Brain-specific heterozygous loss-of-function of ATP2A2, endoplasmic reticulum Ca2+ pump responsible for Darier’s disease, causes behavioral abnormalities and a hyper-dopaminergic state

Abstract A report of a family of Darier’s disease with mood disorders drew attention when the causative gene was identified as ATP2A2 (or SERCA2), which encodes a Ca2+ pump on the endoplasmic reticulum (ER) membrane and is important for intracellular Ca2+ signaling. Recently, it was found that loss-of-function mutations of ATP2A2 confer a risk of neuropsychiatric disorders including depression, bipolar disorder and schizophrenia. In addition, a genome-wide association study found an association between ATP2A2 and schizophrenia. However, the mechanism of how ATP2A2 contributes to vulnerability to these mental disorders is unknown. Here, we analyzed Atp2a2 heterozygous brain-specific conditional knockout (hetero cKO) mice. The ER membranes prepared from the hetero cKO mouse brain showed decreased Ca2+ uptake activity. In Atp2a2 heterozygous neurons, decays of cytosolic Ca2+ level were slower than control neurons after depolarization. The hetero cKO mice showed altered behavioral responses to novel environments and impairments in fear memory, suggestive of enhanced dopamine signaling. In vivo dialysis demonstrated that extracellular dopamine levels in the NAc were indeed higher in the hetero cKO mice. These results altogether indicate that the haploinsufficiency of Atp2a2 in the brain causes prolonged cytosolic Ca2+ transients, which possibly results in enhanced dopamine signaling, a common feature of mood disorders and schizophrenia. These findings elucidate how ATP2A2 mutations causing a dermatological disease may exert their pleiotropic effects on the brain and confer a risk for mental disorders.


Introduction
Bipolar disorder is a mental disorder that shares many clinical features and pathophysiological basis with schizophrenia (1). Calcium signaling abnormalities have been proposed as the most likely hypothesis for the pathophysiology of bipolar disorder (2), which has been supported by genome wide association studies (GWAS) that identified CACNA1C encoding L-type calcium channel (3). However, the effect of the single nucleotide polymorphisms on bipolar disorder is modest with an odds ratio less than 1.2, and modeling in animals is still difficult. In such a situation, animal models of Mendelian diseases that show a bipolar disorder phenotype are therefore a promising strategy. Among such Mendelian diseases, a report of a family of Darier's disease, an autosomal dominantly inherited skin disorder, in which major affective disorders including bipolar disorder co-segregated (4), is of particular interest because the causative gene was found to be ATP2A2, which encodes SERCA2, a Ca 2+ pump on the endoplasmic reticulum (ER) (5)(6)(7)(8). A recently released dataset from the Bipolar Exome (BipEx) sequencing project (https://bipex.broadinstitute.org/) showed an association between damaging missense mutations of ATP2A2 and bipolar disorder (P = 0.005 with odds ratio 10). Similarly, a GWAS of schizophrenia also showed an association with ATP2A2 and schizophrenia (9).
Patients with Darier's disease have a higher prevalence rate of mood disorder (50%) including depression and bipolar disorder (10). Recent evidence has shown that loss of function (LOF) mutations of ATP2A2 confer an increased risk for psychoses including bipolar disorder and schizophrenia (11). The C560R mutation in ATP2A2 in the initial family was shown to drastically decrease expression levels of the ATP2A2 protein (11,12). These findings suggest that heterozygous KO mice of Atp2a2 would be useful to study the role of calcium signaling in bipolar disorder and schizophrenia. However, conventional Atp2a2 heterozygous KO mice showed serious somatic abnormalities (13), which has confounded behavioral analyses. Thus, a brain-specific heterozygous KO mouse of Atp2a2 would be more useful to study bipolar disorder and schizophrenia.
Ca 2+ uptake activity into the ER via ATP2A2 plays an essential role supporting cellular Ca 2+ homeostasis by maintaining a low cytosolic Ca 2+ concentration (14)(15)(16). As the ER is a widely distributed organelle in neurons including the cell body, dendrites, and axon terminals (17)(18)(19), ATP2A2 is thought to play an important role in a variety of neuronal functions. However, the in vivo significance of ATP2A2 in neurons and in the brain has not been studied as extensively compared to other cell types or organs (20). Considering the essential roles of ATP2A2 in Ca 2+ signaling, mood disorders and schizophrenia in Darier's disease patients, which are supposed to result from a pleiotropic effect of the LOF mutations in ATP2A2, a study of brain-specific KO mice for Atp2a2 would be useful to clarify the mechanisms. In this study, we generated and analyzed brain-specific KO mice for Atp2a2. ATP2A2 was shown to be essential for neuronal functions including Ca 2+ homeostasis, and the brain-specific heterozygous KO mice showed characteristic behavioral phenotypes. Enhanced dopamine (DA) neurotransmission was also detected in the nucleus accumbens (NAc). These findings may explain how ATP2A2 confers risk for psychoses including bipolar disorder and schizophrenia.

Generation of Atp2a2 brain-specific heterozygous KO mice
We generated brain-specific Atp2a2 knockout (cKO) mice by crossing Nestin-Cre (NesCre) transgenic mice with Atp2a2 flox/flox mice (Fig. 1A). However, the Atp2a2 cKO mice (Atp2a2 flox/flox ; NesCre tg/• ) were not observed among the newborn pups obtained by mating between the Atp2a2 flox/+ ; NesCre tg/• mice and Atp2a2 flox/flox mice. The fact that no cKO mice were obtained in the 26 offspring from 6 litters suggests embryonic lethality of the cKO mice. When fetal pups were delivered by Caesarean section at embryonic day 14 (E14) and genotyped, cKO mice were identified at the expected ratio. However, these cKO mice were embryonically lethal with severe hemorrhages present in the brain (Arrow, Fig. 1B), though they appeared to have survived longer periods in the uterus than the heart-specific ATP2A2 KO mice (around E11) (21). Western blotting showed greatly reduced expression of ATP2A2 protein in the brain but not in the heart (Fig. 1C). In the brain-specific Atp2a2 heterozygous knockout mice (hetero cKO, Atp2a2 flox/+ ; NesCre tg/• ), ATP2A2 protein expression levels in the brain were approximately half that of control mice (Atp2a2 +/+ ; NesCre tg/• ), and comparable to those of the conventional heterozygotes (Atp2a2 +/− ) ( Fig. 1D and E). The hetero cKO appeared healthy even at 2 years after birth and their body weights were similar among genotypes (Fig. 1F, control, 45.3 ± 1.1 g; hetero cKO, 47.5 ± 1.0 g). In contrast, conventional heterozygotes showed serious somatic phenotypes and greatly reduced body weight (Fig. 1G, +/+, 51.2 ± 2.2 g; +/−, 37.1 ± 1.6 g; * * , P < 0.01, two-tailed unpaired t-test), as observed in a previous study (13). These results showed that the brainspecific heterozygous knockout of Atp2a2 avoided the somatic abnormalities seen in the conventional heterozygotes, allowing behavioral analyses of the hetero cKO mice to be performed. No apparent structural abnormalities were detected in the brains of the hetero cKO mice (Fig. 1H).
To evaluate the significance of the reduced expression levels of ATP2A2 protein at the cellular level, we extracted an ER membrane fraction from the brain tissue and examined the Ca 2+ uptake activity into the ER through ATP2A2 by spectrophotometric assay ( Fig. 2A). As shown in the fluorescence intensity profiles, Ca 2+ was rapidly imported into the ER in both genotypes after Ca 2+ loading by adding ATP. However, the ER membranes from hetero cKO mice showed slower Ca 2+ uptake activity ( Fig. 2A). The uptake activity was most prominent around the first 300 s following the addition of ATP in both genotypes, similar to other reports (22,23). The ER membranes from hetero cKO showed significantly reduced Ca 2+ uptake activity in the period ( Fig. 2B; Control, 835.1 ± 37.0; hetero cKO, 723.9 ± 40.0 a.u.; P = 0.0076, two-tailed paired t-test). The intramembrane Ca 2+ contents of hetero cKO samples were also reduced ( Fig. 2C; Control, 769.0 ± 44.6; hetero cKO, 674.8 ± 64.9 a.u.; P = 0.037, twotailed paired t-test). Next, we analyzed the effects of the impaired Ca 2+ uptake activity on cytoplasmic Ca 2+ dynamics in response to neuronal excitability (Fig. 3). We prepared primary hippocampal neurons from Atp2a2 wild-type (+/+) and heterozygous (+/−) embryos expressing G-CaMP7 and monitored Ca 2+ dynamics in the cytosol by measuring the fluorescence signals (Fig. 3A). We observed that the decay of the cytosolic Ca 2+ levels after

Behavioral abnormalities in Atp2a2 brain-specific heterozygous KO mice
To comprehensively analyze behavioral characteristics in hetero cKO mice, we performed a behavioral test battery (Fig. 4, and Supplementary Material, Figs S1-S3). For the distance traveled in an open field test for 120 min, there was no statistically significant change for the overall period, however, hetero cKO mice showed a tendency to increase for the first 60 min ( No significant changes were found in the center time and stereotypic counts ( Fig. 4C and D). To assess the ability to learn and remember an association between environmental cues and aversive experiences, a contextual and cued fear conditioning test was performed ( Fig. 4E-G). The hetero cKO mice showed similar levels of freezing behavior compared to control mice in a contextual fear test ( Fig. 4F) but they showed impaired fear memory in a cued fear test (Fig. 4G, last 3 min; genotype, P = 0.0079, genotype × time, P = 0.661; two-way repeated measures ANOVA). In other tests that examined sensorimotor functions, learning and memory performance, anxiety-like behaviors, depressionlike behaviors and social interaction, no significant differences were observed (Supplementary Material, Figs S1-S3). Electroencephalographic (EEG) recordings were also carried out for 3 days, but no marked epileptic waveforms were observed in the EEG of hetero cKO mice and the sleep/wake cycles deduced from the analysis of the EEG frequency were not significantly different between genotypes (Supplementary Material, Fig. S4). We measured wheel running activity for more than 3 months to evaluate whether the hetero cKO mice would show behavioral changes relevant to bipolar disorder or recurrent depression (Supplementary Material, Fig. S5) (24,25). The duration or pattern of the wheel running activities appeared similar among genotypes (n = 15 mice for each genotype: Supplementary Material, Fig. S5A). Generally, mice show an increase in wheel running activity immediately after cage changing (every 2 weeks), which was similarly observed in the control mice. However, this behavioral response was not observed in the hetero cKO mice (Supplementary Material, Fig. S5A-C). On the other hand, behaviors other than wheel running activity were not clarified with the equipment. To analyze the behavioral difference after the cage changing more closely, video recordings were performed with a pair of control and hetero cKO mice for 90 min following cage changing over 3 months ( Fig. 5; n = 5 and 6 cage changings for control and hetero cKO, respectively). After the cage changing, for the first 30 min (corresponding to 1800 s), the hetero cKO mouse spent more time moving on the bedding ( Fig. 5B; control, 1092.2 ± 151.4 s; hetero cKO, 1519.8 ± 117.3 s; P = 0.0005, two-tailed unpaired t-test) and accordingly it showed a reduced amount of wheel running activity compared to the control mouse. Both genotypes spent similar time in staying behavior on the bedding (Fig. 5B), in which the mice appeared apparently awake but did not move vigorously and kept staying on the bedding. The hetero cKO mouse showed significantly increased vertical activity (or rearing) after the cage changing ( Fig. 5C; control, 31 ± 1.6; hetero cKO, 55 ± 2.8; P < 0.0001, twotailed unpaired t-test), which was consistent with the result in the open field test (Fig. 4B). Collectively, these results suggest that the hetero cKO mice had altered behavioral responses to novel environments.

Extracellular DA levels were elevated in the NAc of Atp2a2 brain-specific heterozygous KO mice
The increase of vertical activity in novel environments and impaired fear memory (Figs 4 and 5) could be correlated with the dysregulation of the DA system in the NAc (26,27). We performed monoamine analysis using tissue homogenates of brain samples. As shown in the open field tests in Figure 4B, rearing scores in hetero cKO mice were markedly increased during the first 30 min. Thus, hetero cKO mice were subjected to open field tests for 30 min, then sacrificed, and NAc tissues were collected and homogenized. DA and its metabolites (DOPAC and HVA) were quantified by HPLC coupled with electrochemical detection (Fig. 6). DA levels tended to be decreased in the NAc of the hetero cKO mice (Fig. 6, left panel; control, 3401.6 ± 513.1; hetero cKO, 2368.8 ± 249.3 ng/g, P = 0.089, two-tailed unpaired ttest). However, there were no significant changes in the DOPAC and HVA levels.

Discussion
We generated and analyzed Atp2a2 brain-specific heterozygous KO mice to model bipolar disorder and schizophrenia. This is the first report analyzing in vivo significance of ATP2A2 in the central nervous system through a conditional gene-targeting approach. We found that Atp2a2 heterozygous neurons showed a slower decay of the cytosolic Ca 2+ levels, which confirmed an essential role of the ATP2A2 pump for Ca 2+ homeostasis in neurons. Defective DA neurotransmission was observed in the NAc of hetero cKO mice, which suggested a hyperdopaminergic state. Atp2a2 brain-specific hetero cKO mice showed altered behavioral responses to novel environments and impairments in fear memory, which were consistent with mental disorders. However, other phenotypic features related to mood disorders and schizophrenia were absent, which included deficits in prepulse inhibition, working memory, and social behaviors. The genetic background of mutant mice and flanking genes around the targeted locus have been shown to have striking effects, especially when performing behavioral analyses of mouse models of mood disorders and schizophrenia (28,29). Moreover, gene expression, anatomical and electrophysiological phenotypes are also drastically altered by genetic background (30). Previously reported Atp2a2 heterozygous mice were described to have a mixed genetic background of 50% 129/SvJ and 50% Black Swiss (13). However, heterozygous mice are expected to carry more alleles of 129/SvJ and wild-type littermates carry more alleles of Black Swiss near the deleted gene, i.e. flanking gene problem. In other words, the heterozygous and wild-type mice are systematically different in genetic backgrounds, thereby making it difficult to perform and interpret behavioral analyses with these mice. The fact that Swiss Black is an outbred strain may further complicate situations. In the present study, the Atp2a2 mutant mice were originally established using C57BL/6N derived embryonic stem cells with homologous recombination (31) and have been maintained in a C57BL/6N background in our facility. Therefore, their genetic background is relatively clear and phenotypic variations dependent on flanking genes would be minimal in the behavioral abnormalities we observed in the hetero cKO mice. However, the possibility that this homogeneous genetic background of C57BL/6N modified or suppressed phenotypic expression still exists (30). Some phenotypes (e.g. prepulse inhibition, working memory, and social behaviors) might be more susceptible to such suppressant effects than others (e.g. fear memory) in the hetero cKO mice. Another possibility that needs to be considered is whether heterozygosity is equally detrimental in mice and humans. The mouse might be more tolerant to gene dosage reductions than humans. However, in this study, homozygous deletion of ATP2A2 in the whole brain by NesCre transgene resulted in embryonic lethality (Fig. 1B).
Increasing DA transmission by using a dopamine transporter (DAT) inhibitor showed an increase in vertical activity (26), while blockade of DA neurotransmission by a dopamine D1 receptor antagonist showed a reduction in vertical activity in rodents (32). However, these models showed changes in vertical activity that accompanied overall changes in locomotor activity. In contrast, locomotor activity was not significantly changed in the Atp2a2 hetero cKO mice, though the effect was not acute but chronic and would be milder than that in the studies using drugs. Some genetically engineered mouse models showing spontaneous hyperdopaminergic states could be useful to interpret the results in the hetero cKO mice. DAT KO mice had a 5-fold elevation in the extracellular basal levels of DA compared to wild-type and showed increased locomotor activity, while DAT heterozygotes had a 2-fold elevation in the extracellular basal levels of DA but showed no significant changes in locomotor activity (33,34). However, DAT heterozygotes showed some characteristic behavioral abnormalities including associative memory (35,36). DAT A559V heterozygous knock-in mice had a 7-to 8-fold elevation in the extracellular basal levels of DA, while they  showed no significant change in locomotor activity but showed significantly reduced vertical activity (37). The point mutation was derived from patients with mental disorders including bipolar disorder. Considering these conflicting results, the paradoxical findings in the hetero cKO mice, i.e. a 2-fold elevation in the extracellular DA levels (Fig. 7) and no significant change in locomotor activity associated with the abnormal vertical activity (Fig. 4), are also compatible with the findings from other studies. Impairments in fear memory have also been suggested to be related to increased levels of DA in the NAc (27) and thus are relevant to the similar finding observed in human subjects with schizophrenia (38). Thus, increased levels of extracellular DA in the NAc (Fig. 7) may explain a causal relationship between ATP2A2 and mental illness.
Though extracellular DA levels measured by microdialysis were increased, tissue DA levels were not significantly different in this study. Some reports have also shown similar significant differences in extracellular DA levels in the microdialysis without apparent changes in DA levels in the tissue homogenates (37,39). These discrepancies would be partly due to differences in the measurement principles. Microdialysis primarily measures DA concentrations in the extracellular space, whereas in the tissue homogenates DA molecules present in intracellular compartments are also quantified alongside residual extracellular DA. Extracellular DA levels measured by microdialysis would be more appropriate to interpret behavioral phenotypes rather than levels in the tissue homogenates.
DA release was elevated in Atp2a2 hetero cKO mice, as might be expected of mice in which Ca 2+ transients were prolonged. In the presynaptic terminal, the vesicular release was shown to be efficiently regulated via ER-Ca 2+ uptake activity (40). A significant increase in mIPSC frequency, not amplitude, via the presynaptic release of GABA in mouse cerebellar Purkinje cells, was shown to be caused by blocking the activity of the presynaptic ATP2A2 pump with cyclopiazonic acid or thapsigargin (41). It will be important to address how the reduced expression of ATP2A2 affected specifically the DA system in the hetero cKO mice. It may be related to an intrinsic mechanism of DA release in the ventral tegmental area (VTA), as DA neurons have two different release modes, phasic and tonic (42)(43)(44)(45). These firing patterns of DA neurons may make them selectively sensitive, compared to other neuronal cell types, in deficiencies of ER-Ca 2+ uptake activity in the hetero cKO mice. It may also need to be considered that DA release in the NAc can be dynamically modulated without corresponding changes in the spiking of the VTA DA neurons (46). In addition, compensation by increased expression of plasma membrane Ca 2+ ATPase pumps was suggested to have a role in maintaining homeostasis in the vesicular release in Atp2a2 heterozygous pancreatic acini (47). Such a compensatory mechanism could be insufficient in the DA nerve terminals of the hetero cKO mice.
Some patients of Darier's disease have been reported to have epilepsy in addition to psychotic symptoms (48)(49)(50). To examine this, we performed EEG recording but no remarkable epileptic patterns were observed in the waveforms in the hetero cKO mice and the sleep/wake cycles were not significantly different among genotypes (Supplementary Material, Fig. S4). Abnormalities in EEG delta waves during sleep have been reported in schizophrenia patients (51) and its relevance with ATP2A2 has been implicated (52). Thus, a more detailed analysis of EEG could help characterize the neuronal network activity of the Atp2a2 hetero cKO mice.
Among the Darier's disease patients with LOF mutations in ATP2A2, approximately 20% of patients showed psychoses (11). This implies that interactions between LOF mutations in ATP2A2 and other genetic factors, as well as environmental/epigenetic factors, could promote the onset of the disorders. Though many useful mice models for mood disorders and schizophrenia have already been established (1,25,(53)(54)(55), most models exhibit only a few aspects of the disorders, suggesting involvement of multiple genetic factors. It might be beneficial to generate compound mutant mice of Atp2a2 hetero cKO with other established model mouse lines, which could make the phenotypic features more evident. Collectively, our results show that the Atp2a2 hetero cKO mice could be a useful model for some facets of mood disorders and schizophrenia along with other currently used mouse models.

Generation of the Atp2a2 conditional knockout mice
Atp2a2 mutant mice (Atp2a2 tm1a(EUCOMM)Hmgu ) were obtained from the European Conditional Mouse Mutagenesis (EUCOMM) program, which were originally generated by homologous recombination with C57BL/6N-derived embryonic stem cells (31). The allele contained lacZ and neo cassettes inserted in the intron, which was expected to disrupt gene expression and result in a knockout allele, and hence the mouse was regarded as an Atp2a2 heterozygous mouse (KO first allele; hereafter referred to Atp2a2 +/− ). The heterozygous mice have been maintained by crossing with C57BL/6N mice (CLEA Japan Inc.) in our facility, thus they were coisogenic mice with C57BL/6N background. These mice were crossed with FLPe transgenic mice, which had C57BL/6J background (56), to generate mice in which Atp2a2 was flanked by loxP sites. The Atp2a2 flox/+ mice were crossed with the NesCre transgenic mice (B6.Cg-Tg(Nes-cre)1Kln/J [The Jackson Laboratory]; backcrossed and maintained as C57BL/6J congenic strain in our facility), and then backcrossed with C57BL/6N mice three generations (used for behavioral test battery and wheel running activity measurement; Figs 4 and 5) and eight or more generations (used for monoamine and in vivo dialysis; Figs 6 and 7). Considering the crossing with the Flpe transgenic mouse (C57BL/6J background) to obtain the Atp2a2 flox/+ mice, the three and eight generations of backcrossing with C57BL/6N would be expected to reduce genomic material from the two deleter strains, the Flpe and the NesCre transgenic mice, to approximately less than 10% and 0.3%, respectively, while avoiding the flanking gene problem.
All animal experiment protocols were approved by the RIKEN Wako Animal Experiment Committee and all experiments were performed in accordance with the approved guidelines and regulations. All other experimental procedures were approved by the RIKEN Wako Safety Center and were carried out in accordance with the approved guidelines.

Histological analysis
Male mice (12 months old) were transcardially perfused with PBS and fixed with paraformaldehyde (PFA). The fixed samples were dehydrated with ethanol, followed by xylene, and embedded in paraffin. Blocks were cut using a microtome (SM2000R, Leica) into 10 μm serial sections, and mounted onto glass slides, deparaffinized, and subjected to Nissl staining.

ER-Ca 2+ uptake activity assay
The assay was performed as previously described (22). Briefly, to prepare microsomes containing ER membranes, cerebellum tissues were dissected from male mice aged 3-4 months. They were homogenized in a buffer (0.25 M sucrose, 1 mm EGTA, and 50 mm HEPES at pH 7.4) supplemented with protease inhibitor cocktail Halt (Thermo Fisher), centrifuged for 15 min at 4000g at 4 • C, and the supernatants were centrifuged for 30 min at 100 000g at 2 • C. The supernatants were homogenized again and centrifuged for 30 min at 100 000g at 2 • C, and the pellets (microsome fractions) were resuspended in the buffer and kept at −80 • C until the assay. Ca 2+ uptake activity was monitored by measuring the fluorescence of Calcium Green-1 using a spectrophotometer F-2500 (Hitachi). Microsomes were diluted to 600 μl with mobilization buffer containing (in mm: KCl, 110; NaCl, 10; KH 2 PO 4 , 5; MgCl 2 , 2; HEPES, 50 at pH 7.2) supplemented with 2 μm calcium green-1, 10 mm phosphocreatine, 40 U/ml creatine kinase, 1 mm DTT, and 2.5 μg/ml oligomycin. Ca 2+ uptake to microsomes was initiated by adding 0.1 M ATP. After 600 s, 10 μm ionomycin was added to estimate the intra-microsomal Ca 2+ content. Before ending the experiment, 100 μm CaCl 2 and 10 mm EGTA were added to get F max and F min , respectively.

Ca 2+ imaging in primary hippocampal neurons with G-CaMP7
To obtain Atp2a2 wild-type and heterozygous neurons expressing G-CaMP7 (Atp2a2 +/+ ; G-CaMP7 tg/• and Atp2a2 +/− ; G-CaMP7 tg/• , respectively), Atp2a2 heterozygotes were crossed with G-CaMP7 transgenic mice (G7NG817 line) (57). G-CaMP7 expressing hippocampal neurons were prepared from E16.5 mouse embryos according to a standard procedure (58) and cultured in chambered cover glasses. The cells were placed in a thermostatic chamber (Tokai Hit) attached to an inverted laser scanning confocal microscope (FV-3000RS, Olympus) equipped with a resonant scanner and an oil-immersion objective lens (UPLSAPO 40XS). The cells were stimulated with 100 mm KCl to become depolarized. The images were acquired at a frequency of 100 images/s and analyzed. The region of interest (ROI) was manually selected by placing a circular ROI in a cytoplasmic region in the cell body excluding the nucleus. Procedures for imaging with depolarization were established based on the previous reports (59,60), in which ER-Ca 2+ uptake in neurons efficiently worked within seconds after depolarization. Fluorescence intensity decay was defined as the difference of the fluorescence intensities between two time points after depolarization (see legend of Fig. 3 for detail).

Open field test
An open field apparatus (40 × 40 × 30 cm, VersaMax system, AccuScan Instruments) was used, which was illuminated at 100 lx. Total distance traveled, vertical activity (rearing, measured by counting the number of photobeam interruptions), time spent in the center area, and stereotypic counts were recorded for 120 min.

Cued and contextual fear conditioning test
A cued and contextual fear conditioning test was performed as previously described (61). A conditioned stimulus (CS, 55 dB white noise) was presented for 30 s and paired with the unconditioned stimulus (US, 0.3 mA foot shock in the last 2 s of CS). Each mouse received the three CS-US pairings with 2 min intervals on day 1. Twenty-four hours later (On day 2), contextual testing was carried out in the same test chamber for 5 min without CS and US. Subsequent cued testing was carried out 3 h later with altered context (in a novel triangular test chamber) for 6 min. Freezing was recorded during the conditioning, contextual testing and cued testing sessions.

Wheel running activity measurement
Long-term recording of wheel running activity was performed as described previously (24,25). Briefly, male mice (3 months old) were individually housed in cages equipped with a running wheel. The day when the mice were transferred to the cages from conventional breeding cages was designated as day 1, and thereafter cage changings were performed every 2 weeks. Light/dark (12:12 h) cycles were controlled by a PC system, and wheel running activity was recorded by an on-line computer system (O'HARA & CO.). Food and water were constantly available ad libitum. Actograms were drawn using an ImageJ (Fiji) plug-in 'ActogramJ' (62). Video recordings were carried out with a DVD MovieWriter 7 (Corel Corporation).

Monoamine analysis
Male mice (3-4 months old) were individually placed in an open field apparatus (60 × 60 × 30 cm; O'HARA & CO.), which was illuminated at 70 lx. Following 30 min in the apparatus, mice were removed and euthanized by cervical translocation. Whole brains were quickly removed and NAc tissues were excised into ice-cold saline, weighed, frozen with liquid nitrogen, and stored at −80 • C. Tissues were deproteinized, homogenized in a perchloric acid solution, and centrifuged followed by neutralization and HPLC analysis coupled with electrochemical detection as previously described (55,63).

Microdialysis
Male mice (12-17 months old), from the same cohort used in the monoamine analysis, were anesthetized with isoflurane and placed into a stereotaxic frame. A guide cannula (CXG-6, EiCOM) was implanted into the NAc (coordinates relative to bregma in mm, AP +1.3, ML +0.8, and DV −3.9) and fixed onto the skull using acrylic resin PROVINICE (SHOFU INC.). Mice were left to recover for at least 1 week. On the sampling day, a microdialysis probe with an active length of 1 mm (CX-1-6-01, EiCOM) was inserted into the guide cannula. The mouse was placed in a chamber (clear cylindrical container, 14 cm diameter, 22 cm high, Instech Laboratories, Inc.) and continuously perfused with artificial cerebrospinal fluid (aCSF, containing in mm: NaCl, 148; KCl, 4; CaCl 2 , 1.2; MgCl 2 , 0.85) via a syringe pump (CMA/100, Harvard Apparatus) at a flow rate of 1 μl/min for at least 2 h before baseline sampling. Dialysates were collected every 20 min using a refrigerated fraction collector (820, Univentor) at 2 • C with plastic microvials preloaded with perchloric acid to minimize degradation of DA. High potassium stimulation was carried out by perfusing aCSF with 60 mm KCl using a liquid switch (CMA/110, Harvard Bioscience, Inc.). Dialysate samples were kept at −80 • C until use and DA concentration was determined by HPLC coupled with electrochemical detection.
Following microdialysis experiments, mice were anesthetized by isoflurane and transcardially perfused with saline followed by 4% PFA. Brains were dissected, post-fixed in PFA, and sectioned at 50 μm with a vibratome. The sections were mounted onto the slides and stained with hematoxylin and eosin to identify the position of the microdialysis probe. Only mice with probes located in the NAc were included in the analyses.

Experimental Design and Statistical Analysis
Statistical significance was analyzed using GraphPad Prism (GraphPad Software) or Statcel (OMS Publishing Co.). The number of animals used (n) and the specific statistical tests used are indicated in the corresponding part of the main text or figure legends for each experiment. All data are presented as mean ± SEM.

Supplementary Material
Supplementary Material is available at HMGJ online.