4-Phenylbutyrate promoted wild-type γ-aminobutyric acid type A receptor trafficking, reduced endoplasmic reticulum stress, and mitigated seizures in Gabrg2+/Q390X mice associated with Dravet syndrome

Abstract

Objective: γ-Aminobutyric acid type A (GABA_A ) receptor subunit gene mutations are major causes of various epilepsy syndromes, including severe kinds such as Dravet syndrome. Although the GABA_A receptor is a major target for antiseizure medications, treating GABA_A receptor mutations with receptor channel modulators is ineffective. Here, we determined the effect of a novel treatment with 4-phenylbutyrate (PBA) in Gabrg2^+/Q390X knockin mice associated with Dravet syndrome.

Methods: We used biochemistry in conjunction with differential tagging of the wild-type and the mutant alleles, live brain slice surface biotinylation, microsome isolation, patch-clamp whole-cell recordings, and video-monitoring synchronized electroencephalographic (EEG) recordings in Gabrg2^+/Q390X mice to determine the effect of PBA in vitro with recombinant GABA_A receptors and in vivo with knockin mice.

Results: We found that PBA reduced the mutant γ2(Q390X) subunit protein aggregates, enhanced the wild-type GABA_A receptor subunits' trafficking, and increased the membrane expression of the wild-type receptors. PBA increased the current amplitude of GABA-evoked current in human embryonic kidney 293T cells and the neurons bearing the γ2(Q390X) subunit protein. PBA also proved to reduce endoplasmic reticulum (ER) stress caused by the mutant γ2(Q390X) subunit protein, as well as mitigating seizures and EEG abnormalities in the Gabrg2^+/Q390X mice.

Significance: This research has unveiled a promising and innovative approach for treating epilepsy linked to GABA_A receptor mutations through an unconventional antiseizure mechanism. Rather than directly modulating the affected mutant channel, PBA facilitates the folding and transportation of wild-type receptor subunits to the cell membrane and synapse. Combining these findings with our previous study, which demonstrated PBA's efficacy in restoring GABA transporter 1 (encoded by SLC6A1) function, we propose that PBA holds significant potential for a wide range of genetic epilepsies. Its ability to target shared molecular pathways involving mutant protein ER retention and impaired protein membrane trafficking suggests broad application in treating such conditions.

Keywords: 4-phenylbutyrate; ER stress; GABAA receptors; Gabrg2+/Q390X knockin mice; epilepsy.

Figure 1.. 4-phenylbutyrate alone or in combination with diazepam mitigated seizures in the Gabrg2^+/Q390X mice

(A). Schematic depiction of experimental paradigm for EEG recordings and PBA and diazepam treatment. Representative EEG recordings show that the heterozygous Gabrg2^+Q390X (het) KI mice had generalized tonic clonic seizures and frequent absence like spike wave discharges (SWDs) as well as myoclonic jerks during baseline recordings. Comparison of EEG traces recorded after the vehicle (normal saline 100 μl) treated or after treatment with PBA alone (100 mg/kg, ip, single dose, daily) or with a combination of PBA and diazepam (0.3mg/kg) for 7 days. B. Graph showing the total number of generalized tonic clonic seizures (GTCS), myoclonic jerks (MJ) or tonic seizure events calculated by Seizure Pro per 24 hrs. C. Graph showing the total number of 5-7 Hz SWDs calculated by Seizure per 24 hrs. D. Average duration of 5-7 Hz SWDs in the Gabrg2^+Q390X mice treated with vehicle or drugs. In B, ** P< 0.01; *** P< 0.001 PBA vs vehicle. In C, *** P< 0.001 PBA vs vehicle; δδ p < 0.01 PBA+DZP vs PBA treated, Values were expressed as mean ± S.E.M, N=7 mice for vehicle and 6 mice for PBA or PBA plus DZP)

Figure 2.. 4-phenylbutyrate increased the current amplitude of GABA_A receptors in HEK293T cells or hippocampal neurons bearing the GABRG2(Q390X) mutation.

(A) GABA_A receptor currents were obtained from HEK 293-T cells cotransfected with wild-type α1 and β2 subunits and γ2S (1:1:1 cDNA ratio; wt) or heterozygous γ2S/γ2S(Q390X) (1:1:0.5:0.5 cDNA ratio; het) subunits with application of 1 mM GABA applied for 6 sec. The cells were treated with vehicle or PBA for 24 hrs prior to patch clamp recordings (gray trace marked with an arrow for vehicle treated and black trace marked with a star for PBA treated). (B, C) The amplitudes (B) or the extent of desensitization (C) of GABA_A receptor currents in the vehicle or PBA treated cells were plotted. (D-F). Representative traces of GABAergic mIPSCs from 15-17 days old hippocampal neurons cultured from the wildtype (wt) and the Gabrg2^+/Q390X het pups at postnatal day 0 (D). The amplitude (E) or frequency (F) of GABAergic mIPSCs in the hippocampal neurons from wt or het Gabrg2^+/Q390X mice were presented as mean ± S.E.M (n = 6-7 of wildtype or the het neurons cultured from the four different pairs of mouse pups. In B, C, E, ***p < 0.001 vs wt; §§ P<0.01 vs wt vehicle; δδδ p < 0.001 vs het vehicle. Values were mean ± SEM.

Figure 3.. 4-phenylbutyrate increased the cell surface expression of wildtype partnering subunits of GABA_A receptors in the Gabrg2^+/Q390X mice.

(A) The surface protein isolated with cell surface biotinylation from different brain regions (cortex (cor), cerebellum (cb), hippocampus (hip) and thalamus (thal)) from the wildtype (wt) (A) and heterozygous (het) (B) mice at 2-4 months old, untreated or treated with vehicle or PBA (100mg/kg) for 7 days were subjected to SDS-PAGE and immunoblotted with anti-γ2, α1 or β2 antibody. C, D. Integrated density values (IDVs) for GABA_A receptor α1, β2 and γ2 subunits from wild-type and het KI were normalized to the Na⁺/K⁺ ATPase as loading control (LC) in each specific brain region and plotted. N=4 from 4 pairs of mice. In the wt or the het mice, the vehicle treated in each brain region was taken as 1. N=4 from 4 pairs of mice. Values were expressed as mean ± S.E.M. In D, **p < 0.01; ***p < 0.001 vs vehicle treated).

Figure 4.. 4-phenylbutyrate increased the total wildtype partnering subunits of GABA_A receptors and reduced the mutant protein aggregates.

(A) Schematic presentation of human GABRG2-encoded γ2 subunit and the γ2(Q390X) mutant subunit protein topology. The red dot represents the relative locations of the γ2(Q390X) subunit mutation. (B, C) HEK293T cells were cotransfected with α1, β2 and γ2, γ2(Q390X) for 2 days. DMSO (0.1%) or 4-phenylbutyrate (2 mM) in 100 μl of DMEM was applied in cells for 24hrs before harvest. The total cell lysates were analyzed by SDS-PAGE. The membrane was immunostained with polyclonal rabbit anti-γ2 antibody (B) or a mouse monoclonal anti-α1 or anti- β2 antibody (C). The ATPase was used as loading control (LC) for all subunits. The wildtype (wt) designates the cells expressing the wildtype α1, β2 and γ2 subunits at a cDNA ratio of 1;1;1. The “heterozygous” (het) designates the cells expressing the wildtype α1, β2, the γ2 and γ2(Q390X) subunits at a cDNA ratio of 1:1:0.5:0.5. The “homzygous” (hom) designates the cells expressing the wildtype α1, β2 and γ2(Q390X) subunits at a cDNA ratio of 1:1:1 while the “hemizygous” (hemi) designates the cells expressing the wildtype α1, β2, the γ2(Q390X) subunits and the empty vector pcDNA at a cDNA ratio of 1:1:0.5:0.5. The band pointed by the green arrow is a nonspecific band while the one pointed by the purple arrow is the wildtype. Red boxed regions represent the mutant γ2(Q390X) monomer or dimers. (D) The relative total protein integrated density values (IDVs) of the wildtype γ2 subunit in each condition were normalized to those obtained with expression of wildtype γ2 subunit with α1 and β2 subunit cDNAs. E. The IDVs of the high molecular mass protein complexes (HMMPC) in the mutant heterozygous and homozygous conditions were analyzed. (F, G). The relative total protein integrated density values (IDVs) of the wildtype α1 (F) or β2 (G) subunits in the mutant conditions were normalized to those obtained with coexpression of wildtype γ2 subunit with α1 and β2 subunit cDNAs. In D, *** P< 0.001 vs wt, δ P< 0.05 vs “het” vehicle; §§P< 0.01 vs hemi vehicle. In E, *** P< 0.001 vs “het” vehicle; δ P< 0.05 vs “het” vehicle; § P< 0.05 vs “hom” vehicle. In F and G, *P< 0.05; *** P< 0.001 vs wt vehicle, δ P< 0.05 vs “het” vehicle; §P< 0.05 vs “hom” vehicle; & P<0.05 vs “hemi” vehicle. In D to G, n=6 blots from 6 batches of transfections in 6 batches of cells. Mean ± SEM).

Figure 5.. 4-phenylbutyrate reduced the ER stress caused by the mutant γ2(Q390X) subunit protein.

(A). Total lysates from Mouse L929 cells transfected with the wildtype or the mutant γ2(Q390X) subunits in combination with the wildtype α1 and β2 subunits for 48 hrs. Sister cultures were treated with or without tunicamycin (10μg/ml) for 16 hrs before harvest. (B). Total lysates from the cortex of wildtype or heterozygous Gabrg2^+/Q390X mice treated with vehicle or 4-phenylbutyrate (100mg/kg) for 7 days were analyzed by SDS-PAGE. Equal amounts of the protein were loaded and the membranes were immunoblotted with GADD153. B-actin was used as a loading control (LC). (C, D). The total amounts of endogenous GADD153 were normalized to untreated controls (U) (C) for Mouse L929 cells or to vehicle treated wildtype (wt) mice (D). In C, ***p < 0.001 vs U, δδδ p<0.001 vs wt., N=5 batches of transfection in different batches of cells. In D, ***p < 0.001 vs wt; δδδ p<0.001 vs het vehicle. N=6 mice for each group. Mean ± SEM.

Figure 6.. 4-phenylbutyrate increased the folding of the wildtype subunits in the mutant α1β2γ2/γ2(Q390X) receptors.

(A). Schematic depiction of limited proteolysis including microsome isolation, trypsin digestion and SDS-PAGE. (B). Lysates of microsome isolated from HEK293T cells expressing the α1β2γ2/γ2(Q390X) receptors treated with vehicle or 4 phenylbutyrate (PBA 2mM) were undigested or digested with trypsin (100μg/ml). The lysates were analyzed with SDS-PAGE and the membrane was immunoblotted with antibodies that recognizes the wildtype γ2^HA, the total γ2, β2 or α1 subunits. The γ2^HA subunits were run with one main band (panel a); the total γ2 (γ2^HA and γ2(Q390X) subunits were run with four bands (panel b, band 1,2,3,4). The α1 subunits (panel c) were run with one main band while the β2 subunits were run with two bands (panel d). In B panel a,b,d, yellow arrow stands for undigested while the cyan arrow stands for digested protein. All the bands in the purple boxed region were included in quantification. (C-F). Protein IDVs of the wildtype γ2^HA subunits (C), the total γ2 subunit (wildtype γ2^HA and the γ2(Q390X) subunits) (D), the α1 subunit (E) or the β2 subunit (F) after PBA treatment was normalized to vehicle (DMSO) treated. Values were expressed as mean ± S.E.M. In C to F. ***p < 0.001, **p < 0.01; PBA vs vehicle. N=5 batches of transfection in 5 different batches of cells.