Minere M et al. (2025), Presynaptic hyperexcitability reversed by posit...

XB-ART-60811

Brain 2025 Feb 03;1482:533-548. doi: 10.1093/brain/awae232.

Show Gene links Show Anatomy links

Presynaptic hyperexcitability reversed by positive allosteric modulation of a GABABR epilepsy variant.

Minere M , Mortensen M , Dorovykh V , Warnes G , Nizetic D , Smart TG , Hannan SB .

???displayArticle.abstract???
GABABRs are key membrane proteins that continually adapt the excitability of the nervous system. These G-protein coupled receptors are activated by the brain's premier inhibitory neurotransmitter GABA. They are obligate heterodimers composed of GABA-binding GABABR1 and G-protein-coupling GABABR2 subunits. Recently, three variants (G693W, S695I, I705N) have been identified in the gene (GABBR2) encoding for GABABR2. Individuals that harbour any of these variants exhibit severe developmental epileptic encephalopathy and intellectual disability, but the underlying pathogenesis that is triggered in neurons, remains unresolved. Using a range of confocal imaging, flow cytometry, structural modelling, biochemistry, live cell Ca2+ imaging of presynaptic terminals, whole-cell electrophysiology of HEK-293T cells and neurons, and two-electrode voltage clamping of Xenopus oocytes we have probed the biophysical and molecular trafficking and functional profiles of G693W, S695I and I705N variants. We report that all three point mutations impair neuronal cell surface expression of GABABRs, reducing signalling efficacy. However, a negative effect evident for one variant perturbed neurotransmission by elevating presynaptic Ca2+ signalling. This is reversed by enhancing GABABR signalling via positive allosteric modulation. Our results highlight the importance of studying neuronal receptors expressed in nervous system tissue and provide new mechanistic insights into how GABABR variants can initiate neurodevelopmental disease whilst highlighting the translational suitability and therapeutic potential of allosteric modulation for correcting these deficits.

???displayArticle.pubmedLink??? 39028675
???displayArticle.pmcLink??? PMC11788220
???displayArticle.link??? Brain
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: gabra2 kcnj3 myc pam ttn
GO keywords: GABA receptor activity

???attribute.lit??? ???displayArticles.show???

	Figure 1GABABR2 variants align at the heterodimer interface of the activated receptor. (A) Primary amino acid sequences of GABABR2 showing complete (asterisk) or high (colon) conservation of G693, S695 and I705 between species and with the sequences for murine GABABR1 and R2. (B) Cryo-electron microscopy structures of inactive (PDB:7C7S) and active (7C7Q) GABABR heterodimers, depicting transmembrane domain (TMD) α-helix 6 (M6) with locations for G693W, S695I and I705N (orange; wild-type residues depicted). Note that in the inactive structure, GABABR2 M6 is positioned outside the heterodimer interface (left, inset), but due to clockwise rotations of TMDs during activation (black arrows in the active structure), M6 from R1 and R2 form the interface of the active heterodimer (right, inset). The GABABR positive allosteric modulator (PAM) BHFF binds to the inter-subunit interface close to G693 and S695. (C–E) Structural models for G693W (C; cyan side-chains), S695I (D) and I705N (E), based on the active structure, 7C7Q. TMD interface representations (top), highlighting the side-chain position of the three mutations (cyan, with elements N: blue; O: red). The mutation models aligned to wild-type 7C7Q structure (bottom) with selected residues of interest and BHFF (yellow) shown. The models predict varying degrees of side-chain and α-helical shifting (red arrows) due to the mutations. H-bonds are shown as stippled lines (wild-type: grey; epilepsy mutants: pink). Predicted clashes between mutant R2 M6 94 (with sulfur atom shown in yellow) and BHFF are shown as red stippled rings.
	Figure 2Reduced maximal currents of GABABR variants expressed in GIRK cells. (A) GABA-activated currents (IGABA) of wild-type and variant GABABR2 in GIRK cells expressed with GABABR1. Note the negligible current amplitude for S695I. (B) Concentration response relationships, normalized (norm.) to the maximal (max) GABA currents (=100%) and EC50s of wild-type and variant receptors. Normalized maximal currents (%)—wild-type: 100 (n = 12); G693W: 57 ± 9.3 (n = 11); 695I: 4 ± 2.7 (n = 17); I705N: 54.5 ± 8.5 (n = 12). EC50s (µM)—wild-type: 0.36 ± 0.1 (n = 12); G693W: 0.04 ± 0.02 (n = 8); I705N: 0.14 ± 0.04 (n = 11). (C) Western blot (WB) of immunoprecipitated myc-GABABR1 from human embryonic kidney (HEK)-293 cells transiently expressing GABABR1myc and wild-type or mutant GABABR2flag. Immunoprecipitated samples (top) and input—10% of the cell lysate with corresponding expressing receptors (bottom)—were first probed for FLAG-tag with 1:1000 mouse anti-FLAG-tag antibody; the same membranes were stripped and reprobed with anti-GABABR1 antibody. Numbers on the right of each blot = molecular weights (kDa). (D) Bar chart represents normalized band intensity of GABABR2 (R2) to GABABR1 (R1): R1/R2 1.00 ± 0.00 (n = 6), R1/R2S695I 0.79 ± 0.17 (n = 6), R1/R2 IgG control 0.04 ± 0.02 (n = 4), R1 only 0.01 ± 0.004 (n = 6), mock transfected 0.00 ± 0.00 (n = 6). ns = not significant, one-way ANOVA with Tukey–Kramer test. (E) Example two-electrode voltage clamp recordings for 1 mM (wild-type) and 10 mM (S695I) GABA-activated currents for R1R2 wild-type and R1R2S695I expressing oocytes. The bar chart shows mean ± standard error of the mean maximum GABA-activated currents (n = 6–8). **P < 0.001, one-way ANOVA with Tukey–Kramer test
	Figure 3Impaired cell surface expression for R2 variants in human embryonic kidney (HEK-293) cells. (A) Cytofluorograms of cell surface staining for GABABR1a with wild-type and variant GABABR2 (ordinate) in GFP (abscissa) expressing HEK-293 cells. Cells expressing GFP-only and untransfected cells are also shown. Flag-tagged GABABR2 was labelled with an anti-flag antibody followed by an Alexa Fluor 647 secondary antibody (pictogram). Numbers in the corners of the fluorograms are the percentage of cells in each quadrant. (B) Normalized (to R1R2 wild-type = 100%; left ordinate) and raw (right ordinate) cell surface fluorescence for the cells shown in A (top). Bottom: Percentage of fluorescing cells [normalized (left ordinate) to R1R2 = 100%, and percentage of total cells (right ordinate)] in Quadrant 2 (Q2) for the cells shown in A. (C) Cytofluorograms of total (cell surface + intracellular) R1a with wild-type and mutant GABABR2 staining in HEK-293 cells. GABABR2 was labelled after permeabilization (pictogram). (D) As for the bar graphs shown in B, normalized and raw fluorescence and the number of cells (%) in Q2 for wild-type, variant GABABR2 in permeabilized cells, and just GFP-expressing or untransfected permeabilized cells are shown. *P < 0.05, Kruskal–Wallis one-way ANOVA with Dunn’s test.
	Figure 4Impaired cell surface expression and signalling for mutant GABABRs in hippocampal neurons. (A) Confocal images of cell surface GABABR2 labelling in neurons co-transfected with eGFP, GABABR1 and wild-type (WT) or variant GABABR2 (inset) to drive over-expression. (B) Normalized (to R1R2 wild-type = 100%, left ordinate) and raw (right ordinate) cell surface fluorescence intensities of wild-type or variant R1R2 or cells expressing GFP-only from A. (C) Representative K+ currents in response to 10 and 100 μM baclofen recorded from hippocampal neurons expressing eGFP with or without wild-type or variant R2 (pictogram) or untransfected cells. (D) Mean baclofen-activated K+ current density of neurons expressing wild-type or mutant GABABR2. P < 0.05, P < 0.01, **P < 0.001, one-way ANOVA with Tukey–Kramer test or non-parametric ANOVA with Dunn's multiple comparison test.
	Figure 5Excitatory neurotransmission and GABABR2 variants. (A) Representative miniature excitatory postsynaptic currents (mEPSCs) recorded from hippocampal neurons that are untransfected or expressing eGFP alone or with wild-type or S695I GABABR2 (inset). (B) Average frequency and amplitude of mEPSCs for cells treated as in A. (C) Average mEPSC waveforms and mean charge transfer, rise time and T50 of kinetics for mEPSCs recorded from neurons in A. (D) Representative images of neuronal dendrites expressing eGFP alone or in combination with wild-type or S695I (pictogram) showing spines. (E) Density, head size and proportion of mushroom (mushrm.), stubby and thin spines in hippocampal neurons treated as shown in D. (F) Representative spontaneous action potentials recorded by whole-cell current clamp from untransfected (UTF) and wild-type- or S695I- expressing GABABR2. (G) Average resting membrane potential and spiking rate of hippocampal neurons. (H) Average action potential waveforms, peak, rise time, T50 and charge transfer of spikes. *P < 0.05, one-way ANOVA with Tukey test or non-parametric Kruskal–Wallis ANOVA with Dunn's multiple comparison test. Scale bar = 5 μm.
	Figure 6Elevated presynaptic Ca2+ transients due to GABABR variants. (A) Representative images of presynaptic Ca2+ signals in neurons expressing synaptophysin-GCaMP6f alone (left) or with wild-type (middle) or S695I GABABR2 (right). (B) Example recordings of Ca2+ transients from presynaptic terminals in A. (C) Median presynaptic ΔF/F0, frequency of Ca2+ transients and area under the curve for fluorescence change in a 15 s epoch per presynaptic terminal in wild-type and S695I GABABR2 expressing nerve endings. ***P < 0.001, non-parametric Kruskal–Wallis ANOVA with Dunn's multiple comparisons test. Scale bars = 5 μm.
	Figure 7Pharmacological characterization of GABABR positive allosteric modulators (PAMs). (A) Whole-cell GIRK currents activated by ∼EC20 baclofen showing potentiation by GS39783 and rac-BHFF in GABABR1R2-transfected GIRK cells. (B) PAM potentiation curves for GS39783 and rac-BHFF normalized (%) to the maximal baclofen responses from the same cell (=100%). (C) Bar graph comparing maximal potentiation by GS39783 (30 μM) and rac-BHFF (30 μM). (D) Whole-cell GIRK currents activated by ∼EC20 baclofen showing potentiation by GS39783 and rac-BHFF in hippocampal neurons in culture at 14–21 days in vitro. (E) Potentiation curves for GS39783 and rac-BHFF normalized to maximal baclofen responses (=100%) from the same neuron. (F) Bar graph comparing maximal potentiation of GS39783 and rac-BHFF. (G) Example recordings of ∼EC10 baclofen prior to, during and 5 or 10 min after 3 μM GS39783 application in hippocampal neurons. Bar chart showing increased ∼EC10 10 min after the cessation of GS39783 application. (H) Example recordings of baclofen-activated currents in hippocampal neurons before and 15 min after application of 3 μM GS39783. Note the left-shifted concentration response curve and lower EC50 after 15 min wash (inset). (I) Protocol and example recordings showing ∼EC10 and maximum baclofen currents measured before, during and up to 30 min after the application of 3 μM GS39783 along with a %EC10 decay curve (right) for baclofen after GS39783 application. Dotted line shows the initial ∼EC10 value prior to GS39783 application. n = 6–12. P < 0.01, *P < 0.001, two-tailed unpaired t-test, Mann–Whitney test or repeated measures ANOVA with Tukey–Kramer multiple comparisons test (G).
	Figure 8Reversal of presynaptic increased activity with a GABABR positive allosteric modulator (PAM). (A) Images of presynaptic Ca2+ signals in neurons expressing synaptophysin-GCaMP6f with wild-type or S695I GABABR2 in the presence of a vehicle control or 1 μM rac-BHFF. For fluorescing termini, the intensity (ΔF/F0) of the Ca2+ transients was reduced by rac-BHFF for both termini expressing wild-type or variant R2. (B) Example Ca2+ transient recordings from the presynaptic terminals in A. (C) Median presynaptic ΔF/F0, Ca2+ transient frequency and area for the Ca2+ signals recorded in 15 s epochs per presynaptic terminal in wild-type and S695I GABABR2 expressing nerve endings in vehicle or rac-BHFF. P < 0.05, P < 0.01, **P < 0.001, NS = not significant, non-parametric Kruskal–Wallis ANOVA with Dunn's multiple comparisons test. Scale bars = 2 μm.
	Supplementary Fig 1
	Supplementary Fig 2
	Supplementary Fig 3
	Figure 1. GABABR2 variants align at the heterodimer interface of the activated receptor. (A) Primary amino acid sequences of GABABR2 showing complete (asterisk) or high (colon) conservation of G693, S695 and I705 between species and with the sequences for murine GABABR1 and R2. (B) Cryo-electron microscopy structures of inactive (PDB:7C7S) and active (7C7Q) GABABR heterodimers, depicting transmembrane domain (TMD) α-helix 6 (M6) with locations for G693W, S695I and I705N (orange; wild-type residues depicted). Note that in the inactive structure, GABABR2 M6 is positioned outside the heterodimer interface (left, inset), but due to clockwise rotations of TMDs during activation (black arrows in the active structure), M6 from R1 and R2 form the interface of the active heterodimer (right, inset). The GABABR positive allosteric modulator (PAM) BHFF binds to the inter-subunit interface close to G693 and S695. (C–E) Structural models for G693W (C; cyan side-chains), S695I (D) and I705N (E), based on the active structure, 7C7Q. TMD interface representations (top), highlighting the side-chain position of the three mutations (cyan, with elements N: blue; O: red). The mutation models aligned to wild-type 7C7Q structure (bottom) with selected residues of interest and BHFF (yellow) shown. The models predict varying degrees of side-chain and α-helical shifting (red arrows) due to the mutations. H-bonds are shown as stippled lines (wild-type: grey; epilepsy mutants: pink). Predicted clashes between mutant R2 M6 94 (with sulfur atom shown in yellow) and BHFF are shown as red stippled rings.
	Figure 2. Reduced maximal currents of GABABR variants expressed in GIRK cells. (A) GABA-activated currents (IGABA) of wild-type and variant GABABR2 in GIRK cells expressed with GABABR1. Note the negligible current amplitude for S695I. (B) Concentration response relationships, normalized (norm.) to the maximal (max) GABA currents (=100%) and EC50s of wild-type and variant receptors. Normalized maximal currents (%)—wild-type: 100 (n = 12); G693W: 57 ± 9.3 (n = 11); 695I: 4 ± 2.7 (n = 17); I705N: 54.5 ± 8.5 (n = 12). EC50s (µM)—wild-type: 0.36 ± 0.1 (n = 12); G693W: 0.04 ± 0.02 (n = 8); I705N: 0.14 ± 0.04 (n = 11). (C) Western blot (WB) of immunoprecipitated myc-GABABR1 from human embryonic kidney (HEK)-293 cells transiently expressing GABABR1myc and wild-type or mutant GABABR2flag. Immunoprecipitated samples (top) and input—10% of the cell lysate with corresponding expressing receptors (bottom)—were first probed for FLAG-tag with 1:1000 mouse anti-FLAG-tag antibody; the same membranes were stripped and reprobed with anti-GABABR1 antibody. Numbers on the right of each blot = molecular weights (kDa). (D) Bar chart represents normalized band intensity of GABABR2 (R2) to GABABR1 (R1): R1/R2 1.00 ± 0.00 (n = 6), R1/R2S695I 0.79 ± 0.17 (n = 6), R1/R2 IgG control 0.04 ± 0.02 (n = 4), R1 only 0.01 ± 0.004 (n = 6), mock transfected 0.00 ± 0.00 (n = 6). ns = not significant, one-way ANOVA with Tukey–Kramer test. (E) Example two-electrode voltage clamp recordings for 1 mM (wild-type) and 10 mM (S695I) GABA-activated currents for R1R2 wild-type and R1R2S695I expressing oocytes. The bar chart shows mean ± standard error of the mean maximum GABA-activated currents (n = 6–8). **P < 0.001, one-way ANOVA with Tukey–Kramer test.
	Figure 3. Impaired cell surface expression for R2 variants in human embryonic kidney (HEK-293) cells. (A) Cytofluorograms of cell surface staining for GABABR1a with wild-type and variant GABABR2 (ordinate) in GFP (abscissa) expressing HEK-293 cells. Cells expressing GFP-only and untransfected cells are also shown. Flag-tagged GABABR2 was labelled with an anti-flag antibody followed by an Alexa Fluor 647 secondary antibody (pictogram). Numbers in the corners of the fluorograms are the percentage of cells in each quadrant. (B) Normalized (to R1R2 wild-type = 100%; left ordinate) and raw (right ordinate) cell surface fluorescence for the cells shown in A (top). Bottom: Percentage of fluorescing cells [normalized (left ordinate) to R1R2 = 100%, and percentage of total cells (right ordinate)] in Quadrant 2 (Q2) for the cells shown in A. (C) Cytofluorograms of total (cell surface + intracellular) R1a with wild-type and mutant GABABR2 staining in HEK-293 cells. GABABR2 was labelled after permeabilization (pictogram). (D) As for the bar graphs shown in B, normalized and raw fluorescence and the number of cells (%) in Q2 for wild-type, variant GABABR2 in permeabilized cells, and just GFP-expressing or untransfected permeabilized cells are shown. *P < 0.05, Kruskal–Wallis one-way ANOVA with Dunn’s test.
	Figure 4. Impaired cell surface expression and signalling for mutant GABABRs in hippocampal neurons. (A) Confocal images of cell surface GABABR2 labelling in neurons co-transfected with eGFP, GABABR1 and wild-type (WT) or variant GABABR2 (inset) to drive over-expression. (B) Normalized (to R1R2 wild-type = 100%, left ordinate) and raw (right ordinate) cell surface fluorescence intensities of wild-type or variant R1R2 or cells expressing GFP-only from A. (C) Representative K+ currents in response to 10 and 100 μM baclofen recorded from hippocampal neurons expressing eGFP with or without wild-type or variant R2 (pictogram) or untransfected cells. (D) Mean baclofen-activated K+ current density of neurons expressing wild-type or mutant GABABR2. P < 0.05, P < 0.01, **P < 0.001, one-way ANOVA with Tukey–Kramer test or non-parametric ANOVA with Dunn's multiple comparison test.
	Figure 5. Excitatory neurotransmission and GABABR2 variants. (A) Representative miniature excitatory postsynaptic currents (mEPSCs) recorded from hippocampal neurons that are untransfected or expressing eGFP alone or with wild-type or S695I GABABR2 (inset). (B) Average frequency and amplitude of mEPSCs for cells treated as in A. (C) Average mEPSC waveforms and mean charge transfer, rise time and T50 of kinetics for mEPSCs recorded from neurons in A. (D) Representative images of neuronal dendrites expressing eGFP alone or in combination with wild-type or S695I (pictogram) showing spines. (E) Density, head size and proportion of mushroom (mushrm.), stubby and thin spines in hippocampal neurons treated as shown in D. (F) Representative spontaneous action potentials recorded by whole-cell current clamp from untransfected (UTF) and wild-type- or S695I- expressing GABABR2. (G) Average resting membrane potential and spiking rate of hippocampal neurons. (H) Average action potential waveforms, peak, rise time, T50 and charge transfer of spikes. *P < 0.05, one-way ANOVA with Tukey test or non-parametric Kruskal–Wallis ANOVA with Dunn's multiple comparison test. Scale bar = 5 μm.
	Figure 6. Elevated presynaptic Ca2+ transients due to GABABR variants. (A) Representative images of presynaptic Ca2+ signals in neurons expressing synaptophysin-GCaMP6f alone (left) or with wild-type (middle) or S695I GABABR2 (right). (B) Example recordings of Ca2+ transients from presynaptic terminals in A. (C) Median presynaptic ΔF/F0, frequency of Ca2+ transients and area under the curve for fluorescence change in a 15 s epoch per presynaptic terminal in wild-type and S695I GABABR2 expressing nerve endings. ***P < 0.001, non-parametric Kruskal–Wallis ANOVA with Dunn's multiple comparisons test. Scale bars = 5 μm.
	Figure 7. Pharmacological characterization of GABABR positive allosteric modulators (PAMs). (A) Whole-cell GIRK currents activated by ∼EC20 baclofen showing potentiation by GS39783 and rac-BHFF in GABABR1R2-transfected GIRK cells. (B) PAM potentiation curves for GS39783 and rac-BHFF normalized (%) to the maximal baclofen responses from the same cell (=100%). (C) Bar graph comparing maximal potentiation by GS39783 (30 μM) and rac-BHFF (30 μM). (D) Whole-cell GIRK currents activated by ∼EC20 baclofen showing potentiation by GS39783 and rac-BHFF in hippocampal neurons in culture at 14–21 days in vitro. (E) Potentiation curves for GS39783 and rac-BHFF normalized to maximal baclofen responses (=100%) from the same neuron. (F) Bar graph comparing maximal potentiation of GS39783 and rac-BHFF. (G) Example recordings of ∼EC10 baclofen prior to, during and 5 or 10 min after 3 μM GS39783 application in hippocampal neurons. Bar chart showing increased ∼EC10 10 min after the cessation of GS39783 application. (H) Example recordings of baclofen-activated currents in hippocampal neurons before and 15 min after application of 3 μM GS39783. Note the left-shifted concentration response curve and lower EC50 after 15 min wash (inset). (I) Protocol and example recordings showing ∼EC10 and maximum baclofen currents measured before, during and up to 30 min after the application of 3 μM GS39783 along with a %EC10 decay curve (right) for baclofen after GS39783 application. Dotted line shows the initial ∼EC10 value prior to GS39783 application. n = 6–12. P < 0.01, *P < 0.001, two-tailed unpaired t-test, Mann–Whitney test or repeated measures ANOVA with Tukey–Kramer multiple comparisons test (G).
	Figure 8. Reversal of presynaptic increased activity with a GABABR positive allosteric modulator (PAM). (A) Images of presynaptic Ca2+ signals in neurons expressing synaptophysin-GCaMP6f with wild-type or S695I GABABR2 in the presence of a vehicle control or 1 μM rac-BHFF. For fluorescing termini, the intensity (ΔF/F0) of the Ca2+ transients was reduced by rac-BHFF for both termini expressing wild-type or variant R2. (B) Example Ca2+ transient recordings from the presynaptic terminals in A. (C) Median presynaptic ΔF/F0, Ca2+ transient frequency and area for the Ca2+ signals recorded in 15 s epochs per presynaptic terminal in wild-type and S695I GABABR2 expressing nerve endings in vehicle or rac-BHFF. P < 0.05, P < 0.01, **P < 0.001, NS = not significant, non-parametric Kruskal–Wallis ANOVA with Dunn's multiple comparisons test. Scale bars = 2 μm.