Brain dopamine plays a critical role in motor control. This is most clearly exemplified by the motor symptoms of Parkinson Disease (PD), in which brain dopamine levels are reduced. PD is defined by tremor, rigidity, bradykinesia, and postural instability. Bradykinesia and rigidity consistently improve with dopamine replacement, although tremor and postural instability may not. PD patients also experience significant disability from impaired manual dexterity, which causes difficulty with tasks like tying shoelaces, fastening buttons, and handwriting (Pohar and Allyson Jones, 2009). This symptom is distinct from bradykinesia (Foki et al., 2016), but also responds to dopamine replacement (Gebhardt et al., 2008; Lee et al., 2018). Thus, dopamine plays an important, but poorly defined, role in dexterous skill beyond simply regulating movement speed or amplitude.

Two leading hypotheses regarding the role of dopamine in motor control are that it ‘invigorates’ movement and regulates motor learning. The ‘vigor’ hypothesis derives from the exquisite dopa-responsiveness of bradykinesia in PD and is supported by extensive experimental evidence. Intrastriatal infusion of dopamine agonists increases locomotion, and both electrical and optogenetic stimulation of midbrain dopamine neurons cause contraversive turning (Arbuthnott and Ungerstedt, 1975; Saunders et al., 2018). Dopamine signaling increases near movement onset and acceleration bouts (Coddington and Dudman, 2018; Howe and Dombeck, 2016; Jin and Costa, 2010; Schultz et al., 1983) and is correlated with movement velocity (Barter et al., 2015; Saunders et al., 2018). Conversely, dopamine depletion and dopamine receptor blockade slow movement (Leventhal et al., 2014; Panigrahi et al., 2015). These studies used scalar readouts that reflect ‘vigor’ (e.g. movement velocity or numbers of rotations), and therefore could not assess dopaminergic influences on multi-joint coordination.

Dopaminergic roles in reinforcement learning may contribute to ‘non-vigor’ aspects of motor control. Phasic dopamine release patterns are broadly consistent with ‘reward prediction error’ (RPE) signals, or the difference in value between anticipated and realized behavioral states (Glimcher, 2011). In reinforcement learning models, the RPE is used to adjust subsequent behavior. While the details of dopamine’s role in implicit learning remain to be fully elucidated (Schultz, 2019), dopamine signaling clearly influences synaptic plasticity and alters future behavior (Dowd and Dunnett, 2005; Leventhal et al., 2014; Mohebi et al., 2019; Parker et al., 2016; Shen et al., 2008). Most evidence for ‘learning’ models of dopamine function come from behavioral tasks that require no movement (e.g. classical conditioning, Tobler et al., 2005), simple movements (e.g. lever presses, Parker et al., 2016), or innate movements (e.g. locomotion, Howe and Dombeck, 2016). For the most part, such tasks have discrete outcomes (e.g. push the right or left lever, initiate locomotion or not). However, dopaminergic roles in instrumental and classical conditioning may extend to tasks with more degrees of freedom. In support of this hypothesis, dopamine neuron firing patterns consistent with RPEs (more accurately, performance prediction errors) are observed in songbirds receiving distorted audio feedback (Gadagkar et al., 2016). In mice, rotarod performance worsens gradually during dopamine receptor blockade, and improves gradually when the blockade is released (Beeler et al., 2012). These results could be explained by dopamine reinforcing specific, successful actions (e.g. paw adjustments on the rotarod) to gradually improve performance (Beeler et al., 2013). Nonetheless, the role of dopamine in skilled, dexterous movements requiring precise multi-joint coordination remains unclear.

The goal of this study was to determine the effects of precisely timed dopaminergic manipulations on a complex, finely coordinated, and relatively unconstrained motor skill. To do this, we optogenetically stimulated or inhibited midbrain dopamine neurons as rats performed a skilled reaching task. In skilled reaching, rats learn the coordinated forelimb and digit movements to reach for, grasp, and consume sugar pellets. Skilled reaching is readily learned by rats over several sessions (Klein et al., 2012; Lemke et al., 2019), requires precise coordination between the forelimb and digits, and is sensitive to dopamine depletion (Hyland et al., 2019; Whishaw et al., 1986). It is therefore an excellent model for assessing dopaminergic contributions to dexterous skill.

By combining skilled reaching, optogenetics, and measurement of three-dimensional paw/digit kinematics, we addressed the following questions. First, we asked whether dopamine manipulations affect current or subsequent reaches. If dopamine affects only the current movement, reach kinematics should change immediately with dopamine manipulations. Conversely, if dopamine provides a teaching signal for fine motor coordination, reach kinematics should depend on the history of prior dopaminergic activation. Second, we asked how reach kinematics – specifically coordination between forelimb and digit movements – are influenced by dopamine manipulations. If dopamine plays a purely ‘invigorating’ role in movement, altered dopaminergic signaling should affect only the velocity or amplitude of the reaches.

Instead of pure vigor or learning roles for dopamine, we found a complex pattern of dopaminergic influences on skilled reaching. Consistent with a motor learning function, reach kinematics changed gradually with repeated dopamine neuron stimulation or inhibition. In addition to simple kinematic measures (e.g. reach amplitude), coordination between paw advancement and digit movements also changed with repeated stimulation/inhibition. However, once established, rats transitioned between aberrant and baseline reach kinematics within a single trial in a dopamine-dependent manner. These results indicate that dopamine has both immediate and long-term effects on motor control beyond simply invigorating movement, with important implications for understanding dopamine-linked movement disorders.

We optogenetically stimulated or inhibited substantia nigra pars compacta (SNc) dopamine neurons at specific moments during rat skilled reaching. Tyrosine hydroxylase (TH)-Cre⁺ rats were injected bilaterally with a double-floxed channelrhodopsin (ChR2), archaerhodopsin (Arch), or control EYFP construct into SNc (Figures 1A, C and E). Rats were trained on an automated skilled reaching task that allows synchronization of high-speed video with optogenetics (Bova et al., 2019; Ellens et al., 2016). Trials started with rats breaking a photobeam at the back of the chamber, which caused a pellet to be delivered in front of the reaching slot (Figure 1D). Rats could make multiple reaches until the pellet delivery arm descended 2 s after the video trigger event. Following training, optical fibers were implanted over SNc contralateral to the rat’s preferred reaching paw. Immunohistochemistry confirmed that opsin expression was restricted to TH-expressing neurons in SNc projecting to striatum (Figure 1F and Figure 1—figure supplement 3).

Figure 1 with 3 supplements see all

Download asset Open asset

Altered SNc dopamine neuron activity gradually changes skilled reaching outcomes

We stimulated or inhibited SNc dopamine neurons during every reach for ten 30-min sessions (Figure 1D, ‘during reach’). Baseline performance did not differ between groups (Figure 2). Dopamine neuron stimulation did not significantly affect the number of trials attempted (Figure 2B), but caused a progressive decline in performance (Figure 2E and Figure 2—figure supplement 2). Success rate on the first reach within each trial decreased to about half of baseline performance during the first day of testing, then to nearly 0% for the remainder of ‘Laser On’ sessions. The number of reaches per trial increased as success rate declined, but even with multiple attempts per trial, success rate decreased (Figure 2—figure supplement 2). The progressive decline in performance led us to ask whether success rate also changed across trials within individual sessions. Indeed, during the first ‘Laser On’ session, success rate progressively declined (Figure 2G, Figure 2—figure supplement 1). Furthermore, dopamine-dependent changes in reach success persisted into subsequent sessions. Therefore, reaching performance depends on the history of dopamine neuron activation during skilled reaching.

Figure 2 with 2 supplements see all

Download asset Open asset

Figure 2—source data 1

A .mat file containing number of trials (num_trials) and first reach success rate (firstReachSuccess) for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

Data is average per session for each rat. The field ‘experimentInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig2-data1-v2.mat

Download elife-61591-fig2-data1-v2.mat

Figure 2—source data 2

A .mat file containing first reach success rate averages across a moving block of 10 trials for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

The field ‘exptInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig2-data2-v2.mat

Download elife-61591-fig2-data2-v2.mat

Figure 2—source data 3

Statistics.

A.xlsx file containing the statistical output of Wilcoxon ranksum tests for comparisons between ChR2 During and EYFP group averages in Figure 2G.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig2-data3-v2.xlsx

Download elife-61591-fig2-data3-v2.xlsx

Because dopamine stimulation during reaching caused a gradual decline in performance, we asked if reaching performance would recover gradually when dopamine stimulation was removed. Animals were tested for an additional 10 days with the same laser stimulation protocol, but with the patch cable-optical fiber junction physically occluded (‘occlusion’ sessions, Figure 1B). Thus, all cues were identical (e.g., optical shutter noise, visible light) except light penetration into the brain. Reaching performance recovered quickly, but not immediately, to pre-stimulation levels (Figure 2E,G). However, there was significant variability between rats in the rate of recovery (Figure 2—figure supplement 1). On average, recovery to baseline performance was faster than the decline in performance with initial dopamine stimulation (contrast testing, t(583.8) = 2.55, p=0.011). This is further evidence that the history of dopaminergic activation influences subsequent skill execution.

We next asked if dopamine stimulation must occur during reaches to affect success rate. A separate group of ChR2-expressing TH-Cre⁺ rats received laser stimulation during the intertrial interval for a duration matched to ‘during reach’ stimulation (Figure 1D, ‘between reach’, Figure 1—figure supplement 1). Dopamine neuron stimulation between reaches did not significantly affect the number of trials (Figure 3A) or success rate (Figure 3B,C and Figure 3—figure supplements 1,2). Therefore, dopamine neuron stimulation must occur as the rat is reaching to affect subsequent reaching performance. This result has two important implications. First, it suggests that skill performance depends on the history of striatal dopamine levels specifically during performance of that skill. Second, it argues against the possibility that the effects of dopamine neuron stimulation are due to the gradual accumulation of striatal dopamine.

Figure 3 with 2 supplements see all

Download asset Open asset

Figure 3—source data 1

A .mat file containing number of trials (num_trials) and first reach success rate (firstReachSuccess) for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

Data is average per session for each rat. The field ‘experimentInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig3-data1-v2.mat

Download elife-61591-fig3-data1-v2.mat

Figure 3—source data 2

A .mat file containing first reach success rate averages across a moving block of 10 trials for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

The field ‘exptInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig3-data2-v2.mat

Download elife-61591-fig3-data2-v2.mat

Figure 3—source data 3

Statistics.

A .xlsx file containing the statistical output of Wilcoxon ranksum tests for comparisons between ChR2 During and ChR2 Between group averages in Figure 3C.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig3-data3-v2.xlsx

Download elife-61591-fig3-data3-v2.xlsx

Dopamine neuron inhibition during reaching did not affect success rate (Figure 4C,E and Figure 4—figure supplements 2,3). However, dopamine neuron inhibition significantly decreased the number of trials per session (Figure 4A), consistent with a role for midbrain dopamine in motivation to work for rewards (Palmiter, 2008; Salamone and Correa, 2012). It is not clear why dopamine neuron stimulation did not have the opposite effect (Figures 2B,3A), but increases in trial numbers may be limited by a ceiling effect - rats initiate new trials quickly, even in the absence of stimulation. The inhibition-related decrease in trials was gradual, with rats progressively performing fewer reaches across sessions. Dopamine neuron inhibition between reaches had no effect on success rate (Figure 4D and Figure 4—figure supplement 1) or the number of reaches performed in each session (Figure 4B). Control rats injected with constructs expressing EYFP but no opsin did not experience any changes in task performance (Figure 2C,F,G and Figure 2—figure supplements 1,2).

Figure 4 with 3 supplements see all

Download asset Open asset

Figure 4—source data 1

A .mat file containing number of trials (num_trials) and first reach success rate (firstReachSuccess) for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

Data is average per session for each rat. The field ‘experimentInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig4-data1-v2.mat

Download elife-61591-fig4-data1-v2.mat

Figure 4—source data 2

A .mat file containing first reach success rate averages across a moving block of 10 trials for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

The field ‘exptInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig4-data2-v2.mat

Download elife-61591-fig4-data2-v2.mat

Figure 4—source data 3

Statistics.

A .xlsx file containing the statistical output of Wilcoxon ranksum tests for comparisons between Arch During and EYFP group averages in Figure 4E.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig4-data3-v2.xlsx

Download elife-61591-fig4-data3-v2.xlsx

Dopamine manipulations induce progressive changes in reach-to-grasp kinematics

The success rate analysis indicates that repeated dopaminergic stimulation progressively diminished reaching performance, but does not explain why performance worsened. To determine which aspects of reach kinematics were altered by dopaminergic manipulations, we used Deeplabcut to track individual digits, the paw, and the pellet (Figure 5; Bova et al., 2019; Mathis et al., 2018).

Figure 5

Download asset Open asset

Figure 5—source data 1

A .mat file containing paw trajectories (reachData.pd_trajectory) and individual digit trajectories (reachData.dig_trajectory) for each trial in a single ‘retraining’ session for a ‘ChR2 During’ rat.

X, Y, and Z coordinates are provided for each trajectory.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig5-data1-v2.mat

Download elife-61591-fig5-data1-v2.mat

Figure 5—source data 2

A .mat file containing average paw (ratSummary.mean_pd_trajectory) and digit (ratSummary.mean_dig_trajectories) trajectories across all 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’) for the same rat as in Figure 5—source data 1.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig5-data2-v2.mat

Download elife-61591-fig5-data2-v2.mat

Consistent with the success rate analysis, dopamine neuron stimulation during reaching caused progressive changes in reach-to-grasp kinematics. Reach extent (how far the paw extended in the direction of the pellet, z_digit2) became progressively shorter with repeated stimulation during reaches (Figure 6A, Videos 1 and 2). This effect may have been stronger for posteromedially located fibers (Figure 6—figure supplement 1). The progressive change in reach extent occurred both across and within sessions, and did not stabilize until the fifth session of dopamine neuron stimulation (Figure 6D and Figure 6—figure supplement 3). Dopamine neuron stimulation during reaches also gradually narrowed grasp aperture at reach end (Figure 6E,F and Figure 6—figure supplement 4), caused the paw to be more pronated at reach end (i.e. θ decreased) (Figure 6G,H and Figure 6—figure supplement 5), and decreased the maximum reach velocity (Figure 6I,J and Figure 6—figure supplement 6). Kinematic measures continued to change even when success rate had plateaued (compare Figures 2E,6) due to a ‘floor effect’ for success rate – once the rat consistently misses the pellet, no further changes are detectable by this measure. The decrease in reach extent accounted for much of the drop in success rate (Figure 6—figure supplement 7), although reaches matched for reach extent were still less successful during ‘laser on’ than ‘occlusion’ sessions (Figure 6—figure supplement 8). This suggests that paw transport and grasp-related factors both contributed to reach-to-grasp failures. When dopamine stimulation ceased (‘occlusion’ sessions), reach-to-grasp kinematics rapidly returned to baseline. As with success rate, there was individual variability in how quickly rats returned to pre-stimulation kinematics (Figure 6A,D–J and Figure 6—figure supplements 3–6). All reach-to-grasp kinematics were unchanged in rats receiving dopamine neuron stimulation between reaches and EYFP control rats (Figure 6B–D,F,H,J and Figure 6—figure supplements 2–8). In addition to histology, we verified opsin expression and fiber placement by performing ‘during reach’ stimulation in rats previously stimulated between reaches. All rats showed kinematic changes with ‘during reach’ stimulation not observed with ‘between reach’ stimulation. This served as a positive control and reinforces the importance of the timing of dopamine neuron stimulation with respect to specific actions.

Figure 6 with 8 supplements see all

Download asset Open asset

Figure 6—source data 1

A .mat file containing maximum reach extent of digit2 (mean_dig2_endPt), aperture at reach end (mean_end_aperture), paw orientation at reach end (mean_end_orientations), and mean paw velocity (mean_pd_v) for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

Data is average per session for each rat. The field ‘experimentInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig6-data1-v2.mat

Download elife-61591-fig6-data1-v2.mat

Figure 6—source data 2

A .mat file containing digit2 endpoint (digEnd), aperture at reach end (aperture), orientation at reach end (orientation), and velocity (velocity) averages across a moving block of 10 trials for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

The field ‘exptInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig6-data2-v2.mat

Download elife-61591-fig6-data2-v2.mat

Figure 6—source data 3

Statistics.

A .xlsx file containing the statistical output of Wilcoxon ranksum tests for comparisons between ChR2 During and EYFP group averages in Figures 6D, F, H and J.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig6-data3-v2.xlsx

Download elife-61591-fig6-data3-v2.xlsx

Video 1

Download asset

Video 2

Download asset

While dopamine neuron inhibition during reaching did not affect success rate (Figure 4C), it caused subtle changes in reach-to-grasp kinematics. Maximum reach extent lengthened slightly under dopamine neuron inhibition (that is, the paw extended further past the pellet, Figure 7A,C, Videos 3 and 4), in opposition to the effects of dopamine neuron stimulation. This effect almost reached significance in the linear mixed-effect model (p=0.091, see Figure 7 caption), but a contrast test comparing laser day 10 to occlusion day one was significant (t(37) = −3.24, p=0.003). Furthermore, reach extent consistently lengthened at the individual rat level (Figure 7A, gray markers) as well as across trials within sessions (Figure 7C, Figure 7—figure supplement 3). Maximum reach velocity also decreased with dopamine inhibition (Figure 7H,I). This was not quite significant in the linear mixed-effect model (p=0.094, see Figure 7 caption), but there was a significant difference between laser day 10 and occlusion day 1 (contrast testing, t(33) = −2.49, p=0.018). These data suggest that dopamine neuron stimulation and inhibition have roughly opposite effects on reach kinematics. Dopamine neuron inhibition did not significantly affect grasp aperture (Figure 7D,E) or paw orientation (Figure 7F,G), potentially due to ceiling effects. None of these measures differed between successful and failed reaches (Figure 7—figure supplement 7). No kinematic changes were observed in rats that received dopamine neuron inhibition between reaches (Figure 7B and Figure 7—figure supplements 2–8).

Figure 7 with 8 supplements see all

Download asset Open asset

Figure 7—source data 1

A .mat file containing maximum reach extent of digit2 (mean_dig2_endPt), aperture at reach end (mean_end_aperture), paw orientation at reach end (mean_end_orientations), and mean paw velocity (mean_pd_v) for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

Data is average per session for each rat. The field ‘experimentInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig7-data1-v2.mat

Download elife-61591-fig7-data1-v2.mat

Figure 7—source data 2

A .mat file containing digit2 endpoint (digEnd), aperture at reach end (aperture), orientation at reach end (orientation), and velocity (velocity) averages across a moving block of 10 trials for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

The field ‘exptInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig7-data2-v2.mat

Download elife-61591-fig7-data2-v2.mat

Figure 7—source data 3

Statistics.

A .xlsx file containing the statistical output of Wilcoxon ranksum tests for comparisons between Arch During and EYFP group averages in Figures 6D, F, H and J.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig7-data3-v2.xlsx

Download elife-61591-fig7-data3-v2.xlsx

Video 3

Download asset

Video 4

Download asset

Dopamine manipulations disrupt reach-to-grasp coordination

Reach-to-grasp success requires precise coordination of a complex sequence of reach sub-movements. Reaches begin when the rat orients to the pellet with its nose, then lifts and aligns its paw at midline with the digits closed. As the forelimb advances towards the pellet, the digits extend and spread while the paw pronates. After the digits close to grasp the pellet, the forelimb and paw are raised and supinated to bring the pellet toward the mouth (Alaverdashvili and Whishaw, 2010; Whishaw and Pellis, 1990; Whishaw et al., 2008). Because fine motor coordination is impaired in patients with PD, including during reaching-to-grasp (Whishaw et al., 2002), we looked to see if the coordination of reach sub-movements was affected by dopamine neuron stimulation or inhibition.

Dopamine neuron stimulation during reaching altered the coordination of digit spread (aperture) and paw pronation (orientation) with respect to paw advancement (Figures 8,9). Aperture increased earlier (when the paw was further from the pellet) in ‘during reach’ stimulation sessions compared to ‘retraining’ or ‘occlusion’ sessions (Figure 9A,B). Thus, during dopamine stimulation, aperture was smaller at reach end but larger (on average) at matched distances from the pellet (Figures 6C,8,9E). ‘During reach’ dopamine neuron inhibition had the opposite effect – paw aperture began to increase when the paw was closer to the pellet compared to ‘retraining’ or ‘occlusion’ sessions (Figure 9A,B). Similar changes occurred with paw orientation: during sessions with dopamine neuron stimulation, paw pronation began further from the pellet (Figure 9D,E,F). Dopamine neuron inhibition, however, did not affect the relationship between paw orientation and paw advancement. As for other kinematic effects, the changes in coordination progressed across sessions (most evident in Figure 9C,F). No changes were observed in rats that received dopamine stimulation or inhibition between reaches or in EYFP control rats (Figure 9B,C,E,F and Figure 9—figure supplements 1–5). Together, these results suggest that dopamine neuron stimulation accelerates transitions between reach sub-movements, while dopamine neuron inhibition has the opposite effect.

Figure 8

Download asset Open asset

Figure 8—source data 1

A zip file containing .mat files to produce Figure 8.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig8-data1-v2.zip

Download elife-61591-fig8-data1-v2.zip

Figure 9 with 5 supplements see all

Download asset Open asset

Figure 9—source data 1

A .mat file containing aperture (mean_aperture_traj) and paw orientation (mean_orientation_traj) data along reach trajectories (see Materials and methods for details of calculations) for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

Data is average per session for each rat. The field ‘experimentInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig9-data1-v2.mat

Download elife-61591-fig9-data1-v2.mat

Dopamine neuron stimulation establishes distinct reach-to-grasp representations

Dopamine neuron stimulation gradually induced changes in reach-to-grasp kinematics, but kinematics rapidly recovered to baseline when the laser was occluded. We next asked if reinstating dopamine neuron stimulation would again gradually alter reach kinematics. Following testing with the laser occluded, ChR2-injected rats that had received ‘between reach’ stimulation performed additional ‘during reach’ stimulation sessions (Figure 10A). These continued until reach-to-grasp kinematics were impaired (average: 3.17 ± 0.98 sessions) (Figure 10B, C, Figure 10—figure supplements 1,2). Once kinematics were impaired, rats performed an additional one or two 30-min sessions during which the laser alternated every five trials between being off and on during reaches (Figure 10D).

Figure 10 with 2 supplements see all

Download asset Open asset

Figure 10—source data 1

A .mat file containing zdigit2 endpoint (mean_dig2_endPt) data for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

Data is average per session for each rat. The field ‘experimentInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig10-data1-v2.mat

Download elife-61591-fig10-data1-v2.mat

Figure 10—source data 2

A .mat file containing zdigt2 endpoint (mean_dig2_endPt) data for ‘during reach’ sessions in originally ‘ChR2 Between’ rats.

Data is average per session for each rat.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig10-data2-v2.mat

Download elife-61591-fig10-data2-v2.mat

Figure 10—source data 3

A .mat file containing digit2 endpoint (digEnd) averages across a moving block of 10 trials for 22 testing sessions (‘retraining’, ‘laser on’, and ‘occluded’).

The field ‘exptInfo’ provides information on groups.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig10-data3-v2.mat

Download elife-61591-fig10-data3-v2.mat

Figure 10—source data 4

A .mat file containing digit2 endpoint (digEnd) averages across a moving block of 10 trials for ‘during reach’ sessions in originally ‘ChR2 Between’ rats.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig10-data4-v2.mat

Download elife-61591-fig10-data4-v2.mat

Figure 10—source data 5

A .mat file containing ‘5off/5on’ data.

Data includes digit2 endpoint (dig_endPoints), aperture, orientation, and velocity (max_pd_v).

https://cdn.elifesciences.org/articles/61591/elife-61591-fig10-data5-v2.mat

Download elife-61591-fig10-data5-v2.mat

Rats transitioned rapidly between ‘normal’ and ‘impaired’ reach-to-grasp kinematics with the laser off and on, respectively. The mean success rate dropped within a single trial of laser stimulation and improved within one trial when laser stimulation was removed (Figure 10F, Video 5). Similarly, reach kinematics required only one trial to switch between normal and aberrant reaching patterns. Laser stimulation at the beginning of an ‘on’ block (trial 1) caused immediate decreases in maximum reach extent and digit aperture, which remained steady for the remaining ‘Laser On’ trials. Similarly, maximum reach extent and digit aperture immediately increased upon cessation of dopamine neuron stimulation (trial 1, ‘Laser Off’) and remained steady throughout the ‘Laser Off’ block (Figure 10G,H). There was also a significant change in paw orientation at reach end with dopamine neuron stimulation (Figure 10I). However, pronation decreased in these rats unlike in the ‘ChR2 During’ group (Figure 6G). There was no significant difference in maximum reach velocity between ‘Laser On’ and ‘Laser Off’ blocks (Figure 10J). These data indicate that once distinct reaching kinematics have been established by repeated dopaminergic manipulations, current reach kinematics are determined by the activity of nigral dopamine neurons on that trial.

Video 5

Download asset

Dopamine neuron stimulation induces context- and history-dependent abnormal involuntary movements

To verify fiber placement and opsin expression prior to reaching experiments, we placed rats in a clear cylinder and illuminated SNc with blue light of varying intensity (Figure 11). We predicted that rats with well-placed fibers expressing high levels of ChR2 would develop increasingly worse abnormal involuntary movements (AIMs) as laser intensity increased. To our surprise, rats that subsequently developed markedly abnormal reach kinematics during the skilled reaching task appeared unaffected by dopamine neuron stimulation in the cylinder (AIMs Test 1, Figure 11). In ‘post-reaching’ cylinder sessions (AIMS Test 2, Figure 11), however, dopamine neuron stimulation elicited markedly abnormal movements (Figure 11, Video 6). While AIMs were obvious in the context of the cylinder, the same (or higher) stimulation intensities delivered while rats were reaching rarely elicited abnormal movements (other than altered reach kinematics, Figure 11A, Video 7). Only three of six ‘ChR2 during’ rats exhibited any AIMs superimposed on reaches during ‘laser on’ day 10 (2 rats each with AIMs in 1 of 10 reaches, one rat with AIMs in 5 of 10 reaches), and none exhibited dyskinesias during reaching on laser day 2 when success rate was already near zero. Thus, while dyskinesias likely accounted for some reaching deficits, that was not the primary failure mechanism. Furthermore, rats that did not experience dyskinesias in the cylinder still had markedly reduced reaching success, and rats that experienced dyskinesias could have preserved success rates (Figure 11—figure supplement 1). Therefore, the expression of dopamine-dependent AIMs depends not only on current levels of dopamine neuron activation, but the history of prior activation and the current behavioral context.

Figure 11 with 1 supplement see all

Download asset Open asset

Figure 11—source data 1

A .mat file containing number of rotations to the left (leftSpin) and right (rightSpin), axial AIMs scores (axialAmplitude and axialBasis), and limb AIMs scores (limbAmplitude and limbBasic) from AIMs tests 1 and 2.

Field 1 = ChR2 During test 1; Field 2 = ChR2 Between test 1; Field 3 = EYFP test 1; Field 4 = ChR2 During test 2; Field 5 = ChR2 Between test 2; Field 6 = EYFP test 2.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig11-data1-v2.mat

Download elife-61591-fig11-data1-v2.mat

Figure 11—source data 2

A .mat file containing limb and axial amplitude scores for randomly selected skilled reaching trials.

Data is scores for each trial per rat.

https://cdn.elifesciences.org/articles/61591/elife-61591-fig11-data2-v2.mat

Download elife-61591-fig11-data2-v2.mat

Video 6

Download asset

Video 7

Download asset

Our goal was to determine how midbrain dopamine neuron manipulations affect dexterous skill. Our results revealed a role for dopamine in motor learning, as repeated dopamine manipulations induced gradual changes in reach-to-grasp kinematics. These manipulations not only affected gross performance measures (e.g. velocity and amplitude) but also disrupted coordinated execution of reach sub-movements. Once dopamine stimulation-induced changes were established, reach-to-grasp kinematics depended strongly on the dopamine status of the current trial. Furthermore, these effects were temporally specific – only manipulations during reaches influenced forelimb kinematics. Finally, the effect of dopamine on motor control is context-dependent, as the same dopamine stimulation in the cylinder and reaching chamber induced distinct behavioral responses.

The history-dependent effects of dopamine on skilled reaching are superficially consistent with reinforcement learning models (Schultz, 2019). While most evidence for phasic dopamine signals encoding RPEs comes from paradigms in which animals choose between discrete actions (e.g. press a right or left lever), recent studies suggest that dopamine encodes RPE-like ‘performance prediction errors’ for complex behaviors with greater degrees of freedom (Beeler et al., 2010; Gadagkar et al., 2016). It is plausible that dopamine neuron excitation/inhibition creates an artificially reinforcing/discouraging signal that influences subsequent reaches. At baseline and in EYFP controls, failed reaches were slightly (but not significantly) shorter than successful reaches (Figure 6—figure supplement 7). Dopamine neuron stimulation could have gradually reinforced this subtle difference until it became measurable. On the other hand, we stimulated or inhibited dopamine neurons on every trial, so that ‘short reach’ and ‘long reach’ reinforcement should cancel each other. Furthermore, stimulation and inhibition had consistent, almost opposite effects. Since there are many possible failure mechanisms, we would expect variable and unpredictable kinematic changes if dopamine neuron stimulation was artificially reinforcing failed reaches. Together, these findings suggest that kinematic changes do not result purely from performance prediction error signals, but that dopamine intrinsically biases movement kinematics in a consistent direction. Closed loop experiments in which manipulations occur only for ‘long’ or ‘short’ reaches may definitively separate these possibilities (Yttri and Dudman, 2016).

Motion tracking data provide insight into the nature of an intrinsic dopamine bias. A common interpretation of dopamine’s role in movement is that it regulates ‘vigor,’ which has been defined as the speed, frequency, and amplitude of movements (Dudman and Krakauer, 2016). Our dopamine manipulations influenced ‘vigor’ in unexpected ways: dopamine neuron stimulation decreased, and inhibition increased, movement amplitude (reach extent). Furthermore, both stimulation and inhibition decreased movement speed. These effects apparently contradict previous work directly correlating dopaminergic tone with movement velocity and/or amplitude (Carr and White, 1987; Leventhal et al., 2014; Panigrahi et al., 2015).

This discrepancy may be due to different demands on the motor system. ‘Vigor’ assays generally demand movement along one dimension. For example, mice manipulating a joystick (Panigrahi et al., 2015), or humans moving a manipulandum to a target (Baraduc et al., 2013; Mazzoni et al., 2007) make forelimb/arm movements across large joints more or less along a single vector. In such tasks, dopamine-depleted subjects consistently make hypometric, bradykinetic movements. In contrast, skilled reaching comprises a sequence of precisely coordinated submovements (Klein and Dunnett, 2012). Stimulation caused paw pronation and digit spread to occur earlier along the reach trajectory (farther from the pellet), while inhibition delayed digit spread. Similarly, humans with PD reshape their hands closer to the target than control subjects (Rand et al., 2006; Whishaw et al., 2002). This is consistent with dopamine regulating initiation of, and transitions between, movements (da Silva et al., 2018; Hamilos et al., 2020; Howe and Dombeck, 2016). Within a ‘vigor’ framework, dopamine may invigorate the next submovement at the expense of the current one, compressing the overall sequence. In the limiting case, overlapping submovements could manifest as muscle cocontractions and dystonia, a possibility supported by the rare occurrence of abnormal movements intruding into reaches in the most severely affected rats. Simultaneous EMG recordings from multiple muscles could address this possibility (Hyland and Jordan, 1997).

The mechanisms by which forelimb-digit coordination occurs normally are unclear, let alone under dopamine perturbations (Leib et al., 2020). For ballistic movements, it is suggested that internal models predict consequences of motor commands to allow rapid feedback for online corrections (Azim and Alstermark, 2015). Such models necessarily integrate sensory feedback (Azim et al., 2014) with the passage of time (Crevecoeur and Gevers, 2019). Dopamine and striatal circuitry have been implicated in time estimation, though on a slower timescale than skilled reaching (Hamilos et al., 2020; Mello et al., 2015). Accelerated time perception (i.e. perceiving more time has passed than actually has) should cause premature submovement transitions, although some studies indicate that dopamine alters time perception in the opposite direction (Soares et al., 2016). Others suggest that dopamine regulates integration of multiple information streams (including time) for decision making (Howard et al., 2017). It is therefore plausible that dopamine alters reliance on sensory feedback, time perception, the internal model itself, or a combination of these elements.

Kinematic changes developed gradually, but once established they depended on the dopamine status of the current trial (Figure 10). Rats previously stimulated during reaching experienced larger trial-by-trial stimulation-dependent changes in reach kinematics than stimulation-naive rats or rats that had only received ‘between-reach’ stimulation (Figure 10, Video 5). This ‘skilled reaching sensitization’ shares attributes with sensitization to dopaminergic drugs (e.g. amphetamine, cocaine, and levodopa), in which repeated administration renders rodents hypersensitive to subsequent doses (Robinson and Becker, 1986). While the mechanisms of sensitization to dopaminergic drugs are not fully understood, enhanced dopaminergic and glutamatergic signaling are implicated (Cadoni et al., 2000; Zhang et al., 2001). Similar mechanisms could mediate skilled reaching sensitization, possibly in dorsal ‘motor’ striatum.

An important difference between sensitization to drugs and optogenetic stimulation is the specificity of the latter to movements performed during stimulation. Rats that had been stimulated between reaches were initially unaffected by ‘during reach’ stimulation (Figure 10A–C, Figure 10—figure supplements 1,2, Video 5), arguing against a general sensitizing effect. Cortico- or thalamo-striatal transmission may be sensitized specifically at synapses active during elevated dopamine signaling (Yttri and Dudman, 2018). Consistent with this idea, subpopulations of direct pathway medium spiny neurons are associated with dyskinesias after levodopa treatment in dopamine-depleted mice (Ryan et al., 2018), and selective activation of these direct pathway MSNs induces dyskinesias (Girasole et al., 2018). These results suggest that subpopulations of striatal output neurons encode specific dopamine-sensitive movements. In our experiments, the specific population of ‘sensitized’ MSNs would be those active during reaching.

The abrupt transitions between aberrant and baseline reach kinematics are reminiscent of ‘on/off’ motor fluctuations observed in people with PD. With disease progression and prolonged treatment, patients often display sudden transitions between severe bradykinesia, good motor control, and levodopa-induced dyskinesias (Chou et al., 2018). Disease duration, degree of dopamine loss, and magnitude of treatment-related dopamine fluctuations are correlated (Abercrombie et al., 1990; de la Fuente-Fernández et al., 2004). The root causes of motor fluctuations are therefore difficult to identify, though it is suggested that ‘on-offs’ in PD may share mechanisms with drug sensitization (Calabresi et al., 2008). Our results indicate that large, temporally specific dopamine fluctuations are sufficient to cause dramatic dopamine-dependent changes in movement kinematics, even in otherwise healthy subjects. This suggests that large swings in striatal dopamine are sufficient to generate motor fluctuations, independent of the degree of dopamine denervation.

The motor effects of dopamine neuron stimulation also depended on behavioral context. Dopamine neuron stimulation had almost no effect on stimulation-naive rats in a clear cylinder. Rats engaged in skilled reaching during dopamine neuron stimulation continued to engage in the task, with few abnormal involuntary movements during reaching (Figure 11). However, the same stimulation parameters delivered to previously-stimulated rats in clear cylinders induced markedly abnormal limb and body movements (Figure 11 and Video 6). Thus, skilled reaching sensitization may have generalized to the cylinder. However, it is unclear if enhanced AIMs required stimulation in the reaching chamber, or if stimulation during the initial AIMs test was sufficient (a single amphetamine dose can induce sensitization given a sufficiently long drug hiatus, Valjent et al., 2010; Vanderschuren et al., 1999). In either case, this is consistent with the idea that dopamine regulates the ‘vigor’ of movements selected based on the current behavioral context (Yttri and Dudman, 2016). That is, in the reaching chamber, rats approach the reaching slot to perform a (dopamine-modified) reach because that is the appropriate action in that context. Conversely, with no specific goal-directed actions suggested by the cylinder context, dopamine equally invigorates many potential movements. This leads to seemingly random abnormal involuntary movements (Bastide et al., 2015). Interestingly, the severity of experimental levodopa-induced dyskinesias depends on behavioral context (Lane et al., 2011). This context dependence of dopaminergic effects on motor control has parallels in clinical phenomenology: people with PD often can perform goal-directed movements despite the presence of significant levodopa-induced dyskinesias.

There are several limitations of this study. First, we did not record from dopamine neurons or measure dopamine release during optogenetic manipulations. It is therefore not clear how striatal dopamine levels were altered relative to normal reach-related dopamine dynamics, or if repeated stimulation changed spontaneous or optically evoked dopamine release (Saunders et al., 2018). Given the relatively high optical stimulation power (20 mW at the fiber tip) and frequency (20 Hz) used, we suspect that we induced supraphysiologic dopamine release (Patriarchi et al., 2018, but see Hamid et al., 2016). Nonetheless, supra/infraphysiologic manipulations (e.g. lesion studies) can provide important insights into normal function. Furthermore, supraphysiologic dopamine fluctuations are relevant to pathologic states like PD, in which striatal dopamine can transition over minutes to hours between very low and high levels (Abercrombie et al., 1990; de la Fuente-Fernández et al., 2004). Second, we stimulated over SNc. It is therefore unclear how ventral tegmental area (VTA) stimulation would influence skilled reaching, and whether stimulating specific nigral projection fields (e.g. striatal subregions or motor cortex, Guo et al., 2015, Hosp et al., 2011, Zhang et al., 2017) would differentially affect reach kinematics. Finally, while we found that dopamine neuron manipulations during, but not between, reaches affected reach kinematics, the timing of when dopamine manipulations exert their effects could be parsed more precisely. Our ‘during reach’ timing covered approach to the pellet, the reach itself, and immediately after the grasp during pellet consumption. ‘Phasic’ dopamine release at the time of an unpredicted reward is believed to be responsible for reinforcing actions of dopamine, while ‘tonic’ dopamine levels are suggested to regulate motivation/vigor (Niv et al., 2007). Activation of different terminal fields at different times with respect to behavior, as well as continuous monitoring of dopamine release (Patriarchi et al., 2018), may identify distinct roles for tonic and phasic dopamine in dexterous skill.

In summary, temporally specific dopamine signals cause gradual changes in dexterous skill performance separable from pure ‘vigor’ effects. These changes are durable, and expressed in a dopamine-dependent manner on a reach-by-reach basis. This phenomenon has clinical analogy with rapid motor fluctuations in PD patients. It may, therefore, serve as a useful paradigm in which to study the underlying neurobiology of motor fluctuations in PD, as well as address fundamental questions regarding how dopamine and basal ganglia circuits regulate skilled movements.

Original video files and extracted deeplabcut coordinates in csv format are available publicly on figshare (https://doi.org/10.6084/m9.figshare.c.5095484.v1).

1. 1. Leventhal D
   2. Bova A

   (2020) figshare

   Precisely-timed dopamine signals establish distinct kinematic representations of skilled movements.

   https://doi.org/10.6084/m9.figshare.c.5095484.v1

1. 1. Abercrombie ED
   2. Bonatz AE
   3. Zigmond MJ

   (1990) Effects of L-dopa on extracellular dopamine in striatum of normal and 6-hydroxydopamine-treated rats

   Brain Research 525:36–44.

   https://doi.org/10.1016/0006-8993(90)91318-B

   • PubMed
   • Google Scholar
2. 1. Alaverdashvili M
   2. Whishaw IQ

   (2010) Compensation aids skilled reaching in aging and in recovery from forelimb motor cortex stroke in the rat

   Neuroscience 167:21–30.

   https://doi.org/10.1016/j.neuroscience.2010.02.001

   • PubMed
   • Google Scholar
3. 1. Arbuthnott GW
   2. Ungerstedt U

   (1975) Turning behavior induced by electrical stimulation of the nigro-neostriatal system of the rat

   Experimental Neurology 47:162–172.

   https://doi.org/10.1016/0014-4886(75)90244-7

   • PubMed
   • Google Scholar
4. 1. Azim E
   2. Jiang J
   3. Alstermark B
   4. Jessell TM

   (2014) Skilled reaching relies on a V2a propriospinal internal copy circuit

   Nature 508:357–363.

   https://doi.org/10.1038/nature13021

   • PubMed
   • Google Scholar
5. 1. Azim E
   2. Alstermark B

   (2015) Skilled forelimb movements and internal copy motor circuits

   Current Opinion in Neurobiology 33:16–24.

   https://doi.org/10.1016/j.conb.2014.12.009

   • Google Scholar
6. 1. Baraduc P
   2. Thobois S
   3. Gan J
   4. Broussolle E
   5. Desmurget M

   (2013) A common optimization principle for motor execution in healthy subjects and parkinsonian patients

   Journal of Neuroscience 33:665–677.

   https://doi.org/10.1523/JNEUROSCI.1482-12.2013

   • PubMed
   • Google Scholar
7. 1. Barter JW
   2. Li S
   3. Lu D
   4. Bartholomew RA
   5. Rossi MA
   6. Shoemaker CT
   7. Salas-Meza D
   8. Gaidis E
   9. Yin HH

   (2015) Beyond reward prediction errors: the role of dopamine in movement kinematics

   Frontiers in Integrative Neuroscience 9:39.

   https://doi.org/10.3389/fnint.2015.00039

   • PubMed
   • Google Scholar
8. 1. Bastide MF
   2. Meissner WG
   3. Picconi B
   4. Fasano S
   5. Fernagut PO
   6. Feyder M
   7. Francardo V
   8. Alcacer C
   9. Ding Y
   10. Brambilla R
   11. Fisone G
   12. Jon Stoessl A
   13. Bourdenx M
   14. Engeln M
   15. Navailles S
   16. De Deurwaerdère P
   17. Ko WK
   18. Simola N
   19. Morelli M
   20. Groc L
   21. Rodriguez MC
   22. Gurevich EV
   23. Quik M
   24. Morari M
   25. Mellone M
   26. Gardoni F
   27. Tronci E
   28. Guehl D
   29. Tison F
   30. Crossman AR
   31. Kang UJ
   32. Steece-Collier K
   33. Fox S
   34. Carta M
   35. Angela Cenci M
   36. Bézard E

   (2015) Pathophysiology of L-dopa-induced motor and non-motor complications in Parkinson's disease

   Progress in Neurobiology 132:96–168.

   https://doi.org/10.1016/j.pneurobio.2015.07.002

   • PubMed
   • Google Scholar
9. 1. Beeler JA
   2. Cao ZF
   3. Kheirbek MA
   4. Ding Y
   5. Koranda J
   6. Murakami M
   7. Kang UJ
   8. Zhuang X

   (2010) Dopamine-dependent motor learning: insight into levodopa's long-duration response

   Annals of Neurology 67:639–647.

   https://doi.org/10.1002/ana.21947

   • PubMed
   • Google Scholar
10. 1. Beeler JA
    2. Frank MJ
    3. McDaid J
    4. Alexander E
    5. Turkson S
    6. Sol Bernandez M
    7. McGehee DS
    8. Zhuang X

    (2012) A role for Dopamine-Mediated learning in the pathophysiology and treatment of Parkinson’s Disease

    Cell Reports 2:1747–1761.

    https://doi.org/10.1016/j.celrep.2012.11.014

    • Google Scholar
11. 1. Beeler JA
    2. Petzinger G
    3. Jakowec MW

    (2013) The enemy within: propagation of aberrant corticostriatal learning to cortical function in Parkinson's Disease

    Frontiers in Neurology 4:134.

    https://doi.org/10.3389/fneur.2013.00134

    • PubMed
    • Google Scholar
12. 1. Bova A
    2. Kernodle K
    3. Mulligan K
    4. Leventhal D

    (2019) Automated rat Single-Pellet reaching with 3-Dimensional reconstruction of paw and digit trajectories

    Journal of Visualized Experiments 10:59979.

    https://doi.org/10.3791/59979

    • Google Scholar
13. Software

    1. Bova A
    2. Gaidica M
    3. Hurst A
    4. Iwai Y
    5. Hunter J
    6. Leventhal D

    (2020) Bova-Et-Al-2020-Code-Repository GitHub

    Bova-Et-Al-2020-Code-Repository GitHub.

    https://github.com/LeventhalLab/SkilledReaching
14. 1. Cadoni C
    2. Solinas M
    3. Di Chiara G

    (2000) Psychostimulant sensitization: differential changes in accumbal shell and core dopamine

    European Journal of Pharmacology 388:69–76.

    https://doi.org/10.1016/S0014-2999(99)00824-9

    • PubMed
    • Google Scholar
15. 1. Calabresi P
    2. Di Filippo M
    3. Ghiglieri V
    4. Picconi B

    (2008) Molecular mechanisms underlying levodopa-induced dyskinesia

    Movement Disorders 23:S570–S579.

    https://doi.org/10.1002/mds.22019

    • PubMed
    • Google Scholar
16. 1. Carr GD
    2. White NM

    (1987) Effects of systemic and intracranial amphetamine injections on behavior in the open field: a detailed analysis

    Pharmacology Biochemistry and Behavior 27:113–122.

    https://doi.org/10.1016/0091-3057(87)90485-0

    • PubMed
    • Google Scholar
17. 1. Chou KL
    2. Stacy M
    3. Simuni T
    4. Miyasaki J
    5. Oertel WH
    6. Sethi K
    7. Fernandez HH
    8. Stocchi F

    (2018) The spectrum of "off" in Parkinson's disease: What have we learned over 40 years?

    Parkinsonism & Related Disorders 51:9–16.

    https://doi.org/10.1016/j.parkreldis.2018.02.001

    • PubMed
    • Google Scholar
18. 1. Coddington LT
    2. Dudman JT

    (2018) The timing of action determines reward prediction signals in identified midbrain dopamine neurons

    Nature Neuroscience 21:1563–1573.

    https://doi.org/10.1038/s41593-018-0245-7

    • PubMed
    • Google Scholar
19. 1. Crevecoeur F
    2. Gevers M

    (2019) Filtering compensation for delays and prediction errors during sensorimotor control

    Neural Computation 31:738–764.

    https://doi.org/10.1162/neco_a_01170

    • PubMed
    • Google Scholar
20. 1. da Silva JA
    2. Tecuapetla F
    3. Paixão V
    4. Costa RM

    (2018) Dopamine neuron activity before action initiation gates and invigorates future movements

    Nature 554:244–248.

    https://doi.org/10.1038/nature25457

    • PubMed
    • Google Scholar
21. 1. de la Fuente-Fernández R
    2. Sossi V
    3. Huang Z
    4. Furtado S
    5. Lu JQ
    6. Calne DB
    7. Ruth TJ
    8. Stoessl AJ

    (2004) Levodopa-induced changes in synaptic dopamine levels increase with progression of Parkinson's disease: implications for dyskinesias

    Brain 127:2747–2754.

    https://doi.org/10.1093/brain/awh290

    • PubMed
    • Google Scholar
22. 1. Dowd E
    2. Dunnett SB

    (2005) Comparison of 6-hydroxydopamine-induced medial forebrain bundle and nigrostriatal terminal lesions in a lateralised nose-poking task in rats

    Behavioural Brain Research 159:153–161.

    https://doi.org/10.1016/j.bbr.2004.10.010

    • PubMed
    • Google Scholar
23. 1. Dudman JT
    2. Krakauer JW

    (2016) The basal ganglia: from motor commands to the control of vigor

    Current Opinion in Neurobiology 37:158–166.

    https://doi.org/10.1016/j.conb.2016.02.005

    • PubMed
    • Google Scholar
24. 1. Ellens DJ
    2. Gaidica M
    3. Toader A
    4. Peng S
    5. Shue S
    6. John T
    7. Bova A
    8. Leventhal DK

    (2016) An automated rat single pellet reaching system with high-speed video capture

    Journal of Neuroscience Methods 271:119–127.

    https://doi.org/10.1016/j.jneumeth.2016.07.009

    • PubMed
    • Google Scholar
25. 1. Foki T
    2. Vanbellingen T
    3. Lungu C
    4. Pirker W
    5. Bohlhalter S
    6. Nyffeler T
    7. Kraemmer J
    8. Haubenberger D
    9. Fischmeister FP
    10. Auff E
    11. Hallett M
    12. Beisteiner R

    (2016) Limb-kinetic apraxia affects activities of daily living in Parkinson's disease: a multi-center study

    European Journal of Neurology 23:1301–1307.

    https://doi.org/10.1111/ene.13021

    • PubMed
    • Google Scholar
26. 1. Gadagkar V
    2. Puzerey PA
    3. Chen R
    4. Baird-Daniel E
    5. Farhang AR
    6. Goldberg JH

    (2016) Dopamine neurons encode performance error in singing birds

    Science 354:1278–1282.

    https://doi.org/10.1126/science.aah6837

    • PubMed
    • Google Scholar
27. 1. Gebhardt A
    2. Vanbellingen T
    3. Baronti F
    4. Kersten B
    5. Bohlhalter S

    (2008) Poor dopaminergic response of impaired dexterity in Parkinson's disease: Bradykinesia or limb kinetic apraxia?

    Movement Disorders 23:1701–1706.

    https://doi.org/10.1002/mds.22199

    • PubMed
    • Google Scholar
28. 1. Girasole AE
    2. Lum MY
    3. Nathaniel D
    4. Bair-Marshall CJ
    5. Guenthner CJ
    6. Luo L
    7. Kreitzer AC
    8. Nelson AB

    (2018) A subpopulation of striatal neurons mediates Levodopa-Induced dyskinesia

    Neuron 97:787–795.

    https://doi.org/10.1016/j.neuron.2018.01.017

    • PubMed
    • Google Scholar
29. 1. Glimcher PW

    (2011) Understanding dopamine and reinforcement learning: the dopamine reward prediction error hypothesis

    PNAS 108:15647–15654.

    https://doi.org/10.1073/pnas.1014269108

    • PubMed
    • Google Scholar
30. 1. Guo L
    2. Xiong H
    3. Kim JI
    4. Wu YW
    5. Lalchandani RR
    6. Cui Y
    7. Shu Y
    8. Xu T
    9. Ding JB

    (2015) Dynamic rewiring of neural circuits in the motor cortex in mouse models of Parkinson's disease

    Nature Neuroscience 18:1299–1309.

    https://doi.org/10.1038/nn.4082

    • PubMed
    • Google Scholar
31. 1. Hamid AA
    2. Pettibone JR
    3. Mabrouk OS
    4. Hetrick VL
    5. Schmidt R
    6. Vander Weele CM
    7. Kennedy RT
    8. Aragona BJ
    9. Berke JD

    (2016) Mesolimbic dopamine signals the value of work

    Nature Neuroscience 19:117–126.

    https://doi.org/10.1038/nn.4173

    • PubMed
    • Google Scholar
32. Preprint

    1. Hamilos AE
    2. Spedicato G
    3. Hong Y
    4. Sun F
    5. Li Y
    6. Assad JA

    (2020) Dynamic dopaminergic activity controls the timing of self-timed movement

    bioRxiv.

    https://doi.org/10.1101/2020.05.13.094904

    • Google Scholar
33. Book

    1. Hartley R
    2. Zisserman A

    (2003)

    Multiple View Geometry in Computer Vision

    Cambridge University Press.

    • Google Scholar
34. 1. Hosp JA
    2. Pekanovic A
    3. Rioult-Pedotti MS
    4. Luft AR

    (2011) Dopaminergic projections from midbrain to primary motor cortex mediate motor skill learning

    Journal of Neuroscience 31:2481–2487.

    https://doi.org/10.1523/JNEUROSCI.5411-10.2011

    • PubMed
    • Google Scholar
35. 1. Howard CD
    2. Li H
    3. Geddes CE
    4. Jin X

    (2017) Dynamic nigrostriatal dopamine biases action selection

    Neuron 93:1436–1450.

    https://doi.org/10.1016/j.neuron.2017.02.029

    • PubMed
    • Google Scholar
36. 1. Howe MW
    2. Dombeck DA

    (2016) Rapid signalling in distinct dopaminergic axons during locomotion and reward

    Nature 535:505–510.

    https://doi.org/10.1038/nature18942

    • PubMed
    • Google Scholar
37. 1. Hyland BI
    2. Seeger-Armbruster S
    3. Smither RA
    4. Parr-Brownlie LC

    (2019) Altered recruitment of motor cortex neuronal activity during the grasping phase of skilled reaching in a chronic rat model of unilateral parkinsonism

    The Journal of Neuroscience 39:9660–9672.

    https://doi.org/10.1523/JNEUROSCI.0720-19.2019

    • PubMed
    • Google Scholar
38. 1. Hyland BI
    2. Jordan VM

    (1997) Muscle activity during forelimb reaching movements in rats

    Behavioural Brain Research 85:175–186.

    https://doi.org/10.1016/S0166-4328(97)87582-1

    • PubMed
    • Google Scholar
39. 1. Jin X
    2. Costa RM

    (2010) Start/stop signals emerge in nigrostriatal circuits during sequence learning

    Nature 466:457–462.

    https://doi.org/10.1038/nature09263

    • PubMed
    • Google Scholar
40. 1. Klein A
    2. Sacrey LA
    3. Whishaw IQ
    4. Dunnett SB

    (2012) The use of rodent skilled reaching as a translational model for investigating brain damage and disease

    Neuroscience & Biobehavioral Reviews 36:1030–1042.

    https://doi.org/10.1016/j.neubiorev.2011.12.010

    • PubMed
    • Google Scholar
41. 1. Klein A
    2. Dunnett SB

    (2012) Analysis of skilled forelimb movement in rats: the single pellet reaching test and staircase test

    Current Protocols in Neuroscience 8:28.

    https://doi.org/10.1002/0471142301.ns0828s58

    • PubMed
    • Google Scholar
42. 1. Lane EL
    2. Daly CS
    3. Smith GA
    4. Dunnett SB

    (2011) Context-driven changes in L-DOPA-induced behaviours in the 6-OHDA lesioned rat

    Neurobiology of Disease 42:99–107.

    https://doi.org/10.1016/j.nbd.2011.01.010

    • PubMed
    • Google Scholar
43. 1. Lee SH
    2. Lee MJ
    3. Lyoo CH
    4. Cho H
    5. Lee MS

    (2018) Impaired finger dexterity and nigrostriatal dopamine loss in Parkinson's disease

    Journal of Neural Transmission 125:1333–1339.

    https://doi.org/10.1007/s00702-018-1901-5

    • PubMed
    • Google Scholar
44. 1. Leib R
    2. Russo M
    3. d'Avella A
    4. Nisky I

    (2020) A bang-bang control model predicts the triphasic muscles activity during hand reaching

    Journal of Neurophysiology 124:295–304.

    https://doi.org/10.1152/jn.00132.2020

    • PubMed
    • Google Scholar
45. 1. Lemke SM
    2. Ramanathan DS
    3. Guo L
    4. Won SJ
    5. Ganguly K

    (2019) Emergent modular neural control drives coordinated motor actions

    Nature Neuroscience 22:1122–1131.

    https://doi.org/10.1038/s41593-019-0407-2

    • PubMed
    • Google Scholar
46. 1. Leventhal DK
    2. Stoetzner CR
    3. Abraham R
    4. Pettibone J
    5. DeMarco K
    6. Berke JD

    (2014) Dissociable effects of dopamine on learning and performance within sensorimotor striatum

    Basal Ganglia 4:43–54.

    https://doi.org/10.1016/j.baga.2013.11.001

    • Google Scholar
47. 1. Mathis A
    2. Mamidanna P
    3. Cury KM
    4. Abe T
    5. Murthy VN
    6. Mathis MW
    7. Bethge M

    (2018) DeepLabCut: markerless pose estimation of user-defined body parts with deep learning

    Nature Neuroscience 21:1281–1289.

    https://doi.org/10.1038/s41593-018-0209-y

    • PubMed
    • Google Scholar
48. 1. Mazzoni P
    2. Hristova A
    3. Krakauer JW

    (2007) Why don't we move faster? parkinson's disease, movement vigor, and implicit motivation

    Journal of Neuroscience 27:7105–7116.

    https://doi.org/10.1523/JNEUROSCI.0264-07.2007

    • PubMed
    • Google Scholar
49. 1. Mello GB
    2. Soares S
    3. Paton JJ

    (2015) A scalable population code for time in the striatum

    Current Biology 25:1113–1122.

    https://doi.org/10.1016/j.cub.2015.02.036

    • PubMed
    • Google Scholar
50. 1. Mohebi A
    2. Pettibone JR
    3. Hamid AA
    4. Wong JT
    5. Vinson LT
    6. Patriarchi T
    7. Tian L
    8. Kennedy RT
    9. Berke JD

    (2019) Dissociable dopamine dynamics for learning and motivation

    Nature 570:65–70.

    https://doi.org/10.1038/s41586-019-1235-y

    • PubMed
    • Google Scholar
51. 1. Niv Y
    2. Daw ND
    3. Joel D
    4. Dayan P

    (2007) Tonic dopamine: opportunity costs and the control of response vigor

    Psychopharmacology 191:507–520.

    https://doi.org/10.1007/s00213-006-0502-4

    • PubMed
    • Google Scholar
52. 1. Palmiter RD

    (2008) Dopamine signaling in the dorsal striatum is essential for motivated behaviors: lessons from dopamine-deficient mice

    Annals of the New York Academy of Sciences 1129:35–46.

    https://doi.org/10.1196/annals.1417.003

    • PubMed
    • Google Scholar
53. 1. Panigrahi B
    2. Martin KA
    3. Li Y
    4. Graves AR
    5. Vollmer A
    6. Olson L
    7. Mensh BD
    8. Karpova AY
    9. Dudman JT

    (2015) Dopamine is required for the neural representation and control of movement vigor

    Cell 162:1418–1430.

    https://doi.org/10.1016/j.cell.2015.08.014

    • PubMed
    • Google Scholar
54. 1. Parker NF
    2. Cameron CM
    3. Taliaferro JP
    4. Lee J
    5. Choi JY
    6. Davidson TJ
    7. Daw ND
    8. Witten IB

    (2016) Reward and choice encoding in terminals of midbrain dopamine neurons depends on striatal target

    Nature Neuroscience 19:845–854.

    https://doi.org/10.1038/nn.4287

    • PubMed
    • Google Scholar
55. 1. Patriarchi T
    2. Cho JR
    3. Merten K
    4. Howe MW
    5. Marley A
    6. Xiong WH
    7. Folk RW
    8. Broussard GJ
    9. Liang R
    10. Jang MJ
    11. Zhong H
    12. Dombeck D
    13. von Zastrow M
    14. Nimmerjahn A
    15. Gradinaru V
    16. Williams JT
    17. Tian L

    (2018) Ultrafast neuronal imaging of dopamine dynamics with designed genetically encoded sensors

    Science 360:eaat4422.

    https://doi.org/10.1126/science.aat4422

    • PubMed
    • Google Scholar
56. Book

    1. Paxinos G
    2. Watson C

    (1998)

    The rat brain in Stereotaxic coordinates

    Academic Press.

    • Google Scholar
57. 1. Pohar SL
    2. Allyson Jones C

    (2009) The burden of parkinson disease (PD) and concomitant comorbidities

    Archives of Gerontology and Geriatrics 49:317–321.

    https://doi.org/10.1016/j.archger.2008.11.006

    • PubMed
    • Google Scholar
58. 1. Rand MK
    2. Smiley-Oyen AL
    3. Shimansky YP
    4. Bloedel JR
    5. Stelmach GE

    (2006) Control of aperture closure during reach-to-grasp movements in Parkinson's disease

    Experimental Brain Research 168:131–142.

    https://doi.org/10.1007/s00221-005-0073-3

    • PubMed
    • Google Scholar
59. 1. Robinson TE
    2. Becker JB

    (1986) Enduring changes in brain and behavior produced by chronic amphetamine administration: a review and evaluation of animal models of amphetamine psychosis

    Brain Research Reviews 11:157–198.

    https://doi.org/10.1016/0165-0173(86)90002-0

    • Google Scholar
60. 1. Ryan MB
    2. Bair-Marshall C
    3. Nelson AB

    (2018) Aberrant striatal activity in parkinsonism and Levodopa-Induced dyskinesia

    Cell Reports 23:3438–3446.

    https://doi.org/10.1016/j.celrep.2018.05.059

    • PubMed
    • Google Scholar
61. 1. Salamone JD
    2. Correa M

    (2012) The mysterious motivational functions of mesolimbic dopamine

    Neuron 76:470–485.

    https://doi.org/10.1016/j.neuron.2012.10.021

    • PubMed
    • Google Scholar
62. 1. Saunders BT
    2. Richard JM
    3. Margolis EB
    4. Janak PH

    (2018) Dopamine neurons create pavlovian conditioned stimuli with circuit-defined motivational properties

    Nature Neuroscience 21:1072–1083.

    https://doi.org/10.1038/s41593-018-0191-4

    • PubMed
    • Google Scholar
63. 1. Schultz W
    2. Ruffieux A
    3. Aebischer P

    (1983) The activity of pars Compacta neurons of the monkey substantia nigra in relation to motor activation

    Experimental Brain Research 51:377–387.

    https://doi.org/10.1007/BF00237874

    • Google Scholar
64. 1. Schultz W

    (2019) Recent advances in understanding the role of phasic dopamine activity

    F1000Research 8:1680.

    https://doi.org/10.12688/f1000research.19793.1

    • Google Scholar
65. 1. Sebastianutto I
    2. Maslava N
    3. Hopkins CR
    4. Cenci MA

    (2016) Validation of an improved scale for rating l-DOPA-induced dyskinesia in the mouse and effects of specific dopamine receptor antagonists

    Neurobiology of Disease 96:156–170.

    https://doi.org/10.1016/j.nbd.2016.09.001

    • PubMed
    • Google Scholar
66. 1. Shen W
    2. Flajolet M
    3. Greengard P
    4. Surmeier DJ

    (2008) Dichotomous dopaminergic control of striatal synaptic plasticity

    Science 321:848–851.

    https://doi.org/10.1126/science.1160575

    • PubMed
    • Google Scholar
67. 1. Soares S
    2. Atallah BV
    3. Paton JJ

    (2016) Midbrain dopamine neurons control judgment of time

    Science 354:1273–1277.

    https://doi.org/10.1126/science.aah5234

    • PubMed
    • Google Scholar
68. 1. Tobler PN
    2. Fiorillo CD
    3. Schultz W

    (2005) Adaptive coding of reward value by dopamine neurons

    Science 307:1642–1645.

    https://doi.org/10.1126/science.1105370

    • PubMed
    • Google Scholar
69. 1. Valjent E
    2. Bertran-Gonzalez J
    3. Aubier B
    4. Greengard P
    5. Hervé D
    6. Girault JA

    (2010) Mechanisms of locomotor sensitization to drugs of abuse in a two-injection protocol

    Neuropsychopharmacology 35:401–415.

    https://doi.org/10.1038/npp.2009.143

    • PubMed
    • Google Scholar
70. 1. Vanderschuren LJ
    2. Schmidt ED
    3. De Vries TJ
    4. Van Moorsel CA
    5. Tilders FJ
    6. Schoffelmeer AN

    (1999) A single exposure to amphetamine is sufficient to induce long-term behavioral, neuroendocrine, and neurochemical sensitization in rats

    The Journal of Neuroscience 19:9579–9586.

    https://doi.org/10.1523/JNEUROSCI.19-21-09579.1999

    • PubMed
    • Google Scholar
71. 1. Whishaw IQ
    2. O'Connor WT
    3. Dunnett SB

    (1986) The contributions of motor cortex, nigrostriatal dopamine and caudate-putamen to skilled forelimb use in the rat

    Brain 109:805–843.

    https://doi.org/10.1093/brain/109.5.805

    • PubMed
    • Google Scholar
72. 1. Whishaw IQ
    2. Suchowersky O
    3. Davis L
    4. Sarna J
    5. Metz GA
    6. Pellis SM

    (2002) Impairment of pronation, Supination, and body co-ordination in reach-to-grasp tasks in human Parkinson's disease (PD) reveals homology to deficits in animal models

    Behavioural Brain Research 133:165–176.

    https://doi.org/10.1016/S0166-4328(01)00479-X

    • PubMed
    • Google Scholar
73. 1. Whishaw IQ
    2. Whishaw P
    3. Gorny B

    (2008) The structure of skilled forelimb reaching in the rat: a movement rating scale

    Journal of Visualized Experiments 18:816.

    https://doi.org/10.3791/816

    • Google Scholar
74. 1. Whishaw IQ
    2. Pellis SM

    (1990) The structure of skilled forelimb reaching in the rat: a proximally driven movement with a single distal rotatory component

    Behavioural Brain Research 41:49–59.

    https://doi.org/10.1016/0166-4328(90)90053-H

    • PubMed
    • Google Scholar
75. 1. Yttri EA
    2. Dudman JT

    (2016) Opponent and bidirectional control of movement velocity in the basal ganglia

    Nature 533:402–406.

    https://doi.org/10.1038/nature17639

    • PubMed
    • Google Scholar
76. 1. Yttri EA
    2. Dudman JT

    (2018) A proposed circuit computation in basal ganglia: history-dependent gain

    Movement Disorders 33:704–716.

    https://doi.org/10.1002/mds.27321

    • PubMed
    • Google Scholar
77. 1. Zhang Y
    2. Loonam TM
    3. Noailles PA
    4. Angulo JA

    (2001) Comparison of cocaine- and methamphetamine-evoked dopamine and glutamate overflow in Somatodendritic and terminal field regions of the rat brain during acute, chronic, and early withdrawal conditions

    Annals of the New York Academy of Sciences 937:93–120.

    https://doi.org/10.1111/j.1749-6632.2001.tb03560.x

    • PubMed
    • Google Scholar
78. 1. Zhang Y
    2. Larcher KM
    3. Misic B
    4. Dagher A

    (2017) Anatomical and functional organization of the human substantia nigra and its connections

    eLife 6:e26653.

    https://doi.org/10.7554/eLife.26653

    • PubMed
    • Google Scholar

Author details

1. Alexandra Bova

   Neuroscience Graduate Program, University of Michigan, Ann Arbor, United States

   Contribution

   Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing

   Competing interests

   No competing interests declared

2. Matt Gaidica

   Neuroscience Graduate Program, University of Michigan, Ann Arbor, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0191-1899

3. Amy Hurst

   Department of Neurology, University of Michigan, Ann Arbor, United States

   Contribution

   Investigation, Methodology

   Competing interests

   No competing interests declared

4. Yoshiko Iwai

   Department of Neurology, University of Michigan, Ann Arbor, United States

   Contribution

   Investigation

   Competing interests

   No competing interests declared

5. Julia Hunter

   Department of Neurology, University of Michigan, Ann Arbor, United States

   Contribution

   Investigation

   Competing interests

   No competing interests declared

6. Daniel K Leventhal

   1. Department of Neurology, University of Michigan, Ann Arbor, United States
   2. Department of Biomedical Engineering, University of Michigan, Ann Arbor, United States
   3. Parkinson Disease Foundation Research Center of Excellence, University of Michigan, Ann Arbor, United States
   4. Department of Neurology, VA Ann Arbor Health System, Ann Arbor, United States

   Contribution

   Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Methodology, Writing - original draft, Writing - review and editing

   For correspondence

   dleventh@med.umich.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-8174-5933

• 4,781

  views

• 505

  downloads

• 48

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.