Impaired memory and reduced sensitivity to the circadian period lengthening effects of methamphetamine in mice selected for high methamphetamine consumption

Behavioural Brain Research 256 (2013) 197–204

Contents lists available at ScienceDirect

Behavioural Brain Research journal homepage: www.elsevier.com/locate/bbr

Research report

Impaired memory and reduced sensitivity to the circadian period lengthening effects of methamphetamine in mice selected for high methamphetamine consumption Reid H.J. Olsen a , Charles N. Allen a,b , Victor A. Derkach c , Tamara J. Phillips a,d,e , John K. Belknap a,d,e , Jacob Raber a,e,f,g,∗ a

Department of Behavioral Neuroscience, Oregon Health & Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA Center for Research on Occupational and Environmental Toxicology, Oregon Health & Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA c Vollum Institute, Oregon Health & Science University, 8131 SW Sam Jackson Park Road, Portland, OR 97239, USA d Portland VA Medical Center, Oregon Health & Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA e Methamphetamine Abuse Research Center, Oregon Health & Science University, 8131 SW Sam Jackson Park Road, Portland, OR 97239, USA f Department of Neurology, Oregon Health & Science University, 8131 SW Sam Jackson Park Road, Portland, OR 97239, USA g Division of Neuroscience ONPRC, Oregon Health & Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA b

h i g h l i g h t s

g r a p h i c a l

a b s t r a c t

• MALDR mice show spatial memory retention.

• MAHDR mice do not show spatial memory retention.

• MAHDR mice have more GluR2 AMPA receptor subunits in the hippocampus. • During 25 mg/L MA solution access, there is an increase in  in MALDR mice. • During 50 mg/L MA solution access, both lines show an increased .

a r t i c l e

i n f o

Article history: Received 20 June 2013 Received in revised form 2 August 2013 Accepted 6 August 2013 Available online 14 August 2013 Keywords: Circadian Water maze Methamphetamine Glutamate receptors GluA1/2 Spatial memory

a b s t r a c t Drug abuse runs in families suggesting the involvement of genetic risk factors. Differences in addictionrelated neurobiological systems, including learning and memory and circadian rhythms, may exist prior to developing addiction. We characterized the cognitive phenotypes and the free-running circadian period of mouse lines selectively bred for high methamphetamine (MA) drinking (MA high drinking or MAHDR) and low MA drinking (MA low drinking or MALDR). MA-naïve MALDR mice showed spatial memory retention while MAHDR mice did not. MA-naïve MAHDR mice had elevated hippocampal levels of the AMPA receptor subunits GluA2 (old terminology: GluR2), but not GluA1 (old terminology: GluR1). There were no line differences in the free running period () when only water was available. During a 25 mg/L MA solution access period (vs water), there was an increase in  in MALDR but not MAHDR mice, although MAHDR mice consumed significantly more MA. During a 50 mg/L MA solution access period (vs water), both lines showed an increased . There was a positive correlation between MA consumption and  from baseline in MALDR, but not MAHDR, mice. Thus, a heritable proclivity for elevated MA self-administration may be associated with impairments in hippocampus-dependent memory and reduced sensitivity to effects of MA on lengthening of the circadian period. © 2013 Elsevier B.V. All rights reserved.

∗ Corresponding author at: Department of Behavioral Neuroscience, Oregon Health & Science University, 8131 SW Sam Jackson Park Road, Portland, OR 97239, USA. Tel.: +1 503 494 1524; fax: +1 503 494 6877. E-mail address: [email protected] (J. Raber). 0166-4328/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbr.2013.08.015

198

R.H.J. Olsen et al. / Behavioural Brain Research 256 (2013) 197–204

1. Introduction Methamphetamine (MA) is a pyschostimulant, the abuse of which exerts a great physiological [1], social [2], and economic cost [3]. The increasing impacts of MA abuse are offset by prevention and treatment paradigms, which hinge on an understanding of MA addiction itself. In addicted individuals, consumption of a drug leads to changes in behavior that may result in an increase in drug consumption, and subsequently compulsive drug-seeking and drug-taking behavior, over time [4]. Multiple mechanisms are proposed to be involved. They include alterations in motivational states involving rewarding effects of the drug acting as a reinforcer of the behavior [5] and perturbations of learning and memory processes involved in drug-associated stimuli and hedonic and aversive effects of the drug [6]. In response to these perturbations, drug seeking and drug use might become compulsive by hijacking a component of the learning and memory system as part of an adaptive response [5]. Drug abuse runs in families, suggesting common environmental and/or genetic risk factors [7]. Differences in addiction-related neurobiological systems might exist prior to developing addiction. Mouse lines selectively bred for high MA drinking (MA high drinking or MAHDR) and low MA drinking (MA low drinking or MALDR) under two-bottle choice (MA vs water) conditions provide a unique resource for studying pre-existing genetic differences [8,9]. These lines have been validated as a model of genetically-determined differential MA reinforcement, reward and aversion. This validation is based on greater operant oral and intracranial self-administration of MA in MAHDR compared to MALDR mice, MA conditioned place preference that is present in MAHDR mice but completely absent in MALDR mice, and MA conditioned taste aversion that is present in MALDR mice but completely absent in MAHDR mice [8–11]. Other behavioral phenotypes that co-segregate with selection for MA consumption may offer additional insight into common neural systems that play a role in drug abuse and addiction. Based on the putative involvement of learning and memory systems in drug addiction [12] and the impairments in learning and memory associated with MA use [13], we hypothesized that the MAHDR and MALDR lines would differ in cognitive performance. As measures of anxiety and exploratory behavior can potentially affect performance on cognitive tests, they were assessed as well. Substance abuse is associated with sleep disturbances [14] and exposure of the developing brain to MA increases the length of the circadian period in adulthood [15]. Therefore, we hypothesized that the free-running circadian period would be altered following access to a solution containing increasing MA concentrations as part of a two-bottle choice (vs water) condition. We further hypothesized that sensitivity to this disruption could be genetically related to voluntary MA consumption. Finally, we hypothesized that in MA-naïve mice of the MA drinking lines, differences found in cognitive phenotype between the lines would be associated with line differences in hippocampal levels of proteins described below that are known to play an important role in hippocampus-dependent cognition. Expression of glutamate receptors 1 (GluA1; old terminology: GluR1) and 2 (GluA2; old terminology: GluR2) were quantified because of their involvement in circadian rhythmicity [16], the reward system [17], learning and memory [18], MA exposure [19], and addiction [20].

ethanol sensitivity and the alcohol preferring and non-preferring selected lines are examples). The goals for short-term lines are rapid production, and replacement by producing consecutive replicates to follow up interesting initial findings. This is a more feasible approach to limited resources and space for long-term maintenance. Selection for each set of lines was from a population of 120 mice (half of each sex) from the F2 cross of the C57BL/6J and DBA/2J inbred strains. These mice were offered 20 mg and subsequently 40 mg MA HCl/L in water ((+)-MA hydrochloride (Sigma, St. Louis, MO, USA) vs tap water, 18 h each day for 4 consecutive days per concentration. We found that access limited to 18 h each day increases MA intake, but does not result in significant weight loss (Phillips, unpublished data). Selection was based on average consumption (in mg/kg) of the 40 mg/L MA solution. Mass selection was used, which maximizes the response, and is the preferred method for short-term selections in which lines are terminated after 4–5 generations to avoid high rates of inbreeding that could result in the fixation of alleles not associated with the selection trait if selection was continued for additional generations. With mass selection, animals are chosen for breeding that have either the highest or lowest scores for the trait of interest, regardless of what families they belong to; however, mating of animals with common relatives is avoided to reduce inbreeding. Thus, the male and female mice with the highest MA intake were interbred to form the MAHDR (MA high drinking) line, and those mice with the lowest intake were bred to establish the MALDR (MA low drinking) line. Subsequent generations of mice were similarly tested and interbred. In both selections, maximal divergence was associated with intake of about 6 mg MA/kg/18 h in the MAHDR lines and close to 0 intake in the MALDR lines [8,9]. The mice used in the current study were MA-naïve second replicate line mice from the 5th selection generation showing maximal divergence [9]. They did not differ in age. Group sizes were MAHDR; male: N = 9 mice; female: N = 13 mice; and MALDR; male: N = 8 mice; female: N = 8 mice. Mice were maintained on a 12 h light/dark schedule (lights on at 06:00) and laboratory chow (PicoLab Rodent diet 20, # 5053; PMI Nutrition International, St. Louis MO, USA) and water were provided ad libitum. Behavioral testing took place during the light phase. All procedures complied with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the IACUC at Oregon Health & Science University. Mice were weaned and singly housed on postnatal day 20. As line differences in body weight might contribute to line differences in performance on cognitive tests, a subset of the animals were chosen at random and weighed every 3 days, over a period of 16 days (Male MAHDR, N = 8, Male MALDR, N = 7, Female MAHDR, N = 11, Female MALDR, N = 4) starting on postnatal day 30.50 ± 0.21, and ending on post-natal day 46.5 ± 0.21. Growth curve analysis ran concurrently over the span of behavioral testing. 2.2. Behavioral testing The mice were behaviorally tested beginning shortly after weaning on postnatal day 31.6 ± 0.04 in the following order: exploratory behavior in the open field (16 × 16 in.), analyzed as beam breaks, for 10 min in the morning (day 1); measures of anxiety in the elevated zero maze for 10 min in the morning (day 2); novel object recognition (days 1–5); spatial learning and memory in the water maze (days 8–12); and contextual and cued fear conditioning (days 15 and 16). The object recognition test was performed as described [21]. All other tests were performed as described [22], with the following changes for the fear conditioning training paradigm. On the first day of fear conditioning, the mice were placed inside a dark fear-conditioning chamber. Chamber lights (at 100 lx) turned on at 0 s, followed by a 160 s habituation period and a subsequent 20-s (2800 Hz, 80 dB) tone (cue). A 2-s 0.35 mA footshock was administered at 178 s, co-terminating with the tone at 180 s. Chamber lights remained on for 15 s after CS–US pairing, terminating the trial at 195 s. Before each trial, materials used in the test were cleaned with 5% acetic acid, unless otherwise noted. After cognitive testing, mice were used for circadian testing. Due to limitations in terms of available mice and equipment to simultaneously test mice in a single circadian experiment, male mice were first used for circadian rhythm testing. As initial wheel-running data were not sufficient to calculate the free-running period of these mice, the experiment was aborted before administration of MA and the mice were killed by cervical dislocation for western blot analysis of hippocampal tissue (male MALDR, N = 7 mice, male MAHDR, N = 6 mice). Female mice were used for circadian rhythm testing (female MAHDR, N = 8 mice, female MALDR, N = 6 mice). The decision to subdivide animals by sex at this point was made for a number of reasons. Because line did not interact with sex during the current investigation or in the published literature for MA intake or other reward and aversion-related traits in these lines, we anticipated that such an effect in AMPAR and circadian period was unlikely. For all experiments, the researchers were blinded to the line of the animals.

2. Methods

2.3. Circadian rhythm testing

2.1. Animals

Following cognitive testing, a subgroup of female mice was individually housed in Nalgene cages equipped with running wheels and magnetic switches (Minimitter, Bend, OR, USA) to determine the circadian period. Only a subgroup was used, because of limitations on the number of mice the number of mice that can be simultaneously tested with our equipment. The cages were placed into Intellus Control System chambers (Percival Scientific, Perry, IA, USA) maintained at 21–22 ◦ C on a

Two consecutive short-term selective breeding projects for MA drinking have been completed [8,9]. Short-term selection is an alternative approach to the creation of long-term selected lines, which are generally produced with the goal of maintaining the lines for many years (the long-sleep and short-sleep mouse lines bred for

R.H.J. Olsen et al. / Behavioural Brain Research 256 (2013) 197–204

199

12:12 h light:dark cycle. Wheel revolutions were counted and stored on a computer using Vitalview software (Minimitter, Bend, OR, USA). Data were analyzed and actograms generated using ClockLab (Actimetrics Software, Wilmette, IL, USA) and MATLAB software (Mathworks, Natick, MA, USA). ClockLab is a commercially available product that requires MatLab to run (there is no custom software). Wheelrunning behavior was analyzed under 14 days of entrainment to a 12:12 light: dark cycle followed by a period of constant darkness. Constant darkness is used to determine the endogenous circadian period () of the mice. Following ten days in constant darkness, mice were given access to two tubes, one containing water and the other containing 25 mg/L MA in water for 24 h/day. After six days, the MA concentration was increased to 50 mg/L, in an attempt to obtain a higher self-administered dose. The mean amount of water and water containing MA consumed by each mouse per 24 h was recorded daily and the circadian rhythm periods () were calculated using periodograms and F tests. Data for MA consumption are expressed as average daily intake (per 24 h). The group average was used for  for one MALDR mouse in the two periods during which the mouse did not run consistently. 2.4. Western blot analysis As described above, MA-naïve male mice were killed by cervical dislocation and their brains removed. The hippocampus was dissected and immediately frozen in liquid nitrogen for western blot analysis. Tissue was homogenized in 300 ␮l of RIPA buffer (Pierce Pharmaceuticals, Rockford, IL) containing 10% Halt protease inhibitor cocktail (Pierce). Homogenized tissue was spun at 12,000 g for 15 min, and protein concentrations were determined in the supernatant using the Pierce protein assay. Proteins were denatured by boiling for 5 min at 99 ◦ C in a solution of Laemmli’s buffer containing 5% 2-mercaptoethanol (Bio-Rad, Hercules, CA). Equivalent amounts of total protein were loaded on SDS-PAGE (Criterion Bio-Rad Ready Gels, 4–15% Tris–HCl, 18 well) and run with a Bio-Rad Power Pac for 60 min at 170 V. Proteins were transferred to PVDF membranes for 90 min at 30 V or for 18 h at 18 V (GluA1/2) at 4 C. The membranes were placed overnight in AquaBlock blocking buffer (EastCoast Bio Inc., North. Berwick, ME) at 4 ◦ C. Membranes were washed in PBST buffer (4 × 5 min) and incubated in blocking buffer with one of the following primary antibodies for 12 h at 4 ◦ C: anti-GluA1 1:1000 (Millipore AB1504, Bilerica, MA), anti-GluA2 1:1000 (Invitrogen, Grand Island, NY), or anti-␤-tubulin 1:2500 (E7, Developmental Studies Hybridoma Bank, Iowa City). For GluA1 and GluA2, secondary antibodies to the appropriate antibody species were IRDye800 and IRDye700 conjugated (Rockland, Gilbertsville, PA) and used at 1:5000 dilution. Proteins were quantified using the Odyssey Infrared System (LI-COR Biosciences, Lincoln, NE). To account for possible variations in loaded protein, each band of interest was normalized to tubulin as the protein-to-tubulin ratio.

Fig. 1. Body weights in MAHDR and MALDR (A) and male and female (B) mice. (A) MAHDR and MALDR mice exhibited increases in body-weight. MAHDR mice initially exhibited greater body weights compared to MALDR mice, but this difference disappeared by day 10. (B) Both male and female mice exhibited increases in growth by day (male mice:  = 0.026, F(5,10) = 74.1, p < 0.0001; female mice:  = 0.067, F(5,10) = 27.9, p < 0.0001). The last two days of growth differed between male and female mice, p < 0.05. Because no interaction with line was significant, and because male and female mice exhibit different body mass, this relationship was not pursued further. Within group analyses: ˆˆˆˆp < 0.0001, ˆˆˆp < 0.001, ˆˆp < 0.01, ˆp < 0.05, †p = 0.05 for comparison with body weight on prior day. Between group analysis: ***p < 0.001, **p < 0.01, *p < 0.05.

2.5. Statistical analyses All statistical tests were conducted using SPSS 16.0 (SPSS Inc, Chicago, IL, USA) or Statistica (Tulsa, OK) software and were considered significant at P < 0.05. All figures were generated using GraphPad Prism (GraphPad Software, La Jolla, CA). Data are reported as averages ± the standard error of the mean. If data were not normally distributed or violated an assumption of a statistical test, they were transformed using commonly accepted methods or a non-parametric test was used. Data were analyzed as two-way ANOVAs using sex and line as between-subject variables, unless otherwise indicated. For repeated-measures ANOVAs, if Mauchly’s test of sphericity was not satisfied, multivariate ANOVAs were conducted, reporting Wilk’s lambda (). Unless otherwise noted, Bonferroni corrections were used to control for multiple comparisons. Watermaze learning curves were analyzed using repeated-measures ANOVAs with testing day as a within subjects variable and line and sex as between-subjects variables. For the analysis of probe trial data, one-way ANOVAs were used, followed by one-tailed Dunnett’s post hoc tests if there was a significant effect of quadrant, to assess if the time spent in the target quadrant was higher than any other quadrant. Fear conditioning data were analyzed using multiple paired t-tests. All tests were 2-tailed unless indicated otherwise. Results were considered significant at an ˛ level of 0.05. Circadian data were analyzed as described in [15].

F(5,6) = 24.0, p = 0.001) (Fig. 1A). The growth was linear (MAHDR: F(1,18) = 160.5, p < 0.0001; MALDR (F(1,10) = 112.732, p < 0.0001) with a quadratic component (MAHDR: F(1,18) = 59.4, p < 0.0001; MALDR: (F(1,10) = 46.7, p < 0.0001), suggesting that the growth began to plateau. Initially, MAHDR mice weighed significantly more than MALDR (after a Bonferroni correction, Day 1: p < 0.001; Day 2: p < 0.01; and Day 3: p < 0.05) but this difference decreased over time, and parity between groups was reached by the time circadian period was studied (Fig. 1A). 3.2. Open field

3. Results

There were no significant sex or line differences in activity in the open field, but there was a trend toward higher activity in the MAHDR (2128.86 ± 132.78 cm) than the MALDR (1710.43 ± 171.80) line (F(1,34) = 3.9, p = 0.057). There was no effect of sex or line, nor any interaction, for percentage of time spent in the center of the open field (MAHDR: 8.89 ± 1.25%; MALDR: 9.14 ± 1.47%).

3.1. Growth curves

3.3. Elevated zero maze

Weights increased over time for all mice. A main effect of day was observed ( = 0.021, F(5,22) = 203.7, p < 0.0001), a day by line interaction (Fig. 1A,  = 0.368, F(5,22) = 7.542, p < 0.0001), as well as a day by sex interaction (Fig. 1B,  = 0.214, F(5,22) = 16.1, p < 0.0001). There also was a between-subjects effect of line (F(1,26) = 14.1, p = 0.001) but no main effect of sex or interaction of line with sex. Each line exhibited significant growth (MAHDR:  = 0.047, F(5,14) = 56.9, p < 0.0001; MALDR:  = 0.048,

All groups exhibited similar activity levels in the elevated zero maze, i.e., there was no effect of line or sex on total distance moved. Consistent with the time spent in the center of the open field, there were no effects of line on measures of anxiety in the elevated zero maze. However, there was a main effect of sex on time spent in the open areas of the maze (F(1,34) = 8.0, p = 0.008), with females spending less time (5.790 ± 1.267%) than males (11.155 ± 1.417%). There also was a main effect of sex on nose pokes into the open

200

R.H.J. Olsen et al. / Behavioural Brain Research 256 (2013) 197–204

MALDR: 38.8 ± 5.6 s). Nor was there a significant difference in total time spent exploring both objects on day 5 (MAHDR: 42.2 ± 4.5 s; MALDR: 42.0 ± 3.8 s). There were no line differences in novel object recognition, and mice showed a general preference for the novel object on day 5 (t(37) = −6.991, p < 0.0001) (MAHDR: 64.4 ± 2.7%; MALDR: 62.7 ± 2.9%). 3.5. Water maze There was no effect of line or sex on average velocity during the visible or hidden platform trials. All groups learned to locate the visible platform (effect of day, F(1,31) = 195.8, p < 0.0001), with a decrease in latency to locate the platform from day 1 to day 2 (p < 0.0001), but there were no effects of line, sex, or an interaction between line and sex. During the hidden platform training days of the water maze (days 3–5), all groups learned to locate the hidden platform (effect of day, F(2,62) = 4.6, p = 0.013), with decreases in latency occurring on day 4 compared to day 3 (p = 0.017), but there were no effects of line, sex or line by sex interaction (Table 1). When spatial memory retention was assessed in the probe trial following the first day of hidden platform training, neither line showed a spatial bias for the target quadrant (data not shown). Striking line differences were seen, however, in the probe trials following the second and third day of hidden platform training (Fig. 1B and C). MALDR mice showed spatial memory retention in the second (Fig. 1B, effect of quadrant; F(3,64) = 12.5, p < 0.0001) and third (Fig. 1C, effect of quadrant; F(3,64) = 9.7, p < 0.0001) probe trials and spent more time in the target quadrant than any other quadrant. In contrast, MAHDR mice did not show spatial memory retention or even a trend toward spatial memory retention in the second or third probe trial. 3.6. Fear conditioning

Fig. 2. (A) MAHDR and MALDR mice show robust novel object recognition. ****p < 0.0001. (B) Spatial memory retention in the probe trials following the second (Probe 2) day of hidden platform training. (C) Spatial memory retention in the probe trials following the third (Probe 3) day of hidden platform training. MALDR mice showed a preference for the target quadrant compared to all other quadrants in the second and third probe trials (effect of quadrant, #p < 0.0001; quadrant differences: ***p < 0.0001, **p < 0.01, *p < 0.05 versus Target quadrant. T, target; R, right; L, left, and O, opposite.

areas, with males (Md = 3.00, n = 17) showing more nose pokes than females (Md = 1.00, n = 21), U = 107.00, z = −2.187, p = 0.029. 3.4. Novel object recognition Novel object recognition was assessed next (Fig. 2A). There were no significant differences in total exploration time or percent time spent exploring either object on day 4 (MAHDR: 47.5 ± 5.0 s;

To exclude potential effects of anxiety-like or locomotor behavior contributing to group differences in freezing behavior, motion during the 160-s habituation period was analyzed first. There were no line or sex differences in baseline motion. There were also no line differences in response to the shock, but female mice exhibited lower motion in response to the shock than male mice (F(1,36) = 4.3, p = 0.046). There were no line or sex differences in freezing during the initial tone presentation, or after tone-shock pairing (Mann–Whitney tests). Both lines showed contextual fear conditioning (repeatedmeasures ANOVA compared to the baseline;  = 0.235, F(1,34) = 0.235, p < 0.0001) but there were no line differences in contextual freezing. Female mice froze more than male mice in response to the contextual environment (26.32 ± 3.74% freezing versus 14.85 ± 3.61% freezing). All groups demonstrated cued fear conditioning (increase in freezing during the presentation of the CS (tone) as compared to the pre-tone period during the cued test,  = 0.052, F(1,34) = 621.4, p < 0.0001). There were no line or sex differences in cued fear conditioning or line by sex interaction. 3.7. Circadian rhythm Mice were placed in constant darkness (DD) after being entrained on a 12:12 h light/dark cycle for 14 days. Example actograms are shown in Fig. 3A. MAHDR mice consumed a greater quantity of MA than MALDR mice during the 25 mg/L (t(15) = 4.920, p < 0.0001) and 50 mg/L MA solution periods (t(15) = 5.859, p < 0.0001) (Fig. 3B). Comparing MA consumption during the 50 mg/L MA solution period with that in the 25 mg/L MA solution period, revealed a significant increase in MA consumption in MAHDR mice (two-tailed paired t-test, t(8) = 2.686, p = 0.0277) but not in MALDR mice (Fig. 3B).

R.H.J. Olsen et al. / Behavioural Brain Research 256 (2013) 197–204

201

Fig. 3. A. Representative actograms of MALDR and MAHDR mice. Mice were placed into a 12:12 h light:dark cycle then into constant darkness (arrow). Water (1), water or 25 mg/L MA (2), and then water or 50 mg/L MA (3) was supplied in a two-bottle choice paradigm. The individual lines are fit by eye to the onset of each day’s activity. Average daily MA consumption (B) and  (C) in MAHDR and MALDR mice during the 25 mg/L and 50 mg/L MA solution periods. (D) Correlation of MA consumption (mg of 50 mg/L MA solution) and body weight at death in individual MAHDR and MALDR mice. There was a significant negative correlation between MA consumption and body weight at death in MALDR (r2 = 0.60, p = 0.04) but not MAHDR (r2 = 0.25, p = 0.2) mice. (E) Correlation of  and MA consumption. There was a significant positive correlation in the MALDR mice (r2 = 0.80, p = 0.02) but not the MAHDR line (r2 = 0.0003, p = 0.97). White bars and open circles: MAHDR; black bars and closed circles: MALDR. # p < 0.0001;*p < 0.05; + p = 0.09.

202

R.H.J. Olsen et al. / Behavioural Brain Research 256 (2013) 197–204

Table 1 Swim speed and latency to the platform of MAHDR and MALDR mice in the water maze. Velocity (cm/s)

MAHDR MALDR

Visible

Hidden

Probe 1

Probe 2

Probe 3

15.47 ± 0.89 15.90 ± 0.76

18.53 ± 0.52 16.81 ± 0.56

21.52 ± 0.81 21.56 ± 0.74

21.18 ± 0.97 20.08 ± 1.00

19.41 ± 1.06 19.76 ± 1.02

Latency to platform (s) Day

1

2

3

4

5

MAHDR MALDR

20.92 ± 1.48 22.11 ± 0.95

10.02 ± 0.69 9.69 ± 0.80

36.93 ± 2.95 35.14 ± 3.24

29.21 ± 3.37 30.05 ± 4.10

32.41 ± 4.16 27.05 ± 3.94

4. Discussion

Fig. 4. (A) Representative blot of hippocampal GluA1 and GluA2 expression. Tubulin was used as a loading control. (B) MA-naïve MAHDR mice had elevated levels of GluA2, but not GluA1, AMPAR subunits, compared to MALDR mice. N = 4 mice/line. *p < 0.05.

There were no line differences in the free running period () during the period when only water was available (Fig. 3 C). During the 25 mg/L MA solution period, there was an increase in  in MALDR mice (paired t-tests compared to the  when only water was available (Fig. 3C, t(5) = 2.722, p = 0.02). Although MAHDR mice drank much more MA during the 25 mg/L MA solution period (Fig. 3B) there was no increase in  (Fig. 3C). Both lines showed an increased  during the 50 mg/L MA solution period (MAHDR: t(7) = 3.307, p = 0.01; MALDR: t(5) = 3.704, p = 0.007) (Fig. 3C) compared to the water only period. Body weights at death did not differ between the lines (MAHDR mice: 25.13 ± 0.51 g; MALDR mice: 25.53 ± 0.49 g). Average MA consumption (in mg) was negatively correlated with body weight at death in MALDR but not MAHDR mice (Fig. 3D). During the 25 mg/L MA solution period, there was no correlation between MA consumption and . During the 50 mg/L MA solution period, however, there was a positive correlation between MA consumption and  In MALDR, but not MAHDR, mice (Fig. 3E). 3.8. Western blot analysis MA-naïve MAHDR mice had elevated hippocampal levels of GluA2 (F(1,6) = 7.7, p = 0.03), but not GluA1, AMPAR subunits (Fig. 4A and B).

MAHDR mice did not show hippocampus-dependent spatial memory retention but MALDR mice did, indicating a deficit in hippocampus-dependent spatial memory retention in MAHDR mice. This form of memory is linked to theta/gamma oscillations in the hippocampus and theta-burst-dependent synaptic plasticity [23], which relies on the dynamic expression of GluA2lacking AMPARs at hippocampal glutamatergic synapses [24]. MAHDR mice exhibited elevated hippocampal levels of the GluA2, but not GluA1, AMPAR subunit. Interestingly, impairments in cognition, and specifically spatial memory and alterations in hippocampal AMPA receptor subunits, are associated with MA use [19]. Additionally, MALDR mice were more sensitive to the circadian period lengthening effects of MA than MAHDR mice. The association between drug abuse and circadian changes might be bidirectional [14]. Substance abuse might cause sleep disturbances and circadian changes in turn might contribute to the development of substance abuse and increase the risk of relapse [14]. Future studies are warranted to determine whether there are similar effects of MA on the circadian period length in humans. Line differences in growth rates were observed. For the current set of animals, MAHDR mice were initially heavier, though MALDR mice caught up and attained the same body weight within a twoweek period. Some research suggests a link between obesity and drug addiction in terms of a loss of control and compulsive behavior [25]. Whether these behaviors are linked and share common neurobiological processes deserves future research. Both lines self-administered MA at levels that produced a neurobehavioral response. MA consumption levels that lengthened the circadian period in the MALDR line did not do so in the MAHDR line. Previous studies have indicated an overall aversive response to MA in the MALDR line that is largely absent in the MAHDR line [8,9,12]. The data from the current study suggest that the MALDR line might also be more sensitive to the circadian shifting effects of MA. In the current study, a significant line difference in circadian period was seen when the mice self-administered MA over 24 h. In contrast, in previous studies the experimenter administered MA as a single injection (using a dose as low as 1 mg/kg) to demonstrate MA-induced aversion using conditioned taste and place aversion procedures. Therefore, it is conceivable that in MALDR mice MA is aversive following binge exposure but not following a much more gradual intake. Consistent with this idea, MALDR mice continued to consume MA, and in fact exhibited a tendency toward increased consumption over time, although intake remained low. MAHDR mice showed impairments in spatial memory retention in the water maze probe trials. MAHDR mice had greater levels of GluA2 than MALDR mice, which did not show impairments in spatial memory retention. GluA2 plays an important role in synaptic plasticity and learning and memory [26], including spatial memory retention [27]. Because MA was not given to these animals, there is

R.H.J. Olsen et al. / Behavioural Brain Research 256 (2013) 197–204

an innate line difference in GluA2 levels. This line difference might contribute to the spatial retention deficits observed in the MAHDR mice. Synaptic plasticity and LTP are involved in learning and memory [28] and both GluA1 and GluA2 are involved in regulating AMPA-related LTP and LDP [29]. A unique property of GluA2, compared to other AMPAR subunits, is its impermeability to calcium [30] and that it is inwardly rectifying. Abnormal or irregular AMPAR expression, trafficking, or plasticity can affect downstream systems including cognition [31]. While GluA2 knockout mice exhibit major learning impairments [32], likely a consequence of abberant LTP processes [33], there might be developmental effects due to GluA2 deficiency as well, as GluA2 expression follows a consistent developmental pattern [34]. Interestingly, acute administration of MA is associated with impairment of spatial working memory and increased expression of GluA2 in the hippocampus [19]. In MAHDR, tolerance for adverse effects of MA mice and spatial memory impairment may share a common biological mechanism that exists prior to any drug exposure. We recognize that the line difference in GluA2 levels might be global and not be limited to the hippocampus. MA shares a somewhat unique relationship with circadian rhythms in that it is one of two [35] environmental stimuli known to restore rhythmicity in animals with a lesioned master clock (SCN) [36]. The glutamatergic system, which is involved in photic entrainment [37], is perturbed by MA exposure, producing increased glutamate signaling [38]. Blockade of the glutamatergic system prevents photic entrainment [39]. GluA2 incorporation into AMPA receptors in multiple brain regions is associated with MA treatment [19] and reduction in glutamate overflow and associated excitotoxicity [40]. These studies emphasize GluA2 up regulation as an adaptive response, perhaps compensating for the increased glutamatergic drive following MA exposure. Elevated levels of GluA2 in MAHDR mice may therefore represent a pre-existing biological bias to increase the threshold relative to MALDR mice for MA to affect a shift in the circadian period. The glutamatergic system may also modulate other phenotypic differences that have been identified in these mice. The MAHDR line exhibited greater locomotor sensitization to MA (4 mg/kg) than the MALDR line [9]. Valproate and Lithium attenuate MA-induced hyperlocomotion and are associated with a down regulation of hippocampal GluA2 expression [41]. AMPA receptors may play a role in regulating sensitivity to rewarding or aversive effects of stimulants. Elevated GluA2 following intracranial self-stimulation in the nucleus accumbens shell is associated with an increase in the hedonic response, while GluA1 plays an opposing role and is associated with an increase in the sensitivity to aversive symptoms and those associated with withdrawal [42]. Thus, higher GluA2 expression in the MAHDR line may play a role in their greater sensitivity to rewarding and reinforcing effects of MA. There might be similar line-dependent GluA2 expression differences in brain regions other than the hippocampus that may mediate both naïve behavioral differences and those in response to MA treatment. In summary, genetic risk factors for elevated MA selfadministration may be associated with impairments in hippocampus-dependent memory and altered sensitivity to effects of MA on the length of the circadian period. Future studies are warranted to determine the extent to which pre-existing functional impairments in the context of genetic risk may represent risk factors for MA addiction and whether the effects of MA on the circadian period are transient or permanent.

Conflict of interest There are no conflicts of interest to declare.

203

Acknowledgements This work was supported by NS036607, P50 DA018165, the Department of Veterans Affairs, and the development account of Dr Raber. We thank Nathan Rogers for the graphical art.

References [1] Wolkoff D. Methamphetamine abuse: an overview for health care professionals. Hawaii Med J 1997;56:34–6. [2] Sommers I, Baskin D, Sommers A. Methamphetamine use among young adults: health and social consequences. Addict Behav 2006;31:1469–76. [3] Nicosia N, Pacula RL, Kilmer B, Lundberg R, Chiesa J. The economic cost of methamphetamine use in the United States. Rand Corporation; 2005. [4] Everitt BJ, Belin D, Economidou D, Pelloux Y, Dalley JW, Robbins TW. Neural mechanisms underlying the vulnerability to develop compulsive drug-seeking habits and addiction. PhilosTrans R Soc B 2008;363:3125–35. [5] Everitt BJ, Robbins TW. Neural systems of reinforcement for drug addiction: from actions to habits to compulsion. Nat Neurosci 2005;8:1481–9. [6] Peters J, Kalivas PW, Quirk GJ. Extinction circuits for fear and addiction overlap in prefrontal cortex. Learn Mem 2009;16:279–88. [7] Kreek M, Nielsen D, Butelman E, LaForge K. Genetic influences on impulsivity, risk taking, stress responsivity and vulnerability to drug abuse and addiction. Nat Neurosci 2005;8:1450–7. [8] Wheeler JM, Reed C, Burkhart-Kasch S, Li N, Cunningham CL, Janowsky A, et al. Genetically correlated effects of selective breeding for high and low methamphetamine consumption. Genes Brains Behav 2009;8:758–71. [9] Shabani S, McKinnon CS, Reed C, Cunningham CL, Phillips TJ. Sensitivity to rewarding or aversive effects of methamphetamine determines methamphetamine intake. Genes Brains Behav 2011;10:625–36. [10] Shabani S, Dobbs LK, Ford MM, Mark GP, Finn DA, Phillips TJ. A genetic animal model of differential sensitivity to methamphetamine reinforcement. Neuropharmacology 2012;62:2169–77. [11] Shabani S, McKinnon CS, Cunningham CL, Phillips TJ. Profound reduction in sensitivity to the aversive effects of methamphetamine in mice bred for high methamphetamine intake. Neuropharmacology 2012;62:1134–41. [12] Hyman S. Addiction: a disease of learning and memory. Am J Psychiatr 2005;162:1414–22. [13] Meredith C, Jaffe C, Ang-Lee K, Saxon A. Implications of chronic methamphetamine use: a literature review. Harvard Rev Psychiatr 2005;13:141–54. [14] Hasler B, Smith L, Cousins J, Bootzin R. Circadian rhythms, sleep, and substance abuse. Sleep Med Rev 2012;16:67–81. [15] Eastwood E, Allen C, Raber J. Effects of neonatal methamphetamine and thioperamide exposure on spatial memory retention and circadian activity later in life. Behav Brain Res 2012;230:229–36. [16] Michel S, Itri J, Colwell C. Excitatory mechanisms in the suprachiasmatic nucleus: the role of AMPA/KA glutamate receptors. J Neurophysiol 2002;88:817–28. [17] Todtenkopf MS, Parsegian A, Naydenov A, Neve RL, Konradi C, Carlezon Jr WA. Brain reward regulated by AMPA receptor subunits in nucleus accumbens shell. J Neurosci 2006;8(26):11665–9. [18] Sanderson D, Good MA, Seeburg PH, Sprengel R, Rawlins JN, Bannerman DM. The role of the GluR-A (GluA1) AMPA receptor subunit in learning and memory. In: Waxman S, Stein D, Swaab D, Fields H, editors. Progress in Brain Research, vol. 169. Amsterdam, the Netherlands: Elsevier Science; 2008. [19] Simões PF, Silva AP, Pereira FC, Marques E, Grade S, Mihazes N, et al. Methamphetamine induces alterations on hippocampal NMDA and AMPA receptor subunit levels and impairs spatial working memory. Neuroscience 2007;150:433–41. [20] Reissner K, Kalivas P. Using glutamate homeostasis as a target for treating addictive disorders. Behav Pharmacol 2010;21:514–22. [21] Raber J, Bongers G, LeFevour A, Buttini M, Mucke L. Androgens protect against apolipoprotein E4-induced deficits. J Neurosci 2002;22:5204–9. [22] Davis MJ, Haley T, Duvoisin RM, Raber J. Measures of anxiety, sensorimotor function, and memory in male and female mGluR4-/- mice. Behav Brain Res 2012;229:21–8. [23] Lisman J, Buzsáki G. A neural coding scheme formed by the combined function of gamma and theta oscillations. Schizophrenia Bull 2008;34:974–80. [24] Lu W, Shi Y, Jackson AC, Bjorgan K, During MJ, Sprengel R, et al. Subunit composition of synaptic AMPA receptors revealed by a single-cell genetic approach. Neuron 2009;362:254–68. [25] Wang G, Volkow N, Thanos P, Fowler J. Similarity between obesity and drug addiction as assessed by neurofunctional imaging: a concept review. J Addict Dis 2004;23:39–53. [26] Chung H, Steinberg J, Huganir R, Linden DJ. Requirement of AMPA receptor GluA2 phosphorylation for cerebellar long-term depression. Science 2003;300:1751–5. [27] Sacktor TC, Mzeta PK. LTP maintenance, and the dynamic molecular biology of memory storage. In: Waxman S, Stein D, Swaab D, Fields H, editors. Progress in Brain Research, vol. 169. Amsterdam, the Netherlands: Elsevier Science; 2008. [28] Whitlock J, Heynen A, Shuler M, Bear M. Learning induces long-term potentiation in the hippocampus. Science 2006;313:1093–7.

204

R.H.J. Olsen et al. / Behavioural Brain Research 256 (2013) 197–204

[29] Malenka R, Bear M. LTP and LTD. An embarrassment of riches. Neuron 2004;44:5–21. [30] Jonas P, Burnashev N. Molecular mechanisms controlling calcium entry through AMPA-type glutamate receptor channels. Neuron 1995;15:987–90. [31] Makuch L, Volk L, Anggono V, Johnson RC, Yu Y, Duning K, et al. Regulation of AMPA receptor function by the human memory-associated gene KIBRA. Neuron 2011;71:1022–9. [32] Gerlai R, Henderson JT, Roder JC, Jia Zhenping. Multiple behavioral anomalies in GluA2 mutant mice exhibiting enhanced LTP. Behav Brain Res 1998;95:37–45. [33] Jia Z, Agopyan N, Miu P, Xiong Z, Henderson J, Gerlai R, et al. Enhanced LTP in mice deficient in the AMPA receptor GluA2. Neuron 1996;17:945–56. [34] Kumar SS, Bacci A, Kharazia V, Huguenard JR. A developmental switch of AMPA receptor subunits in neocortical pyramidal neurons. J Neurosci 2002;22(8):3005–15. [35] Honma K, Honma S. The SCN-independent clocks, methamphetamine and food restriction. Eur J Neurosci 2009;30:1707–17. [36] Honma K, Honma S, Hioshige T. Activity rhythms in the circadian domain appear in suprachiasmatic nuclei lesioned rats given methamphetamine. Physiol Behav 1987;40:767–74.

[37] Masubuchi S, Honma S, Abe H, Nakamura W, Honma K. Circadian activity rhythm in methamphetamine-treated Clock mutant mice. Eur J Neurosci 2001;14:1177–80. [38] Golombek D, Rosenstein R. Physiol Circad Entrainment. Physiol Rev 2010;90:1063–102. [39] Mark KA, Soghomonian JJ, Yamamoto BK. High-dose methamphetamine acutely activates the atriatonigral pathway to increase striatal glutamate and mediate long-term dopamine toxicity. J Neurosci 2004;24: 11449–56. [40] Vindlacheruvu R, Ebling F, Maywood E, Hastings M. Blockade of glutamatergic neurotransmission in the suprachiasmatic nucleus prevents cellular and behavioural responses of the circadian system to light. Eur J Neurosci 1992;4:673–9. [41] Ikeda M. Calcium dynamics and circadian rhythms in suprachiasmatic nucleus neurons. Neuroscientist 2004;10:315–24. [42] Carlezon WA, Thomas M. Biological substrates of reward and aversion: A nucleus accumbens activity hypothesis. Neuropharmacology 2009;56: 122–32.