Depression is characterized by both anhedonia (the loss of pleasure or lack of reactivity to rewarding stimuli) and increased stress responsiveness, but whether these two promising depressive endophenotypes interact and are modulated by genes remains largely unexplored. As an initial step to address these important issues, the main goal of this dissertation was to examine whether stress and genetic variation independently and interactively influence reinforcement learning, an important behavioral component of anhedonia.

Across the three studies of the current dissertation, participants completed a probabilistic reward learning task that allows for an objective assessment of an individual's ability to modulate behavior according to reinforcement history. In Study 1, we examined how mineralorcorticoid receptor (MR) iso/val genotype (rs5522) and acute laboratory stress impact behavioral reward learning. In Study 2, we probed how genetic variation within corticotrophin-releasing hormone type 1 receptor (CRHR1; rs12938031, rs110402, rs4076452, rs10445364) and stress affect behavioral reward learning and the feedback-related negativity (FRN). The FRN is an event-related potential (ERP) component theorized to reflect phasic dopaminergic bursts critically implicated in reinforcement learning. In Study 3, functional magnetic resonance imaging (fMRI) was used to examine the neural correlates of reward learning and how perceived stress affects reward-related neural activation.

We hypothesized that stress would be associated with: (1) reduced behavioral reward learning, (2) enhanced FRN amplitude (reflective of a reliance on external feedback due to blunted learning), and (3) reduced activation to rewarding stimuli in anterior cingulate and medial prefrontal regions previously implicated in integrating reinforcement history and coding the incentive value of stimuli. Furthermore, we expected that polymorphisms within the CRHR1 and MR genes associated with stress-related psychopathology or a dysregulated stress response would be associated with reduced reward learning, particularly under stress. Lastly, we hypothesized that stimuli predicting more frequent reward as well as unexpected reward delivery would be associated with elevated basal ganglia, anterior cingulate, and orbitofrontal cortex activation.

In line with our hypotheses, acute laboratory stress was associated with behavioral and ERP markers of reduced reinforcement learning (Studies 1 and 2). Furthermore, stress-induced deficits were potentiated by specific MR and CRHR1 genotypes. In Study 3, elevated basal ganglia and orbitofrontal cortex activation was observed in response to reward predicting stimuli and less frequent (and thus, unexpected) reward. Moreover, in Study 3, elevated stress perception was negatively associated with medial PFC activation to reward predicting stimuli and basal ganglia responses to reward feedback. Together, these data indicate that stress and genetic variants regulating the responsiveness of the stress response system individually and interactively impact reward processing. We conclude that: (1) stress-induced deficits in reward processing are a potential mechanism underlying the association between stress and depression, and (2) individuals with certain MR and CRHR1 polymorphisms are more susceptible to stress-induced dysfunction, which may partially explain their increased vulnerability to psychopathology.