All experimental procedures involving the use of animals and the analysis of brain anatomy were approved by the Institutional Care and Use Committees of Tokushima University, Japan. Adult male C57BL/6 mice aged 8–9 weeks were purchased from Nihon SLC Co. (Shizuoka, Japan). Mice were housed in a controlled environment (23 ± 1°C, 50 ± 5% of humidity) with 12 h light/dark cycle. Mice were allowed to take food and tap water ad libitum.

Mice were anesthetized with isoflurane (Sigma-Aldrich, St. Louis, MO, USA) and were mounted on a stereotaxic frame (Narishige, Tokyo, Japan). Each mouse received a stereotaxic injection of 6-OHDA-HCl (8.2 μg) dissolved in 4 μl of saline containing 0.02% ascorbic acid. Two 2-μl injections were administered into the striatum at a rate of 1 μl/min. The needle was left in place for 5 min to allow diffusion away from the injection site. The stereotaxic coordinates according to the mouse brain atlas (Paxinos and Franklin, 2001) were anterior-posterior, +0.5; medial-lateral, +2.4; and dorsal-ventral, −4.0 and −3.0. Mice were allowed to recover for 3 weeks and then apomorphine (Sigma-Aldrich; 0.5 mg/kg)-induced rotation behavior was studied over the course of 60 min. Mice with contralateral rotations (>7 times/min) were chosen and used for further studies.

Three weeks after the 6-OHDA-lesioning, mice received intraperitoneal injections of L-DOPA (Sigma-Aldrich; 20 mg/kg of free base) dissolved in 0.9% saline and intraperitoneal injections of benserazide-HCl (Sigma-Aldrich; 12 mg/kg) dissolved in 0.9% saline 20 min before daily administration of L-DOPA over 10 days. On day 11, the mice underwent behavioral studies and were then sacrificed for histological studies.

AIM scoring was performed according to previous reports (Cenci et al., 1998; Pavón et al., 2006; Santini et al., 2007). AIM scores were obtained after the last injection of L-DOPA for 1 min every 10 min over a period of 140 min. For the evaluation, each mouse was placed in a glass cylinder (diameter of 12 cm). Purposeless movements were classified on the basis of their topographic distribution. The following four subtypes of AIMs were present: locomotive (tight contralateral turns), axial (twisted posturing of the neck and upper body toward the contralateral side), forelimb (jerky movements of the contralateral forelimb, and/or grabbing movement of the contralateral paw), and orolingual (jaw movements and tongue protrusion toward the contralateral side). Each subtype was scored as follows; 0, absent; 1, occasional; 2, frequent; 3, continuous; 4, continuous, not interrupted by sensory stimuli.

Immediately after the last AIM scoring, the mice were intraperitoneally administered a lethal dose of pentobarbital (Sigma-Aldrich). They were then transcardially perfused with 0.01 M phosphate-buffered saline (PBS) at pH 7.2, followed by cold 4% paraformaldehyde in 0.1 M phosphate buffer at pH 7.2. The brains were removed, post-fixed overnight in the same fixative at 4°C, and stored in a 10–30% sucrose gradient in 0.1 M phosphate buffer at 4°C for cryoprotection. Sixteen-micrometre-thick sections were cut on a cryostat and stored in PBS containing 0.05% NaN₃ until use.

Immunostaining was performed on free-floating sections using the tyramide signal amplification (TSA) method, as in our previous report (Okita et al., 2012). After blocking endogenous peroxidase activity, the sections were incubated in PBS containing 3% bovine serum albumin (BSA) for 60 min. They were then incubated with antibodies against one of the following (diluted in PBS-BSA): Gα_olf (rabbit polyclonal, 1:5000; Santa Cruz Biotechnology, Santa Cruz, CA, USA), tyrosine hydroxylase (TH, rabbit polyclonal, 1:100,000) (Sato et al., 2008; Morigaki and Goto, 2016), D₁R (mouse monoclonal, 1:5000; Novus Biologicals, Littleton, CO, USA), A_2AR (mouse monoclonal, 1:5000; Santa Cruz Biotechnology), D₂R (rabbit polyclonal, 1:2000; Merck Millipore, Billerica, MA, USA) or c-Fos (rabbit polyclonal, 1:50,000; Oncogene Science, Cambridge, MA, USA) for 18 h. The bound antibodies were detected using the Histofine Simple Stain Kit (Nichirei, Tokyo, Japan) and the TSA-system with Cyanine3 or Fluorescein (Perkin Elmer, Shelton, CT, USA). For double immunofluorescence staining, the sections stained for Gα_olf using Cyanine3 were incubated in 0.1 M glycine-HCl (pH 2.2) at room temperature for 30 min. After rinsing in PBS for 1 h, the sections were then incubated overnight at room temperature in PBS containing 3% BSA and a rabbit polyclonal antibody against the μ-opioid receptor (MOR; 1:20,000; Millipore, Billerica, MA, USA), a mouse monoclonal antibody against D₁R (1:5000; Novus Biologicals), or a mouse monoclonal antibody against A_2AR (1:5000; Santa Cruz Biotechnology). The bound antibodies were detected using the Histofine Simple Stain Kit (Nichirei) and the TSA-system with Fluorescein (Perkin Elmer).

Dual-antigen recognition PLA experiments were conducted using the Brightfield Duolink PLA kit reagents (Sigma-Aldrich) according to the manufacturer’s recommendations with some modifications. Briefly, after blocking endogenous peroxidases in PBS containing 0.1% H₂O₂ for 30 min, the free-floating sections were incubated in PBS containing 3% normal goat serum for 60 min. They were then incubated in PBS containing 3% normal goat serum and a rabbit polyclonal antibody against Gα_olf (1:500; Santa Cruz Biotechnology) in combination with a mouse monoclonal antibody against D₁R (1:500; Novus Biologicals) or a mouse monoclonal antibody against A_2AR (1:500; Santa Cruz Biotechnology) for 18 h at room temperature. After subsequent secondary labeling with rabbit PLA minus and mouse PLA plus probes, we used the Brightfield Duolink Detection reagents for ligation and amplification and label probe binding according to the manufacturer’s instructions. For final signal visualization, we used the TSA-system with Cyanine3 (Perkin Elmer). After mounting on slides, the stained sections were counterstained with hematoxylin and were cover-slipped using 10% glycerol in PBS.

Digital microscopy images were captured using an Olympus BX51 microscope (Olympus, Tokyo, Japan) equipped with a DP40 digital camera (Olympus). They were imported into Adobe Photoshop CS4 and processed digitally. We adjusted contrast, brightness, and color balance. Using an image analyzer (MetaMorph, Molecular Device, Tokyo, Japan), we measured the optical densities of immunoreactive products and PLA signals in the striatum, which were represented by gray levels on non-colored digital images (Sato et al., 2008; Goto et al., 2013; Morigaki and Goto, 2015). Using the same protocol described above, we also measured optical densities of Gα_olf-immunoreactive products in the striosome and matrix subfields in the striatal sections double-stained for Gα_olf and MOR. We also counted the numbers of neuronal nuclei positive for c-Fos in a 0.5 mm × 0.5 mm field in the striatum and globus pallidus, as in our previous report (Tanabe et al., 2014). These morphometric analyses were carried out in a blind manner.

All experimental values are expressed as means ± SEM. For two-group comparisons, we used a paired two-tailed t-test. Multiple comparisons were analyzed using one-way or two-way analysis of variance (ANOVA), followed by Bonferroni’s post hoc tests for pair wise comparisons. Statistical analyses were performed using Stat View 5.0 (SAS Institute, Cary, NC, USA) software. P-values of less than 0.05 were considered statistically significant.

To model the generation of AIMs in PD following repeated L-DOPA treatments, we employed a well-established PD mouse model in which mice first received unilateral injection of 6-OHDA into the striatum (Santini et al., 2007). After 3 weeks of recovery and an apomorphine test, the mice were subjected to L-DOPA treatment for 10 days according to standardized protocols (Figure 1A). In this study, 6-OHDA-lesioned mice administered daily injections of benserazide-HCl (12 mg/kg) alone for 10 days were designated as “PD” models. Six-OHDA-lesioned mice that received daily injections of L-DOPA (20 mg/kg) and benserazide-HCl (12 mg/kg) for 10 days and finally exhibited LIDs with total AIM scores of more than 20 were designated as “PD with Dyskinesia (PD-D)” models. Among 6-OHDA-lesioned mice that received daily injections of L-DOPA (n = 28), 25 mice (~90%) were grouped into the PD-D model. Mice that received no drug treatment, except for anesthetic drugs, were used as “naïve controls”.

In both PD and PD-D mice, IHC with anti-TH antibody revealed a severe loss of nigral dopaminergic cells (Figure 1B) and striatal dopaminergic afferents (Figure 1C) on the side of the 6-OHDA injection. Quantitative measurements (Figure 1D) revealed a greater-than-90% reduction in TH labeling in the dorsolateral (DL) striatum on the lesioned side when compared to the non-lesioned side (lesion side, 3.9 ± 2.1; non-lesion side, 56.5 ± 8.9; means ± SEM; n = 10; two-tailed t-test, P < 0.01). Figure 1E shows the time course of changes in LIDs as determined by AIM scoring in PD-D mice. AIMs were maximal 40 min after L-DOPA administration, declined after 70 min, and almost disappeared after 120 min.

Figure 2A depicts the known distributional patterns of Gα_olf, D₁R, and A_2AR in a simplified basal ganglia circuit diagram. Note that Gα_olf is mainly localized with D₁R in the D1-cells that form the striatonigral pathway, while it is localized with A_2AR in the D2-cells that form the striatopallidal pathway. Using IHC, we reappraized the localization profile of Gα_olf immunoreactivity (IR) in the mouse striatum. Low-magnification microscopic images show strong Gα_olf labeling in the striatum (Figure 2B), particularly in the DL region (Figure 2B’, arrows). As in our previous report (Sako et al., 2010), Gα_olf IR was differentially concentrated in the different striatal compartments, with heightened Gα_olf labeling in the striosomes relative to the matrix (Figures 2C,D). Optical density measurements (Figure 2E) also revealed that Gα_olf IR in the striosomes was significantly higher than that in the matrix (striosomes, 39.8 ± 5.0; matrix, 23.5 ± 6.0; means ± SEM; n = 10; two-tailed t-test, P < 0.01). Microscopic images with high magnification show numerous tiny dots of Gα_olf IR densely distributed in the DL striatum (Figures 2F–H). In the double-labeling study, Gα_olf-positive dots were frequently localized in MSNs labeled for D₁R (Figure 2G) or A_2AR (Figure 2H). Using dual-antigen recognition in situ PLA, which indicates that two proteins are in close proximity (Söderberg et al., 2006), we also found that dot signals indicating the presence of Gα_olf protein in close proximity to D₁R protein (D₁R-Gα_olf; Figure 2I) or A_2AR protein (A_2AR-Gα_olf; Figure 2J) were abundantly distributed in the DL striatum.

To examine the dopaminergic regulation of Gα_olf protein levels in the striatum, we performed a quantitative IHC using an anti-Gα_olf antibody on striatal sections prepared from naïve control, PD, and PD-D mice. In low-magnification microscopic images, PD mice (Figure 3A) showed dramatic increases in Gα_olf IR in the dorsal striatum on the 6-OHDA-lesioned side when compared to non-lesioned side. In contrast, PD-D mice (Figure 3B) only had a modest increase in striatal Gα_olf IR on the 6-OHDA-lesioned side relative to non-lesioned side. Higher-magnification microscopic images of the DL striatum also show that, when compared to naïve controls (Figure 3C), there is a marked, but, slight increase in Gα_olf IR in the 6-OHDA-lesioned striatal areas of PD (Figure 3D) and PD-D (Figure 3E) mice. As indicated in a previous report (Ruiz-DeDiego et al., 2015), it is likely that dopamine depletion increases Gα_olf IR mainly in the matrix of the 6-OHDA-lesioned striatum, leading to a loss of the striosome-predominant pattern of Gα_olf IR expression in PD mice. However, daily treatment with L-DOPA reverses the lesion-induced increase in Gα_olf IR primarily in the matrix, leading to reappearance of the striosome-predominant pattern of Gα_olf IR expression in PD-D mice. These visual impressions were confirmed by quantitative densitometry analyses of the DL striatum (Figures 3F–H), as follows. We found a significant and marked increase of 101% (P < 0.001, two-way ANOVA) in Gα_olf IR levels in the 6-OHDA-lesioned striatum of PD, when compared to naïve controls. There was a 67% decrease (P < 0.01, two-way ANOVA) in Gα_olf IR levels in the 6-OHDA-lesioned striatum of PD-D mice when compared to that of PD mice (Figure 3F; PD mice: non-lesion, 102 ± 17% and 6-OHDA-lesion, 201 ± 28%; PD-D mice: non-lesion, 103 ± 19% and 6-OHDA-lesion, 134 ± 16%; % of naïve control mice ± SEM; n = 15).

In the striatal compartments of the DL striatum, we found a significant increase of 38% (P < 0.01, two-way ANOVA) Gα_olf IR levels in the striosomes of 6-OHDA-lesioned striatum of PD, when compared to naïve controls. There was no apparent difference (P > 0.05, two-way ANOVA) in striosomal levels of Gα_olf IR in the 6-OHDA-lesioned striatum between PD and PD-D mice (Figure 3G; PD mice: non-lesion, 101 ± 22% and 6-OHDA-lesion, 138 ± 26%; PD-D mice: non-lesion, 102 ± 19% and 6-OHDA-lesion, 122 ± 20%; % of naïve control mice ± SEM; n = 15). We also found a significant increase of 96% (P < 0.001, two-way ANOVA) in matrix levels of Gα_olf IR in the 6-OHDA-lesioned striatum of PD mice, when compared to those of naïve controls. There was a 66% decrease (P < 0.01, two-way ANOVA) in matrix levels of Gα_olf IR in the 6-OHDA-lesioned striatum of PD-D mice when compared to those of PD mice (Figure 3H; PD mice: non-lesion, 99 ± 23% and 6-OHDA-lesion, 196 ± 24%; PD-D mice: non-lesion, 102 ± 12% and 6-OHDA-lesion, 130 ± 22%; % of naïve control mice ± SEM; n = 15). These findings indicate that dopamine depletion causes a dramatic increase in Gα_olf levels in the DL striatum, particularly in the matrix. Daily exposure to L-DOPA induces a down-regulation of this lesion-induced increase in Gα_olf expression.

To examine dopaminergic regulation of D₁R, A_2AR and D₂R expression in the DL striatum, we performed quantitative IHC on sections prepared from 6-OHDA-lesioned striata of PD and PD-D mice (Figure 4). We observed no significant changes (P > 0.05, one-way ANOVA) in the expression levels of D₁R IR in PD or PD-D mice when compared to naïve controls (Figures 4A,B; PD mice, 101 ± 21%; PD-D mice, 98 ± 17%; % of naïve control mice ± SEM; n = 15). The expression levels of A_2AR IR in PD and PD-D mice were not significantly different (P > 0.05, one-way ANOVA) from those in naïve controls (Figures 4C,D; PD mice, 103 ± 15%; PD-D mice, 118 ± 13%; % of naïve control mice ± SEM; n = 15). We found a significant increase in the expression of D₂R in PD (P < 0.01, one-way ANOVA), but not PD-D (P > 0.05, one-way ANOVA), mice when compared to naïve controls (Figures 4E,F; PD mice, 123 ± 15%; PD-D mice, 112 ± 22%; % of naïve control mice ± SEM; n = 15). These findings indicate that dopamine depletion causes a significant increase in striatal D₂R expression, which is reversed by daily treatment with L-DOPA. In addition, dopamine depletion and L-DOPA replacement cause no significant changes in striatal expression of D₁R and A_2AR in the dopamine-denervated striatum.

To examine the dopaminergic regulation of striatal levels of Gα_olf protein in close proximity to D₁R protein, we used a sensitive in situ PLA in sections prepared from 6-OHDA-lesioned striata from PD and PD-D mice (Figure 5). In low-magnification microscopic images, a marked and moderate increase in D₁R-Gα_olf PLA signals was observed in the dorsal striatum in PD and PD-D mice when compared to naïve controls (Figures 5A,B). Higher-magnification microscopic images of the DL striatum also show that compared to naïve controls (Figure 5C), there is a marked and moderate increase in the D₁R-Gα_olf PLA signals in 6-OHDA-lesioned striatal areas in PD (Figure 5D) and PD-D (Figure 5E) mice. Quantitative densitometry analyses of the DL striatum revealed increases of 92% (P < 0.001, one-way ANOVA) and 50% (P < 0.001, one-way ANOVA) in the D₁R-Gα_olf PLA signal in PD and PD-D mice, respectively, when compared to naïve controls. There was a decrease of 42% (P < 0.001, one-way ANOVA) in the D₁R-Gα_olf PLA signal in PD-D mice when compared to PD mice (Figure 5F; PD mice, 192 ± 25%; PD-D mice, 150 ± 21%; % of naïve control mice ± SEM; n = 10). These findings indicate that dopamine depletion causes a marked increase in striatal D₁R-Gα_olf PLA signal, which is downregulated by daily treatment with L-DOPA. However, there is a significant increase of striatal D₁R-Gα_olf PLA signal in PD-D mice compared to naïve controls.

To examine the dopaminergic regulation of striatal levels of Gα_olf protein in close proximity to A_2AR protein, we used a sensitive in situ PLA in sections prepared from 6-OHDA-lesioned striata from PD and PD-D mice (Figure 6). Notably, low-magnification microscopic images show an apparent decrease in the A_2AR-Gα_olf PLA signal in the DL striatum of PD-D mice when compared to both naïve control and PD mice (Figures 6A,B). Higher-magnification images also show the localization patterns of A_2AR-Gα_olf PLA signals in the DL striatum of naïve control (Figure 6C), PD (Figure 6D), and PD-D (Figure 6E) mice. Quantitative densitometry analyses of the DL striatum revealed decreases of 41% (P < 0.001, one-way ANOVA) and 45% (P < 0.001, one-way ANOVA) in A_2AR-Gα_olf PLA signal levels in PD-D mice, when compared to naïve controls and PD mice, respectively (Figure 6F; PD mice, 104 ± 24%; PD-D mice, 59 ± 21%; % of naïve control mice ± SEM; n = 10). These findings indicate that L-DOPA replacement, but not dopamine depletion, causes a significant decrease in the striatal A_2AR-Gα_olf PLA signal in the dopamine-denervated striatum.

To assess changes in striatal responsiveness to dopamine stimulation in PD and PD-D mice, we performed IHC using an antibody against c-Fos, which is known to be induced in the striatum and globus pallidus following the stimulation of D₁Rs and D₂Rs (Marshall et al., 1993; LaHoste and Marshall, 1994). We prepared striatal sections from PD and PD-D mice that received injections of L-DOPA (20 mg/kg) and benserazide-HCl (12 mg/kg) 2 h before sacrifice on day 11 (see Figure 1A). Microscopic images of the DL striatum stained for c-Fos from naïve control, PD and PD-D mice are shown in Figure 7A. Compared to naïve controls that also received the injections of L-DOPA (20 mg/kg) and benserazide-HCl (12 mg/kg) 2 h before sacrifice, we found a marked increase in the densities of c-Fos-positive (c-Fos⁺) nuclei in the 6-OHDA-lesioned striatum in both PD and PD-D mice. Quantitative densitometry analyses also showed a marked increase (P < 0.001, two-way ANOVA) in the density of c-Fos⁺ nuclei in the 6-OHDA-lesioned striatum in both PD and PD-D mice when compared to naïve controls. However, there was a decrease of ~40% (P < 0.001, two-way ANOVA) in the density of c-Fos⁺ nuclei in the 6-OHDA-lesioned striatum of PD-D mice when compared to PD mice (Figure 7B; naïve controls: 20 ± 12; PD mice: non-lesion, 31 ± 10 and 6-OHDA-lesion, 475 ± 55; PD-D mice: non-lesion, 29 ± 12 and 6-OHDA-lesion, 292 ± 49; means ± SEM; n = 10). Microscopic images of the globus pallidus stained for c-Fos obtained from naïve control, PD, and PD-D mice are shown in Figure 7C. Compared to naïve controls, we found increased densities of c-Fos⁺ nuclei in the globus pallidus on the lesioned sides in both PD and PD-D mice. Quantitative densitometry analyses also indicated a significant increase (P < 0.001, two-way ANOVA) in the density of c-Fos⁺ nuclei in the globus pallidus on the lesioned sides in both PD and PD-D mice when compared to naïve controls. Importantly, we found that there was an increase of ~140% (P < 0.01, two-way ANOVA) in the density of c-Fos⁺ nuclei in the globus pallidus on the lesioned side in PD-D mice when compared to PD mice (Figure 7D; naïve controls: 4 ± 2; PD mice: non-lesion, 5 ± 3 and 6-OHDA-lesion, 38 ± 8; PD-D mice: non-lesion, 8 ± 7 and 6-OHDA-lesion, 92 ± 11; means ± SEM; n = 10). These findings indicate that dopamine depletion causes a marked increase in the responsiveness of striatal D1-cells to dopamine stimulation, which is downregulated by daily treatments with L-DOPA. Given the changes in striatal D₂R expression in PD and PD-D mice (see above), it is likely that in the dopamine-denervated striatum, dopamine depletion may cause increased striatal D₂R expression, which then enhances the responsiveness of D2-cells to dopamine stimulation (Cai et al., 2002). Notably, L-DOPA replacement could induce a further increase in the responsiveness of D2-cells to dopamine stimulation despite no obvious increase in striatal D₂R expression in the dopamine-denervated striatum.

We used IHC to demonstrate that Gα_olf IR is highly concentrated in the DL striatum, which corresponds to the motor-sensory territory in rodents and is analogous to the putamen in primates (Graybiel, 2008). This implies that Gα_olf may have a unique position in regulating the activities of the cortico-thalamo-basal ganglia circuit involved in motor functions, i.e., the motor loop (Alexander and Crutcher, 1990), at the striatal level. Although a previous study revealed no obvious compartmental difference in Gα_olf mRNA expression throughout striatal development in rats (Sakagami et al., 1995), we observed differential concentrations of Gα_olf IR in the striosome and matrix compartments, with higher densities of Gα_olf IR in the striosomes relative to the matrix. This finding suggests that Gα_olf may be a key molecule for controlling differential responses of striosome-matrix systems to D₁R activation in adult mice. There is evidence that in experimental animal models with 6-OHDA-lesions (Hervé et al., 1993; Corvol et al., 2004; Alcacer et al., 2012; Ruiz-DeDiego et al., 2015) or in those with a total absence of D₁Rs due to D₁R gene targeting (Hervé et al., 2001), the upregulation of Gα_olf levels in the striatum is not accompanied by a parallel increase in Gα_olf mRNA expression. Thus, homeostatic regulation of striatal Gα_olf protein levels is thought to occur via post-translational mechanisms, wherein the altered expression of Gα_olf protein depends directly on its rate of usage (Hervé, 2011). We suggest that when compared to the matrix, the striosomes might have the lower levels of D₁R/Gα_olf stimulation, which may then lower the Gα_olf degradation rate and lead to accumulation of the protein. Our assumption is supported by the present finding that changes in striatal Gα_olf IR expression were primarily found in the matrix in both PD and PD-D mice.

In this study, we first used highly sensitive dual-antigen recognition in situ PLA using a combination of the Brightfield Duolink PLA kit reagents and the TSA system (see “Materials and Methods” Section). This dual-antigen recognition PLA technique allowed us to obtain specific and efficient fluorescent signals showing Gα_olf protein in close proximity to D₁R or A_2AR protein in the striatum. However, we cannot say that all the PLA signals detected here resulted from the direct interaction (or actual coupling) of Gα_olf protein with D₁R or A_2AR protein. Borroto-Escuela et al. (2013) have shown that PLA can indicate a close proximity between two proteins, which is not always a reflection of direct interaction. This is because in situ PLA signals can be detected when two protein epitopes are in close proximity with ranges of 10–30 nm or more. In addition, although the precise mechanisms by which Gα_olf protein interact with D₁R or A_2AR protein remains unclear, it was also noted that in striatal membrane, the content in Gα_olf protein would be almost one to two orders of magnitude higher than that in D₁R or A_2AR (Hervé, 2011).

As shown in previous studies (Alcacer et al., 2012; Ruiz-DeDiego et al., 2015), we found a marked increase in striatal Gα_olf protein levels in PD mice with 6-OHDA lesions. This is in line with evidence that dopamine depletion may lead to up-regulation of Gα_olf protein expression in the rat striatum (Hervé et al., 1993; Marcotte et al., 1994; Penit-Soria et al., 1997; Corvol et al., 2004; Rangel-Barajas et al., 2011) and in the putamen in patients with PD (Corvol et al., 2004). Given the evidence that striatal levels of D₁R (Shinotoh et al., 1993; Turjanski et al., 1997: Hurley et al., 2001) and other major mediators of D₁R signaling (Girault et al., 1989; Nishino et al., 1993) are unchanged in PD patients, the dramatic increase in striatal Gα_olf protein level may be a key event in the D₁R hypersensitivity that develops in PD (Alcacer et al., 2012). In support of this notion, we detected no obvious changes in striatal D₁R expression in PD mice.

Previous data have suggested that the up-regulation of Gα_olf protein levels in the dopamine-depleted striatum is post-translational (Hervé et al., 1993; Ruiz-DeDiego et al., 2015) and results from the disuse of the D₁Rs (Hervé et al., 2001). Indeed, daily administration of L-DOPA for 10 days resulted in a down-regulation of the increased Gα_olf protein levels in the 6-OHDA-lesioned striatum in PD-D mice. However, we also found a significant increase in striatal Gα_olf levels in PD-D mice when compared to naïve controls. In agreement with the changes in striatal Gα_olf levels in PD and PD-D mice, in situ PLA also revealed that striatal D₁R-Gα_olf PLA signals were dramatically increased in PD mice and moderately increased in PD-D mice. These findings imply an increased responsiveness of D1-cells to D₁R activation in PD-D mice, although this responsiveness is lower than that found in PD mice. Our assumption is also supported by the fact that, compared to naïve controls, a significant increase in the number of striatal c-Fos⁺ nuclei consequent to L-DOPA administration was evident in PD-D mice, although this increase was more pronounced in PD mice. Since Gα_olf represents the rate-limiting factor in D₁R-mediated cAMP production in D1-cells, these findings suggest that striatal Gα_olf level acts as a determinant for the increased responsiveness of D1-cells to dopamine stimulation in LID (see Figure 8).

It has been postulated that repeated exposure to dopaminergic agents leads to increased sensitivity of D2-cells to D₂R activation in the dopamine-depleted striatum in experimental animals (Engber et al., 1989; Asin et al., 1995; Kashihara et al., 2000). However, no obvious increase in striatal D₂R expression has been observed in PD patients treated with dopaminergic drugs (Rinne et al., 1981; Guttman and Seeman, 1985; Antonini et al., 1997; Thobois et al., 2004). In agreement with this notion, we found that pallidal c-Fos induction consequent to L-DOPA administration was more marked in PD-D mice compared to PD mice. On the other hand, there was increased expression of striatal D₂Rs in PD mice, but not in PD-D mice. This indicates that the repeated administration of L-DOPA results in an increased responsiveness of D2-cells to striatal D₂R activation in the dopamine-denervated striatum, and suggests that this phenomenon might underlie LID. Using an in situ PLA, we found that A_2AR-Gα_olf PLA signals were markedly reduced along with Gα_olf protein levels in the 6-OHDA-lesioned striatum of PD-D mice. This novel finding indicates that as in D1-cells, repeated exposure to L-DOPA causes down-regulation of Gα_olf protein levels in D2-cells in the dopamine-depleted striatum. This then leads to the facilitation of the effects of dopamine on D2-cells by reducing A_2AR/Gα_olf signaling-mediated cAMP production (see Figure 8). This may be the reason that PD-D mice display an increased responsiveness of D2-cells to dopamine stimulation. However, the mechanism by which repeated and pulsatile injections of L-DOPA causes a decrease in A_2AR/Gα_olf PLA signals in PD-D remains a matter of speculation, as follows.

A_2AR usage by endogenous adenosine results in a basal rate of Gα_olf degradation (Hervé et al., 2001). It has been shown that in experimental animals with 6-OHDA-lesions, chronic (or persistent) dopamine depletion caused no significant changes (Ballarin et al., 1987; Herrera-Marschitz et al., 1994; Nomoto et al., 2000) or slight decrease (Pinna et al., 2002) in the extracellular adenosine levels in the striatum. In accordance with these findings, our present results also showed no significant changes in striatal levels of A_2AR-Gα_olf PLA signals in PD mice. Thus, we suggest that chronic dopamine depletion per se might cause no obvious changes in A_2AR/Gα_olf signaling activities that depend on the endogenous adenosine levels in striatal D2-cells. However, it is known that endogenous levels of adenosine are increased in response to the activation of N-methyl-D-aspartate (NMDA) receptors (Delaney and Geiger, 1998; Delaney et al., 1998), which can be facilitated by D₁R stimulation (Cepeda and Levine, 2012; Morigaki and Goto, 2015), in the striatum. A landmark report has shown that in the rat striatum, transient (pulsatile) stimulation of D₁Rs facilitates the NMDA receptor-dependent increase in extracellular adenosine levels (Harvey and Lacey, 1997). These findings suggest that in 6-OHDA-lesioned mice with D₁R hypersensitivity, repeated exposure to L-DOPA may lead to a transient activation of D₁Rs, which then enhances the NMDA receptor-dependent increase in adenosine release in the dopamine-denervated striatum. Moreover, Nash and Brotchie (2000) have shown that in striatal slices prepared from rats with 6-OHDA lesions, NMDA receptor activation could cause a marked increase in adenosine release and, thereby, indirectly stimulate A_2ARs. Taken together, we speculate that in the 6-OHDA-lesioned striatum of PD-D mice, decreased Gα_olf levels in D2-cells might be due to increased extracellular adenosine levels caused by the daily pulsatile activation of striatal D₁Rs. If our assumption is correct, striatal D₁R signals might contribute to regulation of the Gα_olf protein levels in not only D1-cells but also D2-cells in the dopamine-depleted striatum.

Because adenosine/A_2AR signaling functionally opposes the actions of D₂Rs on D2-cells by its ability to increase the A_2AR/Gα_olf-dependent cAMP production, it has so far been suggested that A_2AR antagonism may boost the anti-parkinsonian action of D₂R agonists in treating PD symptoms (Jenner, 2003; Schwarzschild et al., 2006; Fuxe et al., 2007; Huot et al., 2013). In addition, based on the evidence that striatal A_2AR expression might be increased in PD patients with dyskinesia (Calon et al., 2004; Ramlackhansingh et al., 2011) and in dyskinetic animal models of PD (Jenner et al., 2009), it has also been suggested that adenosine A_2A sites might be a potential pharmacologic target for reducing LIDs (Jenner et al., 2009; Ramlackhansingh et al., 2011; Huot et al., 2013; Kanda and Uchida, 2014). In Japan, istradefylline, an A_2AR antagonist, is currently used in clinics for treating PD patients (Kondo and Mizuno, 2015). The drug has shown to improve “off” time in patients with advanced PD, but has not shown anti-LID effects in the absence of a reduction in dopaminergic drug dosage. Adjunct use of istradefylline often causes dyskinetic symptoms as a major adverse effect (Kondo and Mizuno, 2015). Considering usage-dependent Gα_olf degradation through adenosine/A_2AR, we assume that in PD patients treated with L-DOPA, adenosine/A_2AR antagonism might be effective in reducing the “priming” of LID. However, once LID is established, adenosine/A_2AR antagonism might exacerbate dyskinetic symptoms. Our assumption may corroborate the notion that A_2AR activation might be required for dyskinesia “priming” mechanism (Brotchie, 2005).