FFAR1−/− mice were generated by homologous recombination (Aizawa et al., 2016; Nakamoto et al., 2017). Briefly, exon 1 of the Ffar1 gene was replaced with a PKG-neo cassette. Frozen Ffar1 −/− fertilized oocytes were inoculated into pseudopregnant foster mother (ICR) at Center for Animal Resources and Development, Kumamoto University (CARD ID 1882) and bred at the Institute of Laboratory Animal Science Research Support Center, Kagoshima University. FFAR1−/− mice were backcrossed to C57BL/6J background for >10 generations and maintained by crossbreeding of heterozygous mice. Only male mice were used in this study. To confirm ablation of Ffar1, we performed PCR on genomic DNA from tail or ear auricle using the following primers: 40-S1: 5′ TAG GAC TGG CTT CTG GTG CT 3′, 40-AS2: 5′ CCT CCT GAG TTG TGG TGG AT 3′, 3 primer-neo: 5′ GGC TAT TCG GCT ATG ACT GG 3’. This primer pair yielded a 1,166-bp DNA fragment for +/+ mice and a 1,300-bp fragment for −/− mice. Primers were obtained from Integrated DNA Technologies, KK (Tokyo, Japan).

We also employed the quantitative real-time PCR method (see below) to evaluate the presence or absence of FFAR1 in the striatum with +/+ mouse pancreas tissue samples as a positive control. As depicted in Supplementary information (Supplementary Figure S1), the transcript level of FFAR1 was high in the pancreas, but significant expression was also detected in the striatum, where, as expected, the expression was undetectable in FFAR1−/− mice. These results are consistent with our previous immunoblot analysis in the mouse CNS (Nakamoto et al., 2012) and qPCR analysis in the human brain by Briscoe et al. (2003).

Mice were housed under controlled temperature (24 ± 1°C) and humidity (55 ± 10%) with a 12 h light-dark cycle with food and water freely available. All mice used aged 8–22 weeks at the start of each experiment. The animal experiments were approved by the Animal Care Committees of Kagoshima University (approval no. MD16077) and were conducted in accordance with the ethical guidelines for the study of experimental pain in conscious animals of the International Association of the Study of Pain.

The microdialysis experiments were carried out on awake FFAR1+/+ and −/− mice following the protocol described elsewhere (Han et al., 2002). Briefly, mice were anesthetized with pentobarbital sodium (50 mg/kg, i.p.). Microdialysis probes (A-I-(-3)-02 probe, 0.22 o.d., 2.0 mm active membrane length, molecular weight cut-off 50,000 Da, EICOM, Kyoto, Japan) were implanted vertically into the dorsal striatum (anterior, 0.0; lateral, 1.8 mm; and vertical, 4 mm from the bregma) (Carboni et al., 2001; Han et al., 2002). After the surgery, animals were allowed a recovery period of 24 h. On the test day, the probe was connected to a syringe pump (EICOM, Kyoto, Japan) and continuously perfused with Ringer’s solution (in mM: 147 NaCl, 4 KCl, and 2.3 CaCl₂) at a rate of 2 μl/min. Samples were collected for 50–75 min to determine the basal levels of dopamine (DA), serotonin (5-hydroxytryptamine; 5-HT), and their metabolites, dihydroxyphenylacetic acid (DOPAC), 3-methoxytyramine (3-MT), homovanillic acid (HVA), and 5-hydroxyindoleacetic acid (5-HIAA). Then, GW9508 (a FFAR1 agonist, MW 347.4; 100 μM; Cayman chemical company, Ann Arbor, MI, United States), GW1100 (a FFAR1 antagonist, MW 520.6; 100 μM; Cayman) or cocaine (20 mg/kg; Dainippon Pharmaceutical Co., Ltd, Osaka, Japan) was administered. In the case of examining the effects of the FFAR1 ligands, GW9508 or GW1100 was added to Ringer’s solution perfusing the microdialysis probe for 50 min and samples were collected for another 200 min. While GW9508 was applied to both FFAR1+/+ and −/− mice, GW1100 was only given to +/+ mice. Cocaine was intraperitoneally (i.p.) injected into both +/+ and −/− mice and samples were collected for another 175–200 min. Stock solutions of these FFAR1 ligands (10 mM each) were prepared by dissolving in ≧ 99.9% dimethyl sulfoxide (DMSO; Sigma-Aldrich, St. Louis, MO, United States) and diluted with Ringer’s solution. The concentrations of the FFAR1 ligands were selected based on our preliminary experiments. The stock solution of cocaine (100 mg/ml) was prepared by dissolving in MilliQ water and diluted with saline. The dose of cocaine was chosen based on previous reports (Reith et al., 1997; Han et al., 2002). All samples were automatically injected into the HTEC-500 microdialysis analysis system (Eicom) every 25 min using a fully automatic online system. Concentrations of DA and 5-HT and their metabolites in the microdialysis samples were measured by HPLC with electrochemical detection. The potential of the glassy carbon working electrode (WE-3G, EICOM) was +700 mV vs. the Ag/AgCl reference electrode. The separation was achieved on a 150 × 2.1 (i.d.) mm Eicompak SC-50DS column (EICOM). The mobile phase was a mixture of methanol and 0.1 M citrate–0.1 M sodium acetate buffer (pH 3.9) (17:83, v/v) containing sodium 1-octanesulfonic acid (140 mg/L) and EDTA-2Na (5 mg/L). The flow rate was 0.23 ml/min. Chromatographic data were acquired and processed by the use of an EPC-500 software (Power Chrom, EICOM). The detection limit for DA and 5-HT was estimated to ∼0.05 pg in 50 μl injected onto the column. At the end of the microdialysis experiments, mice were sacrificed, and their brains were removed and correct probe locations were confirmed by visual inspection.

The open-field test was made of polyvinyl chloride plates and was 50 cm×50 cm×40 cm in size. Each FFAR1+/+ and −/− mouse was transferred to the center of the field, and its locomotor activity was measured for 5 min using a color tracking system (CompACT VAS; Muromachi Kikai, Tokyo, Japan) as described previously (Saegusa et al., 2000).

On the test day, FFAR1+/+ and −/− mice were habituated to the open-field test chamber for 63 min before cocaine (20 mg/kg) injection. The spontaneous locomotion was recorded for 3 min in every 10 min. After i.p. injection of cocaine, locomotor activity was recorded for another 33 min.

In the case of examining the effects of the FFAR1 antagonist (GW1100), vehicle or GW1100 (10 mg/kg) was i.p.-injected into FFAR1+/+ mice after the 63 min habituation, and locomotor activity was similarly measured for 33 min. Then, cocaine (20 mg/kg) was given and activity was monitored for a further 33 min. GW1100 was initially dissolved in ≧ 99.9% DMSO (5 mg/ml) and then diluted with saline (1 mg/ml, final DMSO concentration ∼20%). Thus, ∼ 20% DMSO in saline was employed as vehicle control.

Total RNA was isolated from the striatal tissue including the area in vivo microdialysis experiments performed by using Sepasol RNA I Super G reagent (Nacalai Tesque, Kyoto, Japan) and treated with DNase (Promega, Madison, WI) to avoid DNA contamination. cDNA synthesis was carried out using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) according to manufacturer’s instructions. Quantitative PCR was performed with Thunderbird SYBR qPCR Mix (Toyobo Life Science, Osaka, Japan) using a Thermal Cycler Dice (Real Time System TP800, Takara Bio Inc., Shiga, Japan). Amplification of target cDNA was normalized to GAPDH expression. Relative levels of target mRNA expression were calculated using the standard curve method. At least, two independent qPCR experiments of two individual striatum samples were performed. The sequences of primer pairs are described below. Primers were obtained from Integrated DNA Technologies, KK:

FFAR1 (Ffar1): 5′ GGG CTT TCC ATT GAA CTT GTT AG 3′ (forward), 5′ GCC CAG ATG GAG AGT GTA GAC C 3′ (reverse).

DA transporter (DAT: slc6a3): 5′ ACC ACA CCC GCT GCT GAG TAT T 3′ (forward), 5′ GGT CAT CAA TGC CAC GAC TCT G 3′ (reverse).

5-HT transporter (5HTT: slc6a4): 5′ AGG AAC GAA GAC GTG TCC GA 3′ (forward), 5′ CCA AAC CCA GCG TGA TTA ACA T 3′ (reverse).

D1 receptor (Drd1): 5′ CGG CCT TAT CGG TCA TAT TGG 3′ (forward), 5′ CTG TGG GTA ACG GGT TGG A 3′ (reverse).

D2 receptor (Drd2): 5′ CAG CTC CAA GCG CCG AGT T 3′ (forward), 5′ GGC AGG GTT GGC AAT GAT ACA C 3′ (reverse).

GAPDH: 5′ GAA GGT CGG TGT GAA CGG AT 3′ (forward), 5′ CTC GCT CCT GGA AGA TGG TG 3′ (reverse).

Western blot analysis was conducted as previously described (Karki et al., 2015; Ohnou et al., 2016). Mice were deeply anesthetized with sodium pentobarbital (60 mg/kg, i.p.), and dorsal striatal tissues were quickly dissected out. The tissue was homogenized in a lysis buffer [150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, and 50 mM Tris-HCl, pH 8.0] with a mixture of protease and phosphatase inhibitors (Roche Diagnostics, Mannheim, Germany). Protein concentrations were determined with a Bio-Rad protein assay kit (Bio-Rad, Hercules, CA). Proteins (5–10 μg) were separated by SDS-PAGE (12.5% gel) and then transferred to a polyvinylidene difluoride membrane (Millipore). The following antibodies were used: anti-DA transporter monoclonal antibody (rat, 1:200, Millipore MAB369), anti-5-HT transporter polyclonal antibody (rabbit, 1:2,000, Alomone Labs AMT-004), anti-D1 DA receptor monoclonal antibody (rat, 1:100, Sigma-Aldrich D2994), and anti-D2 DA receptor polyclonal antibody (rabbit, 1:2,000, Millipore AB5084P). Immunoreactivity was detected by using an ECL Prime kit (GE Healthcare, Buckinghamshire, United Kingdom). We observed that the anti-DAT and 5HTT antibodies could exhibit single bands with expected molecular weight (∼80 and ∼60 kDa, respectively) in our experimental condition. For the anti-D1 and D2 receptor antibodies, the specificities were validated with D1 and D2 receptor deficient mice (Stojanovic et al., 2017). An anti-β-actin antibody (mouse monoclonal, 1:2,000; Santa Cruz Biotechnology, Inc., Dallas, TX) was used to normalize protein loading. Relative intensities of the bands were quantified by using an image analysis system with ImageJ software, version 1.46 (National Institutes of Health, Bethesda, MD). At least two independent immunoblot experiments of two individual striatum samples were analyzed.

Experimental data are expressed as mean ± SEM. Single comparisons were made using Student’s two-tailed unpaired t-test corrected for unequal variance (Welch’s correction) when appropriate. All remaining data were compared using one-way analysis of variance (ANOVA) with repeated measures followed by Dunnett’s post hoc test or two-way repeated measures ANOVA, with genotype or drug treatment as the between-subject factor and time as the within-subject factor, followed when necessary, by Tukey-Kramer post hoc test. p < 0.05 was considered statistically significant.

To investigate whether FFAR1 indeed regulates monoamine levels, we employed in vivo microdialysis method in this study. At first, we examined baselines levels of extracellular DA, 5-HT and their metabolites within the striatum of awake normal FFAR1+/+ and −/− mice. We found that baseline DA level in the striatum was significantly higher in −/− mice compared with +/+ mice, while a trend for decreased 5-HT baseline levels was observed in −/− mice (Figures 1A,B). The baseline levels of all DA metabolites, DOPAC, 3-MT, and HVA, also showed significant increases in −/− mice, although the level of 5-HT metabolite, 5-HIAA, was not changed (Figures 1C–F).

The in vivo microdialysis experiments suggested that FFAR1 in fact directly/indirectly regulated at least striatal DA and 5-HT levels. To evaluate how FFAR1 control striatal DA and 5-HT levels, further in vivo microdialysis experiments were performed by in situ application of a FFAR1 agonist, GW9508, and an antagonist, GW1100 (Briscoe et al., 2006; Stoddart et al., 2007) (Figure 2). Although we observed a slight transient reduction of DA level in FFAR1−/− mice after local application of GW9508 (100 μM in the perfusing Ringer’s solution) (Figure 2A), GW9508 perfusion resulted in different changes in 5-HT release between FFAR1 +/+ and −/− mice (Figure 2B). In +/+ mice, GW9508 significantly increased the 5-HT level in the striatum. Intriguingly, the stimulated release of 5-HT was almost completely lost in −/− mice. In contrast to what was found in the GW9508 application, local treatment of GW1100 significantly decreased the 5-HT level in +/+ mice (Figure 2D). Despite the downregulation of 5-HT release by GW1100 being remarkable, DA level was not significantly affected by the FFAR1 antagonist (Figure 2C).

Thus, to further explore the functional significance of striatal FFAR1, we employed a cocaine-induced locomotor enhancement model in this study and examined the behavioral phenotype of FFAR1−/− mice. Moreover, we also tested the effects of GW1100 on cocaine-evoked locomotor enhancement. First, we checked baseline locomotor activity under a novel condition (Figure 3). Naïve FFAR1+/+ and −/− mice were placed in a testing chamber (open-field box) and locomotor activity was recorded for 5 min. In accord with our previous report (Aizawa et al., 2016), FFAR1−/− mice spent more time in the center region of the field when compared with +/+ mice (Figure 3D and Supplementary Figure S2), although there was no significant difference between +/+ and −/− mice in the distance traveled (Figure 3B). In this study, we also analyzed additional parameters of the locomotor activity, the locomotion time, and speed and found that locomotion speed was slightly but significantly slower in −/− mice (Figures 3A,C).

Then, we next studied the functional role of FFAR1 by examining locomotor activity in response to acute cocaine injections in both FFAR1+/+ and −/− mice. Locomotor stimulation is one of the most characteristic effects of psychostimulants including cocaine (Kelly and Iversen, 1976; Joyce et al., 1983; Reith et al., 1997; Cornish and Kalivas, 2001; Han and Gu, 2006; Lüscher and Ungless, 2006).

Both FFAR1+/+ and −/− mice underwent a 63 min habituation in a testing open-field box prior to cocaine (20 mg/kg) injection. Except for locomotion speed, there were no significant differences in the other three parameters (locomotion time, distance traveled, and time spent in the center region) during the habituation period (Figure 4 and Supplementary Figure S3). An acute administration of cocaine significantly increased locomotor activities in both +/+ and −/− mice. However, the overall locomotor responses to cocaine in terms of three parameters (locomotion time, distance traveled, and locomotion speed) were significantly reduced in −/− mice relative to +/+ mice (Figures 4A–C). Interestingly, we observed a change in the pattern of horizontal locomotion. Although we could not detect a significantly increased percentage of the time spent in the center region in −/− mice before cocaine injection with this experimental condition (3 min measurement in every 10 min), we could detect significantly increased ambulation of −/− mice in the center area of the open-field chamber after cocaine injection (Figure 4D).

As shown in Figure 5, pretreatment of FFAR1+/+ mice with GW1100 (10 mg/kg, i.p.) for 30 min also significantly reduced cocaine-induced enhancement of mobility time, distance travelled, and locomotion speed (Figures 5A–C). Although we could not detect a significant increase in the time spent in the center region, there was a tendency toward an increment of this parameter (Figure 5D).

Since impaired cocaine action on striatal monoamine transmission could be responsible for the blunted locomotor responses seen in FFAR1−/− mice, we examined the ability of acute systemic cocaine injection to elevate extracellular DA and 5-HT using in vivo microdialysis approach. As shown in Figures 6A,B, cocaine induced similar magnitude of potentiation in striatal DA and 5-HT levels in both +/+ and −/− mice, suggesting that the function of DA and 5-HT transporters appeared to be normal in −/− mice.

Finally, we examined the expression levels of DA and 5-HT transporters as well as D1 and D2 DA receptors using qPCR and western blot methods. Comparisons of the levels of these mRNAs and proteins between normal +/+ and −/− striata showed no significant differences (Figure 7).