Inhibition of NMDA receptors and other ion channel types by membrane-associated drugs

Abstract

N-methyl-D-aspartate receptors (NMDARs) are ligand-gated ion channels present at most excitatory synapses in the brain that play essential roles in cognitive functions including learning and memory consolidation. However, NMDAR dysregulation is implicated in many nervous system disorders. Diseases that involve pathological hyperactivity of NMDARs can be treated clinically through inhibition by channel blocking drugs. NMDAR channel block can occur via two known mechanisms. First, in traditional block, charged drug molecules can enter the channel directly from the extracellular solution after NMDAR activation and channel opening. Second, uncharged molecules of channel blocking drug can enter the hydrophobic plasma membrane, and upon NMDAR activation the membrane-associated drug can transit into the channel through a fenestration within the NMDAR. This membrane-associated mechanism of action is called membrane to channel inhibition (MCI) and is not well understood despite the clinical importance of NMDAR channel blocking drugs. Intriguingly, a hydrophobic route of access for drugs is not unique to NMDARs. Our review will address inhibition of NMDARs and other ion channels by membrane-associated drugs and consider how the path of access may affect a drug's therapeutic potential.

Keywords: MCI; NMDAR; channel block; hydrophobic; ketamine; memantine; membrane.

FIGURE 1

Cryo-EM structures of ionotropic glutamate receptors and a VGSC with channel blocking drugs bound. (A) Side view of the structure of an open rattus norvegicus (rat) GluN1/2B receptor in complex with the channel blocking drug memantine (green). Inset shows magnified view of bound memantine. GluN1 subunits are in gray, and GluN2B subunits are in blue. A portion of the pre-M4 and M4 regions on the front GluN2B subunit, and of the M3 region on the front GluN1 subunit, have been removed in the inset for visual clarity of the channel blocking site. GluN2A (M630) is a residue that lines the fenestration used by memantine during MCI to access the channel blocking site (Wilcox et al., 2022); the homologous residue GluN2B(M631) is highlighted in orange. There is evidence for at least one additional NMDAR fenestration that connects the central vestibule to the plasma membrane bilayer (Song et al., 2018), although whether this fenestration is involved in MCI is unknown. One of the residues that lines this pathway in Xenopus laevis NMDARs is GluN2B(V637). The analogous residue in rat NMDARs is GluN2B(V640), which is highlighted here in yellow. Protein Data Bank (PDB): 7SAD (Chou et al., 2022). (B) Side view of the structure of an open homomeric rat GluA2 receptor in complex with stargazin (gray), glutamate (white), cyclothiazide (pink), and the channel blocker NASPM (green). The receptor was composed of four GluA2 subunit-stargazin fusion proteins. The A and C subunits are colored in wheat; the B and D subunits are colored in sand. PDB: 6DM1 (Twomey et al., 2018). (C) Side view of the structure of a homomeric rat GluK2 receptor in complex with positive allosteric modulator BPAM-344 (orange) and the channel blocker spermine (green). The A and C subunits are colored in magenta; the B and D subunits are colored in pink. PDB: 9DXS (Gangwar et al., 2024). (D) Side view of a rat Na_v1.5 receptor in complex with the channel blocker flecainide (green). PDB: 6UZ0 (Jiang et al., 2020).

FIGURE 2

Traditional block and MCI of NMDARs by channel blocking drug. (A) In traditional block, charged molecules of the channel blocking drug (teal) enter the open NMDAR channel from the extracellular solution and bind to the deep site. (B) In membrane to channel inhibition (MCI), uncharged molecules of channel blocking drug enter the plasma membrane and then traverse a fenestration (green arrow) when the NMDAR is in the open conformation and bind to the deep site. The uncharged molecule of channel blocking drug gains a charged hydrogen ion when bound to the deep site. Figure created with BioRender.

FIGURE 3

Protocol for study of MCI, modified from Figure 1A in Wilcox et al. (2022). A whole-cell patch-clamp recording at −65 mV from a tsA201 cell transfected to express GluN1/2A receptors is shown. The first (I) and last (IV) 1 mM glutamate (Glu) applications were performed to activate control responses, without pre-exposure to memantine (Mem). The middle application of 1 mM Glu (III) was used to quantify MCI. The middle 1 mM Glu application (III) was preceded by: a 30-s application of 100 μM Mem in the absence of agonist (II), during which Mem entered the cell’s plasma membrane but not the NMDAR; a 1-s wash with control solution (magenta star), which washed away extracellular Mem to prevent subsequent inhibition via traditional block. The observed transient inhibition during the middle 1 mM Glu application (III) therefore resulted from membrane-associated Mem acting through MCI, and not from traditional block, since no extracellular Mem remains during the middle Glu application (III). MCI was quantified by normalizing the current following the middle 1 mM Glu application (III) to the average of the two control responses resulting from the first (I) and last (IV) 1 mM Glu applications.