ABSTRACT Alzheimer’s disease (AD) is the most common cause of dementia; it is increasingly evident that cerebrovascular mechanisms are critically involved in AD pathology. Cerebrovascular disease frequently co- occurs with AD. One driver of this disease is pathologic fibrin deposition (PFD) in the cerebrovasculature caused by resistance to plasmin-mediated fibrin breakdown. PFD is a risk factor for AD, both non genetic (traumatic brain injury) and genetic (ApoE4), and amyloid beta (Aβ) modifies the fibrin clot, making it resistant to plasmin cleavage and worsening PFD. Thrombin cleaves fibrinogen to fibrin and it also activates platelets by cleaving and activating protease-activated receptors PAR1 and PAR4. We have found that platelet PAR4 is responsible for greater platelet activation, thrombin generation, and procoagulant microparticle formation than platelet PAR1. PAR4 is also expressed on endothelial cells, leukocytes, microglia and neurons, where it is activated by thrombin, leading to activation of microglia, blood brain barrier (BBB) breakdown and microbleeds. Lymphocyte PAR4 is required for homing to sites of inflammation and PAR4 expression is inducible under inflammatory conditions, leading to exaggerated local damage. F2RL3, the gene encoding PAR4, is hypomethylated under inflammatory conditions. Our hypothesis is that PAR4 antagonism may minimize PFD by 1) decreasing thrombin generation and fibrin deposition, and 2) blocking lymphocyte infiltration and local tissue damage. In support of our hypothesis, we have found in preliminary data that PAR4 is overexpressed in the cerebrovasculature of 5XFAD mice with 5 familial AD mutations. In addition, we have shown that in aged humans higher levels of F2RL3 expression in the prefrontal cortex are associated with a faster rate of cognitive decline. As a GPCR, PAR4 is an attractive therapeutic target, and we have made significant progress on generating PAR4 antagonists. However, they do not yet have potent activity against the endogenous tethered ligand in either human or mouse platelets, or the DMPK characteristics to be useful in vivo. In Aim 1, we propose to use a computational approach starting with comparative models of both mouse and human PAR4 to screen an ultra-large library using DOCK and ROSETTA, combined with a make-on-demand small molecule library. Based on the success of the Shoichet and Roth laboratories using this approach to generate pM potency, highly selective compounds for GPCRs, we expect to enrich the scaffold diversity of our SAR. We will iteratively screen compounds in mouse and human platelets, then redo the computational screening to further improve SAR. In Aims 2 and 3, traditional medicinal chemistry will then again iteratively improve the properties of antagonists for potency against the tethered ligand, selectivity and in vivo bioavailability. The best antagonist of mouse PAR4 will be tested in 5XFAD mice to determine if it can slow or block progression of AD sympto...