Adolescent Development of Amygdalo-prefrontal Circuit Microstructure and Its Role in Prosocial Behavior of Rhesus Macaques

Adolescence is marked by profound physical and hormonal development, reorganization of the brain, and corresponding changes in behavior. In primates, adolescence is a lengthy process that allows the development of skills to navigate complex hierarchical societies that determine access to resources such as food, shelter, and mates. One such skill is the ability to use an other-oriented reference frame for decision making, leading to more prosocial choices; those that benefit others. The emergence of adult-like prosocial behavior depends on successful remodeling of amygdalo-frontal circuits that support the integration of social and reward related information. The goal of this study is to capture these behavioral and neural changes longitudinally and determine their relationship to each other and to development of the body in rhesus macaque monkeys. To measure prosocial behavior, we longitudinally measured the performance of six male adolescent rhesus monkeys (ages 22-72 months) on a social reward allocation task in which they chose between rewarding only themselves or themselves and a pair-bonded partner in one trial type (self-oriented); and between rewarding only their partner or dispensing a reward into a container both participants can see but cannot access (other-oriented). On the same monkeys, we obtained monthly morphological and endocrine measurements. While age alone was an insufficient predictor of behavior, an increase in prosocial behavior (Wilcoxon signed-rank test, p < 0.05) was found at the onset of puberty, identified in each monkey as the first age where testosterone levels increased by greater than 0.5 ng/mL. We also periodically acquired diffusion-weighted MRI images and explored the relationships between development and diffusion metrics in the amygdala, orbitofrontal cortex (OFC), anterior cingulate cortex (ACC), uncinate fasciculus (UF), and cingulum. We found changes in diffusion metrics of the amygdala, OFC, and ACC that could be indicative of synaptic pruning that were significantly predicted by age and estradiol, but this development was not homogenous throughout amygdala nuclei (t-test, p < 0.05). We also found changes in the cingulum that could be indicative of increased myelination that were significantly predicted by age and estradiol (t-test, p < 0.05). Unexpectedly, changes in diffusion metrics in the UF suggest both decreasing and increasing myelination or tract integrity significantly predicted by age or testosterone, respectively (t-test, p < 0.05). Finally, we selected models of self- and other-oriented behavior predicted by subsets of structural features in all areas by comparing the AIC and BIC of nested models. We determined that behavior in the self-oriented trials was best predicted by the amygdala, cingulum, and ACC; while behavior in the other-oriented trials was best predicted by the amygdala, cingulum, ACC, UF, and OFC. Our findings indicate that development in the amygdala-cingulum-ACC circuit is involved in both types of prosocial behavior, but that other-oriented behavior may also involve the amygdala-UF-OFC circuit. We also find that the behavioral development predicted by microstructure in these circuits may be most prominent early in adolescence in monkeys. The results also underscore that while age is a useful predictor of adolescent development, endocrine status is an important predictor of both behavioral and neural development.