The acquisition of goal-directed actions generates opposing plasticity in direct and indirect pathways in dorsomedial striatum - PubMed (original) (raw)
The acquisition of goal-directed actions generates opposing plasticity in direct and indirect pathways in dorsomedial striatum
Qiang Shan et al. J Neurosci. 2014.
Abstract
A cortical-basal ganglia network involving, particularly, the posterior region of dorsomedial striatum (DMS) has been implicated in the acquisition of goal-directed actions; however, no direct evidence of learning-related plasticity in this striatal region has been reported, nor is it known whether, or which, specific cell types are involved in this learning process. The striatum is primarily composed of two classes of spiny projection neurons (SPNs): the striatonigral and striatopallidal SPNs, which express dopamine D1 and D2 receptors, respectively. Here we establish that, in mice, the acquisition of goal-directed actions induced plasticity in both D1- and D2-SPNs specifically in the DMS and, importantly, that these changes were in opposing directions; after learning, AMPA/NMDA ratios were increased in D1-SPNs and reduced in the D2-SPNs in the DMS. Such opposing plasticity could provide the basis for rapidly rebiasing the control of task-specific actions, and its dysregulation could underlie disorders associated with striatal function.
Keywords: dorsal striatum; electrophysiology; goal-directed action; plasticity.
Copyright © 2014 the authors 0270-6474/14/349196-06$15.00/0.
Figures
Figure 1.
D1R- and D2R-expressing SPN involvement in pERK-related activity associated with action–outcome encoding. A, B, Groups of D1- and D2-GFP mice (n = 4–6) were either trained to lever press for food pellets (trained) or received the food pellet delivery unpaired with lever pressing (yoked). Rate of magazine entries (two-way repeated-measures ANOVA, F(1,48) = 6.8, p < 0.05; _post hoc_ test, _p_ > 0.05) and rate of lever presses (two-way repeated-measures ANOVA, F(1,48) = 101, p < 0.01; post hoc test, **p < 0.01) were recorded and compared between the yoked and the trained mice. C, This training schedule induced a goal-directed outcome association, as a separate group of D1- and D2-GFP mice (n = 9 and 7, respectively) that were trained using the same schedule, demonstrated sensitivity toward outcome devaluation (one-way ANOVA, F(1,30) = 10.4). *p < 0.01. D, The pERK-positive neurons in the pDMS (top) and DLS (bottom) were counted in the indicated regions of coronal sections (dashed line). E, The pERK-positive neurons in the pDMS (top) and DLS (bottom) were expressed as the percentage of D1R-expressing SPNs (GFP-positive cells in D1-GFP mice) and the D2R-expressing SPNs (GFP-positive cells in D2-GFP mice). The percentage of pERK-positive neurons among the D1R-expressing SPNs in the pDMS was significantly increased (F(1,9) = 6.32). *p < 0.05, trained versus yoked in D1 group of the pDMS. Other comparisons in the top and bottom were not significant (all F values <1). F, Representative samples from D1-GFP mice illustrating the proportion of pERK-positive neurons in the pDMS in the yoked and trained groups.
Figure 2.
Ex vivo electrophysiological assessment of plasticity in D1R- and D2R-expressing SPNs in the pDMS after action–outcome learning. A–C, D2-GFP mice were given training sessions in which reward delivery was either paired with lever pressing (trained) or unpaired with lever pressing (yoked). Rate of magazine entries (two-way repeated-measures ANOVA, F(1,117) = 0.5, p > 0.05; n = 22 and 19 for trained and yoked mice, respectively) and rate of lever presses (two-way repeated-measures ANOVA, F(1,117) = 50.5, p < 0.01; _post hoc_ test, **_p_ < 0.01) were recorded and compared between the yoked and the trained mice. **_D_**, Parahorizontal striatal slices were prepared, and the stimulation electrode was placed at the white matter between the cortex and the dorsal striatum, and either GFP-positive or GFP-negative neurons in the pDMS or the DLS were recorded. **_E–H_**, In the trained and yoked groups, the induction of action–outcome learning had a clear effect on plasticity with the D1R-expressing SPNs in the pDMS demonstrating a higher AMPA/NMDA ratio in the trained group than in the yoked or naive groups. Conversely, the D2R-expressing SPNs in the pDMS demonstrated a lower AMPA/NMDA ratio in the trained group than in the yoked or naive groups (**_E_**, **_F_**), indicating opposing changes in plasticity in the D1- and D2-expressing neurons. For D1 neurons in the pDMS, one-way ANOVA revealed no difference between naive versus yoked groups (_F_ < 1), but a significant difference between the trained group and both the naive and yoked groups (_F_(1,46) = 10.9, _p_ < 0.01; _post hoc_ test, **_p_ < 0.01). _n_ = 13–22. For the D2 neurons in the pDMS, the naive and yoked groups again did not differ (_F_ < 1). However, the trained group differed from both the naive and yoked groups (_F_(1,41) = 5.70). *_p_ < 0.05. _n_ = 13–15. Furthermore, these effects were only found in the pDMS and did not extend to the DLS: the AMPA/NMDA ratio of the D1R- and D2R-expressing SPNs in the DLS was not significantly different between the trained and yoked or naive groups (**_G_**, **_H_**: ANOVA, _p_ > 0.05, n = 12–19).
References
Publication types
MeSH terms
LinkOut - more resources
Full Text Sources
Other Literature Sources
Molecular Biology Databases