Moderate-Intensity Aerobic Exercise Enhanced Dopamine Signaling in Diet-Induced Obese Female Mice without Preventing Body Weight Gain

Abstract

Obesity continues to rise in prevalence and financial burden despite strong evidence linking it to an increased risk of developing several chronic diseases. Dopamine response and receptor density are shown to decrease under conditions of obesity. However, it is unclear if this could be a potential mechanism for treatment without drugs that have a potential for abuse. Therefore, the aim of this study was to investigate whether moderate-intensity exercise could reduce body weight gain and the associated decreases in dopamine signaling observed with high-fat diet-induced adiposity. We hypothesized that exercise would attenuate body weight gain and diet-induced inflammation in high-fat (HF)-fed mice, resulting in dopamine signaling (release and reuptake rate) comparable to sedentary, low-fat (LF)-fed counterparts. This hypothesis was tested using a mouse model of diet-induced obesity (DIO) and fast-scan cyclic voltammetry to measure evoked dopamine release and reuptake rates. Although the exercise protocol employed in this study was not sufficient to prevent significant body weight gain, there was an enhancement of dopamine signaling observed in female mice fed a HF diet that underwent treadmill running. Additionally, aerobic treadmill exercise enhanced the sensitivity to amphetamine (AMPH) in this same group of exercised, HF-fed females. The estrous cycle might influence the ability of exercise to enhance dopamine signaling in females, an effect not observed in male groups. Further research into females by estrous cycle phase, in addition to determining the optimal intensity and duration of aerobic exercise, are logical next steps.

Introduction

Diet-induced obesity (DIO) is associated with the development of psychiatric disorders including depression and anxiety.[1,30] The public health burden of obesity and brain related disorders is increased by diet patterns leading to a state of chronic low-grade inflammation that dysregulates the dopamine system.[5] Pharmaceutical therapies to attenuate body weight gain and reduce inflammation may improve brain health, though long term impacts of these treatments are unclear and weight-loss maintenance following treatment removal is poor. A low-cost alternative to pharmaceuticals is moderate exercise which is shown to reduce body weight and decrease chronic inflammation, potentially improving brain health by attenuating the DIO effects on dopamine signaling.

DIO is caused by excess calories, often in the form of ultra-processed foods and other foods high in saturated fat. In neurons, saturated fat is known to bind the same toll-like receptor (TLR4) as lipopolysaccharide (LPS), triggering pro-inflammatory cytokine release from microglial immune cells in the brain.[22,26,59] A review by Felger and Treadway (2017) outline how inflammation, putatively caused by DIO, is linked to disrupted dopamine synthesis and synaptic regulation of dopamine in the brain.[18] Systemic inflammation caused by a high-fat diet is reported to decrease phasic dopamine release[2,16] and change dopamine receptor density.[17] Adipose tissue release pro-inflammatory cytokines when rapid adipocyte expansion promote hypoxia and nutrient deprivation leading to apoptosis.[13] Pro-inflammatory cytokines interleukin-6 (IL-6), tumor-necrosis factor alpha (TNF-α), and interleukin-1beta (IL-1β) are released from enlarged adipocytes[31] and macrophages when an innate immune response is activated.[14] Similarly, microglia release pro-inflammatory cytokines as the resident immune cell type of brain tissue.[29,58] TNF-α and IL-1β are some of the first cytokines released in response to an inflammatory insult (seconds; peaking after minutes) with slower responding cytokines (IL-6) appearing on a longer timescale (minutes; peaking after hours).[49,60] Evidence suggests these cytokines influence dopamine signaling, but questions remain as to the time-course and magnitude of interaction. While several brain regions contribute to dopamine production and release, the nucleus accumbens (NAc) is unique due to its multifaceted role in integrating motivational drive[14] and the salience of palatable foods[55] into a dopamine receptor (D1)-mediated satiety circuit to the lateral hypothalamus.[35] Previous reports of reduced NAc dopamine release and reuptake with DIO [2,15,56] may disrupt the integration of these signals and promote excessive food intake. Inflammation has been shown to reduce dopamine release in the NAc,[15] so strategies to mitigate inflammation could normalize disruptions in this motivation/feeding circuit.

Exercise elevates markers of inflammation in an acute timeframe[38,48]; however, training adaptations following consistent exercise improve immune system efficiency and reduce prolonged immune system activation.[36,42] Exercise also impacts dopamine circuits, as aerobic exercise was shown to increase dopamine signaling during anticipation of a reward in Parkinson’s patients.[41] Similarly, behavioral improvements were observed in children with ADHD who underwent a moderate-intensity exercise regimen.[37] These findings suggest an underlying mechanism that is common between unique neurological disorders with disrupted dopaminergic pathways. However, the potential interaction between pro-inflammatory molecules modulating dopamine neurotransmission is unclear. Exercise has the potential to attenuate inflammation caused by adiposity through the release of brain-derived neurotrophic factor (BDNF). BDNF levels increase following acute aerobic exercise protocols,[44,52] and have been shown to enhance neurogenesis[27] and neuronal plasticity.[12] However, the expression of BDNF after a chronic exercise regimen yielded mixed results,[25,44] confounded by the type of biological sample used for analysis (brain tissue vs plasma).[25,44] Plasma BDNF levels were shown to decrease after long-term exercise in a population of mid-aged men that showed improved memory,[25] suggesting that there might be an interaction between BDNF and dopaminergic pathways due to long-term potentiation in the striatum that is required for dopamine-dependent learning.[3,4,8,40] However, BDNF is increased directly following an exercise bout and might have interactions with dopaminergic pathways, as shown by Bastioli et al., 2022.[6]

A canonical marker of neuroinflammation is activated microglia, measured by a ratio of ionized calcium binding adaptor molecule-1 (Iba1) to cluster of differentiation receptor (Cd11b).[24,47] Cd11b is constitutively expressed by microglia; however, Iba1 expression is elevated when microglia are activated to release pro-inflammatory cytokines.[20] Utilizing a ratio of the mRNA for these proteins, rather than one in isolation, can provide a more reliable picture of the inflammatory state in the NAc. With this in mind, the aim of this study was to investigate if moderate-intensity exercise could normalize reduced dopamine signaling in the NAc by either attenuating adiposity-induced inflammation or by modulating the neuroimmune response which is reported to dampen dopamine release.[15] We used fast-scan cyclic voltammetry to measure pre-synaptic dopamine kinetics in brain slices containing the NAc. We also measured markers of inflammation in mice fed a low-fat (LF) or high-fat (HF) diet with or without a treadmill exercise intervention. We hypothesized that exercise would attenuate body weight gain and diet-induced inflammation in (HF)-fed mice to resemble their (LF)-fed counterparts more closely.

Experimental Procedures

Animals, Diet, and Experimental Design

Six-week-old C57BL/6 mice were purchased from Jackson Laboratories (Bar Harbor, ME) and housed on a reversed 12-hour light/dark cycle (lights off 0600, lights on 1800). Mice were given free access to food and water and randomly assigned to one of four groups: 1) LF sedentary (LF), 2) LF exercised (LFex), 3) HF sedentary (HF), or 4) HF exercised (HFex). LF-fed groups were provided a diet with 10% total kcals from fat (Research Diets, D12450J) while HF-fed groups were provided with a diet 60% total kcals from fat (Research Diets, D12492). The mice had ad libitum access to their prescribed diets and water for six weeks. All experiments were conducted following approval by the UNC Greensboro Animal Care and Use Committee in compliance with the Guide for the Care and Use of Laboratory Animals (NIH).

Treadmill Running

The aerobic exercise protocol consisted of forced treadmill running on a Collins treadmill (Braintree, USA). Exercised groups were acclimated to the treadmill and the onset of the exercise program coincided with the onset of experimental diet consumption. All running bouts started with a warm-up period at a rate of 10m/min for 5 minutes. Speed was gradually increased to 15m/min for the remaining run duration. Mice ran for 30 min/day, 5days/week starting week 1. Duration was progressively increased by 10 minutes/day for each consecutive week until a maximum of 60min/day was reached.

Intraperitoneal Glucose Tolerance Test

Intraperitoneal glucose tolerance tests (IPGTT) were performed to examine blood glucose clearance. Mice were fasted 12 hours prior to IPGTT which began three hours into the dark cycle (0900). Fasting body weights were recorded to calculate volume required for 20% glucose administration. A topical anesthetic (lidocaine 2.5% and prilocaine 2.5% (item #977934 CVS, NC)) was applied to the tail tip, which was then nicked with a surgical blade to produce a droplet of blood for blood glucose measurement with a glucometer (item #323932 CVS, NC). Baseline glucose was measured, then mice received an intraperitoneal injection of 20% glucose (2g/kg bodyweight; Sigma, St. Louis, MO) in sterile saline (0.9% sodium chloride, Hospira; Lake Forest, IL). Blood glucose was measured at 15-, 30-, 60-, and 120-minute intervals. These values were plotted and reported as area under the curve (AUC).

Tissue Collection

After the six-weeks diet and exercise protocol, mice were rendered unconscious with 5% isoflurane and euthanized by rapid decapitation. The brain was removed and hemisected with left brain hemisphere delegated for voltammetry preparation and remaining right hemisphere used to dissect out the NAc, which was flash frozen in liquid nitrogen and stored at −80°C for gene expression analysis.

Real-Time PCR

Frozen tissue samples stored at −80°C were lysed via sonification in RNA lysis buffer (catalog #: R1060-1-100, Zymo Research; CA, USA) and then centrifuged at 13,000rpm for 30 seconds. The resulting supernatant was removed and RNA was further purified through a series of spin centrifuge reactions following manufacturer’s instructions (Quick-RNA^™ MiniPrep, catalog #: R1055, Zymo Research; CA, USA). Complementary DNA (cDNA) was created using a reagent kit from (ThermoFisher Scientific, USA; catalog #: 4368813) according to protocol previously described.[50] RT-PCR was run using TaqMan Advanced Master Mix (catalog #: 4444556, Applied Biosystems; USA) and probes for genes: 18s (Mm03928990_g1), Tyrosine Hydroxylase (Mm00447557_m1), Itgam (Mm00434455_m1), Aif1 (Mm00479862_g1), IL-6 (Mm00446190_m1), IL-1B (Mm00434228_m1), TNFα (Mm00443258_m1), and Slc6a3 (Mm00438388_m1). Gene expression was quantified using the ΔΔ^ct method[28] with 18s used as the endogenous control.

C-Reactive Protein

Trunk blood was collected in (1.5mL Flex-Tubes, catalog #: 022363531; Eppendorf, USA) then centrifuged for 10 min at 3,000 rpm at 4°C and plasma was decanted and stored at −80°C until analysis. ELISA kits were used to analyze CRP protein levels and performed according to manufacturer’s instructions. (Catalog #: MCRP00, R&D Systems, Minneapolis, MN)

Fast Scan Cyclic Voltammetry (FSCV)

Mice from each treatment group proceeded to FSCV in a Latin-square design. Following brain removal, brains were placed in oxygenated artificial cerebrospinal fluid (aCSF) (126mM NaCl, 25mM NaHCO₃, 11mM D-glucose, 2.5mM KCl, 2.4mM CaCl₂, 1.2mM MgCl₂, 1.2mM NaH₂PO₄, 0.4mM L-ascorbic acid) (Sigma product codes: NaCl-1002579927, NaHCO₃-S5761, D-glucose-1002593418, KCl-1002323941, CaCl₂-1002393239, MgCl₂-1002368596, NaH₂PO₄-1002355618, L-ascorbic acid-1002334976) and sliced coronally into 300μm thickness with a compresstome (Precisionary Instruments; Greenville, NC) as previously described.[15] Slices were equilibrated for 60 minutes in oxygenated aCSF flowing at a rate of 100mL/hr before recordings began. A glass capillary-pulled carbon fiber electrode (C005722, batch 4, carbon metal fiber, 10m length; Goodfellow) (A-M Systems, catalog #60200; Sequim, WA) was placed approximately 50μM into the slice within the NAc next to a bipolar stimulating electrode (Plastics One; Roanoke, VA) that evoked terminal release of dopamine with a single monophasic electrical pulse (350μA, 4ms) or a five-pulse train at 20Hz (5p20Hz). Dopamine release and uptake was recorded for 15 seconds with a three-minute recovery window between stimulations. Baseline dopamine recordings were considered stable when the current detected for dopamine peak height was within 10% between three or more recordings. An Ag/AgCl reference (Precision Instruments; Sarasota, FL) was used to scan a triangular waveform between −0.4V and 1.2V at a rate of 400V/s every 100ms. Dopamine current (nA) was converted to concentration (μM) using a calculation of electrode sensitivity through DEMON voltammetry and analysis software.[61] Michaelis-Menten kinetics were used to determine the concentration of maximal dopamine release [DA] and reuptake rate (V_max) with K_m held at 160nM, corresponding with the affinity of dopamine for the dopamine transporter.[32] K_m was only adjusted after the application of amphetamine (AMPH), to measure any diet- or exercise-induced differences in the affinity of dopamine for the DAT.[61] After baseline recordings were established we collected multiple-pulse recordings at 5p20Hz. We then re-established a stable single-pulse baseline prior to the application of the competitive inhibitor AMPH at concentrations of 300nM and 3μM. Dopamine release [DA] and reuptake rate (V_max) were analyzed for differences across groups.

Statistical Analysis

All analyses were performed in GraphPad Prism (v.9.1.1). Three-way ANOVA was used to analyze sex-differences between males and females. Two-way analysis of variance (ANOVA) was used to identify diet or exercise effects in daily food intake, body weight and dopamine related outcomes (release, reuptake rate) within sex. All group differences were assessed using šidák’s or Tukey’s post-hoc tests. Results are expressed as mean ± standard error of the mean (SEM). Outliers were identified and removed using the ROUT method, which is based on false discovery rate, and maximum threshold specified was 1%.

Results

Moderate-intensity exercise did not significantly reduce body weight gain or normalize blood glucose during six-weeks of HF-feeding.

Body weight gain was tracked weekly for the duration of the study and expressed as a percent of initial body weight (Figure 1). All groups gained significant weight from weeks 1-6 (p<0.0001) (Fig. 1–A males, 1C females), consistent with growth and development from adolescence into early adulthood. A main effect of diet was detected for body weight gain in males (p<0.0001), where sedentary HF-fed gained significantly more body weight (13.78g ± 1.45g) than the LF diet (4.64g ± 0.14g) (p<0.0001). Moderate-intensity exercise did not significantly influence body weight gain for either diet group (HFex- 10.68g ± 1.10g and LFex- 3.84g ± 0.6g (Fig. 1B). A similar main effect of diet was observed in females, (p=0.0003), where sedentary HF-fed females gained significantly more weight (8.99g ± 0.84g) than the LF-fed group (1.93g ± 0.21g) (p<0.0001). Exercise did not attenuate body weight gain in females (Hf- 6.67g ± 1.05g and LF- 1.03g ± 0.36g) (Fig. 1C). Body weight gain was also expressed as a percent of weight gain in addition to initial and final body weights with similar results for males (Fig. 1B) and females (Fig. 1D).

Figure 1. Measures of body weight for each experimental group.

Initial body weight from week 1 (Wk1) and final body weight from week 6 (Wk6) are reported for males (A) and females (C) for each of the experimental groups-low fat sedentary (LF), low fat exercised (LFex), high fat sedentary (HF), and high fat exercised (HFex). The changes in body weight are reported as % body weight gain for males (B) and females (D). ***p<0.001; ****p<0.0001

Metabolic phenotyping was assessed using fasting blood glucose measurements and blood glucose clearance challenge via IPGTT (Figure 2). Fasting blood glucose was significantly elevated in sedentary HF-fed males (Fig. 2A) and females (Fig. 2C) (p<0.001) compared to LF-fed counterparts. Additionally, HFex females had significantly elevated fasting blood glucose compared to LFex counterparts (p<0.05) (Fig. 2C). Area under the curve (AUC) was calculated from IPGTT time points to evaluate blood glucose clearance (Fig. 2B males, 2D females). HF-fed groups had significantly larger AUC in comparison to LF-fed males (Fig. 2B) and LF-fed females (Fig. 2D) (p<0.0001, respectively). Exercise did not influence blood glucose clearance with no differences detected between LF and LFex males or females. Similarly, no differences between HF and HFex males or females (Fig. 2B males, 2D females).

Figure 2. Measurements of Glucose Metabolism.

Fasting blood glucose for males (A) and females (C) are shown in addition to area under the curve (AUC) for intraperitoneal glucose tolerance tests (IPGTT) in both males (B) and females (D). *p<0.05; ***p<0.001; ****p<0.0001

Moderate-intensity exercise increased single-pulse dopamine release and reuptake rate in HF-fed female mice.

FSCV was used to measure pre-synaptic dopamine release and reuptake rate (V_max) (Figure 3). No differences in dopamine release from single pulse (Fig. 3A) or five-pulse (5p20Hz) stimulations (Fig. 3B) were observed, and no differences in V_max were observed for males (Fig. 3C). In HF-fed females, exercise significantly increased single pulse dopamine release (Fig. 3D) and increased V_max (Fig. 3F) (p<0.05). However, no differences in dopamine release were seen under evoked multi-pulse conditions (Fig. 3E).

Figure 3. Dopamine Release and Reuptake Rate.

Evoked single pulse μM dopamine [DA] release for males (A) and females (D) for each experimental group. Dopamine release for evoked multi-pulse is expressed as a % of baseline release for both males (B) and females (E). Enzyme kinetics for reuptake of dopamine is shown as V_max (μM/s) for males (C) and females (F). *p<0.05

HF-fed mice displayed faster AMPH-induced depletion of pre-synaptic terminal dopamine and greater sensitivity to inhibition of dopamine transporter reuptake.

After baseline recordings, AMPH (300nM and 3μM) was bathed over brain slices for 30 minutes each (Fig. 4). In sedentary males, AMPH depleted dopamine release from NAc terminals to a greater extent in the HF group compared to the LF group (p<0.05) (Fig. 4A). Similarly, males in the HFex group had greater depletion of dopamine release with AMPH compared to LFex (p<0.05) (Fig. 4D). However, exercise had no effect on dopamine depletion within diet groups LF vs. LFex or HF vs. HFex (Fig. 4C, 4D). In females, no differences in dopamine release were observed following AMPH application for any group comparisons (Fig. 4E–H).

Figure 4. Dopamine release in presence of AMPH.

Single-pulse evoked dopamine release [DA] expressed as a % of baseline for males (A-D) and females (E-H) with application of increasing amphetamine (AMPH) concentrations (300nM, 3μM) over 60-mins. 300nM AMPH was applied over first 30 minutes (light grey panels) and 3μM AMPH during second 30-minute timeframe (darker grey panels). Comparisons are displayed between LF/HF groups (A, E); LFex/HFex groups (B, F); LF/LFex groups (C, G); and HF/HFex groups (D, H). *p<0.05

K_m was calculated to quantify level of inhibition for dopamine reuptake into pre-synaptic terminals during AMPH exposure. In males, the only difference observed was between sedentary LF and HF groups. The HF group had significantly greater uptake inhibition (K_m) than the LF group (p=0.015) at the 60-minute time point with 3μM AMPH (Fig. 5A). Conversely, exercised males showed no differences regardless of diet (p=0.807) (Fig. 5B), suggesting that exercise normalized this effect comparable to sedentary and exercised-LF groups. Likewise, there were no differences within diet group between exercise comparisons for males, LF vs. LFex (Fig. 5B) and HF vs. HFex (Fig. 5D).

Figure 5. Measurement of K_m in presence of AMPH.

Changes in K_m for males (A-D) and females (E-H) with increasing amphetamine (AMPH) doses (300nM, 3μM) over a 60-minute timeframe. 300nM AMPH was applied over first 30 minutes (light grey panels) and 3μM AMPH during second 30-minute timeframe (darker grey panels). Comparisons are shown between groups: LF/HF (A, E); LFex/HFex (B, F); LF/LFex (C, G); HF/HFex (D, H). *p<0.05

In females, there were no differences when comparing diet in sedentary groups (Fig. 5E). However, a significant decrease in K_m at the 60-minute timepoint with 3μM AMPH was observed in the LFex group compared to HFex (p=0.018) (Fig. 5H). Further comparing differences in K_m within diet groups discerned a significantly higher K_m in HFex females than the sedentary HF group (p=0.006) (Fig. 5G). However, no difference between LF-fed females was detected when comparing sedentary and exercised groups (Fig. 5F).

Diet and exercise did not influence gene expression or protein levels of pro-inflammatory biomarkers in male mice.

Gene expression in the NAc was quantified for BDNF (Fig. 5A), TH (5B), DAT (5C), Iba1:Cd11 (5E), IL-6 (5F), IL-1β (5G), and TNFα (5H) all normalized to a housekeeping gene, 18s. Plasma levels of C-reactive protein were measured and normalized to the sedentary LF group (Fig. 5D). No significant differences were detected in male mice for CRP or any gene expression measurements (Figure 5). However, sedentary HF-fed males had a globally enhanced pro-inflammatory gene expression profile compared to sedentary LF males, although statistical significance was not met. A noteworthy increase in the ratio of Iba1 to Cd11 was observed for HF males (949.2 ± 338.9) compared to LF males (46.84 ± 20.11) (Fig. 5E), mirrored trends in pro-inflammatory gene expression, though not significant (p=0.085).

HF-fed female mice exhibited increased protein levels of pro-inflammatory biomarker CRP with or without exercise.

Gene expression in the NAc and plasma CRP protein levels were quantified in females (Figure 7). No differences in gene expression were detected between female groups (Fig. 7A–C; E–H). However, HFex females trended towards increased BDNF levels (3.340 ± 1.66) compared to the HF group (0.793 ± 0.59); although this was not significant (p=0.109) (Fig. 7A). Conversely, the level of C-reactive protein (CRP) was significantly elevated in both sedentary and exercised HF-females in comparison to the sedentary LF female group (p=0.017 and p=0.0008, respectively). However, exercise did not influence this biomarker within diet groups (sedentary vs exercised LF- or HF-diet groups) (Fig. 4D).

Figure 7. Plasma CRP and NAc gene expression data for female groups.

Data are normalized to sedentary low-fat (LF) group values. 18s was used as the housekeeping gene to evaluate gene expression for BDNF (A), Tyrosine Hydroxylase-TH (B), Dopamine transporter-DAT (C), ratio of Iba1:Cd11 (E), IL-6 (F), IL-1β (G), and TNFα (H). C-reactive protein levels are expressed as % of LF group.

Discussion

We examined whether moderate-intensity exercise influenced dopamine terminals by attenuating diet-induced inflammation. Exercise caused a modest reduction in weight gain for HF-fed mice with minimal impact on blood glucose regulation. There was a trend toward increased pro-inflammatory gene expression (IL-6, IL-1β and TNFα) in sedentary HF males and females that was not observed in the HFex groups. Overall, our findings did not support our initial hypotheses largely because we did not observe deficits in dopamine terminal function previously reported after HF feeding,[2,55] nor did we find significantly elevated pro-inflammatory gene expression in HF groups. However, we identified sex-specific effects of exercise on dopamine kinetics and dopamine terminal response to AMPH. Specifically, exercise significantly increased the rate of dopamine reuptake in female mice fed a HF diet. This corresponded with increased potency of AMPH as an uptake inhibitor (Km) in exercised HF-fed female mice. Elevated V_max and increased AMPH potency demonstrate enhanced DAT function in exercised females, while male HF mice were more susceptible to AMPH-induced dopamine depletion. We also observed sex-specific effects on inflammatory markers. The HF diet engendered a greater inflammatory gene expression profile in the NAc of male mice, but systemic inflammation measured by CRP was significantly elevated in HF-fed females. The dissimilarity between male and female inflammatory profile suggests that low-grade inflammation caused by the HF diet (males) and increased systemic inflammation caused by exercise (females) are likely uncoupled from changes in baseline dopamine terminal kinetics caused by diet and exercise.

The exercise protocol employed was adapted from Chen et al (2017)[11] to attenuate body weight gain in the HF-fed groups. Given a modest non-significant attenuation of body weight in the HF groups our exercise prescription may not have been long enough in duration or at an intensity high enough to overcome the excess calories consumed by the HF-fed groups. Despite modestly reduced body weight, exercise did increase dopamine release and reuptake rate in female mice fed a HF-diet. Other pre-clinical studies reported attenuated weight gain with higher intensity exercise [57] or longer duration protocol.[21] Our next step will be to explore stepwise increases in exercise intensity and/or duration to better attenuate weight gain in HF groups. Enhanced exercise intensity may also translate into a larger effect on fasting blood glucose and glucose clearance rate. This underscores the importance of intensity and length of protocol for improving metabolic parameters that develop alongside metabolic syndrome. Had our study extended the exercise protocol until marked improvements in blood glucose were consistently observed, we may have observed a larger effect on body weight and dopamine signaling endpoints.

Another notable finding in exercised females was elevated BDNF gene expression levels, a well-categorized factor involved in neurogenesis[9,62,64] and neurotransmitter modulation.[34,51,52] BDNF could facilitate terminal changes to support enhanced dopamine release and V_max reported in HFex females. Kim et al. (2020) observed no differences in BDNF gene expression between sedentary and exercised mice.[23] The Kim et al. study investigated the coding region of the BDNF gene, like the BDNF probe used in our study. However, we observed elevated BDNF gene expression in the female HFex group, similar to elevated BDNF protein levels observed in high-intensity groups compared to controls reported by Kim et al.[23] There were no differences in moderate intensity exercise groups akin to the protocol used in our study. Interestingly, BDNF was also elevated in male HF sedentary mice. Similar to enhanced release and V_max in the HF female exercise group, the absence of slowed V_max and attenuated phasic dopamine release in the HF males, as reported previously[2,55], could be attributed to elevated BDNF levels. A previous report in Swiss mice fed a HF diet reported an early increase in BDNF followed by a reduction that coincided with an increased inflammatory response.[39] It is possible the male HF-fed mice benefited from the protective effects at the point we measured dopamine, preceding a greater inflammatory decline. Sedentary females fed a HF diet also exhibited a muted rate of dopamine reuptake (V_max) in comparison to LF counterparts. This attenuated reuptake rate was rescued with moderate-intensity aerobic exercise. The consumption of our HF diet is well-documented to influence dopamine neurotransmission through decreasing dopamine receptors[19], and attenuating dopamine release over long-term conditions.[2,45,46,53] Exercise improved the diminished reuptake rate effect in female mice consuming high amounts of saturated fat. Although male mice did trend toward a decrease in evoked phasic release there was not a significant effect for either diet or exercise observed. One explanation could be an issue of timing and differing methylation patterns on the DAT encoding gene. Increased methylation of the DAT promoter was observed in mice fed a high fat diet, which was reversed when HFD was removed.[10]

To test pre-synaptic dopamine terminal function with differing levels of exercise, AMPH was bathed over slices during voltammetry at a concentration of 300nM and 3μM for a duration of 30 minutes each. Male HF-fed mice displayed greater depletion of dopamine from NAc terminals compared to LF counterparts. This could be due to a priming of the pre-synaptic terminals to exacerbate the effects of a dopamine disrupting compound like AMPH. Priming dopamine terminals to respond to exogenous compounds has been previously reported following HF feeding. For example, bathing IL-6 and TNFα during FSCV experiments in the NAc selectively reduced dopamine release in HF-fed mice compared to minimal cytokine effects on LF-fed mice.[15] Together, these findings suggest that a HF diet may prime dopamine terminals in sedentary male mice for an exaggerated response to exogenous insults like recreational drugs or endogenous insults caused by inflammation. The potency of AMPH to inhibit DAT, as measured by apparent K_m, was increased in HF-sedentary males. Conversely, exercised increased the potency of AMPH in the HF-exercised females compared to their HF-sedentary counterparts. This suggests that there might be a sex and diet interaction affecting AMPH’s actions, aligning with previous microdialysis studies showing differences in AMPH-stimulated dopamine release based on estrous cycle.[7] Additionally, dopamine release is altered by estrous cycle stage in rodents, due to estradiol having differing effects on VTA neuron excitability. Specifically, estradiol increases excitability during metestrus stage, while decreasing excitability in proestrus and estrus stages.[43] This further points to the need to investigate these effects under different stages of the estrus cycle in female mice in a model of diet-induced obesity.

The greatest impairment of fasting blood glucose and blood glucose clearance was in sedentary HF males, indicating the greatest metabolic impairment. Exercise was not able to attenuate these outcomes, but exercise did attenuate inflammatory gene expression between HF and HFex males. Since exercised females had increased plasma CRP protein levels, elevated NAc BDNF gene expression, and enhanced V_max but sedentary males displayed a greater inflammatory gene expression with no effect on CRP, BDNF, or V_max there is likely little influence of diet-induced inflammation on baseline dopamine terminal kinetics caused by diet and exercise. It is possible that the chronic low-grade inflammation caused by HF-feeding may alter how dopamine terminals respond to exogenous compounds like dopamine agonists or acute inflammatory insults.

This study was limited in that we examined gene expression in the brain and blood CRP following DIO that engenders low-grade inflammation which can be labile in nature. Exposure to a pro-inflammatory stimulus, like LPS, would provide insight into how the HF diet impacts a known inflammatory response. Adding this positive control would provide a better understanding of inflammation’s direct role on dopamine kinetics under our specific experimental parameters (diet, exercise). The temporal nature of cytokine release from diet-induced inflammation is difficult to capture, but a timed pro-inflammatory challenge could further illuminate any priming chronic low-grade inflammation has on the central immune system response. In our study, the ratio of activated microglia to total microglia (Iba1:Cd11b) was markedly increased in sedentary HF-fed males, while in females the only observed difference was a drastic reduction in LFex females. The increase observed in sedentary HF-males points to the pro-inflammatory effect of DIO, specifically caused from high saturated fat intake. This likely results from saturated fat binding TLR4, initiating activation of NF-kB pathways, ultimately causing release of pro-inflammatory cytokines and inducing microglia to an activated state. Sedentary HF-male mice displayed increased markers of activated microglia (Iba1:Cd11b), suggesting a pro-inflammatory effect of DIO due to high saturated fat intake. Saturated fats have been shown to bind TLR4, activating microglial transcription and release of pro-inflammatory cytokines via the NF-κB pathway. In contrast, exercise had no effects on attenuating gene expression of inflammation-related markers. However, this might be due to temporal differences that were not captured. The initial increase in inflammation within the hypothalamus has been shown to occur after only one day on an HFD[54], while gene expression is initially elevated, then diminished over time.[33,63] There may be an adaptive mechanism regulating this transient release of cytokines to prevent from large accumulations over time. Studies have shown increased Iba1 with chronic long-term feeding, but these experiments fed mice for 5 months.[22]

Conclusions

A 6-week moderate-intensity aerobic exercise protocol did not attenuate body weight gain when consuming a diet high in saturated fat. Dopamine reuptake rate was enhanced in exercised females, suggesting a potential sex-linked effect. This was seen without overall metabolic improvements suggesting a longer protocol or a protocol with higher intensity may be needed to observe a greater impact on neurochemistry. Our gene expression markers of neuroinflammation in the NAc did not show an overt pro-inflammatory profile, but gene expression indicating greater microglial activation in the HF groups suggest the inflammatory response may be raised with HF feeding. However, the pro-inflammatory profile measured in our study had little influence on dopamine neurochemistry. Moderate treadmill exercise independently enhanced the rate of dopamine reuptake in female mice, despite little change in inflammatory gene expression. Future exercise implementation might start at this intensity and duration to initiate neurochemical changes, gradually working up to a high-intensity protocol with longer exercise bouts. Additionally, there is a need to explore intensity and duration combinations to optimize the most sustainable improvements, potentially informing obesity treatments while minimizing use of pharmaceuticals that carry a high potential risk of abuse.

Figure 6. Plasma CRP and NAc gene expression data for male groups.

Data are normalized to sedentary low-fat (LF) group values. 18s was used as the housekeeping gene to evaluate gene expression for BDNF (A), Tyrosine Hydroxylase-TH (B), Dopamine transporter-DAT (C), ratio of Iba1:Cd11 (E), IL-6 (F), IL-1β (G), and TNFα (H). C-reactive protein levels are expressed as % of LF group.

Highlights.

• Moderate-intensity exercise did not prevent significant body weight gain in mice.

• Female mice experienced enhanced dopamine signaling with aerobic exercise.

• Male high-fat fed mice did not show enhanced endpoints with aerobic exercise.

Glossary

V_max

Rate of dopamine reuptake from synapse back into pre-synaptic terminals. Measured as nM/s.

K_m

Measurement of affinity of dopamine for its transporter, (DAT).

List of Abbreviations

AMPH

amphetamine

ANOVA

analysis of variance

AUC

area under the curve

BDNF

brain-derived neurotrophic factor

Cd11

cluster of differentiation

CRP

C-reactive protein

DA

dopamine

DAT

dopamine transporter

DIO

diet-induced obesity

FSCV

fast-scan cyclic voltammetry

LF

low-fat diet (10% total kcals from fat)/ sedentary LF groups

LFex

low-fat diet fed, exercised groups

LPS

lipopolysaccharide

HF

high-fat (60% total kcals from fat)/ sedentary HF groups

HFex

high-fat diet fed, exercised groups

Iba1

ionized calcium binding adaptor molecule 1

IL-1β

interleukin-1 beta

IL-6

interleukin-6

IPGTT

intraperitoneal glucose tolerance test

NAc

nucleus accumbens

SEM

standard error of the mean

TH

tyrosine hydroxylase

TNFα

tumor necrosis factor alpha