Alzheimer’s Top Risk Gene Scars the Brain’s Blood Vessels — and in Mice, It Could Be Reversed

The gene variant most strongly linked to late-onset Alzheimer’s disease, APOE4, appears to actively damage the brain’s small blood vessels by turning their support cells into scar-forming cells — and in aged mice, blocking a single signaling pathway reversed much of that damage. The findings come from two studies led by the Icahn School of Medicine at Mount Sinai, published on September 24, 2026 in Cell and Cell Stem Cell.

Why APOE4 matters

APOE is a gene that helps carry fats such as cholesterol around the body and brain. It comes in three common versions. About a quarter of people carry at least one copy of APOE4, which raises the risk of Alzheimer’s several-fold; carrying two copies raises it much further. Yet decades after that link was discovered, exactly how APOE4 does its damage is still being worked out. Most attention has gone to amyloid plaques and tau tangles in brain cells. These studies focus on something that has received less attention: the brain’s blood supply.

What the researchers found

The team built a single-cell atlas of the human brain’s blood vessels, cataloging the genetic activity of the different cells that make them up. They focused on pericytes — cells that wrap around the brain’s tiniest vessels, keep them stable, help control blood flow and maintain the blood-brain barrier.

In the presence of APOE4, pericytes changed identity. Instead of supporting the vessel, they began to look and act like myofibroblasts, the cells that lay down scar tissue elsewhere in the body. The result was vascular fibrosis — stiff, scarred vessel walls — along with a build-up of amyloid around the vessels. Both could reduce blood flow to brain tissue and, the researchers argue, help drive neurodegeneration.

The reversal in mice

The researchers traced the switch to TGF-β, a signaling molecule that drives scarring in many organs. In aged mice carrying APOE4, blocking TGF-β signaling restored pericyte coverage of the vessels and reduced both fibrosis and vascular amyloid. As senior author Joel Blanchard put it, damage to the brain’s blood vessels is not simply a late consequence of Alzheimer’s but “a biologically active process caused by APOE4 that may be reversible.”

A second study: cholesterol and waste disposal

The companion study used miBrains, lab-grown 3D models of human brain tissue built from stem cells, which include several brain cell types together. In these models, APOE4 caused cholesterol to build up in astrocytes, the brain’s support cells, jamming their lysosomes — the compartments that break down cellular waste. That impaired the clearance of alpha-synuclein, a protein that clumps in Parkinson’s disease and in some Alzheimer’s cases. The team also showed that the models can be frozen and revived, which should make experiments easier to repeat across labs.

Why the blood-vessel angle matters

Vascular problems have long been suspected to contribute to dementia: many people with Alzheimer’s also have damaged small vessels, and earlier research found that APOE4 carriers show signs of a leaky blood-brain barrier before memory problems begin. The new work adds a specific cellular mechanism and a candidate pathway to target.

It may also bear on current treatments. The anti-amyloid antibodies Leqembi and Kisunla carry a risk of brain swelling and small bleeds, known as ARIA, that is highest in people with two copies of APOE4, and amyloid in vessel walls is thought to play a role. A better understanding of how APOE4 damages vessels could eventually help make those treatments safer, or identify who should avoid them.

The caveats

This is early, preclinical research. The reversal was shown in mice, not people, and the human evidence comes from tissue and lab models. TGF-β is involved in many essential processes, including immune regulation and wound healing, so blocking it throughout the body has caused side effects in past drug trials; any therapy would likely need to target it precisely. And it is not yet known how much vascular damage contributes to memory loss compared with plaques and tangles.

What APOE4 carriers can take from this

Carrying APOE4 raises risk; it does not make Alzheimer’s certain, and many carriers never develop it. Genetic testing for APOE is generally not recommended for people without symptoms unless they have talked it through with a doctor or genetic counselor, because the result cannot be changed and there is no APOE-specific treatment. What the new research does reinforce is the long-standing advice that what is good for blood vessels is good for the brain: controlling blood pressure, blood sugar and cholesterol, not smoking, and staying physically active are linked to lower dementia risk in large studies, whatever a person’s genes. None of that has been shown to undo APOE4’s effect on pericytes, but it protects the same small vessels the studies highlight.

What comes next

The logical next steps are to test more targeted ways of blocking the pericyte switch, to see whether restoring vessels in animals improves memory and brain function, and to look for markers that show vascular scarring in living people. For the roughly one in four people who carry APOE4, the study offers a useful message: the gene’s effects may not be fixed. This is research news, not medical advice.