Dopamine in motor cortex is necessary for skill learning and synaptic plasticity

Abstract

Preliminary evidence indicates that dopamine given by mouth facilitates the learning of motor skills and improves the recovery of movement after stroke. The mechanism of these phenomena is unknown. Here, we describe a mechanism by demonstrating in rat that dopaminergic terminals and receptors in primary motor cortex (M1) enable motor skill learning and enhance M1 synaptic plasticity. Elimination of dopaminergic terminals in M1 specifically impaired motor skill acquisition, which was restored upon DA substitution. Execution of a previously acquired skill was unaffected. Reversible blockade of M1 D1 and D2 receptors temporarily impaired skill acquisition but not execution, and reduced long-term potentiation (LTP) within M1, a form of synaptic plasticity critically involved in skill learning. These findings identify a behavioral and functional role of dopaminergic signaling in M1. DA in M1 optimizes the learning of a novel motor skill.

Figure 1. Dopamine (DA) release in M1 is necessary for optimal motor skill acquisition but not for movement execution.

(a) Learning curves for sham-lesioned rats (black, vehicle), rats with dopaminergic terminals destroyed (red, 6-OHDA+D), and rats with noradrenergic terminals destroyed (blue, 6-OHDA+N). Cortical injections (vertical arrows) were performed following an initial training session to determine paw preference. After 3 days of recovery from surgery (horizontal arrow, necessary interval determined in d) rats were trained for 6 successive days. The success rate of skill acquisition was significantly impaired in animals with without dopaminergic terminals but not in animals without noradrenergic terminals (** p<0.05). (b) DA is not required for task performance because elimination of dopaminergic terminals in M1 (red, vertical arrow) in rats that already acquired the reaching skill (black) did not affect reaching performance. (c) Learning impairment is restored with DA substitution (administration of its precursor levodopa). Rats received cortical injections of 6-OHDA+D and were trained comparable to a): As compared with sham-lesioned animals (vehicle-injected, black), the two groups without dopaminergic terminals in M1 (6-OHDA+D-injected, red) demonstrated a learning impairment – phase 1. Rats were then implanted with minipumps (drops): 50% of DA terminal deficient rats received vehicle (grey) and 50% received levodopa (yellow) during the entire second training period, sham-lesioned rats received levodopa – phase 2. Learning was restored in DA-substituted rats underlining the importance of DA for skill acquisition. Minipumps were then removed and all rats were examined for task recall after 6 days of rest. DA is not required to recall an already learned skill as indicated by unchanged performance levels in all groups. (d) Cortical injections independent of whether 6-OHDA or vehicle was used transiently impair locomotor function. Rotarod tests were performed in vehicle- (black) and 6-OHDA (red) injected rats. Parallel deficits indicate that reduced rotarod speed results from injection or surgery and not from the drug itself. Results were used to determine the recovery period following surgery (horizontal arrows in a-c).

Figure 2. Identification of dopaminergic terminals in M1.

(a) Western blot analysis of M1 cortical tissue injected with vehicle (sham-lesioned) and 6-OHDA in conjunction with desipramine (i.p.) using tyroxine hydroxylase (TH) reactivity indicated reduced TH expression after elimination of dopaminergic terminals. (b) Quantification of protein expression in DA-lesioned (6-OHDA+D) and sham-lesioned hemispheres reveals reduced protein expression after elimination of dopaminergic terminals. (c) Immunofluorescence staining of cortical dopaminergic terminals (TH immunoreactivity) in an exemplary vehicle and DA-lesioned hemisphere (6-OHDA injections into M1) indicated almost no staining in layer I and II/III and reduced staining in deeper layers in the lesioned M1. Similar findings were obtained in the other two animals treated analogously.

Figure 3. Functional D1 and D2 receptors in M1 are necessary for optimal motor skill acquisition but not for movement execution.

(a) Blocking D1 receptors with SCH02339 (green) and D2 receptors with raclopride (blue) or sulpiride (orange) on the second and third day (arrows) of motor skill training significantly impaired reaching success compared to vehicle injected animals (black). When antagonists were discontinued, success rate began to increase normally. No significant differences in success rate existed at day 8 between all 4 groups. Inset: exemplary Nissl stain to verify cannula placement. (b) Raclopride injected into M1 (arrows) after the task had been acquired did not affect the performance. Inset: exemplary Nissl stain to verify injection cannula placement. (c,d) To exclude the possibility that the antagonists spread to other brain regions receiving important DA projections thereby causing the observed learning impairment, raclopride was injected into the dorsal striatum (c, blue) and the prefrontal cortex (d, blue) and compared to vehicle injected controls (black). Skill acquisition was not impaired in these animals. Insets: exemplary Nissl stain to verify cannula placement.

Figure 4. Synaptic plasticity but not synaptic transmission depends on DA receptor activity in M1.

(a,b) Exemplary time courses of peak amplitudes of extracellular field potentials (FP) in layer II/III horizontal connections in the M1 forelimb area recorded in brain slices. FP amplitudes at baseline stimulation intensity before (control) and after D1 (a, SCH02339, green) and D1 receptor blockade (b, raclopride, yellow). Antagonists do not modify amplitude or shape of FPs. Insets: each trace represents an average of 10 individual traces at times indicated by numbers. (c) Group data indicate no significant difference before (control) and after antagonist application. (d, e) LTP was induced repeatedly (multiple arrows) until responses were saturated in normal ACSF (control, grey) and in the presence of SCH02339 (d, green) or raclopride (e, yellow). (f) Group data show significantly reduced LTP in the presence of D1 and D2 receptor antagonists compared to controls (grey) for single LTP induction (left) and saturated LTP (right). In the presence of DA antagonists responses are already saturated after the first LTP attempt.