Fig. 5

Memantine inhibition of GluN1–GluN2B NMDARs. a Molecular docking of memantine in the channel at the trapping site. Note memantine binds just above the channel gate with the quaternary amine facing towards the channel pore and H-bonding (dashed lines) with the two N616 residues in GluN1 (red). H-bonding between GluN2B^N615 and GluN1^N616 (pink arrowhead) is also proposed to stabilise memantine coordination. b Upper panel, memantine inhibition of 10 μM glutamate-activated currents at –30 mV in a HEK293 cell expressing GluN1–GluN2B^WT. Glutamate-activated currents in increasing concentrations of memantine are normalised to control currents elicited by 10 μM glutamate and 10 μM glycine. Lower panel, memantine concentration–inhibition relationships for WT and GluN2B^N615I and GluN2B^V618G mutants. The curve fits were generated using the trapping model (black line: WT, red: N615I, blue: V618G). Dashed line shows the calculated memantine concentration–inhibition curve (WT + Mg calc) in the presence of 1.2 mM Mg²⁺ for WT receptors. Currents are normalised to the control current activated by 10 μM glutamate and 10 μM glycine (=100%). The experimental IC₅₀s are: GluN2B^WT: 7.33 ± 1.86 μM, n = 6; GluN2B^N615I: 1.54 ± 0.27 μM, n = 5, p < 0.001; GluN2B^V618G: 52.23 ± 3.67 μM, n = 11, p < 0.001) one-way ANOVA with Dunnett’s post-hoc test. The predicted IC₅₀s from the trapping model are (μM): GluN2B^WT 5.48 μM; GluN2B^N615I 0.89 μM; GluN2B^V618G 52.6 μM; and GluN2B^WT in the presence of 1.2 mM Mg²⁺_o 18.6 μM. c, d I–V relationships for currents activated by 10 μM glutamate and 10 μM glycine in the absence (0 mM) and presence (1.2 mM) of Mg²⁺, with 30 or 300 μM memantine for WT and GluN2B channel mutants. The curves are generated using the trapping block model (black: WT, red: N615I; blue: V618G)