LondonApolipoprotein E ε2 reshapes 192 proteins before dementia begins.
A Nature Aging study finds the strongest upregulations — UNG, VPS29, and BIRC2 — appear in early adulthood, independent of Alzheimer’s diagnosis.
54 of 192 proteins mediate the pathway to Alzheimer’s, suggesting targets for upstream intervention.
Sources: Nature
THE THREE ALLELES
APOE comes in three flavors — ε2, ε3, ε4 — differing by just two amino acids. ε3 is the common variant (~78% of alleles). ε4 roughly triples Alzheimer's risk per copy; ε2 roughly halves it. One gene, three lifetimes of brain aging.
WHAT THE PROTEIN DOES
Apolipoprotein E is the brain's primary lipid shuttle. It packages cholesterol and phospholipids into particles that neurons need to build and repair membranes and synapses. Brain cells cannot import cholesterol from the bloodstream — the blood-brain barrier blocks it — so APOE is the local courier.
THE AMYLOID HYPOTHESIS
For three decades the dominant theory has been that amyloid-beta plaques drive Alzheimer's. Anti-amyloid antibodies (lecanemab, donanemab) do clear plaques but produce only modest clinical benefit. The field is increasingly asking what upstream of amyloid actually matters — which is the gap proteomic studies like this one are trying to fill.
WHY PROTEOMICS
Genomics tells you risk; proteomics tells you mechanism. A genome is fixed at conception, but the proteins it expresses shift across tissues, ages, and environments. Reading 7,000+ plasma proteins in thousands of people lets researchers find which molecules actually move between healthy and diseased states — the layer where drugs can intervene.
THE EARLY-ADULTHOOD CLUE
If ε2's protective proteins appear decades before any dementia onset, the protective mechanism is not a late-life response to damage — it is a lifelong shift in brain maintenance. The therapeutic implication is uncomfortable: meaningful prevention may need to start in the third decade of life, not the seventh.
THE MEDIATION FRACTION
54 of 192 proteins statistically mediate ε2's effect on Alzheimer's risk. Mediation analysis asks: if we held this protein constant, would the genetic effect disappear? A 28% mediation share means most of ε2's benefit flows through measurable, potentially druggable intermediaries rather than through unreachable genetic destiny.