Individuals with substance abuse disorders often continue to seek and use illicit drugs despite the negative consequences that can arise from this behavior. In order to understand the neural basis of this behavioral inflexibility in the face of negative consequences, it is necessary to understand how negative feedback is processed in the healthy brain. Models of decision making have shown that feedback provides information to compare actual to expected outcomes of our actions, and, in this context, presentations of unpleasant outcomes and omissions of valued outcomes can both be thought of as forms of negative feedback. However, these two forms of negative feedback result in different behaviors: response rates decrease when an outcome loses its value, whereas response rates increase for omission of an expected high-value outcome. The neural systems which underlie the effects of these different forms of negative feedback have not been clearly delineated. We hypothesize that separate cortico-limbic-striatal circuits mediate different forms of negative feedback, depending on whether this feedback indicates a change in outcome probability or a change in outcome value. To examine this hypothesis, we will perform an interdisciplinary set of experiments involving research on rodents and humans. In two parallel experiments, rats and humans will perform distinct actions to obtain two types of rewarding food outcomes. Following acquisition of this task, we will manipulate outcome probability by occasionally omitting expected outcomes. In addition, we will manipulate outcome value by selectively devaluing one food outcome by allowing subjects to eat it to satiety. In Experiment 1.1, rats will undergo a neural disconnection procedure to remove basolateral amygdala (BLA) or pre-frontal cortex (PFC) inputs to the striatum, to determine the importance of these circuits in negative feedback processing using the outcome omission and devaluation tests. In Experiment 1.2, we will assess human brain activity using fMRI following the devaluation procedure, as subjects receive the devalued or the valued outcome, or an omission of the expected delivery of these outcomes. Experiment 2.1 extends this work to more abstract behavioral outcomes more typical of experimental work in humans. Positive and negative feedback indicating whether a response is correct or incorrect will be provided in a probabilistic learning task, with this feedback occasionally omitted unexpectedly. This will allow us to determine whether the negative feedback is interpreted by the brain as a devalued outcome or as the lack of a positive outcome. The interdisciplinary nature of these experiments will provide a unique training experience for undergraduate and graduate students. Furthermore, this proposal will help to bridge the gap between behavioral and cognitive neuroscience research and ultimately will aid translational approaches to understanding how neural systems underlying negative feedback processing may be compromised in drug addiction.
Understanding how negative feedback processing operates in the normal brain using natural outcomes will be useful in deciphering how impairments in these processes arise in individuals with substance abuse disorders. Furthermore, such knowledge will aid in developing therapeutic strategies to facilitate recovery from addiction.
|Lempert, Karolina M; Tricomi, Elizabeth (2016) The Value of Being Wrong: Intermittent Feedback Delivery Alters the Striatal Response to Negative Feedback. J Cogn Neurosci 28:261-74|
|Tricomi, Elizabeth; Lempert, Karolina M (2015) Value and probability coding in a feedback-based learning task utilizing food rewards. J Neurophysiol 113:4-13|
|Natsheh, Joman Y; Shiflett, Michael W (2015) The Effects of Methylphenidate on Goal-directed Behavior in a Rat Model of ADHD. Front Behav Neurosci 9:326|
|Dobryakova, Ekaterina; Tricomi, Elizabeth (2013) Basal ganglia engagement during feedback processing after a substantial delay. Cogn Affect Behav Neurosci 13:725-36|