What Protein Causes Alzheimers? Amyloid-Beta and Tau

Two cooperating culprits drive this protein-linked dementia, with amyloid-beta plaques and tau tangles working in tandem to erode memory and cognition. Amyloid-beta clumps outside neurons into plaques, while tau twists into tangles inside neurons, and together they dismantle the memory circuits that define cognitive decline.

This article explains how amyloid-beta and tau misbehave at the cellular level, why certain genes accelerate their buildup, and how that pathology translates into the memory loss clinicians look for during diagnosis.

The Two Proteins Behind Alzheimer’s Disease

For decades, pathologists peering through microscopes have identified two abnormal proteins that consistently mark the cellular landscape of the disease. The first, amyloid-beta, accumulates as sticky deposits between brain cells called amyloid plaques. The second, tau protein, twists into thread-like neurofibrillary tangles inside the neurons themselves.

A brain can show one without the other, yet a definitive Alzheimer’s diagnosis requires both pathologies to appear together. That dual requirement matters because it explains why a single protein marker rarely tells the whole story.

Cognitive decline tracks more tightly with tau tangle burden, while amyloid plaques build up years earlier and set the stage. Treating one without the other misses half of the picture.

Why Both Hallmarks Are Required

Post-mortem studies consistently find that roughly one-third of cognitively healthy older adults carry significant amyloid plaque load without tau pathology. These individuals may never develop dementia. When both proteins appear together, the odds of clinical symptoms rise sharply.

That pattern pushed diagnostic criteria updates in 2011 and again in 2018, allowing biomarker evidence of amyloid and tau to support a diagnosis before death. National Institute on Aging and Alzheimer’s Association guidelines now codify that dual-protein standard.

Amyloid-Beta and the Plaque Formation Process

Amyloid-beta peptides are short fragments, 36 to 43 amino acids long, sliced from a much larger membrane protein called amyloid precursor protein (APP). Under normal conditions, APP is cleaved by alpha-secretase, which releases harmless fragments. In the Alzheimer’s pathway, two other enzymes, beta-secretase and gamma-secretase, make cuts at slightly different spots, producing sticky A-beta peptides that escape the cell.

The Amyloid Cascade Hypothesis

First proposed in the early 1990s, this hypothesis argues that a slow buildup of A-beta triggers a cascade that includes inflammation, synaptic failure, tau tangles, and eventual neuron death. It is not a flawless model, but it has shaped drug development for 25 years.

Most anti-amyloid therapies, including aducanumab (Biogen), lecanemab (Eisai), and donanemab (Eli Lilly), were designed to test whether removing plaques can slow cognitive decline.

Once misfolded, A-beta peptides behave like prions in slow motion. They recruit nearby normal peptides into the same abnormal shape, seeding plaques that grow for years before symptoms appear. Imaging studies show amyloid deposition can begin 15 to 20 years before memory problems surface.

Plaque accumulation may take decades to surface clinically, yet the second protein involved often tracks more tightly with the damage patients actually feel.

“Cleavage matters: the same APP protein produces either harmless fragments or disease-causing ones, depending on which enzyme does the cutting.

Tau Protein and the Neurofibrillary Tangle Mechanism

Tau plays a completely different role than amyloid-beta. Inside healthy neurons, tau acts like scaffolding, binding to microtubules, the structural tracks that transport nutrients and signaling molecules along the axon. When tau becomes hyperphosphorylated (too many phosphate groups attach to the protein), it loses grip on those microtubules and falls free into the cell body.

From Free Tau to Tangled Neurons

Detached tau does not stay inert. It threads together with other tau strands into paired helical filaments, then into dense neurofibrillary tangles that clog the neuron. A 2020 review in Nature Reviews Neuroscience noted that tangle density in the entorhinal cortex and hippocampus correlates more tightly with cognitive decline than amyloid plaque count does, which is why tau biomarkers now play a central role in staging the disease.

Genetic Factors That Drive Abnormal Protein Buildup

Most Alzheimer’s cases are sporadic, meaning they arise without an obvious inherited cause. A small share, under 1%, runs in families as early-onset Alzheimer’s, typically striking people in their 40s or 50s. Three genes explain the bulk of these familial cases: APP (which encodes amyloid precursor protein), PSEN1, and PSEN2 (presenilin mutations that alter the gamma-secretase enzyme complex).

Mutations in any of these shift APP cleavage toward longer, more aggregation-prone A-beta species, which is why families carrying them show near-certain early-onset disease.

APOE4 and Late-Onset Risk

Carrying one copy of the APOE4 allele roughly triples the risk for the late-onset form, making it the single most influential common genetic variant identified so far. Carrying one copy roughly triples lifetime Alzheimer’s risk; carrying two copies raises it roughly eight- to twelve-fold.

APOE4 appears to impair the clearance of A-beta from the brain, allowing plaques to build faster than microglia (the brain’s resident immune cells) can remove them.

Those inherited risks help explain why some families accumulate plaques far faster than others, setting the stage for the clinical picture that follows.

• ▸APP mutations: Directly alter the substrate, producing more aggregation-prone A-beta from birth.
• ▸PSEN1 and PSEN2 mutations: Change gamma-secretase’s cut site, favoring longer, stickier peptides.
• ▸APOE4 alleles: Slow A-beta clearance across the blood-brain barrier.
• ▸MAPT variants: Rare mutations in the tau gene itself can cause frontotemporal dementia with prominent tau pathology.

“Family history is not destiny: APOE4 raises risk but does not guarantee the disease, and many carriers never develop dementia.

From Brain Pathology to Symptoms and Diagnosis

Both amyloid plaques and tau tangles follow a predictable anatomical route. Tau pathology usually appears first in the entorhinal cortex, the gateway to the hippocampus, then spreads to the hippocampus proper (the brain’s memory index), and finally to the neocortex.

Amyloid deposition tends to spread more diffusely but still favors default-mode network regions. The hippocampus is particularly vulnerable because it depends heavily on microtubule-based transport, which tau dysfunction disrupts directly.

Detecting Pathology in Living Patients

Two biomarker tools now let clinicians see this process before death. Amyloid PET scans use a radioactive tracer that binds specifically to fibrillar A-beta, lighting up plaque-rich regions on the scan. Tau PET works similarly, with newer tracers detecting paired helical filament tau.

Cerebrospinal fluid (CSF) tests measure A-beta 42, total tau, and phosphorylated tau ratios. Together, these tools support an Alzheimer’s diagnosis even when memory symptoms are mild.

Prion-Like Spreading

In this model, a single misfolded tau seed slipping out of one neuron can corrupt a healthy neighbor, and similar behavior may apply to A-beta as well. The pattern of progression mirrors the brain’s connectivity map rather than its geography, which explains why memory networks collapse first while motor regions stay intact for years.

Because symptoms follow the disease’s spread along connected networks, therapies must reach those pathways, not merely shrink the proteins themselves.

Treatments Targeting Amyloid and Tau Proteins

Because amyloid-beta was long the easier target, the first generation of disease-modifying therapies aimed at it. Aducanumab (Aduhelm) received accelerated FDA approval in 2021 based on plaque reduction, though clinical benefit was debated. Lecanemab (Leqembi) followed in 2023 with clearer evidence of slowing cognitive decline by about 27% over 18 months in early-stage patients.

Donanemab, developed by Eli Lilly, showed similar results in 2023 trials, with stronger effects in patients with low-to-medium tau burden.

Anti-Tau Approaches in the Pipeline

Anti-tau immunotherapies have lagged behind. Several vaccines and monoclonal antibodies have entered clinical trials, but no tau-targeting therapy has yet matched the amyloid drugs’ clinical endpoint results. Aggregation inhibitors and kinase modulators (drugs that control the enzymes adding phosphate groups to tau) are also under study, though early results have been mixed.

Limits of Protein-Focused Treatments

Anti-amyloid drugs do not reverse damage already done. They slow progression by roughly a quarter to a third in patients treated early, and they carry risks such as amyloid-related imaging abnormalities (ARIA), which include brain swelling and microbleeds visible on MRI.

For everyday risk reduction, the evidence still points to cardiovascular and metabolic health: controlling blood pressure, managing diabetes, staying physically active, and treating hearing loss all correlate with lower dementia incidence, even though none of these directly targets the abnormal proteins in Alzheimer’s disease.

• ▸Anti-amyloid antibodies: Lecanemab and donanemab reduce plaques and modestly slow decline in early disease.
• ▸Tau immunotherapies: Several candidates are in Phase 2 and Phase 3 trials, none yet approved.
• ▸Symptomatic drugs: Cholinesterase inhibitors and memantine help cognition but do not clear plaques or tangles.
• ▸Lifestyle factors: Exercise, sleep quality, vascular health, and social engagement remain the most accessible protective levers.

The Bottom Line

Alzheimer’s disease is a two-protein problem: amyloid-beta seeds the plaques that disrupt signaling between neurons, and tau tangles carry out much of the actual neuron damage once it begins. Both must be present for a definitive diagnosis, and both shape how the disease spreads through connected memory networks.

Current treatments can clear plaques and modestly slow decline, but halting the disease will almost certainly require attacking amyloid and tau at the same time. You can support that effort by staying current on biomarker-based diagnosis and discussing early evaluation with your clinician if symptoms appear.

FAQ