Allosteric modulation and direct activation of glycine receptors by a tricyclic sulfonamide

Abstract

Ionotropic glycine receptors (GlyRs) are chloride-permeable ligand-gated ion channels expressed in the nervous system. Alterations to glycinergic inhibition and the generation of dysfunctional GlyRs have been linked to chronic pain, a widely prevalent disease. Positive allosteric modulators (PAMs) targeting GlyRs exerted analgesic effects, motivating research on glycinergic PAMs as potential pain therapies. Rationally designed tricyclic sulfonamides are novel glycinergic PAMs with analgesic activity. However, detailed electrophysiological studies on these PAMs are still limited, and the GlyR binding site structural data has not been yet validated by mutational studies. Here, we combined electrophysiology and bioinformatics to systematically study the AM-1488 actions, a prototypical tricyclic sulfonamide, on recombinant GlyRs. We determined that AM-1488 is a potent, non-selective PAM of mammalian GlyR subtypes. In addition, the compound displayed agonistic activity, with partial preference for α1GlyRs. Single channel assays revealed that the compound increased the channel open probability without changing conductance. Mutational analyses on the tricyclic sulfonamide site confirm the molecular determinants contributing to functional activity. Our findings further define the mechanistic framework underlying the GlyR modulation by this PAM class, suggesting that further structure-driven exploration within the tricyclic sulfonamide site may originate novel glycinergic modulators for future development.

Keywords: Allosteric modulation; Chronic pain; Electrophysiology; Glycine receptors; Molecular modelling; Pharmacology.

Fig. 1

Allosteric potentiation of recombinant GlyRs by AM-1488. (A) Representative whole-cell currents showing the potentiation generated by the application of 0.5 µM of AM-1488 to HEK293 cells transiently expressing homo and heteromeric GlyRs. The currents were evoked using 15 µM (α1), 30 µM (α2), 40–60 µM (α3), 15 µM (α1β), 25–50 µM (α2β), or 30–40 µM (α3β) of glycine. (B) Concentration response curves of homo and heteromeric GlyRs in the presence of AM-1488. Data are mean ± SEM. (C) The plot summarizes the maximal potentiation of the glycine-evoked currents elicited by the compound. Differences were not significant. (D) The left panel show the AM-1488 binding to homopentameric α3 GlyRs in the presence of glycine. The right panels summarized enhanced views of the predicted binding of AM-1488 to homologous sites of α1, α2 and α3GlyRs. Glycine is shown in purple. AM-1488 is depicted in red. The models showing the AM-1488-α3GlyR complex, in the presence of glycine, were created using the structural information available (PDB:5TIN).

Fig. 2

Direct activation of recombinant GlyRs by AM-1488. (A) Representative traces of the activation of α1, α2 and α3 GlyRs by 50 µM of AM-1488. The maximal glycine current (evoked using 2 mM glycine) is also shown. (B) Summary graph of the percentage of the maximal glycine evoked current activated by different concentrations of AM-1488 (1, 3, 10, 20, 50 and 100 µM) on α1 (red), α2 (green) or α3GlyRs (blue) expressed in HEK293 cells. AM-1488 was applied to cells in the absence of glycine, without pre-application. Each circle represents the mean ± SEM of 6–8 cells. ANOVA followed by Bonferroni post-hoc test; **, p < 0.01; ***: p < 0.001. (C) The representative traces show the blockade of AM-1488 (100 µM) activated currents by the competitive antagonist strychnine (STN, 2 µM, top) or the pore-blocker picrotoxin (PTX, 30 µM, bottom) in α1 (left), α2 (middle) and α3GlyRs (right). (D) Summary graph of the percentage of decrease of AM-1488 gated currents in the presence of glycinergic inhibitors. Each bar represents the mean ± SEM of 3–5 recorded cells. Statistical analysis, ANOVA followed by Bonferroni post-hoc test. α1GlyRs: AM-1488 vs. + STN, p < 0.01; AM-1488 vs. PTX, p = 0.2317. α2GlyRs: AM-1488 vs. + STN, p < 0.05; AM-1488 vs. PTX, p < 0.05; α2GlyRs: AM-1488 vs. STN, p < 0.05; AM-1488 vs. PTX, p < 0.05. (E) The left panel displays the AM-1488 binding to the orthosteric site of homopentameric α3GlyRs. The right panels are enriched outlooks of the predicted binding of AM-1488 to the orthosteric sites of α1, α2 and α3GlyRs. AM-1488 is shown in red. The models were generated from the structural coordinates available (α1, PDB:7TU9; α2, PDB:7KUY and α3, PDB:5CFB).

Fig. 3

Single channel analyses. (A) Representative current traces showing the effects of 0.5 µM of AM-1488 on glycine-activated currents in a cell-patch expressing homomeric α1GlyRs. The channels were activated with 15 µM of glycine. Calibration bar, 5 pA, 1 s. (B). Summary of the normalized open probability (NPo) and of the main conductance in the absence or presence of AM-1488 (0.5 and 10 µM). ANOVA followed by Bonferroni post-hoc test, *, p < 0.05; **, p < 0.01. (C,D) Single channel current traces obtained from cells expressing wild-type (C) or S346E mutated (D) α3GlyRs in the absence or presence of AM-1488 (10 µM). The channels were evoked with 90 µM of glycine. Calibration bar, 5 pA, 2 s. (E) The histograms show the unitary current amplitude distributions of wild-type receptors (left) and S346E mutant receptors before and during the application of AM-1488 (blue). The control condition (i.e. no allosteric modulator) is shown in grey. (F) The plots summarize the normalized open probability and the unitary conductance of wild-type and S346E α3GlyRs constructs in the absence or presence of AM-1488. NPo: ***p < 0.001, unpaired Student t-Test. In all the cases, the conductances were not significantly modified by the compound.

Fig. 4

Mutational analysis of AM-1488 binding site on α3GlyRs. (A) The images describe the AM-1488-α3GlyR complex in the presence of glycine. The center and right panels show enhanced views of the molecular interactions (4Å cutoff). The highlighted drops represent numbered residues of the AM-1488 binding pocket. The color code describes the amino acid biochemical features (light yellow, glycine; blue, positively charged residues; red, negatively charged residues; green, hydrophobic residues; cyan, polar residues). The green lines symbolize a pi–pi stacking. The models were generated from the structural coordinates of PDB:5TIN. (B) Sample current traces obtained from cells expressing wild-type (WT) or mutated α3GlyRs. (C) Concentration response curves to glycine (1-5000 µM) of wild-type (WT) and alanine-mutated α3GlyRs. (D) The current traces and the bar graph describe the allosteric potentiation of AM-1488 (0.5 µM) on the glycine evoked current through wild-type and mutated α3GlyRs. Currents were evoked using 35–40 µM (WT), 700 µM (F13A), 250 µM (L14A), 500 µM (R27A), 250 µM (L83A) of glycine. F13A, L14A and L83A significantly reduced the allosteric potentiation, whereas R27A did not altered the AM-1488 actions. ANOVA followed by Bonferroni post-hoc test (***, p < 0.001; **: p < 0.01). (E) The current traces and the bar graph summarize the direct activation elicited by AM-1488 (100 µM) of wild-type and mutated α3GlyRs. All the mutants tested significantly reduced the agonistic actions of the compound. ANOVA followed by Bonferroni post-hoc test (*, p < 0.05).