A number of recent studies have caught my attention.
At first glance, they seem to be investigating very different problems: cholesterol trapped inside astrocytes, blood-brain barrier dysfunction, fibronectin accumulating around blood vessels, and pericytes changing their identity.
But the more I read, the more I see dots being connected.
Taken collectively, these studies point back to many of the same cells and biological systems. And increasingly, I wonder whether APOE4 creates an environment in which the brain becomes less capable of maintaining homeostasis - the delicate balance necessary for healthy cellular function.
Over time, disruptions in lipid metabolism, cellular cleanup, vascular integrity and inflammatory regulation may gradually erode the brain’s resilience and eventually, the familiar hallmarks of Alzheimer’s disease can become apparent.
But what if those hallmarks are, at least partly, the downstream consequences of a much broader biological problem?
APOE4 is intrinsically involved in lipid biology and transport
This is an important place to start.
APOE is not an Alzheimer’s gene by design.
Apolipoprotein E is a lipid transport protein.
In the brain, astrocytes are the major producers of APOE.
Astrocytes are often described as support cells for neurons, but that really understates their importance.
They supply neurons with lipids, help regulate energy metabolism, recycle neurotransmitters, participate in immune responses and contribute to maintaining the blood-brain barrier.
Several studies have now demonstrated that APOE4 changes how these cells handle cholesterol.
In human APOE4 astrocytes, cholesterol can become trapped inside lysosomes - the cellular recycling centers.
Remarkably, despite accumulating too much cholesterol in these compartments, the cell can behave as though it doesn’t have enough available cholesterol and continue synthesizing more.
Researchers have also identified reduced cholesterol efflux (i.e. removing excess cholesterol) from APOE4 astrocytes.
So the problem isn’t simply “too much cholesterol.”
It appears to be, at least partly, cholesterol in the wrong place and impaired lipid trafficking.
Studies have linked this lysosomal cholesterol accumulation to impaired mitochondrial function, defective autophagy and a reduced ability of astrocytes to dispose of damaged cellular material.
More recent research has extended this concept further, showing that cholesterol-mediated lysosomal dysfunction in APOE4 astrocytes can interfere with the processing of α-synuclein, another misfolded protein associated with neurodegeneration.
Suddenly, lipid metabolism, cellular cleanup and protein aggregation don’t look like entirely separate problems - but may well be connected.
Now look at what is happening around the brain’s blood vessels
This is where the story becomes even more interesting.
Astrocytes don’t simply support neurons.
Their endfeet surround the tiny blood vessels of the brain and form an essential part of what is called the neurovascular unit - the coordinated system of endothelial cells, pericytes, astrocytes and other cells that maintains the blood-brain barrier.
APOE4 appears to disrupt this system through several different mechanisms.
Researchers have known for years that APOE4 can activate an inflammatory pathway in pericytes - cells that wrap around brain capillaries and help maintain blood-brain barrier integrity.
The pathway is usually abbreviated:
CypA → NF-κB → MMP9
Activation of this pathway can degrade components of the blood-brain barrier and make it more permeable.
Importantly, APOE4-associated vascular dysfunction has been observed independently of amyloid pathology, raising the possibility that blood-brain barrier impairment can occur relatively early in APOE4 carriers. The CypA pathway is what the late Dr. Alan Green based his rapamycin theory on to protect APOE4 carriers - and why I’ve been taking this off-label drug since 2021. More about this in a previous post.
But two newer discoveries add fascinating pieces to this puzzle.
The fibronectin connection
Recent research published in Nature Aging reported that APOE4 and amyloid-β42 can drive excessive production of fibronectin by astrocytes.
Fibronectin is an extracellular matrix protein.
We need it. It plays important roles in tissue structure, repair and cellular communication.
The problem appears to arise when too much accumulates around the brain’s blood vessels.
The researchers found that excessive astrocyte-derived fibronectin at the gliovascular interface interfered with communication between astrocytes and endothelial cells and disrupted growth-factor signaling necessary for maintaining the blood-brain barrier.
In other words:
APOE4 astrocyte → excess fibronectin → disrupted astrocyte/blood-vessel communication → impaired blood-brain barrier
This wasn’t simply an observation in mice.
The researchers combined human genetics, human postmortem brain tissue, cerebrospinal fluid, single-cell analyses, human iPSC-derived cerebrovascular models, zebrafish and mouse models.
And there is an especially fascinating piece of human genetic evidence supporting the importance of this pathway.
Some APOE4 carriers appear to have genetic protection
Researchers previously identified a rare variant in FN1, the gene encoding fibronectin, that was associated with approximately a 71% reduction in Alzheimer’s risk among APOE4 carriers.
Think about what that potentially tells us.
These individuals still carry APOE4.
Yet a variation in another gene, affecting a pathway downstream of APOE4, appears to substantially modify their risk.
That is an extraordinarily important concept for those of us who carry this gene.
APOE4 is not destiny.
If altering a downstream biological pathway can dramatically modify the effect of APOE4, understanding those downstream pathways becomes enormously important.
The FN1 finding is an association, not proof that reducing fibronectin in humans will prevent Alzheimer’s. But it provides a compelling reason to investigate the mechanism.
And now pericytes enter the story again
Another new study adds an almost uncanny connection.
Researchers examining brain blood vessels in APOE4 models found fewer normal pericytes and the emergence of cells resembling myofibroblasts - cells associated with scar formation and fibrosis.
The researchers found that APOE4 pericytes were essentially changing their identity.
And what did these new myofibroblast-like cells produce?
Fibronectin.
The process appeared to be driven by increased TGF-β signaling.
The proposed sequence looked something like this:
APOE4 → increased TGF-β signaling → pericyte transition toward myofibroblast → increased fibronectin → vascular fibrosis → increased vascular amyloid
This is particularly interesting because the fibronectin story is now appearing from two directions.
Astrocytes can produce excess fibronectin.
APOE4 pericytes transitioning toward myofibroblasts can also produce fibronectin.
Both converge on the brain vasculature.
And in experimental human brain models, inhibiting TGF-β signaling restored pericyte coverage and reduced both fibronectin and vascular amyloid to levels approaching those seen in APOE3 models.
That does not mean we currently have a proven TGF-β treatment for APOE4 carriers.
These are experimental findings, not a clinical recommendation.
But biologically, they are extremely intriguing.
Put the pieces together
We now have several APOE4-related pathways that appear to converge.
Impaired lipid trafficking → astrocyte dysfunction → disrupted cellular maintenance
Alongside:
APOE4-driven inflammatory and TGF-β signaling → pericyte dysfunction → vascular abnormalities
Both may contribute to:
Dysfunctional neurovascular unit → extracellular-matrix abnormalities, including fibronectin accumulation → blood-brain barrier dysfunction → impaired vascular function and potentially impaired clearance → greater vulnerability to amyloid, tau and other neurodegenerative pathology.
Some mechanisms may operate independently, while others reinforce one another through feedback loops.
But the convergence is becoming difficult to ignore.
And it raises another interesting question.
Could HDL play a bigger role than we thought?
One question that keeps coming back to me as I read these studies is whether we’ve been thinking about HDL too narrowly.
We tend to think of HDL as the good cholesterol, primarily because of its role in transporting cholesterol away from peripheral tissues.
But HDL particles do considerably more than carry cholesterol.
They participate in inflammatory signaling, endothelial function, antioxidant transport and vascular homeostasis.
Increasingly, researchers are looking beyond how much cholesterol HDL carries to the composition and functionality of the particles themselves.
If APOE4 contributes to dysfunctional lipid trafficking, impaired astrocyte function and a compromised neurovascular unit, could improving certain aspects of lipid transport and vascular signaling help counteract some of those vulnerabilities?
This is one reason I’ve been following the obicetrapib research so closely.
As I’ve discussed previously, obicetrapib inhibits CETP, substantially lowers LDL cholesterol and produces unusually large increases in circulating HDL cholesterol.
In clinical studies, HDL-C has more than doubled on average.
But the potentially interesting story may not be the dramatic rise in HDL cholesterol itself.
It may be what happens to the composition and biological activity of HDL particles - and whether those changes influence the environment in which the brain’s blood vessels and supporting cells operate.
Here we need to make an important distinction.
The brain largely manages its own cholesterol economy, and the blood-brain barrier limits the exchange of circulating lipoproteins with the central nervous system.
Consequently, dramatically increasing HDL cholesterol in the bloodstream does not necessarily improve cholesterol trafficking inside APOE4 astrocytes.
Nor does a higher HDL-C concentration automatically mean better HDL function.
But HDL-associated molecules can influence endothelial signaling, inflammation and vascular biology.
And this is where some of the newer obicetrapib research becomes particularly interesting.
Beyond HDL cholesterol: What happens to the particles themselves?
Recent research examining CETP inhibition has reported changes in HDL particle composition, including components such as apolipoprotein A-I, apolipoprotein M and sphingosine-1-phosphate (S1P).
These are interesting molecules because HDL particles are involved in far more than simply transporting cholesterol.
For example, ApoM-associated S1P participates in endothelial signaling and vascular barrier regulation.
Given the growing evidence linking APOE4 to blood-brain barrier dysfunction, this is a biologically interesting connection.
But it is important not to jump ahead of the evidence.
Changes in circulating HDL composition have not been shown to reverse APOE4-related blood-brain barrier dysfunction in humans.
Nevertheless, they raise a question worth investigating:
Could modifying HDL-associated signaling influence the health of the neurovascular unit, even if it doesn’t directly correct cholesterol trapped inside astrocytes?
There is another intriguing observation.
In a small exploratory study involving patients with early Alzheimer’s disease, obicetrapib treatment was associated with changes in certain antioxidant-related compounds measured in cerebrospinal fluid, including carotenoids and tocopherols.
Although preliminary, these findings raise the possibility that modifying peripheral lipid metabolism may influence the availability or transport of certain compounds relevant to the central nervous system.
That is not proof of improved brain health.
But it adds another piece to the biological puzzle.
And what about the Alzheimer’s biomarkers?
In an exploratory analysis of the BROADWAY trial, researchers examined Alzheimer’s-related blood biomarkers, including phosphorylated tau 217 (pTau217).
Among participants carrying two copies of APOE4, obicetrapib was associated with an approximately 7.8% reduction in pTau217 over 12 months, compared with an approximately 12.7% increase in the placebo group.
That’s a difference of roughly 20.5 percentage points between the groups.
Other biomarkers, including GFAP and neurofilament light chain, also showed potentially encouraging differences.
But there are important limitations.
The APOE4/4 subgroup contained only 29 participants, and BROADWAY was primarily a cardiovascular trial, not an Alzheimer’s prevention trial.
And even a reduction in pTau217 cannot, by itself, demonstrate clinical benefit.
Nevertheless, for those of us carrying two copies of APOE4, the findings are difficult to dismiss as uninteresting and are certainly worth investigating further.
Which is exactly what the new SPINOZA clinical trial aims to do.
SPINOZA is investigating obicetrapib in preclinical Alzheimer’s disease, with a particular focus on Alzheimer’s-related biomarkers.
One especially interesting question is whether any observed effects are driven by HDL changes themselves, by other consequences of CETP inhibition, or by mechanisms we haven’t yet fully identified.
We simply don’t know.
But the broader possibility is intriguing.
Perhaps modifying lipid biology could influence several downstream consequences of APOE4 - not necessarily by directly removing cholesterol trapped inside astrocytes, but through changes in vascular signaling, inflammation, lipid transport or other aspects of cellular homeostasis.
That possibility remains a hypothesis, not an established treatment strategy.
Still, it fits remarkably well with the broader picture emerging from these recent studies.
Maybe amyloid isn’t the beginning of the story
None of this makes amyloid irrelevant.
Amyloid clearly participates in Alzheimer’s disease, and some of these pathways may create feedback loops in which amyloid further worsens vascular and astrocyte dysfunction.
But it changes how I think about the disease.
Rather than imagining a relatively simple sequence:
APOE4 → amyloid → Alzheimer’s
Perhaps the biology looks more like:
APOE4 → altered lipid handling + astrocyte dysfunction + pericyte dysfunction + vascular/BBB injury → impaired resilience and clearance → amyloid/tau pathology → neurodegeneration
And there are probably many feedback loops between these processes.
Some may begin before substantial amyloid accumulation. Others may become more damaging as amyloid and tau pathology develop.
But for APOE4 carriers, I actually find it considerably more hopeful.
Why?
Because every mechanism researchers uncover creates another potential intervention point.
The FN1 protective variant is a particularly powerful example.
Those individuals didn’t get rid of their APOE4 allele.
Another gene apparently changed how APOE4-related vulnerability translated into Alzheimer’s risk.
Likewise, researchers have experimentally improved APOE4-associated abnormalities by correcting cholesterol trafficking, restoring lysosomal function, reducing fibronectin accumulation or interrupting TGF-β signaling.
And now, the obicetrapib research raises another possibility: that modifying lipid metabolism and lipoprotein biology might influence some of the downstream processes associated with APOE4.
Most of these approaches are nowhere near ready for clinical use in Alzheimer’s prevention.
And obicetrapib, despite its encouraging biomarker findings, has not yet demonstrated that it prevents Alzheimer’s disease.
But collectively, these studies suggest something important:
The biology downstream of APOE4 is not necessarily fixed.
It can be modified experimentally, and in some circumstances, human genetic differences appear to modify the risks associated with APOE4.
Perhaps the future of Alzheimer’s prevention will involve intervening at several points simultaneously - improving lipid trafficking, preserving astrocyte function, protecting the neurovascular unit, maintaining blood-brain barrier integrity and addressing amyloid and tau pathology.
Perhaps the goal will be less about targeting one pathological protein and more about restoring the brain’s capacity to maintain its own equilibrium.
For those of us living with APOE4, that is an encouraging direction for research.
Because ultimately, our genes may establish vulnerabilities.
They don’t necessarily determine the outcome.