Nna1, Essential for Purkinje Cell Survival, Is also Associated with Emotion and Memory

Nna1/CCP1 is generally known as a causative gene for a spontaneous autosomal recessive mouse mutation, Purkinje cell degeneration (pcd). There is enough evidence that the cytosolic function of the zinc carboxypeptidase (CP) domain at the C-terminus of the Nna1 protein is associated with cell death. On the other hand, this molecule’s two nuclear localization signals (NLSs) suggest some other functions exist. We generated exon 3-deficient mice (Nna1N KO), which encode a portion of the N-terminal NLS. Despite the frameshift occurring in these mice, there was an expression of the Nna1 protein lacking the N-terminal side. Surprisingly, the pcd phenotype did not occur in the Nna1N KO mouse. Behavioral analysis revealed that they were less anxious when assessed by the elevated plus maze and the light/dark box tests compared to the control. Furthermore, they showed impairments in context-dependent and sound stimulus-dependent learning. Biochemical analysis of Nna1N KO mice revealed a reduced level of the AMPA-type glutamine receptor GluA2 in the hippocampal synaptosomal fraction. In addition, the motor protein kinesin-1, which transports GluA2 to dendrites, was also decreased. These results indicate that Nna1 is also involved in emotion and memory learning, presumably through the trafficking and expression of synaptic signaling molecules, besides a known role in cell survival.


Introduction
Nna1 (also known as CCP1 or Agtpbp1) was identified as a causative gene for Purkinje cell degeneration (pcd), an autosomal recessive disorder [1]. Defects in Nna1 also cause degeneration not only in Purkinje cells but also in many other neurons of the central nervous system, including cerebellar granule cells, neurons in the deep cerebellar nuclei, neurons in the inferior olive, as well as retinal photoreceptors, olfactory bulb neurons, and individual subpopulations of thalamic neurons [2][3][4][5][6]. Another report shows that mice deficient in this molecule exhibit male infertility due to defective spermatogenesis [7].
The mouse Nna1 gene encodes a protein of 1218 amino acids (aa), and the molecule contains multiple functional domains. The most investigated region is a zinc carboxypeptidase (CP) domain located in the C-terminus (843-1013aa); mutations in this region induce pcd [8,9]. There are two nuclear localization signals (NLSs) present at the N and C termini (144-151aa and 996-1016aa) [8,10], and studies using fusion proteins with green fluorescent protein have shown that Nna1 exists in both the nucleus and cytoplasm [10]. We have so far conducted various examinations to elucidate the function of the Nna1 protein in

Generation of Nna1 N (Nna1 ∆Ex3 ) KO Mice
Assuming that Nna1 mRNA has multiple transcription or translation start sites, we established mice floxed with exon 3 (Figure 1a,b), which encodes a part of the NLS located on the N-terminal side of Nna1. These mice were crossed with TLCN-Cre mice to generate exon 3-deficient mice (Figure 1a,c,d). We named these exon 3-deficient mice and previously generated exons 21, 22 deficient mice, Nna1 N KO and Nna1 C KO mice, respectively. To examine Nna1 expression in the cerebral cortex, cerebellum, and hippocampus of these mice, we performed Western blot analysis using the anti-Nna1 antibody against the Cterminus (1188-1218 aa). Several ladder-like bands, including a 150 kD band, were detected in WT mouse brains, while in Nna1 C KO mice, these bands were barely detected (Figure 1e). In the Nna1 N KO mice, a weak Nna1 band slightly smaller than the 150 kDa and ladder-like bands were observed. Northern blotting using the Nna1 C-terminal probe (exons 17 to 23) revealed stronger signals in the cerebrum, cerebellum, and hippocampus of Nna1 N KO mice than in WT mice (Supplementary Materials, Figure S1a). Since deletion of the exon 3 of the Nna1 gene causes a frameshift, the presence of truncated Nna1 proteins in the Nna1 N KO mice suggests that Nna1 mRNA has multiple translation start sites and escaped from NMD. The increased amount of mRNA in Nna1 N KO mice could be a compensatory upregulation for the loss of intact Nna1 proteins.

Normal Morphology in Nna1 N KO Mice
We next performed morphological analysis to investigate the brain phenotype of Nna1 N KO mice. Macroscopic images of the adult brains showed no cerebellar atrophy seen in Nna1 C KO and pcd mice (Figure 2a). We next performed histological analyses on the cerebellum by Calbindin-D28K immunohistochemistry, and there were no differences in the lobular and laminar structures between WT and Nna1 N KO mice (Figure 2b,c). Double staining for Car8 and VGluT1, markers in Purkinje cells and parallel fiber terminals, respectively, showed no discernible differences in dendritic arborization and the spine formation of Purkinje cells and in synapse formation with parallel fibers on distal spiny branchlets (Figure 2d,e). Immunohistochemistry for VGluT2, a marker for climbing fiber terminals, indicated no significant differences in the distribution and wiring of climbing fiber synapses on proximal shaft dendrites (Supplementary Materials, Figure S2a-c). In addition, Nissl staining of the hippocampus showed no significant differences in the histology and cellular alignment between WT and Nna1 N KO mice (Figure 2f  . loxP sites are indicated by triangles. Red bars indicate 5′ or 3′ blot analysis. 5′ outer probe by ScaI, Neo probe and 3′ outer probe b for genomic DNAs from wild-type (WT) and chimeric mice. WT: l in each genotype). (c) Genotypes of WT or Nna1 N KO mice were id detecting exon 3 deletion (n = 3 mice in each genotype). (e) Weste antibody (Frontier Institute). Lanes 1, 4, and 7 indicate WT mice, KO mice, and Lanes 3, 6, and 9 indicate Nna1 C KO mice. In the Nn pared to full-length Nna1 protein were detected in Nna1 N KO mi ladder-like patterns (n = 3 mice in each genotype). . loxP sites are indicated by triangles. Red bars indicate 5 or 3 probe regions used for Southern blot analysis. 5 outer probe by ScaI, Neo probe and 3 outer probe by SpeI. (b) Southern blot analysis for genomic DNAs from wild-type (WT) and chimeric mice. WT: l, 3, 5, chimera: 2, 4, 6 (n = 3 mice in each genotype). (c) Genotypes of WT or Nna1 N KO mice were identified by PCR. (d) RT-PCR for detecting exon 3 deletion (n = 3 mice in each genotype). (e) Western blot analysis with anti-Nna1 antibody (Frontier Institute). Lanes 1, 4, and 7 indicate WT mice, Lanes 2, 5, and 8 indicate Nna1 N KO mice, and Lanes 3, 6, and 9 indicate Nna1 C KO mice. In the Nna1 N KO brain, lower bands compared to full-length Nna1 protein were detected in Nna1 N KO mice (asterisks) and faint bands in ladder-like patterns (n = 3 mice in each genotype).

Nna1 N KO Mice Are Impaired in Emotional and Memory Learning
In the open-field test, Nna1 N KO mice were more hyperactive and spent much more time in the center of the open-field area compared with WT mice (Figure 3a-c), while there was no significant difference in their movement speed (Figure 3d). In the next light-dark transition test, Nna1 N KO mice made a more significant number of transitions between the light and dark areas than WT mice (Figure 4a-c). However, there was no significant difference in the total distance traveled (Figure 4b,d). There was no significant difference in the time spent in the light area of the box between WT and Nna1 N KO mice, but Nna1 N KO mice seemed to spend more time in the dark area with more motility than the wild-type mice, indicating a trend of hyperactivity (Figure 4b,e). Furthermore, we performed the elevated plus-maze test to evaluate anxiety (Figure 5a). Although there was no significant difference in the total distance traveled (Figure 5b Mol. Sci. 2022, 23, x FOR PEER REVIEW Figure 2. Immunohistochemical observation in the cerebellum and hippocam (a) Macro image of the brains from Nna1 N KO mice and WT mice in the adult genotype). No significant atrophy on the parasagittal image of the cerebe tween Nna1 N KO and WT mice. (b,c) Calbindin-D28K IHC on parasagittal each genotype). No significant expression difference was observed in the cere KO and WT mice (n = 3 mice in each genotype). (d,d′,d″) and (e,e′,e″) Car8 a also showed no significant difference between KO and WT mice. (f,f′,f″) and on parasagittal sections (n = 3 mice in each genotype). No significant morpho difference was observed between Nna1 N KO and control mice. Scale bars: 2 Scale bars: 0.5mm in (b,c,f,g), Scale bars: 5 µm in (d′,d″).       showed more transition latency than WT mice. (d) No significant difference wa Nna1 N KO and WT mice in the travel distance in the boxes. (e) Nna1 N KO mic time in the dark box than the control mice. WT, n = 6; Nna1 N KO, n = 10, * p < 0 All values presented are means ± SEM. "ns" means not significant. Finally, we performed the contextual-and cued-fear conditionin the effect of N-terminal deletion of Nna1 N KO on learning memory. Ad weeks were used for the conditioning test on day 1, the contextual test cued test on day 3 ( Figure 6a). For the conditioning test, mice were a freely for 3 min and then presented with 55 dB white noise as a conditio for 20 s, including the last 2 s of a foot shock (0.2 mA, 2 s) as an uncon (US). Then similar stimulation, 60 s of CS, including the last 2 s of US, w ( Figure 6b). There was no noticeable difference in response to the stimul Nna1 N KO mice ( Figure 6c). However, in a spatial-dependent learnin considerable differences in freezing behavior between WT and KO m Furthermore, in a sound-dependent learning test conducted the next da in Nna1 N KO mice was lower than in WT mice (Figure 6f,g). Finally, we performed the contextual-and cued-fear conditioning tests to examine the effect of N-terminal deletion of Nna1 N KO on learning memory. Adult mice aged 8-10 weeks were used for the conditioning test on day 1, the contextual test on day 2, and the cued test on day 3 ( Figure 6a). For the conditioning test, mice were allowed to explore freely for 3 min and then presented with 55 dB white noise as a conditioned stimulus (CS) for 20 s, including the last 2 s of a foot shock (0.2 mA, 2 s) as an unconditioned stimulus (US). Then similar stimulation, 60 s of CS, including the last 2 s of US, was repeated twice (Figure 6b). There was no noticeable difference in response to the stimuli between WT and Nna1 N KO mice ( Figure 6c). However, in a spatial-dependent learning test, there were considerable differences in freezing behavior between WT and KO mice (Figure 6d,e). Furthermore, in a sound-dependent learning test conducted the next day, the freezing rate in Nna1 N KO mice was lower than in WT mice (Figure 6f,g).

Nna1 N KO Mice Have a Different Subunit Composition of Glutamate Receptors in the Hippocampus
Since hippocampal function is known to be associated with memory deficits, amined AMPA-type glutamine receptor subunit expression, which is closely rel hippocampal-dependent behavior [16]. Immunohistochemical analysis reveal creased tendency of GluA1 and GluA2 expressions (Figure 7a). Western blots of hippocampal lysates indicated significantly increased GluA2. At the same time, showed an increasing trend but no significant difference (Figure 7b). Interestingly glutamylated tubulin, whose side chains are shortened by the peptidase activity o protein, was significantly increased (Figure 7c

Nna1 N KO Mice Have a Different Subunit Composition of Glutamate Receptors in the Hippocampus
Since hippocampal function is known to be associated with memory deficits, we examined AMPA-type glutamine receptor subunit expression, which is closely related to hippocampal-dependent behavior [16]. Immunohistochemical analysis revealed increased tendency of GluA1 and GluA2 expressions (Figure 7a). Western blots of crude hippocampal lysates indicated significantly increased GluA2. At the same time, GluA1 showed an increasing trend but no significant difference (Figure 7b). Interestingly, polyglutamylated tubulin, whose side chains are shortened by the peptidase activity of Nna1 protein, was significantly increased (Figure 7c, p < 0.05). Furthermore, we observed a mild increase in GluA2 in the Nna1 N KO cerebellum (Supplementary Materials, Figures S3 and S4). Since a change in AMPA-type glutamine receptors content in the synapses may lead to changes in synaptic transmission in the Nna1 N KO hippocampus, we measured the GluA1 and GluA2 in the synaptosome fraction. Surprisingly, there was a significantly decreased concentration of GluA2 in the synaptosome fraction of the Nna1 N KO hippocampus, whereas there was no change in GluA1 (Figure 7d). Furthermore, when kinesin-1 was measured, which is involved in the transport of GluA2-containing vesicles, its concentration was higher in the crude fraction and lower in the synaptosome fraction, like GluA2 (Figure 7d,e). These results suggest that in Nna1 N KO mice, there are defects in the transport of GluA2 vesicles by kinesin-1 complexes. concentration was higher in the crude fraction and lower in the synaptosome fraction, like GluA2 (Figure 7d,e). These results suggest that in Nna1 N KO mice, there are defects in the transport of GluA2 vesicles by kinesin-1 complexes. . Note that significantly increased expression of GluA2 was observed in the crude fraction (crude) of Nna1 N KO mice. Furthermore, a significant reduction was observed in the synaptosome fraction (synapto) as well kinesin-1. There was no significant modification of GluA1. All values presented are means ± SEM from 3 experiments. * p < 0.05, Student's t-test. Β-Actin is the loading control.

Discussion
In this study, we generated Nna1 ΔEx3 KO (Nna1 N KO) mice, in which the N-terminal exon 3 of Nna1 was deleted to analyze the phenotype. We found that Nna1 N KO mice showed no ataxia as seen in Nna1 null [14] or pcd mice lines and no difference in body Note that increased expression of GluA2 and polyG was observed in the Nna1 N KO mice. All values presented are means ± SEM from 3 experiments. * p < 0.05, Student's t-test. (d,e) kinesin-1 and GluA2 showed a similar expression pattern from the crude fraction (crude) to the synaptosome fraction (synapto). Note that significantly increased expression of GluA2 was observed in the crude fraction (crude) of Nna1 N KO mice. Furthermore, a significant reduction was observed in the synaptosome fraction (synapto) as well kinesin-1. There was no significant modification of GluA1. All values presented are means ± SEM from 3 experiments. * p < 0.05, Student's t-test. B-Actin is the loading control.

Discussion
In this study, we generated Nna1 ∆Ex3 KO (Nna1 N KO) mice, in which the N-terminal exon 3 of Nna1 was deleted to analyze the phenotype. We found that Nna1 N KO mice showed no ataxia as seen in Nna1 null [14] or pcd mice lines and no difference in body weight compared to WT mice. Nna1 mRNA is frameshifted by the exon 3 deletion and was expected to be degraded by NMD. However, Northern blot analysis using the Nna1 probe (exons 17-23) showed two prominent bands, similar to the WT (Supplementary Materials, Figure S1). Western blotting with a Nna1 antibody against the C-terminal region showed multiple bands with a maximum molecular weight of approximately 150 kD in the WT mice. In the Nna1 N KO, there was a band with slightly lower molecular weight than the WT band (~150 kD) and multiple bands in a ladder-like pattern (Figure 1). In the Nna1 C KO mice lacking exon 21, 22 encoding carboxypeptidase domain, neither Nna1 mRNA nor protein was detected [14], indicating that mRNA degradation occurs in the Nna1 C KO mice by NMD. However, in the Nna1 N KO mice, the amount of Nna1 mRNA was increased in some brain regions and truncated Nna1 protein was observed. These findings raise some possibilities: (1) the Nna1 gene has multiple sites for transcription start, (2) multiple isoforms with exon skipping, and (3) Nna1 mRNA has multiple sites for translation start. Furthermore, the Nna1 N KO mice lack the N-terminal side of the Nna1 protein but retain the carboxypeptidase domain of the C-terminal region, enabling us to evaluate the function of the N-terminal side of the Nna1 protein. Notably, we cannot exclude the possibility that the low expression of the truncated Nna1 proteins, in addition to the loss of the N-terminus of the Nna1 protein, contributes to the phenotype of the Nna1 N KO mice.
Although the Nna1 N KO mice showed no severe cerebellar phenotypes, such as Purkinje cell death, they exhibited a variety of phenotypes in systematic behavioral analyses, such as hyperactivity, reduced anxiety-like behavior, and impaired memory learning. These findings indicate that Nna1 is involved not only in neuronal survival but also in higher brain functions related to emotion and memory. Furthermore, both contextual and cued fear conditioning tests showed impaired learning ability, suggesting that Nna1 was involved in altered synaptic transmission in the hippocampus (Figure 7d). Thus, we investigated the expression level of AMPA-type glutamate receptors in the hippocampus. Immunohistochemistry showed that GluRA1 and GluRA2 were elevated, while GluRA3 and GluRA4 were unchanged ( Figure 7a); detailed Western blotting analysis revealed that only GluA2 was significantly increased in the hippocampal crude fraction. In the synaptosomal fraction, GluA2 was significantly decreased in concentration, but GluA1 was unchanged. We next measured the amount of kinesin-1 that is involved in AMPA receptor transport [17] and found that, as with GluA2, its concentration was high in the crude fraction and low in the synaptosome fraction (Figure 7b-e), suggesting that there is some impairment occurring in kinesin-1 and GluA2 complex transport. Therefore, we examined the amount of polyglutamylated tubulin, which is associated with Nna1 [18], and found that polyglutamylated tubulin was significantly increased (Figure 7b,c). These findings suggest that Nna1 is involved in GluA2 transport to the synapse by regulating the polyglutamine content of tubulins. Notably, a recent study has reported patients with global developmental delay and hypotonia with a novel homozygous c.3293G>A mutant of the NNA1 gene in a consanguineous family [19].
The post-translational modifications (PTMs) of tubulin occur in the microtubules (MTs) of neurons and play essential roles in the dynamics and organization of MTs, thus exerting a direct effect on cell functioning, which is critical in human health and disease [20]. Nna1 is a cytoplasmic carboxypeptidase that modifies the C-terminal tail of tubulins, such as polyglutamylation, detyrosination, and generation of ∆2-tubulin [21,22]. The carboxyterminus of tubulin is polyglutamylated and is exposed on the outer surface of tubulin during microtubule assembly. The length of the polyglutamate side chain on tubulin is vital for neuronal stability and survival [23]. In this study, we showed that the hippocampus of Nna1 N KO mice expresses higher levels of polyglutamylated tubulin than WT mice (Figure 7b). This observation indicates that the N-terminal loss may result in reduced polyglutamate pruning of Nna1. In addition, the "modified" tubulin rail, where motor protein kinesin-1 binds, makes it difficult for kinesin-1 to move, presumably related to the reduced transport rate of GluA2.
There are two possible reasons for the increased expression of GluA2 in the hippocampal homogenates in the Nna1 N KO mice. One is that the transport of GluA2 from the soma to the synaptic terminals is reduced due to a dysfunction in the axon transport and accumulates in the soma; N-terminal loss could ultimately reduce GluA2 transport to the synapse and induce qualitative changes in synaptic transmission. Another is that the decrease in GluA2 in synaptosomes may provide feedback to the soma to produce more GluA2 and compensate for the decrease in the synapse terminals. The present study demonstrates that Nna1 functions not only in neuronal survival but also in emotion and memory, possibly via synaptic transmission by polyglutamate modification of tubulins.

Generation of Nna1 N Knockout (KO) Mice
Male chimera mice were generated by injection of recombinant ES cells into eight-cell stage embryos from ICR mice (MGI:5462094, SLC Japan), and then heterozygous mice (Nna1 flox(neo+) ) were obtained by natural mating with C57BL/6N female mice (MGI:5657107, Charles River Japan). To generate a Nna1 knockout allele lacking exon 3 (Nna1 ∆Ex3 ), heterozygous F1 mice were crossed with TLCN-Cre mice (MGI:3042494) [26,27], which ubiquitously express Cre recombinase. Double heterozygous mice (TLCN-Cre; Nna1 flox(neo+) ) were crossed with C57BL/6N mice to generate the Nna1 N KO allele (Nna1 ∆Ex3 ). Finally, Nna1 N KO mice (Nna1 ∆Ex3/∆Ex3 mice) were generated by crossing heterozygous pairs (Nna1 ∆Ex3/wt mice). All mice used in this study were maintained in the C57BL/6N background. Animal care and experimental protocols were approved by the Animal Experiment Committee of Niigata University and were carried out under the Guidelines for the Care and Use of Laboratory Animals of Niigata University (approved number: SA00733, SA01091). Animals were handled under the guidelines established by the Institutional Animal Care and Use Committee of Niigata University. The following measures were taken to minimize animals' suffering during experiments: restlessness, vocalizing, loss of mobility, failure to groom, open sores/necrotic skin lesions, guarding (including licking and biting) a painful area, and a change in body color. If these signs were observed, they were excluded from further participation and treated appropriately according to the approved protocol. The mice used were 3-50 weeks old and had 8-29 g body weight. Time-point of each experiment was described in each result. No randomization was performed in this study.

Genotyping PCR
Genotyping by PCR was performed as follows. Genomic DNA was extracted from the tips of the tails of wild-type and Nna1 mutant mice: the tail tissues were incubated with 0.025 N NaOH and 2 mM EDTA for 30 min at 100 • C and then mixed with an equal volume of 40 mM Tris-HCl (pH 8.0) at around 20 • C. The extracted DNA was used as a template for the PCR reaction using the Nna1-lox forward and Nna1-lox reverse primers (Table 1). PCR was performed using Quick Taq HS Dye Mix (Toyobo, Osaka, Japan) under the following PCR conditions: 95 • C for 30 s, 30 cycles of 95 • C for 10 s, 60 • C for 30 s, and 72 • C for 1 min, followed by 72 • C for 5 min. The PCR products were separated by agarose gel electrophoresis to identify the DNA bands; 1050 bp and 560 bp were amplified from the WT and Nna1 KO allele, respectively.

Behavior Tests
Open field test: Locomotor activity was measured using an open field test, similar to the previous one [30]. The chamber was made of a square platform with 50 cm × 50 cm × 40 cm (O'Hara and Co. Tokyo, Japan) and illuminated with a light intensity of 100 lux. Mice were placed in the corner of the field and left for 10 min to allow free exploration. During the test, the total distance traveled and the time spent in the central region were recorded and automatically calculated using Image OFCR software (O'Hara and Co., LTD. Tokyo, Japan; see 'Image analysis for behavioral tests').
Light/dark transition test: A light/dark transition test was conducted as previously described [31]. The apparatus comprised a cage (21 cm × 42 cm × 25 cm) divided into two sections of equal size by a partition with a door (O'Hara Co.). One chamber was brightly illuminated (light chamber), while the other was not (dark chamber). Mice were placed into the dark chamber and allowed to move freely between the two chambers with the door open for 10 min. The total number of transitions, time spent in each compartment, first latency of movement to the light chamber, and distance traveled were recorded automatically using Image LD software (O'Hara and Co., LTD. Tokyo, Japan; see 'Image analysis for behavioral tests').
Elevated plus maze test: Elevated plus maze test was performed as described previously [32]. The elevated plus maze consisted of two open arms (25 cm × 5 cm) and two enclosed arms of the same size, with 15 cm high transparent walls. The arms and central square were made of white plastic sheets, elevated to 60 cm above the floor. To avoid animals falling from the apparatus, 3-mm high Plexiglas sides were used for the open arms. Arms of the same type were arranged on the opposite side. This device was set up under low illumination (center square 100 lux). Each mouse was placed in the central square of the maze (5 cm × 5 cm), facing one of the closed arms, and then behavior was recorded during a 10-min test period. The number of entries and the time spent on open and enclosed arms were recorded. For data analysis, we recorded the percentage of entries onto open arms, the staying time on open arms (seconds), the number of total entries, and the total distance traveled (centimeters). Data acquisition and analysis were performed automatically using Image EP software (O'Hara and Co., LTD. Tokyo, Japan; see 'Image analysis for behavioral tests').
Contextual and cued fear conditioning test: Fear conditioning was performed as described previously [33]. Each mouse was placed in a test chamber (33 cm × 25 cm × 28 cm) and allowed to explore freely for 3 min. Then, a 55-dB white noise, which served as the conditioned stimulus (CS), was presented for 20 s, and during the last 2 s of CS presentation, a foot shock (0.2 mA, 2 s), which served as the unconditioned stimulus (US) to mice was given. Two more CS-US pairings were presented with an inter-stimulus interval of 40 s. Twenty-four hours after the conditioning, a contextual test was performed in the same chamber without CS or US stimulus. Forty-eight hours after the conditioning, a cued fear memory was tested in a triangular chamber (33 cm × 33 cm × 32 cm) made of opaque white plastic and allowed to explore freely for 1 min. Subsequently, each mouse was given CS presentation for 3 min. In each session, data acquisition, and control of stimuli (i.e., shocks) were automatically performed, and the percentage of time spent freezing was calculated using Image FZ software (O'Hara and Co., LTD. Tokyo, Japan; see 'Image analysis for behavioral tests').

Subcellular Fraction and Western-Blot Analysis
Subcellular fractions were prepared following Carlin's method [34] with minor modifications. All procedures were performed at 4 • C. Briefly, WT and Nna1 N KO mice were decapitated after cervical dislocation, and the cerebellum and the hippocampus were immediately dissected, removed, and immersed into the homogenization buffer (320 mM sucrose and 5 mM EDTA, containing complete protease inhibitor cocktail tablet (Complete Mini; Roche, Mannheim, Germany) and centrifuged at 1000× g for 10 min. The supernatant was centrifuged at 12,000× g for 10 min, and the resultant pellet was re-suspended in homoge-nization buffer at the P2 fraction. The P2 fraction was layered over a 1.2 M/0.8 M sucrose gradient and centrifuged at 90,000× g for 2 h. The synaptosome fraction was collected from the interface. The protein concentration was determined using BCA Protein Assay Reagent (Thermo Fisher Scientific Inc. Waltham, MA, USA). An equal volume of SDS sample buffer [125 nm Tris-Cl (pH 6.8), 4% SDS, 20% glycerol, 0.002% BPB, 2% 2-mercaptoethanol] was added to the sample fractions and boiled for 5 min at 95-100 • C.
Dissected brains were post-fixed for 2 h in the same fixative at 4 • C. After dehydration, the brain was embedded with paraffin, and sagittal sections 5 µm thick were stained with Cresyl Violet (Nissl) (Invitrogen).

Image Analysis for Behavior Tests
The application software used for the behavioral studies (Image OFCR, LD, EP, and FZ) was based on the public domain NIH Image program (developed by the U.S. National Institutes of Health and available at http://rsb.info.nih.gov/nih-image, accessed on 1 june 2013) and ImageJ program (http://rsb.ingo.nih.gov/ij/, accessed on 1 june 2013), with some modification for each test (available through O'Hara and Co., Japan)

Statistical Analysis
Results were presented as means ± SEM (standard error of the mean). For Western blotting, statistical analyses were performed using Student's t-test. For mouse behavior tests, statistical analyses were performed using Student's t-test. Values in the graphs are expressed as means ± SEM (standard error of the mean). Statistical significance was set at a value of p < 0.05. Sample calculation and tests for outliers were not performed.

Institutional Review Board Statement:
The study was conducted in accordance with the Animal Experiment Committee of Niigata University and was carried out following the Guidelines for the Care and Use of Laboratory Animals of Niigata University (approved number: SA00733, SA01091). Data Availability Statement: All data generated or analyzed during this study are included in the manuscript and Supporting Files. Raw data have been provided for mean population data shown in figures.