Voluntary oral methamphetamine increases memory deficits and contextual sensitization during abstinence associated with decreased PKMζ and increased κOR in the hippocampus of female mice

Abstract

Background

Female populations exhibit vulnerabilities to psychostimulant addiction, as well as cognitive dysfunction following bouts of abuse.

Aims

The goal for this study was to advance our understanding of the mechanisms that produce sex disparities in drug addiction.

Methods

We used an animal model for voluntary oral methamphetamine administration (VOMA) and focused on male and female mice that consumed 7.6 to 8.2 mg/kg of methamphetamine per day during the last 18d of the paradigm.

Results

VOMA-exposed female mice displayed increased locomotor activity in the drug-administration context compared to males, demonstrating sex-specific changes in contextual sensitization. During two weeks of forced abstinence, mice underwent further behavioral testing. We show that abstinence increased open-arm entries on the elevated plus maze, in both sexes. There were no differences in immobility on the tail suspension test. In a hippocampal-dependent radial arm maze task, VOMA treated females, but not males, showed working memory deficits. Hippocampal tissue was collected and analyzed using western blotting. VOMA-exposed female mice exhibited increased kappa opioid receptor (κOR) expression in the hippocampus compared to males, suggesting a vulnerability toward abstinence-induced dysphoria. Female VOMA mice also exhibited a decrease in the memory protein marker, PKMζ, in the hippocampus.

Conclusions

Our study reveals sex-specific effects following abstinence from chronic methamphetamine consumption on hippocampal κOR and PKMζ expression, suggesting that these neural changes in females may underlie spatial memory deficits and identify an increased susceptibility to dysregulated neural mechanisms. These data validate VOMA as a model sensitive to sex differences in behavior and hippocampal neurochemistry following chronic methamphetamine exposure.

1. Introduction

Methamphetamine (MA) use in the United States has increased steadily over the last decade in adolescents and young adults, with 48% of new MA users aged 12–25 years (SAMHSA, 2019). Treatment admissions data show that 13.1% of all admissions in 2011 came from females aged 12–24 reporting MA as the primary substance of abuse, whereas 4.8% of admissions came from males in the same age and drug group (SAMHSA, 2014). Clinical studies show that adolescent females report higher rates of MA abuse compared to adolescent males (Chen et al., 2014). Epidemiological studies have shown that women begin using MA recreationally at a younger age, and transition more quickly to MA addiction (Rawson et al., 2005) which produces greater psychological burden than observed in men (Simpson et al., 2016). This is evident in increased rates of depression, psychosis, and suicide in female populations with a history of MA use (Glasner-Edwards et al., 2008a; Glasner-Edwards et al., 2008b; Mahoney et al., 2010). In men, but not women, drug craving is correlated with depression and anxiety (Hartwell et al., 2016). Furthermore, women have sustained vulnerability to the neurotoxic effects of MA following prolonged abstinence, showing greater reductions in hippocampal volumes and wide-spread reduction in grey matter compared to female controls, whereas no difference was observed in men (Du et al., 2015; Regner et al., 2015). Thus, human studies show that sex differences contribute significantly to the progression of MA addiction, as well as its long-term effects on the brain and on mental health.

Rodent studies have identified sex-differences in MA self-administration (Dluzen et al., 2008; Becker and Chartoff, 2018). Female rats show accelerated MA self-administration (Roth and Carroll, 2004; Kucerova et al., 2009; Reichel et al., 2012) and more rapidly reinstated drug-seeking (Reichel et al., 2012; Cox et al., 2013) compared to males. While few studies have identified signaling mechanisms associated with sex-dependent drug seeking behaviors (Krasnova et al., 2013; McFadden et al., 2014), even fewer studies have identified the neurochemical effects of long-access or chronic MA, particularly on adolescents or young-adults (Bourque et al., 2011; Manning et al., 2016; Manning and Van Den Buuse, 2013; Buck and Siegel, 2015; Teixeira-Gomes et al., 2015). Accordingly, the present study describes the effects of abstinence after chronic MA exposure, by assessing for a battery of emotive and cognitive behaviors, as well as neurochemical changes in hippocampal κOR and PKMζ expression.

Clinical studies report that chronic adult MA users have cognitive deficits in sustained attention, episodic memory, information processing, and impulse control (Nordahl et al., 2003; Monterosso et al., 2005; Simon et al., 2010; Morgan et al., 2012). Animal studies have identified similar cognitive deficits in female rodents with increased MA consumption and escalating MA administration (Reichel et al., 2012; Westbrook et al., 2020). This suggests that female rodents may be more sensitive to the primary rewarding properties of psychostimulants that may disrupt reward driven memory tests. Protein kinase M zeta (PKMζ) is directly implicated in learning and memory (Ling et al., 2002; Tsokas et al., 2016; Pastalkova et al., 2006; Serrano et al., 2008; Serrano et al., 2005; Yao et al., 2008; Opendak et al., 2018; Sacktor, 2011). Our previous studies in female rodents have shown sex differences in radial 8-arm maze (RAM) performance, a hippocampal dependent working memory task, associated with changes to PKMζ expression (Sebastian et al., 2013b). Additionally, our previous work on the effects of MA on spatial learning and memory in mice showed decreases in PKMζ expression in the hippocampus of male mice with a history of MA exposure (Avila et al., 2018; Braren et al., 2014). Therefore, we hypothesize that MA negatively impacts hippocampal dependent spatial memory via associated decreases in PKMζ expression (Avila et al., 2018).

Abstinence from chronic MA exposure has been reported to change locomotor activity in open field tests and is interpreted as both MA-induced drug-seeking behavior and MA-induced anxiety related behavior (Good and Radcliffe, 2011; Qin et al., 2019). This suggests a potential compounding factor of MA exposure on exploratory behavior. We hypothesize that VOMA increases drug-seeking behaviors by increasing κ-opioid receptor (κOR) expression. κORs are an important mediator of stress-induced propensity to reinstate drug administration in rodents(Redila and Chavkin, 2008), and produces potent dysphoric-responses in rodents (Land et al., 2008; Shippenberg and Herz, 1986) as well as humans (Pfeiffer et al., 1986). MA exposure has been shown to increase behaviors associated with drug-seeking, particularly during abstinence (Carati and Schenk, 2011; Miladi-Gorji et al., 2015), an effect facilitated by opioid receptor binding (Svingos et al., 1999) resulting in decreases in dopamine release(Meshul and McGinty, 2000). Drug-induced dysphoric syndrome has been reported with cocaine dependence (Pliakas et al., 2001; Mague et al., 2003; Todtenkopf et al., 2004; Knoll and Carlezon, 2010), and long access to MA (Whitfield et al., 2015). Selective inhibition of κOR blocks the development and/or expression of escalation of intravenous heroin intake (Schlosburg et al., 2013) excessive alcohol self-administration (Walker et al., 2011) and MA self-administration (Whitfield et al., 2015).

Our current study uses a Voluntary Oral MA Administration model of MA addiction, VOMA, which previously revealed a milieu of molecular shifts in the hippocampus of male mice corresponding to spatial-cognition deficits (Avila et al., 2018). Here, we focus on the sex-differences in various behavioral phenotypes present during abstinence following 28 days of VOMA.

2. Methods

2.1. Animals

All experimental conditions, housing, and drug administration procedures were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Hunter College. Male and female C57/Bl6 mice were purchased from Taconic Biosciences and were received at the Hunter College Animal facility at 6 weeks of age. Mice were individually housed, kept on a 12h/12h light/dark cycle from 7:30h to 19:30h, and fed ad libitum for 1 week prior to handling.

2.2. Food restriction

At 7 weeks of age, mice were transitioned from ad libitum food access to restricted chow access. Mice were fed standard chow that weighed 20% of the average mouse body weight (approx. 4g chow for males and 3.5g chow for females per day, given after VOMA sessions). As previously reported (Avila et al., 2018), this feeding schedule assures that mice are controlled for appetite without interrupting weight gain through the VOMA paradigm. To conduct radial arm maze (RAM) assessments, mice were fed 1–2g of mouse chow per day until they reached and maintained 85% of their baseline body weight. Baseline working memory assessments on the 8-arm radial maze took place over 4d. Following RAM assessments, mice resumed daily feeding of chow weighing 20% of their body weight.

2.3. Radial 8-arm Maze Shaping/Baseline

The RAM was used to assess spatial working memory, as described previously (Avila et al., 2018), using Maypo (Homestat Farm, Dublin, OH) as food rewards. Prior to the start of RAM assessments, mice were food restricted for 3 days until they reached 85% of their baseline body weight. Chow intake during this pre-RAM period was controlled by feeding mice 10% of their BW (~1.5 to ~2.5 grams per mouse), at least 16 hours prior to the start of RAM. On day 1 of RAM shaping, mice were placed on the RAM for 5 min (each) to acclimate to the maze and room cues, with food baits available all around the maze and inside food cups. On day 2 of shaping, mice were given 15 min each to explore the maze and the food baits found only at ends of maze-arms and inside cups. Baseline RAM performance was scored on days 3–4, when mice were given 15 min to collect food rewards found only inside submerged cups at the ends of all 8 arms.

2.4. Groups

Mice were assigned into drug and control groups, balanced across treatments using body weights and baseline working memory performance. VOMA groups contained n=14 mice per sex and Control groups (no meth) contained n=6 or n=5 mice, for males and females, respectively.

2.5. Methamphetamine formulation and voluntary oral administration protocol

A stock solution of 40mg/mL of methamphetamine hydrochloride (Sigma Aldrich) was formulated using de-ionized water. Working-MA solutions were diluted into vanilla-flavored Ensure (Abbot Laboratories), using daily average body weights to calculate MA concentrations. Ensure-MA solutions were prepared containing 0.6 to 4 mg/mL of MA, in order to administer 0.25 to 1 mg/kg of MA in a 7μL volume per MA presentation (MA bait), which we have found is optimal for this administration method. MA presentations were delivered to mice in polystyrene petri dishes. Petri dishes were cleaned with sterile water and sterile cleaning wipes at the end of every MA presentation. Escalation phase of VOMA: mice received 1 dose/day at 0.25 mg/kg MA per bait for 3 days; 4 doses/day at 0.25 mg/kg MA per bait for 3 days; 16 doses/day at 0.25 mg/kg MA per bait for 2 days; and 16 doses/day at 0.5 mg/kg MA per bait for 2 days. From day 11 to day 28 of VOMA: mice received 16 doses/day at 1 mg/kg MA per bait. During VOMA, MA presentations were delivered during a 4h time window. For experimental days involving 16 MA presentations, baits were delivered at 15 min intervals. At the end of every 15 min interval, visual inspection of the petri dish allowed us to track voluntary consumption of the MA baits. For each MA presentation that was delivered, consumption was recorded and tracked throughout the 28d administration period. Control mice had vehicle (Ensure presentations only; 7μL/bait) delivered at the same rate as MA treated mice.

2.6. Contextual Sensitization Behaviors

On days 19, 21, and 27 a subset of VOMA mice (High and Low; n=4 Female, n=6 Male) were videotaped in the drug administration context immediately prior to the start of VOMA to assess for enhanced sensitization. Mice were brought into the VOMA testing room and placed into their respective drug-administration cages. Videos of each sampled mouse cage lasted 3 minutes each. Scorers blind to the experimental conditions tracked the amount of time mice were mobile, at rest, or when they displayed nesting/grooming behavior, throughout the duration of the videos.

2.7. Tail Suspension

After 5 or 6 days of abstinence from 28d VOMA, mice underwent a tail suspension test to assess for dysregulation of escape behavior under threat conditions. Mice were hung by their tails and suspended three feet (approximately 92cm) in the air above a bench to heighten threat responses. Mice were kept suspended for six minutes and videos were recorded of their behavior. Video scorers were blind to the experimental conditions. Scorers quantified the time each mouse spent immobile during the trial. The criteria for immobility was any time a mouse exhibited no voluntary taxis or body movements, including of the torso, limbs, or head.

2.8. Elevated Plus Maze

After 7d of abstinence from VOMA, mice were placed on the elevated plus maze (EPM) to assess anxiety behaviors, and dysregulated ambulatory and novel seeking behaviors. EPM assessments were carried out in a testing room located adjacent to the testing room used for VOMA. Lighting conditions on and around the EPM inside the testing room were kept consistent for all mice. Lux was maintained high (average 200 lux on and around EPM) to measure for anxiolytic effects of chronic MA exposure (Bailey and Crawley, 2009). Following a 15-minute acclimation to the testing room, each mouse was placed onto the center of the EPM, facing an open arm opposite from the experimenter. The mouse was allowed to explore the EPM for 8 minutes. Videos were recorded for each mouse. Video scorers were trained and tested for reproducibility and were blind to the experimental conditions. Scorers quantified the number of open- and number of closed-arm entrances for each mouse, as well the time spent in each arm type. Arm entries were counted when all four mouse paws were in the last 3/4-length of an arm.

2.9. RAM Working Memory Assessment

Post-VOMA working memory assessment occurred 14d after completing VOMA as previously reported (Avila et al., 2018). Mice were tested over 2 days (1 trial/day). Each trial started with all food cups baited. To begin each trial, mice were confined for 30s to the center platform with a plastic cylinder. The sequence of arms entered to retrieve the food rewards from all 8 arms was recorded. To prevent a non-spatial strategy, chaining, mice could collect baits from up to 3 sequential arms before the experimenter interrupted the chaining strategy. Working memory errors were recorded as re-entries into arms where the food reward had already been collected. The maximum latency for each trial was set at 15 min.

2.10. Tissue sample collection and fractionation

Following 28d of VOMA and 15 days of abstinence, animals were euthanized and hippocampi from all animals were collected. Whole hippocampi were flash frozen on dry ice and stored frozen at −80°C, until fractionated into cytosolic and synaptic samples as previously reported (Avila et al., 2018; Braren et al., 2014). Hippocampi were immersed in a homogenization buffer (Tris 50 mM; EDTA 1 mM; EGTA 1 mM), composed of a SigmaFast protease inhibitor cocktail (Sigma Aldrich) which also contained AEBSF (2 mM), Phosphoramidon (1 mM), Bestatin (130 mM), E-64 (14 mM), Leupeptin (1 mM), Aprotinin (0.2 mM), and Pepstatin A (10 mM). Homogenization was accomplished using 200 μl of TEE-homogenization buffer and a motorized pestle. Homogenates were transferred to Eppendorf tubes and centrifuged at 3,000g for 5 min at 4°C. The resulting supernatant underwent a secondary centrifugation at 100,000g for 30 min at 4°C. From this second centrifugation step, the supernatant was collected, labeled and stored as the cytosol. The pellet that remained was resuspended in 100μl TEE buffer that also contained 0.001% Triton X-100 and incubated on ice for 1h before proceeding. This resuspended pellet was then centrifuged once more at 100,000g for 1h at 4°C. The resulting pellet from this third centrifugation step was suspended in 50 μl of TEE buffer, labeled and stored as the synaptic fraction (Nogues et al., 1994; Zanca et al., 2019). We proceeded to use the bicinchoninic acid assay (BCA) (Thermo Scientific, Rockford, IL) to determine protein concentration for each cytosolic and synaptic sample, in order to prepare equivalent loading preparations for western blotting. Samples were finally reduced with 4x Laemmli sample buffer equivalent to 25% of the total volume of the sample, boiled and stored frozen at −80°C until used for western blotting.

2.11. Protein quantification and western blot assessments

Synaptic and cytosolic samples (20 μg) were loaded onto a Tris/Glycine 4–20% midi gel to resolve gapDH (36 kDa), PKCι/λ (68 kDa), PKMζ (55 kDa), and κOR (detected at 58 kDa; predicted 43 kDa). Every gel contained 3–4 lanes loaded with the same control sample, (all brain samples, ABS). ABS were used to standardize protein density signals between gels. Proteins on gels were transferred to nitrocellulose membranes using the IBlot® Dry Blotting System (Life Technologies; Carlsbad, CA) using the 9-minute preset program. Nitrocellulose membranes were then incubated in a solution containing 5% sucrose, Tris Buffered Saline with Tween-20 (TBST; 0.1% Tween-20 in TBS) for 30 minutes at room temperature on a rocker. Samples were incubated with the following primary antibodies for 18–36hrs at 4°C: gapDH (1:2000; Abcam Inc., Cambridge, MA, USA), PKMζ/ PKCι/λ (1:2000; Santa Cruz Biotechnology, Santa Cruz, CA), and κOR (1:2000; Abcam Inc., Cambridge, MA, USA). Then, membranes were washed in TBST for 20 min and probed with Horseradish Peroxidase (HRP) conjugated secondary antibody. Membranes were incubated with Enhanced Chemiluminescence (ECL) substrate and exposed on CL-XPosure Film (Thermo Scientific; Rockford, IL). GapDH was used as a control to standardize for protein concentration loaded on gels. Films were analyzed for protein densitometry with ImageJ (NIH).

2.12. Statistical analyses

For MA consumption and behavioral assessments, ANOVAs were used to analyze differences between sex and treatment and/or across time, where appropriate, using GraphPad’s Prism Ver. 8 (La Jolla, California). Western blot data was analyzed using independent t-tests. Post-hoc analyses used controlled Bonferroni-corrected t test comparisons. Where noted, control data was used to normalize VOMA behavior and western blot data to a % of the average control, within each sex.

3. Results

3.1. Experimental Timeline and Contextual Sensitization Behaviors

A timeline depicts our paradigm and the progression through each in-vivo assessment [Fig1A]. A subset of mice was evaluated for behaviors in the administration context [Fig1B]. Immediately prior to the 4hr administration period, on days 19, 21, and 27 of the 28d VOMA paradigm, mice were videotaped for 3 min prior to the start of the VOMA session, in their respective VOMA or vehicle-Ensure administration cages, containing a wire-cage top and liner-paper. We focused on mobility behaviors (running and jumping), nesting & grooming, and resting. A two-way ANOVA of mobility time [Fig 1C] between male and female VOMA mice shows a significant effect of sex (F_1,8=23.09, p<0.001, **S), no significant effect of day (F_2,16=1.10, p=0.36) and no significant interaction effect of sex by day (F_2,16=0.48, p=0.94). Post-hoc analyses show a significant difference between sexes on day 21 and 27 (p<0.05). Analysis of nesting/grooming time [Fig 1D] shows a significant effect of sex (F_1,8=7.33, p<0.05, *S), no significant effect of day (F_2,16=0.67, p=0.53), and no significant interaction effect of sex by day (F_2,16=0.05, p=1.0). There were no significant effects for resting time [Fig 1E, sex (F_1,8=1.84, p=0.21), day (F_2,16=1.72, p=0.21), interaction (F_2,16=0.36, p=0.70)]. Similar analyses were undertaken for male and female control mice, in their respective control Ensure-only administration context. Analyses of control-mice mobility time [Fig1F] revealed an overall effect of day (F_{1.7, 12.1}=6.67, p<0.05, *D), with no significant effect of sex nor interactions. Analyses of time spent nesting/grooming [Fig1G] and resting [Fig1H] in the same context revealed no significant effects of day, sex, or interactions.

Figure 1.

The Escalation voluntary oral methamphetamine administration (VOMA) model increases contextual sensitization in female mice prior to the start of VOMA sessions. The paradigm lasted 8 weeks, with acclimation and shaping during weeks 1–2; VOMA during weeks 3–6; behavioral assessments during weeks 7–8. Days on which sensitization videos were captured for behavioral analyses are marked with an “x” on the timeline (a). VOMA was carried out in standard mouse cages that included a wire-cage top and a paper liner. Prior to the start of each VOMA session, mice were acclimated to their VOMA cage for 15 min before starting of the VOMA session (b). On study days 33, 35, and 41, videos spanning 3 min during the acclimation period were captured of mice in their respective VOMA cages. Female VOMA mice displayed more anticipatory mobility compared to male mice (**S; (c)) with significant differences on days 35 and 41 (*; (c)). There is a significant overall difference in nesting and grooming behaviors between sexes (*S; (d)), with no significant sex differences in resting behavior (e). Control mice showed no sex differences in anticipatory mobility (f), nesting and grooming (g), nor resting (h), behaviors, but a significant overall effect of day in mobility (*D; (f)). D: overall effect of day; EPM: elevated plus maze; RAM: radial arm maze; S: overall effect of sec; TS: tail suspension; WM: working memory. *p < 0.05. **p < 0.01

3.2. Voluntary Oral Methamphetamine Consumption

A total of 28 mice began the VOMA paradigm and 25 mice completed the VOMA paradigm (n=14 males, n=11females). Results show male and female VOMA mice consumed at either a high consumption rate (n=18 mice; n=9 per sex) or a low consumption rate (n=7 mice; n=5 male and n=2 female). Consumption groups were determined based on the average MA consumption per day (mg/kg MA/day) over the last 3d of VOMA, when we expected for high-MA preferring mice to exhibit sustained consumption above a certain threshold. Mice that consumed less than an average 2.0 mg/kg MA/day on the last 3d of VOMA were considered a low consumer. Mice that consumed more than an average 2.0 mg/kg MA/day on the last 3d of VOMA were considered a high consumer. Our initial analyses compared low and high MA consumers across sex over days 1–10 and days 11–28 [Fig2A], where a two-way ANOVA reveals a significant overall effect of day (F_1,23=85.46, p<0.001, ***), consumer group (F_1,23=33.27, p<0.001) and a significant overall interaction effect of day and consumer group (F_1,23=32.97, p<0.001). Post-hoc analyses revealed that high consumers consumed significantly more MA on d11–28 compared to d1–10 (t₂₃=14.16, p<0.001, #) and more than low consumers on d11–28 (t₄₆=8.13, p<0.001, ^). We next determined whether there are significant differences in VOMA consumption between high MA male vs high MA female consumers [Fig2B]. Analyses show that VOMA females consume 6.01 mg/kg/day over 28d (+/−0.62 SEM) and males consume 5.811 mg/kg MA/day over 28d (+/−0.59 SEM). Over the last 18d of the 28d VOMA paradigm, high male mice consume an average of 7.68 mg/kg MA/day (+/− 0.31 SEM) and high female mice consume an average 8.2 mg/kg MA/day (+/−0.12 SEM). A two-way repeated measures ANOVA of these data reveal a significant overall effect of day (F_1,16=184.1, p<0.001, ***) but no significant effect of sex (F_1,16=0.079, p=0.782) nor a significant overall interaction effect of day and sex (F_1,16=0.55, p=0.47). Post-hoc analyses revealed that each sex within this high consumption group consumed significantly more MA on d11–28 than on d1–10 (males: t₂₃=5.8, p<0.001, #; females: t₂₃=6.34, p<0.001, #). Additionally, we analyzed daily MA-intake across the 28d paradigm using a 2-way repeated measures ANOVA (data not shown). This analysis shows no significant overall effect between sexes (F_1,16=0.08, p=0.782), but both show a significant overall effect of day (F_27,432=56.6, p<0.001) and a significant overall interaction effect of sex and day (F_27,432=1.82, p<0.01). Finally, n=6 male and n=5 female mice underwent vehicle exposure (Ensure only; 7μL/presentation) with presentations occurring at the same rates as VOMA mice. Over the course of the 28d VOMA, control mice ate all Ensure presentation (data not shown). There were no differences in Ensure consumption between control male and control female mice.

Figure 2.

The Escalation voluntary oral methamphetamine administration (VOMA) model produces equivalent MA consumption in male and female mice and reveals consumption preferences. Analyses of consumption revealed two different consumer populations, high-consuming and low-consuming mice (a). High-consuming male and female mice show equivalent consumption during each VOMA phase (b). Body weights were tracked across all weeks, and there were no appreciable differences between high and low consumers (c) and no differences in BW between high-consuming male and female mice (d). #p < 0.001 between days 15–24 and days 25–42. ^p < 0.001 difference between high and low on days 25–42. ***p < 0.001.

3.3. Body weights

Figure 2C shows analyses of daily body weights for high and low MA consumers across 8 weeks of the study normalized as a percent of their controls. A two-way ANOVA of these data reveal no significant effects of consumer type (high vs low; F_1,23=0.028, p=0.87) and week (weeks 1–8; F_2.998,68.95=1.62, p=0.192), but a significant interaction effect of consumer type by week (F_7,161=2.068, p=0.05) with no significant post-hoc differences. Figure 2D shows analyses of body weights for high male and female VOMA consumers across 8 weeks of the study. A two-way ANOVA of these data reveal no significant effect of sex (F_1,16=3.55, p=0.08), week (F_3.04,48.57=2.11, p=0.11), nor an interaction effect of sex by week on body weights (F_7,112=0.87, p=0.54). Importantly, food restriction for all treatment groups throughout the paradigm (fed 20% of BW, daily) as well as for RAM assessments (fed 10% of BW, 3d until reached 85% baseline BW) did not result in significant differences in BW between VOMA and control mice at any timepoint in the study.

3.4. Elevated Plus Maze and Tail Suspension Assessments

Ambulatory and novel seeking behaviors were evaluated on the elevated plus maze (EPM). A two-way ANOVA of closed-arm entries [Fig 3A] reveals a significant effect of sex (F_1,25=5.67, p<0.05, *S), no effect of drug (F_1,25=0.039, p=0.86), and no interaction (F_1,25=0.5, p=0.47). Post-hoc analyses reveal a significant difference between male and female VOMA mice (t₂₅=2.53, p<0.05,*). Analysis of open-arm entries [Fig 3B] shows no significant effect of sex (F_1,25=1.35, p=0.2), a significant effect of drug (F_1,25=5.96, p<0.05, *) and no significant overall interaction effect of sex by drug (F_1,25=0.01, p=0.94). Analyses of time spent in each arm type on the EPM (as a % of control) [Fig3C], reveal a significant overall effect of arm type (F_1,16=8.34, p<0.05, *), with no effect of sex nor interaction. Post-hoc analyses reveal that male VOMA mice show significant changes in time spent in open arms compared to closed arms (t₁₆=2.83, p<0.05, #). These results show that VOMA increased ambulatory behavior in females (increased closed arm entries) whereas both sexes exhibited increased exploratory behavior (i.e., increased open arm entries) following VOMA.

Figure 3.

Abstinence following voluntary oral methamphetamine administration (VOMA) increases ambulatory behavior on the EPM. Female VOMA mice are more ambulatory in closed arms versus male VOMA mice (a) and VOMA mice exhibit more entries into open arms on the EPM compared to controls (b). VOMA increases time spent in open versus closed arms, and male mice spend significantly more time in open arms (c). Escape behavior during threat assessment (tail suspension) reveals no significant differences between sexes (d). *S: overall sex effect. *p < 0.05, *p < 0.01, #p < 0.01 between closed and open arms.

The effects of VOMA on threat-induced escape behaviors were assessed by the tail suspension test in high male and high female MA consumers. Total time immobile during the tail suspension test [Fig3D] did not differ between VOMA males and females (t₁₆=1.05, p>0.05). This result indicates that Escalation VOMA had no effect on dysregulating threat-motivated escape behaviors associated with a learned helplessness behavioral phenotype.

3.5. RAM Working Memory Assessment

Working-memory performance on the RAM was analyzed across bait pairs, i.e. arms visited to collect the first pair of baits (baits 1–2), second set of baits (baits 3–4), third set of baits (baits 5–6) and fourth set of baits (baits 7–8). Within each bait-pair sequence, the number of correct arms (n=2) was divided by the total number of arms visited to collect the bait pair. WM performance is shown as a percent of sex-specific control groups. A two-way repeated measures ANOVA of WM performance between sex and bait pairs [Fig 4A] reveals a significant effect of bait pair (F_3,48=7.57, p<0.001, *Bp), no effect of sex (F_1,16=2.96, p=0.10), and no interaction (F_3,48=1.84, p=0.15). Post-hoc analyses show males perform significantly better on bait pairs 5–6 vs females (p<0.05, *). Analysis of WM errors across each bait pair [Fig4B] show a significant effect of bait pair (F_3,38=6.58, p<0.001, ***Bp), a significant effect of sex (F_1,16=4.67, p<0.05, *S), and no interaction effect. Analyses of total number of working memory errors on the RAM [Fig4C] reveal a significant difference between sexes, with Females committing more WM errors compared to males (t₁₆=2.49, p<0,05, *). These results highlight that Escalation VOMA negatively affects working memory differentially between sexes.

Figure 4.

Abstinence following VOMA increases RAM memory deficits in female mice. Short-term working-memory performance on the RAM during abstinence identified an overall effect of bait pair (Bp; (a and b)) with male mice performing better than female mice for bait pair 5–6 (*; (a)). A similar analysis of the difference in WM errors in each bait pair also reveals an overall effect of sex (*S; (b)). Total WM error in the whole 8-bait assessment shows that female VOMA mice commit more errors compared to male mice (c). RAM: radial arm maze; VOMA: voluntary oral methamphetamine administration; WM: working memory. *Bp: Overall effect of bait pair; *S: overall sex effect. *p < 0.05. ***p < 0.001.

3.6. Hippocampal PKMζ, PKCι/λ, and κOR

Hippocampal cytosolic fractions were analyzed for expression of atypical protein kinase c isoforms, PKMζ and PKCι/λ, which are known to play a role in learning and memory (Hsieh et al., 2016; Tsokas et al., 2016) in order to examine the underlying mechanisms for VOMA-induced spatial memory deficits. Figure 5A shows a significant difference in cytosolic PKMζ expression between sexes (t₄=4.82, p<0.001, ***) with females showing significantly lower levels of PKMζ compared to males. There was no significant difference in PKCι/λ expression between sexes (t₁₃=0.70, p>0.05, Fig 5B). κOR is known to be involved in dysphoria and the negative reinforcement associated with withdrawal (Chavkin et al., 1982; Land et al., 2008; Chavkin and Koob, 2016). Analysis of κOR expression in hippocampal synaptic fraction shows increased expression in females compared to males (t₉=5.408 p<0.001; Fig 5C, ***).

Figure 5.

Hippocampal neurochemistry during abstinence following VOMA. Hippocampal PKMζ expression is significantly lower in female VOMA mice compared to male mice (a), with no differences in PKC ι/λ (b). Female VOMA mice show increased κOR compared to male mice (c). Representative blots are shown for two animals per group. κOR: kappa-opioid receptor; PKMζ: protein kinase M zeta; VOMA: voluntary oral methamphetamine administration. ***p < 0.001.

4. Discussion

Our study investigated the effects of chronic voluntary oral methamphetamine administration (VOMA) in young-adult male and female mice (6–12 weeks of age). We focused our analyses on male and female high VOMA consumers. Our results show sex-specific effects on behavior and on neurochemical changes in the hippocampus. Behavioral data show that during abstinence from VOMA, male and female mice increased exploratory behavior on the EPM. Increased κOR levels in female VOMA hippocampi compared to males suggests abstinence after VOMA may produce an increase in stress and/or dysphoria that promotes context-dependent exploratory behavior. Additionally, females exhibit increased spatial working-memory deficits and decreased levels of hippocampal PKMζ expression compared to males. These results highlight the female-specific behavioral vulnerability of chronic VOMA and the differential neurochemical shifts that occurs between sexes.

4.1. 28d of VOMA reveals differential MA-preferences in mice but does not induce body weight changes

The present study found that 28d of VOMA, comprised of 10d with escalating doses and 18d with static doses, produced two distinct consumption groups in our mice. A similar finding of high and low MA consumers in both males and female rats has recently been reported using an operant self-administration paradigm (Daiwile et al., 2019). In our VOMA study, Low consumers consumed on average 2.09 mg/kg MA/day (+/− 0.44 SEM) over the last 18d of VOMA, whereas high consumers consumed on average 8.11 mg/kg MA/day (+/− 1.24 SEM) over the last 18d of VOMA. Previous studies have found that self-administration of MA gradually increases over a one-week period at the same average rate in rats (Hamor et al., 2018; Krasnova et al., 2010; Rogers et al., 2008), but these reports failed to describe individual differences or MA-preferences in these animals. Recently, an emerging body of work has reported potential genetic contributions to MA-preferences using a drinking model (Shabani et al., 2016). Additionally, other labs have found correlations between MA-seeking lever pressing and orexin receptor expression in female rats with variable preference for the drug (Daiwile et al., 2019). Thus, the finding that our VOMA paradigm reveals MA-preferences in both male and female mice supports the findings from this recent body of work. Finally, the VOMA model strengthens our understanding of MA-preferences by mimicking human MA-abuse patterns that encompass both chronic users that administer variable doses of MA per day and binge users that administer variable dosages of MA over multiple days (Cho and Melega, 2001).

Mice in both high and low consumption groups did not lose weight compared to control mice throughout the 28d VOMA paradigm. The daily consumption rates observed in our study may not be large enough to induce body weight loss throughout the paradigm as seen in previous MA administration models. Animals that self-administer MA at doses of 12–15 mg/kg displayed weight loss compared to saline control animals over a period of one week (Krasnova et al., 2010). Additionally, animals treated with escalating doses of MA from 0.5 mg/kg to 2.0 mg/kg displayed increases in weight not dissimilar to saline control animals over a two-week period and significant body weight loss was only apparent when treated with challenge doses of 60 mg/kg, administered at a rate 5mg/kg per hour (Krasnova et al., 2011). It is possible that higher daily MA consumption rates in our mice would produce shifts in body weights of VOMA mice. Future studies will aim to increase the voluntary daily MA consumption rates using VOMA to determine the effect of this paradigm on body weight changes.

Our data reveal no significant sex-differences in total MA consumed across the 28d VOMA paradigm, although female mice trend to higher levels of MA consumed. However, analyses of daily MA intake rates reveal an interaction effect of sex and day. Thus, access to MA and the methods used to report MA consumption likely play a role in the sex differences of human MA intake reported in the literature. Future studies should investigate the role of different schedules of oral MA-access to produce sex differences in daily and total consumption.

4.2. VOMA increases ambulatory behavior and induces sex-specific upregulation of hippocampal κOR

We observed increases in exploratory behavior on the EPM in VOMA mice, particularly in females. MA addiction has been shown to induce anxiogenic effects during withdrawal in rodents as well as increased drug seeking behaviors (Carati and Schenk, 2011; Miladi-Gorji et al., 2015), suggesting that VOMA may also exacerbate these processes. It is unclear if our data of open arm exploration on the EPM reflect anxiolytic behavior, resulting from decreased anxiety to explore exposed arms, or anxiogenic behavior, arising from a drug-seeking and addiction-like endophenotype. These data may reflect conflict resolution in the context of addiction. Conflict resolution between basic needs and a stressful task is very characteristic of a transgenic mouse model of Down Syndrome where performance on a motor learning task was investigated in thirst deprived mice (Hyde et al., 2001). Our results may reflect a conflict in our animals between drug-seeking and anxiety-inducing conditions thus assessing what an animal is willing to do in the context of addiction when faced with discomfort. Additionally, runway models of drug self-administration, where rodents traverse modified maze alleys to receive drug injections, have previously been used to characterize motivational processes in addiction (Ettenberg, 2009; Pandy and Khan, 2016). These models indicate that drug-seeking behaviors can be assessed in modified behavioral assays using mazes. Furthermore, our EPM assessments were conducted in an adjacent testing room to where VOMA was conducted, thus potentially tapping into spatial and contextual cues that primed VOMA mice to expect MA within the EPM task. This interpretation is supported by our data showing that female mice exhibited greater anticipatory mobility in the administration context. These data of contextual behaviors reveal that male VOMA mice spend significantly less time in mobility in VOMA cages prior to the start of VOMA sessions, and opt to groom, nest, and rest [Fig1 C–E], while females show consistent increased time in mobility [Fig1C]. Future studies will assess the effects of VOMA on home-cage behaviors, particularly nesting, as previous work has shown that MA can manifest deficits in these behaviors differentially across sexes (Jacobskind et al., 2019).

Our results show that VOMA females have higher levels of κOR in the hippocampus compared to males during abstinence. The expression of κOR in the hippocampus may reflect a NAc-dependent circuit associated with the negative affect of drug withdrawal and with compulsive drug-seeking behaviors (Wee et al., 2009; Whitfield et al., 2015). Our results are consistent with studies showing that rodents with a high-preference for and long access to oxycodone exhibit higher protein levels of κOR in the hippocampus (Blackwood et al., 2019), specifically after a prolonged abstinence period. This supports our hypothesis that κOR may underlie drug-seeking during abstinence from VOMA. Future studies that investigate changes in κOR expression in regions like the NAc and mPFC during abstinence from VOMA will aid in understanding how chronic MA changes neural signaling during abstinence-induced drug-seeking. Furthermore, assessing how κOR expression changes at different timepoints within the VOMA paradigm and at additional abstinence time points may allow for a better understanding of how this molecular pathway is modulated across different phases of addiction. We hypothesize that VOMA has downstream consequences on κOR activation via the dopamine D1 receptor which regulates prodynorphin, the precursor of dynorphin, the endogenous ligand for κOR (Chavkin et al., 1982). Our results suggest that females may have a susceptibility to the dysphoria-inducing effects of κOR activity. This hypothesis is consistent with previous studies (Willuhn et al., 2014) demonstrating that escalation of cocaine intake is driven by decreases in phasic dopamine release in the ventral striatum and that normalization of this state by the dopamine precursor L-dopa can eliminate the escalation of cocaine intake. Thus, κOR acts as a presynaptic regulator of dopamine release, and, when the κOR signaling system becomes sensitized by repeated psychostimulant use, it creates a state of deficient presynaptic dopamine release altering not only the rewarding properties of drugs, but also of the food reward earned during spatial learning. Thus, we cannot rule out the possibility that κOR expression in the hippocampus shifts reward salience and drives the appetitive spatial memory deficit.

4.3. VOMA impairs working memory and induces sex-specific downregulation of hippocampal PKMζ

The present study shows working memory deficits in VOMA females but not in males during abstinence. It is well documented that MA produces cognitive deficits following static doses or neurotoxic doses of MA (Avila et al., 2018; Braren et al., 2014). However, escalating doses of MA as we use here have been reported to be neuroprotective compared to chronic or bolus challenges of MA (O’Neil et al., 2006; Segal et al., 2003). Thus, our present results are consistent with the neuroprotective effects of escalating doses of MA on male mice, while promoting vulnerabilities in female mice. While some rodent studies show that both males and females exhibit memory deficits after chronic MA exposure, these studies consistently show that females have higher MA intake, greater relapse to MA-seeking (Reichel et al., 2012) and show enhanced sensitivity to MA (Roth and Carroll, 2004; Roth et al., 2004; Reichel et al., 2012).

Our results show that PKMζ was significantly reduced in female VOMA mice compared to male VOMA mice during abstinence. These results indicate that synaptic plasticity marker expression was differentially affected between sexes to produce the working memory deficit in female VOMA mice. PKMζ and PKCι/λ are synaptic plasticity markers associated with learning and memory (Sacktor, 2011; Opendak et al., 2018). PKMζ is important for both late-phase LTP (Serrano et al., 2005) and long-term memory maintenance across various learning paradigms (Pastalkova et al., 2006; Serrano et al., 2008; Sacktor, 2011; Hsieh et al., 2016). PKMζ also plays a major role in the trafficking of AMPA receptors that sustain LTP processes (Ling et al., 2002; Yao et al., 2008). Previously, it has been shown that PKCι/λ acts as a compensatory AMPAr trafficking enzyme in animal models where PKMζ was knocked-out, but in wildtype mice, PKMζ remains a pivotal enzyme regulating AMPAr trafficking, LTP, and memory(Tsokas et al., 2016). Still unknown is how PKCι/λ may undermine or rescue AMPAr trafficking and LTP deficits in models of stress and drug addiction, where PKMζ is consistently affected (Avila et al., 2018; Braren et al., 2014; Sebastian et al., 2013a). The current study shows that PKCι/λ, unlike PKMζ, remains unchanged following VOMA in both male and female mice. Future studies are needed to determine the degree to which PKCι/λ may act to partially rescue AMPAr and LTP deficits brought on by PKMζ-induced deficits after VOMA. Future studies that investigate the plasticity of PKMζ throughout the VOMA paradigm and at various times during abstinence will provide insight into how learning and memory pathways are modulated throughout the various phases of addiction in this model.

5. Conclusion

To determine how chronic access to MA differentially affects the hippocampus across young-adult male and female mice, we used a VOMA model (Avila et al., 2018). The findings show that with 10 days of escalating MA-dosing followed by 18 days of chronic dosing, VOMA females show working memory deficits as well as increased ambulatory behavior in the drug-context and on the EPM compared to males. These behavioral differences are supported by sex-specific neurochemical changes in hippocampal PKMζ and κOR, which may act to promote drug-seeking and dysregulate spatial memory processing. Our results support the utility of the VOMA model in characterizing the neurochemical changes associated with MA addiction and the corresponding behavioral changes that co-occur during abstinence. Delineating these molecular shifts will be pivotal to prevent cognitive- and emotional-processing deficits in substance addiction, that likely promote relapse during periods of abstinence. Our current study provides evidence that opioid and synaptic plasticity signaling may underlie several behavioral deficits that occurs during abstinence from MA abuse and provide targets to remediate these deficits.

9 Data Availability

The raw data supporting the conclusions of this manuscript will be made available upon reasonable request to the corresponding author.