Tau Filament Length Distribution Reflects End-To-End Annealing

Abstract

Tau is a microtubule associated protein that normally functions as a monomer in league with the microtubule cytoskeleton. However, in Alzheimer's disease (AD), tau aggregates to form filamentous inclusions in cell bodies (neurofibrillary tangles, NFT) and cell processes (dystrophic neurites and neuropil threads). The appearance of tau-bearing lesions correlates with neurodegeneration and cognitive decline, consistent with a connection between tau aggregation and disease progression. Indeed, in biological models, tau aggregates are toxic, with potency inversely proportional to aggregate size. However, the relationship between these species and the aggregation pathway is unknown.Here we investigate the aggregation mechanism of tau protein in vitro in an effort to identify interactions that manifest size dependence. The fits of a mathematical model describing simple nucleation-elongation polymerization to aggregation time-series, consistently overestimated filament growth rate while underestimating filament length distribution, indicating the presence of a secondary process in the pathway. On the basis of filament mixing and shearing experiments, we identified end-to-end annealing as a novel secondary interaction of nascent tau filaments. With the addition of an end-to-end annealing term to the mathematical model in equilibrium with filament fragmentation, we found improved fits for both time series and filament length distributions. In addition to quantifying the intrinsic rate constants for annealing and fragmentation, the model provided evidence for their dependence on filament length. The results indicate that filament ends are active, and that their propensity to engage in homotypic interactions is length dependent. We propose that heterotypic interactions at filament ends are candidate mediators of toxicity in biological models.