Differences in Ethanol Drinking and Nucleus Accumbens Opioid Expression in Male and Female Obesity-Prone and Obesity-Resistant Rats

Abstract

Literature suggests that there may be overlap between proneness to overeating-induced obesity and excessive alcohol drinking, and that this may occur in part via the endogenous opioid system in the nucleus accumbens (NAc). To investigate this, we used rats selectively bred for their susceptibility or resistance to diet-induced obesity (obesity-prone, OP, and obesity-resistant, OR), and examined their (1) ethanol drinking under the intermittent access two-bottle-choice procedure, (2) sucrose drinking under the intermittent access two-bottle-choice procedure, and (3) enkephalin, dynorphin, and opioid receptor mRNA in the NAc core and NAc shell. While we found no major differences in measures of sucrose drinking, male OP compared to OR rats had higher levels of 24-hour ethanol drinking and preference, while female OP compared to OR rats had higher levels of binge-like (30-minute) ethanol drinking and corresponding blood ethanol concentrations. In parallel, both male and female OP compared to OR rats had higher expression of enkephalin but not dynorphin mRNA in the NAc core, and males had higher expression of mRNAs encoding mu, delta, and kappa opioid receptors in the NAc core. Together, these results show that proneness to obesity is associated with excessive ethanol drinking prior to the onset of obesity. Further, they suggest that this could be driven in part through differences in the NAc endogenous opioid system, particularly for males and via the NAc core. This supports the idea that proneness to excessive alcohol drinking may share neurobehavioral mechanisms with susceptibility to diet-induced obesity.

Introduction

Published literature suggests that there may be both genetic and neurobiological overlap between proneness to obesity and excessive alcohol consumption. Genetically, individuals with a family history of alcohol use disorder have been found to have a higher body mass index and increased risk of developing obesity [1, 2]. Similarly, individuals with genetic variants of the fat mass and obesity-associated gene (FTO) have been found to have increased risk of alcohol dependence [3, 4]. Some FTO variants, however, have been found instead to reduce the odds of problem drinking, albeit in men and not women [5]. While both alcohol use disorder and obesity are largely polygenic [6, 7], these findings suggest that there may be common genetic risk factors for the development of obesity and the development of excessive alcohol consumption.

Neurobiologically, one point of overlap between proneness to overeating-induced obesity and excessive alcohol drinking appears to be the endogenous opioid system. Notably, the opioid receptor antagonist, naltrexone, is a U.S. Food and Drug Administration-approved medication used to treat alcohol use disorder that not only reduces binge drinking and the number of alcohol drinking days [8, 9], but can also reduce the frequency of binge eating [10] and, when combined with the antidepressant bupropion, can reduce body weight in people with obesity in the absence of binge eating [11]. Some of the highest density of endogenous opioid expression in the brain is in the nucleus accumbens (NAc) core and shell regions [12, 13]. Within the NAc, the classical endogenous opioids, the enkephalin and dynorphin peptides, are largely found in separate cells [14]. They also differ in their affinity for distinct classical opioid receptors, with the enkephalin peptides primarily binding to the mu opioid receptor (MOR) and delta opioid receptor (DOR), and the dynorphin peptides primarily binding to the kappa opioid receptor (KOR) [15, 16]. It should be noted that other opioid peptides and their receptors, such as the nociceptin/orphanin FQ (N/OFQ) peptide system, have also been identified in the NAc [17]. There is somoatodendritic release of the endogenous opioids from principle (medium spiny) neurons of the NAc [18, 19], but these peptides are also released by afferent inputs to the NAc [20–22]. In relation to food and alcohol, striatal enkephalin levels rise in anticipation of and during a meal [23], and activation of opioid receptors in the NAc (particularly the MOR) enhances food intake, while NAc opioid receptor blockade with naloxone reduces food intake, even in hungry rats [24, 25]. Similarly, activation of MOR, DOR, and KOR receptors in the NAc, and opioid receptor blockade in the NAc, regulates ethanol drinking in rats [26–28]. Thus, the endogenous NAc opioid system is poised as a hub to regulate consummatory behaviors.

Evidence suggests that animals prone to excessive alcohol or dietary fat consumption may exhibit differences in their endogenous NAc opioid system. For example, elevations in enkephalin levels and MOR binding, and reductions in dynorphin levels and KOR binding, in the NAc have been found in inbred and selectively-bred alcohol-preferring compared to alcohol-avoiding rats and mice [29–33]. Similarly, elevations in NAc enkephalin gene expression (proenkephalin) have been found in rats identified as prone to diet-induced obesity compared to controls [34]. These studies support the idea that proneness to ethanol consumption and overeating-induced obesity may share mechanistic links through the NAc endogenous opioid system. However, interpretation of the study by Chang et al. [34] was complicated by the need to allow rats to consume high-fat diet in order to identify them as prone to overconsumption, which could itself alter opioid systems [35]. In addition, relatively few studies have differentiated between the NAc core and shell, the majority used only males, and none has assessed ethanol intake in a model that captures proneness to obesity.

Rat models can be employed to identify links between proneness to obesity and excessive alcohol consumption. For ethanol drinking, one commonly-employed method is the intermittent access two-bottle-choice procedure [36, 37]. In this procedure, rats are offered 10 - 20% ethanol for 24 hours on Monday, Wednesday, and Friday of each week, with water and chow available ad libitum [36–38]. This procedure leads to binge-like ethanol drinking, with the highest intake occurring during the first 30 minutes of daily access, and resulting blood ethanol levels in rats averaging about 60 mg/dl, which is considered pharmacologically-relevant [36, 39, 40]. With this procedure, females typically consume more ethanol than males [40–43], although this can vary depending on the genetic background [44]. A common control for the reinforcement of ethanol is a low-concentration sucrose solution which, although it is not matched for caloric content with the ethanol, is often consumed at similar g/kg levels as 10 - 20% ethanol in the intermittent access procedure [28, 45]. For obesity, selective breeding of outbred rats has been used to identify those with a polygenic predisposition to the development of diet-induced obesity (obesity-prone, OP) and those predisposed to be resistant to this development (obesity-resistant, OR) [46, 47]. Levin and colleagues [46] developed OP and OR rat lines (originally called DIO and DR) by giving outbred Sprague-Dawley rats a sweet-fat high-energy diet for one month, identifying the highest and lowest weight gainers, and then breeding them together. After five generations with this approach, the OP compared to OR rats showed somewhat greater weight gain and food intake while maintained on chow, but dramatically greater weight gain, food intake, and feed efficiency while maintained on a high energy diet [46]. Our (Ferrario) lab maintains colonies of these selectively bred OP and OR lines using the approaches developed by Levin, and has confirmed that they maintain their respective proneness and resistance to diet-induced obesity [47]. Interestingly, NAc neuronal function and motivational processes have been found to differ in OP compared to OR rats prior to any diet manipulation or the development of obesity [47–51], but ethanol consumption and the endogenous opioid system has not previously been examined in these rats.

The purpose of this study was to determine if OP rats would engage in higher levels of ethanol drinking and, prior to ethanol drinking, if they showed differences in their classical NAc opioid system compared to OR rats. Since sex-related differences in behavior and NAc opioid expression have previously been identified across other pairs of selectively-bred rat strains [52, 53], we also sought to determine if there were sex-related differences in these phenomena. We tested male and female OP and OR rats and examined their (1) ethanol drinking under the intermittent access two-bottle-choice procedure, (2) sucrose drinking under the intermittent access two-bottle-choice procedure, and (3) levels of enkephalin, dynorphin, MOR, DOR, and KOR mRNA in the NAc core and NAc shell. We hypothesized that OP compared to OR rats would show elevated drinking of ethanol but not low-concentration sucrose, that they would show differences in their endogenous opioid system, and that some of these differences would be sex-specific.

Materials and Methods

Animals

Selectively bred male and female OP and OR rats were used for all studies (N = 94; OP males n = 24, OP females n = 24, OR males n = 22, OR females n = 24). The original selection was conducted by Levin from outbred Sprague-Dawley rats [46, 54], and current OP and OR colonies are maintained by the University of Michigan (UM) breeding core (see [47, 49] for additional information). Founder OP and OR breeding pairs from UM were shipped to Drexel University, and rats bred in-house at either Drexel or UM were used for the studies. All rats were group-housed in an AAALAC-accredited facility, on a 12-hour reversed light/dark cycle (lights off at 0900 h at Drexel and 0700 h at UM) in a climate-controlled colony room from weaning until 7 weeks old. They were then individually housed for one week prior to the start of experiments. Animals received ad libitum chow (Drexel: Laboratory Rodent Diet 5001, Lab Diet, St. Louis, MO, USA, 3.35 kcal/g; UM: 5LOD - PicoLab^® Laboratory Rodent Diet, Lab Diet; 3.35 kcal/g) and water throughout the study. Experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Drexel University College of Medicine and the IACUC of University of Michigan and followed the NIH Guide for the Care and Use of Laboratory Animals [55]. Estrous cycle was not assessed.

Ethanol or Sucrose Drinking

Under the intermittent-access two-bottle-choice procedure adapted from Wise [37] and Simms [36], rats were given access to unsweetened 20% v/v ethanol (1.4 kcal/g) or 2.5% w/v sucrose (0.1 kcal/g) during three 24-hour-sessions per week in addition to ad libitum water and chow as described [39, 40, 56]. This concentration of sucrose was selected as it has been found to be consumed at similar g/kg levels as 20% ethanol, despite not matching ethanol for its caloric content [28, 45]. Briefly, each Monday, Wednesday, and Friday, one-to-two hours after dark onset, one of their two bottles of water was replaced with a bottle of ethanol or sucrose and, after 24 hours, the ethanol or sucrose bottle was replaced with the second bottle of water. Relative bottle position was alternated each time to prevent side preference, and all bottles were fitted with stainless steel ball-bearing non-drip sipper tubes of equal size. Ethanol intake was calculated as: (weight ethanol solution consumed (g) * (density ethanol * 0.200)) / rat body weight (kg). Sucrose intake was calculated as: (weight sucrose solution consumed (g) * (density sucrose * 0.025)) / rat body weight (kg). Ethanol and sucrose preference was calculated as: volume ethanol or sucrose solution consumed (ml) / (volume ethanol or sucrose solution consumed (ml) + volume water consumed (ml)). To account for spillage, empty cages with ethanol (or sucrose) and water bottles were weighed on testing days, and that number in grams was subtracted from the weight of liquid consumed for the calculation of liquid consumption. Food intake was measured for one ethanol or sucrose access day and one non-access day each week, by weighing the food (g), at the same time as the bottles, on Thursdays, Fridays, and Saturdays. Animals were weighed on Tuesdays and Fridays (g).

Quantitative Real-Time PCR

Rats were euthanized by an experienced investigator via rapid decapitation without anesthesia one-to-two hours after dark onset, and their NAc core and shell were dissected out to examine mRNA levels of endogenous opioids and their receptors. Immediately after the decapitation, the brain was placed in a matrix slicing guide on ice, with the ventral surface facing up. Three coronal cuts were made, starting with the rostral optic chiasm (approx. Bregma −1.0 mm) [57]. The second and third cuts were 2.0 and then 1.0 mm rostral to this, yielding a slice (approx. Bregma 2.0 to 1.0 mm) for microdissection of the NAc core and shell. Then, under a microscope, on a slide placed on a petri dish filled with ice, the NAc core and shell were bilaterally dissected. The NAc core was dissected in a teardrop shape around the anterior part of the anterior commissure, lateral to the lateral ventricle and medial to the lateral stripe of the striatum. The NAc shell was dissected in a crescent moon shape, with the dorsal tip beginning at the lateral ventricle, the medial aspect at the semilunar nucleus, and the ventral edge along the ventral pallidum. These sections were then placed in RNAlater (Qiagen Inc., Valenia, CA, USA) and stored at −20 °C.

To extract RNA, total RNA from each brain section was purified using an RNeasy Mini Kit (Qiagen Inc.), and DNA was removed with RNase-free DNase I (Qiagen). The yield was quantified with a NanoDrop Lite spectrophotometer (Thermo Electron North America LLC, Madison WI), with resulting A₂₆₀/A₂₈₀ ratios between 1.96 and 2.12, indicating high purity. To generate cDNA, cDNA was reverse transcribed using SuperScript^® VILO^™ Master Mix (Invitrogen, Grand Island, NY, USA) in a SimpliAmp^™ Thermal Cycler (Applied Biosystems, Waltham, MA, USA), using 1 μg of RNA from each sample. Negative RT controls were synthesized by denaturing the reverse transcriptase. For quantitative real-time PCR, the SYBR Green PCR core reagents kit (Applied Biosystems, Grand Island, NY, USA) was used in MicroAmp^® Fast Optical 96-Well Reaction Plates (Applied Biosystems), with 12.5 ng of cDNA template in a 25 μl reaction volume. The reaction was carried out on a StepOnePlus Real-Time PCR System (Applied Biosystems), under the conditions of 2 min at 50 °C (primer annealing), 10 min at 95 °C (polymerase activation and sequence extension), and 40 cycles of 15 s at 95 °C (denaturation) plus 1 min at 60 °C (annealing and extension). Each sample was run in triplicate, and each run included a no-template and negative RT control. Target gene expression was quantified relative to cyclophilin-A (CYC) as the reference gene, based on prior findings in the brains of ethanol-exposed Sprague-Dawley rats that this gene is more reliable than β- actin or GAPDH, with no region or treatment specific changes in quantity [58]. One-way ANOVAs (with group or sex as the between-subject factor) of CYC values for each run in this experiment also revealed no significant differences (NAc core: group (F(1, 28) = 0.10, p = 0.758) and (F(1, 28) = 0.01, p = 0.912), sex (F(1, 28) = 1.02, p = 0.321) and (F(1, 28) = 0.40, p = 0.532); NAc shell: group (F(1, 28) = 0.33, p = 0.570) and (F(1, 27) = 1.09, p = 0.306), sex (F(1, 28) = 0.58, p = 0.453) and (F(1, 27) = 3.20, p = 0.085). The relative quantification method (2^{-(CT(target gene)-CT(CYC))}) was used for statistical analysis, with corrections applied when statistical assumptions were violated (see Statistical Analysis). Primers were designed with the NCBI Primer design tool (http://www.ncbi.nlm.nih.gov/tools/primer-blast/) [59], and purchased from Invitrogen at ThermoFisher Scientific (Grand Island, NY, USA). Amplification efficiency of the primers was determined by serial dilutions of cDNA template, with resulting slopes of the threshold cycle versus the dilution factor (using a base-10 semi-logarithmic graph) between −3.47 and −3.57, indicating high amplification efficiency. The specificities of the PCR products were confirmed by visualization of a single peak from the melt curve analysis and, for enkephalin and dynorphin, a single band with a corresponding molecular weight revealed by agarose gel electrophoresis. Primer sequences and concentrations are listed in Table 1.

Table 1.

Primer sequences and concentrations used for quantitative real-time polymerase chain reaction in Experiment 3.

Primers used for quantitative real-time polymerase chain reaction

Primer	Sequence	Concentration
Cyclophilin-A	5′-GTGTTCTTCGACATCACGGCT-3′ (forward) 5′-CTGTCTTTGGAACTTTGTCTGCA-3′ (reverse)	200 nM
Enkephalin	5′-GGACTGCGCTAAATGCAGCTA-3′ (forward) 5′-GTGTGCATGCCAGGAAGTTG-3′ (reverse)	100 nM
Dynorphin	5′-CAGCGGACTGCCTGTCCTT-3′ (forward) 5′-TCAGGGTGAGAAAAGACCAAAAG-3′ (reverse)	100 nM
Mu opioid receptor	5′-AAAATACAGGCAGGGGTCCA-3′ (forward) 5′-AGAAAGCACATACCTGGTGGTT-3′ (reverse)	200 nM
Delta opioid receptor	5′-CCCCGGGATGGAGCAGTGGTATG-3′ (forward) 5′-GGGCACCACGAAGGCGAAGAG-3′ (reverse)	100 nM
Kappa opioid receptor	5′-CCAGTCTTGGAAGGCACAAGT-3′ (forward) 5′-CGACCTCCCTTCCCAAATCAG-3′ (reverse)	100 nM

Experimental Protocols

Experiment 1:

To determine if there are differences between obesity-prone groups and/or sexes in ethanol drinking, male and female OP and OR rats at Drexel (N = 32, n = 8/group), starting at 8 weeks of age, were given 20% ethanol for 6 complete weeks using the intermittent-access two-bottle-choice procedure. The final access session was 30 minutes long and occurred at the start of week 7. Rats were run in two separate cohorts (N = 16/cohort), with 4 animals per group (male, female OP and male, female OR) represented in each cohort. Ethanol and water intake were measured as 24-hour intake for each ethanol access day, with 30-minute intake measured on the final access day. Food intake was measured as 24-hour intake for one ethanol access day and one ethanol non-access day each week. To determine blood ethanol concentration (BEC), trunk blood was obtained immediately after the final 30-minute intake period, and plasma was isolated (10,000 x g, 10 min 4°C) and frozen for later analysis using an Analox AM1 Alcohol Analyzer (Lunenburg, MA, USA).

Experiment 2:

To determine if there are differences between obesity-prone groups and/or sexes in the drinking of another reinforcing substance, sucrose, male and female OP and OR rats at UM (N = 32, n = 8/group), starting at 8 weeks of age, were given 2.5% sucrose for 6 complete weeks using the intermittent-access two-bottle-choice procedure. The final access session was 30 minutes long and occurred at the start of week 7. Rats were run in two separate cohorts (N = 14 - 18/cohort), with 3 - 5 animals per group (male, female OP and male, female OR) represented in each cohort. Sucrose and water intake were measured as 24-hour intake for each sucrose access day, with 30-minute intake measured on the final access day. Food intake was measured as 24-hour intake for one sucrose access day and one sucrose non-access day each week. To determine blood glucose concentration, a small drop of tail capillary blood was collected via tail nick after the final 30-minute intake period and analyzed with a portable glucometer (Contour Next, Ascensia Diabetes).

Experiment 3:

To determine if there are differences between obesity-prone groups and/or sexes in gene expression in the NAc of the endogenous opioid system, a third set of male and female OP and OR rats at Drexel (N = 30, OP males n = 8, OP females n = 8, OR males n = 6, OR females n = 8) was sacrificed at 8 weeks of age (i.e., at an age corresponding to the time of first ethanol or sucrose exposure in Experiments 1 and 2), and expression of enkephalin, dynorphin, MOR, DOR, and KOR mRNA was examined in the NAc core and shell using quantitative real-time PCR. Rats were run in two separate cohorts (N = 14 - 16/cohort), with up to 4 animals per group represented in each cohort.

Statistical Analysis

Twenty-four-hour ethanol/sucrose intake and preference in Experiments 1 and 2 was averaged across the three daily measurements, to obtain a single average 24-hour intake measurement for each week. Body weight in Experiments 1 and 2 was averaged across the two weekly measurements, to obtain a single body weight measurement for each week. Food intake in Experiments 1 and 2 was averaged across weeks, to compare food intake during periods of ethanol/sucrose availability and food intake when it was not available. Data were removed from analysis when there was spillage (Experiments 1 and 2) or Ct values were undetermined (presumably due to poor cDNA quality or pipetting error) (Experiment 3). For all relevant statistical analyses, sphericity of the data was determined using Mauchly’s test, homogeneity of variance was determined using Levene’s test, and normality was determined with a Shapiro-Wilk test. A Greenhouse-Geisser correction or Welch’s test was used when was assumptions were violated. Linear regression was used to determine the relationship between ethanol intake and BEC and between sucrose intake and glucose concentration. Due to the expectation that there would be sex-related differences in the variables of interest [52, 53], an a priori decision was made to directly compare primary measures within each sex. Paired-samples, two-tailed t-tests were used to compare food intake during periods of ethanol or sucrose availability with food intake when ethanol or sucrose was not available. All other data were first analyzed for effects of sex (male vs. female) with a mixed ANOVA (between-subject factor was sex, and within-subject factors were weekly intake/preference, body weight, or food intake in Experiments 1 and 2) and were then analyzed for effects of obesity-prone group (OP vs. OR) within each sex by mixed ANOVA or they were analyzed by one-way ANOVA (all data in Experiment 3 and 30-minute intake and blood levels in Experiments 1 and 2), with sex as the between-subject factor, and then by independent-samples two-tailed t-test for effect of obesity-prone group within each sex. Interaction effects between the primary variables were not significant, possibly due to a low statistical power for these analyses, and are therefore not reported here. Significant main effects were followed by Sidak’s pairwise comparison tests when appropriate. Significance was set at p < 0.05, although we acknowledge the guidelines set forth by the American Statistical Association [60], which hold that scientific conclusions should not be based solely on whether a p value passes a specific threshold, but should consider many contextual factors to derive scientific inferences. All data were analyzed using SPSS (Version 30, IBM, Armonk, NY, USA) and all graphical figures were made using GraphPad Prism (Version 10, Boston, MA, USA). All p values are reported with up to three significant figures. Data are reported as mean ± standard error of the mean (S.E.M.).

Results

Experiment 1: Ethanol drinking in OP and OR male and female rats

To determine if there are differences between obesity-prone groups or sex-specific effects on ethanol drinking, adult male and female OP and OR rats (N = 32, n = 8/group) were given 20% (v/v) ethanol for 6 full weeks using the intermittent-access two-bottle-choice procedure. Examining 24-hour ethanol intake with both sexes, there was a significant main effect of sex (F(1, 24) = 9.64, p = 0.005) (Figure 1A), with females drinking significantly more ethanol per body weight than males. Examining 24-hour ethanol intake among only males, there was a significant main effect of obesity-prone group (F(1, 12) = 7.33, p = 0.019) and no significant main effect of week (F(5, 60) = 1.64, p = 0.164), with OP males drinking significantly more ethanol than OR males. In contrast, examining 24-hour ethanol intake among females, there was no significant main effect of group (F(1, 10) = 0.04, p = 0.855) or week (F(5, 50) = 0.68, p = 0.638). For 24-hour ethanol preference, there was no significant main effect of sex (F(1, 21) = 0.004, p = 0.949) (Figure 1B); however, examining ethanol preference in males, there was a significant main effect of both group (F(1, 9) = 7.11, p = 0.026) and week (F(5, 45) = 5.39, p < 0.001). Male OP rats exhibited higher ethanol preference than male OR rats. Sidak’s pairwise comparisons revealed that ethanol preference in week 1 was significantly lower than preference in week 2 (p = 0.016). In females, there was no significant main effect of group (F(1, 10) = 0.14, p = 0.716) but there was a significant main effect of week (F(5, 50) = 2.51, p = 0.042), and Sidak’s pairwise comparisons revealed that ethanol preference in week 1 was significantly lower than preference in week 5 (p = 0.002). Thus, female rats showed higher 24-hour ethanol intake than male rats, but OP male rats showed higher 24-hour ethanol intake and preference than OR males.

Figure 1.

Experiment 1: Ethanol drinking in obesity-prone (OP) and obesity-resistant (OR) male and female rats. A. Average 24-hour ethanol intake per week. B. Average 24-hour ethanol preference per week. C. Ethanol intake during the 30-minute ethanol access session. D. Blood ethanol concentration (BEC) after the 30-minute ethanol access session. E. Body weight during the six weeks of ethanol access. F. Average 24-hour food intake during periods of ethanol availability. G. Average 24-hour food intake during periods when ethanol was not available. Values are Mean ± S.E.M. ***p < 0.001, **p < 0.01, *p < 0.05.

Examining 30-minute, binge-like ethanol intake, there no significant main effect of sex (F(1, 29) = 0.20, p = 0.661) (Figure 1C); however, independent samples t-tests revealed that while OP males had no significant difference in intake than OR males (t(12) = −1.15, p = 0.271), OP females in fact drank significantly more than OR females (t(11.22) = −3.33, p = 0.007). Examining corresponding BECs, there was again no significant main effect of sex (F(1, 29) = 0.003, p = 0.956) (Figure 1D), but independent samples t-tests revealed that while OP males had no significant difference in BECs at the end of the 30-minute session from OR males (t(12) = −0.36, p = 0.724), OP females did have significantly higher BECs than OR females (t(8.11) = −3.39, p = 0.009). Linear regression revealed a significant model (R² = 0.74, F(1, 25) = 72.35, p < 0.001), where BECs were significantly predicted by 30-minute ethanol intake (β = 0.86, p < 0.001, not shown). Thus, in contrast to 24-hour ethanol intake and preference, OP female rather than male rats showed greater 30-minute ethanol intake and BECs compared to OR females.

For average weekly body weight across the 6 weeks of ethanol access, there was a significant main effect of sex (F(1, 30) = 187.90, p < 0.001) (Figure 1E), with males weighing significantly more than females. Examining average weekly body weight among males, there was no significant main effect of group (F(1, 14) = 0.56, p = 0.466) but there was a significant effect of week (F(1.19, 16.72) = 192.733, p < 0.001). Sidak’s pairwise comparisons revealed that body weight at every week was significantly different than every other week (p < 0.001). Examining average weekly body weight among females, there was a significant difference between groups (F(1, 14) = 5.60, p = 0.033), with OP females weighing more than OR females. There was also a significant effect of week (F(1.06, 14.82) = 11.20, p = 0.004), and Sidak’s pairwise comparisons revealed that body weight at week 1 was significantly lower than body weight at weeks 2, 5, and 6 (p = 0.019 - < 0.001), and body weight at week 3 and later was significantly lower than body weight at all subsequent weeks (p = 0.050 - < 0.001). Thus, at the time of testing, overall body weight did not significantly differ between OP and OR male rats, but OP females did weigh more than OR females.

Examining food intake during the 24-hour periods of ethanol availability, there was a significant main effect of sex (F(1, 30) = 35.83, p < 0.001) (Figure 1F), with the male rats overall eating more chow than the female rats. Examining food intake among males during this period, there was no significant main effect of group (F(1, 14) = 0.02, p = 0.900), but there was a significant main effect of week (F(5, 70) = 6.25, p < 0.001), with significant differences between intake during week 1 compared to week 2 (p = 0.009) and week 5 compared to week 6 (p = 0.008). Examining food intake among females during this period, there was a significant main effect of group (F(1, 14) = 5.10, p = 0.040), with OP females eating more than OR females, but there was no significant main effect of week (F(1.71, 23.95) = 2.00, p = 0.162). Examining food intake during the periods when ethanol was not available, there was again a significant main effect of sex (F(1, 30) = 81.32, p < 0.001) (Figure 1G), with the male rats overall eating more chow than the female rats. Examining food intake among males during this period, there was no significant main effect of group (F(1, 14) = 0.82, p = 0.381), but there was a significant main effect of week (F(5, 70) = 3.68, p = 0.005), with significant differences between intake during week 2 compared to week 6 (p = 0.010). Examining food intake among females during this period, there was there was similarly no significant main effect of group (F(1, 14) = 4.34, p = 0.056), but also no significant main effect of week (F(1.67, 23.34) = 2.29, p = 0.130). Comparing intake during periods of ethanol availability with periods when ethanol was not available, there were significant differences among males (t(15) = −3.43, p = 0.004) but not females (t(15) = 1.08, p = 0.296), with males eating fewer calories during periods of ethanol availability. Thus, female OP rats showed higher levels of 24-hour chow intake than female OR rats but only during periods of ethanol availability.

Experiment 2: Sucrose drinking in OP and OR male and female rats

To determine if there are differences between obesity-prone groups or sex-specific effects on drinking of another reinforcing substance, sucrose, adult male and female OP and OR rats (N = 32, n = 8/group) were given 2.5% w/v sucrose for 6 full weeks using the intermittent-access two-bottle-choice procedure. Examining 24-hour sucrose intake with both sexes, there was a significant main effect of sex (F(1, 30) = 16.91, p < 0.001) (Figure 2A), with females drinking significantly more 24-hour sucrose per body weight than males. Examining 24-hour sucrose intake among males, there was no significant main effect of group (F(1, 14) = 0.18, p = 0.678) but there was a significant main effect of week (F(2.93, 41.06) = 9.94, p < 0.001). Sidak’s pairwise comparisons revealed that sucrose intake in week 1 was significantly lower than intake in weeks 2 and 3 (p = 0.003 – 0.005) and intake in weeks 5 and 6 was significantly lower than in weeks 2 through 4 (p = 0.042 – 0.002). Examining 24-hour sucrose intake among females, there was similarly no significant main effect of group (F(1, 14) = 1.17, p = 0.298) but there was a significant main effect of week (F(3.11, 43.56) = 13.13, p < 0.001). This time, Sidak’s pairwise comparisons revealed that sucrose intake in week 1 was significantly lower than intake in all subsequent weeks (p = 0.044 – 0.003). For 24-hour sucrose preference, there was no significant main effect of sex (F(1, 30) = 0.081, p = 0.778) (Figure 2B). Examining 24-hour sucrose preference among males, there was no significant main effect of group (F(1, 14) = 0.00, p = 0.988) but there was a significant main effect of week (F(1.96, 27.43) = 7.05, p = 0.004). Sidak’s pairwise comparisons revealed that preference for sucrose in week 1 was lower than preference in weeks 3 through 6 (p = 0.029 = 0.16). Examining 24-hour sucrose preference among females, there was similarly no significant main effect of group (F(1, 14) = 4.19, p = 0.060) but there was also no significant main effect of week (F(1.57, 22.04) = 1.95, p = 0.172).

Figure 2.

Experiment 2: Sucrose drinking in obesity-prone (OP) and obesity-resistant (OR) male and female rats. A. Average 24-hour sucrose intake per week. B. Average 24-hour sucrose preference per week. Note that the Y-axis starts at 60%. C. Sucrose intake during the 30-minute sucrose access session. D. Blood glucose concentration after the 30-minute sucrose access session. E. Body weight during the six weeks of sucrose access. F. Average 24-hour food intake during periods of sucrose availability. G. Average 24-hour food intake during periods when sucrose was not available. Values are Mean ± S.E.M. ***p < 0.001.

Examining 30-minute, binge-like sucrose intake, there was a significant main effect of sex (F(1, 30) = 23.72, p < 0.001), with females drinking significantly more 30-minute sucrose per body weight than males (Figure 2C). Independent samples t-tests revealed that there was no significant difference in intake between OP and OR males (t(14) = −0.22, p = 0.832) or between OP and OR females (t(14) = −1.03, p = 0.320). Examining corresponding blood glucose concentrations, there was again a significant main effect of sex (F(1, 30) = 22.33, p < 0.001), with females having significantly lower glucose levels than males (Figure 2D). Independent samples t-tests revealed that there was no significant difference in glucose levels between OP and OR males (t(14) = −1.70, p = 0.111) or OP and OR females (t(14) = 0.16, p = 0.872). Linear regression did not reveal a significant model (R² = 0.06, F(1, 30) = 1.90, p = 0.178), and blood glucose levels were not significantly predicted by 30-minute sucrose intake (β = −0.24, p = 0.178, not shown). Thus, there were no differences between OP and OR rats in measures of sucrose intake or blood glucose concentration, but females overall drank more sucrose per body weight than males, and they exhibited lower blood glucose concentrations after the 30-minute intake session.

For average weekly body weight across the 6 weeks of sucrose access, there a significant main effect of sex (F(1, 30) = 240.04, p < 0.001) (Figure 2E), with males weighing significantly more than females. Examining average weekly body weight among males, there was no significant main effect of group (F(1, 14) = 0.88, p = 0.364) but there was a significant main effect of week (F(1.39, 19.48) = 1114.66, p < 0.001). Sidak’s pairwise comparisons revealed that body weight at every week was significantly different than every other week (p < 0.001). Examining average weekly body weight among females, there was similarly no significant main effect of group (F(1, 14) = 0.35, p = 0.566) but there was a significant main effect of week (F(1.98, 27.71) = 309.76, p < 0.001). Sidak’s pairwise comparisons again revealed that body weight at every week was significantly different than every other week (p = 0.012 - < 0.001). Thus, at the time of testing, there were no significant differences in body weight between OP and OR male rats or OP and OR female rats.

Examining food intake during the 24-hour periods of sucrose availability, there was a significant main effect of sex (F(1, 30) = 72.78, p < 0.001) (Figure 2F), with the male rats overall eating more chow than the female rats. Examining food intake among males, there was no significant main effect of group (F(1, 14) = 0.04, p = 0.854) or week (F(5, 40) = 0.47, p = 0.798). Examining food intake among females, there was similarly no significant main effect of group (F(1, 14) = 0.00, p = 0.990) or week (F(5, 40) = 0.56, p = 0.733). Examining food intake during the periods when sucrose was not available, there was again a significant main effect of sex (F(1, 30) = 123.48, p < 0.001) (Figure 2G), with the male rats overall eating more chow than the female rats. Examining food intake among males, there was no significant main effect of group (F(1, 14) = 0.62, p = 0.445) or week (F(3.13, 43.85) = 2.51, p = 0.069). Examining food intake among females, there was similarly no significant main effect of group (F(1, 14) = 0.73, p = 0.407) or week (F(5, 70) = 0.82, p = 0.537). Comparing intake during periods of sucrose availability with periods when sucrose was not available, there were no significant differences among males (t(15) = −1.24, p = 0.233) or females (t(15) = 1.17, p = 0.260). Thus, OP male and female rats showed no difference in 24-hour chow intake from OR male and female rats, respectively, in either condition of sucrose availability.

Experiment 3: Gene expression in the NAc core and shell of OP and OR male and female rats

To determine if there are differences between obesity-prone groups or sex-specific effects on gene expression of the NAc endogenous opioid system, male and female OP and OR rats (N = 30, OP males n = 8, OP females n = 8, OR males n = 6, OR females n = 8) were sacrificed at 8 weeks of age and their endogenous opioid system was assessed using quantitative real-time PCR. Examining gene expression in the NAc core, enkephalin mRNA showed no significant main effect of sex (F(1, 20) = 0.31, p = 0.583) (Figure 3A), but independent samples t-tests revealed that OP males had significantly higher expression of enkephalin in the NAc core than OR males (t(6.09) = −2.91, p = 0.026) and that OP females had significantly higher expression of enkephalin in the NAc core than OR females (t(9.99) = −2.43, p = 0.036). While dynorphin mRNA in the NAc core also showed no significant main effect of sex (F(1, 20) = 0.18, p = 0.674) (Figure 3B), independent samples t-tests revealed that OP males did not have significantly different expression of dynorphin in the NAc core than OR males (t(8) = −0.14, p = 0.896) and OP females did not have significantly different expression of dynorphin in the NAc core than OR females (t(10) = −0.82, p = 0.434). For the MOR in the NAc core, there was no significant main effect of sex (F(1, 26) = 0.01, p = 0.932) (Figure 3C), but independent samples t-tests revealed that OP males had significantly higher expression of MOR mRNA in the NAc core than OR males (t(9.95) = −2.30, p = 0.044) while OP females did not have significantly different expression of MOR in the NAc core than OR females (t(14) = −1.13, p = 0.277). Similarly, for the DOR in the NAc core, there was no significant main effect of sex (F(1, 27) = 2.51, p = 0.125) (Figure 3D), but independent samples t-tests revealed that OP males had significantly higher expression of DOR mRNA in the NAc core than OR males (t(11.84) = −5.12, p < 0.01) while OP females did not have significantly different expression of DOR in the NAc core than OR females (t(13) = −0.68, p = 0.508). For the KOR in the NAc core, there was also no significant main effect of sex (F(1, 27) = 1.76, p = 0.196) (Figure 3E), and independent samples t-tests revealed that OP males had significantly higher expression of KOR mRNA in the NAc core than OR males (t(11.98) = −2.47, p = 0.029) while OP females did not have significantly different expression of KOR in the NAc core than OR females (t(14) = −1.01, p = 0.329). Thus, while levels of enkephalin mRNA in the NAc core were higher in both male and female OP compared to OR rats, only male OP compared to OR rats showed higher levels of MOR, DOR, and KOR mRNA in this brain region.

Figure 3.

Gene expression in the nucleus accumbens (NAc) core and shell of obesity-prone (OP) and obesity-resistant (OR) male and female rats. A. Enkephalin mRNA in the NAc core. B. Dynorphin mRNA in the NAc core. C. Mu opioid receptor (MOR) mRNA in the NAc core. D. Delta opioid receptor (DOR) mRNA in the NAc core. E. Kappa opioid receptor (KOR) mRNA in the NAc core. F. Enkephalin mRNA in the NAc shell. G. Dynorphin mRNA in the NAc shell. H. MOR mRNA in the NAc shell. I. DOR mRNA in the NAc shell. J. KOR mRNA in the NAc shell. For visualization, data are normalized to the respective OR group within each sex. Values are Mean ± S.E.M. ***p < 0.001, *p < 0.05.

Examining gene expression in the NAc shell, enkephalin mRNA showed no significant main effect of sex (F(1, 26) = 2.06, p = 0.163) (Figure 3F), but independent samples t-tests revealed that OP males had significantly higher expression of enkephalin in the NAc shell than OR males (t(6.17) = −2.48, p = 0.047) while OP females did not have significantly different expression of enkephalin in the NAc shell than OR females (t(12) = −1.78, p = 0.100). As with dynorphin in the NAc core, dynorphin mRNA in the NAc shell showed no significant main effect of sex (F(1, 26) = 0.34, p = 0.567) (Figure 3G), and independent samples t-tests revealed that OP males did not have significantly different expression of dynorphin in the NAc shell than OR males (t(12) = −0.93, p = 0.372) and OP females did not have significantly different expression of dynorphin in the NAc shell than OR females (t(12) = −0.83, p = 0.422). For the MOR in the NAc shell, there was no significant main effect of sex (F(1, 27) = 1.41, p = 0.245) (Figure 3H), and independent samples t-tests revealed that OP males did not have significantly different expression of MOR mRNA in the NAc shell than OR males (t(12) = 0.30, p = 0.766) and OP females did not have significantly different expression of MOR in the NAc shell than OR females (t(13) = 0.25, p = 0.809). For the DOR in the NAc shell, there was also no significant main effect of sex (F(1, 27) = 1.64, p = 0.211) (Figure 3I); however, independent samples t-tests revealed that OP males did have significantly different expression of DOR mRNA in the NAc shell than OR males (t(11.92) = −2.50, p = 0.028) while OP females did not have significantly different expression of DOR in the NAc shell than OR females (t(13) = −1.34, p = 0.205). For the KOR in the NAc shell, there was no significant main effect of sex (F(1, 27) = 3.34, p = 0.079) (Figure 3J), and independent samples t-tests revealed that OP males did not have significantly different expression of KOR mRNA in the NAc shell than OR males (t(12) = −0.84, p = 0.420) and OP females did not have significantly different expression of KOR in the NAc shell than OR females (t(13) = −0.90, p = 0.386). Thus, levels of enkephalin and DOR mRNA in the NAc shell were higher in male OP compared to OR rats, just as in the NAc core but, unlike in the NAc core, there were no differences between OP and OR females or in other opioid receptor mRNA in the NAc shell.

Discussion

In this study, we used male and female OP and OR rats to determine the degree to which proneness to diet-induced obesity is associated with enhanced ethanol consumption and inherent differences in the NAc opioid system. Additionally, given established sex differences in ethanol drinking [40–43], we included sex as a variable in our study. We found that male OP compared to OR rats had higher levels of 24-hour ethanol drinking and preference, while female OP compared to OR rats had higher levels of binge-like (30-minute) ethanol drinking and subsequent BECs. In contrast, there were no differences between OP and OR rats of either sex in sucrose drinking or preference. In parallel, both male and female OP compared to OR rats had higher gene expression of enkephalin but not dynorphin in the NAc core, and male OP rats had higher gene expression of MOR, DOR, and KOR in the NAc core, along with higher expression of enkephalin and DOR in the NAc shell compared to male OR rats. Together, these results support the idea that proneness to obesity, in the absence of obesity or major differences in food intake, is associated with excessive ethanol drinking, and that there are sex-related differences apparent in the manifestation of this excessive drinking. These differences are discussed in light of the caloric content and pharmacological effects of ethanol, as well as the inherent differences in the NAc endogenous opioid system between OP and OR rats found here.

Ethanol consumption is enhanced in OP compared to OR rats, with sex-related differences

We found that female rats drank more 24-hour ethanol than male rats, and that OP rats drank more ethanol than OR rats, although the manner in which OP rats drank more ethanol differed between males and females. Consistent with our results, numerous studies have reported that female rodents consume more ethanol than males under the intermittent access procedure [40–43]. Thus, ethanol consumption during 24-hour access in the OP/OR models used here show sex-related differences in intake that are similar to those seen in outbred rodents. In addition, OP males had higher levels of 24-hour ethanol drinking and preference than OR males (Figure 1A,B); however, OP females had a larger ethanol binge and subsequently higher BECs compared to OR females (Figure 1C,D). While there are many potential reasons for these sex-related differences in the manifestation of increased ethanol drinking (see also below), these data nonetheless demonstrate that ethanol consumption is indeed enhanced in OP compared to OR rats, and they lend support to the idea that there may be overlap in the mechanisms of over-consumption of food and ethanol.

Ethanol is the only abused drug with caloric content (7 kcal/g). Thus, it is possible that the greater 24-hour intake in OP compared to OR males could be due to a substitution of calories from chow (3.35 kcal/g) with calories from ethanol. One argument against this is that work using food reinforcers has shown that OP and OR rats have similar preferences for sugary, fatty foods even when they are calorically matched to chow, and that over-consumption in OP rats primarily develops across time and as weight gain occurs [61]. On the other hand, the unique ability of ethanol to engage endogenous systems that regulate eating through physiologically relevant pathways, compared to other abused drugs which have no caloric density, may contribute to its ability to powerfully promote over-consumption more than its caloric content per se. Consistent with this idea, work with cocaine (which has no caloric content) found that initial acquisition and maintenance of self-administration was similar between OP and OR males, and both groups escalated their cocaine intake to a similar degree when given within-session intermittent access to cocaine [50]. To our knowledge this is the only other study that examined consumption of an abused drug in the OP/OR models. Because the male rats in this study overall consumed less ethanol than the females, and we have previously found that the highest time of ethanol intake across the 24-hour access period occurs in the first 30 minutes of daily access [40], it is unlikely that the OP or OR males achieved pharmacologically-relevant BECs from their ethanol drinking. This supports this idea that the caloric content of ethanol contributed in some degree to the greater 24-hour intake in OP compared to OR males. Of note, the level of chow intake was significantly lower in OP males on days when ethanol was available compared to days when it was not available. This leaves open the possibility that the greater 24-hour intake in OP compared to OR males was due in part to the caloric content of ethanol.

Ethanol as an abused drug also has psychoactive effects. Notably, women more than men in the United States have shown a recent increase in the prevalence of very high-risk drinking, which generally includes binge drinking [62]. It is likely that the higher levels of binge-like drinking in OP compared to OR females were driven by the pharmacological effects of ethanol rather than by its calories. These OP females consumed an average of 1.5 g/kg ethanol in 30 minutes and achieved BECs of 58 mg/dl, which are pharmacologically-relevant [36, 39, 40]. In contrast, neither males nor OR females achieved pharmacologically-relevant BECs in this study. With females overall drinking ethanol at higher levels than males, this indicates that the caloric component of drinking was likely less of a driving factor in the elevated drinking of the OP females, and that the psychoactive effects conceivably were a driving force behind their higher binge-like drinking. Notably, behavior surrounding the use of abused drugs and the neurobiology underlying this behavior has significant overlap with behavior surrounding the eating of highly palatable foods [63–65]. We speculate that the OP female rats may be predisposed to consume reinforcing, caloric substances more compulsively than OR rats, and this would apply both to sugary, fatty foods and to alcohol.

Sucrose consumption is similar between OP and OR rats

In contrast to ethanol, OP rats given intermittent access to a relatively low concentration of sucrose (2.5%) showed no differences from OR rats in 24-hour sucrose intake or preference, 30-minute binge-like intake, or subsequent blood glucose levels. This is consistent with previous studies where preference for a sucrose solution (1% or 4%) was found to be similar between OP and OR rats maintained on standard lab chow ([66] and unpublished observations by CRF). Indeed, work from our group and others has shown that increased food intake in OP compared to OR rats primarily emerges with weight gain and is more pronounced with ad libitum access to diets high in fats and sugars compared to access to standard chow alone [46–48, 66–69]. The same is true of elevations in fasted blood glucose, which also emerge in OP rats alongside weight gain after several weeks of access to high-fat and high-sugar diets [46, 47]. In the current study, body weight did not significantly diverge between OP and OR groups maintained on chow (Figure 2E), and sucrose intake during the 30-minute binge session also did not differ between groups (Figure 2C). Thus, it is not surprising that there were also no differences in post-ingestive blood glucose levels between OP and OR rats. Overall, the absence of differences in sucrose intake or preference during the 24-hour or 30-minute access period in OP compared to OR rats maintained on chow bolsters our interpretation that greater ethanol consumption in OP compared to OR rats is likely related to the unique caloric and pharmacological profile of ethanol.

Gene expression of opioids in the NAc is enhanced in OP compared to OR rats

With no prior exposure to ethanol or sucrose, both male and female OP compared to OR rats had higher gene expression of enkephalin but not dynorphin in the NAc core, and male OP rats had higher gene expression of MOR, DOR, and KOR in the NAc core, along with higher expression of enkephalin and DOR in the NAc shell. This is largely in line with previous results from male inbred and selectively-bred alcohol-preferring rats and mice [29–33] and outbred fat-preferring rats [34]. Our work expands these results to female rats; major divisions of the NAc (core and shell); and multiple opioids and their receptors within the same study. Of note, prior work found that oral administration of fat and ethanol but not sucrose similarly enhanced gene expression of enkephalin in the both the NAc core and NAc shell [70]. Moreover, we have previously shown that injection of a DOR agonist in the NAc shell of male Sprague-Dawley rats stimulated ethanol drinking under a low-level ethanol drinking model [26]. All together, this supports the idea that fat and ethanol consumption may converge on the NAc opioid system.

The majority of the differences in basal opioid gene expression that we observed between OP and OR rats occurred in the NAc core, and in males rather than females for the opioid receptors. In relation to food and abused drugs, the NAc core is traditionally linked to goal-directed motivation and anticipatory behavior, while the NAc shell is more traditionally linked to consummatory and hedonic responses [20, 71–73]. Indeed, the endogenous opioid system does appear to make different contributions to ingestive behavior via the NAc core compared to the NAc shell. For example, in regard to NAc opioids and binge intake, injection of an MOR agonist in the NAc core more than the NAc shell promotes binge intake of a sweet-fat solution [74] and, conversely, injection of an MOR antagonist in the NAc core reduces intake of a high-fat diet but not Ensure, while injection in the NAc shell reduces intake of Ensure but not a high-fat diet [75]. Similarly, injection of a KOR antagonist in the NAc core reduces initial self-administration of heroin while injection in the NAc shell reduces intake over time [76]. Thus, distinct patterns of gene expression differences in the NAc core and shell found here are likely to influence different aspects of ingestive behavior.

The finding that gene expression was different for enkephalin but not dynorphin further supports the idea that differences between OP and OR rats may lie more with reinforcement than with the hedonic components of reward, as it has been proposed that enkephalin is more important in reinforcing effects of drugs like alcohol, whereas dynorphin may be more involved in the affective aspects of drug use [77]. Regarding sex-related differences, our results are supported by several studies, which have noted greater changes in MOR expression or binding in the NAc in males compared to females, both in relation to feeding or obesity and exposure to abused drugs [78–80]. Sex-related differences in the NAc opioid system have not been described in alcohol-preferring compared to non-preferring rodents or in animals without special diet exposure. While elevated NAc enkephalin is a common feature in OP males and females, which both drink more ethanol than OR males and females, albeit in a different manner, it may be that a different peptide system or brain region is more important in driving the higher binge-like ethanol drinking of OP females. All together, although we have not tested causality in this study, these gene expression results suggest that elevated expression in the NAc of the opioid system of obesity-prone subjects, particularly enkephalin in the NAc core, could contribute to their elevated ethanol drinking.

Conclusions

Overall, our results confirm and extend previous findings regarding the neurobiological overlap between proneness to overeating-induced obesity and excessive alcohol drinking. They suggest that the specific manifestation of this excessive drinking may differ between males and females, that the overlap does not extend to all reinforcing substances, and that it could be driven in part by enkephalin in the NAc core. These findings point to the need for further investigation of sex-related differences in overeating-induced obesity and excessive alcohol drinking, and they suggest that medications found to be effective in treating obesity could be similarly effective in treating alcohol use disorder.

Highlights.

• Male obesity-prone vs. -resistant rats drink more ethanol in an intermittent model.

• Female obesity-prone vs. -resistant rats show higher binge-like ethanol drinking.

• Sucrose drinking is similar between obesity-prone and -resistant rats.

• Obesity-prone vs. -resistant rats have more nucleus accumbens proenkephalin mRNA.

• Male obesity-prone rats have more mu, delta, and kappa receptor mRNA in the core.

Data availability statement

All data are available at OSF.