In this proposal, we propose studies to elucidate the basis of the hyper and hypoconnectivity observed in AD. We will combine high resolution in vivo MRI imaging strategies with state-of the-art spatial proteomics. By combining these methods, we will obtain these two types of data in the same animal, allowing us to incorporate machine learning (ML) to build a model to define spatially distinct protein expression changes in regions of hyper and hypoconnectivity over time. A recent MRI/PET imaging study suggested that tau tangles are responsible for hypoconnectivity, while Ab plaques are responsible for hyperconnectivity. This intriguing hypothesis, based on data from a heterogeneous group of patients, needs to be tested under carefully controlled conditions that can only be obtained using animal models of AD. PET imaging identifies tau tangles and Ab plaques but cannot determine which phosphorylated tau (p-tau) species are driving hypoconnectivity, nor can it determine the contribution of soluble Ab42 to the hyperconnectivity. In the studies proposed in this application, we will test three hypotheses: 1) Soluble Ab42 accumulation causes regions of hyperconnectivity. 2) A subset of soluble p-tau species is responsible for hypoconnectivity. 3) Regions of hyper and hypoconnectivity exhibit distinct spatial proteomic profiles that reflect early and late phases of the disease.