Alzheimer’s Disease (AD) is characterized by selective loss of synapses, resulting in decline of cognitive function. The pathology of AD is increasingly recognized to involve neuronal interactions with other brain cell types, notably microglia and astrocytes. Extracellular vesicles (EVs) are secreted by all cells in the brain, and carry protein, lipid and RNA cargo. EVs have the capacity to signal from donor to recipient cells within brain tissue and modify cell functions, as shown in cancers and multiple neurological diseases. EV propagation of pathologic proteins between cells in the brain is a strong candidate mechanism underlying at least some aspects of AD pathology, suggesting that distinct EV cargos may not only serve as biomarkers for disease but also directly induce vulnerability to or protection from pathologic disease states. Yet, little is known about EV cargo diversity and bioactivity, in particular whether EV cargos and signaling bioactivity vary by cell type, physiological state or genetic background of EV donor and recipient cells. To overcome these barriers, we propose to use iPSCs and direct reprogramming to generate human neurons, astrocytes and microglia from iPSCs that vary by their AD-related genetic background. We will then precisely define the EV cargos from these cell types using state-of-the-art proteomic and RNA-Seq methods. In parallel, we will address the functional consequence of EV diversity with two sensitive readouts of EV signaling pertaining to AD pathology: neuronal synaptic imaging, and transcriptomic profiling. These collaborative studies will generate critical insights into the relationship between predicted susceptibilities to AD and EV cargo diversity, EV bioactivity and AD-related functional changes in neurons. Results of these studies will produce a novel cell-type based proteomic and RNA catalog of EV diversity, identify candidate biomarkers to test in human AD studies and uncover candidate pathologic and protective mechanisms linked to specific genetic risks for late onset AD.