Methamphetamine Increases Locomotion and Dopamine Transporter Activity in Dopamine D5 Receptor-Deficient Mice

Dopamine regulates the psychomotor stimulant activities of amphetamine-like substances in the brain. The effects of dopamine are mediated through five known dopamine receptor subtypes in mammals. The functional relevance of D5 dopamine receptors in the central nervous system is not well understood. To determine the functional relevance of D5 dopamine receptors, we created D5 dopamine receptor-deficient mice and then used these mice to assess the roles of D5 dopamine receptors in the behavioral response to methamphetamine. Interestingly, D5 dopamine receptor-deficient mice displayed increased ambulation in response to methamphetamine. Furthermore, dopamine transporter threonine phosphorylation levels, which regulate amphetamine-induced dopamine release, were elevated in D5 dopamine receptor-deficient mice. The increase in methamphetamine-induced locomotor activity was eliminated by pretreatment with the dopamine transporter blocker GBR12909. Taken together, these results suggest that dopamine transporter activity and threonine phosphorylation levels are regulated by D5 dopamine receptors.


Introduction
Dopamine mediates a wide variety of physiological and behavioral functions in the central nervous system (CNS), such as the response to psychomotor stimulants and reward and learning behaviors [1,2,3,4,5,6,7]. These roles of the dopamine system were discovered through the creation and characterization of dopamine receptor-deficient mice ( [8,9,10,11]. The effects of dopamine are mediated through five known subtypes of dopamine receptors in mammals (D1R, D2R, D3R, D4R, and D5R) [12].
Genomic studies found a significant relation between a polymorphism in the D5R gene locus and vulnerability to drug abuse [13,14]. Consistent with this mutation, several studies found that D5Rs play a role in mediating the response to cocaine administration. D5R-deficient mice with a mixed genetic background are less sensitive to acute cocaine administration than control littermates [15]. Furthermore, D5R-deficient mice with a C57/B6 background are more sensitive to chronic cocaine administration than wild-type (WT) littermates [16]. However, it is unknown whether D5Rs contribute to the response to amphetamine-like drugs. To this end, we investigated the effect of D5R deficiency on methamphetamine (METH)-induced behavior. METH is a derivative of amphetamine and is a major psychostimulant that is frequently abused.
We found that D5R-deficient mice were hypersensitive to acute METH challenges. We also found that GBR12909, a dopamine transporter (DAT) blocker, affected the blocking and reversal of monoamine reuptake by METH through monoamine transporters such as DAT. In addition, we evaluated threonine phosphorylation levels in WT and D5R-KO mice because a specific threonine residue in DAT is important for modification of reuptake and release of dopamine [17,18,19], and found that threonine phosphorylation levels were higher in D5R-KO mice than in WT mice. Finally, we measured dopamine levels in the nucleus accumbens (NA) to assess whether this brain region mediated the altered hypersensitivity to METH but failed to detect a significant difference in dopamine levels in this brain region between WT and D5R-KO mice.

Characterization of D5R-KO mice
We created a D5R-KO mice line on a C57/B6 background for this study. The murine D5R gene was disrupted in embryonic stem (ES) cells by homologous recombination that resulted in inactivation of the coding region ( Figure 1a). Consistent with a previous study, the D5R-KO mice were fertile [20]. The authenticity of the D5R-KO line was confirmed by genomic Southern blotting with a 39 region probe (Figure 1b). In addition, Northern blotting showed that D5R mRNA was completely abolished in the D5R-KO mice (Figure 1c). Genomic Southern blotting with a 39 region probe. Genomic DNA was collected from WT (+/+), heterogeneous (+/2), and homogenous (2/2) D5R mice and subjected to electrophoresis and Southern blotting. The bands corresponding to wild-type and mutant DNA are indicated. (c) mRNA was collected from WT (+/+), heterogeneous (+/2), and homogenous (2/2) animals and subjected to electrophoresis and Northern blotting with a D5R cDNA probe. D5R mRNA was absent from the homogenous ( Effects of pharmacological manipulations on ambulation To assess the roles of D5Rs in dopamine-mediated behaviors, we measured open field locomotor activities of WT and D5R-KO mice that were administered 2.5 mg/kg of METH via intraperitoneal injections. METH affects dopamine transmission by blocking dopamine reuptake and reversing dopamine release through the DAT pore. Therefore, we also evaluated the METHinduced locomotor activities after pretreatments with either saline or the DAT blocker GBR12909. Three-way analysis of variance (ANOVA) was employed to analyze METH challenge-induced locomotor activity data from the four groups of mice. The analysis was performed based on the following three factors: 1) pretreatment with saline control or GBR12909; 2) genotype (WT or D5R-KO); and 3) time course.
The three-way ANOVA found a secondary interaction between the three factors (blocker6genotype6time course) (F (11,220) [14,21,22], the METH challenge increased ambulation in D5R-KO mice by approximately 200-260% in animals that were pretreated with saline ( Figure 2a). METH -induced ambulation in D5R-KO mice was significantly greater than in WT mice from 0 to 60 min after the METH challenge. At some of the time points, D5R-KO mice traveled a total distance up to 30% greater than that traveled by WT mice. Thus, pretreatment with the DAT blocker specifically eliminated METH-induced hyperlocomotor activity in D5R-KO mice. By contrast, cocaine (15 mg/kg)induced ambulation was not significantly different between WT and D5R-KO mice ( Figure 2b).

DAT phosphorylation in D5R-KO mice
Because the DAT blocker eliminated METH-induced hyperlocomotor activity in D5R-KO mice and there were no differences in cocaine-induced locomotor activity between D5R-KO and WT mice, we hypothesized that the METH-induced hyperactivity resulted from a change in DAT activity in D5R-KO mice. Previous studies found that DAT activity is regulated by phosphorylation. Phosphorylation of the 53 rd threonine residue in the N-terminus of DAT is required for amphetamine-induced neurotransmitter release that is mediated by monoamine transporters [17,18,19]. Therefore, we assessed threonine phosphorylation of DAT immunoprecipitates that were made from whole brain lysates. The detection of a signal at 75 kDa by anti-phospho-Thr antibody was DAT(+/+)-dependent since no signal was detected in the immunoprecipitant from DAT-KO mice ( Figure 3a). An ANOVA revealed that threonine phosphorylation levels per total DAT protein level were significantly higher in immunoprecipitates made from D5R-KO mouse brain lysates than in immunoprecipitates made from WT mouse brain lysates (

Dopamine transmission in D5R-KO mice
To determine the brain region responsible for the difference in hypersensitivity to METH between WT and D5R-KO mice, we next investigated fluctuations in dopamine levels in the NA. The NA receives inputs from dopamine neurons that regulate locomotor activity [23]. In vivo microdialysis was performed in freely moving mice to measure dopamine levels (Figures 4a, b, c, d). Dopamine levels in the NA were increased by approximately 350% in WT and 400% in D5R-KO mice from 20 to 40 minutes after the METH challenge. By 120 minutes after the METH challenge, the dopamine levels in the NA of both genotypes decreased to slightly more than twice the baseline levels ( Figure 4a). The dopamine fluctuation ratios relative to the baselines were analyzed for the different pretreatment and genotype groups. A three-way ANOVA (blocker6genotype6time course) and subsequent post-hoc analyses found a significant interaction between blocker and time course after the METH challenge (Figure 4a (Figure 4c), and all microdialysis data used for analyses were obtained from the NA (Figure 4d). Thus, dopamine transmission in the NA, which is thought to be required for METH-induced increases in locomotor activity, was blocked by GBR12909. However, the secondary interactions between blocker, genotype, and time course were not significant and there were no differences in dopamine transmission in the NA across genotypes (F (5, 120) = 1.44; and p = 2.14610 21 ). Thus, differences in METH-induced dopamine transmission in the NA do not seem to contribute to the higher METH-induced locomotor activity of D5R-KO mice.

Discussion
Dopamine signaling in the NA mediates psychostimulantinduced locomotion [24,25,26]. In the present study, we found that D5R-KO mice were more sensitive to a METH challenge (2.5 mg/kg) than WT control mice. This greater sensitivity of D5R-KO mice to acute METH administrations was almost completely blocked by pretreatment with the DAT blocker GBR12909 (5 mg/kg). However, pretreatment with GBR12909 did not prevent METH-induced locomotor activity. In addition, our results showed that there were no differences in cocaineinduced locomotor activities between genotypes. These observations are consistent with previous studies, which found that a different line of D5R-KO mice were not hypersensitive to acute cocaine administrations [16]. These observations suggest that alterations in METH-induced dopamine release through the DAT pore likely account for the increased locomotor activity observed in D5R-KO mice. By contrast, other processes such as quantum release-dependent elevations of dopamine levels do not seem to contribute to the enhanced locomotor activity observed in METH-challenged D5R-KO mice. In addition, saline challenge coupled with GBR12909 pretreatment did not cause differences in locomotor activity between genotypes. Furthermore, postsynaptic processes that are induced by synaptic dopamine transmission did not underlie the enhanced locomotor activity observed in D5R-KO mice. Pretreatment with GBR12909, which completely blocks dopamine release through the DAT pore, eliminated the METHinduced increase in locomotor activity in D5R-KO mice. Thus, the DAT pore is required for the observed hypersensitivity to METH in D5R-KO mice. Therefore, one of the crucial consequences of D5R deficiency is alterations in METH-induced DAT activity.
We were unable to determine the precise anatomical targets of D5R deficiency. We did not find differences in the ratios of METH-induced dopamine levels relative to baselines in the NA between genotypes. Thus, the NA is not a direct anatomical target of D5R deficiency. However, neuronal circuits that include dopaminergic terminals are affected by D5R deficiency because DAT blockade specifically eliminated the METH-induced hyperlocomotor activity that was observed in D5R-KO mice. Interestingly, synaptic transmission of norepinephrine in the mPFC influences the dopaminergic signaling in the NA that regulates amphetamine-induced stimulation of ambulation. This influence is likely mediated by direct glutamatergic signaling from the mPFC to the NA [27,28]. Thus, norepinephrine transmission in the mPFC may also be affected by D5R deficiency.
The phosphothreonine levels of DAT were significantly increased in the brains of D5R-KO mice relative to WT mice. These data indicate that D5R activity regulates the phosphorylation status of DAT, which in turn affects METH-induced dopamine release. Indeed, recent studies demonstrated that threonine 53 in DAT is required for amphetamine-induced dopamine release [17,18,19].
Previous studies that utilized a different line of D5R-KO mice found a response to cocaine that we did not observe in the D5R-KO mice described here [15,16]. Differences in the genetic background of the two mouse lines may explain the discrepancies in the responses to cocaine. D5R-KO mice with a mixed genetic background (hybrid of 129/SvJ16C57BL/6J) are less sensitive to cocaine than WT mice [15]. By contrast, when this mixed background line of D5R-KO mice is bred onto a C57B6 genetic background, there are no differences in sensitivity to cocaine between D5R-KO and WT mice [16]. Our results were consistent with the findings of the later study. Moreover, other studies that use crossbred rather than genetically engineered mice also find that sensitivity to METH is dependent upon the genetic background of the mice [29,30,31]. More specifically, a crossbred line of mice that has an extreme sensitivity to METH also exhibits hypersensitivity to cocaine. Because these results contrast with the genetically engineered D5R-KO line used in our study, it seems likely that the locus responsible for the hypersensitivity to cocaine The clinical relevance of our findings is unclear because METH is an exogenous substance. However, epidemiological studies find a potential relation between D5R deficiency and behavior. For example, a small population of people has microsatellites in the D5R gene locus that may attenuate D5R gene expression [32]. These individuals with a microsatellite polymorphism in the D5R gene locus (148 bp allele) are also at high risk of becoming drug abusers [13,14].
In summary, the results presented here provide the first evidence of a functional relationship between the D5R gene and DAT. This link between D5R gene functions and DAT activity might help to elucidate more complex multigene dysfunctions that underlie diseases that cannot be attributed to alterations in one gene.

Ethics Statement
The Animal Care and Use Committee of the Tokyo Metropolitan Institute for Neuroscience approved all surgical and experimental protocols that were used in this study (Permit Number: 08-1815).

Generation of D5R-KO mice
The WT D5R gene consists of a 9.0 kb genomic fragment that contains the receptor coding region and a 9.5 kb genomic fragment that contains the upstream region. The mutant D5R construct ''pD5KO'' consisted of part of the 9.0 kb coding region genomic fragment and the complete 9.5 kb upstream genomic fragment. The fragments were isolated from a 129/SvJ1 genomic library and subcloned as EcoRV-NheI (9.0 kb) and XbaI-XbaI (9.5 kb) fragments into the same gene targeting vector. The targeting vector was constructed so that the entire coding sequence was deleted by using the following DNA fragments: a 1.2 kb MC1 promoter-diphtheria toxin-A fragment gene (DT-A) for negative selection; the XbaI-XbaI fragment containing the upstream region of the D5R gene; a 2.3 kb PGK promoter-Escherichia coli xanthine-guanine phosphoribosyltransferase gene (gpt); a 1.1 kb MC1 promoter-neomycin gene (neo); a 4.0 kb XbaI-NheI fragment containing the 39-untranslated region; and a pBluescript plasmid (Figure 1a). This targeting vector was then used to generate D1 dopamine receptor-deficient mice [33].
The ES cell line used for generation of the D5R-KO mice was previously described [33]. Briefly, CCE ES cells (E14TG2a IV 129-derived ES cells, acquired from Dr. E. Robertson) were cultured as described [34]. The cultured ES cells (2.5610 7 cells) were transfected with the linearized targeting vector (50 mg) by electroporation, and then cells that contained the targeting vector were selected with G418. A total of 120 drug-resistant colonies were collected, and then the genomic DNA was digested with EcoRV and NheI and subjected to Southern blot analysis for confirmation of homologous recombination (Figure 1b). D5R-KO mice were then generated using the homologous recombinant ES cells. The mice were backcrossed with a C57BL/6J strain for 10 generations and were then maintained on the C57BL/6J genetic background. D5R-KO heterozygous males and females with a C57BL/6J genetic background were crossed to generate D5R-KO homozygous lines and WT homozygous lines. All of the mice used in the experiments were generated by matings between D5R-KO homozygous mice and between WT mice.
Brain mRNA was prepared according to standard protocols and subjected to Northern blot analysis to confirm that the D5R-KO mice did not contain the D5R transcript (Figure 1c).

Locomotor activity measurements
The horizontal movements of the animals were captured with a CCD camera connected to a computer with a video-capture board. The camera was operated using Linux software (IO Data, Japan), and the animals were tracked with an I-rec webcam system [35]. The travel distances were calculated with custom-made scripts. Mice were familiarized to an open field (40640 cm) for 1 h. The next day, mice were injected with saline and the locomotor activities were recorded for 1 h. Then, the mice were exposed to either a METH (2.5 mg/kg) or a cocaine challenge (15 mg/kg) and the locomotor activities were recorded for an additional 1 h. For the blocking experiments, either saline or GBR12909 (5 mg/kg) was administered by intraperitoneal injections that occurred 80 min before the METH challenge. Six time points (six time bins of 10 min each) were analyzed after the challenges.

Microdialysis
Microdialysis probes (type AG; outer diameter: 0.5 mm; inner diameter: 0.4 mm; Eicom, Kyoto, Japan) were stereotaxically implanted under anesthesia to target the NA and were then fixed to the skull. The coordinates used were: anterior-posterior 1.1 mm; medial-lateral 1.2-1.6 mm; and dorsal-ventral 3.8-4.4 mm. Experimental dialysate samples were collected on days 1 and 2 after the surgeries. The microdialysis probes were perfused with artificial cerebrospinal fluid at a flow rate of 2 ml/min and with a sampling time of 20 min.
Once the perfusions began, the mice were then left alone for either 2 h (day 1) or 1 h (day 2) to stabilize the baseline measurements. Immediately following stabilization on day 1, saline was injected and samples were collected during a baseline period of 60 min. Next, METH (2.5 mg/kg) was injected and samples were collected during an experimental period of 120 min. On day 2, mice were pretreated with GBR12909 (5 mg/kg) for 20 and 80 min prior to the saline and METH injections, respectively. Nine and thirteen dialysate samples were obtained from the NA on day 1 (baseline and METH treated) and on day 2 (baseline, GBR12909, and METH treated), respectively. Forty microliters (2 ml/min for 20 min) of dialysates were collected at 4uC in a polypropylene sample tube during each session. The dialysate samples were stored under acidic conditions (pH 3.5). Dialysate samples were also collected from D5R-KO mice that were only treated with GBR12909. All of the samples were immediately frozen and stored at 220uC until use. The dopamine content of the samples was assessed as previously described [36].

Immunoprecipitations and Western blotting
Whole brain lysates from WT and D5R-KO mouse brains were prepared for Western blotting. For each experimental and control condition, eight hemispheres from a total of six mice that were 12 weeks old were homogenized in ice-cold buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.1% sodium dodecyl sulfate (SDS), 1% sodium deoxycholate, 1% Triton, 50 mM NaF, and 2 mM EDTA) that was supplemented with 2 mM Na 3 VO 4 , a phosphatase inhibitor cocktail (Sigma Chemical Company, St. Louis, MO, USA) and protease inhibitors (Roche Diagnostics, Indianapolis, IN, USA). The insoluble proteins were removed by centrifugation at 10,0006g for 10 min. The lysates (supernatants) were then incubated for 4 h with protein G agarose beads (50 ml) at 4uC.
Immunoprecipitations were then performed overnight at 4uC using 1 ml of rat DAT antibody (Chemicon MAB369, Chemicon, Temecula, CA, USA) and 2 mg of protein lysate. The volume of antibody-coated agarose beads used per mg of protein was approximately 0.5 ml. The beads were washed three times with 1 ml of immunoprecipitation buffer. Proteins were eluted by addition of 26SDS sample buffer (25 ml; 125 mM Tris-HCl pH 6.8, 10% 2-mercaptoethanol, 4% SDS, 10% sucrose, and 0.01% bromophenol blue) and 5 min of boiling. The eluted samples (10 ml) were then separated on 10% SDS-PAGE gels and transferred to polyvinylidene difluoride (PVDF) membranes at 100 V for 1 h. The membranes were exposed to a phosphothreonine primary antibody (Cell Signaling Technology #9381; 1:1000; Cell Signaling Technology, Beverly, MA, USA) and then a horseradish peroxidase-coupled donkey anti-rabbit IgG secondary antibody (1:200; GE Healthcare, Waukesha, WI, USA). The bands on the Western blots were visualized with an enhanced chemiluminescence kit (GE Healthcare) and exposure to Hyperfilm ECL (GE Healthcare).

Statistical analyses
Data were analyzed with the ''R'' programming environment and software language created by the R Foundation for Statistical Computing (Vienna, Austria). The changes in ambulatory activities between baselines and METH challenges were calculated by subtracting the average ambulation that occurred during the last 30 min of the baseline periods from the average ambulation that occurred during each of the 10 min time bins after the METH challenges. The ambulatory activities after the METH challenges were analyzed using a three-way ANOVA for blocker, genotype, and time (26266). Post-hoc analyses were then used to determine secondary interactions between the three factors. Simple interactions between blocker and genotype were subjected to further post-hoc analyses. Specifically, the simple main effects of blocker (saline or GBR12909) and genotype on locomotor activity were assessed for each 10 min time bin.
Microdialysis data were also analyzed using a three-way ANOVA for blocker, genotype, and time (26266). The post-hoc analyses were performed as described above for the ambulation experiments. DAT phosphorylation data were analyzed using a two-tailed paired t-test. Because it was difficult to control for the variance in the background levels across the different Western blot experiments, the data from each experiment were treated as a pair (WT and D5R-KO) in the t-test. Data are presented as mean 6 s.e.m. throughout the manuscript.