Induction and Assessment of Levodopa-induced Dyskinesias in a Rat Model of Parkinson's Disease

Although it has been more than 50 years since levodopa (L-DOPA) was first introduced as a treatment for individuals with PD^1,2, it remarkably remains the most effective therapy for parkinsonian motor symptoms. The clinical motor symptoms associated with PD stem from the loss of dopamine (DA) neurons in the substantia nigra (SN) pars compacta, resulting in the dramatic decrease in available dopamine in the striatum. L-DOPA effectively restores striatal DA levels, resulting in motor benefit early in the disease^3,4. Inopportunely, with long-term treatment, most individuals with PD will develop L-DOPA induced dyskinesias (LID), including chorea, dystonia, and athetosis, which often significantly impact activities of daily living^5,6,7.

While several behavioral models of LID in rodents exist, differences in modeling and behavioral assessment of LIDs have called into question the reproducibility of results between labs as well as the reliability of these experimental tools for pre-clinical PD research. Developed in association with a clinical movement disorder specialist⁸, the current protocol is a straightforward method for LID induction and rating and is appropriate for use in a rat model of PD utilizing 6-hydroxydopamine (6-OHDA)-induced unilateral nigral lesioning^9,10. The LID rating scale provided here includes scoring for both the intensity and frequency of dyskinetic behavior in various individual body parts. Pertinent information regarding workflow optimization of experiments and the appropriate care and handling of parkinsonian and dyskinetic animals is also provided.

The animals presented here were maintained and handled in compliance with the institutional guidelines. All animal procedures were approved by the Michigan State University Institutional Animal Care and Use Committee (IACUC) in compliance with federal and state regulations.

1. Drug-free confirmation of 6-OHDA lesion status

1. Postural Tail hang test^11,12,13
   NOTE: Assess lesion status at least 1 week following 6-OHDA lesion induction (See^9,10,35 for details on lesioning) in experimental subjects (e.g., male or female, adult Sprague Dawley or Fisher 344 rats). 
   1. Suspend the rat approximately 6 cm above its cage, firmly holding at the base of the tail, for ~5 s.
   2. Record the direction of body contortion as + for a successfully lesioned animal twisting contralateral to the lesioned side and - for lack of twisting or twisting in both directions.
   3. NOTE: These tests are optional but recommended. See^14,15,16,17 for additional drug-free testing options/variations.
2. Step adjusting drag test (adapted from¹⁶)
   NOTE: Assess lesion status at least 1 week following unilateral 6-OHDA lesion induction in experimental subjects (e.g., male or female, adult Sprague Dawley or Fisher 344 rats).
   1. Hold the rat by the base of its tail, elevating the back feet off the surface by ~6 cm; drag backward across a flat, smooth but not slippery surface, ~75 cm, over 5–10 s.
   2. Observe and record the number of tapping/step adjusting movements of each forepaw over three repeated tests.
   3. Score the subject as + for successful unilateral lesioning when 0–2 forepaw taps are observed contralateral to the lesioned side, together with rapid tapping (~10 taps) from the forepaw ipsilateral to the lesioned side (e.g., animals unilaterally lesioned on the left side show tapping deficit (0–2 taps) with the contralateral right forepaw).
   4. Conversely, score moderate to rapid tapping (5–10 taps) from both forepaws as - to indicate incomplete or unknown lesion status.
      NOTE: An anxious animal can show rapid tapping/step adjusting even if successfully lesioned. If this is suspected, place the rat back in their home cage and re-test ≥30 mins later.

2. Preparation of reagents and supplies

1. Determine L-3,4-dihydroxyphenylalanine methyl ester hydrochloride (levodopa or L-DOPA) and benserazide hydrochloride, a peripheral decarboxylase inhibitor (see Table of Materials) dose, rating frequency, and experimental timeline that is appropriate for the investigational question^12,18,19,20 (Figure 1).
   NOTE: Investigational questions can be posed to seek any number of questions ranging from asking whether a specific therapy might reduce existing LID or prevent induction of LID. They can also explore whether therapeutic efficacy is dependent on the dose of levodopa or whether LID expression and/or therapeutic efficacy varies depending on the sex, species, and age of the subject.

Figure 1: Example of treatment timeline. Example L-DOPA dose-escalation timeline of 12 weeks in total length, with 8 weeks of L-DOPA injections beginning 3 weeks after 6-OHDA lesioning and 4 weeks following experimental treatment. In this example, L-DOPA is subcutaneously injected 5x per week (Monday–Friday) at approximately the same time each day, for 2 weeks at each prescribed L-DOPA dose (3, 6, 9, and 12 mg/kg). Behavioral LID ratings take place on days 1, 6, and 10 of each L-DOPA dosage level. Please click here to view a larger version of this figure.

2. Weigh rats weekly to calculate the appropriate drug quantity based on the ongoing weight changes during the study.
   NOTE: Due to increased activity in LID+ rats, there is potential for weight loss with long-term L-DOPA treatments. If weight loss occurs provide rats with nutritionally complete, highly palatable treats (see Table of Materials) following L-DOPA injections.
3. Calculate the amount of L-DOPA and benserazide required for each weekly concentration, weighing out lyophilized aliquots for each day of injection, and storing in combination for 1–2 weeks at -20 °C in glass amber vials until the day of treatment.
   NOTE: Target dose is 12 mg/kg or 12 mg L-DOPA/1000 g body weight. Example of calculations for determining the amount of L-DOPA and saline needed for each day of a week using 12 mg L-DOPA/kg body weight at an injection volume of 1 cc/kg of rat weight is given in Supplementary File 1.

3. Room and cage set up

1. On the first day of L-DOPA treatment 3-4 weeks following 6-OHDA lesion surgery, transfer rats to single housing, including IACUC-approved enrichment.
2. Maintain in single housing throughout the study to avoid peer interference with behavioral assessments.
3. On LID rating days, place the home cages on a steel wire rack, turned at an approximately 45^° angle for optimal viewing of the rat (Figure 2A). Flip the identification tags (Figure 2B) upward and remove water bottles, food racks, and all types of enrichment in the cage (Figure 2C) to avoid interference with behavioral assessments.

Figure 2: Example of the cage set up for LID ratings of large-scale rat experiments. (A) Multiple cages can be set up for LID rating using large metal racks that allow for optimal viewing of each animal. Cages should be spread apart at a 45° angle with ID cards flipped upward (B), food, water bottles, nesting materials, and other enrichment removed to limit visual obscureness of the rat and distractions to the rat while examining dyskinetic behaviors (C). The metal racks need to be a few feet away from any wall to allow the rater to examine the rat at the front or back of the cage as needed. It is essential to label enrichment apparatuses (e.g., C- red rat retreat houses) with individual animal IDs to replace them into the same cage from which they came. This is particularly important when using animals of different sexes not to increase stress to the experimental subjects. Please click here to view a larger version of this figure.

4. Levodopa injections and dyskinesia rating

1. Subcutaneous injections of L-DOPA^21,22
   1. Immediately preceding the daily injection of L-DOPA, add the appropriate volume of sterile saline to the pre-weighed lyophilized L-DOPA and benserazide mix in the amber vial and shake well for 10 s (step 2.3).
      NOTE: Target injection volume is 1 mL/1000 g body weight (with 12 mg L-DOPA per mL). The volume of the sterile saline will depend on the number of animals per study.
   2. Fill individual syringes (e.g., 1.0 or 0.5 mL with 26 G needle) with the required volume for each animal (1 mL/kg rat weight) and label each syringe with individual animal identification.
      NOTE: Keep the filled syringes protected from light in sterile pouches until the time of injection. L-DOPA rapidly oxidizes in the presence of oxygen and light in an aqueous environment^23,24,25,26.
   3. Bring the first cage to the injection bench.
   4. Remove the rat from its cage and place it on the injection surface.
   5. Gently restrain the head and shoulders against the surface on which the rat is resting with the palm of the non-dominant hand.
   6. Gently, scruff the skin on the back overlying the scapulae with the thumb and forefinger of the non-dominant hand, inject L-DOPA volume with the dominant hand into the subcutaneous space between/below the fingers, keeping the needle as parallel to the body as possible to avoid intramuscular injection.
      ​NOTE: The rats are not anesthetized before injection.
   7. Dispose of each used individual syringe in a sharp's container.
   8. Replace the rat into its individual cage and add nutritionally complete treats except on LID rating days to avoid interference with behavioral assessments until after ratings are complete.
   9. Set the timer for 1–2 min depending on the rating time desired and the number of rats in the study on rating days. Retrieve the next cage and inject the next rat when the timer indicates.
   10. Repeat this, injecting one rat every 1–2 min, until all the rats are injected.
2. Levodopa-induced dyskinesia rating post-injection
   1. Rate the intensity (Table 1) and frequency (Table 2) of dystonic and hyperkinetic dyskinesia movements at the desired number of timepoints, which should include the initial onset of LID behavior, peak behavior, and the phase of decline (see Supplementary File 2 for an example of LID rating log sheet).
   2. For male and female adult Sprague Dawley or Fisher 344 rats, and a sample size of N = 40 rats, begin the dyskinesia ratings 20 min after the first L-DOPA injection, and then at 50 min intervals until 220 or 270 min post-injection, depending on when LID behaviors have discontinued in 90%–100% of rats.
   3. If using 1 min rating intervals, set a timer for 1 min. Rate the first rat for one minute. Move to the next rat and rate it for 1 min. Continue through all the rats, rating for 1 min intervals.
   4. Have a timer positioned next to the cage visibly so that the LID behavior intensity (Table 1) can be observed while estimating the frequency of any given behavior (Table 2) during the rating period.
   5. After ratings for the first time point are completed start again with the first rat at the next time point (e.g; 70 min post-injection) and continue at the desired interval (e.g., every 50 min) until all time points are completed.
      NOTE: Due to the overlap of L-DOPA injection and LID rating tasks, two persons are needed on the rating days, one for injecting and one for behavioral ratings.

LIDs in parkinsonian rats can manifest as a range of abnormal involuntary movements (AIMs), including dystonic, hyperkinetic, and stereotypic behaviors. LID rating criteria for such behaviors are presented here to include both intensity (Table 1) and frequency (Table 2). This provides an overall LID severity score for each rat that reflects both the quality (intensity) and quantity of time spent engaging (frequency) in these behaviors at each rating timepoint. The final LID severity score is calculated by multiplying the intensity score by the frequency score for each behavioral component. Scoring criteria for individual attributes of LID behaviors are provided here as written descriptions (Table 1–2), and examples are shown as still images with detailed reports (Figure 3) and as videos (Animated Videos 1–4). Additional descriptive information about individual movements and scoring is provided in the figure legends.

A comprehensive assessment of the impact of any treatment on amelioration of LID behaviors over time can be observed using a L-DOPA dose-escalation approach, 12 weeks in total length, with 8 weeks of L-DOPA injections following 6-OHDA lesioning (Figure 1). In this scenario, L-DOPA is given 5x per week (Monday–Friday) for 2 weeks at each prescribed L-DOPA dose. Behavioral ratings that take place on days 1, 6, and 10 of each L-DOPA dose provide a robust approach for treatment efficacy assessment (Figure 1). However, researchers may find that altered timelines and dosing schemes better answer their experimental questions. Indeed, the dose of L-DOPA and benserazide, the peripheral decarboxylase inhibitor, can vary depending on the investigator's experience with doses and/or the hypotheses to be tested. It is of note that the most common doses of benserazide are between 10–15 mg/kg, which is supported by the report by Tayarani-Binazir et al. (2012)²⁰ showing that optimal behavioral effects of L-DOPA are found with 10 mg/kg of benserazide, with no additional benefit at 15 mg/kg.

There are multiple ways to present and analyze LID data when determining whether a given treatment has a meaningful impact in ameliorating this aberrant behavior. Figure 4A provides example of data presented as peak dose LID over the entire experimental time line. Peak dose LID can be defined as the time point at which the group average LID score is the greatest. Alternatively, one can report ‘absolute peak’, which is the maximal peak LID score for an animal regardless of when it occurs during a rating period. Additionally, LID rating data can be examined over each daily rating time course and presented either as experimental group data and/or plotted to show individual subject's LID severity at each time point (Figure 4B)¹⁹. This latter approach allows assessment of whether some treatments can reduce, for example, the total amount of time a subject displays LID, but not necessarily impact peak dose severity. Finally, one can quantify total LID severity scores over the entire rating session. For a given subject, the LID severity score for each time point is summed to produce a total score for that rating period. This approach emulates the area under the curve estimations, which also can be calculated.

It is essential to appreciate that LID/AIMs scoring data, created using rating scales for assigning values to dyskinesia severity and/or duration (Tables 1 and 2), are ordinal data. Thus the most appropriate statistical tests are non-parametric. While there is no direct equivalent non-parametric test for two-way analysis of variance, LID data can be analyzed with non-parametric alternatives to the one-way ANOVA. Specifically, the Kruskal-Wallis test is used for between-subject comparisons of two or more independent groups. The Friedman test is a non-parametric alternative to a one-way ANOVA with repeated measures used for comparisons within subjects. Both are used with post-hoc tests (i.e., Dunn or Dunn-Bonferroni) following a significant Kruskal-Wallis or Friedman's test. For examining whether there are meaningful differences between two independent groups, the Mann-Whitney U test is considered the non-parametric equivalent of the independent t-test^28,29.

Figure 3: Freeze-frame images showing examples of LID intensity scoring in parkinsonian rats. (A) Representative image of a non-dyskinetic parkinsonian rat displaying the often-typical hunched parkinsonian posture in the absence of dystonia or hyperkinesia. (B) Parkinsonian rat with overall mild LID severity, indicating mild-to-moderate right forelimb dystonia and a lack of neck, trunk, or hindlimb dystonia; in the accompanying video (Animated Video 1), small-amplitude right forepaw dyskinesia (RFPD) and orolingual movements can be seen. (C) Parkinsonian rat with moderate neck and trunk dystonia (bold text corresponds with directional arrows indicating trunk twisted at 90° angle). (D) Dyskinetic parkinsonian rat with mild neck and moderate trunk dystonia, moderate-to-severe right forelimb dystonia; in the accompanying video (Animated Video 3), mixed amplitude RFPD can be observed. (E) Parkinsonian rat with moderate-to-severe neck and severe trunk dystonia, severe right hindlimb, and moderate right forelimb dystonia (bold text corresponds with directional arrows indicating the trunk is twisted ~180°; this is especially notable in example #2). Please click here to view a larger version of this figure.

Figure 4: Example LID data complied over 8 weeks of LID ratings with escalating L-DOPA dosage. (A) Peak dose LID severity (80 min post-L-DOPA) across time and doses. (B) Daily time course (20–170 min post-L-DOPA). For each dose and day, the top graphs reflect mean ± SEM; the bottom graphs show individual subject responses over time. Statistics were calculated using Kruskal-Wallis with Dunn's multiple comparison tests (for between-subjects tests) and Friedman tests with Dunn's multiple comparison tests (for within-subjects tests). Abbreviations: Control (Ctr) (n = 7); Treatment (Tx) (n = 10). This figure is reprinted and adapted with permission from Reference¹⁹. Please click here to view a larger version of this figure.

Table 1: LID rating criteria for Intensity of dystonic or hyperkinetic behaviors in parkinsonian rats. These rating criteria provide a range of intensity measures related to the quality/severity of the rat's abnormal involuntary postures and/or movements. Specific attributes described for postures and behaviors are generally classified as mild, moderate, or severe, with specific descriptors provided here. A final severity score is determined as the product of Intensity x Frequency (Table 2).

Table 2: LID Rating criteria for Frequency of dystonic or hyperkinetic behaviors. These criteria provide a quantification measure for the rate at which any given behavior occurs or is repeated during the observation period. This is important in assessing the overall severity, given that infrequent behavior warrants a less severe score than a persistent one. A final severity score is determined as the product of Intensity (Table 1) x Frequency.

Animated Video 1: Mild LID with small amplitude RFPD and chewing of the left forepaw. Dyskinetic parkinsonian rat with no appreciable neck or trunk dystonia, but moderate right forepaw dystonia, small amplitude RFPD noted as rapid repeated flickering movement of the right forepaw. This rat also displays orolingual behavior involving vacuous chewing directed at the left forepaw. L-DOPA dose: 12 mg/kg; video recorded ~70 min post-injection. Please click here to download this Video.

Animated Video 2: Moderate trunk and neck dystonia. Dyskinetic parkinsonian rat with moderate trunk and neck dystonia exemplified by the 90° difference in the position of the forelimbs and hindlimbs (see diagram in Figure 2C). This rat also displays prominent stereotypic sniffing and occasional vacuous biting at the litter. L-DOPA dose: 12 mg/kg; video recorded ~70 min post-injection. Please click here to download this Video.

Animated Video 3: Moderate-to-severe right forepaw dystonia with mixed amplitude RFPD. Dyskinetic parkinsonian rat with mild neck and moderate trunk dystonia, moderate-to-severe right forepaw dystonia, mixed amplitude RFPD seen as smaller wiping movements near the mouth mixed with larger amplitude downward pulling of the right forepaw. This rat also has frequent rotational behavior that can be quantified in addition to LID profiles. L-DOPA dose: 12 mg/kg; video recorded ~70 min post-injection. Please click here to download this Video.

Animated Video 4: Head bob and orolingual movements with tongue protrusions. Dyskinetic parkinsonian rat with severe trunk and forepaw dystonia (hindlimb dystonia not visible), moderate-to-severe neck dystonia, and a constant head bob with ongoing tongue protrusions. The head bob resembles the 4 Hz tremor seen in PD; however, in rats, it is only noted in dyskinetic parkinsonian rats following L-DOPA administration. This rat also has mixed amplitude RFPD, seen as more significant amplitude movements early in the video and small wiping-like motions near the face, characteristic of small amplitude RFPD seen later. Slight movement was observed on the severely dystonic/extended forelimb early in the video; however, this is related to the head and neck tremor activity and would not be classified as small-amplitude RFPD. When the observer taps strongly on the cage, these LID behaviors are not interruptible, suggesting that this is a severe LID that cannot be overcome with a startle. Please click here to download this Video.

Supplementary File 1: Example for calculating the amount of L-DOPA. Example of calculations for determining the amount of L-DOPA and saline needed for each day of a week using 12 mg L-DOPA/kg body weight at an injection volume of 1 cc/kg of rat weight. Please click here to download this File.

Supplementary File 2: Example LID rating log sheet. This sheet can be used or adapted to log LID intensity and frequency scores during rating sessions.The final severity score is determined as the product of Intensity (Table 1) x Frequency (Table 2). These rating scales can also be used to quantify LID in mice³⁰. Please click here to download this File.

Presented here are details for the reproducible induction and rating of LIDs in a parkinsonian rat model following unilateral 6-OHDA lesioning of the nigrostriatal DA system. While it was once thought that rodents did not develop LID and that rotational asymmetry may be the analog of LID in rats³¹, rat and mouse models have been characterized over the past two decades and are a well-accepted tool for LID research^15,32,33,34. The protocol presented here is specifically helpful for larger-scale experimental designs^18,19. It provides details for rating the intensity and frequency of a range of dyskinetic behaviors, both dystonic and hyperkinetic. Notably, images and videos with corresponding detailed notes and scoring of dystonic and hyperkinetic LID behaviors will assist the experimenter's assessment following L-DOPA administration. While an example of every possible AIM is not provided, a range of behaviors varying from mild to moderate to severe is offered as a framework for LID assessment. This protocol has been documented to help evaluate various therapies in parkinsonian rats^19,35,36,37. It augments other published rodent rating scales that focus primarily on the duration of prescribed LID behaviors^33,38,39. Further, the outlined procedure seeks to align scoring methods in a rat PD model with current practices assessing LIDs in non-human primates (NHPs), the benchmark PD model^40,41.

Commonly found in the parkinsonian rodent literature are LID, or AIM rating scales that involve exclusively examining the occurrence or frequency of (1) limb dyskinesias (e.g., rhythmic jerks of the forelimb contralateral to the lesion³³, or rapid, purposeless movement of the forelimb controlled by the lesioned hemisphere³⁹); (2) axial dyskinesias (e.g., torsion of axial muscles affecting the neck, trunk and tail³³, or dystonic twisting of the neck and torso contralateral to the lesioned hemisphere³⁹); and (3) orolingual dyskinesias (e.g., masticatory movements of the empty mouth with tongue protrusions³³, or repetitive mastication or tongue protrusions when the rat's mouth was empty and not in contact with any object³⁹). These three behavioral categories are given a severity score based on their frequency during the observation period. The severity score representing the amount of time an animal spent exhibiting these behaviors, frequently without characterization of the quality of behavior, is then combined into the final ALO (axial, limb, and orolingual) score.

In contrast, the current protocol provides a more detailed scale that allows not only the evaluation of global AIMs, but also enables examination of AIMs in independent body parts (i.e., neck, trunk, forelimb, hindlimb, and mouth), the intensity/quality of the AIM being assessed, differentiation of dystonia from hyperkinetic behaviors, and, as with other scales, incorporates the frequency of any given AIM's occurrence. The overall goal is to allow a comprehensive assessment of the LID behavior expressed in each subject. For example, for the single behavioral readout of trunk dystonia, a rat exhibiting mild trunk dystonia (i.e., 45° difference between upper and lower torso, not predictive of turning behavior) for an entire observation period is very different than a rat exhibiting severe trunk dystonia (i.e., constant severe dystonia/twisting, corkscrew-like posturing, approaching 180°; head and feet in opposite directions, unable to ambulate) for the entire observation. Further, given the topographic representation of the rat's body in the striatum, differentiating individual attributes of LIDs can be informative³⁶.

Another component of this protocol to consider carefully is the experimental design of escalating L-DOPA dosing over an extended (e.g., 8-week) treatment time course (Figure 1). A study employing only low doses of L-DOPA (e.g., 3–6 mg/kg) and/or administering the dose of choice for a short period (e.g., 2–4 weeks) may show evidence of LID amelioration with therapeutic intervention in parkinsonian rats; however, these studies' relevance to individuals with PD, where chronic administration and dose escalation are often necessitated, is questionable. Such divergent treatment protocols between pre-clinical and clinical investigations could be suggested to underlie the lack of efficacy seen in clinical trials undertaken with drugs that showed promise in pre-clinical rodent studies.

While the protocol presented here provides instructions on a comprehensive approach for rating LID in parkinsonian rats and the rat model of LID has been established as a reliable model of clinical LID^{3,15,38,42,43}, there are inherent limitations to any model. Animal models are tools that are useful in emulating particular attributes of human disease and making predictions about, for example, the impact that therapeutic intervention might have on a set of disease features. One feature of pre-clinical LID models commonly debated as a potential limitation is that rats require significant DA depletion to express LID. In addition, it is often cited that LID develops more rapidly in rodent and NHP models than in patients and thus these model characteristics do not reflect clinical LID. However, it is often underappreciated that regardless of species (i.e., human, non-human primate, or rodent), near-complete loss of striatal DA innervation is generally required for LID to manifest. Once L-DOPA is administered to subjects with severe striatal DA depletion, LID is induced^{1,19,38,44,45,46}. A second feature that has fostered skepticism toward rodent models of LID comes from the expectation that the neurological signs of LID should resemble the physical manifestations of LID seen in primates (i.e., human and non-human). Indeed, the appearance of the abnormal movements classified as LIDs in rats varies in features from that seen in primates (e.g., choreoathetosis in primates vs. stereotypy/hyperkinesia in rodents). While it is beyond this article to comprehensively address this topic, briefly, such differences in signs between species are based on the fact that humans are habitual bipedal, non-primates frequent bipedal, and rodents quadrupedal^43,46. Thus, various species have specific repertoires of behaviors manifest by specific osteoarticular and muscular structures and modulated by species-specific neural systems⁴³. Accordingly, animal modeling of human-like symptoms should be based primarily on an expectation of functional similarity rather than on physical identity⁴³. Interpretation of findings from animal models with due prudence has, and will continue to provide valuable predictions for disease treatment.

While there is currently one FDA-approved drug for the treatment of LID in PD, amantadine, and the surgical intervention of DBS can ameliorate LID, the efficacy and tolerability are not optimal, and not all patients will qualify for DBS surgery. As eloquently reviewed by Cenci and colleagues⁴⁷, there is a notion expressed by some that individuals afflicted with PD would rather be ON (i.e., experiencing the motor benefit of L-DOPA) with dyskinesias than OFF. As these authors poignantly state: "The deeper truth is that patients would very much prefer to be ON without dyskinesia. As researchers and clinicians, we should aspire to make that goal a reality. To this end, translational research on LID is to be encouraged and persistently pursued." In pursuing this goal, we present our method of induction and rating of LID. Our rating scale is designed to combine a range of intensity measures related to the quality/severity of the AIMs and indicates the amount of time a variety of attributes of dyskinesia are displayed. The intention of this model is to allow the determination of a numerical value that accurately reflects LID severity while accounting for both intensity and frequency. Furthermore, this value indicates LID severity not only systemically but in reference to specific body parts. With this model, we aim to ensure that experimental therapeutics can be rigorously tested for their ability to modify levodopa-induced dyskinesias.