PROJECT SUMMARY Preclinical animal models of neurodevelopmental disorders (NDDs), including autism spectrum disorders (ASDs) and those of intellectual disabilities, have yielded insight into candidate genetic etiologies, but circuit-level hypotheses underlying behavioral alterations (particularly social impairments) have remained elusive. A significant challenge for developing biomarkers and diagnostic tools has been technological – because canonical neuronal tract-tracing techniques require histology of post-mortem tissue, longitudinal assessments of circuit development are few and far between. To address this critical unmet need, we propose to develop a novel, transient, and repeatable method that circumvents invasive histochemistry altogether and allows for longitudinal axonal-level tracing in developing marmoset monkeys. The common marmoset (Callithrix jacchus) is an ideal preclinical modeling species for studying circuit trajectories related to aberrant social development. We recently discovered that marmosets, like humans, possess socially selective face processing patches in frontal cortex. As such, marmosets can recapitulate primate-specific aspects of social behaviors that form the diagnostic criteria for many NDDs. In Aim 1, we will optimize our approach, which uses transcranial focused ultrasound (FUS) to deliver MnCl2, a magnetic resonance imaging (MRI)-visible axonal tracer to any prescribed region of the brain. In Aim 2, we seek to demonstrate that the FUS-MEMRI technique is a reliable screening tool for developmental circuit-level aberrations with test-retest reliability. We will leverage our genetically diverse breeding colony of marmosets to study the structural underpinnings of naturally occurring differences in social face processing from an early age. From two months into adulthood, we will employ the FUS-MEMRI method to track the axonal connectivity of our previously identified socially selective face patch in medial frontal cortex. This circuity is ostensibly involved in assigning social salience to faces, a classic hallmark deficit in ASDs. By employing our established fMRI paradigm – which is passive and can be used with very young marmosets – we will link axonal development to functional variability in social processing of faces. Further, we will employ trained machine- learning algorithms to isolate gaze patterns to relevant facial features, providing a behavioral index of face processing. With this noninvasive multimodal approach, we will establish parameters of healthy variability of this circuitry in our marmoset population, laying the groundwork for genetic and pharmacological models currently in development by our laboratory and others. This work will lead to the establishment of a novel noninvasive approach for longitudinal axonal tracing that is transient, targeted, and repeatable. This technique will have broad application in translational research, especially for neurodevelopmental applications, providing a mean...