Alzheimer’s Disease

What goes wrong in Alzheimer's

Alzheimer's disease destroys neurons in the brain regions responsible for memory and thought. Two molecular hallmarks drive that destruction: extracellular plaques of misfolded amyloid-β, and intracellular tangles of misfolded tau.

Ribbon diagram of an amyloid-beta fibril, PDB 2BEG — An Aβ42 fibril — the stacked cross-β arrangement that aggregating peptides assemble into. (Wikimedia Commons / Boku wa Kage, CC BY-SA)

Amyloid-β (Aβ42) is a short 42-amino-acid peptide that, in a healthy brain, gets cleared. In Alzheimer's, it instead sticks to other copies of itself — first into small soluble oligomers, then into long fibrils that pile up as plaques. The visible plaques used to be considered the toxic species, but it's now clear the small early oligomers do the real damage to neurons.

Tau is a structural protein inside neurons that helps stabilize their internal scaffolding. When tau misfolds and clumps into tangles, that scaffolding collapses and the neuron dies from the inside.

Neurofibrillary tangles of misfolded tau inside neurons in Alzheimer's disease — Neurofibrillary tangles of misfolded tau inside neurons — the second molecular hallmark of Alzheimer's. (Wikimedia Commons / NIA, public domain)

What Folding@home is doing

The small Aβ42 oligomers that kill neurons are notoriously hard to study in the lab — they're short-lived, structurally heterogeneous, and form on timescales between what experiments and conventional simulations can reach. Folding@home was built for exactly this problem.

Donated compute lets us run enough trajectories in parallel to map the kinetics of how Aβ42 monomers find each other and assemble into the toxic oligomers. From those maps, we can identify the points along the assembly pathway where a small molecule could intervene — and look for candidates that block aggregation before it starts.

Every CPU and GPU joining Folding@home pushes those simulations further. It's how a long-timescale problem gets solved by a lot of people, in parallel, in their off-hours.