Glioma Hijacks Nerve Signals to Speed Its Own Growth

A team from Stanford University published a paper on August 13 in the journal Nature Neuroscience describing, for the first time, the full chain by which glioma cells, the deadliest family of primary brain tumors, hijack normal nerve signaling and use it to multiply. The paper carries the digital object identifier 10.1038/s41593-026-02397-8. The senior author is Michelle Monje of the university’s Department of Neurology, and the co-first authors, who contributed equally, are Sean Gillespie and Jun Seok Kim.

That glioma grows faster when the nerve cells around it are active has been known for about a decade. So had the go-between: neuroligin 3, a molecule that active neurons release into their surroundings and that speeds up the division of tumor cells. What was missing was the first step. Nobody knew what receptor the tumor cell used to catch that molecule, or what happened inside the membrane once it bound.

What Binds to What

The answer is the protein CSPG4, also known as NG2, found on the surface of glioma cells as well as on healthy oligodendrocyte precursor cells, immature support cells in the brain that later give rise to the myelin sheath around nerve fibers.

The researchers describe a four-step sequence. Neuroligin 3 binds to CSPG4. The binding physically stretches the cell membrane and changes its tension. That change in tension opens the mechanosensitive channel PIEZO1, which responds to mechanical pressure, and the membrane depolarizes. The depolarization then activates the enzyme ADAM10, which acts as molecular scissors and cleaves CSPG4 off the cell surface, resetting the system and readying it for the next signal.

The same loop, the authors argue, underlies normal development too. In healthy oligodendrocyte precursor cells it keeps them in an immature, stem-like state. In glioma, that same loop pushes the cell toward accelerated division and invasion of healthy brain tissue.

Measurements and the Model

The binding partner was found through affinity chromatography: a purified fragment of neuroligin 3 was attached to a carrier, membrane proteins from patient-derived tumor cultures were passed through it, and whatever stuck was identified by mass spectrometry.

The PIEZO1 channel was measured electrophysiologically. After exposure to neuroligin 3, current through the channel at a pressure of minus 40 millimeters of mercury rose from 7.9 picoamperes to 17.1 picoamperes. Blocking the ADAM10 enzyme completely stopped the neuroligin 3 induced cleavage of CSPG4, tying the last step of the chain to the ones before it.

The mechanism was not checked on a single model. The researchers tested the diffuse midline glioma cell lines SU-DIPG13 and SU-DIPG6, the adult glioblastoma line MGG8, and a mouse model of optic pathway glioma carrying an NF1 mutation. They also used cerebrospinal fluid from patients: four with diffuse midline glioma and four with leukemia as a negative control group. Both proteins turned up in the fluid of only two of the four tumor patients, and in none of the control group.

What Happened When the Genes Were Switched Off?

Using CRISPR, the researchers knocked out the genes for CSPG4 and PIEZO1 in human glioma cells and injected those cells into mouse brains. Four weeks later, tumor growth was significantly slowed, and the cells divided markedly more slowly than controls.

The reverse experiment showed the flip side. When the gene for neuroligin 3 was removed in mice, the number of healthy oligodendrocyte precursor cells fell in the corpus callosum, the structure that connects the brain’s two hemispheres. The same signal, in other words, also maintains the normal supply of those young support cells.

Limits

This is work done in cell cultures and in mice, not in patients, and it does not show that tumor growth in humans can be stopped by switching off these two genes. The samples are small: cerebrospinal fluid was drawn from four people in each group, and individual experiments were repeated three to four times.

Another limitation is that the target is not exclusively the tumor. CSPG4 and PIEZO1 also sit on healthy precursor cells, and the chain that links them is something the brain needs for myelin repair. The authors themselves leave open the question of whether this pathway can be blocked selectively and temporarily, just enough to slow the tumor while leaving normal nerve tissue repair intact. Answering that will require further experiments, and they will have to be done in a living organism.