Role of 5-HT1A Receptor in Vilazodone-Mediated Suppression of L-DOPA-Induced Dyskinesia and Increased Responsiveness to Cortical Input in Striatal Medium Spiny Neurons in an Animal Model of Parkinson’s Disease

L-DOPA therapy in Parkinson’s disease (PD) is limited due to emerging L-DOPA-induced dyskinesia. Research has identified abnormal dopamine release from serotonergic (5-HT) terminals contributing to this dyskinesia. Selective serotonin reuptake inhibitors (SSRIs) or 5-HT receptor (5-HTr) agonists can regulate 5-HT activity and attenuate dyskinesia, but they often also produce a loss of the antiparkinsonian efficacy of L-DOPA. We investigated vilazodone, a novel multimodal 5-HT agent with SSRI and 5-HTr1A partial agonist properties, for its potential to reduce dyskinesia without interfering with the prokinetic effects of L-DOPA, and underlying mechanisms. We assessed vilazodone effects on L-DOPA-induced dyskinesia (abnormal involuntary movements, AIMs) and aberrant responsiveness to corticostriatal drive in striatal medium spiny neurons (MSNs) measured with in vivo single-unit extracellular recordings, in the 6-OHDA rat model of PD. Vilazodone (10 mg/kg) suppressed all subtypes (axial, limb, orolingual) of AIMs induced by L-DOPA (5 mg/kg) and the increase in MSN responsiveness to cortical stimulation (shorter spike onset latency). Both the antidyskinetic effects and reversal in MSN excitability by vilazodone were inhibited by the 5-HTr1A antagonist WAY-100635, demonstrating a critical role for 5-HTr1A in these vilazodone actions. Our results indicate that vilazodone may serve as an adjunct therapeutic for reducing dyskinesia in patients with PD.


Introduction
Parkinson's disease (PD) is a devastating neurodegenerative disorder that is caused by a progressive loss of nigrostriatal dopamine (DA) neurons. Current therapeutic approaches typically focus on restoring central DA function, and treatment with the DA precursor levodopa (L-DOPA) remains the most effective pharmacological strategy to alleviate motor symptoms in PD [1][2][3]. However, long term L-DOPA treatment also produces debilitating motor side effects characterized by involuntary movements known as L-DOPA-induced dyskinesia [4][5][6]. In fact, the incidence of L-DOPA-induced dyskinesia is estimated to reach 90% after 10 years of treatment [7][8][9], which significantly reduces the therapeutic window

Evaluation of 6-OHDA Lesion
Stepping tests were performed before and after the 6-OHDA lesion to assess the impact of DA cell loss on forelimb movements. Only animals exhibiting a severe loss of forelimb movements, with a drop from a pre-surgery rate of approximately 11 adjusting steps with the forepaw contralateral to the lesion to three or fewer steps at 4 weeks post-surgery, were considered significantly impaired and were included in this study ( Figure 1A). This approach has previously been shown to predict near-total DA lesions [55][56][57]. Tyrosine hydroxylase (TH) immunohistochemistry and cell counts were performed to confirm the lesions. The number of DA neurons in the substantia nigra pars compacta (SNc) ipsilateral (lesion) and contralateral (intact) to the side of 6-OHDA infusion was determined. Our results show that rats with three or fewer adjusting steps displayed a loss of DA neurons with a range of 87.9-98.6% (mean ± SEM, 93.95 ± 0.72% of intact side; Figure 1B).
Molecules 2021, 26, x FOR PEER REVIEW 3 of 1 attenuated all types of L-DOPA-induced AIMs, as well as the increased MSN responsive ness to cortical drive, and that these effects were dependent on 5-HTr1A activation.

Evaluation of 6-OHDA Lesion
Stepping tests were performed before and after the 6-OHDA lesion to assess the im pact of DA cell loss on forelimb movements. Only animals exhibiting a severe loss of fore limb movements, with a drop from a pre-surgery rate of approximately 11 adjusting step with the forepaw contralateral to the lesion to three or fewer steps at 4 weeks post-surgery were considered significantly impaired and were included in this study ( Figure 1A). Th approach has previously been shown to predict near-total DA lesions [55][56][57]. Tyrosin hydroxylase (TH) immunohistochemistry and cell counts were performed to confirm th lesions. The number of DA neurons in the substantia nigra pars compacta (SNc) ipsilatera (lesion) and contralateral (intact) to the side of 6-OHDA infusion was determined. Ou results show that rats with three or fewer adjusting steps displayed a loss of DA neuron with a range of 87.9-98.6% (mean ± SEM, 93.95 ± 0.72% of intact side; Figure 1B). Forelimb stepping scores (mean ± SEM) in tests performed pre-and four weeks post-surgery. These tests revealed that the 6-OHDA lesion produced a stepping deficit in the forelimb contralateral to the lesion, while stepping with the ipsilateral (intact) forelimb was not affected. The stepping scores given are from the included animals that showed three or fewer adjusting steps with the contralateral forelimb (n = 33). (B) Number of TH-positive cells in the SNc ipsilateral to the lesion, expressed as percentage of the TH-positive cells on the intact side. The included animals displayed a loss of 87.9-98.6% of TH-positive (DA) neurons on the side of the lesion (mean ± SEM 93.95 ± 0.72% of intact side).

Vilazodone Suppresses Established AIMs, but does not Affect Improved Stepping Performance, after L-DOPA Treatment
During the first week of drug treatment, all rats received L-DOPA (6/Veh/LD) an displayed axial, limb, and orolingual AIMs on the side of their body contralateral to th 6-OHDA lesion. These AIMs typically lasted for up to 3 h, with a peak in severity occur ring around 30-90 min after L-DOPA administration ( Figure 2). AIM subtypes wer scored on the last three days of the week, and 3-day averages for each subtype and tota scores are presented ( Figure 2). (A) Forelimb stepping scores (mean ± SEM) in tests performed pre-and four weeks post-surgery. These tests revealed that the 6-OHDA lesion produced a stepping deficit in the forelimb contralateral to the lesion, while stepping with the ipsilateral (intact) forelimb was not affected. The stepping scores given are from the included animals that showed three or fewer adjusting steps with the contralateral forelimb (n = 33). (B) Number of TH-positive cells in the SNc ipsilateral to the lesion, expressed as percentage of the TH-positive cells on the intact side. The included animals displayed a loss of 87.9-98.6% of TH-positive (DA) neurons on the side of the lesion (mean ± SEM, 93.95 ± 0.72% of intact side).

Vilazodone Suppresses Established AIMs, but Does Not Affect Improved Stepping Performance, after L-DOPA Treatment
During the first week of drug treatment, all rats received L-DOPA (6/Veh/LD) and displayed axial, limb, and orolingual AIMs on the side of their body contralateral to the 6-OHDA lesion. These AIMs typically lasted for up to 3 h, with a peak in severity occurring around 30-90 min after L-DOPA administration ( Figure 2). AIM subtypes were scored on the last three days of the week, and 3-day averages for each subtype and total scores are presented ( Figure 2).  The video analysis revealed that vilazodone co-administration in week 2 significantly attenuated the expression of AIMs compared to L-DOPA-only treatment in week 1, for axial, limb and orolingual AIMs (individual animals) (D) and for total AIM scores (mean ± SEM) across 180 min after L-DOPA administration (E). *** p < 0.001, 6/VIL/LD (week 2) vs. 6/Veh/LD (week 1). (F) Forelimb stepping scores after these drug treatments. Scores (mean ± SEM) from 4 weeks after the lesion (before the start of the treatment protocol, "baseline") and 1 h after L-DOPA administration on the second day of treatment week 2 are shown. The forelimb stepping test revealed that vilazodone, while inhibiting L-DOPA-induced AIMs, did not negate the prokinetic effects of L-DOPA in stepping behavior (6/VIL/LD vs. 6/Veh/LD, p > 0.05). ### p < 0.001, vs. ipsilateral (INTACT) side; *** p < 0.001, vs. contralateral (LESION) baseline. In week 2, rats received pretreatment with either vehicle or vilazodone 30 min prior to L-DOPA administration (Figure 2A). In vehicle-pretreated rats (6/Veh/LD, n = 8), AIM scores did not differ between week 2 and week 1, for axial, limb, orolingual, or total AIMs (all Z ≤ −1.26, p > 0.05; Wilcoxon matched-pairs signed-rank test) ( Figure 2B,C). In contrast, vilazodone administration before L-DOPA (6/VIL/LD, n = 14) almost completely suppressed established axial, limb, orolingual, and total AIMs, compared to week 1 (all Z = −3.29, p < 0.001) ( Figure 2D,E).

Effects of Vilazodone and 5-HTr 1A Blockade on Striatal MSN Activity in Dyskinetic DA-Depleted Animals
We used in vivo single-unit extracellular recordings to assess drug effects on cortically evoked activity in MSNs of the sensorimotor striatum ipsilateral to the 6-OHDA lesion ( Figure 4). While the vast majority (≥95%) of striatal neurons are MSNs, we also infrequently encountered fast-spiking interneurons. These interneurons can be distinguished from MSNs by their short onset latency and duration of action potential responses to low stimulus intensities (<0.95 ms), and burst-like activity following cortical stimulation (see [58,59]). Only cells that exhibited an action potential duration of 1 ms or higher following cortical stimulation were included in this study. Tonically active interneurons typically did not respond to our stimulation protocol.
guished from MSNs by their short onset latency and duration of action potential responses to low stimulus intensities (<0.95 ms), and burst-like activity following cortical stimulation (see [58] and [59]). Only cells that exhibited an action potential duration of 1 ms or higher following cortical stimulation were included in this study. Tonically active interneurons typically did not respond to our stimulation protocol.  results. Spike onset latency ( Figure 4B, left) showed a tendency for a main effect of drug treatment (F(2,87) = 2.143, p = 0.12), no main effect of stimulation intensity (F(2,87) = 0.344, p > 0.05) and no significant interaction (F(4,87) = 0.053, p > 0.05). Analysis of spike probability ( Figure 4B, right) revealed a significant main effect of drug treatment (F(2,87) = 3.976, p < 0.05), but no significant main effect of stimulation intensity (F(2,87) = 2.036, p > 0.05) and no significant interaction (F(4,87) = 0.132, p > 0.05). The 6/VIL/LD group showed tendencies for a reduction in onset latency and spike probability compared to the other groups; however, post-hoc tests did not reveal significant differences between individual groups ( Figure 4B).
In cells recorded after drug administration ("challenge"; 6/Veh/LD, n = 21; 6/VIL/LD, n = 19; 6/W/VIL/LD, n = 16), two-factor ANOVAs of stimulation-evoked activity demonstrated the following. Spike onset latency ( Figure 4C Given the effects of the drug treatments on baseline activity (see above), we also expressed activities recorded after the drug challenge relative to baseline values (percent of baseline; Figure 4C, bottom). Two-factor ANOVAs for these data ("challenge", % of baseline) revealed, for spike onset latency ( Figure 4C In summary, consistent with previous findings demonstrating a significant increase in MSN activity in DA-depleted animals following L-DOPA treatment (e.g., [60][61][62]), our analysis showed that L-DOPA-only treatment (6/Veh/LD) reduced the MSN spike onset latency to 75% of baseline values. Vilazodone reversed this L-DOPA-induced facilitation of MSN responses, and this reversal was blocked by the 5-HTr 1A antagonist, WAY-100635.

Discussion
The present study investigated, in the 6-OHDA rat model of PD, the antidyskinetic effects of the multimodal serotonergic drug, vilazodone, the 5-HT receptor subtypes involved, and the electrophysiological correlates in the striatum of these drug actions. Our main results can be summarized as follows. First, our findings confirm and extend previous results by us [54] and others [53], demonstrating a powerful inhibitory effect of vilazodone on the various subtypes of L-DOPA-induced dyskinesia (AIMs) observed in this model. Second, importantly, in contrast to other serotonergic modulatory agents, vilazodone co-Molecules 2021, 26, 5790 9 of 17 administration did not compromise the therapeutic efficacy of L-DOPA, as shown by our outcomes in the forelimb stepping test. Third, also in agreement with previous findings [53], these antidyskinetic effects of vilazodone were blocked by the selective 5-HTr 1A antagonist WAY-100635, demonstrating a critical role for 5-HTr 1A in this vilazodone action. Fourth, in line with the behavioral effects of vilazodone, our in vivo electrophysiological studies revealed that vilazodone prevented the abnormal L-DOPA-induced facilitation of corticostriatal transmission (reflected by a decrease in onset latency of cortically evoked spikes in MSNs), and that these vilazodone effects were also attenuated by blocking 5-HTr 1A . These results complement our previous findings showing that vilazodone suppresses abnormal L-DOPA-induced gene regulation in MSNs in this model [54]. Collectively, these findings indicate that vilazodone co-treatment is capable of "normalizing" aberrant MSN activities and corticostriatal transmission that contribute to L-DOPA-induced dyskinesia, and that 5-HT 1A serotonin receptors mediate these vilazodone effects.

Characterization of Dopamine Lesion
The degree of DA cell loss after 6-OHDA infusion was determined by stereological quantification of the number of TH+ cells in the SNc. Our findings show that the DAdepleted animals that were included in this study, after meeting the inclusion criterion of three or fewer contralateral forelimb steps, had a near-total (average >93%) reduction in the TH+ cell numbers in the SNc ipsilateral to the 6-OHDA infusion. This is consistent with previous findings showing that rats with such a robust deficit in stepping performance had a 90% or greater loss of DA cell bodies in the SN [55,56], or an 80-100% loss of DA tissue content [63] or TH immunoreactivity [54,57] in the ipsilateral striatum.

Vilazodone Attenuates L-DOPA-Induced AIMs, but Does Not Block Prokinetic Effects of L-DOPA
In this study, as in previous studies (e.g., [54,57,64,65]), extensive 6-OHDA-induced striatal DA depletion, followed by repeated daily L-DOPA treatment, produced robust development and expression of AIMs, the rodent equivalent of L-DOPA-induced dyskinesia observed in patients with PD. Our recent work [54,57] demonstrated that L-DOPA given at the relatively low dose of 5 mg/kg once daily for 2-4 weeks was sufficient to induce AIMs in this PD model. Consistent with this finding, in the present study, AIMs emerged as early as one or two days after the first L-DOPA administration, and these AIMs stabilized during the last three days of week 1 and did not further increase between weeks 1 and 2.
A recent study [53] first demonstrated that vilazodone (10 mg/kg), when combined with L-DOPA, significantly reduced established AIMs in 6-OHDA-lesioned rats, an effect we confirmed for our model [54]. In agreement with these outcomes, we here report that vilazodone (10 mg/kg) pretreatment in week 2 almost completely abolished total AIMs as compared to week 1. Moreover, in this study, we provide a detailed analysis of the impact of vilazodone co-administration on the different AIM subtypes (i.e., axial, limb, and orolingual), in addition to the time course of AIM scores across 3 h after L-DOPA administration. Our results demonstrate that all AIM subtypes were dramatically suppressed, with the most robust inhibition seen for axial AIMs and a somewhat lesser effect for limb and orolingual AIMs.
Previous work that assessed SSRIs or 5-HTr 1A agonists as antidyskinetic agents found beneficial effects of those compounds, but also reported potentially problematic side effects, including 5-HT syndrome-like effects and a reduction in L-DOPA-induced motor improvement (e.g., [37,38,42,43,47,66]). In contrast, studies investigating vilazodone did not observe symptoms of 5-HT syndrome, even at higher doses [49,53,67]. Moreover, importantly, our findings demonstrate that vilazodone co-administration, at the present intermediate dose (10 mg/kg), which largely suppressed L-DOPA-induced abnormal gene regulation in striatal MSNs [54], did not interfere with the prokinetic efficacy of L-DOPA, as assessed in the forelimb stepping test (present results; [54]), although higher doses may lose some of this advantage [53].
The loss of the prokinetic effects of L-DOPA (e.g., [38,47]) has been attributed to a strong inhibition of 5-HT neurons innervating the striatum following a high-dose treatment with SSRIs or 5-HTr 1A full agonists, which might lead to a near-complete shutdown in striatal DA release from 5-HT terminals [68]. It can be speculated that the unique pharmacological profile of vilazodone as an SSRI together with its 5-HTr 1A partial agonist property, which is thought to desensitize 5-HTr 1A on the serotonergic cell bodies that regulate the firing activity of these neurons [68][69][70], may account for the efficacy seen with this multimodal agent. The impact of vilazodone may be sufficient to moderate serotonergic activity and DA release from these terminals, thus avoiding abnormal DA spikes in the striatum and their molecular [54] and behavioral (AIMs) consequences, without completely shutting down this DA input, and thus enabling prokinetic effects of L-DOPA.
Consistent with previous work, our results show that chronic L-DOPA treatment produced an increase in MSN responsiveness to cortical stimulation, reflected by a decrease in spike onset latency. Enhanced responsiveness to corticostriatal drive after L-DOPA treatment has been related to aberrant hyperactivation of intracellular signaling pathways in MSNs (e.g., [44,71,72,75,76]). Importantly, in our study, this increase in MSN responsiveness was prevented when vilazodone was combined with L-DOPA treatment, a drug combination that also attenuated abnormal molecular signaling in MSNs [54]. Our findings suggest that vilazodone's modulatory effects on 5-HT neurons produce a tempered DA release from striatal 5-HT terminals, resulting in an attenuation of hyperactive intracellular signaling, thus allowing a "normalization" of neurophysiological (present study) and molecular [54] activities in these neurons (see also [44]), both critical for avoiding L-DOPA-induced motor abnormalities such as dyskinesia (see [54], for discussion). Future work with cell type-specific experimental manipulations will be necessary to provide more detailed insights into the specific cellular mechanisms underlying the vilazodone effects on striatal neuronal activity.

The Effects of Vilazodone Are Mediated by 5-HTr 1A
Previous work first indicated that vilazodone, a 5-HTr 1A partial agonist [50,52], indeed acts via stimulation of 5-HTr 1A to inhibit L-DOPA-induced dyskinesia [53]. Our present results confirm and extend these earlier findings by showing that a selective 5-HTr 1A antagonist (WAY-100635) strongly attenuated the antidyskinetic effects of vilazodone for all subtypes of AIMs. Moreover, we investigated the impact of blocking 5-HTr 1A on the vilazodone effects on cortical stimulation-evoked MSN activities in dyskinetic animals. Our results show that the ability of vilazodone to ameliorate aberrant corticostriatal signaling was inhibited by the 5-HTr 1A antagonist, underscoring the importance of 5-HTr 1A for vilazodone's impact on cellular and behavioral effects. These findings thus provide mechanistic insights into the impact of the serotonergic innervation and 5-HTr 1A on corticostriatal activation of MSNs in this dyskinetic PD rat model.
Many 5-HT receptor subtypes, including 5-HTr 1A , have a fairly wide distribution in the brain [77]. In our study, all drugs, including the 5-HTr 1A antagonist, were given systemically, thus precluding conclusions regarding their specific sites of action (receptor location). However, in line with our reasoning, a recent study reported a near-complete suppression in 5-HT neuron firing following administration of another 5-HTr 1A agonist (±8-OH-DPAT), an effect that was reversed by WAY-100635 [68]. These findings are consistent with an involvement of the 5-HT transmission to the striatum in vilazodone's impact on the cellular and behavioral effects of L-DOPA. Future studies using local drug administration will be necessary to ascertain the role of 5-HTr 1A on 5-HT neurons in these effects.

Conclusions
Our results indicate that vilazodone co-treatment has the ability to "normalize" aberrant corticostriatal transmission and striatal circuit activity following repeated L-DOPA administration via modulating 5-HT activity in a 5-HTr 1A -dependent manner. Vilazodone may help temper L-DOPA-mediated DA input by enabling a more physiological-like release of DA from 5-HT terminals in the striatum. This SSRI/5-HTr 1A partial agonist thus appears to be superior to agents acting as SSRIs only or as 5-HTr 1A full agonists. Future clinical trials will be necessary to confirm vilazodone's potential clinical efficacy.

Animals
Adult male Sprague-Dawley rats (225-249 g upon arrival; Harlan, Indianapolis, IN, USA) were housed 2-3 per cage under standard laboratory conditions (12 h light/dark cycle, lights on at 07:00 h; with ad libitum access to food and water). All procedures met the NIH guidelines for the care and use of laboratory animals and were approved by the Rosalind Franklin University Animal Care and Use Committee (protocol # 17-05; approved on 19 April 2017).
The 6-OHDA lesion was assessed by performing a forelimb stepping test [79] presurgery and then 4 weeks post-surgery. In this test, the rat is held by an experimenter and moved sideways, with its forelimb on the side opposite to the movement direction touching the bench surface. Normally, the rat will perform adjusting steps during this lateral movement, in our settings, typically 10-14 steps [54,55,57]. Following a >90% DA depletion, the number of adjusting steps with the forelimb contralateral to the lesion drops to three steps or fewer, two to four weeks after the 6-OHDA lesion, while stepping with the forelimb ipsilateral to the lesion is unaffected [55,56]. Only rats that displayed a stepping deficit of three or fewer steps with the contralateral forelimb following the 6-OHDA lesion were selected for this study. The lesion was further characterized by measuring TH immunoreactivity (see below).

Behavioral Analysis
On the second day of each treatment week, a stepping test was performed 60 min after L-DOPA treatment. Dyskinesias were assessed during the last three days of each treatment week (Wed-Fri), using an established and well-characterized rat dyskinesia scale to measure AIMs [64,65]. Briefly, rats were individually placed in clear plastic cylinders, and AIMs were videotaped and their frequency and severity scored during a 1-min period at 30-min intervals, 30 to 180 min after L-DOPA injection. AIMs are classified as axial, limb (forelimb), or orolingual. Their frequency was assessed using the following scale: 0 = absent; 1 = occasional (1 to 29 s); 2 = frequent (30 to 59 s); 3 = continuous but interrupted by external sensory stimuli; and 4 = continuous, not interrupted by strong sensory stimuli) [64].
Additionally, the AIM severity (amplitude) was assessed as follows: Axial AIMs (1 = 30 • angle lateral deviation of head and neck; 2 = 30 • < angle ≤ 60 • lateral deviation of head and neck; 3 = 60 • < angle ≤ 90 • lateral deviation of head, neck, and upper trunk; 4 = > 90 • angle torsion of head, neck, and trunk, often causing the rat to lose balance), forelimb AIMs (1 = minor involuntary movements of the distal forelimb; 2 = low amplitude movements causing translocation of both distal and proximal forelimb; 3 = involuntary movements of the whole limb including shoulder muscles; 4 = strong, ballism-like limb and shoulder movements), and orolingual AIMs (1 = involuntary movements of the orofacial muscles with no tongue protrusion; 2 = involuntary movements of the orofacial muscles with tongue protrusion).
Blinded scorers were allowed to give partial scores such as 0.5, 1.5, 2.5, and 3.5 in order to increase the accuracy of AIM ratings. A severity score for each AIM subtype was calculated by multiplying frequency and amplitude scores for each assessment period (i.e., 30, 60, 90, 120, 150, and 180 min), and these values were added for a total AIM score for each subtype. An overall total AIM score was calculated by adding total axial, limb, and orolingual scores.

In Vivo Single-Unit Electrophysiological Recordings
Electrophysiological recordings were performed after the behavioral studies. All animals were maintained on the same daily treatment regimen as in the last week of their behavioral studies and received their last treatment on the day of the recordings. Thus, MSNs were recorded in rats treated with L-DOPA following an injection of either vehicle (6/Veh/LD), vilazodone (6/VIL/LD), or WAY + vilazodone (6/W/VIL/LD). In each group, several MSNs were recorded before the last drug treatment was received to determine "baseline" responses. Responses after the last drug treatment in these animals are designated "challenge" responses; these were recorded 20-180 min after the last drug administration.
Cortically evoked MSN activity was recorded as previously described [55,59,80,81]. Briefly, rats were deeply anesthetized with urethane (1.5 g/kg in physiological saline), and their temperature was maintained at 37 • C using a heating pad. A bipolar cortical stimulation electrode was implanted ipsilateral to the lesion (coordinates, from bregma: AP +3.0 mm, ML -2.5 mm, DV -1.6 mm; [78]) to target the sensorimotor cortex. Cortical local field potentials on the side contralateral to the lesion (AP +3.0 mm, ML +2.5 mm, DV -1.6 mm) were monitored for the presence of slow, large-amplitude waves to ensure that animals were in a deeply anesthetized state during recordings [59]. Recordings began 1 h after electrode implantation.
Microelectrodes for extracellular recordings were manufactured from 2.0 mm outer diameter borosilicate glass capillary tubing (World Precision Instruments, Sarasota, FL, USA) using a vertical micropipette puller (Narishige, Tokyo, Japan). The microelectrode tip was broken to~1 µm in diameter by pushing against a glass rod, and the electrode was filled with 2 M NaCl solution. Striatal MSN activity was recorded ipsilateral to the cortical stimulation (and lesion) at the following coordinates: AP 0.0 to +0.75 mm, ML -3.3 to -3.9 mm, DV -3.0 to -6.5 mm. These coordinates targeted the sensorimotor striatum, where the most robust L-DOPA-induced pathophysiological changes occur [54,57,74].
MSN activity was assessed across stimulation trials (50 pulses/trial) by measuring probability and onset latency of action potentials evoked by cortical stimulation at four different current intensities (400 µA, 600 µA, 800 µA, and 1000 µA in separate trials) as described previously [80,82]. The order of cortical stimulation intensity was counterbalanced between cells (i.e., either 400-1000 or 1000-400 µA). Cortically evoked MSN action potentials were amplified (Neuro Data Instruments, Delaware Water Gap, PA, USA), filtered, digitized via a Digidata 1440a (Molecular Devices, San Jose, CA, USA), acquired using Axoscope software (Molecular Devices), and analyzed using Clampfit 10 software (Molecular Devices). Upon the completion of the experiment, rats were quickly perfused with 4% paraformaldehyde, and their brains were extracted for postmortem assessment of DA cell loss by TH immunohistochemistry staining and DA cell counting in the SNc, using stereological techniques (described below).

Tyrosine Hydroxylase Immunohistochemistry
Rat brains were sliced coronally into 50 µm thick sections, using a sliding microtome (SM2010 R, Leica Microsystems, Wetzler, Germany) as previously described [83]. Sections containing the substantia nigra (from bregma: approximately −4.8 to −6.1 mm) were incubated in rabbit anti-TH antibody (1:500; Pel-Freez Biologicals, Rogers, AR, USA) for 24 h followed by a 2-h incubation with biotinylated goat-anti-rabbit secondary antibody (1:200; Vector Laboratories, Burlingame, CA, USA). Sections were then incubated with avidin/biotinylated complex (ABC; Vector Laboratories), and bound complexes were visualized using 3,3 -diaminobenzidine and hydrogen peroxide tablets as previously described [56]. The number of TH-positive neurons was estimated by stereological means (Stereo Investigator, MBF Biosciences, Williston, VT, USA). Briefly, the SNc region in 6 coronal sections (collected at 200 µm intervals) was carefully outlined under 4× magnification using a rat brain atlas [78]. TH+ cells from the SNc on the ipsilateral (lesioned) and contralateral (intact) sides were counted at 100× magnification. Cells were only included if the nucleus and soma were visible and under focus. The extent of the lesion was then calculated as the number of TH+ cells on the lesioned side relative to that on the intact side.

Statistical Analysis
Statistical analysis was performed using GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA, USA). The differences in AIM scores between treatments in within-subject design experiments were assessed using Wilcoxon matched-pairs signedrank tests, or Friedman ANOVAs, followed by Dunn's post-hoc tests to identify differences between individual treatments. Stepping scores were compared with two-factor ANOVAs with Tukey post-hoc tests. For electrophysiological recordings, the differences in spike probability and onset latency of cortically evoked responses were assessed using two-factor ANOVAs, followed by Tukey post-hoc tests to describe differences between individual groups. Differences were considered significant if p < 0.05. Institutional Review Board Statement: All procedures met the NIH guidelines for the care and use of laboratory animals and were approved by the Rosalind Franklin University Animal Care and Use Committee (protocol # 17-05).

Informed Consent Statement: Not applicable.
Data Availability Statement: The data generated during the current study are available from the corresponding author on reasonable request.