Premotor cortex is critical for goal-directed actions - PubMed (original) (raw)

Premotor cortex is critical for goal-directed actions

Christina M Gremel et al. Front Comput Neurosci. 2013.

Abstract

Shifting between motor plans is often necessary for adaptive behavior. When faced with changing consequences of one's actions, it is often imperative to switch from automatic actions to deliberative and controlled actions. The pre-supplementary motor area (pre-SMA) in primates, akin to the premotor cortex (M2) in mice, has been implicated in motor learning and planning, and action switching. We hypothesized that M2 would be differentially involved in goal-directed actions, which are controlled by their consequences vs. habits, which are more dependent on their past reinforcement history and less on their consequences. To investigate this, we performed M2 lesions in mice and then concurrently trained them to press the same lever for the same food reward using two different schedules of reinforcement that differentially bias towards the use of goal-directed versus habitual action strategies. We then probed whether actions were dependent on their expected consequence through outcome revaluation testing. We uncovered that M2 lesions did not affect the acquisition of lever-pressing. However, in mice with M2 lesions, lever-pressing was insensitive to changes in expected outcome value following goal-directed training. However, habitual actions were intact. We confirmed a role for M2 in goal-directed but not habitual actions in separate groups of mice trained on the individual schedules biasing towards goal-directed versus habitual actions. These data indicate that M2 is critical for actions to be updated based on their consequences, and suggest that habitual action strategies may not require processing by M2 and the updating of motor plans.

Keywords: action selection; goal-directed actions; habitual actions; premotor cortex; value-based decision making.

PubMed Disclaimer

Figures

Figure 1

Figure 1

M2 lesions disrupt the ability to shift between automatic and goal-directed actions in the same animal. Representative picture of an M2 lesion (A), with the area of lesion outlined in the box in the left panel enlarged in the right panel. Bottom panels are illustrated examples showing approximately the largest (black) and smallest (grey) extent of the lesions observed. (B) Schematic of experimental design. During acquisition, mice were concurrently trained under random interval (RI) and random ratio (RR) reinforcement schedules to press a similar lever for the same outcome. A separate outcome was provided daily in the home cage. Mice then underwent outcome revaluation testing comprising Valued (pre-fed home-cage outcome) and Devalued (pre-fed operant outcome) days. (C–F) Lever-pressing behavior during acquisition of concurrent RI (left panel) and RR (right panel) schedules, showing the effects of M2 lesions on the number of lever presses made (C), the rate of lever pressing (D), the number of rewards earned (E), or head entries (F) during RI and RR schedule training. Mice then underwent subsequent outcome revaluation testing. (G) Shown is normalized lever pressing [Lever presses for each Revaluation state (Valued or Devalued state)/total Lever presses (Valued + Devalued states)] during outcome revaluation testing for Sham and M2 lesion mice. Non-reinforced lever pressing in previously RI and RR training contexts was examined on both Valued (black bars) and Devalued (grey bars) days. Error bars = ± SEM. * = Bonferroni corrected p < 0.05.

Figure 2

Figure 2

M2 lesions disrupt goal-directed actions but spare habitual actions. Separate groups of mice were trained to lever press for an outcome under only RI or only RR schedules of reinforcement, and then underwent subsequent outcome revaluation testing. (A) Schematic of experimental design. Mice were trained to press a lever only under a random interval (RI) reinforcement schedule, and then underwent outcome revaluation testing. (B–F) Effect of M2 lesions on acquisition under RI schedule on the average number of lever-presses made (B), the average rate of lever-pressing (C), the average rewards earned (D), and the average head entries performed (E). (F) Effect of outcome revaluation on normalized lever-pressing following RI schedule training for Sham and M2 lesion mice. (G) Schematic of experimental design. Mice were trained to press a lever only under a random ratio (RR) reinforcement schedule, and then underwent outcome revaluation testing (H–K) Effect of M2 lesions on acquisition under RR schedule on the average number of lever-presses made (H), the average rate of lever-pressing (I), the average rewards earned (J), and the average head entries performed (K). (L) Effect of outcome revaluation on normalized lever-pressing following RR schedule training for Sham and M2 lesion mice. For revaluation testing, Valued days = black bars, and Devalued day = grey bars. Error bars = ± SEM. * = Bonferroni corrected p < 0.05.

Similar articles

Cited by

References

    1. Adams C. D. (1982). Variations in the sensitivity of instrumental responding to reinforcer devaluation. Q. J. Exp. Psychol. B 34, 77–98. 10.1080/14640748208400878
    1. Adams C. D., Dickinson A. (1981). Instrumental responding following reinforcer devaluation. Q. J. Exp. Psychol. B 33, 109–121. 10.1080/14640748108400816
    1. Balleine B. W., O’Doherty J. P. (2009). Human and rodent homologies in action control: corticostriatal determinants of goal-directed and habitual action. Neuropsychopharmacology 35, 48–69 10.1038/npp.2009.131 - DOI - PMC - PubMed
    1. Balleine B. W., Ostlund S. B. (2007). Still at the choice-point: action selection and initiation in instrumental conditioning. Ann. NY Acad. Sci. 1104, 147–171 10.1196/annals.1390.006 - DOI - PubMed
    1. Colwill R. M., Rescorla R. A. (1985). Postconditioning devaluation of a reinforcer affects instrumental responding. J. Exp. Psychol. Anim. Behav. Process. 11, 120–132 10.1037//0097-7403.11.1.120 - DOI

LinkOut - more resources