Pramipexole restores depressed transmission in the ventral hippocampus following MPTP-lesion

The hippocampus has a significant association with memory, cognition and emotions. The dopaminergic projections from both the ventral tegmental area and substantia nigra are thought to be involved in hippocampal activity. To date, however, few studies have investigated dopaminergic innervation in the hippocampus or the functional consequences of reduced dopamine in disease models. Further complicating this, the hippocampus exhibits anatomical and functional differentiation along its dorso-ventral axis. In this work we investigated the role of dopamine on hippocampal long term potentiation using D-amphetamine, which stimulates dopamine release, and also examined how a dopaminergic lesion affects the synaptic transmission across the anatomic subdivisions of the hippocampus. Our findings indicate that a 1-methyl-4-phenyl-1, 2, 3, 6-tetrahydropyridine induced dopaminergic lesion has time-dependent effects and impacts mainly on the ventral region of the hippocampus, consistent with the density of dopaminergic innervation. Treatment with a preferential D3 receptor agonist pramipexole partly restored normal synaptic transmission and Long-Term Potentiation. These data suggest a new mechanism to explain some of the actions of pramipexole in Parkinson´s disease.

It is likely, therefore, that DA and its regulation may be important in variety of hippocampal functions. This is particularly relevant for conditions such as Parkinson's Disease which is characterized by the loss/dysfunction of midbrain dopaminergic neurons with associated deficits in fine movement control and motor learning. In the last few years, however, there has been a growing interest in the study of non-motor symptoms such as emotional disturbances and cognitive decline, whose causes have remained largely unknown. Clinical and experimental findings have shown the existence of cognitive decline from early stages, and even before Parkinson's Disease is diagnosed. Whereas research on non-motor symptoms has mainly focused on frontostriatal functions, several studies in the last 20 years found hippocampal atrophy in Parkinson's Disease patients, some of which presented with memory impairment 17 . Depression and anxiety, mood disorders which have been related to vHip disturbances, are highly prevalent in Parkinson's Disease patients 18 . A recent cohort study involving a half million people, concluded that there is a direct correlation between depression and the risk of developing Parkinson's Disease 19 . Despite of the importance of these findings to the quality of life of people living with Parkinson's Disease and the evidence of the importance of the vHip, little is known about the mechanisms underlying this.
The neurotoxin 1-methyl-4-phenyl-1, 2, 3, 6-tetrahydropyridine (MPTP) is a commonly used animal model for parkinsonism. This model causes a degeneration of the dopaminergic system, depleting DA and their metabolites in dorsal striatum and prefrontal cortex. However, due to the difficulty of measuring DA in hippocampus and the question of which region was studied, the effects are controversial. Some studies found no decrease in DA levels after MPTP administration 20 while others showed a depletion after multiple MPTP injections 21,22 . Despite this uncertainty, MPTP intoxication has been shown to impair several types of memory and also damages synaptic transmission in mice 22,23 .
In this study we used electrophysiological techniques to investigate how DA pathways modulate hippocampal-dependent functions, using D-amphetamine (Amph) and R-pramipexole (PPX), a dopamine agonist in clinical use in Parkinson's Disease. We also utilized the MPTP model of Parkinsonism in order to examine the effect of dopaminergic denervation on hippocampal function. We demonstrate that pharmacological elevation of DA with Amph results in an increase in LTP, depression of the Input/Output (I/O) curve and an augmentation in Paired Pulse Facilitation (PPF) ratio under both baseline and MPTP lesion conditions. These effects were consistent across the hippocampus, although larger in vHip than dHip. Moreover, PPX treatment succeeded in restoring the already depressed I/O curve of MPTP-lesioned mice and increased LTP and reduced the PPF ratio in vHip. These findings provide insight into the cognitive deficits and mood disorders that affect Parkinson's Disease patients and partially explain the antidepressant properties previously described for PPX.

LTP varies between the dorsal and ventral sections of the hippocampus.
To investigate the heterogeneous nature of the hippocampus, we selected sections from the dHip and vHip ( Fig. 1A and B).
High frequency stimulation of the Schaffer collateral pathway resulted in larger LTP in dHip than vHip (control dHip, 151% vs. control vHip, 134%, p ≤ 0.0001) (Fig. 1C) when recorded from the proximal stratum radiatum of CA1. This was not accompanied by a change in excitability (I/O curve) (Fig. 1D) or probability of neurotransmitter release (PPF ratio) (Fig. 1E). We chose the 50% of maximum response (S 50 ) to compare the I/O curves in all experiments, since this was the threshold for PPF, baseline and LTP recording.
Amphetamine effects are more pronounced in ventral hippocampus. Amphetamine is a highly addictive psychostimulant that mainly acts through the dopamine transporter (DAT), the latter which controls DA reuptake from dopaminergic terminals and can modulate dopaminergic transmission 24 . Amphetamine reverses DAT transport and increases DA release 25 . We were interested in how this DA release modulates synaptic activity in hippocampus. A previous study found that DA caused an inverted-U shape action on cognition 26 , with both low dose and high doses resulting in degraded cognitive responses. Because of this we used two different concentrations (0.1 μ M and 10 μ M). Our results indicate that Amph evoked an increase in LTP in the both the dorsal and ventral hippocampus. When hippocampal sub-regions were analyzed we found that only the lower dose of Amph altered LTP in dHip (control, 151% vs. 0.1 μ M Amph, 166%; p ≤ 0.0001; Fig. 2A). Both doses of Amph increased LTP in vHip (control, 134% vs. 0.1 μ M Amph, 192%; p ≤ 0.0001; vs. 10 μ M Amph, 185%; p ≤ 0.0001; Fig. 2B).
Synaptic excitability is directly determined by the frequency of activation, with excitable synapses producing more and larger action potentials at lower stimulation thresholds 4 . Aberrant excitatory activity of neurons has been described in Parkinson's Disease. Counterintuitively, previous studies showed that Amph and other psychostimulants depressed the synaptic excitability of some brain structures like VTA and nucleus accumbens (NAc) 27,28 . In concordance with those studies, we found that Amph decreased the synaptic excitability in the hippocampus, but there were different effects along the dorsal and ventral slices. Amph had modest effects in the dHip, the lower dose produced a significant reduction (control, 40% vs. 0.1 μ M Amph, 35%; p ≤ 0.05; Fig. 2C). However, vHip exhibited a more pronounced reduction after Amph exposure (control, 41% vs. 0.1 μ M Amph, 27%; p ≤ 0.01; vs. 10 μ M Amph, 31%; p ≤ 0.05; Fig. 2D and F).
As far as we know, the effect of Amph upon PPF in the different regions of the hippocampus has not been previously studied. However, a previous report showed that Amph increases the PPF ratio in the NAc 29 . Neuromodulators directly regulate the probability of neurotransmitter release from presynaptic boutons by a complex interplay of different presynaptic properties including the ability of altering the size and properties of the vesicle pool, the location of voltage-gated calcium channels at the active zone or the magnitude of calcium influx 30 . The data showed that Amph did not alter the PPF ratio in the dHip (Fig. 2G), but did decrease the probability of neurotransmitter release in the vHip (control PPF ratio, 1.37 vs. 0.1 μ M Amph PPF ratio, 1.47; p ≤ 0.05; vs. 10 μ M Amph PPF ratio, 1.56; p ≤ 0.01; Fig. 2H).  Pramipexole enhances synaptic transmission in hippocampus. PPX is a preferential D 3 receptor agonist 31 . In contrast to the Amph-induced effect on LTP (Fig. 2), PPX enhanced LTP in the dHip (control, 151% vs. 10 μ M PPX, 182%; p ≤ 0.0001, Fig. 3A), and vHip (control, 134% vs. 10 μ M PPX, 152%; p ≤ 0.0001; Fig. 3B). These data agree with previous published data using a different D 3 -preferential agonist (7-OH-DPAT) 16 .
The effects of PPX on synaptic excitability were explored, and found not to alter it in the dHip (Fig. 3C). PPX had the opposite effect when compared with AMPH in the vHip (shown in Fig. 2D and F), evoking a significant increase in synaptic excitability (control, 41% vs. 10 μ M PPX, 53%; p ≤ 0.01; Fig. 3D).
The effects of recovery after MPTP lesion in ventral hippocampus. Stereological cell counts of the SNpc revealed that MPTP injections caused a 35% loss of nigral neurons (control, 5989 ± 72 SEM, n = 12; MPTP, 3897 ± 148, n = 21; p < 0.0001), and 28,7% loss of VTA (control, 5375 ± 206 SEM, n = 7; MPTP, 3732 ± 223 SEM, n = 5; p < 0.0001). Our aim was to investigate short-and long-term effects of MPTP lesion in the hippocampus. Seven days after MPTP, TH immunolabelling showed a dramatic denervation of the dorsal striatum (Fig. 4A) and the vHip (Fig. 4C). Sixty days after the MPTP lesion the remaining neurons activated sprouting mechanisms to partially reinnervate these regions, and a modest recovery in the density of innervation in the striatum and vHip was observed ( Fig. 4A and C, respectively). These new terminals had larger varicosities, consistent with synaptic plasticity in the striatum 32 . Western blot revealed that the compensatory sprouting was not enough to fully restore the TH levels at 60 days post-MPTP (Fig. 4D).

MPTP lesion triggers an abnormal hippocampal LTP.
We demonstrated that an MPTP nigral lesion increased LTP. When hippocampal sub-regions were separated, the data showed that dHip increased LTP (control, 151% vs. MPTP 60d, 162%; p ≤ 0.01; Fig. 5A). The vHip normally has a greater dopaminergic innervation, 7 days after the MPTP lesion we found enhanced LTP (control, 134% vs. MPTP 7d, 172%; p ≤ 0.0001; Fig. 5B). Sixty days after the MPTP lesion, LTP was still augmented (control, 134% vs. MPTP 60d, 152%; p ≤ 0.0001; Fig. 5B). These data suggest that the reduction of the dopaminergic innervation caused by MPTP lesion results in enhanced LTP, which is only partially reversed by reinnervation. These data suggest that dopamine innervation is required for normal LTP.

Amphetamine had opposite effects on LTP in the dorsal and ventral hippocampus after MPTP
lesion. The majority of currently available treatments are aimed at restoring DA levels; the most common of which is the DA precursor, L-Dopa. As it is recognized that excess DA can trigger emotional and cognitive impairments 12 , we examined the effect of elevated DA on LTP in the MPTP-lesioned mouse model.
Amphetamine, which rapidly increases the levels of extracellular dopamine, caused opposite effects along the dorso-ventral hippocampal axis of MPTP-lesioned mice. Specifically, there was a reduction of LTP in the dHip (MPTP 60d control, 162% vs. MPTP 60d + 0.1 μ M Amph, 147%; p ≤ 0.05; vs. MPTP 60d + 10 μ M Amph 119%; p ≤ 0.0001; Fig. 5A) and an enhancement of LTP in the vHip (MPTP 60d 152% vs. MPTP 60d + 0.1 μ M Amph, 192%; p ≤ 0.0001; Fig. 6B), the latter which is consistent with the dopaminergic terminal reinnervation shown in Fig. 4C. The main reason to perform the 7 day experiment after MPTP experiments was to check if the amphetamine-induced increase of LTP that we found in vH after 60 days of MPTP lesion was in response to the reinnervation and plasticity of the vH. Amphetamine did not induce any augmentation of LTP after 7 days in the vHip, (MPTP 7d control, 172% vs. MPTP 7d + 0.1 μ M Amph, 164%; Fig. 6B). This data suggests that a reinnervation of hippocampus happened, partially recovering the ability for amphetamine to induce LTP.

Discussion
In this study we investigated the role of dopaminergic pathways in the function of the dorsal and ventral hippocampus. These results confirm that the dopaminergic system is a key regulator of normal hippocampal function and the MPTP data suggests that the dopaminergic lesion has different effects on synaptic transmission along the dorso-ventral axis in the hippocampus, having more profound effects on the vHip where there is a larger dopaminergic innervation 14 . Furthermore, the MPTP lesion resulted in time-dependent changes in dopamine: a short-term dopaminergic denervation and longer-term compensatory mechanisms that resulted in dysfunctional synapses 35,36 .
Our results confirm that there is a differential capability to induce LTP across the dorso-ventral hippocampal axis. In agreement with these results we found that under normal conditions, LTP is larger in the dHip than vHip   as have been described previously 37 . This phenomenon can be partly explained by differences in density of dopaminergic innervation along the hippocampal dorso-ventral axis 14 . The question of how this difference could be influenced in a pathological condition such as Parkinson's Disease however, has yet to be answered. To address this question, we used a neurotoxin (MPTP) that preferentially causes a degeneration of dopaminergic neurons. We were interested in the short-term, but also particularly in the long-term effects of the lesion as this is more comparable to the type of chronic degeneration and compensatory repair that characterizes Parkinson's Disease. We hypothesize that under normal conditions the dopaminergic innervation suppresses hippocampal excitability and LTP. During the course of Parkinson's Disease the altered dopaminergic innervation results in hippocampal dysfunction and may partially explain some of the enhanced neuropsychiatric manifestations that occur in the disease. We found a distinct increase in LTP along the dorso-ventral hippocampal axis that corroborated our hypothesis. However, following MPTP lesion, the enhancement of LTP was more pronounced in vHip than dHip, coincident with the proportions of dopaminergic innervation. Interestingly, LTP induction after MPTP lesion was time-course dependent. Seven days after MPTP lesion, we found a robust increase in LTP in the vHip and the largest dopaminergic denervation. At 60 days post-MPTP lesion, LTP was still larger than control in both vHip and dHip, but downregulated compared with the short-term lesion in vHip. This moderation of LTP occurred in parallel with the dopaminergic reinnervation of vHip (Fig. 8). Also, we found that MPTP lesion caused a time-dependent depression in synaptic excitability, as indicated by alterations in PPF and the I/O curve. Recent works suggesting that a dopaminergic lesion could potentiate GABA inhibition could explain this phenomenon, but further experiments are needed 38 .
A theoretical model explains LTP as a mechanism for short term modification of the synapse and a longer term mechanism for stimulating spine formation 9 . Glutamate release is required to induce LTP, and induces de novo growth of functional spines 11 . Dopamine release from the VTA and SN inputs to the hippocampus are required for LTP at Schaffer collateral synapses where it controls the excitability of CA1 pyramidal neurons through direct modulation of GABAergic interneuron excitation 39 .
Moreover, it has been demonstrated that the late phase of hippocampal LTP (after 60 min of HFS) can be blocked in presence of D1 antagonists or D1 KO mice [40][41][42][43][44] . Although studies about synaptic transmission in hippocampus following MPTP lesion did not find significant differences of LTP from longer time of 30 min. post HFS or TBS 22,[45][46][47] , further studies are needed to check whether our results are only related to the short phase of LTP.
We suggest a model (Fig. 8) where a short-term lesion causes an enhanced LTP as direct result of the loss of dopamine innervation 23 . There is a discordance amongst the studies that have examined LTP following MPTP lesion, which is most likely caused by different experimental approaches e.g. varying timeframes post-MPTP, different schedule of MPTP treatment, different protocol to induce LTP and different origins/orientations of the slices along the dorso-ventral axis of the hippocampus 22,47,48 . No previous studies have differentiated between dHip and vHip 22,[45][46][47] . Further other studies found an enhancement of fEPSP when MPTP was applied to the hippocampus in the bath 23 . The theoretical model explains LTP as a mechanism for short term modification of the synapse and in the longer term a mechanism for stimulating spine formation. Other authors have found enhanced LTP induced by an injury. Injury of visual cortex is followed by processes of enhanced neuroplasticity like LTP 49 . Other studies found an increase in LTP was observed after closed head injury in hippocampal CA1 50 . These data support the view that that a compensatory mechanism for LTP or a mechanism of remodeling the neuronal networks is plausible following injury. Our results indicate that function is gradually and only partially restored in 60 days when the hippocampal DA networks are gradually recovered. This suggests that these newly formed terminals only partially conserve the normal function of DA networks, and this could be aided by using drugs that modulate dopaminergic actions (Amph-like or PPX).
Amphetamine is a commonly used cognitive enhancer prescribed for attention-deficit hyperactivity disorder (ADHD). It is known that Amph promotes LTP in CA1 region of the hippocampus 51 . Whilst the cellular response to Amph depends on the level of DAT expression, with its activity and potency varying across striatal sub-regions 52 , the mechanisms underlying the differential effects of Amph in the hippocampus has yet to be clarified. Therefore, we sought to determine whether Amph had a larger effect in vHip because of the larger dopaminergic innervation or a because of differential effects on cathecolamines and serotonin 24 . Our data indicate that Amph enhances LTP in the whole hippocampus, but its potency varies along the dorso-ventral axis, with the vHip showing the largest increase in LTP. Also, we found that Amph depressed the synaptic excitability in the vHip, as previously described in NAc 28 and VTA 27 . Moreover, Amph increased the PPF in the vHip, similar to that shown Amph (Dunnett´s post hoc test following a one-way ANOVA, not significant). (D) vHip vs. MPTP 7d (Unpaired t-test, S 50 repeated data shown in Fig. 5). vHip MPTP 7d vs. MPTP 7d 0.1 μ M Amph (Unpaired t-test at S 50 not significant). vHip MPTP 7d vs MPTP 7d 0.1 μ M Amph (Unpaired t-test at S 50 ****p < 0.0001). vHip vs. MPTP 60d (Unpaired t-test at S 50 shown in Fig. 5). vHip MPTP 60d vs. MPTP 60d 0.1 μ M or 10 μ M Amph (Dunnett´s post hoc test following a one-way ANOVA, not significant difference). (E) I/O curve comparing vHip vs. MPTP 60d (Unpaired t-test at S 50 , repeated data shown in Fig. 5). vHip MPTP 60d vs. MPTP 60d 0.1 μ M or 10 μ M Amph (Dunnett´s post hoc test following a one-way ANOVA, *p < 0.01). (F) PPF ratio in dHip vs. MPTP 60d (Unpaired t-test, *p < 0.05). dHip MPTP 60d vs. MPTP 60d 0.1 μ M or 10 μ M Amph (Dunnett´s post hoc test following a one-way ANOVA, *p < 0.05). (G) PPF ratio in vHip vs. MPTP 7d (Unpaired t-test, ****p < 0.0001). vHip MPTP 7d vs. MPTP 7d 0.1 μ M Amph (Unpaired t-test, ****p < 0.0001). (H) PPF ratio in vHip vs. MPTP 60d (Unpaired t-test, *p < 0.05). vHip MPTP 60d vs. MPTP 60d 0.1 μ M or 10 μ M Amph (Dunnett´s post hoc test following a one-way ANOVA, not significant difference). I/O curves and PPF ratio were recorded before high frequency stimulation. in the NAc 29 . Amph could also have dopamine independent effects, that is, acting directly on serotonin receptors or cannabinoid receptor type 1 27,53 .
We used multiple methods to assess the effect of the lesion; stereology of SNpc neurons (assessed as a 35% lesion); stereology of VTA neurons (29% lesion), TH immunohistochemistry of CPu (80% decrease in densitometry); western blot of hippocampal slices for tyrosine hydroxylase (show 75% decrease in TH that persists even up to 60 days, Fig. 4d). The appears to be no standard way to produce or assess dopaminergic lesions following a toxin lesion in rodents. The complexity includes; the dose, the method used to assess denervation and the timing of the assessment after intoxication 36 . Given these limitations, it appears that the lesions produced for this study are equivalent to those previous studies.
Utilising the MPTP-lesioned mice we confirmed that low dose Amph caused enhancement of LTP and this was dependent on the dopaminergic system. Seven days after MPTP lesion is coincident with profound dopaminergic denervation, Amph did not induce any LTP increase in the vHip. In contrast, after 60 days of MPTP lesion, when the dopaminergic innervation is partly restored, Amph did recover its ability to induce LTP. This data indicate that the dopaminergic system is crucial for low dose Amph-mediated LTP induction in the vHip. The observation that Amph caused a decrease in LTP in the dHip of MPTP-lesioned mice may explain cognitive and emotional alterations found in individuals treated with L-Dopa 12 . It is reasoned that Amph is acting as a DA releaser, increasing DA in the synaptic cleft to give the same functional result as current dopamine therapeutics. We want to emphasise an apparent parallelism between MPTP and Amph-induced LTP, where in both treatments synaptic excitability was depressed and PPF was increased in vHip.
We also examined the effects of PPX, a D 3 -preferent agonist, which was found to increase LTP in the hippocampus in a similar manner to that shown with another D 3 dopamine receptor agonist (7-OH-DPAT) 16 . The differences with Amph are possibly due to D 3 R being expressed not only in the presynaptic dopaminergic neurons, but also in the postsynaptic neuron where it modulates GABAergic transmission 54 . However, further experiments using KO mice or antagonists are needed to check whether this effect is D 3 R-mediated. We also found that PPX evoked a disinhibition in vHip, as demonstrated by either a decrease in PPF ratio or by an increase in the synaptic excitability (I/O curve). Hammad and Wagner described this phenomenon using another D 3 receptor agonist (PD 128907) 54 , although they did not separate slices along the dorso-ventral axis and our data indicate that this effect is vHip specific. This disinhibitory ability of PPX led us to question whether PPX could restore normal hippocampal function in MPTP-lesioned mice. Our results indicate that PPX succeeded, not only increasing the I/O curve and decreasing the PPF of MPTP-lesioned mice, but increasing the LTP in vHip. These data add a new mechanism of action for PPX, and could partly explain the antidepressant effects described for PPX in both Parkinson's Disease 33 and models of Parkinsonism 55 . Moreover, we propose a possible mechanism that could be considered to counteract the depressing effect caused by Amph and other psychostimulants 27,28 . However, whether this disinhibitory effect of PPX in the vHip could be used for restoring the depressed excitability induced by Amph and its impact over drug addiction should be further explored. Also, regarding to the mechanism of action of PPX, experimental studies carried out over the last two decades indicate that D 3 R agonists can mediate their neuroprotective effects through D 3 R-dependent and D 3 R-indepenendent mechanisms 56,57 . Therefore, further experiments would be needed to fully characterize all the dopaminergic receptors action in these models.
Our findings shed light on the role of dopaminergic system on hippocampal function, and particularly could help for the better understanding of the sequential mood changes associated with Parkinson's Disease. The main findings are: 1) the dopaminergic system has a prominent effect on the synaptic activity in vHip but lesser in the dHip; 2) The discovery of a differential susceptibility to dopaminergic MPTP lesion across the dorso-ventral axis in hippocampus. These lesions have distinct acute and chronic effects; 3) We propose a novel mechanism of PPX that should be further explored for cognitive and mood disorders associated with Parkinson's Disease.

Methods
Animals. Male C57BL/6 J mice aged 11 weeks weighing between 20 and 25 g were used for this study. All procedures involving mice conformed to the Australian National Health and Medical Research Council code of practice for the care and use of animals for scientific purposes and were approved by the Florey Institute animal ethics committee. All experiments were designed to minimize the number of animals used, pain and discomfort. MPTP Intoxication protocols. C57BL/6 J mice were administered MPTP (Sigma, USA) in an acute dosing regimen of 60 mg/Kg given over four injections (15 mg/Kg each) 2 hours apart 58 . Each experimental trial contained MPTP-lesioned animals that were randomly subdivided into a sham group and MPTP lesion group. Experimenters were blinded to the assignment for each of the groups. After either 7 or 60 days post-MPTP administration, mice were either culled and brains collected for LTP or they were deeply anaesthetized and 60 days of MPTP lesion vs. after 60 days of MPTP lesion treated with 10 μ M PPX (Unpaired t-test, not significant difference). (D) I/O curve comparing vHip control vs. after 60 days of MPTP lesion (Unpaired t-test at 50% of maximum stimulation, S 50 data shown in Fig. 3). I/O curve comparing vHip after 60 days of MPTP lesion vs. after 60 days of MPTP lesion treated with 10 μ M PPX (Unpaired t-test, ***p < 0.001). (E) PPF ratio indicating the probability of neurotransmitter release in dHip control vs. after 60 days of MPTP lesion (Unpaired t-test, *p < 0.05). PPF ratio indicating the probability of neurotransmitter release in dHip control vs. dHip after 60 days of MPTP lesion treated with 10 μ M PPX (Unpaired t-test, **p < 0.01). (F) PPF ratio indicating the probability of neurotransmitter release in vHip control vs. vHip after 60 days of MPTP lesion (Unpaired t-test, *p < 0.05). PPF ratio indicating the probability of neurotransmitter release in vHip control vs. vHip after 60 days of MPTP lesion treated with 10 μ M PPX (Unpaired t-test, **p < 0.01).
tissues collected for either western blot or stereology 58 . The size of the lesion was assessed using stereology. The total number of DA neurons in the SNpc was estimated using a fractionator sampling design 32,58,59 . Counts were made at regular predetermined intervals (x = 140 μ m, y = 140 μ m). Systematic samples of the area occupied by the nuclei were made from a random starting point. An unbiased counting frame of known area (45 μ m × 35 μ m) was superimposed on the image of the tissue sections using stereology software (MBF, Stereo Investigator) utilizing a 63x objective lens (Leica, N.A. 1.36). Experimenters were blinded to the treatments of each of the groups. The dopamine axonal innervation of the dorsal striatum was assessed by measuring the densitometry of TH immunoreactivity using Image J (v 1.49, NIH).
Microelectrode array electrophysiology. We study synaptic activity parameters such as; The I/O curve and PPF. The majority of hippocampal synapses terminate on dendrites that integrate with multiple inputs that compute to produce an output (field excitatory postsynaptic potential, fEPSP). I/O curve serves as an index of synaptic excitability of large neuronal populations 60 . PPF ratio (fEPSP2/fEPSP1) is used as an easy measure of the probability of neurotransmitter release, where an increase of PPF ratio is interpreted as a decrease of probability of neurotransmitter release, and vice versa 61 .
The animals were killed and brain removed. The brain was cut into 300 μ m Horizontal sections with a vibratome (Leica VT1200S) according to the lamellar orientation of hippocampus. We chose slices belonging to dHip, that serves cognitive function, and the vHip that corresponds to the affective hippocampus. Intermediate slices were discarded as it has partly overlapping characteristics with its neighbours 7 (see Fig. 1A). After sectioning, ventral or dorsal hippocampal slices were pre-incubated for 40 min at 34˚C with either D-amphetamine (0.1 μ M or 10 μ M; Sigma, USA, pramipexole (10 μ M; Sigma) in carbogen bubbled aCSF solution, or carbogenated aCSF solution alone (control).
One acute hippocampal slice was transferred from slice-holding container (60 min pre-incubation in carbogenated aCSF at 34 °C) to a 3D-MEA chip with 60 electrodes spaced 200 μ m apart (60 MEA 200/30 iR-Ti: MCS GnbH, Reutlingen, Germany). The slice was immobilized with a harp grid (ALA Scientific Instruments, New York, USA) and was continuously perfused with carbogenated aCSF (3 ml/min at 32˚C). The Schaffer-collateral pathway was stimulated by injecting a biphasic current waveform (100 μ s) through one selected electrode at 0.033 Hz. Care was taken to choose the stimulating electrode in the same region from one slice to the other. The peak-to-peak amplitude of fEPSP at the proximal stratum radiatum of CA1 was analyzed using LTP-Analyzer (MCS GnbH, Reutlingen, Germany). Following a 20 min incubation period, slices were continuously stimulated with medium-strength stimuli. When stable evoked fEPSPs were detected (after at least 20 min), the stimulus threshold was determined, and a stimulus strength-evoked response curve (i.e. input-output, stimulation voltage vs evoked response amplitude) was recorded by gradually increasing stimulus intensity until the maximal fEPSP  62 . Regulation of this synaptic organization is now considered pivotal in understanding the synaptic plasticity that underpins CPu dysfunction in addiction and Parkinson´s disease 63 . However, these specialized structures have been poorly studied in hippocampus, where a recent study suggests that DA terminals also directly control GABA inhibitory interneurons. Our data and other studies suggest that, DA indirectly modulates pyramidal neurons activity 39,62 . We propose that seven days after MPTP dopaminergic terminal loss disrupted synaptic transmission, as the I/O curve decreased and PPF increased. After sixty days of lesion, LTP was downregulated by the newly innervated synapses. However, the synaptic activity was not fully restored. PPX succeeded on restoring the impaired synaptic transmission, and increasing LTP in MPTPlesioned mouse.