BRAIN NETWORKS FOR PUNISHMENT LEARNING

We constantly navigate choices, weigh risks, and learn from outcomes. When a choice leads to reward, we are more likely to repeat it; when it leads to punishment, we are not. This ability to learn from mistakes, essential to survival, is called punishment learning. Impairments in punishment learning whereby individuals struggle to learn from mistakes, have been implicated in several disorders such as depression, anxiety, substance use disorders, and behavioural addictions. Despite this, our knowledge of the brain mechanisms of punishment learning, and how these can go awry, is impoverished, especially compared with other forms of learning such as fear and reward. Here we studied the brain mechanisms of punishment learning. First, we used whole-brain, single-cell c-Fos imaging to identify the brain-wide networks for punishment learning. Graph theory and network analyses identified three brain regions - basolateral amygdala, zona incerta, and rostral linear nucleus- as key nodes in this punishment learning network. Next, we used multiplexed chemogenetic manipulation to assess the causal role of these key nodes in punishment learning. We expressed the inhibitory hM4Di DREADD in these three regions and showed that concurrent chemogenetic inhibition of basolateral amygdala, zona incerta, and rostral linear nucleus profoundly impaired punishment learning as shown by a lack of selective suppression of the punished action during chemogenetic inhibition and the following day when tested without chemogenetic inhibition. These results provide the first map of whole brain networks for punishment learning and demonstrate a causal role for this network in punishment learning.