An enzyme in the endothelial cells lining brain blood vessels keeps a vessel-remodelling factor in check. Remove it in mice and the cerebellum forms too few synapses, leaving the animals clumsy.

Blood vessels in the brain are usually described in terms of plumbing and security: they deliver oxygen and glucose, and the blood-brain barrier they form keeps unwanted molecules out. A study in Science Signaling adds a third job. The cells lining those vessels appear to help determine how the brain’s motor circuits are wired.

Researchers led by Jinjin Lu report that an enzyme in brain endothelial cells, PRMT5, indirectly controls the formation of synapses in the cerebellum, and that mice lacking it are poorly coordinated.

“Beyond its critical barrier and circulatory functions, the vasculature thus contributes molecular instructions” that shape neural circuit connectivity, wrote Dario Bonanomi of the IRCCS San Raffaele Scientific Institute in a related Focus article accompanying the paper.

The cerebellum and fine control

The cerebellum sits at the back and base of the brain and contains more neurons than the rest of the brain combined. It does not initiate movement so much as refine it, coordinating timing and precision, which is why damage to it produces not paralysis but clumsiness: unsteady gait, imprecise reaching, tremor during deliberate movement.

Its circuitry is unusually well characterised, which makes it a favoured place to study how synapses form. Much of its function depends on inhibitory connections, which dampen rather than excite the cells they contact, and the balance between excitation and inhibition is what allows fine control rather than crude activation.

The cerebellum’s most distinctive cells, the Purkinje neurons, are among the largest in the brain and carry elaborate branching dendrites that receive enormous numbers of inputs. They are themselves inhibitory, and they form the sole output of the cerebellar cortex, which is why the balance of inhibition reaching them determines what the structure sends on.

An enzyme in the vessel wall

PRMT5 is a protein arginine methyltransferase, an enzyme that attaches methyl groups to arginine residues on other proteins. Such modifications can alter how proteins behave and how genes are read, which is why the paper describes the mechanism as epigenetic.

In brain endothelial cells, the team found, PRMT5 suppresses a factor called ANGPTL4, angiopoietin-like 4, which is involved in remodelling blood vessels. The enzyme keeps ANGPTL4 production turned down.

When the researchers deleted Prmt5 in mice, ANGPTL4 rose. The consequence was not primarily vascular but neural: excess inhibitory signalling, fewer synapses formed in the cerebellum, and motor deficits in the animals.

Deleting a gene only in a particular cell type is now standard practice, usually achieved by engineering animals so that the deletion occurs wherever a chosen cell-specific gene is active. It is what allows a claim about endothelial cells specifically, rather than about the animal as a whole.

The causal chain is the interesting part. A change inside a blood vessel cell altered the output of a signalling molecule, and that molecule changed how neurons nearby connected to one another. The vessel was not merely supplying the circuit; it was influencing its construction.

Vessels that talk to neurons

The idea that blood vessels participate in brain development rather than just servicing it has gathered support over the past two decades. Blood vessels and nerves grow through tissue along parallel paths, guided in part by the same molecular cues, and some factors originally identified as vascular turn out to act on neurons as well.

Vascular endothelial growth factor is the best-known case: described as a stimulator of blood vessel growth, it also affects neuronal survival and growth. The broader concept, sometimes called the neurovascular unit, treats neurons, glia and blood vessels as a functional ensemble rather than separate systems.

This study extends that logic to synapse formation, and specifically to inhibitory synapses, which have had less attention than excitatory ones.

What the study does not establish

The work is preclinical and used male mice, a detail worth noting because sex differences in both vascular biology and cerebellar development are documented, and a result in male mice may not hold in females.

The release reports no quantitative results at all: no counts of synapses, no measures of motor performance, no effect sizes. What it conveys is the shape of the mechanism rather than its magnitude, so how large these effects are, and how much of normal cerebellar wiring depends on this pathway, is not stated.

Nor does deleting a gene throughout development tell you what the pathway does in an adult brain. A mouse that never had endothelial PRMT5 develops differently from the start, and the deficits observed could reflect accumulated developmental consequences rather than an ongoing requirement.

Why it might matter clinically

The authors suggest the findings may inform future research into motor control disorders, which is a cautious claim, and appropriately so.

Still, the logic has implications. If vascular cells help wire motor circuits, then conditions affecting small blood vessels in the brain might impair circuits in ways not explained by reduced blood flow alone. Cerebral small vessel disease is associated with gait disturbance and slowed movement in older people, usually attributed to damaged white matter and poor perfusion. A signalling role for endothelial cells would be a different kind of explanation.

It also matters for drug development. ANGPTL4 has been pursued as a target in metabolic and cardiovascular disease, since it influences lipid processing and vessel behaviour. A compound that alters ANGPTL4 levels systemically might, on this evidence, have effects in the brain that nobody looked for.

Methyltransferase inhibitors are also in development as cancer drugs, and PRMT5 inhibitors in particular have reached clinical trials for tumours with specific genetic deletions. A role for the enzyme in vascular signalling to neurons is the kind of finding that becomes relevant when such compounds are given to patients.

The expert contacts listed for the study are Jun Wang, of the National Center of Protein Sciences Beijing, Hebei University and Anhui Medical University, and Xiao Yang, of Tsinghua University and the National Center of Protein Sciences Beijing. The release does not state the study’s funding sources.

---

Notes

1. Lu, J., et al. “A PRMT5-ANGPTL4 axis in brain endothelial cells regulates cerebellar inhibitory synaptogenesis and motor coordination.” Science Signaling, 2026. DOI: 10.1126/scisignal.aed6605.
2. American Association for the Advancement of Science. “The brain’s maze of blood vessels also shapes fine motor control in the cerebellum, a new study suggests.” EurekAlert!, 20 October 2026. https://www.eurekalert.org/news-releases/1148356.

Purkinje cells in the cerebellum, stained and magnified 63 times. (CREDIT: Laboratory of Developmental Neurobiology at The Rockefeller University)