Cancer as a Neural Circuit Disease

# Cancer cells join neural circuits *Evidence guide: Functional recordings and preclinical perturbations. Transmitter effects depend on cell context; electrical integration does not imply cognition.* Paracrine signals diffuse through tissue, but neurons also use synapses. In 2019, the Monje Lab reported that some glioma cells receive functional synapses from neurons, and a German group independently found the same. Patch-clamp recordings showed the synapses work: stimulating nearby axons evoked fast inward currents within milliseconds, blocked by the AMPA-receptor blocker NBQX and by tetrodotoxin, which silences neurons. These glutamatergic synapses use AMPA receptors, partly calcium-permeable ones. Only a subset of cells received this input. Across 643 pediatric glioma cells recorded, about five to ten percent showed synaptic currents. About forty percent showed slower currents lasting over a second, driven mostly by potassium from active neurons and amplified through gap junctions, channels that couple glioma cells into an electrical network. So tumors become electrically integrated, and the influence runs both ways: gliomas make nearby neurons more excitable. Recordings during surgery in three adult patients showed heightened high-frequency activity in tumor-infiltrated cortex. Seizures are common in glioma, and researchers describe a possible feedback loop of activity and growth. In 2023, Kathryn Taylor and colleagues showed these synapses can strengthen. Activity-released BDNF, acting through its receptor TrkB, moved more AMPA receptors to the glioma cell surface, boosting synaptic currents, and increased the number of neuron-to-glioma synapses. This resembles the synaptic plasticity of learning, and removing TrkB, or the TrkB-blocking drug entrectinib, extended survival of mice with glioma xenografts. But tumors do not think, and glioma cells do not become neurons: they remain malignant glial cells exploiting neural machinery. So far, glutamate, the main excitatory transmitter. GABA is the mature brain’s main inhibitory transmitter: opening GABA-A channels usually lets chloride flow in, making a neuron less likely to fire. In 2025, Tara Barron and colleagues found GABAergic synapses from local interneurons onto diffuse midline glioma cells, seen by electron microscopy; about forty percent of recorded DMG cells had GABA-A currents. But in these cells, GABA depolarizes, because of chloride handling. DMG cells express the chloride importer NKCC1 and very little of the exporter KCC2, so chloride builds up inside; when GABA-A channels open, it flows out, depolarizing the membrane. Blocking NKCC1 in brain slices shifted this toward the usual pattern. Stimulating these interneurons increased DMG proliferation in mice, and the sedative lorazepam, which enhances GABA-A currents, did so in a dose-dependent way. Hemispheric high-grade gliomas without the histone mutation showed little GABA current and no lorazepam effect. These are mouse experiments, not medical advice. A third transmitter is acetylcholine. In 2025, Rebecca Drexler and colleagues showed that midbrain cholinergic neurons promote DMG growth with anatomical precision: the pedunculopontine nucleus drove tumor proliferation in the pons, and the laterodorsal tegmental nucleus in the thalamus. Acetylcholine acted directly on DMG cells through muscarinic M1 and M3 receptors, and blocking either abolished the activity-driven proliferation in mice. Whether these cholinergic contacts are functional synapses remains to be tested. The lesson: tumors do not respond to one generic thing called neuronal activity. Circuits, transmitters, receptors, and tumor subtypes interact in different ways. Sources: [venkatesh2019](https://pmc.ncbi.nlm.nih.gov/articles/PMC7038898/), [venkataramani2019](https://doi.org/10.1038/s41586-019-1564-x), [winkler2023](https://pmc.ncbi.nlm.nih.gov/articles/PMC10107403/), [taylor2023](https://doi.org/10.1038/s41586-023-06678-1), [barron2025](https://doi.org/10.1038/s41586-024-08579-3), [drexler2025](https://doi.org/10.1016/j.cell.2025.05.031).