We infused the right internal carotid artery with MPTP in anesthetized animals using angiographic guidance as previously described (6). Briefly, animals fasted overnight and underwent anesthesia induction with ketamine and glycopyrrolate followed by intubation with a soft-cuffed endotracheal tube. Anesthesia was maintained with isoflurane inhalation. A 20-gauge catheter was placed into a femoral artery followed by fluoroscopic guidance of an 80-cm catheter guided into the right internal carotid artery over a guidewire. Cerebral angiogram prior to infusion confirmed proper placement of the catheter and good distal blood flow.

To obtain a dose response curve, a randomized dose between 0 and 0.25 mg/kg of MPTP (Sigma, St. Louis, MO) diluted in normal saline was infused IC in a concentration of 0.1 mg/mL at a rate no faster than 1 mL per minute. The mean absolute dose of MPTP administered was 1.28 mg ± 0.71 across monkeys injected (10). A repeat angiogram was performed following infusion of MPTP to document maintenance of proper positioning and distal blood flow. After the procedure, animals were continuously observed until able to care for themselves. Safety precautions were followed for handling MPTP and all contaminated tissues and waste products (13).

Video was obtained at the same time of day (early morning) in the first 3 days following MPTP infusion, then several times per week after MPTP until euthanasia. The behavioral protocol included a) walking 15 feet toward, then away from the camera three times in a straight corridor, b) walking in a circle of approximately 7 feet diameter, 3 times clockwise then 3 times counterclockwise, and c) reaching 10 times with each hand for a food item held by the trainer at arm’s length from the animal. A single observer blinded to each animal’s treatment status rated each video clip, which was presented in a random order, using a scale previously developed and validated for non-human primate studies (7). Behavioral evaluations of parkinsonism and dystonia are consistent with those previously described (6, 7, 9).

Dystonic posturing was defined as extensor posturing of an upper or lower extremity and was clearly distinguished from the more typical parkinsonian flexed posture of the limb. Briefly, ratings for dystonia included a 0–3 scale (0 = unaffected, 1 = mildly affected, 2 = moderately affected, and 3 = severely affected) for dystonia occurring in each limb: 1) with the animal sitting at rest, 2) with movement of the rated limb (i.e., during gait or when reaching for a food item), and 3) movement of the contralateral limb. The ratings were scored for each side separately, producing a composite with a maximum dystonia rating of 18 per side. Ratings for parkinsonism included a 0–3 scale (0 = unaffected, 1 = mildly affected, 2 = moderately affected, and 3 = severely affected) for bradykinesia, tremor, and/or flexed posturing. The maximum score for parkinsonism ratings was also 18 per side. Behavioral raters were blind to in vitro measures (described below) and MPTP dose. Parkinsonism and dystonia ratings from some animals have been previously published (6, 9), but rating scores were reestablished to maintain rater consistency across all animals as the prior rater was no longer available to extend ratings to newly acquired animals.

Animals were euthanized with intravenous pentobarbital (100 mg/kg; Somnasol Euthanasia, Butler Schein Animal Health, Dublin, OH) at completion of the parent study. The brain was removed within 10 min. Details regarding animal euthanasia, brain retrieval and preparation, tyrosine hydroxylase immunohistochemistry and stereology, and striatal dopamine measurement protocols are described elsewhere (9), but we provide a brief overview here.

For substantia nigra tyrosine hydroxylase immunohistochemistry, the midbrain was fixed in 4% paraformaldehyde in 0.1 M phosphate buffered solution for 7 days, transferred to 30% sucrose for 7 days, and then cut on a freezing microtome in serial 50 μm-thick transverse sections after 7 days. Every 6th section was processed for tyrosine hydroxylase immunocytochemistry and sections were counterstained with cresyl violet for unbiased stereological counts of nigral neurons obtained via microscope digital image acquisition. Substantia nigra pars compacta was defined as the region ventral to medial lemniscus and lateral to the third cranial nerve fibers, and TH-stained nigral cells were counted using the Visiopharm Integrator System (version 3.2.10.0).

For striatal dopamine measures, separated hemispheres were sliced in the coronal plane at the time of euthanasia followed by collection of standard punch biopsies from caudate and putamen bilaterally which were frozen on dry ice snow and stored at −80°C until dopamine was assayed. High-performance liquid chromatography with electrochemical detection was applied to measure striatal dopamine levels, expressed in microgram per Gram of wet tissue.

All in vitro measurements were performed by a rater blinded to MPTP dose and behavioral measures. Select in vitro measures from animals included here have been published in-part elsewhere (9).

Mean, standard deviation, is reported for continuous variables of interest, while median is reported for ordinal data. Two-sample paired t-tests were used to compare continuous variables of interest between the MPTP-treated and control sides. We correlated non-parametric validated blinded ratings of transient peak dystonia scores with that of subsequent peak parkinsonism, in vitro measures of striatal dopamine, and unbiased stereologic counts of nigral neurons with TH by calculating a one-tailed Spearman’s coefficient (rho) in SPSS (IBM, version 22). Prior to data analysis, we conservatively defined the threshold of behavioral dystonia as a score of at least two points to avoid possible overlap with normal behavior, similar to our prior strategy with respect to parkinsonism (9).

The typical behavioral response to unilateral MPTP injection included recovery from anesthesia followed within hours to days by contralateral transient hemi-dystonia. Hemi-dystonia was observed in 16/18 MPTP injected monkeys as extension of the fore- and hind-limbs at the shoulder/hip, elbow/knee, and wrist/ankle with external rotation at the shoulder/hip that often worsened with attempted use of the affected limbs (either walking-worse when walking in circles with the affected side on the inner diameter- or reaching for objects). There was no postural tremor at onset of dystonic features in any of the observed monkeys. Four animals (two that did not receive MPTP) did not exhibit any dystonia or parkinsonism following MPTP administration, before euthanasia. The mean times following MPTP to onset of dystonic manifestations, peak dystonia and duration of dystonia are presented in Table 1. Hemi-dystonia gradually resolved after peak dystonia with overlapping onset of hemi-parkinsonism (Figure 1). At euthanasia, no animals had dystonic manifestations.

Peak dystonia score correlated with peak parkinsonism ([18],r = 0.82, p < 0.001, Figure 2A). Mean time following MPTP to onset and plateau of parkinsonism at peak parkinsonism severity are presented in Table 1. The parkinsonian phase typically began with flexed posturing of the affected upper and lower extremity with bradykinesia. Some ambiguity existed with regard to scores of bradykinesia since severe dystonia often caused bradykinesia. In some animals, extensor dystonic posturing of the lower or upper extremity persisted when bradykinesia developed. Only two animals had tremor during the parkinsonian phase.

Quantification of TH-stained nigral neurons was obtained a mean of 53 days following MPTP, in all cases acute transient dystonia resolved. The percent of residual nigral cell counts correlates with the prior peak dystonia rating scores ([18],r = −0.63, p = 0.002, Figure 2B). The threshold of nigral cell loss remains unknown given timing of TH-nigral measures in relation to dystonia, but monkey 6 represents the lowest threshold of injured nigra at 14% with associated dystonia, a similar level to that previously observed for parkinsonism (9).

The peak dystonia rating score correlated linearly with the percent residual striatal dopamine obtained a mean of 53 days after MPTP (r = −0.81, p < 0.001), when nigral injury was less than 50%. When nigral injury exceeded 50%, the percent residual striatal dopamine was essentially zero across the entire measured range of dystonia scores. Data for residual striatal dopamine are displayed in Figure 2C; we only correlate data where nigral injury is less than 50% to avoid false representation of what appear as independent clusters. The threshold of striatal dopamine loss that corresponded to transient dystonia also remains unknown given timing of striatal dopamine measures in relation to dystonia, but the lowest threshold of striatal dopamine depletion was 14%–36% with associated prior transient dystonia, similar to that previously reported for parkinsonism (9). We averaged post-mortem striatal measures of dopamine for the caudate and putamen as the mean caudate dopamine concentrations did not differ from putamen measures on the control side (mean 11.9 ± 6.5 ug/g versus 12.1 ± 5.1 ug/g of wet tissue, respectively; two-sample paired t-test: t = −0.13, p = 0.90) and MPTP-lesioned side (5.4 ± 7.3 ug/g versus 5.4 ± 6.2 ug/g of wet tissue, respectively; two-sample paired t-test: t = −0.03, p = 0.97). Similarly, the mean caudate percent residual dopamine concentration did not differ from the putamen concentrations (41 ± 42% versus 43.7 ± 44%, respectively; two-sample paired t-test: t = −0.53, p = 0.61).

Dystonia is not consistently reported before parkinsonism in primates treated with MPTP. Variability in MPTP primate models may contribute to this, including use of different ages of old- and new-world primates and variable administration and dosing methods that includes intracarotid, intramuscular, subcutaneous, or intravenous as a single dose or repeatedly over a few days, weeks, or months (14–22). Regardless, IC MPTP administration in baboons, Macaca nemestrina, and Macaca fasicularis has consistently caused transient dystonia followed by chronic parkinsonism (6, 7). It remains unclear whether differences in observed behavior results from methodology, or rather, lack of recognition and/or reporting of the early dystonic manifestation. The high correlation of dystonia and parkinsonian severity strengthens the association of MPTP with transient dystonia, strengthening support of this model for the study of pathophysiologic mechanisms and nigrostriatal function.

It is unlikely that subjective rating errors occurred when distinguishing phenomenology of dystonia and parkinsonism (i.e., mistaking early, more severe parkinsonism for dystonia) since the scores on these two rating scales clearly diverged as demonstrated in Figure 1. Also, a key rating observation helped to distinguish the dystonic from parkinsonian phase; we defined dystonia as extensor posturing of an extremity while flexion indicated parkinsonism.

Animal studies have previously implicated the basal ganglia in dystonia [see reviews: (23–25)]. Selective striatal lesions with 6-hydroxydopamine or 3-nitropropionic acid and MPTP caused dystonia in rodents (26, 27) and non-human primates (6, 7, 28, 29). MPTP demonstrates selective preference to injure dopaminergic neurons (30–32). During the dystonic phase in baboons previously treated with IC MPTP, putaminal dopamine D2-like specific binding sites decreased about 30% (measured as soon as 10 days after MPTP) with associated striatal dopamine deficiency of about 97%–98% (33). In macaques, during the dystonic phase, striatal dopamine decreased about 66% in the first day (n = 2), 98% at day 3 (n = 1) and 99% at 2–4 months (n = 5) (6). Striatal FDOPA uptake decreased approximately 70%, a loss maintained 2–4 months following infusion (6). Thus, the dystonic phase begins, continues and ends during a period of sustained striatal dopamine deficiency. Based on our observations that transient dystonia occurs in context of incomplete, low intensity persistent striatal dopamine depletion is of particular interest regarding pathophysiology given that striatal healing is unlikely. Our findings indicate that striatal dopamine alone does not reflect dystonia severity in this model through the full range of severity.

Macaque nemestrina and Macaque fascicularis animals that develop dystonia following IC MPTP (0.1–0.4 mg/kg) have quicker drops in striatal dopamine during the dystonic phase (6) than squirrel monkeys systemically administered MPTP for which dystonia was not reported (34). The squirrel monkeys have only modest drops in putamen dopamine at day 1 following subcutaneous administration of 2.5 mg/kg followed by further decreases at day 5 in both caudate and putamen. Thus, the rapid changes induced by IC MPTP that is completely extracted on first pass through the brain circulation may be critical to induce dystonia, above and beyond striatal dopamine deficiency alone.

The temporal dissociation of transient dystonia and biochemical measures limits our ability to speculate regarding the correlation of dystonia severity with residual nigral cell counts. This association may reflect changes in the effects of dopamine transmission and/or receptor configuration in downstream striatal terminals, for which future studies are required to understand the timing of nigral injury in relation to transient dystonia.

A critical question in our non-human primate IC MPTP model relates to what causes dystonia to recover whereas parkinsonism does not. Alterations in striatal dopaminergic signaling relate to the pathophysiology of both dystonia and parkinsonism (35, 36). Reports regarding the effects of MPTP on dopamine receptors vary. In general, there appears relatively little change in D1-like receptor numbers. Some report decreased dopamine D2-like receptors, but others report no change or an increase (37–41). As multiple dopamine receptors are the products of separate genes (42), it is possible that changes in D2-like receptor binding described in prior studies may be attributable to changes in the expression of a single receptor subtype or to complex changes in the expression of D2, D3, or D4 receptors. Future studies focused on potentially adaptive relationships of various dopamine receptors during the transition from dystonia to parkinsonism, particularly differentiating the direct and indirect pathway, may be valuable to understand the role of such receptors in altered neural plasticity and phenotypic conversion from dystonia to parkinsonism. Changes in other modulators of the direct and indirect pathways such as such as dynorphin, enkephalin, substance P and PDE10a may also play a role in development of dystonia, and future molecular studies measuring their levels in this NHP model during and after transient dystonia may prove informative regarding pathophysiologic mechanisms of transient dystonia.

Acute transient dystonia in the current NHP model may also relate to abnormal cortico-striatal neural plasticity (43). In rodents, MPTP induced nigro-striatal injury produces immediate striatal release of dopamine and its metabolites (44, 45). Early profound dopaminergic efflux likely stimulates D1 and D2-like dopamine receptors simultaneously, possibly causing acute alterations in glutamatergic sensitivity via altered synaptic long term potentiation and long term depression at medium spiny neurons (46). Thus, IC MPTP may cause a sudden flood of available dopamine followed by marked reduction (6). This rapid dopaminergic striatal efflux followed by relative dopaminergic paucity mimics the clinical scenario of human cocaine exposure (i.e., striatal dopaminergic priming), where those exposed are more prone to drug-induced dystonia following dopamine receptor blockade (47). Acute dopamine depletion may produce more excitable medium-sized spiny neurons, less excitable GABAergic interneurons and increased cholinergic cell excitability (48, 49); thus, dopamine depleted striatal circuits exhibit pathological hyperactivity. Longitudinal neurophysiological, PET, and functional neuroimaging experiments flanking development and resolution of transient dystonia in this NHP MPTP model may unravel potential links between biochemical and circuit-level dysfunction contributing to altered cortico-striatal plasticity.

Each of these potentially contributing mechanisms of altered striatal plasticity requires future study, but the most intriguing relates to cholinergic interneurons (ChI). ChIs show hyperactivity with bursts and silences in the dopamine depleted striatum, where hyperactivity also relates to extrinsic synaptic inputs targeting ChIs via glutaminergic and GABAergic inputs. The latter is essential for sustaining ChI hyperactivity, but MPTP induced dopamine loss alters substantia nigra pars compacta signaling which reduces glutamate co-excitation of dorsolateral striatal ChIs due to downregulation of mGluR1 (50). As the relative ratio of striatal acetylcholine/dopamine relates to acute dystonia in mice treated with acute reserpine, one might anticipate that this ratio would normalize with downregulation of mGluR1, potentially translating to the end of the dystonic phase. This speculation supports growing data that ChIs play a role in dystonia and may serve as the main drivers of pathological hyperactivity in the striatum secondary to dopamine depletion and altered extrinsic synaptic inputs (51). Post-mortem histopathological measures of choline acetyltransferase (ChAT) or in vivo PET measures of [18F]fluoroethoxybenzovesamicol, (-)-[18F] FEOBV,(-)-(2R,3R)-trans-2-hydroxy-3-(4- phenylpiperidino)-5-(2-[18F]fluoroethoxy)-1,2,3,4-tetralin (52) or (-)-(1-(8-(2-[(18)F)fluoroethoxy)-3-hydroxy-1,2,3,4-tetrahydronaphtalen-2-yl)-pi peridi n -4 -yl) (4-fluorophenyl)methanone([18F]VAT) (53) may provide clues with respect to altered ChI activity during the transient dystonia phase in this NHP MPTP model.

Future studies are critical to tease out mechanistic components in the NHP model of transient dystonia which may include shifting neurotransmitter levels, immunomodulatory effects secondary to cellular injury, direct nigrostriatal influence versus indirect downstream effects, receptor auto regulation, altered transcription pathways, or shifts in network-level synchronicity.

Given the acute nature of IC MPTP in our animals, one might speculate that initial terminal field injury causes dystonia with recovery of function (i.e. sprouting of terminal fields), relieving dystonia. However, there is no evidence in our animals to support sprouting of terminal fields that produces any functional activity or neuronal recovery during the dystonic phase or beyond (54).