Male C57BL/6J mice (8–10 weeks old, weighing 20–27 g) were purchased from Charles River Laboratories Japan (RRID:IMSR_JAX:000664; Kanagawa, Japan). Upon arrival, the mice were maintained in the facility with controlled humidity (50 ± 20%) and temperature (23 ± 3°C), under a 12-h light/dark cycle (light on at 7:00 a.m.), with free access to food (MF chow pellets, Oriental Yeast, Tokyo, Japan) and water. The mice were housed in plastic cages (182 × 260 × 128 mm, CLEA Japan Inc., Tokyo, Japan) with sterilized paper bedding (Paperclean, Nihon SLC, Shizuoka, Japan), and acclimated for at least 4 days prior to experiments. The mice were identified by tail marking and grouped based on computer-based randomization by body weight stratification. All animal procedures were carried out in accordance with the applicable international, national, and institutional guidelines for the care and use of animals. All experiments were approved by the Committee of Fujimoto Pharmaceutical Corporation on Animal Experimentation (the approval number: AC-F-2666).

Selegiline hydrochloride (0.3, 1, 3, and 10 mg/kg; Fujimoto Pharmaceutical Corp., Osaka, Japan), rasagiline mesylate (0.1, 0.3, 1, and 3 mg/kg; Sigma-Aldrich, St. Louis, MO, United States; catalog number: SML0124), and pramipexole dihydrochloride monohydrate (0.1, 0.3, and 1 mg/kg; Wako Pure Chemical Industries, Osaka, Japan; catalog number: 169-26183) were dissolved in saline and administered subcutaneously (s.c.) 60 min before behavioral tests or HFS in the electrophysiological study. Benserazide hydrochloride (Sigma-Aldrich, St. Louis, MO, United States; catalog number: B7283) was dissolved in saline and administered intraperitoneally (i.p.) at a dose of 2.5 or 6.25 mg/kg 15 min before injection of 3,4-dihydroxy-L-phenylalanine (L-Dopa; Sigma-Aldrich, St. Louis, MO, United States; catalog number: D9628). L-Dopa was suspended in saline containing 0.25% (w/v) carboxymethylcellulose sodium and administered i.p. at a dose of 10 or 25 mg/kg 30 min before measurement of motor functions. All compounds were administered to mice in a volume of 10 mL/kg.

1-Methyl-4-phenyl-1,2,3,6-tetrahydropyridine hydrochloride (Sigma-Aldrich, St. Louis, MO, United States; catalog number: M0896; two independent batches) was dissolved in saline. The mice received four i.p. injections of saline or MPTP at a dose of 16, 17.5, 19, or 20 mg free base/kg at 2-h intervals, according to the method described previously by Mori et al. (2005). The mice were kept warm at 25 ± 1°C in metal cages (170 × 300 × 100 mm, Natsume Seisakusho Co., Tokyo, Japan) until the morning after the third MPTP injection to prevent fatal hypothermia, according to the method described previously by Jiao et al. (2015). In the present study, 20 of 513 mice died owing to unexpected adverse events, probably cardiac abnormalities after administration of MPTP. All experiments were performed 2 days after MPTP treatment (Figures 1A, 4A). Behavioral tests were performed with two independent batches of mice.

Locomotor activities of each mouse were automatically measured by an infrared-linked activity sensor (Neuroscience, Tokyo, Japan) at 5 min intervals for 30 min in a plastic cage (13 × 18 × 26 cm). The horizontal bar test (HBT) was performed according to the method described by Deacon (2013). Briefly, the mice were allowed to grasp the center of a horizontal wooden bar (diameter: 0.6 cm, width: 38 cm; 20 cm above the bench surface) with their forelimbs. Then, the latency to fall from the bar was manually measured within 30 s by an observer blinded to the treatment conditions.

Depression-like behavior was assessed by the tail suspension test (TST), according to the methods described by Steru et al. (1985) and Can et al. (2012). Briefly, each mouse was suspended from a bar by the tail with an adhesive tape in a three-walled compartment (55 × 15 × 15 cm). The mouse tail base area was covered with a transparent plastic tube (3 cm length) to prevent tail climbing behavior. The duration of immobility (defined as the time spent without any limb movement) was manually measured for 6 min by an observer blinded to the treatment conditions.

Anxiety-like behavior was assessed by the elevated plus maze (EPM) test, according to the method described by Lister (1987). Briefly, the mice were placed on the intersection (neutral zone) of an opaque white plastic plus maze (four arms: width: 4 cm, length: 25 cm; height: 50 cm). Two opposing arms were enclosed by transparent plastic walls (height: 10 cm), and the two open arms had transparent edges (height: 0.3 cm). The time spent in the open arms and the distance traveled were measured for 10 min by the tracking software EthoVision 3.0.13 (Noldus, Wageningen, Netherlands). The treatment conditions were not blinded.

Immediately after the HBT, the mice were sacrificed by cervical dislocation. After decapitation, their brains were dissected, fixed in 10% (v/v) neutral buffered formalin for 2 days, and embedded in paraffin. Coronal sections (thickness: 20 μm), including the substantia nigra (ranging from 2.8 to 3.2 mm posterior to the bregma) or the striatum (ranging from 1.5 to 0.7 mm anterior to the bregma), were cut with a microtome, deparaffinized with xylene, and subjected to antigen retrieval with 0.1% (v/v) trypsin. The sections were incubated in 0.3% (v/v) hydrogen peroxide in methanol, then in a blocking buffer [5% (v/v) normal goat serum in phosphate-buffered saline containing 0.2% (v/v) Triton-X] for 30 min, and further incubated at 4°C with a rabbit polyclonal antibody against tyrosine hydroxylase (TH; EC 1.14.16.2) diluted in the blocking buffer (RRID:AB_390204; 1:500, Millipore, Billerica, MA, United States; Shin et al., 2014). After incubation with biotinylated goat anti-rabbit IgG antibody in blocking buffer (RRID:AB_2313606; 1:500, Vector, Burlingame, CA, United States), the sections were incubated with an avidin-biotinylated horseradish peroxidase (HRP; EC 1.11.1.7) complex using VECTASTAIN ABC Standard Kit (RRID:AB_2336818; Vector, Burlingame, CA, United States). TH immunoreactivity was visualized by the 3,3′-diaminobenzidine staining method. TH-positive cells in the substantia nigra were counted inside a 500 × 500 μm counting frame by an observer blinded to the treatment conditions. The striatal TH-positive fiber density was analyzed in a 500 × 500 μm frame of the dorsolateral striatum by using the image analysis software WinROOF 7.0.0 (Mitani, Tokyo, Japan). The optical density in the dorsolateral striatum was corrected for non-specific background staining in the cerebral cortex by subtracting such staining.

Immediately after the TST or EPM, the mice were sacrificed by cervical dislocation. After decapitation, the cerebrum was dissected and stored at −80°C until measurement. MAO-A and -B activities in the cerebral mitochondria were measured by using serotonin (5-HT) and benzylamine, respectively, as substrates, together with the EnzyChrom MAO Assay Kit (BioAssay Systems, Hayward, CA, United States), according to the manufacturer’s instructions with a slight modification (substrate alteration).

The mice subjected to the TST were sacrificed by cervical dislocation, and their brains were rapidly removed after decapitation. The striatum and cerebral cortex were dissected and stored at −80°C until sample preparation. The sample preparation was carried out according to the method previously described by Kasai et al. (2017). Briefly, tissues were homogenized in 0.2 M perchloric acid containing isoproterenol (for striatum: 100 pg/mL; cerebral cortex: 10 pg/μL) as an internal standard. The homogenates were kept on ice for 30 min and centrifuged for 20 min at 15,000 × g at 4°C. The supernatants were passed through a 0.45-μm filter membrane. The filtered supernatants were stored at −80°C until high-performance liquid chromatography (HPLC) measurement. The tissue content of DA and its metabolites, 3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA), was measured in a HPLC-electrochemical detection system (ECD-700, Eicom Corp., Kyoto, Japan). 5-HT and its metabolite 5-hydroxyindoleacetic acid, and norepinephrine (NA) and its metabolite 3-methoxy-4-hydroxyphenylglycol, were also measured. Each sample was injected into a C18 reverse-phase column (Eicompak SC-5ODS: 3.0 mm × 150 mm, Eicom Corp., Kyoto, Japan) conditioned at 25°C. The mobile phase comprising 0.1 M acetic acid–citric acid buffer (pH 3.5), 15% (v/v) methanol, 190 mg/L sodium 1-octanesulfonate, and 5 mg/L ethylenediaminetetraacetic acid was delivered at a flow rate of 0.5 mL/min. The applied potential was set at +750 mV versus Ag/AgCl. The content of monoamines and their metabolites was calculated using standard curves, and expressed as μg/g wet tissue.

Electrophysiological tests were performed by using the method described previously by Izaki et al. (2001) and Hiraide et al. (2012). Under urethane anesthesia (1.5 g/kg, i.p.), a bipolar stimulating electrode was stereotaxically placed in the CA1/subicular region of the ventral hippocampus [Anterior–Posterior (AP): −3.6 mm, Medial–Lateral (ML): +3.4 mm, Dorsal–Ventral (DV): −2.0 to −3.5 mm from the bregma], and a recording electrode (stainless steel, 100-μm diameter) was lowered into the mPFC (AP: +1.9 mm, ML: +0.4 mm, DV: −1.3 to −1.8 mm from the bregma) according to the mouse brain atlas (Franklin and Paxinos, 2008). The population spike amplitude (PSA) in the mPFC was obtained from seven stimuli at 30-s intervals, and was recorded every 5 min by the data analysis software Scope 3.7.6 (AD Instruments, Sydney, Australia). The intensity of stimulation was adjusted for each mouse to elicit a PSA of approximately 60% of maximum amplitude (pulse duration: 250 μs). The PSA was expressed as a percentage of the baseline value before application of two series of HFS (10 stimuli/train at 10-s intertrain intervals, 50 trains at 4-ms intervals) at a 10 min interval. Area under the curve (AUC) of PSA for 60 min after the first HFS was calculated. The treatment conditions were not blinded.

One hour after a single injection of the drugs, mice were sacrificed by cervical dislocation and decapitated. The frontal cortices were dissected by using a mouse brain slicer matrix and stored at −80°C until use. Tissues were homogenized in radio-immunoprecipitation assay buffer comprising 25 mM Tris–HCl (pH 7.6), 150 mM NaCl, 1% (v/v) NP-40, 1% (w/v) sodium deoxycholate, 0.1% (w/v) sodium dodecyl sulfate (SDS), 1× phosphatase/protease inhibitor cocktail (Thermo Scientific, Waltham, MA, United States), and 1 mM phenylmethylsulfonyl fluoride. After centrifugation at 1,000 × g for 10 min, the supernatants were collected and centrifuged at 15,000 × g for 20 min at 4°C. The resultant supernatants were collected, and their total protein concentrations were determined by the Lowry method (DC Protein Assay, Bio-Rad, Hercules, CA, United States). The samples were denatured by boiling for 5 min in Laemmli sample buffer. Ten micrograms of protein was separated by electrophoresis through 9% (w/v) SDS–polyacrylamide gel and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, United States). The membranes were blocked with 5% (w/v) skim milk in Tris-buffered saline containing 0.05% (w/v) Tween 20, and then incubated with the following primary antibodies: a polyclonal antibody against rabbit CaMKII (RRID:AB_2067912; 1:5,000, Santa Cruz, Dallas, TX, United States) or against threonine-286 phosphorylated-CaMKII (RRID:AB_310282; 1:1,000, Millipore, Billerica, MA, United States; Anderson et al., 2008) and a monoclonal antibody against mouse β-actin (RRID:AB_476697; 1:10,000, Sigma-Aldrich, St. Louis, MO, United States). After incubation with HRP-conjugated secondary antibodies (RRID:AB_228341 or AB_228427), immunoreactive signals were detected by using Luminata Forte Western HRP Substrate (Millipore, Billerica, MA, United States), and the band intensities were quantified using the image analysis software ImageLab 3.0 (Bio-Rad, Hercules, CA, United States).

Statistical analyses were performed with SPSS 23.0 (IBM Corp., Armonk, NY, United States). Welch’s t-test was used to evaluate the difference between two groups with unequal variance. One-way ANOVA was used to compare more than two groups, followed by post hoc Dunnett’s, Tukey’s (equal variance), or Games–Howell (unequal variance) tests. Non-parametric data were analyzed by the Kruskal–Wallis test, followed by a post hoc Dunn’s test with Bonferroni adjustment. Possible correlation between immobility time in the TST and monoamine content or its turnover rate was analyzed by Pearson’s correlation test or Spearman’s non-parametric rank correlation test. Difference and correlation were considered significant at p < 0.05. In the present study, 20 of 493 animals were excluded before conducting analyses owing to incomplete testing due to technical insufficiency. The exact numbers for all experiments are presented in the figure legends.

Four repeated i.p. injections of MPTP to C57BL/6J mice at 20 mg/kg, but not 17.5 mg/kg, resulted in a significant reduction in suspension time in the HBT (20 mg/kg, p = 0.026). This motor impairment in 20 mg/kg-MPTP mice was ameliorated in a dose-dependent manner through combined treatment with L-Dopa and benserazide (Figure 1B). Administration of MPTP at a dose of 17.5 or 20 mg/kg significantly reduced the TH-positive fiber density in the striatum (17.5 mg/kg, p = 0.002; 20 mg/kg, p < 0.001). MPTP at 20 mg/kg, but not 17.5 mg/kg, induced a reduction in the number of TH-positive nigral cells (20 mg/kg, p = 0.029; Figure 1C). The mice that received MPTP at a dose of 17.5 or 19 mg/kg showed significantly longer immobility time than saline-treated mice (control mice) in the TST, a standard behavioral paradigm indicative of depression (17.5 mg/kg, p = 0.046; 19 mg/kg, p = 0.026; Figure 1D). In the 17.5 mg/kg-MPTP mice, a significant reduction in time length spent in the open arms of the EPM, a standard behavioral paradigm indicative of anxiety (p = 0.043; Figure 1E), was observed; however, the reduction did not correlate with the distance traveled (r = 0.383, p = 0.086). Although 17.5 mg/kg-MPTP treatment did not induce any apparent motor impairment, as shown in the results of the HBT (Figure 1B) and EPM (Figure 1E, distance traveled), it led to the extension of immobility time in the TST (Figure 1D) and to the reduction in time spent in the open arms of the EPM, suggesting that the 17.5 mg/kg-MPTP mice exhibited depression- and anxiety-like behavior. Thus, we chose 17.5 mg/kg MPTP for preparation of a premotor PD model in the subsequent experiments.

An anti-PD drug, pramipexole (a DA D₂ and D₃ receptor agonist), has been reported to be effective in treating depression associated with PD in randomized clinical trials (Barone et al., 2010; Seppi et al., 2011). Thus, to confirm the validity of the present MPTP-treated animal model for mimicking psychiatric symptoms in PD and to investigate any involvement of DAergic actions in depression-like behavior, pramipexole was administered to MPTP mice as a reference drug. A single s.c. administration of pramipexole shortened the extended immobility time of MPTP mice in the TST (0.3 mg/kg, p = 0.001; 1 mg/kg, p < 0.001; Figure 1F) in a dose-dependent manner, while 0.1–1 mg/kg doses did not increase spontaneous locomotor activities (Supplementary Figure 1). These findings indicate that pramipexole alleviates depression-like behavior in MPTP mice. For the next set of experiments, pramipexole at a dose of 0.3 mg/kg, which had shown a substantial recovery potential for MPTP-induced extended immobility time, was selected to compare its effects on NMS-like behavior with those of MAOBIs.

When 0.3–10 mg/kg selegiline was administered by a single s.c. injection at 60 min before the TST, the 10 mg/kg dose shortened the extended immobility time of MPTP mice (MPTP versus control mice, p = 0.026; 10 mg/kg selegiline-treated versus saline-treated MPTP mice, p = 0.001). This recovery effect of 10 mg/kg selegiline on the immobility time of MPTP mice was comparable to that of 0.3 mg/kg pramipexole (pramipexole-treated versus saline-treated MPTP mice, p < 0.001; Figure 2A). Similar to pramipexole, selegiline at any dose did not significantly increase the spontaneous locomotor activities in MPTP mice (Figure 2B).

Rasagiline, another MAOBI, has been reported to be 3–15 times more potent than selegiline in terms of MAO-B inhibition in a rodent brain in vivo (Youdim et al., 2001). Moreover, the clinical dose of rasagiline is approximately one-tenth of the selegiline dose in patients with PD (Tomlinson et al., 2010), and based on this ratio, pharmacological effects of selegiline and rasagiline in rodents have been compared at doses of 10 and 1 mg/kg, respectively (Abassi et al., 2004; Kasai et al., 2017). In the cerebrum of MPTP mice, selegiline (0.3–10 mg/kg) and rasagiline (0.1–3 mg/kg) markedly and comparably inhibited MAO-B activities (selegiline: 89.8–93.4%, rasagiline: 71.4–95.2%). In addition, selegiline and rasagiline at doses greater than or at 3 mg/kg, respectively, suppressed MAO-A activities significantly (3 or 10 mg/kg selegiline-treated versus saline-treated MPTP mice, p = 0.003 or p = 0.003; 3 mg/kg rasagiline-treated versus saline-treated MPTP mice, p < 0.001). The MAO-A-inhibitory activity of 10 mg/kg selegiline (36.2 ± 4.4%) was comparable to that of 3 mg/kg rasagiline (32.5 ± 1.9%; Figure 2C). Hence, we compared the effects of 0.3–10 mg/kg selegiline with those of 0.1–3 mg/kg rasagiline on depression-like behavior to clarify whether these MAOBIs administered for PD treatment exert comparable antidepressant efficacies. Rasagiline did not show antidepressant-like effects at any dose in MPTP mice. Selegiline at a dose of 10 mg/kg showed more potent efficacy in treating depression-like behavior in MPTP mice than rasagiline (10 mg/kg selegiline versus 1 or 3 mg/kg rasagiline, p < 0.05; Figure 2A).

In the EPM test, a single administration of either selegiline (1–10 mg/kg) or rasagiline (0.1–1 mg/kg) did not significantly recover MPTP-induced reduction in time spent in the open arms (MPTP versus control mice, p = 0.017) or the distance traveled (MPTP versus control mice, p < 0.001; Supplementary Figure 2). These results suggest that a single administration of selegiline has antidepressant-like effects, but not anxiolytic-like effects in MPTP mice, and its effect is neither a consequence of MAO inhibition nor of an increase in motor activities.

We next evaluated the effects of MAOBIs and pramipexole on the striatal DA content and its turnover rate, and on the content of cortical DA, 5-HT, and NA, and their turnover rates in MPTP mice subjected to the TST. Selegiline (10 mg/kg) and pramipexole (0.3 mg/kg) significantly lowered the striatal DA turnover rate that had been elevated by MPTP, whereas rasagiline (0.1–1 mg/kg) did not (ratio of DOPAC and HVA to DA: MPTP versus control, p < 0.001; pramipexole-treated versus saline-treated MPTP mice, p = 0.006; 10 mg/kg selegiline-treated versus saline-treated MPTP mice, p = 0.013; Figure 3A). A positive correlation was observed between the striatal DA turnover rate (ratio of DOPAC and HVA to DA) and the immobility time in the TST in each treatment group (selegiline: r = 0.676, p < 0.001; pramipexole: r = 0.541, p = 0.006; rasagiline: r = 0.529, p = 0.001; Figure 3A). MPTP treatment decreased DA and 5-HT content (p < 0.05), but did not significantly influence turnover rates of DA, 5-HT, and NA in the cortex (Figure 3B and Supplementary Figure 3).

The cortical DA turnover rate was markedly reduced by selegiline (10 mg/kg), but not by rasagiline or pramipexole (ratio of DOPAC to DA: selegiline-treated versus saline-treated MPTP mice, p < 0.001), and positively correlated with the immobility time in the TST in selegiline-treated MPTP mice (r = 0.514, p = 0.001; Figure 3B). Moderately positive correlations were also observed between the cortical 5-HT or NA turnover rates and the immobility time of mice that had received selegiline (5-HT: r = 0.568, p < 0.001; NA: r = 0.356, p = 0.033; Supplementary Figure 3). These results suggest that the antidepressant-like effects of selegiline in MPTP mice are attributable to the normalization of impaired nigrostriatal DAergic systems and to the enhancement of cortical DAergic systems. In addition, these results suggest that the antidepressant-like effect of pramipexole is mediated partly by normalization of the striatal DA turnover.

We aimed to clarify whether the decreased cortical monoamine content in MPTP mice and whether the enhancement of monoaminergic transmission by selegiline or pramipexole influence synaptic plasticity, because monoamine concentrations in the synaptic cleft modulate the threshold of synaptic plasticity (Goto et al., 2010). We performed electrophysiological recordings from the mPFC in MPTP mice after a single s.c. administration of selegiline (10 mg/kg, as an effective dose for depression-like behavior), rasagiline (1 mg/kg, based on the clinical dose ratio against selegiline), pramipexole (0.3 mg/kg, as an effective dose for depression-like behavior), or saline (Figure 4A). Among these groups, there were no differences in the baseline synaptic transmission rates assessed by the input–output curve for PSA in the mPFC (Figure 4B). As can be seen in Figures 4C,D, HFS in the hippocampus led to a sustained increase in PSA in the mPFC of control mice, indicating LTP induction. In contrast, corresponding LTP induction in the mPFC was not detected in MPTP mice after application of HFS (p = 0.005), suggesting that MPTP treatment induced impairment in cortical synaptic plasticity. Both selegiline (10 mg/kg) and pramipexole (0.3 mg/kg) significantly recovered the impaired LTP and the decreased AUC of PSA in the mPFC of MPTP mice up to the levels comparable to those in control mice (pramipexole-treated versus saline-treated MPTP mice, p = 0.017; selegiline-treated versus saline-treated MPTP mice, p = 0.002). However, rasagiline (1 mg/kg) failed to restore the impaired LTP in the mPFC of these mice.

Phosphorylation of CaMKIIα is crucial for LTP induction (Giese et al., 1998; Chang et al., 2017). In the present study, a significant decrease in CaMKIIα phosphorylation was detected in the frontal cortex of MPTP mice compared with that of control mice (p = 0.007). Selegiline (10 mg/kg) and pramipexole (0.3 mg/kg) normalized the MPTP-induced decrease in CaMKIIα phosphorylation (pramipexole-treated versus saline-treated MPTP mice, p = 0.003; selegiline-treated versus saline-treated MPTP mice, p = 0.002; Figure 4E), but rasagiline (1 mg/kg) did not. None of these drugs affected CaMKIIβ phosphorylation or the total expression of both CaMKII isoforms (Figure 4E). These results suggest that the MPTP-induced impairment of LTP is associated with reduced CaMKIIα activation, and that the restorative effects of selegiline and pramipexole on impaired LTP in MPTP mice are possibly mediated by CaMKIIα activation.