Differential effects of D-cycloserine and amantadine on motor behavior and D2/3 receptor binding in the nigrostriatal and mesolimbic system of the adult rat

D-cycloserine (DCS) and amantadine (AMA) act as partial NMDA receptor (R) agonist and antagonist, respectively. In the present study, we compared the effects of DCS and AMA on dopamine D2/3R binding in the brain of adult rats in relation to motor behavior. D2/3R binding was determined with small animal SPECT in baseline and after challenge with DCS (20 mg/kg) or AMA (40 mg/kg) with [123I]IBZM as radioligand. Immediately post-challenge, motor/exploratory behavior was assessed for 30 min in an open field. The regional binding potentials (ratios of the specifically bound compartments to the cerebellar reference region) were computed in baseline and post-challenge. DCS increased D2/3R binding in nucleus accumbens, substantia nigra/ventral tegmental area, thalamus, frontal, motor and parietal cortex as well as anterodorsal and posterior hippocampus, whereas AMA decreased D2/3R binding in nucleus accumbens, caudateputamen and thalamus. After DCS, ambulation and head-shoulder motility were decreased, while sitting was increased compared to vehicle and AMA. Moreover, DCS increased rearing relative to AMA. The regional elevations of D2/3R binding after DCS reflect a reduction of available dopamine throughout the mesolimbocortical system. In contrast, the reductions of D2/3R binding after AMA indicate increased dopamine in nucleus accumbens, caudateputamen and thalamus. Findings imply that, after DCS, nigrostriatal and mesolimbic dopamine levels are directly related to motor/exploratory activity, whereas an inverse relationship may be inferred for AMA.

In rats, systemic AMA (100 mg/kg i.p.) augmented striatal acetylcholine (ACh) and nigral and striatal GABA, starting immediately post-injection 21 . Increases of striatal DA and serotonin (5-HT) concentrations were detectable, but not significant 22 . Also Maj et al. 19 observed no effect on striatal DA after application of lower AMA doses (10 to 80 mg/kg i.p.). Other scientific groups, however, reported a significant augmentation of striatal DA after systemic AMA (40 mg/kg s.c. 27 ; 46 or 92 mg/kg i.p. 28 ; 100 mg/kg i.p. 29 ), also starting immediately post-injection 29 . Likewise, intrastriatal infusion of AMA (0.1 mM or 1 mM) elevated the striatal release of both DA and GLU 29 .
After systemic DCS, so far, no in vivo imaging studies of D 2 R-like binding have been performed on either humans or rats. After chronic treatment with AMA (200 mg/day for at least 10 days), two in vivo imaging studies of striatal D 2 R-like binding have been conducted on Parkinsonian patients, using [ 11 C]raclopride as radioligand 30,31 . Both studies reported a significant elevation of striatal D 2 R-like binding, implying that, at least in Parkinsonian patients, AMA did not increase synaptic DA to an extent sufficient to effect a detecable competition with the exogenous radioligand. Contrarily, in our previous study on healthy rats 32 , AMA challenge with 40 mg/ kg i.p. reduced D 2 R-like binding in nucleus accumbens (NAC), caudateputamen (CP) and thalamus (THAL) relative to baseline, while 10 mg/kg diminished D 2 R-like binding in the anterodorsal HIPP (aHIPP).
In the present investigation, we assessed the effects of systemic DCS (20 mg/kg i.p.) on motor/exploratory behaviors and on D 2 R-like binding in regions of the rat nigrostriatal and mesolimbic systems, which are related to motor as well as cognitive and emotional functioning ( [33][34][35] . Effects of DCS on D 2 R-like binding were compared to our previous findings on the NMDAR antagonist AMA 32 .

Materials and Methods
Animals. Imaging studies of D 2 R-like binding sites were conducted on 38 adult male Wistar rats (ZETT, Heinrich-Heine University, Düsseldorf, Germany), weighing 397 ± 49 g (mean ± standard deviation [SD]; age: 3-4 months). The animals underwent morphological MRI, SPECT measurements in baseline and after injection of DCS or AMA, and behavioral testing after injection of DCS (20 mg/kg: n = 16) or AMA (40 mg/kg: n = 22). MRI and SPECT measurements in baseline and after pharmacological challenges were performed in randomized order and were separated by at least 3 days. Due to seizures or cardial arrest after the administation of the anaesthetic, 5 rats merely underwent behavioral measurements without subsequent D 2 R-like imaging. Behavioral data obtained after DCS and AMA were compared to the behavioral data obtained after vehicle (0.9% saline) in 16 further male rats of the same strain, age (3-4 months) and weight (418 ± 63 g). Behavioral data obtained after 40 mg/kg AMA and saline have been previously published 32,36 .
Rats were kept as previously described (e.g. 32 ). The study was performed in accordance with the German Law on the Protection of Animals and the National Institutes of Health guide for the care and use of laboratory animals (NIH Publications No. 8023, revised 1978). The protocol was approved by the regional authority (Landesamt für Natur, Umwelt und Verbraucherschutz, Nordrhein-Westfalen, Recklinghausen, Germany).
SPECT studies. D 2 R-like imaging in baseline and after challenge with DCS or AMA was conducted as previously described 32,36,37,40 . Also the employed small animal SPECT ("TierSPECT"; field of view: 90 mm; sensitivity: 22 for 123 I; spatial resolution: 3.4 mm for 123 I) employed in the present study was described in detail elsewhere 41 .
Behavioral studies. Immediately after administration of DCS, AMA or saline, motor and exploratory behaviors were assessed in an open field (Phenotyper ® , Noldus Information Technology, Wageningen, The Netherlands; dimensions: 45 × 45 × 56 cm) with EthoVision XT (Noldus Information Technology, Wageningen, The Netherlands) as previously described 32,36,37,40 . Durations (s) and frequencies (n) of ambulation, sitting (as a measure of "passive immobility" 45 ), rearing, head-shoulder motility and grooming were rated in blocks of 5 min for a total of 30 min. Subsequent to behavioral tests, rats were anaesthetized as described above and injected [ 123 I] IBZM.
Evaluation of SPECT imaging studies. D 2/3 R imaging data were evaluated with PMOD (version 3.5, PMOD Technologies Ltd., Zürich, Switzerland) as previously described 32,37 . Briefly, for each rat, SPECT and MR images were coregistered. Then, the MR image of each rat was coregistered with the Paxinos standard rat brain MRI 46 provided by PMOD. The necessary mathematical transformations were used to re-import the SPECT image previously coregistered with the MRI. On the individual overlays of each rat brain SPECT with the Paxinos standard rat brain MRI volumes of interest (VOIs) were defined, as previously described 32,37 . Thereby, the maximum VOI diameters were either in the range of or beyond the spatial resolution of the employed small animal SPECT. Regional BPs were estimated according to the simplified reference tissue model 47 by computing ratios of radioactivity counts obtained in the specifically-bound compartments (NAC, CP, THAL, SN/VTA, FC, MC, PC, aHIPP and pHIPP) to radioactivity counts in the cerebellar reference VOI.

Statistical analysis. D 2/3 R imaging studies.
Distributions of both regional BPs and behavioral data were tested for normality with the non-parametric Kolmogorov-Smirnov test (α ≤ 0.05). Regional BPs were neither uniformly distributed in baseline, nor after DCS or AMA (0.002 ≤ p ≤ 0.200).
Medians and interquartile ranges (25-/75-and 5-/95-percentiles) of regional BPs were computed for both compounds. Moreover, percentual differences of BPs after DCS or AMA relative to baseline were calculated. Regional BPs were compared between baseline and challenge (20 mg/kg DCS or 40 mg/kg mg/kg AMA) with the non-parametric Wilcoxon signed rank test for paired samples (two-tailed, α ≤ 0.05).
Behavioral studies. Behavioral variables (duration and frequencies of ambulation, sitting, rearing, head-and-shoulder motility and grooming) were evaluated with two-way analyses of variance (ANOVAs) with the factors "time" (denoting the individual 5-min time bins) and "treatment" (denoting challenge with either 20 mg/kg DCS, 40 mg/kg AMA or saline). In the majority of comparisons over time bins and treatments, the Shapiro-Wilk normality test failed (p < 0.050). Post hoc pairwise comparisons between treatment groups were performed for each variable in the individual time bins with the Holm-Sidak test (overall α ≤ 0.05). Furthermore, Spearman rank correlation coefficients (r; α ≤ 0.05) were calculated for regional radioligand binding and behavioral parameters in the individual time frames (min 1-5, 6-10, 11-15, 16-20, 21-25 and 26-30).
Statistical analysis was performed with IBM SPSS Statistics 23 (IBM SPSS Software Germany, Ehningen, Germany). and SigmaStat (version 3.5, Systat Software Inc., Erkrath, Germany). Figures 1 and 2 show images of the Paxinos standard rat brain MRI atlas 46 at different positions from Bregma together with the standard VOI templates provided by PMOD (left columns). The next columns show characteristic images of regional [ 123 I]IBZM accumulations on coronal slices in baseline (middle) and after challenge (right) with 20 mg/kg DCS ( Fig. 1) and 40 mg/kg AMA (Fig. 2), respectively, at the positions from Bregma depicted in the left columns 48 . Baseline and post-challenge scans after both treatments stem from the same rat.
In precedent studies on rats, systemic treatment with the DA precursor L-DOPA 36 , the DA reuptake inhibitor methylphenidate 49 , and the GABA A R agonist muscimol 37,40 diminished [ 123 ]IBZM binding to the rat D 2/3 R. The augmentation of DA concentrations in the synaptic cleft is common to all of these compounds. Since [ 123 ]IBZM competes with endogenous DA molecules for D 2/3 R binding sites, the observed decreases of D 2/3 R binding may be conceived to reflect increased levels of synaptic DA 50 . Therefore, it may be surmised that, also in the present study, the AMA-induced regional reductions of D 2/3 R binding were due to elevated DA concentrations in these areas, whereas the observed regional increases of D 2/3 R binding after pre-treatment with DCS indicate reductions of available DA. www.nature.com/scientificreports www.nature.com/scientificreports/ This is the first study, which assessed the effects of DCS on subcortical and neocortical DA in rats with a non-invasive in vivo imaging approach. Until now, the effect of DCS challenge on DA has only been studied in the rat CP, where either no effect 23 or a significant elevation of DA efflux 22 was observed. The latter is in contrast with our findings, which did not show an alteration of D 2/3 R binding in the CP after 20 mg/kg DCS. Likely reasons for this inconsistency are the differences in methods: firstly, we performed in vivo SPECT, while Bennett and Gronier 22 assessed striatal homogenates with high pressure liquid chromatography; secondly, we administered DCS systemically, while Bennett and Gronier 22 incubated striatal slices; and, thirdly, we used adult rats with a mean weight of 397 ± 49 g, while Bennett and Gronier 22 employed adolesent animals, weighing between 250 and 350 g.
The present result of significantly decreased ambulation after challenge with DCS compared to saline contradicts previous studies, which reported either no effect or merely a slight depression of spontaneous locomotor activity after systemic treatment with 0.3 to 65 mg/kg DCS [15][16][17][18] . Also the finding of decreased ambulation after challenge with AMA does not agree with previous fndings, showing an elevation of locomotor activity after systemic treatment with 40 to 100 mg/kg AMA [19][20][21] . Only the present finding of unaffected grooming duration confirms previous results obtained after 0.3 and 3 mg/kg DCS 17 .
Pharmacological effects on motor function in rodents are strongly dependent on age: the DA precursor L-DOPA, for instance, increased motor activity in neonatal (5 to 8 days old 51 ) and immature rats (18 to 20 days old 52 ) www.nature.com/scientificreports www.nature.com/scientificreports/ after doses of 12.5 to 50 mg/kg and 150 mg/kg, respectively, whereas motor activity was diminished in adult animals (25 to 30 days of age 52 ). In the present study, rats were considerably older (approximately 4 months old and weighing 397 ± 49 g) compared to the other investigations on DCS and AMA, in which adolescent (250-300 g 15 ; 250 ± 50 g 16 ; 200-250 g 17 ; 200 g 18 ; 110-115 g 19 ; 100-120 g 20 ; average weight of 250 g 21 ) animals were used. As a consequence, the present discrepancies may be accounted for by the difference of ages between samples in conjunction with NMDAR agonistic and antagonistic action on synaptic DA levels. The effects of DCS and AMA on both regional D 2/3 R binding and motor/exploratory parameters must be assessed in future investigations in rats of different ages in order to shed further light on this matter.
It is not surprising that the NMDAR agonist DCS and the NMDAR antagonist AMA exert opposite actions on D 2/3 R binding. Striking, however, are the differences in regional contributions: while AMA affects merely NAC, CP and THAL, DCS acts on the site of origin of DA fibers (SN/VTA) as well as on target regions of DAergic projections throughout the mesolimbic and nigrostriatal system (NAC, THAL, neocortex, HIPP) with the exception of the CP. www.nature.com/scientificreports www.nature.com/scientificreports/ In rodents, DCS increased GABA efflux in the mouse whole brain 18,24 . Moreover, decreases of GLU levels were observed in the rat amygdala 25 as well as in the mouse whole brain 18 , whereas no effect was detected in the rat FC 26 . This implies that overall alterations of GABAergic and GLUergic input to the nigrostriatal and mesolimbic target regions of ascending and descending fibers incurred a net decline of DA, reflected by the observed increases of D 2 R binding in NAC, SN/VTA, THAL, neocortex and HIPP.
As far as can be inferred from precedent investigations, the major difference between DCS and AMA action on cortical and subcortical DA levels is that GLU appears to be either unaffected or decreased by the former 18,25,26 but increased by the latter 29 . As previously outlined in more detail 32 , it may be assumed that AMA (contrarily to DCS) increases GLUergic input to the target regions of corticostriatal and corticomesolimbic projections, thus augmenting DA efflux in CP, NAC and THAL. It may be hypothesized that these differences in GLUergic and DAergic activation are related to the observed behavioral differences, namely decreased rearing duration and frequency (min 1-10, each) and increased sitting frequency (min [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] after AMA compared to DCS. Moreover, also in the second half of the testing time, after AMA, exploration was primarily performed by the sitting animal merely moving its head and shoulders (min [11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]. In the direct pathway (CP -pars reticulata of the SN/internal globus pallidus) DA disinhibits GABAergic neurons, incurring an activiation of the mesencephalic, diencephalic and brainstem motor centers, whereas, in the indirect pathway (CP -external globus pallidus/subthalamic nucleus -pars recticulata of the SN/internal globus pallidus), GABAergic neurons are inhibited by DA, resulting in a suppression of motor activity 53 . Moreover, the NAC with its afferents to the limbic system and its efferents to the GP acts as a limbic-motor interface, which is pivotal for the translation of emotional and motivational states into action 54 . Correlation analysis revealed that, after DCS, high sitting frequency (at 15 min post-challenge) predicted high D 2/3 R binding (and low DA) in CP, FC and MC (at 75 min post-challenge). Moreover, a low frequency of head-shoulder motility (immediately post-challenge) predicted high D 2/3 R binding (and low DA) in NAC, CP, THAL, SN/VTA, FC and pHIPP (at 75 min post-challenge). Contrarily, after AMA, a high sitting frequency (at 6 min post-challenge) predicted low D 2/3 R binding in NAC and CP (at 75 min post-challenge), whereas low durations and/or frequencies of ambulation and head-shoulder motility (immediately post-challenge) predicted low D 2/3 R binding in NAC, CP, THAL and SN/VTA (at 75 min post-challenge). This infers that after DCS and AMA, the altered levels of DA in the individual regions of the nigrostriatal and mesolimbic pathway within the first 15 min post-injection differentially affected motor neurons in conjunction with emotional/motivational states. Thereby, after DCS, lower regional DA concentrations produced the effect that rats were less able (and/or less "motivated") to ambulate or explore their environment by head-shoulder movements, but more able (and/or more "motivated") to exhibit rearing behavior. In contrast, after AMA, higher regional DA levels induced a general behavioral depression characterized by decreased rearing and increased sitting, while, consistently, exploration mainly consisted of head-and-shoulder movements. The question still remains to be solved, however, in as much the regional BPs (and DA levels) at the time of in vivo imaging correspond to the DA levels at the time of data acquisition in the open field. Future investigations are needed, in which behavior are assessed for a longer time than 30 min post-challenge. Moreover, regional D 2/3 R binding should be determined in different sets of animals at various times after [ 123 I]IBZM application.
In the present study, findings may have been biased by the employment of the NMDAR antagonist ketamine as anaesthetic. Since ketamine has previously been shown to enhance DA release in rats (e.g. [55][56][57], it can not be excluded that DA release elicited by ketamine actually reduced the amounts of visible regional D 2/3 receptor binding after both DCS and AMA. Effects on neostriatal and/or ventrostriatal DA, however, are exerted by practically all known anaesthetics, including pentobarbital, propofol, halothane, chloral hydrate and isoflurane (for review see 58 ). Therefore, we decided to maintain the usage of ketamine, which was employed in all our previous investigations. Since also this possible pitfall concerns the outcome of SPECT measurements both in baseline and post-challenge, the obtained BPs remain comparable between conditions.

Conclusion
Taken together, in adult rats, DCS increased D 2/3 R binding in NAC, SN/VTA, THAL, FC, MC, PC, aHIPP and pHIPP, whereas AMA decreased D 2/3 R binding in NAC, CP and THAL. The elevations of D 2/3 R binding after DCS reflect a reduction of available DA throughout the nigrostriatal and mesolimbic system, while the reductions of D 2/3 R binding after AMA indicate an increased availability of DA in NAC, CP and THAL. Findings imply a direct relationship between nigrostriatal and mesolimbic DA levels and motor/exploratory activity after DCS, but an inverse relationship after AMA.