Systemic inflammation triggers long-lasting neuroinflammation and accelerates neurodegeneration in a rat model of Parkinson’s disease overexpressing human α-synuclein

Increasing efforts have been made to elucidate how genetic and environmental factors interact in Parkinson’s disease (PD). In the present study, we assessed the development of PD-like symptoms on a genetic PD rat model overexpressing human α-synuclein (Snca+/+) at a presymptomatic age, exposed to a pro-inflammatory insult by intraperitoneal injection of lipopolysaccharide (LPS), using immunohistology, high-dimensional flow cytometry, electrophysiology, and behavioral analyses. A single injection of LPS to both WT and Snca+/+ rats triggered long-lasting increased activation of pro-inflammatory microglial markers, infiltrating monocytes and T-lymphocytes. However, only LPS Snca+/+ rats displayed dopaminergic neuronal loss in the substantia nigra pars compacta (SNpc), associated with a reduction of evoked dopamine release in the striatum. No significant changes were observed in the behavioral domain. We propose our double-hit animal as a reliable model to investigate the mechanisms whereby α-synuclein and inflammation interact to promote neurodegeneration in PD.

Most cases of PD are sporadic.However, genome-wide association studies (GWAS) have identified over 90 genetic risk loci 7 , responsible for an overall heritable component of the disease estimated at around 35% 8 .Missense mutations and overexpression of the α-synuclein gene (Snca), one of the most common mutations in monogenic PD, are widely used to model PD in rodents; however, no single animal model replicates all pathogenic and clinical PD features, and some of them fail to develop nigral DAergic neurodegeneration.Hence, there is common agreement that environmental factors significantly contribute to PD pathogenesis.Indeed, it is worth noting that patients exhibiting the same genetic mutation may not have analogous clinical presentations, suggesting that a complex interaction exists among environmental, age-associated, and genetic factors responsible for the aetiology of the disease 9,10 .
It has been proposed that endotoxins may participate in PD pathogenesis 11 .The lipopolysaccharide (LPS) is the most used endotoxin to elicit an acute immune response.It is a major component of the outer membrane of gram-negative bacteria and a potent inducer of inflammation in peripheral tissues and the central nervous system via activation of toll-like receptor 4 (TLR4).Multiple studies support LPS treatment as an experimental model to reproduce PD-like symptoms in the animal.Progressive DAergic neurons' degeneration and glial cell activation are reported following LPS intraperitoneal (i.p.) injection in wild-type animals [12][13][14][15] .Repeated i.p. injections of LPS over four consecutive days in 3-month-old male mice triggered a reduction in the number of SNpc tyrosine hydroxylase-positive (TH + ) and NeuNpositive neurons 19 days after the first LPS injection 15 ; however, no further cellular reduction was found at 36 days post-injection 16 .The shift in cytokine production, with increased anti-inflammatory and reduced pro-inflammatory cytokines, indicates that, at later stages, animals develop cellular and molecular strategies to stop the ongoing neurodegenerative process 16 .In line with this, TH protein levels decreased in the SN of Wistar rats 15 days after a single dose of LPS, followed by recovery 30 and 60 days post-injection 17 .
In the present investigation, we tested the hypothesis that PD development occurs through a double-hit mechanism involving the combined interplay of elevated endotoxin and aggregated α-synuclein, with the outcome of neuronal degeneration in the SNpc.To this aim, we used rats overexpressing human αsynuclein (Snca +/+ ) characterized by a prodromal asymptomatic phase, with the first functional signs of DAergic pathology starting at four months of age, later resulting in a 25-27% depletion of TH and somadendritic neuronal modification at 12 months, accompanied by altered firing properties of DAergic neurons [18][19][20][21] .In this animal model, we induced systemic inflammation at a presymptomatic age (2 months) via a single intraperitoneal injection of LPS, and we evaluated the effects on neuroinflammation and the DAergic system three months later.

Time course of sickness behavior and body weight
Since LPS stimulates the release of several pro-inflammatory cytokines that are associated with fever, sickness behavior, and body weight loss 22,23 , we first measured sickness signs and body weight over time (2h, 18h, 24h, 42h, 48h, 72h, 96h, and 168h) after LPS i.p. injection (Supplementary Fig. 1A-D).No sickness signs were observed in saline (SAL)-treated WT and Snca +/+ rats except for slightly decreased exploration and locomotor activity in 10% of Snca +/+ rats (score =2) 24h after saline injection.The LPStreated groups exhibited clear signs of sickness compared to controls.The maximum score was reached at 24h, when rats manifested decreased locomotion and explorative behavior, curled body posture, closed eyes, piloerection, and irregular fur (Supplementary Fig. 1 A, B).SAL Snca +/+ rats increased their body weight by 20 g ± 6 g over seven days, whereas LPS injection resulted in acute weight loss.Post hoc analysis demonstrated that LPS Snca +/+ rats lost weight in the first 72h compared with SAL-treated rats, decreasing their body weight by 42 g ± 5 g at 18h and then gradually recovering weight during the following weeks (Supplementary Fig. 1 C, D).

Effects of peripheral LPS on long-lasting neuroinflammation
We next sought to investigate whether a single systemic dose of LPS injection was sufficient to elicit long-lasting inflammation in both SNpc and striatum, by using the sophisticated approach of high dimensional flow cytometry, enabling us to identify all the different brain resident and infiltrated immune cells (Fig. 1A).
We observed that the percentage of CD45 high CD11b high infiltrated monocytes/macrophages was higher in Snca +/+ rats compared with WT rats, both in the SNpc and the striatum.This percentage was even higher in LPS-treated groups, especially Snca +/+ rats (Fig. 1B).
On the other hand, no significant changes were observed in the percentage of the resident myeloid cells of the brain, i.e., CD45 low CD11b + microglia in the LPS-treated groups compared to the SAL-treated group (Fig. 1C).Similar results were obtained with unbiased stereological cell count; no significant increase in the number of IBA1 + cells was found in the SNpc and the striatum of LPS Snca +/+ compared with SAL Snca +/+ rats (Supplementary Fig. 2A-B).
Although microglial cell count did not change, we next sought to further characterize the activation state of CD45 low CD11b + microglia by assessing the expression of pro-inflammatory M1-like markers CD86 and MHC-II (Fig. 1D).In WT rats, CD86 was not significantly altered in the two brain regions at three months after LPS administration (Fig. 1D).MHC-II was significantly increased in the LPS WT group compared to the SAL-treated groups in the SNpc (Fig. 1E, top) but not in the striatum (Fig. 1E, bottom).
On the other hand, expression levels of CD86 were significantly higher in Snca +/+ rats in both SNpc and striatum regardless of the treatment and were even higher upon LPS stimulation (Fig. 1D).In contrast, the levels of MHC-II were increased in the Snca +/+ rats compared to WT rats only in the SNpc (Fig. 1E, top) and such levels were even higher following LPS treatment in both brain areas (Fig. 1E).Since microglia's resting/activation state is associated with its morphological changes, we next performed a Sholl analysis of IBA1 + cells.Notably, microglia displayed a decreased complexity, shorter branching, decreased number of intersections, endings, and nodes in LPS Snca +/+ compared with the SAL-treated group in SNpc and striatum (Supplementary Fig. 2 E-F).
These data proved that a single systemic LPS dose triggers a neuroinflammatory response in the SNpc and the striatum that likely involves microglia activation.

Effects of peripheral LPS on leukocyte infiltration
Given that peripheral-derived immune cells are also present in brain tissue and contribute to disease pathology 24 , we investigated whether a single systemic LPS administration affects peripheral immune cell infiltration into the SNpc and the striatum.To this end, the rest of the CD45 high CD11b low cells (Fig. 1A) were further gated by high dimensional flow cytometry to identify CD3 + T-lymphocytes, CD45RA + B-lymphocytes and CD161 + NK cells.We observed that only infiltrating T-lymphocytes were significantly increased in the SNpc of LPS Snca +/+ rats compared to SAL-and LPS-treated WT rats (Fig. 2A, top).No changes were found in the percentage of infiltrating B-lymphocytes and NK cells (Fig. 2B-C).On the other hand, in the striatum, increased percentages of T-lymphocytes were also observed in WT rats upon LPS treatment (Fig. 2A, bottom), along with a significant increase in infiltrating NK cells in Snca +/+ rats compared with WT rats, but not following LPS treatment (Fig. 2C, bottom panel).
Since we reported an increase in T lymphocytes within the brain of Snca +/+ rats and recent evidence supports a key role for T cells in PD pathogenesis 25 , we further investigated the immunophenotype of peripheral blood mononuclear cells (PBMCs) (Supplementary Fig. 3A) to assess whether differences of immune infiltrates reflect changes in the peripheral blood.No significant changes were observed in the percentage of B cells (Supplementary Fig. 3B), NK cells (Supplementary Fig. 3C), and granulocytes (Supplementary Fig. 3D).At the same time monocytes were higher only in LPS Snca +/+ rats compared to the other experimental groups (Supplementary Fig. 3E).Interestingly, the proportion of T cells was significantly increased in both genotypes following LPS stimulation (Supplementary Fig. 3F).

Effects of peripheral LPS on nigrostriatal alterations in Snca +/+ rats
To test the hypothesis that endotoxin may enhance nigral DAergic neuron degeneration in Snca +/+ rats, we evaluated the immunohistological changes in nigral neurons three months after the injection via the unbiased stereological cell count, by staining midbrain coronal sections with TH as a DAergic marker.
The unbiased stereological count confirmed the absence of TH + neuronal loss in SAL Snca +/+ compared to the SAL WT rats (Fig. 3A), as previously reported 19 .Three months after a single LPS i.p. injection, both WT and Snca +/+ groups showed a statistically significant decreased number of TH + neurons in the SNpc, by almost 28% (±7% SEM) and 49% (±8% SEM) compared to SAL WT rats (Fig. 3A).Since a study by Heo and colleagues 26 demonstrated that some SNpc DAergic neurons could lose their TH expression and become so-called dormant neurons, we tested whether the observed decrease in TH + neurons in LPS-treated rats could reflect TH downregulation, rather than an actual neuronal loss, by using DOPA decarboxylase (DDC) as an additional DAergic marker.We found TH-negative (TH -) neurons that were DDC-positive (DDC + ) mainly in LPS WT but not in the LPS Snca +/+ rats (Fig. 3B), indicating that in LPS WT rats some SNpc DAergic neurons lose their TH-phenotype, and could be surviving or bound to die neurons.In contrast an actual loss of TH + neurons only occurs in LPS Snca +/+ rats.SNpc DA neurons have extensive dendritic arborization that extends ventrally towards the SN pars reticulata (SNpr).In Snca +/+ rats injected with LPS this TH + dendritic projection showed changes in morphology and density.Optical densitometry showed a significant loss of TH + dendrites at three months after LPS injection, compared to SAL-treated groups (Fig. 3C).Dendrites from the remaining DAergic neurons appeared truncated with a distorted morphology (swollen and bubble-shaped) compared to the other experimental groups, where they showed a lengthened and smooth profile (Fig. 3C).
The densitometric analysis of multiple sections of DAergic terminals in the dorsal striatum showed no significant decrease in TH immunoreactivity in LPS-treated rats compared with the SAL-treated groups, despite the decrease of SNpc TH + neurons and altered dendritic arborization (Fig. 4D).
In addition, constant potential amperometry measurements of DA outflow in the dorsolateral striatum revealed a significant decrease of evoked DA amplitude in LPS Snca +/+ rats compared with SAL-treated rats (Fig. 4A), with no significant change in the half-decay time (192.7 ± 1.7 ms vs. 195.9± 1.9 ms, respectively).In previous studies in four-month-old naïve Snca +/+ rats 19 , we found a reduced DA outflow in the dorsal striatum in SAL Snca +/+ compared with WT rats.Here, we found a more significant decrease in DA release when combining genetic and endotoxic risk factors.No significant differences were found in LPS WT rats compared with SAL WT (Fig. 4A) in either amplitude or half-decay.Next, to further explore possible alterations of DA release, we analyzed the modifications of DA transporter (DAT) function in Snca +/+ rats via superfusion of striatal coronal slices with the DAT blocker cocaine.As expected, cocaine (0.3-1 μM) increased the amplitude of the amperometric signal in a concentration-dependent manner and increased its half-decay time 27 .However, we did not find any significant difference between the two experimental groups either at low (0.3 µM) or high (1 µM) concentrations of cocaine (Fig. 4B).
Possible functional alterations of the DAergic transmission were also evaluated by assessing the animal's general locomotor activity and depressive-like behavior via the rotarod and open field tests, and the sucrose preference test, respectively.No significant differences were observed in these behavioral tests in the animals three months after a single LPS dose (Supplementary Fig. 4A-C).Since the two-way ANOVA revealed a significant treatment effect in the sucrose solution intake, we then performed Tukey's multiple comparison post hoc test; however, no difference was pointed out among the four experimental groups (Supplementary Fig. 4C).

DISCUSSION
There is now a wide consensus that PD is a multisystem disorder involving genetic and environmental risk factors.However, understanding the complex mechanisms underlying gene and environment interactions presents a significant challenge.Our present data propose human α-synuclein expressing rats injected with LPS as a reliable model to investigate how these two factors cooperate in the genesis and progression of the neurodegenerative processes in the DAergic system typical of PD, pointing to inflammation as a connective element of this harmful cooperation.
A prodromal symptomatic phase usually characterizes PD animal models based on SNCA gene mutation and overexpression.Previously, we described the human α-synuclein overexpressing rat model, used in this study, showing a presymptomatic phase preceding early pathological signs occurring at four months.
At this age, rats exhibit a significant reduction of striatal DA release despite the integrity of DAergic terminals and in the absence of nigral neuronal loss.These early alterations are paralleled by neuroinflammation in the CNS, with microglial activation and increased levels of pro-inflammatory cytokines in the cerebrospinal fluid (INFγ), while in the blood granulocytes, T, and B cells are not altered 19 .Snca +/+ rats exhibit robust, long-lasting central and peripheral inflammatory activation when combined with an early event of acutely induced inflammation.Most notably, microglia were markedly activated, polarizing towards a more pro-inflammatory phenotype, and accompanied by monocyte and Tlymphocyte infiltration into the brain parenchyma.In contrast, LPS WT rats displayed mild differences from SAL-treated WT, with no macrophagic infiltration and only increased levels of microglia MHC-II pro-inflammatory activation markers in the SN and few infiltrating T-lymphocyte into the striatum.
Interestingly, enhanced nigral microgliosis was also reported in a transgenic mouse model expressing mutated α-synuclein, i.p. injected with LPS at the early symptomatic age 30 .This study shows a delayed chronic and progressive degeneration of nigral TH + neurons, with a more prominent effect five months after LPS injection.The increased infiltration of T-lymphocytes into the brain parenchyma reported in the present study is of particular interest since T cells are known to play a key role in both the CNS and periphery, leading to a profound imbalance of the immune network in PD patients 31 .T lymphocytes were reported in postmortem brain specimens from PD patients 29,30 and a mouse model of PD 32 ; additionally, CD4 + T cell knock-out on a PD mouse model resulted in a markedly reduced DAergic neuronal death 32 .These data strengthen the hypothesis that T cells are crucial for neurodegeneration during PD and support the contribution of autoimmune-based pathogenesis for this disease 25,31,32 .
Since inflammation triggers neurodegeneration, we hypothesized that LPS could also affect DAergic neuron viability.Our findings show that LPS injection in Snca +/+ rats significantly degenerates the DAergic neurons.This was observed at an age (5 months) when these animals do not normally present signs of neurodegenerative processes in the SNpc.While the observed reduction of TH + neurons in both WT and Snca +/+ rats takes place independently from α-synuclein, the decrease of DDC + neurons that only occurred in LPS Snca +/+ rats suggests that the DAergic neuron death only happens in response to a double-hit of genetic human α-synuclein overexpression and endotoxin-induced inflammation.The decreased TH + neurons in LPS WT rats could reflect TH downregulation.Indeed, Heo and colleagues defined neurons without TH as dormant DAergic cells that survive or are bound to die at later stages 26 .
Future studies are needed to assess whether these neurons will die in LPS-injected WT rats at later phases.
It is also worth mentioning that a significant portion of TH -/DCC + neurons was demonstrated in the SNpc of various PD animal models and post-mortem PD brains 26 .Similar results were reported on the parkin - /-mice, another PD model, after six months of repeated low-dose intraperitoneal LPS injections.Both wild-type and parkin -/-mice show substantial TH + neuronal loss, but only in LPS-treated parkin -/-mice there is a significant reduction of NeuN-positive neurons in the SNpc 33 .
Another important feature highlighted in the present study is the dystrophic TH + dendritic arborizations, branching from SNpc surviving DAergic neurons towards the SNpr in LPS Snca +/+ rats, anticipating by almost seven months similar alterations in the naϊve Snca +/+ animals 21 .The loss of neuronal complexity and the decreased dendritic arborization have been linked to α-synuclein overexpression in virally infected DAergic neurons in vitro and in vivo, preceding their eventual death 34,35 and resembling PD patient's postmortem alterations 36 .
The loss of DAergic neurons is associated with a greater extent of reduced evoked striatal DA release in LPS Snca +/+ compared with SAL Snca +/+ rats.However, no significant difference in TH immunoreactivity in the striatum was observed.In other genetic α-synuclein-based models, TH immunoreactivity in the striatum is unaltered at early phases 20,19,37 and in the parkin -/-PD model under repeated LPS administrations 38 .A possible explanation for the lack of reduction in striatal TH immunoreaction despite DAergic neuronal loss in the SNpc could be the presence of compensatory TH upregulation due to the sprouting of the surviving DAergic terminals.Besides T lymphocytes, NK cells are increased in the striatum and this could account for their neuroprotective role in PD that has been proposed based on in vitro and in vivo that suggest they might clear α-synuclein aggregates [39][40][41][42][43] , thus protecting the surviving DAergic terminals, at least in an initial phase of pathology progression.
The increased expression of activations markers on microglia like MHC-II and CD86, which provide respectively the first and costimulatory signals necessary for T cell activation, coupled with the increased percentages of T cells in both brain areas suggest the instauration of an immunological synapse at key areas involved in PD pathology which could drive and sustain neuroinflammation and lead to neurodegeneration.
According to our results, the decrease of DAergic neurons in the SNpc and evoked striatal DA of LPStreated Snca +/+ rats did not alter motor behavior, evaluated by rotarod and open-field tests.This result may not be surprising, because a high degree of neuronal degeneration is required for motor symptoms to be evident, in animals 37 and humans 44,45 .Similarly, the decreased DA release in LPS Snca +/+ rats did not alter the preference for a sweet-tasty solution, commonly used to assess the presence of a depressivelike behavior, such as anhedonia.In a recently published work, reduced sucrose consumption has been reported shortly after systemic LPS administration 46 .However, this condition could reflect a more acute phase in response to the LPS-induced inflammatory activation.

Concluding remarks
Our data support a dual-hit hypothesis for PD, whereby elevated endotoxin under a genetic predisposition may interact or synergize, driving neurodegenerative processes that underlie PD neuropathology.We cannot discern whether the synergy between α-synuclein overexpression and LPS-induced inflammation is limited to an acceleration of PD-like symptoms typically expressed at later stages in Snca +/+ rats or exacerbates their expression.Experiments performed at later stages from LPS injection are needed to clarify this issue.In conclusion, our double-hit model could be relevant to familiar forms of PD, in which the development of a PD profile may depend on the subjective exposure to risk environmental factors through inflammatory responses.More importantly, our findings provide further evidence for the role of the immune system in playing a key role in the onset and progression of PD.

Animals
Homozygous transgenic rats (Sprague-Dawley background) overexpressing the human full-length SNCA locus under the control of the endogenous human regulatory elements (Snca +/+ ) were used 18 .
Two or three rats per cage were housed at standard conditions (23 ± 1°C room temperature, 45-60% relative humidity, 12-h light/dark cycle) with food and water ad libitum.Following species-specific behavior, red plastic tubes and paper are added to the cages.All procedures follow the guidelines on the ethical use of animals from the Council Directive of the European Communities (2010/63/EU) and were approved by the Italian Ministry of Health (Authorization N°617-2019PR).Experimental animals were obtained by crossing heterozygous males with heterozygous female rats and were confirmed as WT or Snca +/+ following genotyping with quantitative PCR using DNA from ear biopsies and the primers for copy numbers of the α-synuclein transgene: SynProm-F: 5′-cgctcgagcggtaggaccgcttgttttagac-3′ and LC-SynPromR: 5′-cctctttc cacgccactatc-3′, normalized to the rat β-actin reference gene with primers: β-actin-F: 5′agccatgtacgtagccatcca-3′ and β-actin-R: 5′-tctccggagtccatcacaatg-3′ 18,19,21 .

Treatment protocol
A single dose of lipopolysaccharide (LPS) from Escherichia coli O111:B4 (Sigma-Aldrich, #L2630) at the concentration of 5mg/kg dissolved in sterile saline (NaCl 0.9%) was intraperitoneally (i.p.) injected in wild-type (WT) and Snca +/+ rats at two months of age 47 .Control groups consist of a single i.p. injection of sterile saline in Snca +/+ rats at the same age.Animals were randomly divided into four experimental groups.
Sickness Behavior and body weight I.p. administration of LPS induces a sickness behavior in mice or rats 22 .In this study, sickness signs consisting of absent exploration and locomotion, curled body posture, irregular fur, piloerection, and closed eyes were evaluated in the LPS-treated and control (saline) group over time after the intraperitoneal injection of LPS, as previously described by 22,23 .Animals were individually placed in a cage and scored on a four-point scale: 0 = no signs, 1 = one sign, 2 = two signs, and 3 = three or more signs.The experimenter quantifying sickness signs was blind to experimental and control conditions.Sickness behavior and body weight were monitored after 2 h, 18h, 24h, 42h, 48h, 72h, 96h, and 168h after the i.p. injection.

Tissue collection
Immunohistochemical and cytofluorimetric analyses were carried out on brain tissues.Animals were transcardially perfused with 1% heparin in 0.1 M sodium phosphate buffer (PB), after deep anesthesia (Rompum; 20mg/ml, 0.5 ml/kg; Bayer, and Zoletil; 100mg/ml, 0.5 ml/kg; Virbac).The brain was taken from the animal's skull, and the two hemispheres were divided into two halves along the medial sagittal line.The left hemisphere was treated for subsequent immunohistological analysis, and the right hemisphere was freshly dissected and processed for high dimensional flow cytometry.The left hemisphere was post-fixed in 4% paraformaldehyde (PFA) in phosphate buffer (PB; 0.1 M, pH 7.4) at 4 °C, and kept for 72 hours in 4% PFA and replaced every 24 hours with a fresh solution.Then, samples were rinsed three times in PB and immersed in 30% sucrose and 10% glycerol solution at 4 °C until sinking for cryoprotection.Afterward, the brains were embedded into OCT, frozen in isopentane, cooled into liquid nitrogen, and stored at -80 °C.Coronal sections (30 μm) were cut from the anterior part of the brain to the midbrain using a cryostat (Leica) at -20°C.The slices were stored at 4°C in 0.1 M PB containing 0.02% sodium azide before being further processed for immunohistochemical staining.

Dissociation of substantia nigra and striatum
After animal sacrifice, substantia nigra and striatum were collected and immersed in D-PBS with high glucose buffer at 4°C.According to the manufacturer's protocol, the tissue was immediately dissociated to single-cell suspension by enzymatic degradation and myelin removal using the adult brain dissociation kit and GentleMACS dissociator (Miltenyi Biotec.).The isolated tissues were placed in C-tubes, and a mix containing Enzyme P and Buffer Z was added to the samples.A second mix containing Enzyme A and Buffer Y is then added, C-tubes tightly closed and attached upside down onto the sleeve of the gentleMACS Octo Dissociator with Heaters (Miltenyi Biotec.).The appropriate program 37C_ABDK_02 was run for 30 minutes, and the c-tube was detached and centrifuged to collect the sample, which was then resuspended and filtered to a 70 um MACS SmartStrainer with 10 ml of cold D-PBS with high glucose.After centrifugation at 300g for 10 minutes, the sample was resuspended in the debris removal solution, gently overlayed in cold D-PBS with high glucose and centrifuged at 3000g for 10 minutes with full acceleration brake.The two top phases were aspirated and discarded, and the sample was resuspended in cold D-PBS with high glucose, centrifuged at 1000g for 10 minutes, and then resuspended in the staining buffer ready f or cell count and flow cytometry staining.

Immunophenotyping of immune cells of substantia nigra and striatum by high dimensional flow cytometry
Cellular phenotypes were assessed using multiparametric flow cytometry panels containing markers to identify cell types and activation states.These markers allowed us to exclude all cells of no interest based on physical parameters (side and forward scatter) and to gate on specific cells of interest.For the immunophenotyping of substantia nigra and striatum, single suspension cells were stained at the cell surface in different panels with PerCP5.Miltenyi Biotec Cat# 130-108-084, RRID:AB_2651886) for 15 minutes at 4°C in the dark.After the staining, cells were washed and resuspended in PBS, ready to be acquired.These markers allowed us to identify the different cell populations correctly.Briefly, cells were identified as CD45+ total leucocytes/resident immune cells.Inside this gate, microglial cells were identified as CD45 low CD11b/c + cells and infiltrated myeloid cells as CD45 high CD11b high , which were subsequently identified as macrophages according to the expression of F4/80.The remaining of CD45 high CD11b low cells were further gated to identify T-lymphocytes as CD3 + , B-lymphocytes as CD45RA + , and NK cells as CD161 + .The expression of CD68 and MHC-II was further assessed inside the microglial cell population.All samples were acquired on a 13-color Cytoflex (Beckman Coulter).For each analysis, at least 0.2x10 6 live cells were acquired by gating on aqua Live/Dead negative cells and then analyzed by the Flowjo analysis software (FlowJo 10.0 (RRID:SCR_008520); Tre Star; https://www.flowjo.com/solutions/flowjo),as reported 48 .

Immunophenotyping of immune cells of the peripheral blood by flow cytometry
Peripheral blood was collected from the animals and whole blood was lyzed with the RBC (1X) solution for 10 minutes at 37°C.Upon lysis, cells were washed and resuspended in cold PBS for immunophenotyping using two different panels.Cells were stained at the cell surface with FITC- Tre Star; https://www.flowjo.com/solutions/flowjo),as reported 19,48,49 .

Immunofluorescence
Sections were first incubated with a blocking solution made with 5% normal donkey serum and 0.3% TritonX-100 (Sigma-Aldrich) in a PB solution for 1h to prevent the non-specific binding of antibodies.
The primary antibodies (TH/DDC, TH/IBA1; listed in Table 1) were incubated over-weekend in PB containing 0.3% Triton X-100 at 4°C and then incubated for 2h at room temperature with the adequate secondary antibodies (listed in table 1).Sections were counterstained with DAPI (4',6-diamidin-2fenilindolo).The specificity of the immunofluorescence labeling was confirmed by omitting primary antibodies and using normal serum instead (negative controls).Confocal laser scanner microscopy (LSM800, Zeiss, Oberkochen, Germany) was used to acquire images.

Stereological cell count
The three-dimensional optical fractionator stereological probe was used to estimate the number of TH + neurons in the SNpc and the Ionized calcium-binding adapter molecule 1-positive (IBA1 + ) cells in the SNpc and striatum.We used the Stereo Investigator System (MicroBright-Field, Vermont, USA).An optical microscope (Axio Imager M2, Zeiss, Oberkochen, Germany) equipped with a motorized stage, a MAC 6000 controller (Ludl Electronic Products, Ltd) and a camera are connected to software Stereoinvestigator 2019.1.3(RRID:SCR_018948; https://www.mbfbioscience.com/stereo-investigator).
The region of interest was defined by TH staining and the area distinction was performed according to the rat brain Paxinos and Franklin's atlas and outlined with a 5x objective.The three-dimensional optical fractionator stereological probe (x, y, z dimension of 50×50×30 μm, with a guard zone of 5 μm along the z axis) was used.Cells were marked with a 40x objective.All quantifications were performed by an investigator blinded to the experimental groups.

Morphological analyses of tyrosine hydroxylase fibers
Three free-floating sections (30 μm) per animal, including the dorsolateral striatum or the SNpr, were treated as for stereological cell count.Sections were photographed with a light microscope (Axio Imager M2, Zeiss, Oberkochen, Germany).Densitometric analysis (OD) of TH + fibers was analyzed using the Java image processing and plugin analysis program in ImageJ (NIH, USA; RRID:SCR_003070; https://imagej.net/).Densitometry values were corrected for non-specific background staining by subtracting densitometric values from the cortex for the striatum and from the surrounding neuropil for SNpr in the same images.All quantifications were performed by an investigator blinded to the experimental groups.

Sholl analysis
IBA1 + cells were imaged with an optical microscope (Axio Imager M2, Zeiss, Oberkochen, Germany) equipped with a motorized stage and a camera connected to Neurolucida 2020.1.2(MicroBright-Field, Vermont, USA; RRID:SCR_001775; http://www.mbfbioscience.com/neurolucida)that allows for quantitative 3D analysis of the entire cell body and branching.Only cells with intact processes unobscured by background labeling or other cells were included in cell reconstruction.The cell body area and perimeter, number of intersections, number of nodes (branch points) and endings, and total length of processes were traced and exported to Neurolucida Explorer 2019.2.1 (MicroBright-Field, Vermont, USA; RRID:SCR_017348; https://www.mbfbioscience.com/neurolucida-explorer).To account for changes in the cell's complexity concerning distance from the soma, concentric rings (radii) were spaced 10 µm apart around the cell, centered at the centroid of the cell body and radiating outward.Branching originating from the soma and the number of branch points and endings, processes intersecting the radii and process length were measured as a function of the distance from the cell soma for each radius.
Overall, fifteen cells per animal were selected randomly for analysis, and all data were subsequently averaged for each rat.All quantifications were performed by an investigator blinded to the experimental groups.

Constant Potential Amperometry
Amperometric detection of electrically evoked DA release was performed in acute brain slices containing the striatum.The carbon fiber electrode was gently positioned into the tissue, and the voltage (MicroC potentiostat, World Precision Instruments) was imposed between the carbon fiber electrode and the Ag/AgCl pellet at 0.60 V.The DA-recording carbon fiber electrode (diameter 30 μm, length 100 μm, World Precision Instruments) was positioned near a bipolar Ni/Cr stimulating electrode to a depth of 50-150 μm into the coronal slice.For stimulation, a single rectangular electrical pulse was applied using a DS3 Stimulator (Digitimer) every 5 min along a range of stimulation intensities (20-1000 μA, 20-80 μs duration).Under stimulation, DA-containing vesicles were released from presynaptic terminals.When DA molecules in the synaptic milieu hit the carbon surface, electrodes were transferred, and a current can be measured.In response to a protocol of increasing stimulation, a plateau of DA release was reached at maximal stimulation (1000 μA, 80 μs).To measure changes in DA synaptic activity caused by drugs affecting the dopamine transporter (DAT) or D23 receptors, we analyzed the effects on evoked DA release by applying cocaine or quinpirole, respectively.Drugs were superfused to the striatal slice when the extracellular DA response to electrical stimulation was stable for four or five successive stimulations.
The evoked release of DA was evaluated following bath perfusion of 0.3 o 1 μM Cocaine dissolved in aCSF 27 .The perfusion was done for 10 minutes when the maximum effect on the DAT by cocaine was reached, and then it was washed out and followed for at least 1 h.The activity of D2 receptors on DA release was evaluated following bath perfusion of 0.03 μM quinpirole, an agonist of the D2 receptor.As for cocaine, quinpirole was perfused for 5 minutes and washed out for 30 minutes.Signals were digitized with Digidata 1440A coupled to a computer running pClamp 10 (Molecular Devices; RRID:SCR_011323; http://www.moleculardevices.com/products/software/pclamp.html).At the end of each experiment, electrode calibration was performed by bath-perfused DA at known concentrations (0.3-10 μM).

Behavioral tests
Three months after the single injection, rats were tested for the Rotarod, open field, and sucrose preference tests.
Animals were habituated for 1 h to the testing room before the beginning of Rotarod and open field tests.
At the end of each session/trial, all apparatus were cleaned with 70% ethanol to remove olfactory cues.

Rotarod test
Motor coordination and balance were tested using an accelerating Rotarod protocol (LE8355, Panlab, Spain).The apparatus consisted of a rod suspended horizontally at 47 cm from the floor.All animals were accustomed to being placed on the rod rotating at low speed (4 rpm) for at least 30 seconds before the first session of the test, which consisted of four sessions.Sessions were performed on two consecutive days (with an inter-session interval of 3.5 h) with an accelerating rod from 4 to 40 rpm in 300 s.Each session consisted of three trials, with at least 5 min of rest between trials.For each trial, the time the rat fell from the rod was recorded (maximum 300 s).The first three sessions are considered pre-training on the Rotarod apparatus to reach a stable performance.The fourth session was considered the final test, and the latency to fall from the rod was analyzed by averaging the three trial records.

Open field test
The open field is a simple test to evaluate general motor and explorative behavior in rodent models of CNS disorders.The apparatus consisted of a cylindrical plastic arena (100 cm in diameter) with dark walls (35 cm high) and floor, placed in a room with dim lighting.During the test, each animal was placed in the center of the arena and freely allowed to explore the apparatus for 10 min.Experiments were recorded with a video camera suspended above the arena, and data were analyzed with EthoVision XT15 (Noldus, The Netherlands; RRID:SCR_000441; https://www.noldus.com/ethovision)video tracking software, and the total distance traveled was measured.