The goal of the Disease Model Development and Phenotyping Project (DMDPP) is to produce and characterize the next generation of animal models for AD that accurately model the pathology of late onset AD (LOAD) and to provide predictive models for therapeutic development. These models will be generated under transparent and open intellectual property conditions. The Jackson Laboratory will conduct 2nd-site validation of LOAD models and will ensure their broad availability and rapid dissemination to all researchers. The foundation of this endeavor is our innovative APP knock-in (APP-KI) mouse that expresses humanized A at physiological levels, and which exhibits amyloid plaque deposition (see introduction). We will complete our base platform for development of mouse models of LOAD, by humanizing the mouse Tau (Mapt) locus by replacing coding exons of mouse Tau with those from the human Tau (MAPT) locus. To double homozygous APP-KI and hTau (APP-KI+/+/hTau+/+) mice, we will add the major risk factor for LOAD, APOE4. To expedite analysis of additional factors influencing LOAD, the DMDPP will make extensive use of CRISPR/Cas9 technology to generate animal models that express polymorphisms in risk factors identified from genome wide association studies (GWAS), including TREM2, PICALM, BIN1, CD2AP, ABCA7 and EPHA1. The effect of each GWAS allele in generating a LOAD phenotype will be determined when combined with APP-KI, and hTau. Subsequently we will combine specific GWAS alleles to investigate for synergistic effects on LOAD pathology. The incorporation of multiple GWAS polymorphisms on the APP-KI+/+/hTau+/+ background and production of cohorts of mice for analysis will be accelerated using IVF and embryo transfer instead of standard breeding. The molecular pathological phenotypes in these LOAD models will be characterized at an unprecedented level of detail through a novel and innovative immunological approach, using conformation dependent and aggregation specific monoclonal antibodies that distinguish eight different types of amyloid deposits in humans and transgenic mice. The molecular phenotype will also be comprehensively determined by quantifying neurons and microglia, synaptic loss, soluble and insoluble tau and A and markers of phospho tau. The core will also characterize the gene expression profile of the models by RNAseq and epigenetic markers, as well as structural, functional, and diffusion magnetic resonance imaging. The behavioral phenotype will be characterized by elevated maze, open field, and novel object recognition. The extensive data generated will be used to compare the molecular pathology to that of LOAD and to decide which lines to advance for further development and testing to produce the next generation of animal models.