Fig. 3: Mutation of key residues in the ligand binding domain of GluE1αA narrows its ligand-selectivity by switching to glutamate.

a Multiple sequence alignment of deterministic residues that interact with ligand amino acid side chains and mediate ligand specificity in mammalian NMDA, AMPA, and kainate receptors (numbers correspond to amino acids positions in rat GluA2). The residues in green were mutated in the wildtype (wt) GluE1αA receptor to resemble the glutamate-activated GluN2A and GluA2 receptors from human (the single, double and triple mutant variants are referred to as m1, m2 and m3 respectively). b Sample whole-cell currents of wt, m1, m2 and m3 variants of GluE1αA elicited by application of different amino acids at 10 mM. c Plot of average normalized peak inward currents of wt, m1, m2 and m3 variants of GluE1αA in response to different amino acid ligands at 10 mM (n = 5-7). Letters above the bars denote statistically significant differences for ligand type based on post hoc Tukey tests (p < 0.05) after One Way ANOVAs (glycine: F = 290.12, p = 1.11E−16; glutamate: F = 336.17, p = 1.11E−16; D-serine: F = 209.73, p = 2.78E−15; L-serine: F = 534.65, p = 0; alanine: F = 279.39, p = 2.78E−15; valine: F = 20.76, p = 2.35E−6). Specifically, means/bars with the same letters above them are not statistically different, while those with different letters are different based on Tukey tests. d Dose response curves of wt and m1 variants GluE1αA normalized current responses to glycine (n = 8), and of the m2 and m3 variants to glutamate (n = 6). One Way ANOVA confirmed significant differences among the EC₅₀ values (F = 343.15, p = 0), and Tukey post hoc tests revealed that all three were different from each other (p < 0.01). e AlphaFold 3-predicted structure of the GluE1αA tetramer. Three of the four subunits are shown as surfaces colored pale goldenrod, black, and gray, while the fourth subunit is depicted as a ribbon structure with different structural regions colored differently as indicated by the legend. NTD: N-terminal domain; LBD S1: ligand binding domain segment 1; LBD S2: ligand binding domain segment 2; M1 to M4: transmembrane helices 1 to 4; P-loop: pore-loop; CTD: C-terminal domain. f Homology modeling and docking of glycine in the putative ligand binding pocket of the wildtype GluE1αA receptor. g Homology modeling and docking of glycine in triple mutant (m3) variant of GluE1αA. h Homology modeling and docking of glutamate in the wt GluE1αA receptor. i Homology modeling and docking of glutamate in triple mutant (m3) variant of GluE1αA. In f to i, predicted hydrogen bonds are depicted by dashed grey lines.