Amyloid protein deposited in cerebral vessel walls and diffuse plaques of patients with hereditary cerebral hemorrhage with amyloidosis, Dutch type (HCHWA-D), is similar to the 40-42 residues amyloid beta (Abeta) in vessel walls and senile plaques in brains of patients with Alzheimer's disease (AD), Down's syndrome, and familial and sporadic cerebral amyloid angiopathy (CAA). In 1990 we sequenced the amyloid beta-protein precursor (AbetaPP) gene from HCHWA-D patients revealing a single mutation that results in an amino acid substitution, Abeta E22Q. Subsequent identification of additional mutations in the AbetaPP gene in familial AD (FAD) pedigrees revealed that whereas substitutions in the middle of Abeta, residues Abeta21-23, are predominantly vasculotropic, those found amino- or carboxyl-terminal to the Abeta sequence within AbetaPP enhance amyloid parenchymal plaque deposition. Studies of transfected cells showed that substitutions amino- or carboxyl-terminal to Abeta lead to either greater Abeta production or to enhanced secretion of the more hydrophobic thus more fibrillogenic Abeta1-42. Substitutions in the center of Abeta facilitate rapid aggregation and fibrillization, slower clearance across the blood-brain barrier and perivascular drainage to the systemic circulation, possibly higher resistance to proteolysis, and enhanced toxicity towards endothelial and smooth muscle cells. However, most AD patients have no genetic defects in AbetaPP, indicating that other factors may alter Abeta production, conformation, and/or clearance initiating the disease process.