Reversal of dopamine system dysfunction in response to high-fat diet

Authors

  • JesseLea Carlin,

    1. Department of Pharmacology, Institute of Translational Medicine and Therapeutics, Perelman School of Medicine, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Tiffany E. Hill-Smith,

    1. Department of Psychiatry, Institute of Translational Medicine and Therapeutics, Perelman School of Medicine, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Irwin Lucki,

    1. Department of Pharmacology, Institute of Translational Medicine and Therapeutics, Perelman School of Medicine, Philadelphia, Pennsylvania, USA
    2. Department of Psychiatry, Institute of Translational Medicine and Therapeutics, Perelman School of Medicine, Philadelphia, Pennsylvania, USA
    Search for more papers by this author
  • Teresa M. Reyes

    Corresponding author
    1. Department of Pharmacology, Institute of Translational Medicine and Therapeutics, Perelman School of Medicine, Philadelphia, Pennsylvania, USA
    Search for more papers by this author

  • Disclosure: The authors declared no conflict of interest.

  • Full financial disclosures and author notes may be found in the online version of this article.

  • Funding agencies: This research was supported by the grants MH087978, MH86599, and T32 GM008076 from the National Institutes of Health (NIH). This work was supported by the following grants from the National Institutes of Health (NIH): MH087978 (T.M.R.), MH86599 (I.L.), and T32 GM008076 (J.L.C).

Abstract

Objective

To test whether high-fat diet (HFD) decreases dopaminergic tone in reward regions of the brain and evaluate whether these changes reverse after removal of the HFD.

Design and Methods

Male and female mice were fed a 60% HFD for 12 weeks. An additional group was evaluated 4 weeks after removal of the HFD. These groups were compared with control fed, age-matched controls. Sucrose and saccharin preference was measured along with mRNA expression of dopamine (DA)-related genes by Real Time-quantitative PCR (RT-qPCR). DA and 3,4-dihydroxyphenylacetic acid (DOPAC) were measured using high-performance liquid chromatography. DNA methylation of the dopamine transporter (DAT) promoter was measured by methylated DNA immunoprecipitation and RT-qPCR.

Results

After chronic HFD, sucrose preference was reduced, and then normalized after removal of the HFD. Decreased expression of DA genes, decreased DA content and alterations in DAT promoter methylation, was observed. Importantly, response to HFD and the persistence of changes depended on sex and brain region.

Conclusions

These data identify diminished DA tone after early-life chronic HFD with a complex pattern of reversal and persistence that varies by both sex and brain region. Central nervous system changes that did not reverse after HFD withdrawal may contribute to the difficulty in maintaining weight-loss after diet intervention.

Introduction

Overconsumption of widely available, calorie-dense palatable food is considered a major factor contributing to the high rates of obesity in the U.S [1]. Because palatable foods are often consumed after energy requirements have been met, the rewarding properties of palatable foods may override homeostatic satiety signals. Many neurotransmitters play a role in feeding behavior (e.g., opioids, dopamine (DA), gamma-aminobutyric acid (GABA), serotonin) as well as the integration of peripheral nutrient signals (e.g., leptin, insulin, ghrelin). DA signaling is a key mediator in both food reward and reward-seeking behavior, as DA in the mesolimbic/mesocortical region is associated with the rewarding properties of food, sex, and drugs of addiction [2]. Acutely, palatable food causes a burst of DA in the central reward system [3, 4]. With chronic consumption of rewarding food, the increased DA release over time may lead to adaptations that are associated with reward hypo-function.

Several lines of evidence support the hypothesis of altered DA function in obesity. Human imaging studies revealed blunted activation in reward regions of obese patients while drinking a highly palatable solution (milkshake) [5]. The blunted reward response was associated with less brain dopamine receptor D2 (DRD2) availability. In fact, mutations in the human dopamine D2 receptor have been linked with both obesity and addiction [6]. DA content in the synapse is largely controlled by dopamine transporter (DAT) uptake. DAT levels are negatively correlated with body mass index and genetic variants of DAT are also associated with obesity [7, 8]. Animal models of obesity have demonstrated decreases in basal extracellular DA and reduced DA neurotransmission in the nucleus accumbens (NAc) and ventral tegmental area (VTA) [9]. Decreases in DA-related genes after chronic high-fat diet (HFD) suggest decreased signaling in reward regions [12]. This decrease in DA activity after chronic HFD may reduce the sensitivity to natural rewards and facilitate continued overconsumption and further weight gain.

Early life is a critical period in brain development, and the early nutritional environment can influence brain pathways controlling food intake and energy metabolism. Early exposure of mice to a HFD for as little as 1 week altered adult caloric intake and expression of DA-related signaling molecules [16]. Further, early postnatal over nutrition in mice, driven by a small litter number throughout lactation, predisposes the offspring to adulthood obesity by altering hypothalamic development [17]. While it is clear that early life nutrition can affect brain development and obesity risk, little is known about the relative permanence of these changes across the lifespan. Additionally, previous studies have been done in male animals but females have been rarely studied in this context. To these ends, both male and female mice were studied for changes in gene expression and DA metabolism after they were made obese in early life through chronic consumption of a HFD from birth through 12 weeks of age. The DA system was also evaluated 4 weeks after removal of the HF diet, to examine whether the changes persisted or reversed.

Methods

Animals and experimental model

C57BL/6J females were bred with DBA/2J males (The Jackson Laboratory, Bar Harbor, ME). All dams were fed standard control diet (#5755; 18.5% protein, 12% fat, 69.5% carbohydrate) up until parturition when half the dams/litters were placed on HFD (Test Diet, Richmond, IN #58G9; 18% protein, 60% fat, and 20.5% carbohydrate). Offspring were weaned at 3 weeks of age and remained on either the control diet or the HFD until 12 weeks of age. Body weights were recorded weekly, and both male (n = 5-10) and female (n = 5-10) mice were used. The Institutional Animal Care and Use Committee of the University of Pennsylvania approved all procedures.

Sucrose and saccharin preference

In separate experiments, mice were individually housed (n = 8-10/group) in standard cages for 3 days with one bottle of 200 mL of the test solution (4% sucrose or 1% saccharin solution (w/v)) and another bottle with 200 mL of tap water. House chow was available ad libitum. Sucrose (mL), water (mL), and food consumption (g) were measured and the placement of the bottles was reversed daily. Preference was calculated using the average of the measurements from the last 2 days as follows: preference % = [(sucrose consumption/sucrose + water consumption) × 100].

Genomic DNA and total RNA isolation from the brain

Animals (n = 5/group) were euthanized with an overdose of carbon dioxide, followed by cervical dislocation; a method recommended by the Panel on Euthanasia of the American Veterinary Medical Association. Brains were then rapidly removed and placed in RNAlater (Ambion, Austin, TX) for 4-6 h before dissection. Brain dissections to isolate the prefrontal cortex (PFC), the nucleus accumbens (NAc), and the VTA were preformed as previously described [18]. Genomic DNA and total RNA were isolated simultaneously using AllPrep DNA/RNA Mini Kit (Qiagen).

Gene expression analysis by quantitative real-time PCR

For each individual sample, 500 ng of total RNA was used in reverse transcription using High Capacity Reverse Transcription Kit (ABI, Foster City, CA). Expression of target genes was determined by quantitative RT-PCR using gene specific Taqman Probes with Taqman gene expression Master Mix (ABI) on the ABI7900HT Real-Time PCR Cycler. Gene probes are listed in Supporting Information*. Relative amount of each transcript was determined using ΔCT values as previously described in [21]. Changes in gene expression were calculated against an unchanged GAPDH standard.

Ex vivo dopamine and dopamine metabolites

High-performance liquid chromatography (HPLC) was used to measure the content of DA and its metabolites in the mesolimbic reward areas of the brain (n = 8-12), as described previously [18, 22]. Brains were collected from animals and bisected into right and left hemispheres. The NAc and PFC were dissected out and quickly frozen by dry ice and stored at −80°C. The tissue was prepared for analysis by homogenization in 0.1 N perchloric acid, centrifuged at 15,000 rpm for 15 min at 2-8°C, and the supernatant filtered. Samples were analyzed by a Bioanalytical Systems HPLC (West Lafayette, IN) using a LC-4C electrochemical detector. Samples (12 μL) were injected onto a reverse phase microbore column at a flow rate of 0.6 mL/min and electrodetection set at +0.6 V. Separation for DA and DA metabolites was accomplished by a mobile phase consisting of 90-mM sodium acetate, 35-mM citric acid, 0.34-mM ethylenediamine tetraacetic acid, 1.2-mM sodium octyl sulfate, and 15% methanol v/v at a pH of 4.2. Peak heights of samples were measured and compared with standards for DA and its metabolite 3,4-dihydroxyphenylacetic acid (DOPAC).

Methylated DNA immunoprecipitation assay

Methylated DNA immunoprecipitation (MeDIP) assay was preformed using MagMeDIP kit (Diagenode, Denville, NJ). Methylated DNA was immunoprecipitated using 0.15 μL of magnetic beads coated with anti-5methylcytidine antibody (Diagenode) or mouse preimmune serum. Enrichment in MeDIP fraction was determined by quantitative RT-PCR using ChIP-qPCR Assay Master Mix (SuperArray) on the ABI7900HT Real-Time Cycler. For all genes examined, primers were obtained from SuperArray (ChIP-qPCR Assays (-01) kb tile, SuperArray) for the amplification of genomic regions spanning the CpG sites located ∼300-500 bp upstream of the transcription start sites. MeDIP results were expressed as fold enrichment of immunoprecipitated DNA for each site. To calculate differential occupancy fold change (% enrichment), the MeDIP DNA fraction CT values were normalized to Input DNA fraction CT values.

Statistics

Gene expression analysis was performed using Student's t-test comparing aged matched controls to HF and HF + recovery groups. The alpha level was adjusted for the multiple brain regions surveyed. Significance of a gene used in one brain region was P = 0.05; for two regions, P = 0.025, for three brain regions P = 0.016. Sucrose preference, saccharin preference, HPLC and MEDIP, body weights, and corticosterone assay analyzed using one-way ANOVA to compare control, HF, and HF + recovery groups. Post hoc Bonferonni Multiple Comparison Tests were used to compare pair-wise differences between groups. Significance for these tests was set at an alpha level of P = 0.05.

Results

Mice had continuous access to control diet (control) or 60% HFD until 12 weeks of age. At 12 weeks of age, half of the HF-fed animals were placed on the house chow for 4 weeks (HF +recovery). In both males and females, HFD animals (circles) were heavier than controls beginning at 9 weeks of age (P < 0.05) and remained heavier than controls throughout the recovery period (Supporting Information Figure 1).

Sucrose and saccharin preference tests were administered to assess animals' response to natural and nonnutritive rewarding stimuli. Sucrose preference but not saccharin preference was altered after HFD exposure and returned to normal levels after HFD recovery in males and females. One-way ANOVA revealed sucrose preference was significantly decreased in males (Figure 1A) and trended toward a decrease in females (Figure 1B) after HFD exposure (F(2,16) = 4.82, P < 0.05; F(2,16) = 5.41, P < 0.06, respectively). After removal of the HFD, this behavior normalized and sucrose preference no longer differed from controls. Saccharin preference was not altered in either males (Figure 1C) or females (Figure 1D) as a result of HFD exposure.

Figure 1.

Sucrose preference but not saccharin preference is altered after high-fat diet (HFD) exposure and returns to control levels after HFD recovery in males and females. Sucrose preference (A, B) and saccharin preference (C, D) were evaluated in counter balanced order of control (white bars) HFD (black bars) and HFD + recovery (patterned bars) males and females (n = 8/group). *P < 0.05 significantly different from control group.

Because DA is a key regulator of reward behavior, DA-related gene expression was examined within the reward circuitry of a separate cohort of males and females after 12 weeks on the HFD, and in an additional group, after 4 weeks recovery from the HFD. Table 1 summarizes the gene expression patterns and statistical analysis in the VTA, PFC, and NAc. In the VTA, three genes important in regulating DA levels at synaptic terminals were measured: catecholamine methyl transferase (COMT) involved in inactivation of catecholamine neurotransmitters; DAT, membrane spanning pump that clears DA from the synapse, and tyrosine hydroxylase (TH), the rate-limiting enzyme for DA synthesis. Fold change values for each group were determined using aged matched controls (e.g., both control time points are set to 1, and for clarity, only the control for HFD is depicted on the graph). Student's t-test (n = 5/group) revealed in the male VTA that COMT, DAT, and TH mRNA were significantly decreased by HFD exposure (Figure 2A) and returned to or exceeded the control levels after a recovery period off the diet (HF + recovery).

Figure 2.

Chronic high-fat diet (HFD) and recovery after HFD alters dopamine-related gene expression in males and females. Gene expression was measured in the ventral tegmental area (VTA; A, D), prefrontal cortex (PFC; B, E), and nucleus accumbens (NAc; C, F) of control (white bars), HFD (black bars) and HFD + recovery (patterned bars) males and female mice (n = 5/group). HFD and HF + recovery groups are fold change from their own aged matched controls. Only one control group (12 weeks) is shown and set to 1. *P < 0.05, **P< 0.01, ***P< 0.001, ****P< 0.0001 significantly different from own control group.

Table 1. Gene expression summary and statistics in males
Gene12 Weeks HFD (fold change)P value4 Weeks recovery (fold change)P value
  1. Gene expression results for male ventral tegmental area (VTA), prefrontal cortex (PFC), and nucleus accumbens. Summary of fold change, significance, and P values are presented for HF and HF + recovery groups as compared with their age matched control.
VTA Males  
COMT<Control (0.36)0.0007>Control (1.93)0.0036
DAT<Control (0.42)0.0008>Control (1.46)0.0132
TH<Control (0.38)0.0037= Control (1.15)0.1454
PFC
COMT= Control (0.74)0.066<Control (0.58)0.004
DARPP-32>Control (1.62)0.0064>Control (1.55)0.03
DRD1<Control (0.37)0.0004<Control (0.55)0.0052
DRD2<Control (0.52)0.0096<Control (0.45)0.0042
NAc
COMT<Control (0.40)0.0026<Control (0.13)<0.0001
DARPP-32>Control (2.03)0.023<Control (0.23)0.0005
DRD1<Control (0.52)0.0053<Control (0.18)0.0014
DRD2<Control (0.51)0.0004<Control (0.19)0.009

In the PFC and NAc, genes important for DA signaling and DA turnover were examined (n = 5/group): COMT; protein phosphatase 1 regulatory subunit 1B (DARPP-32), a downstream signaling protein regulated by receptor stimulation; dopamine receptor D1 (DRD1), a postsynaptic G-protein coupled receptor that stimulates adenylyl cyclase; and DRD2, a postsynaptic G-protein coupled receptor that inhibits adenylyl cyclase. In the male PFC (Figure 2B), DARPP-32 was increased, whereas DRD1 and DRD2 were decreased after HFD exposure, and these changes persisted after the removal of the HFD (although the increase in DARPP-32 mRNA was not statistically reliable). In the male NAc (Figure 2C), COMT, DRD1, and DRD2 were decreased by HFD exposure, and remained below control levels after removal of the HFD. DARPP-32 levels were increased by HFD, but significantly decreased from controls after 4 weeks off the HFD.

The same brain regions and genes were examined in female mice (n = 5/group). As shown in Table 2, there were significant differences observed in the pattern of gene expression in response to the HFD, as well as to the recovery off the diet. Similar to males, in the VTA, mRNA levels of COMT, and TH were significantly decreased after HFD exposure (Figure 2D). However, unlike the males, these changes persisted after the removal of the HFD. Further, in direct opposition to the pattern observed in males, HFD exposure increased DAT mRNA expression in the VTA in females, and after removal of the HFD levels were even lower than age matched controls. In the PFC, only DARPP-32 was affected by chronic HFD, with a significant increase in mRNA levels after 12 week HFD, and a return to control levels after removal of the HFD. Both COMT and D1R mRNA were significantly decreased after 4 weeks off the HFD. In the female NAc, COMT, DRD1, and DRD2 were all decreased after HFD exposure (Figure 2F). DRD1 and DRD2 recovered to control levels after diet removal, whereas COMT levels remained significantly decreased after 4-week recovery.

Table 2. Gene expression summary and statistics in females
GENE12 Weeks HFD (fold change)P value4 Weeks recoveryP value
  1. Gene expression results for male ventral tegmental area (VTA), prefrontal cortex (PFC), and nucleus accumbens. Summary of fold change, significance, and P values are presented for HF and HF + recovery groups as compared with their age matched control.
VTA Females  
COMT<Control (0.50)<0.0001<Control (0.07)<0.0001
DAT>Control (1.93)0.0004<Control (0.17)<0.0001
TH<Control (0.64)0.01<Control (0.27)<0.0001
PFC
COMT<Control (0.77)0.028<Control (0.37)<0.0001
DARPP-32>Control (1.98)0.0002= Control (0.85)0.2113
DRD1= Control (1.21)0.398<Control (0.56)0.0214
DRD2= Control (1.14)0.6225= Control (0.67)0.1566
NAc
COMT<Control (0.67)0.0027<Control (0.17)<0.0001
DARPP-32= Control (0.83)0.296>Control (2.05)0.052
DRD1<Control (0.47)<0.0001= Control (0.76)0.1861
DRD2<Control (0.48)<0.0001= Control (0.74)0.063

Given the consistent decrease in gene expression for DA regulating genes in the VTA, DA and DA metabolites were quantified in regions that receive projections from the VTA, PFC, and NAc. Figure 3 shows DA and the dopamine metabolite (DOPAC) from the PFC and NAc in males (Figure 3A and 3C) and females (Figure 3B and 3D). In males, exposure to the HFD produced a decrease in DA levels in PFC (Figure 3A) and NAc (Figure 3C) (F(2,13) = 3.95; F(2,18) = 3.536, P < 0.05), which recovered after HFD removal only in the NAc. Dopamine turnover (DOPAC:DA ratio) increased in male PFC (F(2,12) = 3.85, P < 0.05) and NAc (F(2,17) = 4.69, P < 0.05). In contrast, the effect of HFD on DA and DOPAC in females was qualitatively different than in males. In the PFC, HFD did not affect DA or DOPAC levels. In the NAc, DA levels were decreased in HFD-fed animals and remained decreased even after removal of the HFD (Figure 3D, F(2,23) = 4.79, P < 0.05). DOPAC levels were unchanged in the NAc of females, which resulted in an increase in DA turnover (DOPAC:DA ratio) (F(2,23) = 7.00, P < 0.01).

Figure 3.

Decrease in dopamine levels in PFC and NAc after HFD from birth and mixed recovery after HFD removal. Dopamine and dopamine metabolites were measured in the PFC (A, B) and NAC (C, D) of control (white bars), HF from birth (black bars) and HF + recovery (patterned bars) (n = 8-10/group). *P< 0.05, **P< 0.01 significantly different from control group.

Given that DAT transcription can be regulated by differential DNA methylation and the observation of a notable sex difference in the expression of DAT in the VTA, DNA methylation in the promoter region of DAT was examined. In Figure 4A and 4C, DAT gene expression in the VTA is presented again for clarity (taken from Figure 2A and 2D). DAT promoter methylation was significantly increased in males (Figure 4B) after HFD and returned to control levels in HFD + recovery males (F(2,11) = 23.64, P < 0.01). In females, DAT promoter methylation trended toward decrease in HFD animals (D) and was significantly decreased in HFD + recovery females (Figure 4D, F(2,12) = 5.70, P < 0.05).

Figure 4.

Changes in DNA methylation status of DAT promoter parallel changes in gene expression in the VTA. Genomic DNA was isolated from the dissected VTA of control (white bars), HF from birth (black bars), and HF + recovery (patterned bars). DNA was digested by restriction enzyme MSE1 and immunoprecipitated by a 5-methylcytosine antibody. The enrichment of DNA methylation relative to input DNA in the promoter region of DAT was quantified by qPCR (B, D). DAT mRNA levels, (A, C) previously presented in Figure 1) are shown for comparison (n = 5/group). *P< 0.05, **P< 0.01, ***P< 0.001 significantly different from control group, ####P< 0.0001 significantly different than HFD group.

To assess if removal of the HFD in the recovery period was a stressor, baseline plasma corticosterone levels (μg/dL) were taken in control, HFD exposed (12 weeks), HFD + 1-week recovery and HFD + 4-week recovery groups (n = 5/group, Supporting Information Figure 2). One-way ANOVA revealed no significant differences between groups in male animals (F(3,16) = 3.21, n.s.).

Discussion

Chronic consumption of a HFD beginning in early life was used to establish diet-induced obesity in mice. Mice displayed decreased sucrose preference and evidence of reduced dopaminergic tone in reward regions of the brain. After 4 weeks off the HFD, sucrose preference normalized in both males and females; however, some DA gene expression changes persisted. These experiments provide important new data describing the effect of chronic HFD on the brain reward system, highlighting the capacity for recovery and key sex differences between male and female mice.

In the HFD fed animals, a decreased sucrose preference was observed, which reversed after a recovery period. These findings extend our previous report of HFD intake driving a reduced sucrose preference [23] by demonstrating that this can occur with a shorter duration of HFD exposure (12 weeks vs. 22 weeks), and importantly, that the response recovers in the absence of HFD. Female mice demonstrated the same response patterns as males. These findings are consistent with others in the literature which have shown through the inclusion of a pair fed group that chronic HFD, and not obesity per se, attenuates the response for sucrose in an operant task [24]. Similarly, in the current study, sucrose preference recovered after 4 weeks off the HFD, while body weight remained significantly elevated, supporting the conclusion that decreased sucrose preference was driven by the HFD exposure and not the accompanying body weight gain. It was particularly interesting that there was no change in saccharin preference. This may indicate that chronic HFD differentially affects the response to caloric and noncaloric sweet rewards. Post-ingestive effects have been shown to influence preference independent of palatability, as sucrose intake has been shown to induce DA release in “sweet-blind” taste receptor knockout mice [25], nutritional value is required for reward and reinforcement [26] and taste-independent metabolic sensing pathways have been defined in Drosophila [27]. Saccharin is significantly sweeter than sucrose, so an effort was made to establish equivalency in sweetness (typically 4-10× higher concentration of sucrose [26]); however, the overall preference for saccharin was lower than that for sucrose in these animals. Therefore, an alternative explanation may be that HFD differentially affected the sucrose preference because it was relatively more rewarding than saccharin (high vs. low value reward), although animals still evinced a strong preference for saccharin (∼75-80% preference for saccharin compared with ∼85-90% preference for sucrose).

Overall, dopaminergic gene expression within the VTA, NAc, and PFC was decreased in male mice following chronic HFD. These findings are consistent with other studies that observed decreases in DA-related genes in response to HFD [12, 13, 15]. Decreases in dopamine D2 receptor expression and function have been observed in human imaging studies [5, 28] and rodent obesity models [14, 29]. Decreased DA signaling reduces the sensitivity to natural rewards and may therefore facilitate continued overconsumption of palatable foods and further weight gain [30, 31]. Further, disrupted DA homeostasis driven through decreased DAT surface expression is known to drive increased intake of HFD [32]. An exception to this pattern was seen with DARPP-32, a DA- and cyclic AMP-regulated phosphoprotein, which increased after HFD in NAc and PFC. DARPP-32 plays a key role in integrating a variety of biochemical and behavioral responses controlled by DA. It may be that DARPP-32 upregulation was compensatory in response to the chronic downregulation of D1R. In a similar model (12-week HFD in mice), it has been shown that D1R downregulation was matched by an increase in phosphorylation of DARPP-32 in NAc [33].

Few studies have examined the capacity for recovery of these changes after removal of the HFD. However, in two recent reports, gene expression changes and reward system defects persisted after a short withdrawal period (14-18 days) [13, 14]. In contrast, studies in obese patients before and after gastric bypass surgery have shown a reversal of dopaminergic changes after a longer period of weight loss [34]. In males, the pattern of recovery varied by brain region. In the VTA, the observed decreases in COMT, DAT, and TH were all normalized with the removal of the HFD. In contrast, all gene expression changes observed in the NAc and PFC did not normalize. In the current study, chronic HFD led to significant weight gain and after 4 weeks off the diet, animals were still significantly heavier than controls. Therefore, the subsequent metabolic and hormonal changes that accompany obesity (e.g., increased leptin, elevated adipokines) were likely still present at 4 weeks off the diet. Therefore, gene expression changes that normalized (e.g., in the VTA) may have been primarily driven by the HFD, while those that were maintained (in NAc and PFC) may be more tightly coupled to obesity. Maintenance of weight loss by dieting is characteristically low (with 67% [35] to 80% [36] of patients regaining the lost weight). This persistence of gene expression changes in reward regions could be important in partly explaining this common occurrence. It is also important to note that the observed behavioral and gene expression changes are not likely to be due to stress associated with changing diets, as there were no significant changes in basal plasma corticosterone levels on the HFD or after 1-week or 4-week recovery.

Interesting sex differences were revealed, both in the response to chronic HFD, as well as in response to diet removal. Females were similar to males in showing an overall decrease in DA-related genes that would predict a decrease in DA activity, particularly in the VTA and the NAc. One noteworthy sex difference was the increase in DAT mRNA expression in the female VTA after HFD. This difference in gene expression, coupled with similar decreases in TH gene expression in both sexes, would suggest significant differences in DA neurotransmission within the NAc, both at the end of the HFD exposure as well as after the recovery period. A greater appreciation for the functional significance of these differences is an important focus of future research.

Additionally, while COMT and TH decreases recovered in the male VTA, these decreases persisted in the females after 4-week off the HFD. It is yet to be determined whether these differences would reverse with a longer time off the diet; however, it supports the conclusion that females are at the very least slower to recover, if they recover at all. Further, gene expression changes of D1R and D2R in NAc and PFC were quite different between males and females. In males, there was a general decrease in gene expression in both regions that largely persisted after diet removal. In females, D1R and D2R were decreased in the NAc and then recovered, but there was no effect of HFD on DA receptors in PFC. In the current studies, female animals were sacrificed without accounting for estrus stage. While some of the observed endpoints are known to vary across the estrus cycle, female animals in this study did not demonstrate increased variance across the endpoints, particularly when compared with the effect of the diet manipulations.

To complement the gene expression findings, DA was measured in the primary projection regions of the VTA, namely, the PFC and the NAc. DA levels tended to parallel changes seen in TH mRNA in the VTA. In the NAc of both males and females, levels of DA decreased in response to the HFD diet; a response which recovered in males, but not females. In the PFC, DA levels were also decreased by HFD; however, there was no recovery off the diet in the PFC. Additionally, females had lower levels of DA in the PFC than males. Sex differences in DAT expression and function are well known in the literature, with females demonstrating increased DAT expression [37] and function [38], and these differences may contribute to the different baseline levels of DA between males and females. Examination of the DOPAC:DA ratio is informative as well. An increase in this ratio may have reflected a compensatory response driven by decreases in DA. The long-term functional significance of these changes in DA metabolism would be illuminated by measuring changes in DA release using in vivo microdialysis.

Moreover, these data identify dynamic regulation of DNA methylation within the promoter of the DAT gene, particularly in the males. Recently, we have demonstrated that DAT expression can be dynamically regulated by differential DNA methylation in response to HFD [12], and that increased DAT promoter methylation correlates with a decrease in gene expression. Here, we identify the plasticity of this response, as the increased DNA methylation (and decreased expression of the mRNA) seen in males reverses upon removal of the HFD. Epigenetic gene regulation, for example, through changes in DNA methylation, presents a pathway whereby organisms can readily adapt to environmental challenges. Epigenetic marks can be maintained across the lifespan [39], and in cultured embryonic stem cells, both reversible and persistent patterns of differential DNA methylation were observed in response to changing environmental conditions [40]. These data are the first to demonstrate in vivo a dynamic methylation pattern that changes with the presence or absence of an environmental challenge. It was notable that this same pattern was not observed in females. While the initial response to the HFD was as predicted (decreased DNA methylation driving increased gene expression), this pattern was not maintained throughout the recovery period. This suggests that DNA methylation and gene expression may become uncoupled during the 4 weeks off the HFD or it may suggest that DAT mRNA is regulated by other means in females.

In males, sucrose preference, DA-related gene expression in the VTA, and DA in the NAc follow a consistent pattern, of suppression in response to the chronic HFD that recovers after removal of the diet. Interestingly, while the behavioral responses to sucrose are similar in the females, both the gene expression pattern and NAc DA levels show a lack of recovery upon removal of the HFD. Reward-related behaviors are clearly influenced by additional neurotransmitter systems such as the opioids, and perhaps in females, the behavioral response to sucrose is more strongly associated with changes in opioids. Overall, the present data suggest that sex differences in the both the initial response to HFD, as well as to recovery after removal of the HFD, with regard to DA-related gene expression represent an important direction for future research directed at how chronic consumption of a HFD impacts the brain reward system. Most notably, these data identify significant plasticity in the dopaminergic response to HFD, suggesting that while the adverse effects of chronic HFD consumption and/or obesity are significant, the potential for recovery exists.

Ancillary