We will isolate human tau proteins overexpressed in E.coli and mouse brain to study the effect of various chemical and physical perturbation of tau filament formation. We will test whether the colloidal aggregation model for amyloid fiber formation occurs in mouse brain. We introduced the colloidal aggregation model to the amyloid community one decade and a half ago, based on our own study of fibers formed by five different proteins, namely human tau40, A-beta, a-synuclein, mouse serum amyloid A, and yeast Sup35, we introduced a two-stage colloidal aggregation model to elucidate protein aggregation and formation of amyloid fibers. Other evidence supporting the colloidal aggregation model comes from the study of two non-amyloidogenic particles, colloidal gold and low-density lipoprotein, that are capable of linear assembly into fibers. It is well known that seeding and agitation promote colloidal aggregation; for example, seeding clouds. AFM, TEM, confocal microscopy will be used in combination with mass spectrometry, FPLC, automated biofluid chemistry analyzer, PCR, dynamic light scattering, Circular Dichroism, and conventional biochemistry and molecular biology approaches. These methods allow the analysis both of a population of molecules as well as an individual or single molecule or particle level. Theoretical calculations of the energy involved, the application of DLVO theory to colloidal aggregation and behavior, provide a deeper understanding of the mechanism of amyloid fiber formation. Such a theoretical approach has allowed us to elucidate the conformational change to cross-beta sheet during amyloid fiber formation. Energy calculation provides prediction for the events and pathways involved in the process.