Rit2 silencing in dopamine neurons drives a progressive Parkinsonian phenotype

Parkinson’s disease (PD) is the second most prevalent neurodegenerative disease and arises from dopamine (DA) neuron death selectively in the substantia nigra pars compacta (SNc). Rit2 is a reported PD risk allele, and recent single cell transcriptomic studies identified a major RIT2 cluster in PD DA neurons, potentially linking Rit2 expression anomalies to a PD patient cohort. However, it is still unknown whether Rit2 loss itself is causative for PD or PD-like symptoms. Here we report that conditional Rit2 silencing in mouse DA neurons drove a progressive motor dysfunction that was more rapid in males than females and was rescued at early stages by either inhibiting the DA transporter (DAT) or with L-DOPA treatment. Motor dysfunction was accompanied by decreases in DA release, striatal DA content, phenotypic DAergic markers, and a loss of DA neurons, with increased pSer129-alpha synuclein expression. These results provide the first evidence that Rit2 loss is causal for SNc cell death and a PD-like phenotype, and reveal key sex-specific differences in the response to Rit2 loss.


Introduction
(Supplemental Figure 2A,E,I,M) and females (Supplemental Figure 2C,G,K,O). However, there 117 was a differential impact on gait in males and females following LT Rit2 KD. LT shRit2 males 118 completed significantly fewer gait analysis trials (Supplemental Figure 2B) and had significantly 119 narrower forelimb stride widths (Supplemental Figure 2J), whereas LT shRit2 females had 120 significantly wider hindlimb stride widths (Supplemental Figure 2L). Despite the observed 121 coordination and gait deficits, all LT shRit2 mice also had significantly increased four-limb grip 122 strength (Supplemental Figure 2R,T). Taken together, the behavioral data suggests that DAergic 123 Rit2 is specifically required for male motor learning and that prolonged Rit2 suppression leads 124 to progressive motor coordination and gait deficits in both males and females. 125 126

Short-term DAergic Rit2 silencing blunts DA release in males 127
Motor learning deficits in males in response to ST DAergic Rit2 silencing could be due to altered 128 DA release. To test this possibility, we leveraged fast-scan cyclic voltammetry (FSCV) to 129 measure both DA release and clearance in ex vivo dorsal striatal slices (Figure 2A,B). Given the 130 viability issues inherent to acute brain slices prepared from older animals, we limited our FSCV 131 studies to ST shRit2-and control-injected males, in which we observed a motor learning deficit. 132 We(9) and others(35) previously reported that DRD2 autoreceptors significantly blunt DA 133 transient amplitudes and accelerate DA clearance. Indeed, in control mice DA transient 134 amplitudes were significantly smaller when evoked in ACSF as compared to those evoked in the 135 presence of L-741,626 (25nM, Figure 2C), a DRD2-specific antagonist, as we previously 136 reported(9). In shRit2 mice, DA transient amplitudes recorded in ACSF were not significantly 137 different transients from control mice ( Figure 2C). However, unlike control DA transients, DA 138 amplitudes recorded in the presence of L-741,626 were not significantly greater than those 139 recorded in ACSF, and were significantly smaller than amplitudes recorded in L-741,626 from 140 control mice. Moreover, we previously reported that ST Rit2 silencing in males decreased DAT 141 surface levels by ~50% in dorsal striatum (DS). Despite this reduction in DAT, Rit2 silencing did 142 not significantly affect DA clearance times ( Figure 2D), and we still observed DRD2-mediated 143 enhancement of DA clearance in slices from both control and shRit2 mice, suggesting that DRD2 144 signaling was intact ( Figure 2D). Taken together, these results suggest that males locomotor 145 deficits following ST Rit2 silencing may be, in part, due aberrant DA signaling. 146 147 Rit2 silencing suppresses the DAergic phenotype with earlier manifestation in males than 148 females 149 Given our FSCV results, we hypothesized that motor deficits observed in ST males, and in both 150 males and females following LT Rit2 silencing, could potentially be due to a loss in DAergic tone. 151 To test this possibility, we first measured striatal DA content using mass spectroscopy in male 152 and female dorsal (DS) and ventral (VS) striata following ST and LT Rit2 silencing. In ST shRit2 153 mice, total DA content was not significantly affected in either DS or VS from either male or female 154 mice as compared to their respective controls ( Figure 3A,C). However, DA content was 155 significantly reduced in LT shRit2 male DS ( Figure 3B) and LT shRit2 female VS and DS ( Figure  156 3D) as compared to controls. Importantly, total striatal GABA content was not altered in male or 157 female VS or DS at any timepoint ( Figure 3E-H), demonstrating specific changes in DANs and 158 not global changes in striatal neurotransmitter content. 159 Given Rit2's association with PD and the profound changes motor function and DAergic tone 160 observed with LT Rit2 silencing, we asked whether Rit2 silencing impacted DAN viability. We 161 first assessed DAergic gene and protein expression in isolated ventral midbrain (vMB) and 162 striatum, respectively, following ST and LT Rit2 silencing. In males, RT-qPCR studies revealed 163 that ST Rit2 KD significantly decreased tyrosine hydroxylase (TH) and DAT mRNA in vMB 164 ( Figure 4A,B), and quantitative immunoblotting revealed that striatal TH and DAT protein were 165 also significantly reduced ( Figure 4I We further asked whether Rit2 silencing impacted TH activation, by measuring pSer40-TH via 172 immunoblot. When normalized to actin, pSer40-TH was significantly reduced in ST shRit2 male 173 striatum (Supplemental Figure 3A) and trended to decrease in LT shRit2 males (Supplemental 174 Figure 3B). However, proportion of pSer40-TH to total TH was not significantly different in ST or 175 LT shRit2 male mice as compared to controls (Supplemental Figure 3E,F), suggesting that 176 functional regulation of the TH pool is intact. In females, ST Rit2 silencing had no effect on 177 pSer40-TH or the fraction of pSer40-TH (Supplemental Figure 3C,G), however we detected a 178 drastic loss pSer40-TH loss following LT Rit2 silencing in females (Supplemental Figure 3D), as 179 well as the fraction of pSer40-TH (Supplemental Figure 3H), suggesting the TH is dysregulated 180 in this population. 181 Given the profound losses in TH and DAT expression, as well as DA content, we further asked 182 whether other characteristic ventral midbrain DAergic mRNAs were affected by shRit2. In both 183 ST and LT shRit2 males, DRD2 and Pitx3 mRNA were significantly decreased (Supplemental 184 Figure 4A,B,E,F), and Nurr1 was significantly diminished following LT, but not ST, Rit2 silencing 185 (Supplemental Figure 4I,J). In females, ST Rit2 silencing did not significantly affect DRD2, Pitx3,186 or Nurr1 mRNA levels (Supplemental Figure 4C,G,K). However, by the LT timepoint all three 187 DAergic markers were significantly diminished (Supplemental Figure 4D,H,L). Taken together 188 these data demonstrate that DAergic Rit2 silencing results in downregulation of all DAergic 189 genes consistent with the progressive loss in DAergic tone. 190 We additionally tested whether gene silencing in response to Rit2 KD was specific to DAergic 191 genes or whether pan neuronal and/or ubiquitous genes are also affected by Rit2 silencing 192 (Supplemental Figure 5). We measured vMB expression of the ubiquitously expressed Rit2 193 homolog, Rit1, and Vps35, a core retromer component that is also associated with PD. Given Rit2's association with PD, and the progressive motor dysfunction we observed following 212 Rit2 silencing, we next asked whether Rit2 silencing was accompanied by a change in a-213 synuclein, a major component of Lewy bodies, which are the hallmark of PD neuropathology 214 (ref). To test this, we probed striatal lysates from all mouse cohorts for aSyn and pSer129-aSyn, 215 which is markedly increased in idiopathic PD(36). Total aSyn levels were not significantly 216 affected in either ST or LT shRit2 males ( Figure 6A,B) nor in LT females ( Figure 6D), but were 217 significantly increased in ST shRit2 females ( Figure 6C). Importantly, shRit2 drove a significant 218 increase in pSer129-aSyn in ST and LT shRit2 males ( Figure 6E,F) and in LT females ( Figure  219 6H), and strongly trended for an increase in ST females ( Figure 6F). Together, these data 220 indicates that in addition to profound motor deficits, and DAN degeneration, LT DAergic Rit2 221 silencing drives an increase in modified a-synuclein. 222 223

Male motor learning is rescued with Parkinson's therapeutics. 224
The most widely used treatment strategy for PD is to increase DA availability by providing the 225 DA precursor, L-DOPA. Moreover, recent studies suggest that the DAT inhibitor, 226 methylphenidate (Ritalin), may have therapeutic potential in PD(37,38). We asked whether such 227 pharmacological intervention could rescue the motor deficits observed on the rotarod due to Rit2 228 silencing. We first tested whether increasing extracellular DA levels by inhibiting DAT with 229 methylphenidate could rescue motor learning. ST shRit2 males were assessed on the 230 accelerating rotarod, injected ±methylphenidate (MPH, 5mg/kg, I.P.), and were reassessed 15 231 min post-injection (see schematic, Figure 7A). MPH treatment significantly improved rotarod 232 performance as compared to vehicle-injected mice ( Figure 7B). MPH (aka Ritalin) is a 233 therapeutic psychostimulant that is equipotent at DAT and the norepinephrine transporter (NET) 234 (39). Therefore, to rule out any adrenergic contributions to motor learning rescue, we tested 235 whether rotarod performance was improved with desipramine (DMI), a NET-specific inhibitor. 236 DMI treatment had no significant effect on shRit2 mouse performance ( Figure 7B), suggesting 237 that DAT inhibition was specifically responsible for rescued rotarod performance in ST shRit2 238 males. L-DOPA is the immediate chemical precursor to DA and has been the prevailing 239 treatment for PD since the 1960s(40). Therefore, we next asked whether treating ST and LT 240 shRit2 mice with L-DOPA could rescue motor learning deficits. Mice were assessed on the 241 accelerating rotarod, injected ± L-DOPA (20mg/kg, I.P.), and reassessed 1-hour post-injection. 242 In ST shRit2 mice, L-DOPA robustly and significantly improved rotarod performance ( Figure 7C). 243 However, in LT shRit2 mice, L-DOPA treatment had no effect on rotarod performance ( Figure  244 7D). Taken together, these data demonstrate, that while pharmacological intervention can 245 rescue motor deficits exhibited by ST shRit2 mice, the progressive loss of DAergic tone and 246 DANs caused by LT Rit2 KD drives deficits that are not rescuable by pharmacological means. Rit2 was identified as a PD risk allele in multiple GWAS and, although critical SNPs have been 252 identified, it was unknown whether decreased Rit2 expression itself is detrimental to DAN 253 function and/or viability (15-26). Rit2 mRNA is one of the more highly downregulated genes in 254 postmortem PD patient substantia nigra (31), and defines a specific transcriptomic cluster in 255 single-cell RNAseq studies from post-mortem patients(32). However, whether that diminished 256 Rit2 levels is causal or consequential for PD progression has yet to be determined. Our results 257 demonstrate that prolonged conditional DAergic Rit2 silencing leads to PD-like phenotypes. We Many of these phenotypes we observed significantly affected males at an earlier timepoint than 264 females, consistent with sex-specific differences in PD prevalence and onset. Interestingly, 265 female performance on the accelerating rotarod was completely resistant to Rit2 loss, even at 266 the LT timepoint, despite significant deficits in gait, fix-speed rotarod, and challenge balance 267 Surprisingly, despite coordination deficits, conditional Rit2 silencing did not perturb horizontal 284 locomotion, even with prolonged silencing (Supplemental Figure 1). PD is a late onset 285 neurodegenerative disorder and motor symptoms are often not overtly apparent until >75% of 286 SNc DANs have died (4). Indeed, our stereological data suggest that at the LT timepoint (~25 287 weeks post-injection), there is ~32% loss of TH+ neurons in the SNc ( Figure 5). Thus, it is 288 possible that longer Rit2 silencing would lead to even further losses in the TH+ population, and 289 more pronounced baseline motor deficits. 290 While Rit2 silencing diminished DA neuron numbers, the mechanism(s) downstream of Rit2 loss 291 that lead to decreased DAN viability are unknown. To date, the function of Rit2 in neurons 292 remains poorly defined. Rit2 is required for EGF-and NGF-mediated neurite outgrowth in cell 293 culture models (11-13), and is required for NGF-mediated ERK phosphorylation and cell viability 294 (12,14). Our lab previously reported that Rit2 binds directly to the DAT (10) and is required for 295 both PKC- (8) and mGluR5-mediated (9) DAT internalization in ex vivo striatal slices. Indeed, 296 conditional mGluR5 silencing in DANs blocked DAT internalization, increased DAT plasma 297 membrane presentation, and likewise resulted in male inability to perform on the accelerating 298 rotarod(9). Thus, ST shRit2 effects may be due, in part, to DAT dysregulation, while LT shRit2 299 effects may progressively diminish DAN viability. It is worth noting our previous shRit2 study, in 300 which we found that ST shRit2 differentially modulates acute cocaine locomotor responses 301 based on sex, wherein male shRit2 mice exhibit increased cocaine sensitivity, and females 302 exhibit a loss in cocaine sensitivity (34). Thus, although we only observed motor dysfunction in 303 males following ST Rit2 silencing in our current study, there is clearly still an impact in females 304 that does not manifest as a motor behavior. 305 Despite the observed decreases in striatal DAT protein in ST shRit2 males ( Figure 4J), we did 306 not measure any significant change in DA clearance in parallel FSCV studies ( Figure 2D). 307 Interestingly, we instead observed that Rit2 silencing dampened DA release as compared to 308 controls ( Figure 2C). Decreased DA release was accompanied by decreases in both pSer40-TH 309 (Supplemental Figure 3A) and DRD2 mRNA (Supplemental Figure 4A), raising the possibility 310 that DA synthesis may be altered due to Rit2 silencing. However, DRD2-mediated regulation of 311 DA clearance remained intact following Rit2 KD ( Figure 2D), suggesting that DRD2 312 dysregulation is not likely to mediate the diminished DA release we observed. Rit2 may play 313 some previously undefined role in DA synthesis and/or release that is independent of DRD2 314

regulation. 315
In this study, we assessed vMB expression of five DAergic genes which are critical for DA 316 signaling and gene regulation: TH, DAT, DRD2, Pitx3 and Nurr1. Expression of all of these genes 317 was significantly diminished in male and female LT shRit2 mice and all except Nurr1 were 318 decreased in ST shRit2 males (Supplemental Figure 4). Nurr1 is associated with PD progression 319 and conditional ablation of DAergic Nurr1 results in decreased, DAT and TH expression, reduced 320 striatal DA content and locomotor deficits (45,46). Surprisingly, we saw upregulation in the 321 ubiquitously expressed genes, Rit1 and Vps35 (Supplemental Figure 5). Rit1 is the closest 322 homolog to Rit2 and may be upregulated or stabilized to compensate for Rit2 loss, however, 323 whether they functionally overlap is unknown. We also observed increases in the PD biomarker 324 pSer129-aSyn ( Figure 6). Previous studies have demonstrated that pSer129-aSyn accumulates 325 in DAergic nuclei and negatively regulates Nurr1 expression (47). Whether Rit2 directly regulates 326 DAergic gene expression, or whether the observed changes are consequences of viability and 327 cell death will need to be determined. 328 As discussed above, Rit2 silencing results in hypodopaminergic tone, which likely contributes to 329 observed motor learning and coordination deficits. We used two pharmacological approaches to 330 test whether boosting DA availability could rescue motor dysfunction: DAT inhibition with MPH, 331 and L-DOPA treatment. Both MPH and L-DOPA treatments rescued male accelerating rotarod 332 performance following ST Rit2 KD ( Figure 7B,C), but was unable to rescue performance 333 following LT Rit2 KD ( Figure 7D). Whether L-DOPA is completely ineffective at LT timepoints or 334 whether, as in PD patients, increasing doses are required to overcome DA depletion remains to 335 be tested. 336 Our study, for the first time demonstrates that DAergic Rit2 is required for DA neuron viability

RNA Extraction and RT-qPCR 368
Bilateral 1.0mm 2 tissue punches were obtained from 300µm coronal ventral midbrain slices of 369 experimental mice. Punches were collected while visualizing GFP on an inverted fluorescence 370 microscope and RNA was extracted immediately, or following tissue storage at -70°C, using 371 RNAqueous®-Micro Kit RNA isolation (Thermo Fisher Scientific). Extracted RNA was reverse 372 transcribed using RETROscript® reverse transcription kit (Thermo Fisher Scientific). 373

Quantitative PCR was performed using the Applied Biosystems® 7500 Real-Time PCR System 374
Machine and software or using the Bio-Rad C1000 Touch Thermal Cycler with CFX96 Real-375 Time system and software using Taqman® gene expression assays for mouse Rit2 376 Diego Instruments) as previously described (34). Horizontal, vertical, and fine movements were 384 measured in 5-minute binds for 90 minutes total. 385 Accelerating and Fixed-Speed rotarod: Mice were habituated to the testing room in home cage 386 for >30min with ambient lighting and the rotarod (UgoBasile 47600) running at 4 RPM. 387 Accelerating rotarod: Mice were placed on the rod moving at 4 RPM and rod speed was 388 increased linearly from 4 to 40 RPM over 5 minutes. Trials were terminated and latency 389 determined by either triggering the strike plate during a fall or if the mouse made >1 consecutive 390 passive rotation. Fixed-speed: Mice were placed on the rod moving at the indicated speeds (20,391 25,30,35,40,45 RPM) for 60 second trials. Latency to fall was measured or trial was stopped 392 following >1 passive rotation. Two consecutive trials were performed per speed and latencies 393 were averaged per animal. 394 Challenge Balance Beam: Mice were habituated to the testing room for >30min with overhead 395 lights off and only a single light source placed approximately 1.5 feet over the beam origin 396 illuminated. On day one, mice were trained over 5 trials to traverse a 1.0m, step-wise tapered 397 (widths: 35mm, 25mm, 15mm, 5mm) elevated beam (#80306, Lafayette Neuroscience) at an 398 incline of 15°. A dark box with home-cage bedding was placed at the far end of the beam. On 399 day two, a challenge grid with 1cm x 1cm openings (custom 3D-printed, Thingiverse #4869650) 400 was placed over the beam and mice traversed the beam in 3 independent trials. Traversal 401 initiation and completion were determined by breaking an IR beam at each end of the beam. Tissue harvesting and immunoblotting 420 Striata were collected by preparing 300µm coronal sections on a Vibratome as previously 421 described (8,9). Sections were collected through the entire striatum, dorsal and ventral striata 422 were subdissected, and slices encompassing each region were pooled for each independent 423 mouse. Tissue was lysed in RIPA buffer (10mM Tris, pH 7.4; 150mM NaCl; 1.0mM EDTA; 0.1% 424 SDS, 1% Triton X-100, 1% Na deoxycholate) supplemented with protease inhibitors (1.0mM 425 phenylmethylsulfonyl fluoride and 1.0g/mL each leupeptin, aprotinin, and pepstatin) and 426 Phosphatase inhibitor cocktail V (EMD Millipore). Mechanical tissue disruption was also 427 performed by triturating sequentially through a 200µL pipette tip, 22-, and 26-gauge tech tips 428 and solubilized by rotating (30min 4°C). Insoluble material was removed by centrifugation 429 (15min, 18K x g, 4°C). Lysate protein concentrations were determined by BCA protein assay 430 (Thermo Fisher Scientific). Protein samples were denatured in an equal volume of 2x Laemmli 431 sample buffer and were either rotated (30min, RT) for membrane protein immunoblots or boiled 432 (5min) for soluble protein immunoblots. Proteins were resolved by SDS-Page, transferred to 433 nitrocellulose membranes, and the indicated proteins were detected and quantified by 434 immunoblotting with the following antibodies: rat anti-DAT (MAB369, Millipore; 1:2000), rabbit 435 anti-TH (AB152, Millipore, 1:10000), rabbit anti-pSer40 TH (AB5935, Millipore, 1:5000), rabbit 436 anti-aSyn, rabbit anti-pSer129-aSyn, anti-LRRK2, anti-pSer935 LRRK2, mouse anti-actin 437 (Santa Cruz, 1:5000). Secondary antibodies conjugated to horseradish peroxidase were all from 438 Jackson ImmunoResearch and immunoreactive bands were visualized by chemiluminescence 439 using SuperSignal West Dura (Thermo Scientific). Immunoblotting solutinos were prepared in 440 either PBS-T, or TBS-T (137mM NaCl, 2.7mM KCl, 19mM Tris base, ph7.4, 0.1% Tween20) 441 when probing for phosphoproteins. Non-saturating immunoreactive bands were detected using 442 either VersaDoc 5000MP or Chemidoc imaging stations (Bio-Rad) and were quantified using 443 Quantity One software (Bio-Rad). Representative blots shown for a given condition were 444 cropped from the same exposure of the same immunoblot and spliced together for presentation 445 purposes only. Splice margins are indicated with a line. Brightness and contrast settings were 446 identical for all immunoblot images presented. 447

Fast-Scan Cyclic Voltammetry 448
Mice were sacrificed by cervical dislocation and rapid decapitation. Heads were immediately 449 submerged in ice-cold NMDG cutting solution, pH 7.3-7.4 (20mM HEPES, 2.5mM KCl, 1.25mM 450 NaH2PO4, 30mM NaHCO3, 25mM glucose, 0.5mM CaCl2·4H2O. 10mM MgSO4·7H2O, 92mM N-451 methyl-D-glucamine, 2mM thiourea, 5mM Na + -ascorbate, 3mM Na + -pyruvate). Brains were 452 removed, glued to the stage of a VT1200S Vibroslicer (Leica) and submerged in ice-cold, 453 oxygenated cutting solution. 300µm slices were prepared and were hemisected along the 454 midline prior to recovering in ACSF (125mM NaCl, 2.5mM KCl, 1.24mM NaH2PO4, 26mM 455 NaHCO3, 11mM glucose, 2.4mM CaCl2·4H2O,1.2mM MgCl2·6H2O, pH 7.4) at 31°C for a 456 minimum of 1 hour prior to recording. Hemislices were moved to the recording chamber and 457 were perfused with oxygenated ASCF supplemented with 500µM Na-Ascorbate. Glass pipettes 458 containing a 7µm carbon-fiber microelectrode were prepared and preconditioned in ASCF by 459 applying triangular voltage ramps (−0.4 to +1.2 and back to −0.4 V at 400 V/s), delivered at 60Hz 460 for 1 hour. Recordings were performed at 10Hz. Electrodes were calibrated to a 1µM DA 461 standard prior to recording. Electrodes were positioned in DS and DA transients were electrically 462 evoked with a 250µA rectangular pulse every 2 min, using a concentric bipolar electrode placed 463 ~100µm from the carbon fiber electrode. Data were collected with a 3-electrode headstage, 464 using an EPC10 amplifier (Heka) after low-pass filter at 10 kHz and digitized at 100 kHz, using 465 Patchmaster software (Heka). A stable baseline was achieved after evoking six consecutive DA 466 transients, after which experimental data were collected. Each biological replicate is the average 467 of three evoked DA transients/slice, and a minimum of 3 independent mice were used to gather

Stereological analysis 496
SNc total and TH+ neurons were quantified as previously described(50). Briefly, mice were 497 perfused and fixed with freshly made 4% paraformaldehyde (PFA) in PBS. Brains were removed 498 immediately and fixed again in 4% PFA followed by equilibration in 30% sucrose in PBS. Fixed 499 brains were imbedded in the OCT-compound media (Sakura) and frozen in liquid nitrogen. 40 500 µm cryosections were prepared through the midbrain a Leica CM3050s cryostat, and were 501 stored in an antifreeze media containing 30% ethylene glycol, 25% glycerol, and 5% phosphate 502 buffer. For stereology counting, 1 in every 5 sections was selected with a random start and a 503 total of 6 brain slices on average were used for each mouse for IHC labeling for TH, including 504 DAB enhancement, followed by Cresyl violet staining to reveal all neurons. Substantia nigra pars 505 compacta was imaged using a Zeiss Axioplan 2 microscope equipped with a 20X objective, and 506 Stereo Investigator was used to estimate the total number of neurons in the region of interest 507 using the following parameters: frame sizes, 150 X 150 µm; grid sizes, 250 X 250 µm; top guard 508 zone height, 2 µm; and optical dissector height, 8 µm. These parameters yielded a coefficient of 509 error < 10% throughout the analysis. Total cell numbers measured were weighted to section 510 thickness for each mouse and were averaged across each cohort. Investigators performing 511 stereological counting were blinded to mouse identity. 512

Statistics 513
Data analysis was performed with GraphPad Prism software. All data were assessed for 514 normality and nonparametric tests were applied if data distribution was non-Gaussian. Outliers 515 in each data set were identified using either Grubb's or Rout's outlier tests, with a or Q values 516 set at 0.05 or 5%, respectively, and were removed from further analysis. Significant differences 517 between two values were determined using either a one-tailed, two-tailed, or paired Student's t 518 test, as indicated. Differences amongst more than two conditions were determined using one-519 way or two-way ANOVA, as appropriate, and significant differences among individual values 520 within the group were determined by post-hoc multiple comparison tests, as described for each 521 experiment. 522

Data Availability 523
All data generated or analysed during this study are included in this published article and its 524   were bilaterally injected with either AAV9-TRE-eGFP or AAV9-TRE-shRit2 and brains were 849 fixed, sectioned and stained with cresyl violet and TH-specific antibodies at the LT timepoint. 850 Total and TH+ cell numbers were counted as described in Methods. LT male and female data 851 were pooled. LT Rit2 KD significantly decreased total Nissl+ (A. *p=0.02) and TH+ (B. **p=0.008) 852 cells, but did not significantly affect the %neurons that were TH+ (C. p=0.64) in SNc. Unpaired, 853 two-tailed Student's t test, n=7. 854 855 856