GABAB receptor signaling in the caudate putamen is involved in binge-like consumption during a high fat diet in mice

Previous studies suggest that signaling by the gamma-aminobutyric acid (GABA) type B receptor (GABABR) is involved in the regulation of binge eating, a disorder which might contribute to the development of obesity. Here, we show that intermittent access to a high fat diet (HFD) induced binge-like eating behavior with activation of dopamine receptor d1 (drd1)-expressing neurons in the caudate putamen (CPu) and nucleus accumbens (NAc) in wild-type (WT) mice. The activation of drd1-expressing neurons during binge-like eating was substantially increased in the CPu, but not in the NAc, in corticostriatal neuron-specific GABABR-deficient knockout (KO) mice compared to WT mice. Treatment with the GABABR agonist, baclofen, suppressed binge-like eating behavior in WT mice, but not in KO mice, as reported previously. Baclofen also suppressed the activation of drd1-expressing neurons in the CPu, but not in the NAc, during binge-like eating in WT mice. Thus, our data suggest that GABABR signaling in CPu neurons expressing drd1 suppresses binge-like consumption during a HFD in mice.

www.nature.com/scientificreports/ eating in rodents [16][17][18] as well as binge eating in humans [19][20][21] . Our recent study has shown that GABA B R signaling in the mesolimbic system is involved in the suppression of the binge-like consumption of a HFD in mice 22 , although the precise sites of action in the system still remain unclear.
In the present study, we employed corticostriatal neuron-specific GABA B R-deficient mice, and subjected them to intermittent access to a HFD in order to clarify critical sites of GABA B R signaling in the suppression of binge-like consumption with a HFD.

Results
Intermittent access to a HFD induces binge-like eating with neural activations in the striatum. The experimental protocol used is described in Supplemental Fig. 1A. As shown in Supplemental   Fig. 1B, BW on the experimental day did not differ between control and binge groups. On the experimental day, mice in the binge group only consumed a HFD over 2 h, with consumed calories higher than in the control group (Supplemental Fig. 1C). We evaluated cFos expression at 30 and 120 min in the limbic system in a bingelike eating model based on previous studies 10, 23 . The mRNA expression level of cFos in the CPu was significantly increased 30 and 120 min after the consumption of a HFD compared to values at time 0 in the binge group (Fig. 1A). While the cFos mRNA expression level in the CPu was increased at 30 min in the control group, absolute values were lower in the control group compared to the binge group (Fig. 1A). The cFos mRNA expression levels in the CPu were proportional to HFD consumption at 30 min in the binge group (Supplemental Fig. 1D). The mRNA expression level of ΔFosB in the CPu was increased significantly 120 min after the consumption of a HFD in the binge, but not control group (Fig. 1B). The phosphorylation of dopamine-and cyclic-AMP-regulated phosphoprotein of molecular weight 32,000 (DARPP32), a prominent mediator of dopamine signaling in the striatum 24 , in the CPu was significantly increased at 30 min in the binge group (Fig. 1C), but not in the control group (Fig. 1D). The mRNA expression levels of both cFos and ΔFosB in the NAc were significantly increased by the consumption of a HFD at 30 and 120 min in the binge, but not control group ( Fig. 1E and F). The phosphorylation of DARPP32 was also significantly increased at 30 min in the binge (Fig. 1G), but not control group (Fig. 1H). The cFos mRNA expression levels in the PFC were significantly increased at 30 min compared to values at 0 min in both binge and control groups (Supplemental Fig. 1E). The ΔFosB mRNA expression levels in the PFC were significantly increased at 120 min compared to values at 0 min in both binge and control groups  www.nature.com/scientificreports/ (Supplemental Fig. 1F). No significant differences in cFos and ΔFosB mRNA expression levels at all measurement times were found between binge and control groups (Supplemental Fig. 1E and F). The consumption of a HFD for 30 min did not affect dopamine receptor d1 (drd1) expression in CPu and NAc of mice in the binge group (Supplemental Fig. 1G and H).
GABA B R deficiency in corticostriatal neurons increases binge-like eating with activation of drd1-expressing neurons in the CPu. Consumption of a HFD for 30 and 120 min was significantly increased in male knockout (KO) mice compared to WT mice in the binge group (Supplemental Fig. 2A), as reported previously 22 . The mRNA expression levels of both cFos and ΔFosB in the CPu after the consumption of a HFD were significantly higher at 30 min in corticostriatal neuron-specific GABA B R-deficient KO mice compared to wild-type (WT) mice in the binge group ( Fig. 2A and B). The phosphorylation of DARPP32 in the CPu was also significantly increased in KO mice compared to WT mice 30 min after the consumption of a HFD (Fig. 2C). In comparison, significant differences were not noted in the mRNA expression levels of cFos and ΔFosB, and the phosphorylation of DARPP32 in the NAc between genotypes 30 min after the consumption of a HFD ( Fig. 2D-F). There were no significant differences in drd1 expression levels in the CPu and NAc between KO and WT mice 30 min after the consumption of a HFD in the binge group (Supplemental Fig. 1B and C). Immunohistochemistry in the striatum showed that while cFos-positive cells in drd1-positive neurons (cFos + /drd1 +) in the CPu were significantly increased 30 min after the consumption of a HFD in both WT and KO mice in the binge group, the numbers were significantly greater in KO than in WT mice ( Fig. 3A and B). In contrast, no significant differences were observed in the number of cFos-positive cells in drd1-positive neurons (cFos + / drd1 +) in the NAc between genotypes ( Fig. 3C and D). Significant differences in the number of cFos-positive cells in drd1-negative neurons (cFos + / drd1−) in the CPu and NAc were not found between genotypes (Supplemental Fig. 3A and B).
Baclofen suppresses binge-like eating and neural activation in the CPu. As reported previously 22 , treatment with baclofen, a GABA B R agonist, significantly reduced HFD consumption compared to vehicle in WT mice, but not in KO mice (Supplemental Fig. 4A). Baclofen suppressed elevation of cFos and ΔFosB mRNA expression levels in the CPu compared to vehicle at 30 and 120 min after the consumption of a HFD ( Fig. 4A and B). The phosphorylation of DARPP32 was also significantly suppressed by baclofen compared to vehicle (Fig. 4C). In contrast, baclofen did not affect the mRNA expression levels of cFos or ΔFosB, or the phosphorylation of DARPP32 in the NAc, after the consumption of a HFD (Fig. 4D-F). Immunohistochemical staining in the striatum showed that baclofen significantly reduced the number of cFos-positive cells in drd1-positive neurons (cFos + /drd1 +) in the CPu 30 min after the consumption of a HFD in WT, but not in KO mice ( Fig. 6A and B). Significant differences were not found in the number of cFos-positive cells in drd1-positive neurons (cFos + /drd1 +) in the NAc between vehicle and baclofen groups in both WT and KO mice ( Fig. 6C and D). Also, no significant differences were evident in the numbers of cFos-positive cells in  Table 2  www.nature.com/scientificreports/ drd1-negative neurons (cFos + /drd1−) in the CPu and NAc between vehicle and baclofen groups in both WT and KO mice (Supplemental Fig. 5A and B).

Discussion
In the present study, we showed that drd1-expressing neurons in the CPu and NAc were significantly activated in WT mice after binge-like consumption on a HFD. Our data also showed that activation of drd1expressing neurons was significantly enhanced in the CPu, but not in the NAc, in corticostriatal neuron-specific GABA B R-deficient mice compared to WT mice after binge-like consumption on a HFD. Furthermore, the GABA B R agonist, baclofen, suppressed the activation of neurons expressing drd1 in the CPu and decreased binge-like eating by WT mice on a HFD, but not that of KO mice. These data suggest GABA B R signaling in the CPu is crucial for the suppression of binge-like eating while on a HFD. Previous studies using binge-like eating models in rats showed that the number of c-Fos-positive cells in the CPu, as well as ΔFosB expression and DARPP32 phosphorylation in the NAc, were increased after the intake of palatable food 9,25,26 . These data are consistent with the findings in the current study that neurons in the CPu and NAc were activated in WT mice with intermittent access to a HFD. Previous studies also showed that the dorsal medial PFC is associated with binge-like eating 23 , whereas the ventral medial PFC is associated with resistance   www.nature.com/scientificreports/ to binge-like eating 27 . While our data did not reveal any difference in dopamine receptor signaling in the PFC between binge and control groups, our analyses of whole PFC do not exclude the role of PFC in binge-like eating. We previously reported that GABA B R signaling in the mesolimbic system is involved in the suppression of binge-like consumption under a HFD 22 ; this was confirmed in the current study showing that binge-like eating during a HFD was enhanced in corticostriatal neuron-specific GABA B R-deficient mice. In addition, we clearly demonstrated that the suppression of binge-like eating during a HFD by baclofen was accompanied by decreased activity of drd1-expressing neurons in the CPu, but not in the NAc. The site of activation in the striatum, reportedly, depended on the type of food or drink ingested: ethanol (EtOH) infusion, but not fat intake, significantly increased NAc dopamine signaling, while fat intake, but not EtOH infusion, significantly increased CPu dopamine signaling 28 . It was also shown that taste quality regulated dopamine release in the NAc, while dopamine in the CPu was released when a nutritive solution, but not a non-nutritive solution, was ingested 29 . Taken together with our findings, it is suggested that the CPu plays an important role in regulating the binge-like consumption of calorie-enrich palatable food such as when on a HFD, where GABA B R signaling suppresses the activity of drd1-expressing neurons.
Our data showed that the consumption of a HFD did not affect drd1 expression levels in the CPu, and that these levels did not differ between WT and KO mice. Taken together with the findings that a GABA B R agonist suppressed mRNA expression levels of cFos and ΔFosB as well as the phosphorylation of DARPP32 in the CPu in WT mice, and that they were significantly increased in KO mice compared to WT mice, it is suggested that GABA B R signaling suppressed signaling downstream of drd1 in the CPu (Fig. 7).
The first limitation in the present study is that we cannot rule out the possibility that GABA B R signaling in the medial PFC and orbitofrontal cortex (OFC), which are important sites for decision making and related to reward-based behavior 30,31 , might be affected in KO mice, since GPR88-positive neurons include not only striatal neurons but also those in the medial PFC and OFC 22 . Second, we did not measure concentrations of dopamine in the striatum but used surrogate markers for dopamine signaling. Third, the systemic injection of baclofen   Table 2  www.nature.com/scientificreports/ was employed in the current study. To clarify the site of action, a local infusion of baclofen would be warranted in future studies. Fourth, while our study focused on dopamine receptor d1 signaling, this does not exclude the possibility that dopamine receptor d2 signaling in the CPu and NAc might also be involved in binge-like consumption that occurs during a HFD.
In conclusion, our data suggest that GABA B R signaling in CPu neurons expressing drd1 suppresses binge-like consumption during a HFD in mice.

Methods
Mice. All experiments were performed using 9-week-old male mice. All animal procedures were approved by the Animal Care and Use Committee of Nagoya University Graduate School of Medicine and performed in accordance with National Institutes of Health animal care guidelines. The study was carried out in compliance  Mice with GPR88-sepecific deletion of GABA B1 receptor. We crossed GABA B1 R lox511 /lox511 mice with GPR88-Cre heterozygous mice to generate GABA B1 R lox511/lox511 GPR88-Cre mice (KO mice), GABAB +/lox511 GPR88-Cre mice, and GABA B1 R lox511/lox511 littermate controls (WT mice). GABA B1 R lox511/lox511 mice have been generated previously 32  Training and assessment of binge-like eating behavior. We used a published protocol 22 to train and assess binge-like eating behavior in male mice (Supplemental Fig. 1A). After acclimation, WT mice (9 weeks of age) were exposed to both a chow diet (CD; CE-2, CLEA Japan, Tokyo, Japan; 24.9% protein, 4.6% fat, and 70.5% carbohydrate) and a HFD (Test Diet 58Y1; PMI Nutrition International, KS, USA; 18.3% protein, 60.9% fat, and 20.1% carbohydrate) for 2 days, and then exposed to only CD for 5 days. Next, the mice were divided into two groups: ''Binge group'' and ''Control group'' , on the experimental day. The mice in the binge group were given free access to HFD and CD for 120 min (zeitgeber time [ZT] 12 to 14: 21:00 to 23:00), while the mice in the control group were continuously fed only CD. Then HFD and CD intakes were measured for 120 min (ZT 12 to 14). The food intake of mice on both the CD and HFD were assessed by multifeeders (Shinfactory, Fukuoka, Japan). Mice in the binge group were sacrificed at 30 and 120 min after the start of access to HFD and CD (i.e., 21:30 and 23:00, respectively) and those in the control group were also sacrificed at 21:30 and 23:00. The GABA B R agonist, baclofen (Sigma-Aldrich, St Louis, MO, USA), dissolved in 0.9% saline, or vehicle (saline) were administered intraperitoneally 30 min before the beginning of the dark cycle (ZT 12) on the experimental day. The dose of baclofen was 3 μg/10 μL/g body weight.
Extraction of brain tissues. After mice were sacrificed, the CPu, NAc, and PFC were rapidly dissected www.nature.com/scientificreports/ Determination of mRNA levels by qRT-PCR. Total RNA was extracted from samples using TRIzol (Invitrogen, Carlsbad, CA, USA) and an RNeasy kit (QIAGEN, Hilden, Germany). Copy DNA was synthesized from 60 to 150 ng total RNA using a ReverTra Ace qPCR RT Kit (TOYOBO, Osaka, Japan). Quantitative reverse transcriptase (qRT)-PCR reactions were carried out using Brilliant III Ultra-Fast SYBR Green QPCR Master Mix (Agilent Technologies, Santa Clara, CA, USA), and samples were processed using a CFX Connect Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA). The relative mRNA levels of cFos and ΔFosB were assessed by qRT-PCR using glyceraldehyde 3-phosphate dehydrogenase (Gapdh) as an internal control. The qRT-PCR reactions were carried out, and relative mRNA expression levels were calculated using a comparative Ct method as described previously 34,35 . The sequences of primers are described in Supplemental Table 1.
Immunohistochemistry. Free-floating sections were blocked with Triton X-100 (0.3%; 30 min) and bovine serum albumin solution (5%; 1 h) at room temperature. They were then incubated in diluted blocking buffer (Blocking one, NACALAI TESQUE, Japan) and 0.2% sodium azide for 5 days at 4 °C with the following antibodies: rabbit anti-cFos (226 003, Synaptic Systems, Goettingen, Germany, RRID: AB_2231974), and mouse anti-drd1 (sc-33660, Santa Cruz Biotechnology, Dallas, TX, USA, RRID: AB_668813). The sections were then incubated with Tween-20 (0.05%; 40 min) followed by Alexa Fluor 488-conjugated anti-rabbit IgG secondary antibody (Invitrogen; RRID: AB_221544), Alexa Fluor 594 conjugated anti-mouse IgG secondary antibody (Invitrogen; RRID: AB_2650601), and DAPI (D523; DOJINDO, Kumamoto, Japan) for 2 h at room temperature. After washing in Tween-20 (0.05%), sections were placed on slides, air dried, and Vectashield (Vector Labs, Peterborough, UK) used to fix coverslips in place. All fluorescently stained sections were examined with either a confocal laser microscope (TiEA1R; NIKON INSTECH, Tokyo, Japan) or a fluorescence microscope (BZ-9000; Keyence, Osaka, Japan; RRID:SCR_015486), and viewed using NIS-Elements software (NIKON INSTECH; RRID:SCR_014329). Cells labeled for cFos colocalization, with or without drd1, were counted bilaterally in a blinded fashion. cFos was merged with nuclear marker (DAPI), while drd1 was expressed in the cytoplasm. For analysis, samples from 4-10 mice from each group were used for staining and the mean values from two or three serial sections from each mouse were calculated. The sections included the CPu and NAc, which were located 1.10-1.42 mm from the bregma based on coordinates in a mouse brain atlas. The anatomical boundaries of each brain region were also identified using the mouse brain atlas 37 .
Statistical analysis. The statistical significance of differences between groups was analyzed by either unpaired t-test, two-way factorial analysis of variance (ANOVA) or two-way ANOVA assessed by repeated measures followed by Sidak's multiple comparison by using SPSS Statistics 26 (IBM, Endicott, NY, USA; RRID:SCR_002865). Results are expressed as mean ± standard error of the mean (SEM), and differences were considered significant at p < 0.05.

Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.