The brain was walled off from immunity — its skull bone marrow had other plans

There is an immune organ inside your skull that nobody put in an anatomy textbook.

It lives in the spongy bone marrow of the skull, contains structures resembling lymph nodes, and trains antibody-producing cells to recognize dangers coming from the brain — firing a response before the rest of the immune system has any idea something is wrong. In mice with brain cancer, disrupting these hidden hubs made tumors grow faster. Allowing them to work, then boosting them through the scalp with a gel, improved survival.

The discovery comes from the lab of Jonathan Kipnis at Washington University School of Medicine in St. Louis, and it doesn’t sit lightly beside what textbooks say about how the immune system and the brain relate. The prevailing model for most of the twentieth century was that the brain was immunologically privileged — the blood-brain barrier sealed it off, it had no lymph drainage worth noting, and immune surveillance there was minimal at best. That picture has been updated, piece by piece, over the last decade. This is the most dramatic update yet.

Recommended ReadingLost brain support cells were thought irreplaceable — adult brains are replacing them

“The skull bone marrow harbors previously unrecognized hubs for brain-specific immune responses,” Kipnis said.

How the hubs form and where they sit

The skull is not just a calcium case around the brain. It has channels — tiny physical passages — connecting the bone marrow inside it to the dura mater, the outermost membrane wrapping the brain. Immune cells and proteins can move through those channels. Brain waste products can exit through them. Kipnis’s group had mapped some of this previously.

What Jang Hyun Park, the study’s lead author, found was that the skull bone marrow contains structures resembling lymph nodes: clusters of T follicular helper cells alongside B cells. Together, these cell types do something specific: they run antibody production targeted at the brain. The T follicular helper cells train the B cells to recognize particular threats from brain tissue. The result is a local immune-training center, positioned directly adjacent to the brain, capable of generating antibody responses calibrated to what is happening in the skull’s closest neighbor.

Nothing like this had been described in healthy bone marrow before. Lead author Jang Hyun Park noted that these structures were entirely unexpected — standard anatomy gives no account of them.

What the glioblastoma experiments showed

To test whether these hubs actually mattered, the research team turned to glioblastoma, the most aggressive primary brain tumor. They used a mouse model and systematically disrupted the skull’s immune hubs with drugs. When the hubs were impaired, tumors grew faster and survival dropped.

The next step was the therapeutic implication: if the hubs exist and protect the brain, can they be strengthened? The team developed a gel mixture containing three immune-boosting proteins and applied it directly under the scalp, in physical contact with the skull. This targeted the bone marrow hubs locally, without systemic treatment. Mice receiving the gel showed better tumor rejection and lived longer than controls.

Critically, similar immune cells were found in human skull bone marrow — not only in mice. That doesn’t confirm the hubs function identically in humans, but it establishes that the architecture exists.

Why this matters beyond cancer

The implications extend well past glioblastoma. Alzheimer’s disease, Parkinson’s disease, schizophrenia, and long COVID all carry an immune component in the brain. If the skull bone marrow holds a standing immune station for the brain — one that samples brain proteins through the channels Kipnis’s group has been mapping — then it is a potential target for every neurological condition with an inflammatory dimension.

More concretely, the scalp-gel approach represents something that has been elusive in neurology: a way to get an immune intervention close to the brain without forcing it across the blood-brain barrier. The barrier exists precisely to protect the brain from circulating molecules; it also keeps most drugs out. The skull route sidesteps that problem entirely.

What this finding doesn’t settle

Nearly everything shown here was demonstrated in mice, and the translation from mouse brain biology to human brain biology is the field’s perennial challenge. The skull channels have been confirmed in humans; the immune hub structures have been found in human bone marrow samples; but the full functional role — whether they fight tumors in humans with the same urgency as in the mouse experiments — has not been tested.

The gel therapy is pre-clinical. No human trial exists. The protein mixture used in the experiments has not been shown to be safe or effective in people, and the question of what happens when the skull hubs are boosted in a brain that already has an inflammatory disease — rather than a cancer — is entirely open.

The timing claim also deserves scrutiny: the finding that the skull hubs respond before distant lymph nodes was measured in a glioblastoma model. Whether the same speed advantage holds across other brain conditions is not yet established.

Common questions about skull bone marrow and brain immunity

What is the blood-brain barrier and why does it matter for treatment?

The blood-brain barrier is a layer of tightly packed cells lining blood vessels in the brain. It blocks most substances in the bloodstream from reaching brain tissue — protecting the brain from pathogens and toxins, but also preventing most drugs from getting through. The skull-channel route to bone marrow bypasses this barrier, which is why it is therapeutically significant.

Is skull bone marrow different from bone marrow elsewhere in the body?

Ordinary bone marrow — in the femur, pelvis, sternum — produces blood cells and runs systemic immune functions. What this study found is that skull bone marrow has additional specialized structures, directly linked to the brain through physical channels in the bone, giving it a brain-specific immune role that other marrow sites don’t share.

Could this change how Alzheimer’s disease is treated?

Potentially, though that remains hypothetical for now. If the skull bone marrow monitors the brain for protein signatures — and Alzheimer’s is a disease of abnormal protein accumulation — then the hubs might be an entry point for therapeutic antibodies that IV-infusion approaches cannot efficiently deliver. No clinical trial tests this yet.

How does this fit with what Kipnis’s lab found before?

The Washington University group discovered previously that the dura mater contains lymphatic vessels — overturning the idea that the brain had no lymph drainage. They then described physical channels in the skull through which immune cells transit between bone marrow and dura. The current paper adds a third layer: the bone marrow itself is an active immune-training center with structures that were missed entirely.

The research team’s next priority is to test whether the scalp-gel approach can move toward human safety trials, and to characterize how the skull’s immune hubs change in age-related neurological disease — the conditions where the therapeutic payoff would be largest.

Reference: Park, J.H. et al., “Functional lymphoid structures within the skull bone marrow and their role in central nervous system immunosurveillance and immune responses to brain disease,” Nature, 2026. DOI: 10.1038/s41586-026-10951-4

Tags: neuroscience, blood-brain barrier, brain immunity, glioblastoma, Jonathan Kipnis, skull bone marrow

More Like This

Immune surprise: recently evolved alarm molecule drives inflammation

Blocking one protein turns your immune cells into cancer-killing machines

The sleeping brain replays bad memories with more precision than good ones

Mice carrying a naked mole rat gene lived 4.4% longer with less inflammation

Ultra-processed food is quietly dismantling your cellular architecture

Human Brain Project study offers insights into neurotransmitter receptor organisation