{"paper_id":"2579ee63-fc2b-4fff-897c-eabbd642c783","body_text":"Motor sequence analysis as a sensitive biomarker of dopaminergic \ndegeneration in a non-human primate model of parkinsonism \nLaís Resque Russo Pedrosa1#, Leon Claudio Pinheiro Leal1,2#, José Augusto P. C. \nMuniz1,2, Arthur Gonsales da Silva1, Deiweson Souza-Monteiro3, Rafael Rodrigues \nLima3, Bruno D. Gomes1,4*, Lane V. Krejcová1  \n1 Laboratório de Neurofisiologia Eduardo Oswaldo Cruz, Instituto de Ciências \nBiológicas, Universidade Federal do Pará, Belém, Brasil.  \n2 Centro Nacional de Primatas, Instituto Evandro Chagas, Ananindeua, Brasil. \n3 Laboratório de Biologia Estrutural e Funcional, Instituto de Ciências Biológicas, \nUniversidade Federal do Pará, Belém, Brasil.  \n4 Laboratório de Simulação e Biologia Computacional, Centro de Computação de \nAlto Desempenho, Universidade Federal do Pará, Belém, Brasil.  \n# These authors contributed equally \n * Corresponding author \n \nCorreponding author: \nName: Bruno Duarte Gomes \nAddress: Avenida Perimetral, 2-224, Room 238 \nGuamá, Belém – PA \nBrazil, 66077-830 \nMail: brunodgomes@ufpa.br \n  \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nAbstract  \nParkinson's disease is characterized by progressive dopaminergic degeneration, yet \nmotor symptoms emerge only after substantial neuronal loss - a dissociation that \nchallenges the sensitivity of conventional behavioral endpoints in preclinical models. \nHere, we present a proof -of-principle study establishing a graded hemiparkinsonism \nmodel in adult male capuchin monkeys ( Sapajus apella ) through unilateral, MRI -\nguided stereotaxic injection of 6 -hydroxydopamine into the substantia nigra pars \ncompacta. Three tox in concentrations (4, 10, and 40 mg/mL; n = 3) were tested \nalongside a vehicle -injected sham control (n = 1). Motor function was assessed \nlongitudinally before and after surgery using a three -task battery comprising the \nStaircase test, Tube test, and Brinkman board, capturing complementary dimensions \nof motor functions, including gross lateralization, forelimb use asymmetry, and fine \ndigit coordination. Critically, we introduce a novel sequence -deviation metric applied \nto Brinkman board performance data to quantify disruption in the spatial organization \nof pellet retrieval independently of task success. Post -surgical tyrosine hydroxylase \nimmunohistochemistry combined with optical fractionator stereology revealed \nipsilateral dopaminergic cell losses of 47%, 59%, and 44% relative to the contralateral \nhemisphere across the three treated animals, with the sham showing no meaningful \nhemispheric difference. Behavioral impairments were heterogeneous and strategy -\ndependent: task completion rates were largely preserved, whereas fine motor strategy \nanalysis revealed post-lesion increases in retrieval sequence disorganization in two of \nthree animals. Exploratory regression analyses suggested that strategy -level metrics \nwere more sensitive to nigrostriatal degeneration than global performance measures. \nThese findings demonstrate that capuchin monkeys subjected to unilateral 6 -\nhydroxydopamine lesions reproduce clinically relevant features of hemiparkinsonism \nand that motor sequence analysis constitutes a sensitive readout o f subclinical \ndopaminergic dysfunction, and can outperform conventional performance -based \nmetrics detecting early motor alterations, therefore a potential biomarker of subclinical \ndopaminergic dysfunction, with implications for early detection paradigms in  \nParkinson's disease research.  \nKeywords: Parkinson’s Disease; 6 -hydroxydopamine; Nigrostriatal degeneration; \nmotor sequence analysis; hemiparkinsonism; non-human primates.  \n  \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n1. Introduction \nParkinson's disease (PD) is the second most prevalent neurodegenerative disorder \nglobally, yet its defining motor manifestations emerge only after the nigrostriatal \nsystem has sustained substantial, largely irreversible damage. Approximately 50–60% \nof dopaminergic neurons in the substantia nigra pars compacta and up to 80 –85% of \nstriatal nerve terminals are lost before overt motor dysfunction becomes clinically \napparent - a threshold that reflects the remarkable compensatory plasticity within the \nnigrostriatal system (Braak et al., 2003; Zigmond et al., 1990). Consequently, behavioral \nmeasures based solely on task success or completion rate may systematically \nunderestimate the severity and organization of underlying motor dysfunction, precisely \nbecause preserved performance can mask substantial reorganization of motor control \nstrategies. \nRodent models provide valuable mechanism insights for understanding the neural \nbasis of PD, but the behavioral repertoire is limited for translating many clinically \nrelevant outcome measures (Prasad & Hung, 2020) .Fine motor assessments, upper \nlimb dexterity tasks analogous to human clinical evaluations, and many clinically \nrelevant motor and non-motor PD features cannot be adequately modeled in rodents \nwith the same validity achievable in Non-Human Primates (NHPs) (Bezard et al., 2025). \nAmong NHP species, c apuchin monkeys ( Sapajus apella ) present a particularly \ncompelling model for fine motor investigation. These animals possess a highly \norganized motor cortex with well -defined digit representations, exhibit spontaneous \ntool-use behavior in natural conditions, and have dopaminergic neuron s containing \nneuromelanin, a feature shared with humans and associated with selective \nvulnerability to oxidative stress (Pedrosa et al., 2024) . These features support the \nimplementation of high-resolution fine motor tasks with a high degree of translational \nrelevance. \n6-Hydroxydopamine (6 -OHDA) was the first neurotoxin used to model Parkinson's \ndisease experimentally and remains one of the most mechanistically characterized \ntools for inducing selective dopaminergic loss. Unlike MPTP, which can be \nadministered systemically and exploits species-specific monoamine oxidase B activity, \n6-OHDA does not cross the blood –brain barrier and therefore requires direct \nstereotaxic injection into dopaminergic target structures (Emborg, 2007) . When \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\ndelivered to the substantia nigra (SN), 6 -OHDA is selectively taken up via the \ndopamine transporter and triggers retrograde nigrostriatal degeneration through \noxidative stress and mitochondrial complex I inhibition. Unilateral injection produces \nan asymmet ric hemiparkinsonism (HP) phenotype in which the contralateral \nhemisphere remains intact. This provides a within -subject internal control for both \nbehavioral and histological comparisons, an experimental advantage that bilateral \nmodels and systemic neuroto xins cannot offer. The ability to titrate lesion severity \nthrough concentration adjustments further enables investigation of the relationship \nbetween graded dopaminergic loss and motor impairment across a continuum that \nspans sub-threshold, threshold, and suprathreshold degeneration. \nMotor behavior is inherently multidimensional, encompassing not only the ability to \ncomplete the task successfully but also the way the task has been completed - \ncoordination, sequencing, and spatial organization of movements (Schmidlin et al., \n2011). Dopaminergic degeneration may affect these distinct components differently, \nhence neither reductions in task success nor speed fully capture disruptions in motor \nplanning and control. A comprehensive characterization of motor function, therefore, \nrequires integrating complementary behavioral dimensions, rather than relying on a \nsingle endpoint measure. In this study, we operationalized this multidimensional \nframework using a battery of three complementary motor tasks - the Staircase test, \nthe Tube test, and the Brinkman board - each capturing a distinct behavioral dimension \nranging from gross forelimb lateralization to fine digit coordination and motor sequence \norganization. Critically, a novel sequence -deviation metric specifically designed to \ncapture disruptio ns in motor sequence organization independently of task success \nwas applied to Brinkman board data to quantify disruption in the spatial organization \nof pellet retrieval, independently of total task performance. \nGiven the translational motor phenotype fidelity of capuchin monkeys and the \nmechanistic tractability of focal 6 -OHDA lesions, we aimed to establish a graded HP \nmodel in adult male Sapajus apella using MRI-guided stereotaxic 6-OHDA injections \nat three concentrations (4, 10, and 40 mg/mL) into the SN. This proof-of-principle study \ncombined longitudinal behavioral assessments across three complementary motor \ntasks with Tyrosine Hydroxylase (TH) immu nohistochemistry and stereological \nanalysis through the optic al fractionator technique to quantify dopaminergic neurons \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nin the SN. By integrating measures of limb asymmetry, task latency, grasping strategy, \nand motor sequence organization, we aimed to characterize how distinct dimensions \nof motor performance are differentially affected by graded nigrostriatal degeneration, \nand test the hypothesis that strategy -level metrics are more sensitive than \nconventional performance measures for detecting dopaminergic dysfunction. \n \n2. Methods \n2.1 Animals \nFour adult male capuchin monkeys (Sapajus apella) were obtained from the National \nPrimate Center (Centro Nacional de Primatas, IEC) Ananindeua, Pará, Brazil. The \nanimals had a mean age of 19.66 ± 5.75 years and a mean body weight of 4.30 ± 0.99 \nkg. All animals were clinically healthy with no prior histo ry of neurological disease. \nThey were single-housed in standard cages (2.5 × 2.0 × 2.5 m) under a 12 h light/dark \ncycle. Their diet included laboratory chow specifically formulated for non -human \nprimates, supplemented with fresh fruits and natural juice. Water was available ad \nlibitum. Environmental enrichment was provided in accordance with the International \nGuidelines for non-human primate’s welfare. \nAll experimental procedures were conducted in accordance with Directive 2010/63/EU \nof the European Union and ARRIVE guidelines and were approved by the Ethics \nCommittee for the Use of Animals at the Evandro Chagas Institute (CEUA/IEC), under \nprotocol numbers 45/2016 and 37/2018. \n \n2.2 Behavioral Assessment \nAnimals underwent behavioral assays to evaluate bilateral motor function, including \nboth fine and gross motor coordination, and side dominance. Subjects were tested \nindividually three times per week, with sessions scheduled alternately and always \nconducted prior to daily feeding. Behavioral assessments were conducted both before \n(baseline) and after surgery to analyze changes in motor performance. The three-task \nbattery was the Staircase, Tube, and Brinkman board tasks. All behavioral \nassessments were perfo rmed by the same experimenter in order to minimize inter -\nrater variability. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \n2.2.1 Staircase test. An adapted staircase apparatus for non -human primates was \nused (J. Marshall et al., 2002; J. W. Marshall & Ridley, 2003). It corresponds to the \nValley version of the apparatus, in which a central opening for accessing the reward \nis placed on ascending steps toward the sides. A total of 120 sessions were conducted \nacross pre- and post-surgical phases, each limited to a maximum duration of 3 minutes \nfor reward retrieval. The following parameters were scored for each side: first-retrieval \nlatency, time to retrieve all rewards, number of cross-hand reaches (using the opposite \nhand for a given side), and number of dropped rewards. \n \n2.2.2 Tube test. The tube test was performed using a PVC tube with two upper \nopenings, filled with an edible reward of creamy texture (e.g. peanut butter). Each \nanimal was given 1.5 minutes per session to retrieve the reward using either forelimb. \nTwenty-five sessions were  conducted before and after surgery. The following \nparameters were assessed: time spent using the dominant and non -dominant hand, \nnumber of grasps for each hand, frequency of index finger use, and frequency of use \nof other fingers. \n \n2.2.3 Brinkmann board test. The modified Brinkman board was used to quantitatively \nassess fine motor coordination by measuring the ability to retrieve small pellets \n(Rouiller et al., 1998; Yamanaka et al., 2021). The apparatus consists of an acrylic \nboard (22 x 12 cm) with 50 rectang ular wells —25 arranged vertically and the \nremaining 25 horizontally—each filled with a flavored pellet. \nThe board was positioned at a 45° angle on a wooden table in front of the animal, \nwhich was isolated in the upper compartment of its cage without unilateral restriction. \nNon-human primates were trained to retrieve using a precise pincer grip (thumb \nopposed to index finger). Twenty -five sessions were conducted in both the pre - and \npost-surgical periods. No time limit was imposed for completing the test. Motor \nperformance was evaluated by the number of pellets collected per session using the \ndominant and non-dominant hands, as well as through a quantitative index of motor \nsequence organization adapted from a previous work (Schmidlin et al., 2011). \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \n2.3 Lesions \nAnimals were initially anesthetized with an intramuscular injection of xylazine (100 \nmg/Kg) and ketamine (100 mg/Kg). After animal positioning in the stereotaxic frame, \nanesthesia was maintained with an intravenous infusion of the same agents under \nconstant monitoring by the veterinary staff. \nMRI-guided stereotaxic targeting was performed individually for each subject based \non anatomical landmarks and behavioral lateralization, targeting the dominant side \nidentified during baseline behavioral testing. Subsequent 6 -OHDA injections were \nexecuted using subject-specific MRI data, following the methodology and stereotaxic \ncoordinates described by Pedrosa et al., 2024. \nAnimals in the HP group received unilateral 6 -OHDA injections dissolved in 0.01% \nascorbic acid in saline. A volume of 2 µL was delivered at each of four equidistant sites \nwithin the SN (injection rate: 0.5 µL/min). The SHAM animal received vehicle -only \ninjections (0.01% ascorbic acid in saline) at the same stereotaxic coordinates. The 6-\nOHDA concentrations administered to each animal are provided in Table 1. \n \nSubject Condition 6-OHDA \nConcentration \n(mg/ml) \nLesioned \nside \nAM-BEN HP 4  Left \nAM-BEG HP 10 Right \nAM-AOR HP 40 Left \nAM-AQX SHAM - Left \n \nTABLE 1. The HP animals received different concentrations of 6 -OHDA (4, 10, and \n40 mg/mL) to evaluate potential motor impairments. Lesions were induced in the \ncontralateral hemisphere to the preferred hand assessed previously. The SHAM \nanimal received vehicle injections only (0.01% ascorbic acid).  \n \n2.4 Tyrosine Hydroxylase (TH) Immunohistochemistry  \nAfter the completion of the behavior experiments, the animals were euthanized by \nanesthetic overdose using a combination of ketamine hydrochloride (100 mg/kg) and \nxylazine hydrochloride (50 mg/kg). To fix the nervous tissue, the subjects were \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\ntranscardially perfused with 0.9% saline, followed by 4% paraformaldehyde (PFA). \nBrains were extracted, post-fixed, and coronally sectioned (50µm) with a vibratome. \nFor TH immunohistochemistry, the sections were pre -treated for antigen retrieval \nusing 0.3% boric acid solution and washed three times in PBS for 5 minutes each. \nSubsequently, they were incubated for 20 minutes in 10% serum for blocking. The \nsections were then incubated in a solution containing anti -tyrosine hydroxylase \nprimary antibody (1:1000 in PBS, pH 7.2-7.4) for 72 hours (Anti-TH (Ab-5) Rabbit pA, \nPC38-100ul – Santa Cruz Biotechnology, Inc.). \nAfter incubation, the sections were removed from the primary antibody solution and \nwashed four times in 0.1M phosphate buffer for 5 minutes each. They were then \nincubated overnight with the secondary antibody. The next day, the sections were \nwashed three t imes in 0.1M PBS for 5 minutes each, followed by blocking of \nendogenous peroxidase using 0.3% hydrogen peroxide (H ₂O₂) in 0.1M phosphate \nbuffer for 10 minutes. After three more washes, the sections were incubated in an \navidin-biotin enzyme complex solution (ABC kit, PK-4000, Vector, Burlingame, CA) for \n1 hour. Then, they were rewashed and subjected to the peroxidase detection reaction \nusing DAB (diaminobenzidine – Sigma-Aldrich, Inc) as the chromogen. Finally, the \nsections were washed in a 0.1M phosphate buffer, dehydrated, cleared, and mounted. \n2.5 Microscopic Analysis and Cell Counting Techniques \nDopaminergic neurons in the SN were quantified using the optical fractionator method, \na precise stereological approach for quantifying cell populations that combines the \nfunctionality of an optical dissector and a fractionator (Bonthius et al., 2004; M. J. West, \n1993, 1999). One of the main advantages of this method is its resistance to histological \nalterations such as tissue shrinkage or expansion induced by lesions (M. West et al., \n1991). \nIn the histological sections, we precisely identified the layers of SN, positioning \ncounting probes and capturing digital images. For this, we used a low -magnification \nobjective lens (4×) on a NIKON Eclipse 80i microscope (Nikon, Tokyo, Japan) \nequipped with a motorized stage (MAC200, Ludl Electronic Products, Hawthorne, NY, \nUSA). The system was connected to a computer running StereoInvestigator software \n(MicroBrightField, Williston, VT, USA, \nhttps://www.mbfbioscience.com/products/stereo-investigator, access ed on February \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n15, 2025), allowing the precise digital recording and analysis of the x, y, and z \ncoordinates of the selected points. \nTo ensure accurate neuron identification with the dissector probe, we replaced the \nlow-magnification objective with a high -resolution oil immersion objective (100×, \nNikon, NA 1.3, DF = 0.19 µm). This adjustment enabled unequivocal neuron cell \ncounting. \nMoreover, the thickness of each section was carefully measured at each counting site \nusing the high-resolution objective, allowing precise delimitation of the upper and lower \nplanes. Given the variability in thickness and cell distribution across sections, the total \nnumber of cells of interest was adjusted based on that thickness. Only neuronal cell \nbodies clearly visible within the counting frame or crossing the acceptance line without \ntouching the rejection line were included, following previously establi shed \nmethodological criteria (Gundersen & Jensen, 1987). To ensure comprehensive and \nunbiased sampling, the counting frames were distributed systematically and randomly \nwithin a grid. \nAll sections from the first appearance to the decussation of the SN were used, \ngenerating between 6 and 8 sections with intervals of 6 for each technique and each \nanimal. Depending on the technique, the counting box size ranged from 50 to 150 µm², \nwith 15 to 20 boxes per ROI (region of interest). The optical dissector was adjusted \nbased on the actual section thickness, which could vary depending on the shrinkage \nproduced by each technique. A Gundersen coefficient error < 0.07 was adopted. \nCounting boxes wer e consistently positioned within a grid in a randomized yet \nsystematic manner to ensure comprehensive coverage and unbiased sampling. \nPercent cell loss was computed in percent loss relative to the contralateral side:  \n𝐶𝑜𝑛𝑡𝑟𝑎𝑙𝑎𝑡𝑒𝑟𝑎𝑙−𝐼𝑝𝑠𝑖𝑙𝑎𝑡𝑒𝑟𝑎𝑙\n𝐶𝑜𝑛𝑡𝑟𝑎𝑙𝑎𝑡𝑒𝑟𝑎𝑙  x100 \n \n2.6 Behavioral statistical analysis \n \n2.6.1 Staircase and Tube test analysis \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nData were tested for normality using the D’Agostino –Pearson test. Non -normally \ndistributed data were analyzed with the Wilcoxon signed -rank test for within -subject \ncomparisons and with the Kruskal –Wallis test, followed by Dunn’s post hoc test with \nBonferroni correction, for inter-group comparisons. Data are reported as median and \ninterquartile range. \n \n2.6.2 Brinkman sequence analysis  \nThese data were visualized as heatmaps to assess the consistency of the retrieval \nstrategy across sessions and to evaluate its alteration following the intervention. The \nretrieval order was represented as a session-retrieval-step matrix. It contains the pellet \nnumber retrieved at the corresponding step.  \nTo quantify sequence disorganization, a deviation score was computed for each \nretrieval event as the difference between the retrieved pellet number and the ordinal \nretrieval step. Therefore, the value of zero indicated perfect agreement with the \ncanonical sequence, whereas nonzero values reflected deviations from that order.  \nFor each session, the deviation variance metric was calculated across all retrieval \nevents and used as sequence variability, with higher values indicating greater \nsequence instability. Session-level deviation variance values were compared between \npre- and post-intervention conditions for each animal. Statistical comparisons were \nperformed using the Mann–Whitney U test.  \n \n2.6.3 Behavioral Data Processing and Strategic Analysis \nBehavioral performance across tasks (Staircase success rate, Tube test asymmetry, \nBrinkman board variability) was normalized to each animal’s baseline. Functional \nimpairment was calculated as a percentage change from baseline: \n𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑝𝑟𝑒 − 𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑝𝑜𝑠𝑡\n𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑃𝑟𝑒  x100 \n \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nNegative values indicate animals that maintained or exceeded baseline performance, \npotentially reflecting motor adaptation or task-related learning. \nTo quantify alterations in motor organization, a variability metric was derived from the \nBrinkman task. For each trial, spatial deviation was defined as the difference between \nthe observed retrieval sequence and the canonical board order (positions 1 –50). \nAbsolute deviations were computed to avoid directional cancellation, and the mean \nabsolute deviation was calculated per session. These values were normalized to \nbaseline using the same formula, providing an index of post -lesional increases in \nsequence variability. \n2.7 Correlation with Dopaminergic Degeneration \nDopaminergic degeneration was quantified as the percentage loss of tyrosine \nhydroxylase (TH) -positive neurons in the SN, based on comparisons between \nipsilateral (lesioned) and contralateral (not-lesioned) hemispheres. \nAnimal-level behavioral changes were compared with TH loss using simple linear \nregression. Given the small sample size, analyses were considered exploratory, with \nemphasis on effect sizes and consistency across measures. The coefficient of \ndetermination (R ²) was used to describe the variance explained. An animal -level \nheatmap was generated to summarize behavioral and histological measures. \n \nA schematic representation of the experimental timeline, including behavioral \nassessment and surgical intervention, is shown in Figure 1. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \nFIGURE 1. Experimental timeline for the hemiparkinsonism induction model in \nSapajus apella. Baseline motor assessments (Staircase, Tube, and Brinkman board \ntests) were conducted during the pre -surgical period (approximately 2 weeks before \nsurgery; not drawn to scale). At week 0 (W0), animals underwent MRI -guided \nstereotaxic surgery for unilateral 6-OHDA injection (2 µL at four equidistant SN sites; \ndissolved in 0.01% ascorbic acid saline) or vehicle -only injection (SHAM). Post -\nsurgical motor assessment began at week 2 (W2). Animals were euthanized at week \n4 (W4) for TH immunohistochemistry and optical fractionator stereology. Subjects: HP-\n4 (6 -OHDA 4 mg/mL), HP -10 (6 -OHDA 10 mg/mL), HP -40 (6 -OHDA 40 mg/mL), \nSHAM (vehicle). \n \n3. Results \n \n3.1 Behavioral adaptations reveal preserved performance despite underlying \nmotor deficits \nTo determine whether graded dopaminergic lesions induce detectable motor deficits, \nanimals were assessed using three -task battery before and after 6 -OHDA \nadministration.  \nIn the Staircase test, subjects were required to retrieve ten single-placed rewards. All \nthe animals maintained their consistency in performing the task independently of the \nsurgical condition, which indicates that overall task performance was largely preserved \nacross animals (see Supplementary Figure 1A). However, strategy -level analysis \nrevealed asymmetric motor adaptations, including increased reliance on the non -\ndominant limb and altered retrieval latency patterns following lesion (Figures 2 and 3):  \nBEN HP 4 exhibited a compensatory behavior toward the non -dominant side, \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\ncharacterized by increased first -retrieval latency on the dominant side, decreased \nlatency on the non -dominant side, and a significant post -surgical switch in hand \npreference. In contrast, BEG HP10 did not adopt a comparable compensatory \nstrategy, despite an increased loss of rewards on the dominant side after surgery. \nNonetheless, AOR HP 40 showed clear non -dominant compensation by a significant \nreduction in dominant-side performance. \nThe Tube test probes motor execution under a limited time per session. In this task, \nanimals exhibited post -lesion shifts in forelimb use, characterized by reduced \nengagement of the dominant limb and compensatory recruitment of the non-dominant \nside (Figure 4): BEN HP 4 and BEG HP 10 exhibited such post -surgical strategies, \nwith reduced time spent and fewer grips on the dominant side, concomitant with \nincreased use of the non-dominant side. However, AOR HP40 increased grip counts \non both sides but exhibited a post-surgical bias, spending more time on the dominant \nside. \nCollectively, these findings indicate that 6 -OHDA lesions induce heterogeneous but \nconsistent strategy-level motor adaptations, in which compensatory recruitment of the \nnon-dominant side coexists with subtle signs of bradykinesia on the dominant side. \nImportantly, preserved task completion masked underlying alterations in motor \nstrategy. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \nFIGURE 2. Staircase test: reaction latency and retrieval time. (A) First-retrieval latency \non the dominant side. (B) First -retrieval latency on the non -dominant side. (C) Total \nretrieval time per session on the dominant side. (D) Total retrieval time per session on \nthe non-dominant side. Blue bars: pre -surgical baseline; orange bars: post -surgical \nassessment. Data are presented as mean ± SEM (seconds). P-values from Wilcoxon \nsigned-rank tests (pre vs. post within-subject) are shown below each subject label. N \n= 120 sessions per phase. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \nFIGURE 3. Staircase test: reward loss and cross -hand reaching. (A) Number of \ndropped rewards on the dominant side. (B) Number of dropped rewards on the non -\ndominant side. (C) Number of cross-hand reaches on the dominant side. (D) Number \nof cross-hand reaches on the non-dominant side. Blue bars: pre -surgical baseline; \norange bars: post -surgical assessment. Data are mean ± SEM. P -values from \nWilcoxon signed-rank tests are indicated below each subject label. Note: the SHAM \nanimal exhibited a significant post -surgical increase in dominant -side cross -hand \nreaches (panel C, p=0.008), which may reflect a non -specific effect of the surgical \nprocedure independent of dopaminergic lesion. \n \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \nFIGURE 4. Tube test: forelimb grasping and duration. (A) Number of grasps using the \ndominant hand per session. (B) Number of grasps using the non -dominant hand per \nsession. (C) Total time (seconds) spent using the dominant hand per session. (D) Total \ntime (seconds) spent using the non -dominant hand per session. Blue bars: pre -\nsurgical baseline; orange bars: post -surgical assessment. Data are mean ± SEM. P -\nvalues from Wilcoxon signed -rank tests are indicated below each subject label. N = \n25 sessions per phase. \n \n \nFine motor performance was further evaluated using the Brinkman board task, with \nparticular emphasis on motor sequence organization. For the total score in the \nBrinkman board task, Wilcoxon analyses revealed significant pre –post differences \nwithin the experimental groups (Supplementary Figure 1B). While overall performance \nshowed limited changes across animals, sequence -level analysis revealed marked \nchanges in motor organization after lesion: significant changes were observed in the \nBEG HP 10 and AOR HP 40  for both right and left sides, indicating a clear effect of \nthe intervention in these groups. No significant pre–post differences were found in the \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nBEN and SHAM groups (p > 0.05). Additionally, right–left comparisons demonstrated \nsignificant asymmetry in the post-intervention condition for the BEG HP 10 and AOR \nHP 40 (p < 0.05). In contrast, no significant differences between sides were observed \nin the BEN and SHAM (p > 0.05). \nFine motor function was assessed using the Brinkman board task, in which the \nsequence of pellet retrieval across 50 board positions was recorded for each session. \nSessions were categorized into pre - and post-intervention phases according to the \nexperimental design. Sequence variability differed across individuals, with some \nanimals exhibiting increased post -lesion disorganization, while others displayed high \nbaseline variability: BEG HP 10 showed greater variance in deviation than the other \nsubjects, suggesting that this animal adopted a distinct retrieval strategy at baseline. \nComparisons between pre- and post-intervention phases were significant for BEN HP \n4 and AOR HP 40, but not for BEG HP 10 (Figure 5A). \nTo visualize individual pellet retrieval strategies within each session, heatmaps were \ngenerated to represent the spatial distribution of retrieval order across board positions, \nrelative to the default task strategy. More organized performance was characte rized \nby smoother spatial gradients and localized clusters of similar values, indicating a \nconsistent progression of pellet retrieval across neighboring slots.  \nHeatmap visualization revealed a clear disruption of spatial retrieval organization \nfollowing lesion, with loss of structured sequential patterns in affected animals: In BEN \nHP 4 and AOR HP 40, post -lesion heatmaps became less spatially coherent, with a \nmore irregular and fragmented distribution of retrieval order, consistent with reduced \nmotor organization after intervention. In contrast, BEG HP 10 showed a relatively \ndispersed and weakly organized pattern already during the pre -intervention phase, \nwith only minor post-intervention changes after intervention, consistent with its higher \ndeviation variance (Figure 5B) \n \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \nFIGURE 5. Brinkman board: motor sequence analysis. (A) Left: schematic of the 50 -\nslot Brinkman board (22 × 12 cm acrylic board; 25 vertically and 25 horizontally \noriented wells). Right: session -level deviation variance (mean ± SEM) before (blue) \nand after (orange) s urgical intervention. Deviation variance quantifies within -session \ndeviations between actual and expected retrieval order (see Methods 2.7); higher \nvalues indicate greater sequence disorganization. P -values from Mann -Whitney U \ntests (pre vs. post ) are shown below each subject label. (B) Retrieval sequence \nheatmaps for the dominant hand across pre - (left) and post -lesion (right) sessions. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nEach row represents one session; each column represents an ordinal retrieval step \n(1–50). Color encodes the pellet number retrieved at each step (color scale: blue = \nlow pellet number, red = high pellet number; range 10–50). The white dashed diagonal \nrepresents the ideal sequential strategy, where pellet number equals retrieval order. \nOrganized performance appears as a smooth blue -to-red diagonal gradient; \ndisorganization manifests as color fragmentation across the heatmap. N = 25 sessions \nper phase. Note: BEG HP10 showed high baseline sequence variability in both \nphases; white cells in AOR HP40 post-lesion indicate excluded sessions.  \n \n3.2 Reduction of dopaminergic neurons in 6-OHDA Hemiparkinsonian non-\nhuman primates \nAfter behavioral assessment, TH staining was performed to evaluate the integrity of \ndopaminergic neurons in the SN. Using stereology and the mean section thickness \nestimator, we observed that unilateral 6 -OHDA injections resulted in substantial and \nasymmetric dopaminergic cell loss across all treated animals, ranging from \napproximately 44% to 59% relative to the contralateral hemisphere.   BEN HP 4 \nshowed 150,022.61 TH+ cells bodies contralaterally versus 80,080.39 ipsilaterally \n(46.6% reduction). BEG HP 10 showed 89,326.05 contralaterally versus 36,759.49 \nipsilaterally (58.8% reduction). The AOR HP 40 showed 105,697.83 contralaterally \nversus 59,488.58 ipsilaterally (43.7% reduction). No significant hemispheric difference \nwas observed in the SHAM animal, confirming the specificity of the lesion (110,176.22 \ncontralateral vs 110,334.87 ipsilateral; −0.1%), see Figure 6. \n \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n \nFigure 6.  Unilateral 6-OHDA injection induces ipsilateral dopaminergic cell loss in the \nsubstantia nigra of adult Sapajus apella. Left: representative TH photomicrographs of \nthe ipsilateral (lesioned) and contralateral (non -lesioned) SN from coronal vibratome \nsections (50 µm). Images were acquired at 2.5× magnification. Scale bar = 500 µm \n(shown in A; applies to all photomicrographs). Rows correspond to individual subjects: \n(A) BEN HP4 (6 -OHDA, 4 mg/mL); (B) BEG HP10 (6 -OHDA, 10 mg/mL); (C) AOR \nHP40 (6-OHDA, 40 mg/mL); (D) SHAM (vehicle-injected control, 0.01% ascorbic acid \nin saline). Right: subject-level estimates of the total num ber of TH+ cell bodies in the \nSN, obtained by the optical fractionator method with actual measured section \nthickness (Gundersen CE ≤ 0.07; for full stereological parameters see Methods). Black \nfilled circles = contralateral hemisphere; red filled circles =  ipsilateral hemisphere; \nconnecting lines indicate the direction and magnitude of interhemispheric change. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nPercentage values indicate ipsilateral cell loss relative to the contralateral hemisphere \n[(contralateral − ipsilateral)/contralateral × 100]. TH, tyrosine hydroxylase; SN, \nsubstantia nigra; HP, hemiparkinsonism. \n \n3.3 Exploratory association between behavioral impairment and histological \nlesion severity  \nTo explore the relationship between dopaminergic degeneration and behavioral \noutcomes, we compared histological and motor behavior metrics across subjects (see \nthe graphs in Figure 7). As previously mentioned, conventional performance -based \nmetrics for mot or assessment (e.g. score, time) showed limited association with \ndopaminergic cell loss, therefore is not sufficient for assessing complex motor \nimpairment (Braak et al., 2003), seen in Staircase Impairment, and Tube Asymmetry. \nOn the other hand, the Strat egy Chaoticity metric revealed substantial changes in \nmotor organization.  \nThe Animal -level summary can be interpreted as an exploratory multidimensional \nguide in PD models (Figure 7D). Changes in Tube Asymmetry can be interpreted as \nimproved motor skills (AMAQX -SHAM), or compensatory nigrostriatal degeneration \nstrategy (AMBEN, and AMBEG). It is important to note that sequence -based metrics \nderived from the Brinkman task showed greater sensitivity to nigrostriatal \ndegeneration.  \n \n \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nFigure 7. Exploratory association between behavioral impairment and nigrostriatal \ndegeneration. Scatter plots (panels A –C) show the linear regression between SN \ntyrosine hydroxylase (TH) -positive cell loss (x -axis; %) and baseline -normalized \nchange in behavioral met rics (y -axis; %), individually for each subject. Regression \nlines (red dashed), 95% confidence intervals (pink shading), R² values, and p -values \nare displayed within each panel. Given the small sample size (n = 4), analyses are \nstrictly explorat ory; effect sizes and directional consistency rather than statistical \nsignificance are the primary interpretive focus. (A) Staircase Impairment: baseline -\nnormalized change in reward retrieval success rate on the dominant side (R² = 0.48, \np = 0.307). A moderate, non-significant negative trend suggests that greater TH loss \nis associated with greater staircase impairment, although the relationship is not linear \nacross the concentration range tested. (B) Tube Asymmetry: baseline -normalized \nchange in forelimb use asymmetry between the dominant and non-dominant sides (R² \n= 0.01, p = 0.889). No meaningful linear association was observed between TH loss \nand tube asymmetry, consistent with the interpretation that forelimb asymmetry \nreflects individual compensatory st rategies rather than degeneration severity per se. \n(C) Strategy Chaoticity (Brinkman board deviation variability): baseline -normalized \nchange in session-level deviation variance (R² = 0.18, p = 0.571). The absence of a \nclear linear relationship is consiste nt with the heterogeneous nature of motor \nsequence reorganization across subjects, particularly given BEG HP10's high pre -\nlesion baseline variability. (D) Animal -level integrated summary heatmap. Each row \nrepresents one subject (AMBEN = BEN HP4; AMBEG = BE G HP10; AOR = AOR \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nHP40; AMAQX = SHAM); each column represents one metric: SN TH Loss (%), \nStaircase Impairment (baseline -normalized change %), Tube Asymmetry (baseline -\nnormalized change %), and Strategy Chaoticity/Variability (baseline -normalized \nchange %). Color scale refl ects the magnitude of each value (darker red = higher \nabsolute value; lighter yellow = values near zero or negative change). Numerical \nvalues are displayed within cells. Negative values in behavioral metrics indicate \nmaintained or improved performance rela tive to baseline, potentially reflecting motor \nlearning, task adaptation, or compensatory dopaminergic upregulation. The heatmap \nis intended as a qualitative, multidimensional overview to support hypothesis \ngeneration rather than statistical inference.  \n \n4. Discussion \n \n4.1 Proof-of-principle for a precise 6-OHDA model of Parkinsonism in NHP \nThis study establishes a graded non-human primate model of hemiparkinsonism and \ndemonstrates that motor sequence analysis provides a sensitive readout of \ndopaminergic dysfunction. There is a need for standardized model systems capable \nof capturing the heterogeneity of Parkinson’s disease. Current preclinical approaches \nhave relied heavily on rodent models, particularly the 6 -OHDA-lesioned rat and the \nMPTP-treated mouse, as well as transgenic lines overexpressing wild-type or mutant \nα-synuclein, or carrying loss -of-function mutations in PD -associated genes (LRRK2, \nPINK1, Parkin, DJ -1) (Emborg, 2007; Zhang et al., 2025). Although rodent models \nhave been instrumental for  mechanistic insights, their limited behavioral repertoire \nconstrains the assessment of complex motor functions relevant to Parkinson’s \nDisease. Several therapeutic candidates for PD —including trophic factors, glutamate \nreceptor modulators, anti -dyskinetic agents, and cell -based therapies —have \ndemonstrated efficacy in rodent models but have shown limited success in Phase II or \nIII clinical trials (Bezard et al., 2025; Vermilyea & Emborg, 2018).   \nIn this context, non -human primates offer a biologically and functionally relevant \nmodel, particularly due to the presence of neuromelanin -containing dopaminergic \nneurons, and a highly developed motor system, thereby more closely reflecting human \ncellular vulnerability and therapeutic response. In addition, the well -developed \nprefrontal cortex in primates —disproportionately expanded relative to rodents —\nsupports higher -order cognitive functions, including executive control, working \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nmemory, social behavior, and decision-making. This complexity enables NHP models \nto capture both motor and non -motor dimensions of PD, including cognitive and \nneuropsychiatric manifestations.  \nThe availability of capuchin monkey ( Sapajus apella ) colonies in Brazil reduces \nlogistical and financial constraints related to breeding and regulatory compliance. The \nspecies exhibits a complex motor system, including the ability to use objects from the \nenvironment as tools. Consistent with this behaviora l repertoire, the motor cortex is \nhighly organized, with well-defined representations of individual digits that support fine \nmotor control. Together, these anatomical and behavioral features enable \nsophisticated motor performance, as demonstrated in tasks such as the Brinkman \nboard (see Figure 5). \nThe unilateral model provides the advantage of an internal control, as the contralateral \nhemisphere serves as a reference within each animal. Based on internal control, TH \nhistological investigation permitted a link between the percentage of neuronal loss and \nmotor behavioral impairments. Performing three different 6 -OHDA concentrations (4, \n10, and 40 mg/ml), the observed reduction losses were 47%, 59%, and 43%, \nrespectively. Overall, the motor assessment revealed dominant -side motor \ndysfunction across all tasks, manifested as either increased time to retrieve the reward \n(Figure 2A) and/or loss of the reward immediately after retrieval (Figure 3A). \nImportantly, a compensatory strategy was observed in the animals showing the \ngreatest TH reduction (Figure 3D). Furthermore, motor deficits were not fully captured \nby conventional performance measures, reinforcing the need for more sensitive \nbehavioral metrics. \n \n4.2 Behavioral deficits were heterogeneous and captured clinically relevant \nfeatures of Parkinsonian motor dysfunction \nThe results demonstrate marked heterogeneity in behavioral deficits following \nunilateral 6-OHDA induction, despite the use of a standardized surgical protocol. This \nvariability should not be interpreted as experimental noise, but rather as a biologically \nmeaningful feature of the neurobiological response to dopaminergic depletion. In NHP \nmodels, the response to partial dopaminergic lesions is well recognized as non -\nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nuniform, reflecting inter -individual differences in motor compensatory mechanisms \n(Emborg, 2007; Teil et al., 2021). \nDopaminergic degeneration does not translate linearly and into motor impairment, as \ndifferent neuronal subpopulations exhibit selective vulnerability, while pre - and \npostsynaptic compensatory mechanisms may sustain motor function even in the \npresence of substantial degeneration (Bezard & Gross, 1998; Brotchie & Fitzer-Attas, \n2009; Kish et al., 1988; McGregor & Nelson, 2019; Santana -Román et al., 2025) . \nNigrostriatal dopaminergic neurons exhibit selective and heterogeneous vulnerability, \ninvolving subpopulations with distinct molecular profiles and differential susceptibility \nto oxidative stress and mitochondrial dysfunction ( Blesa & Przedborski, 2014; \nLebowitz & Khoshbouei, 2020a) . This non -uniform pattern of degeneration directly \ncontributes to the functional variability observed across individuals under similar \nexperimental conditions. \nIn this context, the observed heterogeneity may be interpreted as the expression of \ndistinct compensatory strategies that vary in efficiency and organization across \nindividuals, as shown in Figures 3C and 3D. Evidence indicates that the motor system \ncan sustain function through pre - and postsynaptic adaptations, including increased \nefficiency of residual dopaminergic signaling, modulation of neurotransmitter release, \nand reorganization of basal ganglia circuits(Berardelli et al., 2001; Bezard & Gross, \n1998; Brotchie & Fitzer -Attas, 2009; Perez et al., 2008). However, this adaptive \ncapacity is not uniform, resulting in distinct behavioral patterns even under controlled \nexperimental conditions. \nSupporting this perspective, both classical and contemporary studies demonstrate that \nmotor symptoms in Parkinson’s disease become clinically apparent only after \nsubstantial striatal dopamine loss, typically exceeding 70%, indicating the existence \nof a fun ctional compensatory phase (Kish et al., 1988; Nutt et al., 2004). Preserved \nperformance in global motor tasks can mask substantial alterations in motor \norganization and motor strategies, as task execution is maintained through \ncompensatory adjustments that are not captured by conventional performance metrics \n(Espay et al., 2019). \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nPreserved task execution does not imply the absence of motor deficits as seen in \nSupplementary Material 1 but rather reflects behavioral adaptations that partially \ncompensate for dopaminergic dysfunction (Perez et al., 2008; Przedborski, 2017). \nThis finding reinforces the importance of analytical approaches that consider not only \nfinal performance outcomes but also movement organization and execution patterns, \nwhich are more sensitive to the detection of early motor deficits (Espay et al., 2019). \nTaken together, these findings support the notion that inter -individual variability is an \ninherent feature of hemiparkinsonian models in non-human primates and represents, \nin fact, a translational advantage. PD itself is characterized by substantial \nheterogeneity in both dopaminergic loss and clinical manifestation, driven by \ndifferences in neuronal vulnerability, molecular factors, and circuit-level reorganization \n(Blesa & Przedborski, 2014; Santana-Román et al., 2025). Hence, the results suggest \nthat the experimental model employed may capture this variability, representing a \npromising approach for investigating neural compensation mechanisms and for \ndeveloping therapeutic strategies targeting not only neuroprotection but also the \nmodulation of adaptive motor processes. \n \n4.3 Brinkman board performance provided a sensitive readout of fine motor \nimpairment \nThe Brinkman board task provided the most sensitive readout of motor impairment in \nthis study, as it constituted a particularly sensitive measure for detecting fine motor \ndeficits, especially because it enabled the analysis not only of overall task outcome  \nbut also of spatiotemporal organization of motor execution. Unlike conventional \nperformance measures, based solely on the total number of pellets retrieved, the \nmodified Brinkman board has been widely used as a quantitative tool for assessing \nmanual dexterity, requiring precise grip control, digital coordination, and sequential \norganization of actions (Savidan et al., 2017; Schmidlin et al., 2011). Furthermore, \nsequence analysis of pellet retrieval, particularly when represented through heatmaps, \nenhances the interpretative capacity of the task by revealing disruptions in motor \norganization even when task success was preserved (Figure 5B). \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nThe Brinkman board allows the dissociation of different components of fine motor \ncontrol, including precision, coordination, and sequential organization, and is capable \nof identifying alterations in execution strategy even in the absence of evident \nquantitative deficits (Badoud et al., 2016; Hoogewoud et al., 2013; Maetzler et al., \n2024) . In experimental models involving NHP, the modified Brinkman board has been \nwidely employed not only as a quantitative measure of manual dexterity but also as a \nsensitive tool for investigating functional recovery and motor reorganization following \ncentral nervous system lesions(Badoud et al., 2016; Darling et al., 2018; Hoogewoud \net al., 2013; Lebowitz & Khoshbouei, 2020b; Nudo, 2013; Savidan et al., 2017; \nSchmidlin et al., 2011). In this regard, evidence suggests that recovery of performance \nin manual dexterity tasks does not necessarily correspond to restoration of the original \nmotor pattern but may instead occur through the adoption of alternative compensatory \nstrategies (Hoogewoud et al., 2013; Schmidlin et al., 2011). \nA key finding of this study is the dissociation between preserved task performance and \ndisrupted motor sequence organization. Despite the maintenance of task execution in \nparadigms such as the Staircase test, detailed analysis of the Brinkman task revealed \nconsistent reorganization of motor sequences, suggesting that compensatory \nmechanisms sustain performance at the expense of movement efficiency and \nstructural organization. This pattern is consistent with the literature on dopaminergic \ncompensation, in which pre- and postsynaptic adaptations, as well as reorganization \nof basal ganglia circuits, allow preservation of motor function despite significant \nneuronal loss (Berardelli et al., 2001; Brotchie & Fitzer -Attas, 2009; McGregor & \nNelson, 2019; Perez et al., 2008). Thus, preserved global performance does not imply \nabsence of motor deficit, but rather reflects behavioral adaptations that mask \nunderlying alterations in movement organization. \nFrom a translational perspective, findings related to fine motor coordination are \nparticularly relevant, as accumulating evidence indicates that subtle motor alterations \nemerge in very early stages of PD, even before clinical diagnosis. Studies have shown \nthat fine motor deficits can be detected years before the onset of classic symptoms, \nparticularly involving alterations in finger tapping, handwriting, and manual \ncoordination (Panyakaew et al., 2023; Schaeffer et al., 2024). Consistently, high -\nresolution digital approaches, such as touchscreen typing analysis, have \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\ndemonstrated the ability to detect subtle motor impairments associated with \nParkinson’s disease, even in early or at-risk stages, revealing alterations in movement \nkinematics and variability (Iakovakis et al., 2020). Furthermore, deficits in fine motor \norganization and movement sequencing are already present in the prodromal stage \nand can be detected through sensitive quantitative metrics even in the absence of \novert functional impairment (Yang et al., 2025). \nTherefore, the ability of this task to detect alterations in motor sequence organization, \neven in the absence of deficits in global performance, closely parallels the pattern \nobserved in early stages of Parkinson’s disease in humans, in which task execution is \npreserved but movement kinematics, timing, and efficiency are already impaired \n(Berardelli et al., 2001; Espay et al., 2019).Taken together, these findings reinforce \nthe value of the Brinkman board as a sensitive tool for detecting nigrostriatal \ndysfunction in intermediate or subclinical stages of dopaminergic degeneration, \ncontributing to the development of experimental approaches with greater translational \nrelevance. \n \n \n4.4 Study limitations and model constraints \nParkinson’s disease is characterized by both the progressive degeneration of \ndopaminergic neurons in the SN and the widespread distribution of Lewy pathology, \nreinforcing the need for models capable of capturing this complexity (Herculano -\nHouzel, 2009). Neurotoxin-based models of Parkinsonism, including 6-OHDA, do not \nfully recapitulate α-synuclein pathology or the progressive nature of the disease. Still, \nthe 6-OHDA model in NHPs induces retrograde dopaminergic depletion, which may \noffer a valua ble window for investigating mechanisms of neurodegeneration and \ntesting neuroprotective strategies (Eslamboli et al., 2005). These limitations should be \nconsidered when interpreting the translational scope of the model. \nSince 6-OHDA does not cross the blood -brain barrier, the model depends on direct \nintracerebral delivery. Even minimal surgical deviations may affect the final distribution \nof the toxin within the target region, thereby contributing to variability in its sp read, \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nneuronal uptake, and ultimately the extent of dopaminergic depletion. MRI -based \ntargeting is essential to improve the accuracy of injection placement, particularly in \nnon-human primates, where inter -individual neuroanatomical variation may limit the \nprecision of stereotaxic coordinates derived solely from atlases (Pedrosa et al., 2024). \nAs a proof-of-principle, the present study nonetheless provides a reliable foundation \nfor future methodological refinement, supporting the optimization of targeting \nprecision, lesion consistency, and reproducibility in larger cohorts. Together, our \nfindings highlight the importance of integrating multidimensional behavioral metrics in \npreclinical models of Parkinson’s disease. \n \n4.5 Future directions \nFrom a broader translational perspective, the findings align with proof -of-principle \nevidence supporting the need for preclinical models capable of capturing the \ncomplexity and heterogeneity of PD beyond dopaminergic loss alone. Disease-related \ndysfunction involves subtle and progressive alterations in motor control, circuit -level \nreorganization, and integration between motor and cognitive domains, features that \nare not adequately reproduced by conventional rodent models (Bezard et al., 2025; \nEmborg, 2007). In this context, recent position statements, including those from the \nPD-AGE task force, reinforce that NHP models occupy a unique and indispensable \nrole in translational neuroscience, as they preserve key anatomical and functional \ncharacteristics of cortico-basal ganglia circuits, enabling the investigation of fine motor \ncontrol, behavioral adaptation, and progressive dysfunction with greater fidelity to the \nhuman condition (Bezard et al., 2025; Teil et al., 2021). Furthermore, the extended \nlifespan of NH P and their ability to model aging -related processes provide an \nadditional advantage for studying prodromal stages of neurodegeneration and disease \nprogression (Colman, 2018; Mattison & Vaughan, 2017). \n \n5. Conclusion \nThe present proof-of-principle study demonstrates that unilateral stereotaxic injection \nof 6 -OHDA into SN of adult Sapajus apella produces a graded and histologically \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nvalidated HP model, with ipsilateral tyrosine hydroxylase -positive neuron losses \nranging from 44 to 59% relative to the intact contralateral hemisphere, and no \nmeaningful hemispheric difference in the vehicle -injected control animal. This within -\nsubject design, enabled by the unilateral lesion paradigm, provided a robust internal \nreference for both behavioral and histological comparisons. \nCritically, the behavioral findings reveal a dissociation between preserved global task \nperformance and disrupted motor sequence organization: while gross outcome \nmeasures such as task completion rates remained largely intact, analysis of fine motor \nsequence organization using a deviation variance metric applied to the Brinkman \nboard identified post -lesion increases in retrieval sequence disorganization, \ndemonstrating that conventional behavioral measures may underestimate underlying \nmotor dysfunction. This  pattern, consistent with dopaminergic compensation \nmechanisms previously described in both clinical and experimental contexts , \nemphasizes the inadequacy of task -success metrics as sole endpoints in preclinical \nmodels of Parkinson's disease and highlights motor sequence analysis as a sensitive \nbiomarker of subclinical nigrostriatal dysfunction. \nTogether, these findings support the translational utility of capuchin monkeys as a \nbiologically relevant NHP model for Parkinson's disease research and establish a \nmethodological framework integrating quantitative stereology with multidimensional \nbehavioral analysis, as well as highlight motor sequence analysis as a promising \nbiomarker for detecting subclinical nigrostriatal dysfunction. Future studies employing \nlarger cohorts, extended post-lesion timelines, and expanded non-motor assessments \nwill be essential to consolidate these proof -of-principle observations and to evaluate \nthe responsiveness of the motor sequence biomarker to pharmacological or \nneuroprotective interventions. \n \n6. Conflict of interest \nThe authors declare that the research was conducted in the absence of any \ncommercial or financial relationships that could be construed as a potential conflict of \ninterest.  \n \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n7. Author contributions \n(1) Research project: A. Conception. B. Organization and execution. \n(2) Data analysis: A. Design. B. Refinement and statistical analysis. C. Review.  \n(3) Manuscript: A. Writing. B. Review and Critique.  \nL.R.R.P.: 1B, 2A, 2B, 2C, 3A, 3B. \nL.C.P.L: 1B, 2A, 2B, 2C, 3A, 3B. \nJ.A.P.C.M: 1B, 2C, 3C. \nA.G.S: 2B, 2C. \nR.R.L: 1B, 3B. \nD.SM: 1B, 3B. \nB.D.G: 1B, 2C, 3B. \nL.V.K: 1A, 1B, 2A, 3A, 3B. \n \n8. Acknowledgments \nWe would like to express our gratitude to the staff of the Medical Clinic in Castanhal -\nPA. We extend our thanks to the technicians of the National Center of Primates in \nAnanindeua-PA for their skillful assistance with MRI imaging acquisition, and to \ncollaborators of the Eduardo Oswaldo -Cruz Neurophysiology Laboratory of the \nFederal University of Pará (UFPA). We are also deeply grateful to Professor Cristovam \nWanderley Picanço Diniz and the team of the Laboratory of Investigations in \nNeurodegeneration and Infection (LNI), Institute of Biological Sciences, Federal \nUniversity of Pará, for generously providing access to their microscopy infrastructure \nand for their expert technical support during the histological imaging procedures. \n \n9. Data Availability Statement  \nThe datasets supporting the findings of this study are available as supplementary files \naccompanying the manuscript. These include: raw and session-level Brinkman board \nperformance metrics (brinkman_session_metrics.xlsx), deviation variance summary  \nstatistics for the Brinkman board sequence analysis  \n(brinkman_deviation_variance_summary.csv), Wilcoxon test results for Brinkman  \nboard sequence variability comparisons \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\n(brinkman_variability_wilcoxon_summary.xlsx), Brinkman board retrieval sequence  \nmatrices used to generate heatmap visualizations  \n(brinkman_heatmap_matrices.xlsx), Wilcoxon test results for Staircase test  \nparameters (staircase_wilcoxon_results.csv), and Wilcoxon test results for Tube test  \nparameters (tube_wilcoxon_summary.csv). The custom analysis algorithms and  \nscripts used for sequence deviation scoring and statistical processing are not publicly \ndeposited but are available upon reasonable request to the c orresponding author, \nBruno Duarte Gomes (brunodgomes@ufpa.br), Laboratório de  Neurofisiologia \nEduardo Oswaldo Cruz, Instituto de Ciências Biológicas,  Universidade Federal do \nPará, Avenida Perimetral, 2-224, Room 238, Guamá, Belém – PA, Brazil, 66077-830. \n \n10. References \nBadoud, S., Nicastro, N., Garibotto, V., Burkhard, P. R., & Haller, S. (2016). Distinct \nspatiotemporal patterns for disease duration and stage in Parkinson’s disease. \nEuropean journal of nuclear medicine and molecular imaging, 43(3), 509–516. \nBerardelli, A., Rothwell, J. C., Thompson, P. D., & Hallett, M. (2001). \nPathophysiology of bradykinesia in Parkinson’s disease. Brain, 124(11), 2131–2146. \nBezard, E., Anderson, R. M., Badin, R. A., Bergman, H., Boehringer, A., Borgognon, \nS., Emborg, M. E., Kordower, J. H., Li, J.-Y., & Martel, A.-C. (2025). Position paper: \nLeveraging non-human primate (NHP) specificities to accelerate Parkinson’s disease \nand ageing research. npj Parkinson’s Disease, 11(1), 227. \nBezard, E., & Gross, C. E. (1998). Compensatory mechanisms in experimental and \nhuman parkinsonism: Towards a dynamic approach. Progress in neurobiology, \n55(2), 93–116. \nBlesa, J., & Przedborski, S. (2014). Parkinson’s disease: Animal models and \ndopaminergic cell vulnerability. Frontiers in neuroanatomy, 8, 123289. \nBonthius, D. J., McKim, R., Koele, L., Harb, H., Karacay, B., Mahoney, J., & \nPantazis, N. J. (2004). Use of frozen sections to determine neuronal number in the \nmurine hippocampus and neocortex using the optical disector and optical \nfractionator. Brain Research Protocols, 14(1), 45–57. \nBraak, H., Tredici, K. D., Rüb, U., de Vos, R. A. I., Jansen Steur, E. N. H., & Braak, \nE. (2003). Staging of brain pathology related to sporadic Parkinson’s disease. \nNeurobiology of Aging, 24(2), 197–211. https://doi.org/10.1016/S0197-\n4580(02)00065-9 \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nBrotchie, J., & Fitzer-Attas, C. (2009). Mechanisms compensating for dopamine loss \nin early Parkinson disease. Neurology, 72(7_supplement_2), S32–S38. \nColman, R. J. (2018). Non-human primates as a model for aging. Biochimica et \nBiophysica Acta (BBA) - Molecular Basis of Disease, 1864(9, Part A), 2733–2741. \nhttps://doi.org/https://doi.org/10.1016/j.bbadis.2017.07.008 \nDarling, W. G., Ge, J., Stilwell-Morecraft, K. S., Rotella, D. L., Pizzimenti, M. A., & \nMorecraft, R. J. (2018). Hand motor recovery following extensive frontoparietal \ncortical injury is accompanied by upregulated corticoreticular projections in monkey. \nJournal of Neuroscience, 38(28), 6323–6339. \nEmborg, M. E. (2007). Nonhuman primate models of Parkinson’s disease. ILAR \njournal, 48(4), 339–355. \nEslamboli, A., Georgievska, B., Ridley, R. M., Baker, H. F., Muzyczka, N., Burger, \nC., Mandel, R. J., Annett, L., & Kirik, D. (2005). Continuous Low-Level Glial Cell \nLine-Derived Neurotrophic Factor Delivery Using Recombinant Adeno-Associated \nViral Vectors Provides Neuroprotection and Induces Behavioral Recovery in a \nPrimate Model of Parkinson’s Disease. Journal of Neuroscience, 25(4), 769–777. \nhttps://doi.org/10.1523/JNEUROSCI.4421-04.2005 \nEspay, A. J., Vizcarra, J. A., Marsili, L., Lang, A. E., Simon, D. K., Merola, A., \nJosephs, K. A., Fasano, A., Morgante, F., & Savica, R. (2019). Revisiting protein \naggregation as pathogenic in sporadic Parkinson and Alzheimer diseases. \nNeurology, 92(7), 329–337. \nGundersen, H. J. G., & Jensen, E. (1987). The efficiency of systematic sampling in \nstereology and its prediction. Journal of microscopy, 147(3), 229–263. \nHerculano-Houzel, S. (2009). The human brain in numbers: A linearly scaled-up \nprimate brain. Frontiers in Human Neuroscience, Volume 3-2009. \nhttps://doi.org/10.3389/neuro.09.031.2009 \nHoogewoud, F., Hamadjida, A., Wyss, A. F., Mir, A., Schwab, M. E., Belhaj-Saif, A., \n& Rouiller, E. M. (2013). Comparison of functional recovery of manual dexterity after \nunilateral spinal cord lesion or motor cortex lesion in adult macaque monkeys. \nFrontiers in neurology, 4, 101. \nIakovakis, D., Chaudhuri, K. R., Klingelhoefer, L., Bostantjopoulou, S., Katsarou, Z., \nTrivedi, D., Reichmann, H., Hadjidimitriou, S., Charisis, V., & Hadjileontiadis, L. J. \n(2020). Screening of Parkinsonian subtle fine-motor impairment from touchscreen \ntyping via deep learning. Scientific Reports, 10(1), 12623. \nhttps://doi.org/10.1038/s41598-020-69369-1 \nKish, S. J., Shannak, K., & Hornykiewicz, O. (1988). Uneven pattern of dopamine \nloss in the striatum of patients with idiopathic Parkinson’s disease. New England \nJournal of Medicine, 318(14), 876–880. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nLebowitz, J. J., & Khoshbouei, H. (2020a). Heterogeneity of dopamine release sites \nin health and degeneration. Neurobiology of disease, 134, 104633. \nLebowitz, J. J., & Khoshbouei, H. (2020b). Heterogeneity of dopamine release sites \nin health and degeneration. Neurobiology of disease, 134, 104633. \nMaetzler, W., Mirelman, A., Pilotto, A., & Bhidayasiri, R. (2024). Identifying Subtle \nMotor Deficits Before Parkinson’s Disease is Diagnosed: What to Look for? Journal \nof Parkinson’s Disease, 14(s2), S287–S296. https://doi.org/10.3233/JPD-230350 \nMarshall, J., Baker, H., & Ridley, R. (2002). Contralesional neglect in monkeys with \nsmall unilateral parietal cortical ablations. Behavioural Brain Research, 136(1), 257–\n265. \nMarshall, J. W., & Ridley, R. M. (2003). Assessment of cognitive and motor deficits in \na marmoset model of stroke. ILAR journal, 44(2), 153–160. \nMattison, J. A., & Vaughan, K. L. (2017). An overview of nonhuman primates in \naging research. Experimental Gerontology, 94, 41–45. \nhttps://doi.org/https://doi.org/10.1016/j.exger.2016.12.005 \nMcGregor, M. M., & Nelson, A. B. (2019). Circuit mechanisms of Parkinson’s \ndisease. Neuron, 101(6), 1042–1056. \nNudo, R. J. (2013). Recovery after brain injury: Mechanisms and principles. Frontiers \nin human neuroscience, 7, 887. \nNutt, J. G., Carter, J. H., & Sexton, G. J. (2004). The dopamine transporter: \nImportance in Parkinson’s disease. Annals of Neurology: Official Journal of the \nAmerican Neurological Association and the Child Neurology Society, 55(6), 766–\n773. \nPanyakaew, P., Duangjino, K., Kerddonfag, A., Ploensin, T., Piromsopa, K., \nKongkamol, C., & Bhidayasiri, R. (2023). Exploring the Complex Phenotypes of \nImpaired Finger Dexterity in Mild-to-moderate Stage Parkinson’s Disease: A Time-\nSeries Analysis. Journal of Parkinson’s Disease, 13(6), 975–988. \nhttps://doi.org/10.3233/JPD-230029 \nPedrosa, L. R. R., Leal, L. C., Muniz, J. A. P., Bastos, C. de O., Gomes, B. D., & \nKrejcová, L. V. (2024). From imaging to precision: Low cost and accurate \ndetermination of stereotactic coordinates for brain surgery Sapajus apella using MRI. \nFrontiers in Neuroscience, 18, 1324669. \nPerez, X. A., Parameswaran, N., Huang, L. Z., O’Leary, K. T., & Quik, M. (2008). \nPre‐synaptic dopaminergic compensation after moderate nigrostriatal damage in \nnon‐human primates. Journal of neurochemistry, 105(5), 1861–1872. \nPrasad, E. M., & Hung, S.-Y. (2020). Behavioral tests in neurotoxin-induced animal \nmodels of Parkinson’s disease. Antioxidants, 9(10), 1007. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nPrzedborski, S. (2017). The two-century journey of Parkinson disease research. \nNature Reviews Neuroscience, 18(4), 251–259. \nRouiller, E., Yu, X., Moret, V., Tempini, A., Wiesendanger, M., & Liang, F. (1998). \nDexterity in adult monkeys following early lesion of the motor cortical hand area: The \nrole of cortex adjacent to the lesion. European Journal of Neuroscience, 10(2), 729–\n740. \nSantana-Román, E., Ortega-Robles, E., & Arias-Carrión, O. (2025). Longitudinal \ndynamics of clinical and neurophysiological changes in parkinson’s disease over four \nand a half years. Scientific Reports, 15(1), 27284. \nSavidan, J., Kaeser, M., Belhaj-Saif, A., Schmidlin, E., & Rouiller, E. M. (2017). Role \nof primary motor cortex in the control of manual dexterity assessed via sequential \nbilateral lesion in the adult macaque monkey: A case study. Neuroscience, 357, \n303–324. \nSchaeffer, E., Kluge, A., Schulte, C., Deuschle, C., Bunk, J., Welzel, J., Maetzler, \nW., & Berg, D. (2024). Association of Misfolded α-Synuclein Derived from Neuronal \nExosomes in Blood with Parkinson’s Disease Diagnosis and Duration. Journal of \nParkinson’s Disease, 14(4), 667–679. https://doi.org/10.3233/JPD-230390 \nSchmidlin, E., Kaeser, M., Gindrat, A.-D., Savidan, J., Chatagny, P., Badoud, S., \nHamadjida, A., Beaud, M.-L., Wannier, T., & Belhaj-Saif, A. (2011). Behavioral \nassessment of manual dexterity in non-human primates. Journal of visualized \nexperiments: JoVE, (57), 3258. \nTeil, M., Arotcarena, M.-L., & Dehay, B. (2021). A New Rise of Non-Human Primate \nModels of Synucleinopathies. Biomedicines, 9(3), 272. \nhttps://doi.org/10.3390/biomedicines9030272 \nVermilyea, S. C., & Emborg, M. E. (2018). The role of nonhuman primate models in \nthe development of cell-based therapies for Parkinson’s disease. Journal of Neural \nTransmission, 125(3), 365–384. \nWest, M. J. (1993). New stereological methods for counting neurons. Neurobiology \nof aging, 14(4), 275–285. \nWest, M. J. (1999). Stereological methods for estimating the total number of neurons \nand synapses: Issues of precision and bias. Trends in neurosciences, 22(2), 51–61. \nWest, M., Slomianka, L., & Gundersen, H. J. G. (1991). Unbiased stereological \nestimation of the total number of neurons in the subdivisions of the rat hippocampus \nusing the optical fractionator. The Anatomical Record, 231(4), 482–497. \nYamanaka, H., Takata, Y., Nakagawa, H., Isosaka-Yamanaka, T., Yamashita, T., & \nTakada, M. (2021). An enhanced therapeutic effect of repetitive transcranial \nmagnetic stimulation combined with antibody treatment in a primate model of spinal \ncord injury. PLoS One, 16(6), e0252023. \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint \n\nYang, N., Liu, J., Sun, D., Ding, J., Sun, L., Qi, X., & Yan, W. (2025). Motor \nsymptoms of Parkinson’s disease: Critical markers for early AI-assisted diagnosis. \nFrontiers in Aging Neuroscience, Volume 17-2025. \nhttps://doi.org/10.3389/fnagi.2025.1602426 \nZhang, S., Yuan, L., Wu, Z., Du, X., Kubiak, J. Z., Yue, F., Yan, X., Jiang, G., & \nHuang, Y. (2025). Non‐human primate models of Parkinson’s disease: Decoding \npathogenesis and advancing therapies. Brain‐x, 3(2), e70032. \nZigmond, M. J., Abercrombie, E. D., Berger, T. W., Grace, A. A., & Stricker, E. M. \n(1990). Compensations after lesions of central dopaminergic neurons: Some clinical \nand basic implications. Trends in neurosciences, 13(7), 290–296. \n \nreuse, remix, or adapt this material for any purpose without crediting the original authors. \npreprint (which was not certified by peer review) in the Public Domain. It is no longer restricted by copyright. Anyone can legally share, \nThe copyright holder has placed thisthis version posted April 27, 2026. ; https://doi.org/10.64898/2026.04.23.720333doi: bioRxiv preprint","source_license":"Public-Domain","license_restricted":false}