The involvement of dopamine within the nucleus accumbens in the formation and maintenance of pair bonds was assessed in a series of experiments using the monogamous prairie vole. We show that dopamine transmission that promotes pair bond formation occurs within the rostral shell of the nucleus accumbens, but not in its core or caudal shell. Within this specific brain region, D1- and D2-like receptor activation produced opposite effects: D1-like activation prevented pair bond formation, whereas D2-like activation facilitated it. After extended cohabitation with a female, male voles showed behavior indicative of pair bond maintenance-namely, selective aggression towards unfamiliar females. These voles also showed a significant upregulation in nucleus accumbens D1-like receptors, and blockade of these receptors abolished selective aggression. Thus, neuroplastic reorganization of the nucleus accumbens dopamine system is responsible for the enduring nature of monogamous pair bonding. Finally, we show that this system may also contribute to species-specific social organization.

The goal of the present study was to characterize the effects of valence in facial cues and object targets on event-related potential (ERPs) indices of gaze-directed orienting. Participants were shown faces at fixation that concurrently displayed dynamic gaze shifts and expression changes from neutral to fearful or happy emotions. Emotionally-salient target objects subsequently appeared in the periphery and were spatially congruent or incongruent with the gaze direction. ERPs were time-locked to target presentation. Three sequential ERP components were modulated by happy emotion, indicating a progression from an expression effect to a gaze-by-expression interaction to a target emotion effect. These effects included larger P1 amplitude over contralateral occipital sites for targets following happy faces, larger centrally distributed N1 amplitude for targets following happy faces with leftward gaze, and faster P3 latency for positive targets. In addition, parietally distributed P3 amplitude was reduced for validly cued targets following fearful expressions. Results are consistent with accounts of attentional broadening and motivational approach by happy emotion, and facilitation of spatially directed attention in the presence of fearful cues. The findings have implications for understanding how socioemotional signals in faces interact with each other and with emotional features of objects in the environment to alter attentional processes.

Multiple sources of information from the face guide attention during social interaction. The present study modified the Posner cueing paradigm to investigate how dynamic changes in emotional expression and eye gaze in faces affect the neural processing of subsequent target stimuli. Eventrelated potentials (ERPs) were recorded while participants viewed centrally presented face displays in which gaze direction (left, direct, right) and facial expression (fearful, neutral) covaried in a fully crossed design. Gaze direction was not predictive of peripheral target location. ERP analysis revealed several sequential effects, including: (1) an early enhancement of target processing following fearful faces (P1); (2) an interaction between expression and gaze (N1), with enhanced target processing following fearful faces with rightward gaze; and (3) an interaction between gaze and target location (P3), with enhanced processing for invalidly cued left visual field targets. Behaviorally, participants responded faster to targets following fearful faces and targets presented in the right visual field, in concordance with the P1 and N1 effects, respectively. The findings indicate that two nonverbal social cues-facial expression and gaze direction-modulate attentional orienting across different temporal stages of processing. Results have implications for understanding the mental chronometry of shared attention and social referencing.