Cesium and Glycine Switch On the Brain’s Inhibitory Glycine Receptor Through the Same Molecular Machinery, Study Finds

In a finding that reshapes how neuroscientists think about one of the brain’s core inhibitory switches, researchers at the Technical University of Braunschweig have mapped the amino acid positions that control the activation of the RNA-edited glycine receptor α3L, and the results point to a surprising conclusion: cesium ions, which can act as an unexpected agonist of this chloride channel, appear to hijack the very same structural machinery that glycine uses to open the pore. The study, published in Frontiers in Molecular Neuroscience, systematically mutated five candidate positions in or near the receptor’s glycine binding pocket and measured how each change altered responses to both the natural neurotransmitter and the monovalent cation.

Glycine receptors belong to the pentameric Cys-loop receptor family, a class of ligand-gated ion channels that also includes the receptors for GABA, acetylcholine, and serotonin. In the central nervous system, GlyRs mediate a major share of fast inhibitory synaptic transmission by opening a chloride-conducting pore whenever glycine binds. The α3 subunit exists in short and long splice variants, generated by alternative inclusion of exon 8A of the Glra3 gene within the large intracellular loop between transmembrane domains three and four. Although this loop does not touch the ligand binding site or the ion conduction pathway directly, splice-dependent differences influence how the receptor is regulated and trafficked to the neuronal surface.

On top of alternative splicing, glycine receptor transcripts undergo a rare but potent post-transcriptional modification: cytidine-to-uridine RNA editing. In α2- and α3-GlyRs, this editing swaps a proline for a leucine within the extracellular domain, at positions P185 and P192 respectively. The substitution increases the receptor’s sensitivity to glycine, producing a gain-of-function phenotype. Elevated levels of Glra3-encoded mRNA have been observed in temporal lobe epilepsy, and previous work has suggested that RNA-editing-dependent changes in GlyR function may contribute to pathological shifts in neuronal network homeostasis, with consequences ranging from cognitive deficits to anxiety, depending on which neuronal cell types are affected.

The cesium story adds an unusual twist. While sodium, potassium, and lithium ions do not induce measurable GlyR activation, cesium has been shown to activate at least α2- and α3-containing receptors, and ammonium behaves similarly. Earlier work by the same group demonstrated that the RNA-edited variants respond more strongly to both Cs+ and NH4+ at millimolar concentrations than their unedited counterparts. In 2025, the team characterized a predicted allosteric cesium binding site involving positions D141, E192, and D194 near the editable P185 site in the F-loop region, identifying D194 as essential for activation of α3L receptors at the cell surface and E192 as relevant for desensitization of the current response.

For the new study, the researchers turned to five positions predicted to participate in cesium binding closer to the orthosteric glycine site: the acidic residues D91, D97, E110, and D114, plus the neutral polar tyrosine Y161. Using PCR-based mutagenesis, they replaced each of these residues with either alanine, which carries a neutral hydrophobic side chain, or lysine, which carries a positive charge. The mutated constructs were expressed in HEK293T cells together with EGFP, and receptor function was assessed by whole-cell patch-clamp electrophysiology, with glycine applied at a subsaturating concentration of 100 micromolar followed by successive applications of cesium at 0.5, 5, 50, and 150 millimolar. An HA-epitope tag inserted near the mature N-terminus allowed quantitative surface staining by confocal microscopy, so that losses of function could be distinguished from losses of trafficking.

The electrophysiological data revealed a striking hierarchy of importance. Substitutions at D97, whether alanine or lysine, abolished receptor function in response to both agonists, producing only negligible residual currents and no measurable concentration-response parameters. The same was true for Y161K, where the introduction of a positive charge at this position within the B-loop of the binding site proved catastrophic for glycine-dependent activation. By contrast, mutations at D91, E110, and D114 left detectable currents in place, though with clear quantitative consequences: D91A/K variants reduced peak amplitudes for both glycine and cesium and shifted apparent affinities upward, with EC50 values for glycine rising from 8.6 micromolar in the wild-type receptor to 13.3 and 18.1 micromolar for the alanine and lysine mutants respectively, accompanied by reduced Hill slopes indicating diminished cooperativity of ligand gating.

Perhaps the most intriguing result came from the E110 mutants. Both E110A and E110K primarily impaired glycine-evoked currents while leaving cesium-dependent activation essentially intact, and the normalized data showed cesium-evoked currents that were comparable to or even exceeded the glycine response at high cesium concentrations. The E110A substitution also reduced apparent glycine affinity and cooperativity. Meanwhile, D114A showed reduced responses at low cesium concentrations and reduced apparent affinities for both agonists, while D114K retained near-wild-type behavior. At Y161, the neutral alanine substitution nearly eliminated glycine responses but preserved at least some cesium-dependent responsiveness at concentrations above 5 millimolar, indicating that a positive charge at this position is specifically detrimental to glycine-dependent activation.

The surface staining experiments ruled out the simplest explanation for these losses of function. Although the Y161 mutants showed significantly reduced surface expression, the D97 mutants were clearly detectable at the cell membrane despite being completely non-functional, and Y161A, despite its reduced surface levels, still generated significant cesium-evoked currents. This dissociation between surface presence and functional responsiveness indicates that the mutations act by altering the receptor’s activation machinery rather than by preventing the channel from reaching the membrane. The authors note that changes in receptor surface expression do not appear to be responsible for the observed effects on receptor activation.

Structurally, the mutated positions map onto well-defined elements of the Cys-loop receptor extracellular domain. D91 and D97 sit between β-sheets 3 and 4, just before the A-loop, while E110 and D114 lie between β-sheets 5 and 5′, immediately after the A-loop and before the E-loop. Y161 occupies the B-loop directly within the orthosteric glycine binding site. In the established gating model for this receptor family, agonist binding triggers local electrostatic shifts that break resting-state salt bridges, setting off a conformational wave in which the β-sheets twist around the Cys-loop axis. This motion generates mechanical and electrostatic pulling forces that are transmitted through the linker between transmembrane domains two and three, ultimately opening the central ion pore. The new findings bring loops A and B into the structural framework for GlyR activation, complementing earlier work that implicated the F-loop and β-sheet 9.

The authors draw two main conclusions from the complete dataset. First, because no mutant was identified that responded to glycine while lacking cesium-evoked activity, cesium-mediated activation likely involves the same structural framework required for glycine-induced channel opening rather than representing an independent activation mechanism. They propose that cesium, and by extension other monovalent cations of similar ionic radius such as ammonium, may exert a kind of electrostatic pressure on the receptor protein structure, pushing it toward the open-channel conformation through multiple binding sites distributed across the extracellular ligand-binding domain. Second, the alternative possibility remains that positions or combinations of positions not yet examined could be selectively involved in cesium-dependent activation. The study also acknowledges a limitation: only homomeric α receptors were tested, so the contribution of the GlyR β-subunit, which is known to shape ligand binding and current-voltage relationships in synaptic heteromeric receptors, remains to be explored. Together with the earlier identification of D194 as essential and E192 as a desensitization determinant, this work assembles an increasingly detailed electrostatic map of how a chloride channel central to inhibitory neurotransmission can be switched on by two chemically very different keys.

Subject of Research: Molecular determinants of glycine- and cesium-dependent activation of the RNA-edited glycine receptor α3L

Article Title: Identification of critical sites for glycine- and cesium-dependent activation of the RNA-edited glycine receptor α3L

Article References: Novykova, Y., Fricke, S., & Meier, J. C. (2026). Identification of critical sites for glycine- and cesium-dependent activation of the RNA-edited glycine receptor α3L. Frontiers in Molecular Neuroscience, 19, Article 1918246. https://doi.org/10.3389/fnmol.2026.1918246

Image Credits: AI Generated

DOI: 10.3389/fnmol.2026.1918246

Keywords: glycine receptor, cesium, RNA editing, ligand-gated ion channel, patch-clamp, Cys-loop receptors, GlyR α3, receptor activation, inhibitory neurotransmission, HEK293T cells, mutagenesis, molecular neuroscience