Intermittent dietary supplementation with fish oil prevents high fat diet-induced enhanced sensitivity to dopaminergic drugs

Abstract

Eating a high fat diet can lead to obesity, type 2 diabetes, and dopamine system dysfunction. For example, rats eating high fat chow are more sensitive than rats eating standard chow to the behavioral effects (e.g., locomotion and yawning) of dopaminergic drugs (e.g., quinpirole and cocaine). Daily dietary supplementation with 20% (w/w) fish oil prevents high fat diet-induced enhanced sensitivity to quinpirole-induced yawning and cocaine-induced locomotion; however, doctors recommend that patients take fish oil just 2-3 times a week. To test the hypothesis that intermittent (i.e., 2 days per week) dietary supplementation with fish oil prevents high fat diet-induced enhanced sensitivity to quinpirole and cocaine, rats eating standard chow (17% kcal from fat), high fat chow (60% kcal from fat), and rats eating standard or high fat chow with 20% (w/w) intermittent (e.g., 2 days per week) dietary fish oil supplementation were tested once weekly with quinpirole (0.0032-0.32 mg/kg, i.p.) or cocaine (1.0-17.8 mg/kg, i.p.) using a cumulative dosing procedure. Consistent with previous reports, eating high fat chow enhanced sensitivity of rats to the behavioral effects of quinpirole and cocaine. Intermittent dietary supplementation of fish oil prevented high fat chow-induced enhanced sensitivity to dopaminergic drugs in male and female rats. Future experiments will focus on understanding the mechanism(s) by which fish oil produces these beneficial effects.

Introduction

The prevalence of obesity remains a major public health concern with more than 40% of adults in the United States diagnosed as overweight or obese (Hales et al., 2020). Overconsumption of high fat or high sugar foods has been associated with obesity and type 2 diabetes (U.S. Department of Agriculture, 2010) and can also cause dysfunction to dopamine systems (Baladi et al., 2011; Baladi et al., 2012a), the same systems that are targeted by drugs of abuse (Wise and Robble, 2020). For example, patients diagnosed with obesity have less dopamine receptor binding as compared to controls (Tomasi and Volkow, 2013), suggesting that dopamine systems are altered among individuals with a chronic history of consuming high fat or high sugar foods. Studies with rodents have also demonstrated that eating a diet that is high in fat or sugar can dramatically alter dopamine systems (South and Huang, 2008; Fordahl and Jones, 2017; Speed et al., 2011). For example, rats eating high fat chow display decreases in dopamine transporter function and expression (Speed et al., 2011) as well as changes to dopamine receptor expression (South and Huang, 2008). Further, rodents eating high fat chow are more sensitive to the behavioral effects of drugs that act on dopamine receptors (e.g., dopamine D₂/D₃ receptor agonist quinpirole; Baladi et al., 2011) and dopamine transporters (e.g., cocaine; Baladi et al., 2012b; Collins et al., 2015). This change in drug sensitivity, as a result of diet, could have implications for vulnerability to substance use disorder, but also for the effectiveness of therapeutic drugs that target dopamine systems.

Recent evidence suggests that while eating a high fat diet could serve as a potential vulnerability factor, dietary supplements that are rich in omega-3 polyunsaturated fatty acids, such as those found in fish oil, might help to protect individuals against the development of substance use disorder (Darcey and Serafine, 2020). Several recent publications using animal models also demonstrate that dietary supplementation with fish oil, can prevent or reverse high fat chow-induced effects, including enhanced sensitivity to dopaminergic drugs (Serafine et al., 2016; Hernandez-Casner et al., 2017; Hernandez-Casner et al., 2019). These studies involved daily dietary supplementation with fish oil, despite the fact that doctors recommend only 2 servings (e.g., 3.5 oz of fish or 1000 mg of fish oil) per week to experience health benefits (Lichtenstein et al., 2006; American Heart Association, 2017). While intermittent dietary supplementation might be sufficient for the beneficial effects of fish oil on cardiovascular health, is not known if intermittent (e.g., 2 days per week) fish oil dietary supplementation is also sufficient to prevent high fat diet-induced enhanced sensitivity to dopaminergic drugs. Understanding a minimally effective amount of fish oil consumption is especially critical, given recent reports linking excessive fish oil consumption with prostate cancer (Brasky et al., 2011; Brasky et al., 2013).

Sex differences have been revealed regarding the impact of eating high fat chow on sensitivity of rats to dopaminergic drugs. For example, while eating high fat chow can dramatically change sensitivity of male rats to quinpirole-induced yawning (Hernandez-Casner et al., 2017), dopaminergic drugs do not induce comparable yawning in females (Ramos et al., 2019; Serafine et al., 2014) and this behavioral effect appears to be androgen-mediated (Berendsen and Nickolson, 1981). In contrast, females are generally more sensitive to the locomotor-stimulating effects of psychostimulants like cocaine and methamphetamine (Anker and Carroll, 2011) and this effect is further enhanced in female rats eating high fat chow (Baladi et al., 2012b) as compared to males (Baladi et al., 2015). Therefore, in order to provide a comprehensive assessment of the beneficial effects of fish oil, it is necessary to study different drug-induced behavioral effects in females versus males.

While there is a growing literature examining how drug sensitivity is impacted by different diets, many laboratories use a variety of types (e.g., manufactured high fat chows versus cafeteria diets) and amounts (e.g., free access versus restricted access) of food in these assessments. Some modifications can provide experimental control for extraneous variables such as weight gain. For example, in one report assessing sensitivity of female rats to cocaine, one group of rats had free (e.g., ad libidum) access to high fat chow, while a second group had restricted access (Baladi, et al., 2012b). Rats with restricted access to high fat chow were not food restricted such that they lost weight, but rather they consumed enough high fat chow daily to maintain their body weight based on a normal growth curve for standard chow-fed control females (Baladi et al., 2012b). In contrast, female rats with free access to high fat chow gained weight throughout the study (e.g., similar to animal models of diet-induced obesity; Baladi et al., 2012b). Among females, this restricted access paradigm actually yielded more robust effects on drug sensitivity as compared to free access (Baladi et al., 2012b), suggesting that food consumption itself, rather than weight gain, is critical for changing sensitivity to dopaminergic drugs among females. In contrast, male rats eating high fat chow (either free or restricted access) were equally more sensitive to dopaminergic drugs than standard chow-fed controls (Baladi et al., 2011; Baladi et al., 2015).

Given these sex differences, in the present report male rats had free access to high fat chow, and females had either free access or restricted access to high fat chow, to approximate the conditions used in these previous studies (Baladi et al., 2011; Baladi et al., 2012b). The present study examined the potential beneficial effects of intermittent (e.g., 2 days per week) dietary supplementation with fish oil on high fat chow-induced enhanced sensitivity to dopaminergic drugs, in 2 experiments. Experiment 1 examined quinpirole-induced yawning in male rats, and experiment 2 examined cocaine-induced locomotion in female rats.

Methods

Subjects

Male (n = 32) and female (n = 48) Sprague-Dawley rats (Envigo, Indianapolis, IN, USA) arrived at approximately postnatal day (PND) 20-21. On average, male rats weighed 42.4 g ± 0.6 and female rats weighed 42.4 g ± 0.5 upon arrival. The experimental timeline varies between sexes slightly (by approximately 5 days) due to the nature of the behavioral assays being conducted and limitations on how many subjects could be tested on a given day (see table 1). However, this range of 5 days falls within the age range of behavioral testing published in previous reports (Baladi et al., 2012b; Baladi et al., 2011; Hernandez-Casner et al., 2017). Testing spanned the duration of adolescence into adulthood for both sexes. Rats were housed individually in an environmentally controlled room (20-26°C, 30-70% relative humidity) with a 12/12 light dark cycle and lights on at 0800 hours. Except during experimental testing, all rats had access to food and water in their home cages. Specific feeding conditions (type and amount of food available) for each group of rats are outlined below. Experiments were conducted in accordance with the Institutional Animal Care and Use Committee at The University of Texas at El Paso, and in accordance with the 2011 Guide for Care and Use of Laboratory Animals (Institute of Laboratory, Animal Resources on Life Sciences, the National Research Council, and the National Academy of Sciences).

Table 1:

Experimental timeline and sample weekly timeline for quinpirole induced-yawning in experiment 1 and cocaine induced-locomotion in experiment 2. Experimental timeline describing animal age during different stages of the experiments. The sample weekly timeline is an example, and not representative of all animals (e.g., some rats were tested with quinpirole (males) or cocaine (females) on days other than Tuesdays and were staggered by 5 days); however, fish oil supplementation always occurred 2 days after behavioral testing (with cocaine or quinpirole) for individual rats, and individual rats were always tested with drug on the same day of the week.

Experimental Timeline
Experiment 1	PND 20	PND 20-26	PND 27-64
Males	Arrival	Standard chow / habituation	Feeding conditions assigned (standard or high fat), quinpirole testing (once weekly)
Experiment 2	PND 20	PND 20-25	PND 26-108
Females	Arrival	Standard chow	Feeding conditions assigned (standard or high fat)
				Habituation (PND 61-67)	Cocaine testing (once weekly; PND 68-100)	Insulin testing (PND 104-108)
Sample Weekly Timeline
Monday	Tuesday	Wednesday	Thursday	Friday	Saturday	Sunday
Standard or High Fat	Standard or High Fat	Standard or High Fat	Standard or High Fat	Standard or High Fat	Standard or High Fat + 20% w/w Fish oil	Standard or High Fat + 20% w/w Fish oil

Feeding conditions

In experiment 1, male rats had free access to standard chow for approximately one week starting at arrival (approximately PND 20). One week later (staggered across PND 27-29) male rats were tested with quinpirole under baseline conditions (e.g., when all rats ate standard chow) and were then rank assigned to dietary conditions based on individual maximal effect for quinpirole-induced yawning (as has been reported previously, Hernandez-Casner et al., 2017). These feeding conditions continued for the remainder of the experiment (e.g., for 5 consecutive weeks through PND 64; see table 1). Specifically, separate groups of rats (n = 8 per feeding condition; see table 1) either continued to have free access to standard chow (Envigo Teklad 7912) or were given free access to high fat chow (Envigo Teklad 06414). Two additional groups of rats had free access to standard chow or high fat chow 5 days per week, and for the remaining 2 days per week they had free access to their respective diet (standard or high fat chow) supplemented with 20% (w/w) fish oil (Nordic Naturals Omega-3 Pet; Nordic Naturals, Watsonville, California, USA; see table 1). The concentration of 20% (w/w) fish oil was selected based on previous literature demonstrating that this concentration is sufficient to prevent or reverse the impact of a high fat diet on dopaminergic drug sensitivity (Serafine et al., 2016; Hernandez-Casner et al., 2017; Hernandez-Casner et al., 2019; and high fat diet-induced inflammation, Pimentel et al., 2013).

In experiment 2, female rats had free access to standard chow for approximately one week starting at arrival (approximately PND 20). Approximately one week later (staggered across PND 26-30). Female rats were then randomly assigned to dietary conditions (similar to previous reports, Baladi et al., 2012b; Serafine et al., 2016). These feeding conditions continued for the remainder of the experiment (e.g., for 11 consecutive weeks through PND 108; see table 1). Specifically, separate groups of rats (n = 8 per feeding condition; see table 1) either continued to have free access to standard chow or had free or restricted access to high fat chow. Rats in the group with restricted access were only provided enough high fat chow daily to maintain their body weight based on a normal growth curve derived from standard chow-fed females (see Baladi et al., 2012b; Serafine et al., 2014). Three additional groups of female rats had free access to standard chow, high fat chow, or restricted access to high fat chow for 5 days per week, and for the remaining 2 days per week they had free or restricted access to their respective chow supplemented with 20% (w/w) fish oil.

Body weight and food consumption were monitored daily and food was replaced daily between 0800-1000h. For intermittent fish oil access, the 2 consecutive days each week when fish oil was provided remained constant for individual rats (e.g., fish oil was provided on the same days each week). Quinpirole or cocaine tests always occurred on the same days each week for all groups, which was 2 days after fish oil was removed for rats in intermittent access groups. The nutritional content of the standard chow (Envigo Teklad 7912) was 5.7% fat, 44.3% carbohydrate, and 19.9% protein with energy content of 3.1 kcal/g. The high fat chow (Envigo Teklad 06414) contained 34.3% fat, 27.3% carbohydrate, and 23.5% protein with a calculated energy content of 5.1 kcal/g. The calculated energy content for the standard chow with 20% (w/w) fish oil was 4.28 kcal/g, whereas the supplemented high fat chow with 20% (w/w) fish oil had a calculated energy content of 5.88 kcal/g (see table 2).

Table 2:

Feeding conditions for male rats in experiment 1 and female rats in experiment 2 based on total percentages of fat, carbohydrates, protein, and calculated energy content (kcal/g). For experiment 2 only, there was also a restricted access high fat condition, in which female rats had access to a restricted amount of high fat chow (e.g., enough to maintain body weight according to a standard growth curve). The macronutrient breakdown was the same for rats in the restricted access condition; however, they received fewer total grams of food daily than rats in free access conditions.

Feeding Conditions
Diet	% fat	% carbohydrates	% protein	kcal/g
Standard	5.7	44.3	19.9	3.1
High fat	34.3	27.3	23.5	5.1
Standard + intermittent fish oil	7.0	44.3	19.9	4.28
High fat + intermittent fish oil	39.5	27.3	23.5	5.88

Experiment 1: Yawning and Body Temperature

Yawning was defined as an opening and closing of the mouth (for ~1 second) such that the lower incisors were visible (Baladi and France, 2010). Male rats were habituated to experimental procedures on PND 22-24 and quinpirole testing began on PND 27-29, while all rats still had access to standard chow (e.g., baseline conditions). Quinpirole-induced yawning was examined once per week in rats beginning under baseline conditions and continuing until the end of the study (e.g., after eating their respective diets for 5 consecutive weeks). On test days, rats were removed from their home cages and placed in identical cages with an acrylic plexiglass divider to observe yawning. During quinpirole tests, no access to bedding, food, or water was provided. Following a 15-minute habituation period, body temperature was assessed before injections of saline and cumulative doses of quinpirole (0.0032, 0.01, 0.032, 0.1, and 0.32 mg/kg; i.p.) as has been done previously (Hernandez-Casner et al., 2017; Baladi et al., 2011). Injections (saline followed by increasing cumulative doses of quinpirole) occurred every 30 minutes, and yawning was observed in 10-minute intervals, starting 20 minutes after each injection.

Body temperature was measured in a temperature-controlled room (20-26°C, 30-70% relative humidity). During quinpirole experiments, rectal body temperature was assessed by inserting a lubricated thermal probe attached to a thermometer ~3cm into the rectum prior to each injection of saline or cumulative doses of quinpirole for a total of 6 measurements. This was done initially with saline during PND 22-24 prior to quinpirole testing (as described above) in order to habituate animals to the procedure and to obtain baseline body temperature time-courses for individual subjects under drug-free conditions. Quinpirole was administered once weekly, for 5 consecutive weeks, starting on PND 27-29 and ending on PND 62-64 (represented as “Week 5” in Fig. 4 & 5 to designate the last week of quinpirole testing).

Figure 4.

Mean ± SEM number of yawns for male rats observed in a 10-minute period for rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles), high fat chow (squares), and high fat chow with intermittent fish oil supplementation (diamonds) after rats ate their respective diets for 5 consecutive weeks in experiment 1; n = 8/group. Vertical axis: yawns in 10-minute observation period. Horizontal axis: cumulative doses of quinpirole (mg/kg).

Figure 5.

Mean ± SEM body temperature for male rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles), high fat chow (squares), and high fat chow with intermittent fish oil supplementation (diamonds) after rats ate their respective diets for 5 consecutive weeks in experiment 1; n = 8/group. Vertical axis: temperature °C. Horizontal axis: cumulative doses of quinpirole (mg/kg).

Experiment 2: Locomotion and Insulin Sensitivity

Cocaine-induced locomotion was examined in chambers using Med Associates open field chambers with Activity Monitor 7 Software (Med Associates Inc, Fairfax, VT). Female rats were habituated to experimental procedures after rats had been eating their respective diets for 5 weeks (e.g., habituation was staggered between PND 61-65) and cocaine testing began after rats had been eating their respective diets for 6 weeks (e.g., staggered between PND 68-72; represented as “Week 1” in Fig. 7 to designate the first week of cocaine testing, also represented in Fig. 6). To measure locomotor activity, female rats were placed into individual chambers and were allowed to habituate to the environment for 30 minutes. After this initial habituation period locomotor activity was tracked during 15-minute intervals, after an injection of saline followed by increasing cumulative doses of cocaine (1.0, 3.2, 10.0, 17.8 mg/kg; i.p.), for a total of 5 injections. The last 5 minutes of activity for each 15-minute interval was measured in order to capture the peak effect of cocaine-induced locomotion (Catlow and Kirstein, 2005), to avoid including any motor behavior that might have been related to grooming/hyperactivity following handling and injections (e.g., occurring just after the injection interval) and matching previous reports (Baladi et al., 2012b; McGuire et al., 2011; Serafine et al., 2016). Locomotor activity was measured via ambulatory counts (e.g., individual beam breaks) occurring within this 15-minute period. Cocaine was administered once weekly, for 5 consecutive weeks, starting after rats had been eating their respective diets for 6 weeks (e.g., staggered between PND 68-72) and ending after rats had been eating their respective diets for 10 weeks (e.g., ending between PND 96-100).

Figure 7.

Mean ± SEM area under the cocaine dose-response curve (AUC) for female rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles), restricted access to high fat chow (squares), and restricted access to high fat chow with intermittent fish oil supplementation (diamonds). Female rats had access to these respective diets for 6 weeks prior to (e.g., before “Week 1) and during cocaine testing, which continued once weekly for 5 consecutive weeks (e.g., through “Week 5”) in experiment 2; n = 8/group. Vertical axis: area under the curve. Horizontal axis: week in study.

Figure 6.

Mean ± SEM locomotor activity counts recorded in a 5-minute period for female rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles), restricted access to high fat chow (squares), and restricted access to high fat chow with intermittent fish oil supplementation (diamonds) after rats ate their respective diets for 6 consecutive weeks (also represented as the first week of cocaine testing in Fig. 7) in experiment 2; n = 8/group. Vertical axis: locomotor activity. Horizontal axis: cumulative doses of cocaine (mg/kg).

Previous reports have demonstrated that female rats eating high fat chow for 5 weeks do not become insulin resistant (Serafine et al., 2016). However, longer access to high fat chow could result in changes to insulin sensitivity. Therefore, insulin sensitivity was measured one week following completion of behavioral assessments (PND 104-108; e.g., after rats ate their respective diets for 11 consecutive weeks). After rats fasted for a minimum of 6 hours, a sample of blood was collected from the end of the tail and placed onto a blood-test strip and examined using a commercial blood glucose meter (Accu-Checc Aviva, CVS). A baseline measurement was collected, and blood glucose was examined 15, 30, 45, and 75 minutes after an injection of insulin (2.0 U/kg, i.p.) as has been done previously (Baladi et al., 2011; Serafine et al., 2016).

Drugs

Quinpirole hydrochloride, cocaine hydrochloride, and insulin (Humulin R; 100U) were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). Quinpirole, cocaine and insulin were dissolved in 0.9% sterile saline and were injected intraperitoneally (i.p.) at a volume of 1 ml/kg body weight.

Data analyses

Quinpirole-induced yawning results are expressed as the average (±S.E.M.) number of yawns observed in each 10-minute observation period plotted as a function of dose. Small doses of quinpirole increased yawning, while large doses decreased yawning, resulting in an inverted U-shaped dose-response curve. To examine changes in quinpirole-induced yawning, several calculations were performed. First, the maximal effect was calculated for individual rats to examine changes in the magnitude of yawning between groups. Mean (±S.E.M.) maximal effect for rats in different groups were analyzed using a two-way mixed model analysis of variance (ANOVA) with Tukey’s post hoc comparisons where appropriate. Additionally, in order to examine shifts in the dose-response curve between groups, the ascending and descending limbs of the dose-response curve were analyzed separately (Hernandez-Casner et al., 2017; Serafine et al., 2014). Each limb was analyzed separately because previous work has demonstrated that the two limbs are mediated by different receptor subtypes (Collins et al., 2005). The data were expressed as a percentage of the maximal yawning effect (for individual rats), the linear portion of each limb of the dose-response curve was determined, and then linear regression was used to estimate the log ED₅₀ values for the ascending and descending limbs of each individual dose-response curve. The linear portion of each limb of the dose-response curve included doses that spanned the 50% level of effect and included not more than one dose producing greater than 75% effect, and not more than one dose producing less than 25% effect. Differences in log ED₅₀ values and in maximal effects were analyzed using an ordinary one-way ANOVA and Tukey’s multiple comparisons tests to compare group differences. Quinpirole-induced hypothermia was reported as the average (±S.E.M.) body temperature expressed in degrees Celsius (C°) and analyzed using a two-way mixed model ANOVA with Tukey’s multiple comparisons tests.

Body weight results were reported as the average (±S.E.M.) body weight expressed in grams (g). Results were analyzed using a two-way mixed model ANOVA with Tukey’s multiple comparisons tests. Food consumption, expressed in both g and kilocalories (kcal), were reported as the average (±S.E.M.) analyzed using a two-way mixed model ANOVA with Tukey’s multiple comparisons tests. Food consumption data is presented using both g and kcal consumed, because the different diets contain a different amount of kcal per g of chow. Cocaine-induced locomotion results were expressed as the average (±S.E.M.) number of ambulatory locomotor counts occurring in each last 5-minute observation period plotted as a function of dose. Locomotor activity was analyzed using a two-way mixed model ANOVA with Tukey’s multiple comparisons tests. Area under the curve (AUC) was analyzed using a two-way mixed model ANOVA with Tukey’s multiple comparisons tests. Insulin sensitivity results are expressed as the average (±S.E.M.) blood glucose concentration measured at differing time intervals analyzed using a two-way mixed model ANOVA with Tukey’s multiple comparisons tests. For all assessments, the a priori criteria for statistical significance was < 0.05.

Results

Body weight

In experiment 1, all 32 male rats gained weight throughout the study, regardless of chow (see Fig. 1A and 1C). Under baseline conditions, rats in all groups weighed 87.65 g ± 1.61 on average. At the end of the study (e.g., after rats ate their respective diets for 5 consecutive weeks), male rats with free access to standard chow weighed 299.6 g ± 7.01, male rats with free access to high fat chow weighed 326.12 g ± 7.42, male rats with free access to standard chow with intermittent fish oil supplementation weighed 300.13 g ± 5.0, and male rats with free access to high fat chow with intermittent fish oil supplementation weighed 312.0 g ± 7.03. A mixed model ANOVA revealed a significant main effect of day [F(34, 952) = 4133, p < 0.0001], a significant day by diet interaction effect [F(102, 952) = 3.181, p < 0.0001], but no significant main effect of diet. Tukey’s multiple comparisons tests revealed that at the end of the study (e.g., after rats ate their respective diets for 5 consecutive weeks), male rats with free access to high fat chow weighed significantly more than male rats with free access to standard chow (p < 0.01) and male rats with free access to standard chow with intermittent fish oil supplementation (p < 0.01).

Figure 1.

Mean ± SEM body weights for male and female rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles) (a-b), free or restricted access to high fat chow (squares), and free or restricted access to high fat chow with intermittent fish oil supplementation (diamonds) (c-d); n = 8/group. Days when fish oil supplementation occurred are shaded darker than non-supplementation days. Baseline (BL) refers to when all rats were eating standard chow at the start of the study (e.g., prior to dietary manipulation). Vertical axis: body weight in grams (g). Horizontal axis: week in study.

In experiment 2, all 48 female rats gained weighed throughout the study, regardless of chow (see Fig. 1B and 1D). Under baseline conditions, female rats in all groups weighed 91.06 g ± 0.9 on average. At the end of the study (e.g., after rats ate their respective diets for 10 consecutive weeks), female rats with free access to standard chow weighed 226.63 g ± 8.24, female rats with free access to high fat chow (data not shown) weighed 232.25 g ± 4.37, female rats with restricted access to high fat chow weighed 225.13 g ± 2.2, female rats with restricted access to high fat chow with intermittent fish oil supplementation weighed 226.5 g ± 3.9, female rats with free access to standard chow with intermittent fish oil supplementation weighed 226.5 g ± 3.98, and female rats with free access to high fat chow with intermittent fish oil supplementation (data not shown) weighed 251.0 g ± 8.23. A mixed model ANOVA revealed a significant main effect of day [F(69, 2898) = 2158, p < 0.0001], a significant main effect of diet [F(5, 42) = 2.847, p < 0.05], and a significant day by diet interaction effect [F(345, 2898) = 1.896, p < 0.0001]. Tukey’s multiple comparisons tests revealed that at the end of the study (e.g., after rats ate their respective diets for 10 consecutive weeks), female rats with free access to high fat chow with intermittent fish oil supplementation (data not shown) weighed significantly more than female rats with free access to standard chow (p < 0.01), female rats with free access to high fat chow (data not shown; p < 0.05), female rats with restricted access to high fat chow (p < 0.01), and female rats with restricted access to high fat chow with intermittent fish oil supplementation (p < 0.01).

Feeding conditions

In experiment 1, food consumption in g (see Fig. 2A and 2C) and kcal (see Fig. 3A and 3C), were examined as a function of day following assignment to dietary conditions. A mixed model ANOVA revealed a significant main effect of day [F(34, 952) = 22.92, p < 0.0001], a significant main effect of diet [F(3, 28 = 131.9, p < 0.0001], and a significant day by diet interaction effect [F(102, 952) = 2.255, p < 0.0001]. Tukey’s multiple comparisons tests revealed that at the end of the study (e.g., after rats ate their respective diets for 5 consecutive weeks), male rats with free access to standard chow consumed significantly more g on average than male rats with free access to high fat chow (p < 0.0001) and male rats with free access to high fat chow with intermittent fish oil supplementation (p < 0.01). Tukey’s multiple comparisons tests also revealed that male rats with free access to standard chow with intermittent fish oil supplementation consumed significantly more g on average than male rats with free access high fat chow (p < 0.0001) and male rats with free access high fat chow with intermittent fish oil supplementation (p < 0.01).

Figure 2.

Mean ± SEM food consumption for male and female rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles) (a-b), free or restricted access to high fat chow (squares), and free or restricted access to high fat chow with intermittent fish oil supplementation (diamonds) (c-d); n = 8/group. Days when fish oil supplementation occurred are shaded darker than non-supplementation days. Baseline (BL) refers to when all rats were eating standard chow at the start of the study (e.g., prior to dietary manipulation). Vertical axis: food consumption in grams (g). Horizontal axis: week in study.

Figure 3.

Mean ± SEM food consumption for male and female rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles) (a-b), free or restricted access to high fat chow (squares), and free or restricted access to high fat chow with intermittent fish oil supplementation (diamonds) (c-d); n = 8/group. Days when fish oil supplementation occurred are shaded darker than non-supplementation days. Baseline (BL) refers to when all rats were eating standard chow at the start of the study (e.g., prior to dietary manipulation). Vertical axis: food consumption in kilocalories (kcal). Horizontal axis: week in study.

After converting g consumed to kcal consumed (see Fig. 3A and 3C), a mixed model ANOVA revealed a significant main effect of day [F(34, 952) = 30.12, p < 0.0001], a significant main effect of diet [F(3, 28) = 29.9, p < 0.0001], and a significant day by diet interaction effect [F(102, 952) = 4.456, p < 0.0001]. Tukey’s multiple comparisons tests revealed that at the end of the study (e.g., after rats ate their respective diets for 5 consecutive weeks), there were no significant differences among the groups of male rats.

In experiment 2, food consumption in g (see Fig. 2B and 2D) and kcal (see Fig. 3B and 3D), were examined as a function of day following assignment to dietary conditions. A mixed model ANOVA revealed a significant main effect of day [F(69, 2829) = 3.539, p < 0.0001], a significant main effect of diet [F(5, 41) = 279.4, p < 0.0001], and a significant day by diet interaction effect [F(345, 2829) = 2.017, p < 0.0001). Tukey’s multiple comparisons tests revealed that at the end of the study (e.g., after rats ate their respective diets for 10 consecutive weeks), female rats with free access to standard chow consumed significantly more g on average than female rats with free access to high fat chow (data not shown; p < 0.0001), female rats with free access to high fat chow with intermittent fish oil supplementation (data not shown; p < 0.0001), female rats with restricted access to high fat chow (p < 0.0001), and female rats with restricted access to high fat chow with intermittent fish oil supplementation (p < 0.0001). Female rats with free access to standard chow with intermittent fish oil supplementation consumed significantly more g on average than female rats with free access to high fat chow (data not shown; p < 0.0001), female rats with free access to high fat chow with intermittent fish oil supplementation (data not shown; p < 0.01), female rats with restricted access to high fat chow (p < 0.0001), and female rats with restricted access to high fat chow with intermittent fish oil supplementation (p < 0.0001). Further, female rats with free access to high fat chow with intermittent fish oil supplementation (data not shown) consumed significantly more g on average than female rats with restricted access to high fat chow with intermittent fish oil supplementation (p < 0.01).

After converting g consumed to kcal consumed (see Fig. 3B and 3D), a mixed model ANOVA also revealed a significant main effect of day [F(69, 2829) = 5.904, p < 0.0001, a significant main effect of diet [F(5, 41) = 32.61, p < 0.0001], and a significant day by diet interaction effect [F(345, 2829) = 2.416, p < 0.0001). Tukey’s multiple comparisons tests revealed that at the end of the study (e.g., after rats ate their respective diets for 10 consecutive weeks), female rats with free access to high fat chow consumed significantly more kcal on average than female rats with restricted access to high fat chow with intermittent fish oil supplementation (data not shown; p < 0.05). Tukey’s multiple comparisons tests also revealed that female rats with free access to high fat chow with intermittent fish oil consumed significantly more kcal on average than female rats with restricted access to high fat chow with intermittent fish oil supplementation (data not shown; p < 0.01).

Experiment 1: Yawning and Body Temperature

In experiment 1 beginning with baseline testing, small doses of quinpirole increased yawning and larger doses decreased yawning, resulting in an inverted U-shaped dose-response curve (see Fig. 4) similar to previous reports (Baladi and France, 2009; Collins et al., 2005; Hernandez-Casner et al., 2017). At the end of the study (e.g., after rats ate their respective diets for 5 consecutive weeks), the quinpirole-induced yawning dose-response curve was shifted to the left for male rats with free access to high fat chow (see Fig. 4). That is, a one-way ANOVA examining the ED₅₀ values revealed significant differences among the ascending [F(3, 28) = 3.221, p < 0.05] and descending [F(3, 28) = 6.96, p < 0.01] limbs of the quinpirole-induced yawning dose-response curve (see table 3). Tukey’s multiple comparisons tests revealed significantly smaller ED₅₀ values for the ascending limb for male rats with free access to high fat chow compared to male rats with free access to standard chow (p < 0.05). Tukey’s multiple comparisons tests also revealed significantly smaller ED₅₀ values for the descending limb for male rats with free access to high fat chow compared to male rats eating free access to standard chow (p < 0.01), male rats with free access to standard chow with intermittent fish oil supplementation (p < 0.01), and male rats with free access to high fat chow with intermittent fish oil supplementation (p < 0.05). When examining maximal effect across 5 weeks of testing, a mixed model ANOVA revealed a significant main effect of week [F (4, 140) = 6.320, p < 0.0001], a significant main effect of diet [F (3, 28) = 3.707, p < 0.05], but no significant week by diet interaction effect. Tukey’s multiple comparisons tests revealed that at week 4, rats eating high fat chow had a significantly greater maximal effect than rats eating standard chow with intermittent fish oil supplementation (p < 0.05). However, by week 5, there were no significant differences in maximal effect among groups.

Table 3.

ED₅₀ values (mg/kg) for the ascending and descending limbs of the quinpirole-induced yawning dose-response curve for male rats after 5 consecutive weeks of quinpirole testing in experiment 1.

ED50 values (mg/kg)
Diet	Ascending	Descending
Standard	0.007	0.055
High fat	0.003 *	0.030 ^
Standard + intermittent fish oil	0.006	0.055
High fat + intermittent fish oil	0.005	0.050

^*denotes significantly different than rats eating standard chow for the ascending limb.

^{^}denotes significantly different than all other groups for the descending limb.

In experiment 1 beginning with baseline testing, larger doses of quinpirole decreased body temperature (see Fig. 5) in male rats. At the end of the study (e.g., after rats ate their respective diets for 5 consecutive weeks), a mixed model ANOVA revealed a significant main effect of dose [F(5, 140) = 1443, p < 0.0001], a significant dose by diet interaction effect [F(15, 140) = 2.188, p < 0.01], but no significant main effect of diet. Tukey’s multiple comparisons tests revealed that male rats with free access to standard chow with intermittent fish oil had a significantly warmer body temperature than male rats with free access to high fat chow after 0.1 mg/kg quinpirole (p < 0.01). However, linear regression revealed no significant group differences among dietary conditions for the slopes or for the intercepts.

Experiment 2: Locomotion and Insulin Sensitivity

In experiment 2 beginning with baseline testing, cocaine-induced locomotion increased as function of dose (see Fig. 6) in female rats. At the beginning of the experiment (e.g., after rats ate their respective diets for 6 consecutive weeks) when analyzing ambulatory counts, a mixed model ANOVA revealed a significant main effect of dose [F(4, 168) = 96.73, p < 0.0001], a significant dose by diet interaction effect [F(20, 168) = 1.689, p < 0.05], but no significant main effect of diet. Tukey’s multiple comparisons tests revealed that female rats with restricted access to high fat chow had significantly greater locomotor activity at the cumulative dose of 10.0 mg/kg compared to female rats with free access to standard chow (p < 0.05), female rats with free access to standard chow with intermittent fish oil supplementation (p < 0.01), female rats with free access to high fat chow (data not shown; p < 0.01), and female rats with restricted access to high fat chow with intermittent fish oil supplementation (p < 0.05). When examining across all 5 weeks of cocaine testing plotted as the AUC (see Fig. 7) a mixed model ANOVA revealed a significant main effect of week [F(4, 168) = 31.01, p < 0.0001], no significant main effect of diet, and no significant week by diet interaction effect. Tukey’s multiple comparisons tests revealed that at week 2, female rats with free access to high fat chow had a significantly greater AUC than female rats with free access to standard chow with intermittent fish oil supplementation (data not shown; p < 0.05). Sensitization occurred in all dietary groups (e.g., linear regression slope that is significantly non-zero). That is, all female rats had increased sensitivity to the locomotor stimulating effects of cocaine throughout the 5 weeks of repeated testing.

In experiment 2, at the end of the study (e.g., after female rats ate their respective diets for 11 consecutive weeks) blood glucose concentrations were obtained. A mixed model ANOVA reported a significant main effect of time [F(4, 168) = 201.8, p < 0.0001], a significant time by diet interaction effect [F(20, 168) = 2.263, p < 0.01], but no significant main effect of diet. All female rats experienced hypoglycemia in response to insulin injections (see Fig. 8).

Figure 8.

Mean ± SEM percent change in blood glucose of female rats eating standard chow (circles), standard chow with intermittent fish oil supplementation (triangles), restricted access to high fat chow (squares), and restricted access to high fat chow with intermittent fish oil supplementation (diamonds). Female rats had access to these respective diets for 11 total weeks prior to insulin testing after rats ate their respective diets for 11 consecutive weeks in experiment 2; n = 8/group. Vertical axis: percent change in glucose (mg/dL). Horizontal axis: time (minutes).

Discussion

Eating a high fat diet leads to negative health consequences such as obesity and type II diabetes (U.S. Department of Agriculture, 2010), and can also cause dysfunction to dopamine systems (Baladi et al., 2011; Baladi et al., 2012a). Preclinical studies have demonstrated that daily dietary supplementation with fish oil prevents high fat chow-induced enhanced sensitivity to the behavioral effects of dopaminergic drugs (Hernandez-Casner et al., 2017; Serafine et al., 2016). The present study examined the impact of intermittent (e.g., 2 days a week) supplementation with fish oil on high fat chow-induced enhanced sensitivity to quinpirole (experiment 1) and cocaine (experiment 2). Different behavioral assays were used to examine males and females, based on previous work from our lab and others demonstrating that quinpirole-induced yawning works well to examine the impact of diet in males (Hernandez-Casner et al., 2017; Baladi and France, 2009), but not in females (Serafine et al., 2014; Ramos et al., 2019), whereas cocaine-induced locomotion works better to examine the impact of diet in females as compared to males (Baladi et al., 2012b; Baladi et al., 2015; Serafine et al., 2016).

In experiment 1, at the end of the study (e.g., after rats ate their respective diets for 5 consecutive weeks), quinpirole-induced yawning was enhanced among male rats with free access to high fat chow, as evidenced by a leftward-shift in the ascending and descending limb of the dose-response curve (see Fig. 4; Baladi et al., 2011; Baladi and France, 2009; Hernandez-Casner et al., 2017). This leftward shift did not occur in male rats eating high fat chow with intermittent dietary fish oil supplementation (see Fig. 4). These results suggest that intermittent dietary supplementation with fish oil prevented high fat chow-induced enhanced sensitivity to quinpirole, since quinpirole-induced yawning for male rats eating high fat chow with intermittent fish oil and male rats eating standard chow was not significantly different. Intermittent fish oil had no effect on its own (i.e., in the absence of high fat chow) since quinpirole-induced yawning was not significantly different between male rats eating standard chow with intermittent fish oil supplementation and male rats eating standard chow alone (see Fig. 4).

Male rats with free access to high fat chow were not obese (see Fig. 1A and 1C), demonstrating that similar to previous reports (Baladi et al., 2011; Baladi et al., 2012b), drug sensitivity is altered by type of food consumed rather than actual obesity or weight gain. Male rats with free access to high fat chow consumed less g of food on average per day than rats with free access to standard chow but consumed an equal number of kcal on average daily (see Fig. 2A, 2C, 3A, and 3C). Male rats with free access to standard chow with intermittent fish oil supplementation consumed more kcal on average than male rats in all other dietary conditions (see Fig. 3A). As has been shown previously, decreases in food consumption occurring weekly correspond to the days immediately following each quinpirole test (Terry et al., 1995). Food consumption recovered by 2 days following quinpirole administration, suggesting the food suppressant effects of quinpirole were not long lasting.

In experiment 2, at the beginning of the experiment (e.g., after rats ate their respective diets for 6 consecutive weeks) female rats with restricted access to high fat chow were more sensitive to cocaine-induced locomotion than female rats eating free access to standard chow (Baladi et al., 2012b). While some reports have demonstrated a similar enhanced sensitivity of female rats with free access to high fat chow (Baladi et al., 2012b; Serafine et al., 2016), this was not replicated in the present study, as there were no significant differences in cocaine-induced locomotion among female rats with free access to high fat chow as compared to female rats with free access to standard chow (data not shown). Female rats with restricted access to high fat chow had a significantly greater locomotor activity than rats eating free access to standard chow (see Fig. 6). Rats with intermittent fish oil dietary supplementation were protected from the effects of restricted access to high fat chow, since cocaine-induced locomotion among rats with restricted access to high fat chow with intermittent fish oil supplementation was not significantly different than rats eating standard chow (see Fig. 6). This effect is prevented when supplementing the chow with fish oil intermittently (e.g., 2 days a week). Intermittent dietary fish oil supplementation had no impact in the absence of high fat chow, since cocaine-induced locomotion among rats eating standard chow and rats eating standard chow with intermittent fish oil was not significantly different (see Fig. 6). These significant differences were only apparent when cocaine administration was acute (e.g., during the first week of cocaine testing). Repeated weekly testing with cocaine resulted in cocaine-induced sensitization that was comparable across groups (see Fig. 7). Although female rats with restricted access to high fat chow were more sensitive to the acute effects of cocaine induced-locomotion, insulin sensitivity levels were similar across groups of female rats eating different diets. That is, all female rats were comparably sensitive to insulin injections, and no insulin resistance developed in female rats eating high fat chow (see Fig. 8).

One consideration worth noting is that in the present report, there were some slight differences between experiment 1 and experiment 2 that could impact direct comparisons being made between sexes. First, while diet began roughly around the same age for both males and females, male quinpirole testing ended after 5 weeks of eating these diets, whereas females started behavioral testing with cocaine after 6 weeks of eating their diets, and were tested once weekly for an additional 5 weeks. In other words, the age of rats at the end of the experiment was different, even though for the time points represented in figures 4 and 6, rats were approximately the same age. It is possible that continued exposure to diet beyond 5 weeks might have produced additional changes; however, this is not supported by the literature (see Ramos et al., 2019; Hernandez-Casner at el., 2019 for examples with longer access to high fat chow). Further, while the present results demonstrate that generally, intermittent fish oil has an attenuating effect on the high fat diet-induced effects on behavioral sensitivity to dopaminergic drugs in both sexes, both assays were not conducted in both sexes precluding our ability to make a direct comparison between sexes. Previous reports have demonstrated that quinpirole does not induce comparable frequencies of yawning among females (Ramos et al., 2019; Serafine et al., 2014) and that eating high fat chow enhances sensitivity of females more so than males to cocaine-induced locomotion (Baladi et al., 2012b; Baladi et al., 2015; Serafine et al., 2016), limiting the ability to test both sexes in both assays with a comparable degree of effectiveness.

The type and amount of food consumed plays an important role in the phenomena of dopamine system dysfunction. This can be examined with drugs that act on dopamine receptors (e.g., quinpirole) and dopamine transporters (e.g., cocaine). The present report replicated previous findings, demonstrating that male rats with free access to high fat chow (experiment 1) and female rats with restricted access to high fat chow (experiment 2) are more sensitive to the behavioral effects of dopaminergic drugs than rats eating standard chow. In both experiments, this enhanced sensitivity among rats eating high fat chow was prevented when the chow was supplemented with fish oil for just 2 days per week. Although the mechanisms driving these high fat diet-induced increases in sensitivity to dopaminergic drugs remain unknown, previous reports have demonstrated that diet can directly impact the function and expression of dopamine transporters (Speed et al., 2011), as well as expression of dopamine receptors (South and Huang, 2008; Ramos et al., 2020) that are important sites of action for cocaine and quinpirole, respectively. Further, the mechanisms by which fish oil produces protective effects (e.g., prevents high fat diet-induced enhanced sensitivity to drugs) also remain unknown, though omega-3 fatty polyunsaturated acids have been shown to modulate several relevant targets (including dopamine transporters, vesicular monoamine transporters, dopamine D1 and D2 receptors) involved in dopamine synthesis, release, and homeostasis (Metz et al., 2019; Chitre et al., 2020). Future directions will investigate the mechanisms by which fish oil produces these beneficial effects by examining the two main omega-3-fatty polyunsaturated acids found in fish oil; docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) and the relative contribution of specific fatty acid receptor subtypes.