Dorsomedial prefrontal neural ensembles reflect changes in task utility that culminate in task quitting

Abstract

When performing a physically demanding behavior, sometimes the optimal choice is to quit the behavior rather than persist and waste time and energy. The dorsomedial prefrontal cortex (dmPFC), consisting of the anterior cingulate cortex and secondary motor area, likely contributes towards such utility assessments. Here, we examined how rodent dmPFC single unit and ensemble level activity corresponded to changes in motivation and quitting in an effortful weight lifting task. Rats carried out two task paradigms: one that became progressively more physically demanding over time and a second fixed effort version. Rats could quit the task at any time. Dorsomedial PFC neurons were highly responsive to each behavioral stage of the task, consisting of rope pulling, reward retrieval, and reward area leaving. Activity was highest early in sessions, commensurate with the highest relative task utility, then decreased until the point of quitting. Neural ensembles showed stable task representations across the entirety of sessions. However, these representations drifted and became more distinct over the course of the session. These results suggest that dmPFC neurons represent behavioral states that are dynamically modified as behaviors lose their utility, culminating in task quitting.

Neural activity in the anterior cingulate cortex is required for effort-based decision making

“…ACC neurons can dynamically encode potential actions with multiplexed representation of value, cost, and reward probability [49][50][51][52] . In rats, ACC activity in both single units and at the population level encode choice value that integrates reward magnitude and effort cost 11,14,15,36 . We find that silencing ACC during a choice, or when preparing for choice trials in the start box, disrupts effortful action selection.…”

“…These studies find clear encoding of effort-related value, but also find high degrees of heterogeneity in neuron responding suggesting that ACC subpopulations may perform unique functions. Other studies have begun to explore the encoding of effort-related behaviors at the ensemble level using miniaturized head-mounted microscopes, or multielectrode recordings 20,26 . Similarly, local field potential recordings between ACC and reward related regions implicate oscillatory activity in spatial effort-based decision making [27][28][29] .…”