analysis of the leptin promoter expressed in adipose tissue to investigate the epigenetic effects of MSEW.

Oral glucose tolerance test.

At 11 wk of either LF or HF diet, mice were fasted for 8 h to perform an oral glucose tolerance test (OGTT), as previously reported (42). After an oral gavage of d-glucose (2 g/kg body wt), blood glucose measurements were taken via tail prick at 0, 15, 30, 60, and 120 min using the second drop of blood on the glucometer (Accu-check; Roche). Area under the curve was calculated using GraphPad Prism (GraphPad Software, San Diego, CA).

Plasma analysis.

Five-hundred microliters of whole blood was collected using 7.5% EDTA-coated tubes and centrifuged, and then plasma supernatant was stored at −80. Leptin analysis was measured by ELISA (Cayman Chemical, Ann Arbor, MI), following the manufacturer’s protocol.

Leptin expression in adipose tissue.

Total RNA was extracted, from gWAT in male and female mice, using Purzol Reagent (Bio-Rad, Hercules, CA) and the Aurum Total RNA mini kit (Bio-Rad). NanoDrop 2000/2000c spectrophotometer (Applied Biosystems, Waltham, MA) was used to determine RNA quality and quantity. One-hundred nanograms of RNA was used to perform NanoString analysis on isolated RNA to probe for genes and transcription factors in gWAT (Nanostring Technologies, Seattle, WA).

Adipose tissue explants protocol.

Under anesthesia, mice were perfused (heparinized NaCl, 0.9%), and adipose tissue was collected. Then, adipose tissue explants [gWAT and subcutaneous (sc) WAT, 40 mg] were incubated (1 h at 37°C) in DMEM (200 μl) containing free fatty acid-BSA (1%). Leptin content in incubation media were measured by ELISA, following the manufacturer’s instructions (Millipore, Burlington, MA).

Bisulfite conversion and pyrosequencing.

To analyze the DNA methylation of the candidate genes in gWAT frozen tissue from male and female mice, freshly isolated genomic DNA was digested with HindIII, and 1 μg of digested DNA was used in bisulfite conversion according to the instructions with the kit (EZ DNA Methylation-Lightning Kit; Zymo Research). Converted DNA was eluted in 12 μl, and 2 μl was used in PCR reactions. Primers were designed using Zymo Research Bisulfite Primer Seeker around the CpG sites identified as differentially methylated by Methyl-Mini Seq (http://www.zymoresearch.com/tools/bisulfite-primer-seeker). PCR conditions for Bisulfite using ZymoTaq PreMix were as follows: 1) 95°C for 10 min, 2) 95°C for 30 s, 3) 50–58°C for 40 s, 4) 72°C for 45 s, 5) repeat steps 2–4 for 40 cycles, and 6) 72°C for 7 min. After amplification, PCR products were separated on a 1.5% agarose gel, and appropriate bands were cut and purified. Purified PCR product was then blunt-end modified (End-It DNA End-Repair Kit; Epicenter) and subcloned into pUC19 vector predigested with HincII. Ten to fifteen clones from each product were sequenced (University of Chicago Comprehensive Cancer Center DNA Sequencing and Genotyping Facility). Sequences were aligned with the genomic sequence using BiQ Analyzer. This program aligns the sequences, determines the converted cytosine residues, and calculates percent methylation at specific CpG sites in the sequence.

qRT-PCR.

RNA was isolated from 100 mg of gWAT or scWAT with RNeasy Mini-Kit (Qiagen 74134). One microgram of RNA was used in a reverse transcriptase reaction with iScript Reverse Transcriptase (Bio-Rad no. 1708891) to prepare cDNA. Twenty-five nanograms of the cDNA was used in the quantitative (q)RT-PCR reactions. The reaction protocol was as follows: 1) 94°C for 5 min, 2) 94°C for 30 s, 3) 56°C for 30 s, 4) 72°C for 45 s, 5) repeat steps 2–4 for 40 cycles, and 6) 72°C for 10 min.

Primers for the housekeeping gene (GAPDH) and primers used for genes with DMRs were created for this study using PrimerBank (Table 1).

Table 1.

Primers for Mouse RT-PCR in adipose tissue

Primer	Sequence	Primer No.
mmLep 115 F	GTGGCTTTGGTCCTATCTGTC	Primer Bank ID: 34328437c1
mmLep115 R	CGTGTGTGAAATGTCATTGATCC
mmGap95 F	AGGTCGGTGTGAACGGATTTG	Primer Bank ID: 126012538c1
mmGap95 R	GGGGTCGTTGATGGCAACA
mmDNMT179 F	CCGTGGCTACGAGGAGAAC	Primer Bank ID: 313661498c1
mmDNMt179 R	TTGGGTTTCCGTTTAGTGGGG
mmDNMT3B120 F	CTGTCCGAACCCGACATAGC	Primer Bank ID: 172088107c2
mmDNMT3B120 R	CCGGAAACTCCACAGGGTA
mmDMNT3A101 F	GATGAGCCTGAGTATGAGGATGG	Primer Bank ID: 114145580c1
mmDNMT3A101 R	CAAGACACAATTCGGCCTGG
mmMR286 F	GTGGACAGTCCTTTCACTACCG	Invitrogen: 296422D09
mmMR286 R	TGACACCCAGAAGCCTCATCTC	Invitrogen: 296422D10

F, forward; R, reverse.

Chromatin immunoprecipitation assay.

A chromatin immunoprecipitation (ChIP) assay (MagnaChIP, no. 17-20000, Millipore) was performed using specific antibodies for RNA Pol II (ab5096; Abcam), PPARγ-binding protein (mAb no. 2435; Cell Signaling Technology), and SP1 (ab13370). Briefly, adipose tissue samples were glass-glass homogenized, and cross-link chromatin was performed by adding formaldehyde (1% final concentration) for 10 min at 37°C. Pellet was washed with PBS and resuspended and incubated 5 min at 4°C in Cell lysis buffer. After 2,000 rpm of centrifugation, the supernatant was removed, and the procedure was repeated using Nuclear lysis buffer. Then, pellets were sonicated to a range of 200–500 bp and centrifuged at 13,000 rpm at 4°C for 10 min and ran on agarose gel to visualize the sample bp size. Chromatin precipitation was accomplished by adding 2-μl antibodies RNA Pol II (ab5096; Abcam), PPARγ-binding protein (mAb no. 2435; Cell Signaling Technology, Danvers, MA), SP1 (ab13370), and GAPDH (Sigma G9545; Sigma, St. Louis, MO) and 20 µl of magnetic beads to 50 µl of sonicated supernatant 10-fold in ChIP dilution buffer and incubating overnight at 4°C. Then, the beads and supernatant were removed, and the pellet was washed and eluted. DNA was recovered using spin columns and was purified and used for real-time PCR analysis using primers flanking the site 3 (forward = TACCGCTGCTCAGTAGCTG, reverse = CGCTTGGTATGAGCCGGTA) and site 15 (forward = AGTTGGCGCTCGCAGG, reverse = GCCTGCCCCTCTTATAACTGC). The sample inputs were normalized using GAPDH (IDT DNA, Coravilla, IA). PCR products were run on a 2% agarose gel with Texas Red dye. Images were captured on a Syngene PXi UV Transilluminator (Syngene International). The sequence for this gene is available through National Center for Biotechnology Information (NCBI) at the National Institutes of Health (NIH) website under the reference sequences NC_000072.5, covering nucleotides (nts) 29009221–29010220.

Statistical analysis.

All data are expressed as means ± SE. Unpaired Student’s t test was used between groups during different time points. Evaluations between groups at the end points were made by two-way ANOVA, followed by a Tukey’s post hoc test, using a Benjamini-Hochberg correction to control the false positive rate. P < 0.05 was considered statistically significant. Statistical analyses were conducted using GraphPad Prism 7 (GraphPad Software, Inc., La Jolla, CA).

RESULTS

Short-term effect of MSEW on body composition.

At 8 wk of age, the average BW and fat mass were similar in MSEW and control male mice fed a regular chow (24.2 ± 0.6 vs. 24.5 ± 0.3 g and 6.8 ± 0.9 vs. 7.1 ± 0.4% BW, respectively). One week after the switch to special diets, there was an increase in BW due to MSEW (F_1,63 = 4.474, P < 0.05) and HF (F_1,63 = 25.77, P < 0.05) (Table 2). In addition, MSEW male mice fed a HF showed increased total body adiposity when compared with control counterparts (F interaction_1,63 = 7.093, P < 0.05; Table 2). Similarly, the average BW and fat mass were not different between MSEW and control female mice fed a regular chow at 8 wk of age (19.1 ± 0.2 vs. 19.1 ± 0.2 g and 9.0 ± 0.4 vs.10.0 ± 0.6% BW, respectively). After 1 wk of HF, both MSEW and control female mice displayed increased BW (Table 2); however, total body adiposity was elevated by diet (F_1,58 = 31.26, P < 0.05) and MSEW (F_1,58 = 11.81, P < 0.05) (Table 2).

Table 2.

Effects of MSEW on body composition and metabolic function after 1 wk or 12 wk of either a LF or HF in control and MSEW male and female mice

Metabolic Parameters	Control LF	MSEW LF	Control HF	MSEW HF	P (Int)	P (MSEW)	P (Diet)
	1 Week
BW, g
Males	24.4 ± 0.4 (n = 15)	25.0 ± 0.4 (n = 14)	26.4 ± 0.6 (n = 13)	27.8 ± 0.5 (n = 12)	0.400	<0.05	<0.0001
Females	18.8 ± 0.2 (n = 11)	19.6 ± 0.3 (n = 14)	21.2 ± 0.5 (n = 12)	21.7 ± 0.6 (n = 14)	0.722	0.127	<0.0001
Fat mass (%BW)
Males	7.3 ± 0.6 (n = 15)	8.1 ± 0.4 (n = 14)	11.0 ± 1.3 (n = 13)#	15.8 ± 0.7 (n = 12)*#	<0.05	<0.05	<0.0001
Females	8.8 ± 0.6 (n = 11)	12.8 ± 0.9 (n = 14)	14.6 ± 0.8 (n = 12)	16.5 ± 1.1 (n = 14)	0.228	<0.05	<0.0001
	12 Weeks
BW, g
Males	28.5 ± 0.9 (n = 8)	30.9 ± 1.6 (n = 9)	41.2 ± 1.7 (n = 12)	43.1 ± 1.7 (n = 13)	0.880	0.198	<0.0001
Females	20.7 ± 0.3 (n = 11)	21.5 ± 0.3 (n = 12)	29.2 ± 1.6 (n = 11)	35.0 ± 1.8 (n = 12)	0.064	<0.05	<0.0001
Fat mass, %BW
Males	13.1 ± 1.6 (n = 8)	16.4 ± 2.6 (n = 9)	26.0 ± 3.7 (n = 12)	32.6 ± 2.5 (n = 13)	0.534	0.235	<0.0001
Females	9.0 ± 0.5 (n = 11)	10.85 ± 1.2 (n = 12)	30.3 ± 1.6 (n = 11)#	39.5 ± 2.5 (n = 12)*#	<0.05	<0.05	<0.0001
Food intake, kcal/day
Males	10.7 ± 1.5 (n = 6)	9.6 ± 1.1 (n = 8)	15.1 ± 2.0 (n = 9)	13.5 ± 2.0 (n = 9)	0.890	0.461	<0.05
Females	10.7 ± 1.1 (n = 9)	11.1 ± 1.1 (n = 5)	10.9 ± 1.0 (n = 10)	11.5 ± 1.6 (n = 7)	0.937	0.694	0.813
OGTT (AUC)
Males	10,896 ± 1,692 (n = 5)	9,279 ± 1,161 (n = 6)	14,127 ± 1,211 (n = 9)	13,700 ± 1,934 (n = 8)	0.734	0.5611	<0.05
Females	8,614 ± 1,344 (n = 7)	10,084 ± 1,024 (n = 6)	15,334 ± 2,037 (n = 8)	18,009 ± 1,512 (n = 6)	0.735	0.2511	<0.05

Values are means ± SE. AUC, area under the curve; BW, body weight; Int, intervention; MSEW, maternal separation and early weaning; HF, high-fat; LF, low-fat; OGTT, oral glucose tolerance test. Statistically significant comparisons are in boldface.

^*P (MSEW) < 0.05 vs. control;

^#P (diet) < 0.05 vs. LF.

Short-term effect of MSEW on plasma and tissue leptin expression.

One-week of HF diet induced increases in plasma leptin in male (F_1,25 = 9.617, P < 0.05; Fig. 1A) and female (F_1,21 = 10.01, P < 0.05; Fig. 1B) mice. The ratio of plasma leptin/adiposity was increased by diet in males (F_1,23 = 4.302, P < 0.05; Fig. 1C) but not in females (Fig. 1D). Leptin mRNA expression in gWAT was increased by diet (F_1,21 = 22.15, P < 0.05) and MSEW in males (F_1,21 = 7.547, P < 0.05) (Fig. 2A); however, MSEW female mice showed further increases in leptin mRNA expression compared with controls (F interaction_1,19 = 12.10, P < 0.05; Fig. 2B).

Fig. 1.

Short-term effect of maternal separation and early weaning (MSEW) on plasma leptin. Plasma leptin (A and B) and plasma leptin normalized to body adiposity (C and D) in male and female control or MSEW mice fed either a low-fat (LF) or high-fat (HF) diet for 1 wk. Control LF males, n = 8; MSEW LF males, n = 8; control HF males, n = 5; MSEW HF males, n = 8; control LF females, n = 5; MSEW LF females, n = 7; control HF females, n = 6; MSEW HF females, n = 7. BW, body weight.

Fig. 2.

Short-term effect on tissue leptin expression in maternal separation and early weaning (MSEW). Leptin mRNA expression in gonadal white adipose tissue (gWAT) in male (A) and female (B) control or MSEW mice fed either a low-fat (LF) or high-fat (HF) diet for 1 wk. *P < 0.05 vs. control; #P < 0.05 vs. LF. Control LF males, n = 4; MSEW LF males, n = 9; control HF males, n = 5; MSEW HF males, n = 7; control LF females, n = 6; MSEW LF females, n = 6; control HF females, n = 5; MSEW HF females, n = 6.

In MSEW female mice fed a HF diet, leptin receptor A mRNA expression was increased in gWAT when compared with controls (2.6 ± 0.2 vs. 1.0 ± 0.1-fold, P < 0.05); however, both receptors were expressed similarly in scWAT. Male mice fed a HF did not show any significant difference in Leptin receptor A between groups in either fat depot.

Long-term effect of MSEW on body composition.

MSEW and control male mice fed a HF diet for 12 wk displayed similar increases in total body adiposity (F_1,39 = 30.11, P < 0.05), food intake (F_1,28 = 5.277, P < 0.05), and glucose tolerance (F_1,24 = 4.873, P < 0.05) (Table 2). However, MSEW female mice showed an exacerbated increase in adiposity compared with control counterparts (F interaction_1,37 = 5.569, P < 0.05; Table 2). These increases in fat mass displayed by MSEW female mice were present despite no changes in food intake or glucose tolerance compared with control mice (Table 2).

Long-term effect of MSEW on plasma and tissue leptin expression.

Twelve weeks of HF diet increased plasma leptin similarly in MSEW and control male mice (F_1,34 = 4.166, P < 0.05; Fig. 3A). Contrarily, HF-fed females showed exacerbated plasma leptin by MSEW (F interaction_1,24 = 7.0, P < 0.05; Fig. 3B). Plasma leptin normalized to body adiposity was increased in HF-fed male groups (F_1,28 = 4.747, P < 0.05; Fig. 3C); however, this ratio was increased in MSEW female mice compared with controls (F interaction_1,21 = 6.022, P < 0.05; Fig. 3D). In males, whereas leptin mRNA expression in gWAT was not changed as a result of diet or MSEW (Fig. 4A), leptin protein measured in gWAT media explants increased similarly in MSEW and control HF-fed mice (F_1,18 = 7.802, P < 0.05; Fig. 4C). In females, HF increased leptin mRNA expression in gWAT, whereas MSEW further increased this expression (F interaction_1,21 = 4.349, P < 0.05; Fig. 4B). Also, leptin measured in gWAT media explants in females was increased only in MSEW mice fed a HF (F interaction_1,24 = 6.916, P < 0.05, Fig. 4D).

Fig. 3.

Long-term effect of maternal separation and early weaning (MSEW) on plasma leptin. Plasma leptin (A and B) and plasma leptin normalized to body adiposity (C and D) in male and female control or MSEW mice fed either a low-fat (LF) or high-fat (HF) diet for 12 wk. *P < 0.05 vs. control; #P < 0.05 vs. LF. Control LF males, n = 7; MSEW LF males, n = 10; control HF males, n = 11; MSEW HF males, n = 12; control LF females, n = 7; MSEW LF females, n = 8; control HF females, n = 7; MSEW HF females, n = 6.

Fig. 4.

Long-term effect of maternal separation and early weaning (MSEW) on gonadal white adipose tissue (gWAT) leptinexpression. Leptin mRNA expression (A and B) and leptin protein release from media (C and D) from gWAT in male and female control or MSEW mice fed either a low-fat (LF) or high-fat (HF) diet for 1 wk. *P < 0.05 vs control; #P < 0.05 vs. LF. Control LF males, n = 4; MSEW LF males, n = 5–9; control HF males, n = 6; MSEW HF males, n = 7; control LF females, n = 6–7; MSEW LF females, n = 5–7; control HF females, n = 7–8; MSEW HF females, n = 6–7.

Overall, leptin expression in scWAT was increased by HF diet. In males, despite leptin mRNA expression being increased by both diet (F_1,23 = 22.35, P < 0.05) and MSEW (F_1,23 = 6.775, P < 0.05; Fig. 5A), leptin measured in media explants showed a similar HF-induced increase in both groups (F_1,21 = 36.71.57, P < 0.0001; Fig. 5C). In females, HF-fed MSEW and control mice displayed similar increases in both leptin mRNA expression in scWAT (F_1,31 = 50.59, P < 0.05; Fig. 5B) and scWAT media explants (F_1,26 = 9.084, P < 0.05; Fig. 5D).

Fig. 5.

Long-term effect of maternal separation and early weaning (MSEW) on subcutaneous white adipose tissue (scWAT) leptin expression. Leptin mRNA expression (A and B) and leptin protein release from media (C and D) from scWAT in male and female control or MSEW mice fed either a low-fat (LF) or high fat (HF) diet for 1 wk. Control LF males, n = 4–7; MSEW LF males, n = 6–7; control HF males, n = 7–8; MSEW HF males, n = 7–8; control LF females, n = 6–10; MSEW LF females, n = 7–8; control HF females, n = 8–10; MSEW HF females, n = 8.

MSEW-induced site-specific changes in methylation of leptin promoter in gWAT.

Because of the combination of increased circulating leptin, leptin/adiposity, and leptin mRNA and protein levels in gWAT from MSEW female mice fed a HF diet, we sought to investigate the epigenetics of the leptin promoter. First, we validated the increases in gWAT leptin expression in a subset of samples from female mice fed a HF diet (Fig. 6A). Indeed, leptin mRNA expression (F = 365577, DFn = 2, Dfd = 2, P < 0.05), along with the de novo DNA methyltransferase (DNMTs), DNMT3A (F = 14317, DFn = 2, Dfd = 2, P < 0.05) and DNMT3B (F = 3747, DFn = 2, Dfd = 2, P < 0.05), were increased in gWAT from MSEW mice compared with controls. In these same samples, we determined 16 CpG sites within the leptin promoter (Fig. 6B). We found that overall, gWAT from HF-fed MSEW mice showed increased methylation across the promoter when compared with controls; however, three sites (CpG 3, 15 and 16) displayed hypomethylation (P < 0.05, Fig. 6B). Furthermore, we validated the site-specific changes in methylation in gWAT from male mice. Unlike females, control male mice fed a HF diet showed site 3 completely unmethylated, whereas MSEW male mice displayed this site ∼30% methylated (0 vs. 0.33 ± 0.15%5mC, respectively). Likewise, site 15 showed similar levels of methylation in both groups of mice fed a HF (0.75 ± 0.14 vs. 0.89 ± 0.25%5mC, respectively).

Fig. 6.

Maternal separation and early weaning (MSEW)-induced site-specific changes in methylation of leptin promoter in gonadal white adipose tissue (gWAT). The methylation landscape of the leptin promotor (A and B) in female control or MSEW mice fed either a low-fat (LF) or high-fat (HF) diet for 16 wk. *P < 0.05 vs. control. Control LF females, n = 3; MSEW LF females, n = 3; control HF females, n = 3; MSEW HF females, n = 3. DNMT, DNA methyltransferase.

The sites we identified as hypomethylated in tissue from female mice are in regions of the promoter in which adipogenic transcription factors have the capability to bind and activate leptin expression (33, 56). Specifically, we investigated the occupancy on the these hypomethylated sites by the master regulator of the adipogenic differentiation process, PPARγ (Fig. 7). The occupancy of the transcription factor PPARγ was increased in MSEW female mice by ∼1.7-fold at site 3 (F = 16.65, DFn = 6, Dfd = 7, P < 0.05; Fig. 7B) and ∼2.8-fold at site 15 compared with control female mice (F = 37.25, DFn = 5, Dfd = 5, P < 0.05; Fig. 7B). Figure 7C shows the PCR products including the inputs for each sample.

Fig. 7.

Binding of peroxisome proliferator-activated receptor-γ (PPARγ) to the leptin promoter is increased at the hypomethylated CpG site 3 in high-fat (HF)-fed maternal separation and early weaning (MSEW) female mice. The relative degree of binding of PPARγ to CpG site 3 (A) and site 15 (B) and a representative image of PPARγ binding to site 3 of the leptin promoter as determined by chromatin immunoprecipitation (ChIP) assay (C) in female mice fed either a low-fat (LF) or HF diet for 16 wk. *P < 0.05 vs. control (C);n = 6/group.

DISCUSSION

Our study shows for first time that MSEW, a mouse model of ELS, exacerbates body adiposity and leptin gene expression in a sex- and fat depot-specific manner. Specifically, MSEW male mice displayed increased fat mass and leptin mRNA expression in response to a short-term HF diet; however, this effect weaned over time. Contrarily, MSEW female mice displayed increased fat mass, plasma leptin, and leptin mRNA expression after 1 wk of HF diet, an effect that remained significant after 12 wk. Interestingly, MSEW female mice, and not males, exhibited a specific increase in gWAT-derived leptin expression and protein release. Furthermore, along with the increased leptin mRNA and protein expression, female MSEW mice fed a HF diet displayed hypomethylation of CpG sites 3, 15, and 16 of the leptin promoter specifically expressed in gWAT. Thus, the increases in gWAT leptin expression could be secondary to the hypomethylation of the leptin promoter in regions that are known to be important for the binding of adipogenic transcription factors such as PPARγ. Therefore, the hypomethylation of specific sites, combined with the increased occupancy of PPARγ, could be a potential mechanism by which MSEW induces hyperleptinemia and promotes visceral fat deposition. Taken together, our data reveal a sex- and depot-specific upregulation of leptin expression by MSEW in response to HF.

ELS, or adverse childhood experiences, is described as abuse, neglect, or loss during childhood that results in negative health outcomes long after the stressor has ended (6, 7). It is well established that exposure to ELS results in an increased risk for obesity, cardiovascular disease, type 2 diabetes, and early mortality (6, 26, 51, 59). Clinical and experimental evidence support the notion that ELS enhances the risk for chronic disease by directly altering the sensitivity of the HPA axis, neuroendocrine system, energetic metabolism regulation, and the epigenome (25, 28). Although numerous studies suggest that ELS increases the risk of obesity development, the sex-specific differences underlying this relationship are not well understood (53, 64). Thus, our study provides new insights regarding the mechanisms by which stress increases adiposity in a sex-specific fashion.

The negative implications of ELS are well documented to be mediated via the actions of glucocorticoids (1, 49). Glucocorticoids, also referred to as the stress hormones, have been shown to exert control over adipose tissue homeostasis via the glucocorticoid (GC) and mineralocorticoid (MC) receptors (31). In general, glucocorticoids are implicated in the regulation of adipogenesis, the differentiation process of preadipocytes into mature adipocytes that can fill with lipids and have endocrine actions, henceforth resulting in hypertrophy (47). Utilizing mouse models, adipose tissue-specific overexpression of 11β-hydroxysteroid dehydrogenase 1 (11bHSD1) results in visceral adipose tissue expansion (36). During obesity, glucocorticoids are the more biologically relevant ligand of the GR and MC receptors present in adipose tissue since they typically circulate at a higher concentration than aldosterone at the tissue level. Moreover, it is well known that MR inhibition impairs adipogenesis, and the degree of this dysfunction is higher than that of GR inhibition (8, 9). In support of these findings, experimental studies performing adrenalectomy and 11bHSD1 inhibition in rodents have been shown to prevent the development of obesity, resulting in a lean phenotype (17, 27, 29). Thus, glucocorticoids play an important role controlling adipocyte differentiation via both the GC and MR receptors (9, 24, 32).

In previous studies conducted in a rat model of ELS, we have shown that the treatment with a glucocorticoid synthase inhibitor during postnatal life can impair female rats to show adipose tissue expansion in response to HF (43). Therefore, these data indicate a glucocorticoid-dependent mechanism linking ELS and fat expansion in female rodents. Later, using the mouse paradigm, we have shown that female control and MSEW mice fed a HF diet display a similar increase in the levels of circulating corticosterone and aldosterone compared with LF-fed mice (14). However, at the tissue level, we found differences in 11bHSD1, GC, and MC receptor mRNA expression in fat and liver (14). These data suggest that the local metabolism of glucocorticoids is most likely altered by MSEW. Specifically, in gonadal fat, MSEW females showed unchanged GC receptor expression but increased MC receptor expression compared with control counterparts. Similarly, MSEW female mice fed a HF diet for 12 wk in the current study showed an approximately threefold MR receptor upregulation in gWAT compared with controls (3.04 ± 0.5 vs. 1.25 ± 0.3, respectively, P < 0.05) but not in scWAT. Yet MSEW male mice did not show MR expression differences in either fat depot (data not shown). Overall, the greater susceptibility of glucocorticoid-induced visceral adipose tissue expansion could also contribute to increased release of adipokines (22, 35). Therefore, increased stress-induced hypertrophy may stimulate the release of leptin in the differentiated adipocyte. However, whether exacerbated leptin expression is glucocorticoid dependent in our model of ELS remains under investigation.

There is a growing body of literature showing that women exposed to abuse/neglect show a greater association with increased BMI and waist circumference (18–23). Leptin displays a fat depot-specific release, with a majority of its production derived from scWAT (18, 21). Considering that scWAT is a larger fat depot compared with the gWAT, our study shows that control and MSEW males display similar increases in adiposity, circulating leptin, and both gWAT and scWAT-derived leptin production. Conversely, female MSEW mice had increased gWAT, circulating leptin, and gWAT-derived leptin production compared with control mice. Thus, the increased leptin levels we found in females are more likely associated with an enhanced stimulatory effect of HF diet on gWAT in mice that were exposed to MSEW. This finding is particularly relevant since visceral adiposity has been associated with increased cardiometabolic risk in women (14). Interestingly, a study by Schlitt and Schulz (54) showed that leptin expression from scWAT in a nonpregnancy state switches to primarily visceral adipose tissue expression during pregnancy. In agreement with these results, Kronfeld-Schor et al. (30) reported that pregnant mice show increased visceral leptin production via a corticosterone-dependent mechanism. Overall, these data show that sex hormones can mediate adaptations to different physiological and pathophysiological states (i.e., early life stress), and this may influence the source of leptin expressed in different fat depots.

To address the mechanism connecting MSEW with exacerbated adiposity in females fed a HF diet, we need to consider the effects of leptin in caloric intake and energy expenditure regulation. Female MSEW mice show increased circulating leptin without differences in food intake, as we have reported in previous a study (14) and the current study. Thus, whether MSEW reduces energy expenditure needs to be assessed by calorimetry to have a better understanding of the contribution of energetic metabolism on MSEW-induced obesity. In addition, it is important to tease out whether increased leptin expression is a cause or effect of MSEW on the mature adipocyte.

Individuals exposed to ELS often present changes in gene expression mediated by epigenetic modifications of DNA structure (11, 40). Numerous studies have focused on the brain methylation landscape associated with changes in behavior (15, 50, 63). These studies report that early-life stress-induced changes in methylation patterns of genes in the hippocampus result in a vulnerability to stress and behavioral abnormalities in adult life. Other studies looking specifically at differential methylation patterns of vasopressin and corticotropin-releasing factor, similar to this study, also reported hypomethylation at CpG sites of the promoter of these genes. This hypomethylation was associated with increased corticosterone secretion and hype responsiveness to stress (39–41). Ultimately, this hypomethylation resulted in a decreased ability to silence target genes and contributed to ELS-induced dysfunction. Our data provide evidence that MSEW influences the methylation of genes expressed in the periphery that are involved in metabolic function and body weight balance. We focused on epigenetic modifications of the leptin promoter, such as DNA methylation at the carbon-5 positions of cytosines in CpG dinucleotides by DNA methyltransferase. As aforementioned, these changes in methylation have been shown to control transcription at particular loci by modulating access of transcription factors and machinery to gene promoters (7, 19, 58). Furthermore, Shen et al. (56) detailed the epigenetic modifications of the leptin promoter during diet-induced obesity, showing hypermethylation of the leptin promoter in obese male mice along with increased DNA methyltransferase (DNMT) binding and decreased RNA PolII binding. These authors concluded that hypermethylation needs to be a possible feedback mechanism to inhibit further leptin synthesis (56). Because this work only described the CpG sites of the leptin promoter in male mice, it remained unclear how diet-induced obesity affects the methylation landscape in female mice when fed a HF. Thus, our study is the first to look at the sex-specific changes in leptin promoter methylation. Whereas DNA methylation is typically associated with gene silencing because it is thought to prevent transcription machinery from assembling at the transcription start site, Shen et al. (56) reported an inverse relationship between DNA methylation of the leptin promoter and leptin expression. Although there were several sites of the leptin promoter hypermethylated by MSEW, we focused our study on the hypomethylated CpG sites 3, 15, and 16 because these regions of the promoter have been shown to bind pro-adipogenic transcription factors, including PPARγ, KL4, CEBPα, and Sp1 (56, 62). Nevertheless, there remain gaps in the literature regarding the full transcriptional mechanism of activation for leptin. However, in line with our data, it has been shown that PPARγ has a specific leptin-binding site and that adipose-specific deletion of PPARγ reduces leptin expression (65, 68). Additionally, male MSEW mice display different leptin promoter methylation landscape, which suggests that postnatal stress may prime the epigenetic regulation of leptin expression in a sex-specific manner. Future studies will address how site-specific hypermethylation in female MSEW mice fed a HF diet may contribute to the activity of the leptin promoter.

In conclusion, our study reports a depot- and sex-specific effect of MSEW on exacerbated body weight gain, adiposity, and leptin expression. Whereas both male and female MSEW mice fed a short-term HF diet display increases in adiposity, only female mice showed long-lasting effects. Thus, the current study suggests that MSEW may determine the increased risk for obesity and fat deposition in female mice via an epigenetic mechanism, leading to the activation of the leptin promoter in visceral adipose tissue. Together, our data suggest that ELS in humans may result in persistent changes in the adipose tissue epigenome, possibly contributing to the development of obesity and obesity-induced hypertension. Future studies will be focused on determining the effect of MSEW on molecular mechanisms associated with preadipocyte differentiation and mature adipocyte lipogenesis to understand the origin of exacerbated obesogenic response in females.

GRANTS

This study was supported by funds from the National Heart, Lung, and Blood Institute (R00-HL-111354 and R01-HL-135158 to A.S.L., R01-HL-135158-S1 to J.R.L.) and the pilot project from the University of Kentucky Center of Research in Obesity and Cardiovascular Disease COBRE P20 GM103527 to A.S.L.

DISCLOUSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.R.L. and A.S.L. conceived and designed research; J.R.L., M.R., D.M.C., X.X., and A.S.L. performed experiments; J.R.L., M.R., D.M.C., Y.F., and A.S.L. analyzed data; J.R.L., M.R., D.M.C., Y.F., and A.S.L. interpreted results of experiments; J.R.L. and A.S.L. prepared figures; J.R.L. and A.S.L. drafted manuscript; J.R.L., M.R., D.M.C., and A.S.L. edited and revised manuscript; J.R.L., M.R., D.M.C., Y.F., and A.S.L. approved final version of manuscript.