Global knockdown of glutamate decarboxylase 67 elicits emotional abnormality in mice

Reduced expression of glutamate decarboxylase 67 (GAD67), encoded by the Gad1 gene, is a consistent finding in postmortem brains of patients with several psychiatric disorders, including schizophrenia, bipolar disorder and major depressive disorder. The dysfunction of GAD67 in the brain is implicated in the pathophysiology of these psychiatric disorders; however, the neurobiological consequences of GAD67 dysfunction in mature brains are not fully understood because the homozygous Gad1 knockout is lethal in newborn mice. We hypothesized that the tetracycline-controlled gene expression/suppression system could be applied to develop global GAD67 knockdown mice that would survive into adulthood. In addition, GAD67 knockdown mice would provide new insights into the neurobiological impact of GAD67 dysfunction. Here, we developed Gad1tTA/STOP−tetO biallelic knock-in mice using Gad1STOP−tetO and Gad1tTA knock-in mice, and compared them with Gad1+/+ mice. The expression level of GAD67 protein in brains of Gad1tTA/STOP−tetO mice treated with doxycycline (Dox) was decreased by approximately 90%. The GABA content was also decreased in the brains of Dox-treated Gad1tTA/STOP−tetO mice. In the open-field test, Dox-treated Gad1tTA/STOP−tetO mice exhibited hyper-locomotor activity and decreased duration spent in the center region. In addition, acoustic startle responses were impaired in Dox-treated Gad1tTA/STOP−tetO mice. These results suggest that global reduction in GAD67 elicits emotional abnormalities in mice. These GAD67 knockdown mice will be useful for elucidating the neurobiological mechanisms of emotional abnormalities, such as anxiety symptoms associated with psychiatric disorders.


Introduction
γ-Aminobutyric acid (GABA), a major inhibitory neurotransmitter, regulates a variety of biological functions. GABA is synthesized from glutamate by glutamate decarboxylase (GAD) existing two isoforms with different molecular weights, 67 kDa (GAD67) and 65 kDa (GAD65), which are independently encoded by the Gad1 and Gad2 genes, respectively [1,2]. Since GAD67 and GAD65 proteins have different subcellular distributions and a cofactor association [3][4][5][6], the physiological roles of GAD67 and GAD65 may be different in the brain.
Decreased expression of the full-length GAD1 transcript and GAD67 protein is one of the most consistent findings in the brains of subjects with several psychiatric disorders, including schizophrenia, bipolar disorder and major depressive disorder [7][8][9][10][11][12]. Alternative splicing of GAD1 and the epigenetic state may play roles in brain development and the risk of schizophrenia [12].
Recent studies with whole exome sequencing of schizophrenic patients identified missense mutation mapping at the GAD1 gene, which caused a reduction in GAD67 enzymatic activity by ~ 30% due to impaired homodimerization [13,14]. Therefore, the functional impairment of GAD67 associated with genetic mutations is involved in the neurobiological mechanisms of several psychiatric disorders. To elucidate the physiological roles of GAD67 in mature brains, a study using transgenic animals with reduced GAD67 expression would be helpful. Because homozygous Gad1 knockout (Gad1 −/− ) is lethal in newborn mice [15], conditional Gad1 −/− mice generated with a Cre-loxP strategy have often been used for neurobiological studies [16][17][18][19][20]. These studies have provided much information about the function of GAD67 in targeted cells. However, behavioral and neurochemical consequences of the global dysfunction of GAD67 in mature brains are not fully understood. Several studies have been performed to investigate GAD67 function using mice with GAD67 haplodeficiency; however, physiological changes, such as GABA reduction in the brain and behavioral abnormalities, were mild in those mice [15,[21][22][23][24].
The tetracycline-controlled gene expression/suppression system allows reduction in a gene of interest by administration of antibiotic tetracyclines and its derivative doxycycline (Dox) [25]. This system requires two distinct transgenic mice: one has a cell type-specific promoter driving a tetracycline-controlled transcriptional activator (tTA)-expressing allele, and the other has a tetracycline operator site (tetO) binding to tTA and driving the expression of the gene of interest. By crossing them, a biallelic knock-in mouse (tTA/tetO) can be obtained with a tTA-mediated gene induction system, which can be turned off by the administration of Dox.
We hypothesized that the tetracycline-controlled gene expression/suppression system could be applied to generate transgenic mice able to postnatally knockdown GAD67. In brief, two knock-in mice were generated: (1) Gad1 STOP−tetO knock-in mice, which were generated by inserting the tetO sequence following the STOP sequence upstream of the Gad1 translation initiation site (Fig. 1a), and (2) Gad1 tTA knock-in mice [26], which express tTA proteins under the control of an endogenous Gad1 promoter (Fig. 1b). The Gad1 tTA/ STOP−tetO mice can produce GAD67 protein (Fig. 1c) and survive into adulthood. In addition, the expression of GAD67 protein in the Gad1 tTA/STOP−tetO mice is suppressed by treatment with Dox (Fig. 1d). Therefore, Dox-treated Gad1 tTA/STOP−tetO mice can be used as GAD67 knockdown mice for studying the behavioral and neurochemical consequences of the global loss of GAD67 in mature brains.
The major aim of this study was to examine the phenotypes of mice with GAD67 suppression at adult stage. In this study, we successfully developed Gad1 STOP

Ethics
This study was performed in accordance with the Guidelines for Animal Experimentation at Gunma University Graduate School of Medicine and was approved by the Gunma University Ethics Committee (Permit number: 14-006 and 19-009). Every effort was made to minimize the number of animals used and their suffering.

Experimental design
We first generated Gad1 STOP−tetO knock-in mice and assessed the Gad1 knockout phenotypes (GAD67 deletion, neonatal death and cleft palate) of homozygous Gad1 STOP−tetO/STOP−tetO mice to confirm the elimination of Gad1 gene function by inserting the Neo-STOP-tetO cassette. We then crossed heterozygous Gad1 tTA/+ and Gad1 STOP−tetO/+ mice to obtain Gad1 tTA/STOP−tetO mice and confirmed tTA-mediated GAD67 expression. Afterward, we evaluated whether tTA-mediated GAD67 expression was suppressed by treatment with Dox. It has been reported that GAD67 haplodeficient mice demonstrate an approximately 40% reduction in GAD67 protein in the brain compared with wild-type mice [27]. Therefore, we judged that GAD67 knockdown mice were successfully developed when the expression level of GAD67 protein in the brain was reduced by more than 40%. These experiments were performed in mice of both sexes. In the behavioral tests, male mice were only used. We prepared two independent cohorts comparing Gad1 +/+ mice and Gad1 tTA/STOP−tetO mice at the ages of 8 to 10 weeks. One cohort was used for assessing the body weights, motor coordination performance, and GABA and glutamate contents in their brains. The body weights and motor coordination performance of the mice were measured 3 weeks after starting Dox treatment. Immediately after the motor coordination test, the mice were killed by decapitation, and the frontal cortex (FCX), hippocampus (HIP) and cerebellum (CER) were quickly dissected. The collected tissues were immediately frozen in liquid nitrogen and stored at − 80 °C until use. The frozen tissues were used for measuring GABA and glutamate contents. Another cohort was used for the open-field test and PPI test. Three weeks after treatment with Dox, the open-field test was conducted. After testing, the mice were returned to their home cage, and treatment with Dox was continued. Two days after the open-field test, the acoustic startle responses and prepulse inhibition (PPI) responses were assessed in the mice.
The generation of Gad1 tTA knock-in mice has already been described [26]. In these mice, the tTA2 cDNA followed by the SV40 polyadenylation signal was inserted into exon 1 of the Gad1 gene in frame with the translation initiation codon, and the tTA protein was expressed under the control of an endogenous Gad1 promoter (Fig. 1b).
The animals were housed at 2-3 mice per cage (16.5 × 27 × 12.5 (H) cm) and had free access to food and water. The animal rooms for breeding and experiments were maintained at 22 ± 3 °C with a 12-h light-dark cycle (lights on at 6:00, lights off at 18:00).

Palate formation
Mouse neonates were killed by decapitation, and the lower jaw was removed. The cleft palate of the mouse was determined under a stereoscopic microscope.

Immunoblot analysis
The mice were killed by decapitation. The brain hemispheres of neonates and the FCX, HIP and CER of adult mice were quickly dissected on an ice-cold stainless plate. The tissues were immediately frozen in liquid nitrogen and stored at − 80 °C until use. The frozen tissues were homogenized in ice-cold buffered sucrose (0.32 M) solution containing 20 mM Tris-HCl (pH 7.5) and protease inhibitor cocktail (P8340, Sigma-Aldrich, Inc.). The homogenates were centrifuged at 1000×g for 10 min at 4 °C, and the supernatants were collected as the protein samples. The protein concentrations were determined using a TaKaRa BCA Protein Assay Kit (T9300A, Takara Bio Inc., Japan).
The protein samples were diluted with electrophoresis sample buffer. Proteins (1.5 μg) were separated by 8% SDS-polyacrylamide gels and transferred to a PVDF membrane. Blots were probed with respective antibodies to GAD65/67 (1:1000, rabbit polyclonal antibody) [29] and GAD67 (1:1000, mouse monoclonal antibody, Millipore, Code No. MAB5406). Immunoblots were developed using horseradish peroxidase-conjugated secondary antibodies (GE Healthcare) and then detected with chemiluminescence reagents (ECL prime, GE Healthcare) and visualized by the Light Capture AE-9672 (ATTO Co., Ltd.). After the detection of immunoblots, the blotting membranes were washed with PBS several times and reprobed with a mouse monoclonal antibody to β-actin (1:10,000, Medical & Biological Laboratories Co. Ltd., Code No. M177-3). The immunoblots of β-actin were developed and visualized by the same protocol described above. The density of the bands was determined using ImageJ software. The band densities of β-actin were used as the loading control. The relative expression level of GAD67 to β-actin was calculated and used for comparisons between the genotypes.

Double-label immunofluorescence analysis
Deeply anesthetized mice by continuous inhalation of isoflurane were fixed by perfusion with Mildform 10 N (containing 3.7-4.3 w/w% formaldehyde; FUJIFILM Wako Pure Chemical Co., Osaka, Japan) through the left ventricle. The brain was removed and postfixed in Mildform 10 N overnight at 4 °C. The brain hemispheres were cut into 50-μm-thick sagittal sections by a vibrating blade tissue slicer (Neo-LinearSlicer MT, Dosaka EM Co., Ltd., Kyoto, Japan).
Free Three independent mice in the respective groups were assessed. The names of brain regions were referenced to the Allen Mouse Brain Atlas (https ://allen insti tute.org/).

Motor coordination test
The performance of motor coordination in mice was tested by a rotarod apparatus (Ugo Basile, Comerio, Italy) according to a previous report [16]. Briefly, each mouse was placed in a separate lane of the apparatus on a rotating cylinder (3 cm diameter) at 20 rounds per minute. The latency until the mouse fell from the cylinder (up to 120 s) was recorded in three consecutive trials with 2-3 min intervals, and the median latency was used for the following analysis. If the mouse did not fall within 120 s, the latency to fall was recorded as 120 s.

Open-field test
Each mouse was placed in the center of an open-field apparatus (50 cm × 50 cm × 40 (H) cm) that was illuminated by light-emitting diodes (30 lx at the center of the field) and allowed to move freely for 5 min. The data were collected and analyzed using ImageJ OF4 (O'Hara & Co., Ltd., Tokyo, Japan), which is modified software that is also based on the public domain ImageJ program. The procedure was performed according to our previous report [30].

GABA and glutamate contents in the brains
The frozen tissues were weighed and then homogenized by BioMasher II (Nippi, Inc., Tokyo, Japan) in 500 μL of 0.1% formic acid in acetonitrile (Wako, Tokyo, Japan) containing an internal standard 2-morpholinoethanesulfonic acid (2-MES; Dojindo, Tokyo, Japan). The standard was spiked at a final concentration of 10 μM. The homogenates were centrifuged at 15,000×g for 15 min at 4 °C, and then, the supernatants were collected and filtered through an ISOLUTE PLD + column (Biotage Japan Ltd., Tokyo, Japan). The 40 μL filtrates were lyophilized and stored at − 20 °C.
At the time of analysis, the lyophilized samples were dissolved in 1.25 mL of ultrapure water. The prepared sample solutions (3 μL) were then injected on ultraperformance liquid chromatograph coupled to triplequadrupole mass spectrometer (LC/MS) (LCMS-8050; Shimadzu, Kyoto, Japan). The chromatographic conditions were according to the Shimadzu method package using a pentafluorophenylpropyl (PFPP) column. The MS settings, data acquisition and data analysis were in accordance with the manufacturer's instructions for analyzing Primary Metabolites version 2.0 (Cat. #: 225-24865A, Shimadzu). The relative values of metabolites from the internal standard 2-MES and the weight of corresponding tissues were calculated and used for the following data analysis.
Standard solutions containing GABA (A2129, Sigma-Aldrich Co. LLC., USA) and l-glutamic acid (G1251, Sigma-Aldrich) at dose ranges of 0.01-3 μmol/L and 0.03-10 μmol/L, respectively, with internal standard 2-MES (10 μM) were also applied to the LC/MS system. The concentrations of GABA and glutamate in the sample solutions were determined by the peak heights of the chromatogram. The GABA and glutamate contents per the corresponding tissue weights were calculated.

Statistical analysis
Statistical analyses were conducted using BellCurve for Excel ver. 3.20 (Social Survey Research Information Co., Ltd., Tokyo, Japan). Significant differences among the multiple groups were analyzed by the Bonferroni multiple comparison test after one-way analysis of variance (ANOVA). Significant differences between two groups were analyzed by Student's t-test. The factorial comparisons in some experiments were performed by two-way ANOVA. Data are expressed as the mean with standard error (SE). Statistical significance was defined as a p value less than 0.05.

Generation of Gad1 STOP−tetO knock-in mice
To confirm the elimination of function of the Gad1 gene by inserting the Neo-STOP-tetO cassette, we first generated homozygous Gad1 STOP−tetO/STOP−tetO mice by crossing heterozygous Gad1 STOP−tetO/+ parents and assessed whether the Gad1 STOP−tetO/STOP−tetO mouse showed the Gad1 knockout phenotypes. Gad1 STOP−tetO/STOP−tetO mice, Gad1 STOP−tetO/+ mice and Gad1 +/+ mice were born at the expected Mendelian frequency (Fig. 2a). Then, mouse pups were divided into two groups to determine the survival rates and palate formation. All Gad1 STOP− tetO/STOP−tetO mice died within 1 day after birth, but Gad-1 STOP−tetO/+ mice and Gad1 +/+ mice survived (Fig. 2b). All Gad1 STOP−tetO/+ and Gad1 +/+ mice formed normal palate. However, 57% of Gad1 STOP−tetO/STOP−tetO mice exhibited a cleft palate (Fig. 2c). These phenotypes are the same as those of Gad1 −/− mice previously reported [31]. Western blot analyses demonstrated that the expression of GAD67 protein in the brain was abolished in Gad-1 STOP−tetO/STOP−tetO mice with or without cleft palate (Fig. 2d, e). These observations indicate that the insertion of the Neo-STOP-tetO cassette following the Gad1 promoter eliminates the function of the Gad1 gene in mice.
Next, the expression levels of GAD67 protein in the brains of Gad1 tTA/STOP−tetO mice were examined in the absence and presence of Dox treatment. The Dox treatment was performed on the mice at 8-10 weeks of age. We noticed that some Gad1 tTA/STOP−tetO mice died during the 3 weeks after starting the Dox treatment. Approximately 44% of Gad1 tTA/STOP−tetO mice survived just after 3 weeks of Dox treatment (Fig. 4a). The expression levels of GAD67 protein in the FCX (F(3,8) = 20.563, p < 0.001, one-way ANOVA), HIP (F(3,8) = 189.298, p < 0.001, one-way ANOVA) and CER (F(3,8) = 22.760, p < 0.001, one-way ANOVA) were significantly decreased by treatment with Dox in Gad1 tTA/STOP−tetO mice compared with Gad1 +/+ mice (Fig. 4b-d). Importantly, the expression level of GAD67 protein in the CER of Gad1 tTA/STOP−tetO mice was markedly decreased in the presence of Dox compared with in the absence of Dox (Fig. 4d). By immunofluorescence, GAD67 immunoreactivity in Gad1 +/+ mice was detected widely in the brain, particulary at high levels in the olfactory bulb, globus pallidum, olfactory  tubercle, substantia nigra, superior and inferior colliculi, and deep cerebellar nuclei (Fig. 4e, upper panel). In brains of Gad1 tTA/STOP−tetO mice, the overall immunoreactivity was reduced moderately without Dox treatment and severely with the treatment. Dox treatment to Gad1 +/+ mice did not affect GAD67 immunoreactivity (data not shown). PV is expressed in a major subclass of GAD67positive inhibitory neurons [32].  Representative western blot analysis of the expression of GAD65/67 and β-actin proteins in the frontal cortex (d), the hippocampus (e) and the cerebellum (f) of Gad1 +/+ , Gad1 tTA/STOP−tetO , Gad1 tTA/+ and Gad1 STOP−tetO/+ mice. β-Actin was used as a loading control. Three independent mice of the respective genotypes were examined. The band densities of GAD67 and β-actin were quantified by ImageJ software. GAD67 protein levels were normalized to β-actin protein expression, and the % change was calculated relative to Gad1 +/+ mice.  , the hippocampus (c) and the cerebellum (d) of Gad1 +/+ and Gad1 tTA/STOP−tetO mice treated and not treated with Dox. β-Actin was used as a loading control. Three independent mice in the respective groups were examined. The band densities of GAD67 and β-actin were analyzed by ImageJ software. GAD67 protein levels were normalized to β-actin protein expression, and the % change was calculated relative to Gad1 +/+ mice. The means with SE are demonstrated as columns. *p < 0.05 vs. the value of nontreated Gad1 +/+ mice (Bonferroni test). # p < 0.05 vs. the value of nontreated Gad1 tTA/STOP−tetO mice (Bonferroni test). e Representative immunoreactivities of GAD67 and parvalbumin (PV) proteins in the brain sections of nontreated Gad1 +/+ mice, nontreated Gad1 tTA/STOP−tetO mice and Dox-treated Gad1 tTA/STOP−tetO mice. The white bars in the images indicate 2 mm length in PV immunoreactivity were found between Gad1 +/+ mice and Gad1 tTA/STOP−tetO mice with or without Dox treatment (Fig. 4e, lower panel). We assessed 3 independent mice in the respective groups and observed similar findings. These results suggest that Dox treatment globally suppresses the expression of GAD67 in the brains of Gad1 tTA/STOP−tetO mice.
From these experimental results, GAD67 protein levels were found unexpectedly low in several brain regions in Gad1 tTA/STOP−tetO mice. Therefore, we stopped comparing Dox-treated and Dox-untreated Gad1 tTA/STOP−tetO mice. Instead, in the following experiments, we compared Dox-treated Gad1 tTA/STOP−tetO mice and Gad1 +/+ mice.
We measured the brain contents of GABA and glutamate in Dox-treated Gad1 tTA/STOP−tetO mice and Gad1 +/+ mice. The GABA content was significantly lower in Doxtreated Gad1 tTA/STOP−tetO mice than in Gad1 +/+ mice in the FCX, HIP and CER (Table 1). On the other hand, the glutamate content in the respective brain regions was comparable between these genotypes (Table 1).

Behavioral abnormalities in GAD67 knockdown mice
We compared the behavioral phenotypes of Gad1 tTA/ STOP−tetO mice with those of Gad1 +/+ mice in the presence of Dox treatment. To avoid the effects of sex differences, male mice were only used in the following experiments.
We first investigated the body weights and the performance of motor coordination in Dox-treated Gad1 tTA/ STOP−tetO and Gad1 +/+ mice. No difference was observed in the body weights (t(17) = 1.066, p = 0.301, Student's t-test, Fig. 5a) or the latency to fall from the cylinder in the rotarod test (t(17) = 0.772, p = 0.451, Student's t-test, Fig. 5b) between the two genotypes.
We next conducted the open-field test, which is a wellaccepted behavioral test to evaluate the anxiety-like state of rodents [33]. The total distance, total duration of movement, moving speed, distance per movement, and duration per movement were significantly increased in Dox-treated Gad1 tTA/STOP−tetO mice compared with Dox-treated Gad1 +/+ mice ( Table 2). In contrast, the total number of movement episodes was significantly decreased in Dox-treated Gad1 tTA/STOP−tetO mice compared with Dox-treated Gad1 +/+ mice ( Table 2). These observations indicate that Gad1 tTA/STOP−tetO mice walk longer distances with less frequency. In addition, Doxtreated Gad1 tTA/STOP−tetO mice walked a long time in the wall side and a short time in the center region compared with Dox-treated Gad1 +/+ mice ( Table 2 and Fig. 5c), indicating that Gad1 tTA/STOP−tetO mice exhibited anxietylike behavior in the open-field test.

Table 1 Brain GABA and glutamate contents
Dox-treated Gad1 +/+ (n = 10) and Gad1 tTA/STOP−tetO (n = 9) mice were used to determine the brain contents of GABA and glutamate. Data represent the mean ± SE. % Reduction was calculated by 100 − [(the mean value of Gad1 tTA/STOP−tetO mice)/(the mean value of Gad1 +/+ mice)] × 100. A p value less than 0.05 was considered a significant difference (Student's t-test)

Discussion
We first generated Gad1 STOP−tetO knock-in mice. All homozygous Gad1 STOP−tetO/STOP−tetO mice died on the day of birth, and 57% of Gad1 STOP−tetO/STOP−tetO mice exhibited a cleft palate. The expression of GAD67 protein was lacking in the brains of Gad1 STOP−tetO/STOP−tetO mice with or without the cleft palate. These phenotypes in Gad1 STOP−tetO/STOP−tetO mice are consistent with those in Gad1 −/− mice [15,31]. Therefore, the function of the Gad1 gene was eliminated by the insertion of the Neo-STOP-tetO cassette in the 5′-untranslated region of the Gad1 gene in mice. It has been reported that neonatal death in Gad1 −/− mice is caused by respiratory failure rather than impairment of suckling [15,35]. Therefore, neonatal death in Gad1 STOP−tetO/STOP−tetO mice may also be caused by respiratory failure. We next developed Gad1 tTA/STOP−tetO biallelic knockin mice by crossing Gad1 tTA/+ and Gad1 STOP−tetO/+ parents. Approximately 40% of Gad1 tTA/STOP−tetO mice died on the day of birth, and the number of Gad1 tTA/STOP−tetO mice at P21-P28 in our breeding colony was smaller than the numbers of mice with the other genotypes. None of the Gad1 tTA/STOP−tetO mice demonstrated a cleft palate. Unexpectedly, some adult Gad1 tTA/STOP−tetO mice died by treatment with Dox. Therefore, GAD67 is important for survival not only in the neonatal period but also in adulthood. However, the cause of death in Dox-treated Gad1 tTA/STOP−tetO mice is currently unknown. To resolve this question, pathological examination is required in a future study.

Dox-treated
Adult mice with Gad1 haplodeficiency demonstrated an approximately 40% reduction in GAD67 protein levels in the whole brain compared with Gad1 +/+ mice [27]. Consistently, we observed that heterozygous Gad1 tTA/+ and heterozygous Gad1 STOP−tetO/+ knock-in mice exhibited a 30-50% reduction in GAD67 protein levels in the FCX, HIP and CER compared with Gad1 +/+ mice. In the absence of Dox treatment, the expression level of GAD67 protein in Gad1 tTA/STOP−tetO mice relative to Gad1 +/+ mice was dependent on the brain regions. In the immunoblotting analysis, the expression of GAD67 protein in the CER was comparable between Gad1 tTA/STOP−tetO mice and Gad1 +/+ mice. However, the expression of GAD67 protein in the FCX and HIP was significantly lower in Gad1 tTA/STOP−tetO mice than Gad1 +/+ mice. Importantly, in the presence of Dox treatment, GAD67 expression was reduced by approximately 90% in the brains of Gad1 tTA/ STOP−tetO mice, compared with Dox-treated Gad1 +/+ mice. The brain-wide reduction of GAD67 expression in Dox-treated Gad1 tTA/STOP−tetO mice was also observed in the immunofluorescence analysis. These findings suggest that GAD67 expression is suppressed by treatment with Dox in the brains of Gad1 tTA/STOP−tetO mice.
In adult mice with Gad1 haplodeficiency, the GABA content in the brain was reduced by 7-20% from those in wild-type control mice [15,32]. In this study, we found that the GABA contents in the FCX, HIP and CER of Dox-treated Gad1 tTA/STOP−tetO mice were reduced by 27.0-56.8% from those in Dox-treated Gad1 +/+ mice. This result is inferred that the amount corresponding to the reduced GABA content compared to the GABA content in the brain of Gad1 +/+ mice is derived from the amount biosynthesized by GAD67 expression in the brain of Gad1 tTA/STOP−tetO mice. Therefore, the GABA reduction in the brains of Dox-treated Gad1 tTA/STOP−tetO mice was larger than that in Gad1 haplodeficient mice. Because approximately half of the brain GABA is produced by GAD65 in adulthood [36], the remaining GABA in the brain of Dox-treated Gad1 tTA/STOP−tetO mice is mainly synthesized by GAD65. The brain glutamate contents in Dox-treated Gad1 tTA/STOP−tetO mice were comparable to those in Dox-treated Gad1 +/+ mice. Therefore, the glutamatergic system may be normal in Dox-treated Gad1 tTA/STOP−tetO mice.
GAD67 haplodeficient mice demonstrated several abnormal behaviors, such as hyper-locomotor activity, reduced interactions with an unfamiliar mouse, and aggressive behavior. However, the emotional behaviors in Gad1 haplodeficient mice were normal in the open-field test, the light-dark avoidance test and the elevated plusmaze test [21,23]. In the current open-field test, Doxtreated Gad1 tTA/STOP−tetO mice walked longer distances than Dox-treated Gad1 +/+ mice. In addition, Dox-treated Gad1 tTA/STOP−tetO mice preferentially walked for more time along the walls and for less time in the center region. These observations indicate that Dox-treated Gad1 tTA/ STOP−tetO mice exhibited behavioral abnormalities, including the hyper-locomotor activity and anxiety-like  [33]. Therefore, the reduction in brain GABA in Dox-treated Gad1 tTA/STOP−tetO mice may cause the induction of anxiety-like behavior. The mice lacking GAD67 in protein phosphatase 1 regulatory subunit 2 (Ppp1r2)-expressing cells, in which Cre recombinase expression is largely confined to GABA interneurons of the cerebral cortex and the hippocampus, demonstrated the hyper-locomotor activity and anxiety-like behavior in the open-field test [37]. In addition, we recently reported that mice lacking GAD67 in somatostatin-expressing GABA interneurons demonstrated anxiety-like behavior in the open-field test without affecting locomotor activity [20]. Therefore, the anxiety-like behavior in Dox-treated Gad1 tTA/STOP−tetO mice may be due to GAD67 knockdown from somatostatin-expressing GABA interneurons in cortical and hippocampal areas. In addition, the hyperlocomotor activity in Dox-treated Gad1 tTA/STOP−tetO mice may be associated with GAD67 knockdown from the other subtypes of GABA interneurons. In this study, Dox-treated Gad1 tTA/STOP−tetO mice showed a normal response to PPI. Therefore, global knockdown of GAD67 does not affect the PPI response. However, we previously reported that mice with conditional Gad1 heterozygous knockout predominantly in parvalbumin-positive cells demonstrated an impaired response to PPI [19]. It is possible that a mild reduction in GAD67 in parvalbumin-positive cells might be required for impairing the PPI response. Interestingly, the startle responses elicited by large acoustic stimuli were impaired in Dox-treated Gad1 tTA/STOP−tetO mice. The acoustic startle response is a simple reflex of animals that results in a whole body motor response elicited by hearing a loud sound. Multiple neural circuits from the cochlear nucleus to motoneurons contributes to elicit the acoustic startle reflex [38]. Therefore, it is possible that Dox-treated Gad1 tTA/STOP−tetO mice might exhibit the impairment of the acoustic startle reflex associated with those neural circuits. In addition, GABAergic interneurons expressing GAD67 proteins constitute auditory neural networks and contribute to auditory function [39][40][41]. Therefore, global knockdown of GAD67 in the brain might induce the impairment of auditory function. To note that the behavioral responses by intensity of pre-pulse in the control mice were different between the present study and the previous study [19]. There are several possible reasons to explain the different observations. First, it may be related to the difference in the environment during development between these two types of mice because all mice in the present study were born from GAD67 haplodeficient (Gad1 STOP−tetO/+ ) mothers and cared by the same mothers. We previously reported that wildtype (Gad1 +/+ ) mice born from GAD67 haplodeficient (Gad1 GFP/+ ) mothers exhibited the vulnerability to stress at the adult age [42]. In addition, it has been reported that the anxious state in mother mice affects behavioral phenotypes in their pups at the adult age [43]. Second, we used wild-type (Gad1 +/+ ) mice as the control, but Fujihara et al. [19] used Gad1 flox/+ mice as the control. The difference in genotypes might affect the reaction of PPI in mice. Third, there is difference whether the mice were subjected to Dox treatment or not. It can not be excluded the possibility that the Dox treatment affected the PPI reaction in mice. Further experiments are necessary to resolve the difference in the baseline reaction of prepulse between the studies.
The GAD1 gene and GAD67 protein have often been targeted in human studies to elucidate an association with the pathophysiology of psychiatric disorders [44][45][46]. Reduced full-length GAD1 transcript and GAD67 protein is a consistent finding in the postmortem brains of patients with several psychiatric disorders including schizophrenia, bipolar disorder and major depressive disorder [7][8][9][10][11][12]. Because GAD67 reduction was predominantly observed in parvalbumin-positive GABAergic interneurons in the postmortem brains of schizophrenic patients [47][48][49], mice with conditional knockout of GAD67 in parvalbumin-positive cells have been used as an animal model of schizophrenia [18,19,50,51]. On the other hand, the specific subsets of GABAergic interneurons reducing GAD67 expression have not yet been identified in other psychiatric disorders, which might indicate that GAD67 expression is globally reduced in the brain of subjects with the other psychiatric disorders. In this study, global knockdown of GAD67 elicited anxiety-like behavior in mice. Therefore, we suggest the possibility that the global reduction in GAD1 transcript and GAD67 protein in the brain might be related to the occurrence of anxiety symptoms frequently comorbid in several psychiatric disorders.
We did not examine whether administration of anxiolytic, antidepressant or antipsychotic drugs improves anxiety-like behavior in Dox-treated Gad1 tTA/STOP−tetO mice. Because of the high lethality rate, we could not immediately prepare a sufficient number of animals for conducting a pharmacological study. Therefore, we will determine the utility of Dox-treated Gad1 tTA/STOP−tetO mice as a tool for screening potential medications for anxiety symptoms in a future study.
In summary, Gad1 tTA/STOP−tetO biallelic knock-in mice showed GAD67-knockdown phenotypes when treated with Dox. We suggest that the global reduction in GAD67 elicits emotional and auditory abnormalities in mice. The use of GAD67 knockdown mice will provide new insights into the neurobiological impact of GAD67 dysfunction and elucidate the neurobiological mechanisms of emotional abnormalities associated with psychiatric disorders.