Combining Valosin-containing Protein (VCP) Inhibition and Suberanilohydroxamic Acid (SAHA) Treatment Additively Enhances the Folding, Trafficking, and Function of Epilepsy-associated γ-Aminobutyric Acid, Type A (GABAA) Receptors*

Background: The α1 subunit harboring the A322D mutation is subject to excessive ERAD. Results: VCP inhibition using Eeyarestatin I reduces the ERAD of α1(A322D) subunits, and coapplication of SAHA additively enhances their proteostasis. Conclusion: Combining ERAD inhibition and folding enhancement yields significant functional rescue. Significance: This combination represents a new, promising strategy to treat epilepsy resulting from GABAA receptor misfolding. GABAA receptors are the primary inhibitory ion channels in the mammalian central nervous system. The A322D mutation in the α1 subunit results in its excessive endoplasmic reticulum-associated degradation at the expense of plasma membrane trafficking, leading to autosomal dominant juvenile myoclonic epilepsy. Presumably, valosin-containing protein (VCP)/p97 extracts misfolded subunits from the endoplasmic reticulum membrane to the cytosolic proteasome for degradation. Here we showed that inhibiting VCP using Eeyarestatin I reduces the endoplasmic reticulum-associated degradation of the α1(A322D) subunit without an apparent effect on its dynamin-1 dependent endocytosis and that this treatment enhances its trafficking. Furthermore, coapplication of Eeyarestatin I and suberanilohydroxamic acid, a known small molecule that promotes chaperone-assisted folding, yields an additive restoration of surface expression of α1(A322D) subunits in HEK293 cells and neuronal SH-SY5Y cells. Consequently, this combination significantly increases GABA-induced chloride currents in whole-cell patch clamping experiments than either chemical compound alone in HEK293 cells. Our findings suggest that VCP inhibition without stress induction, together with folding enhancement, represents a new strategy to restore proteostasis of misfolding-prone GABAA receptors and, therefore, a potential remedy for idiopathic epilepsy.

GABA A receptors are the primary inhibitory ion channels in the mammalian central nervous system. The A322D mutation in the ␣1 subunit results in its excessive endoplasmic reticulum-associated degradation at the expense of plasma membrane trafficking, leading to autosomal dominant juvenile myoclonic epilepsy. Presumably, valosin-containing protein (VCP)/p97 extracts misfolded subunits from the endoplasmic reticulum membrane to the cytosolic proteasome for degradation. Here we showed that inhibiting VCP using Eeyarestatin I reduces the endoplasmic reticulum-associated degradation of the ␣1(A322D) subunit without an apparent effect on its dynamin-1 dependent endocytosis and that this treatment enhances its trafficking. Furthermore, coapplication of Eeyarestatin I and suberanilohydroxamic acid, a known small molecule that promotes chaperone-assisted folding, yields an additive restoration of surface expression of ␣1(A322D) subunits in HEK293 cells and neuronal SH-SY5Y cells. Consequently, this combination significantly increases GABA-induced chloride currents in whole-cell patch clamping experiments than either chemical compound alone in HEK293 cells. Our findings suggest that VCP inhibition without stress induction, together with folding enhancement, represents a new strategy to restore proteostasis of misfolding-prone GABA A receptors and, therefore, a potential remedy for idiopathic epilepsy.
Proteostasis, an optimal state of the cellular proteome (1,2), is constantly challenged by intrinsic stress such as inherited misfolding-prone proteins (3), the environment (4), and aging (5). Proteostasis deficiency in ion channels leads to a variety of ion channel diseases called channelopathies, which are often caused by excessive endoplasmic reticulum-associated degradation (ERAD) 2 and inefficient membrane trafficking of corresponding ion channel proteins harboring misfolding-prone mutations (3).
Valosin-containing protein (VCP), also called p97 in mammals, Cdc48p in yeast, CDC-48 in Caenorhabditis elegans, or TER94 in the fly, is a highly conserved, abundant ATPase of the ATPases associated with diverse cellular activities family (6). VCP is a homohexamer, forming a ring around a central pore (7). VCP has diverse physiological functions in ubiquitin-mediated degradation, autophagy, chromatin remodeling, cell cycle control, and more (8,9). Probably the most prominent and studied role of VCP is in the ERAD regulation of a variety of substrates (10). The ERAD machinery is responsible for the degradation of terminally misfolded proteins in the endoplasmic reticulum (ER) (11)(12)(13)(14). ERAD contains distinct but coupled steps, including substrate recognition, retrotranslocation, polyubiquitination, and proteasome targeting and degradation (11,15). VCP plays an essential role in driving the retrotranslocation process, extracts misfolded proteins from the ER to the cytosol using energy, and targets them to the proteasome for degradation (10,15).
VCP has been reported to differentially influence the maturation of various established ERAD substrates. RNA interference of VCP led to the accumulation of both mature and immature forms of the ⌬F508 cystic fibrosis transmembrane conductance regulator, resulting in partial rescue of the mature ⌬F508 cystic fibrosis transmembrane conductance regulator in IB3-1 cells (16). VCP inhibition rescued trafficking of lysosomal glucocerebrosidase harboring the L444P mutation in Gaucher disease patient-derived fibroblasts (17). In contrast, RNA inter-ference of VCP did not increase the steady-state level of ␣1-antitrypsin Hong Kong variant or T-cell receptor T3 ⌬ chain (gene name CD3D) in HeLa cells (18). Therefore, the exact role of VCP in regulating protein maturation might depend on a specific ERAD substrate. Of particular interest is that the role of VCP in regulating GABA A receptor proteostasis has not yet been established. GABA A receptors are the primary inhibitory neurotransmitter-gated ion channels in the mammalian central nervous system (19). They belong to and share common structural characteristics within the Cys loop receptor superfamily (20 -22). The most common GABA A receptors in the human brain are a heteropentamer, consisting of two ␣1, two ␤2 , and one ␥2 subunits. Each subunit has four transmembrane (TM) helices (TM1-TM4, with the TM2 domain lining the interior of the pore), a large extracellular (or the ER luminal) N terminus, and a short extracellular (or the ER luminal) C terminus (Fig. 1A). The large cytosolic loop between TM3 and TM4 serves as a prominent candidate to interact with cytosolic molecular chaperones and degradation factors during the cellular protein folding and degradation steps.
GABA A receptors have a strong genetic association with idiopathic epilepsy (23)(24)(25)(26)(27). The missense A322D mutation in the TM3 domain of the ␣1 subunit of GABA A receptors leads to autosomal dominant juvenile myoclonic epilepsy, a common form of idiopathic generalized epilepsy representing 5-10% of all epilepsy cases (28). The A322D mutation results in the misfolding and, therefore, rapid degradation of the ␣1(A322D) subunit, mainly by ERAD (29). The consequence is that few ␣1(A322D) subunits are transported to the plasma membrane, reducing the level of functional pentameric GABA A receptors in the cell membrane. The A322D mutation leads to substantially reduced GABA-induced current in electrophysiological experiments. The few mutant receptors that reach the plasma membrane generate GABA-induced currents with different kinetics properties compared with WT receptors (30,31). The cellular ERAD machinery regulating the rapid degradation of ␣1(A322D) subunits, however, is largely unexplored in the literature.
Presumably, the misfolded ␣1(A322D) subunit is recognized by the ER quality control machinery, polyubiquitinated, extracted from the ER membrane to the cytosol, and targeted to the proteasome for degradation. Here, we studied VCP as our first step to characterize the ERAD network for GABA A receptors because VCP plays an essential role in the substrate extraction step. Moreover, VCP appears in the interactome list for the 1 subunit of GABA A receptors (32), and human 1 and ␣1 subunits share high sequence homology (32.5% identity and 63.8% similarity).
We hypothesized that VCP extracts misfolded ␣1(A322D) subunits for their fast degradation, which outcompetes their folding and trafficking. Therefore, inhibiting VCP allows ␣1(A322D) subunits to have more time to fold in the ER for subsequent trafficking to the plasma membrane. We have demonstrated previously that SAHA, a potent histone deacetylase inhibitor, increases functional ␣1(A322D) subunit cell surface levels, partially by promoting BiP and calnexin-assisted folding (33). In this study, we investigated how VCP inhibition influences the degradation and trafficking of ␣1(A322D) subunits. Furthermore, we determined whether ERAD inhibition and folding enhancement by using SAHA have an additive effect to restore the function of epilepsy-associated GABA A receptors.

EXPERIMENTAL PROCEDURES
Reagents-Eeyarestatin I (EerI) and Dynole 34-2 were obtained from Tocris Bioscience. SAHA and lactacystin were from Cayman Chemical, and thapsigargin was from Enzo Life Science. The pCMV6 plasmids containing the human GABA A receptor ␣1, ␤2 (isoform 2), and ␥2 (isoform 2) subunits and the pCMV6 entry vector plasmid (pCMV6-EV) were obtained from Origene. The human GABA A receptor ␣1 subunit missense mutation A322D was constructed using the QuikChange II site-directed mutagenesis kit (Agilent Genomics), and the cDNA sequences were confirmed by DNA sequencing.
The mouse monoclonal anti-␣1 (clone BD24) and anti-␤2/3 (clone 62-3G1) antibodies were obtained from Millipore, and the rabbit polyclonal anti-␥2 antibody was from R&D systems. The mouse monoclonal anti-␤-actin antibody came from Sigma. The rabbit polyclonal anti-calnexin and anti-Hsp70, mouse monoclonal anti-Hsp90, and rat polyclonal anti-Grp94 antibodies were obtained from Enzo Life Sciences. The rabbit monoclonal anti-VCP and anti-BiP antibodies were obtained from Epitomics. The rabbit polyclonal anti-ubiquitin antibody was obtained from Cell Signaling Technology.
Western Blot Analysis-Cells were harvested with Trypsin 0.05% and then lysed with lysis buffer (50 mM Tris (pH 7.5), 150 mM NaCl, and 1% Triton X-100) supplemented with Roche complete protease inhibitor mixture. Lysates were cleared by centrifugation (15,000 ϫ g, 10 min, 4°C). Protein concentration was determined by MicroBCA assay (Pierce). Endoglycosidase H (endo H) and peptide-N-glycosidase F (New England Biolabs) digestion was performed according to the published procedure (29). Aliquots of cell lysates were separated in an 8% SDS-PAGE gel, and Western blot analysis was performed using the appropriate antibodies. Band intensity was quantified using ImageJ software from the National Institutes of Health.
Immunoprecipitation-Cell lysates (500 g) were precleared with 30 l of protein A/G Plus-agarose beads (Santa Cruz Biotechnology) and 1.0 g of normal mouse IgG for 1 h at 4°C to remove nonspecific binding proteins. The precleared cell lysates were incubated with 2.0 g of mouse anti-␣1 antibody (clone BD24, Millipore) or normal mouse IgG (negative control for nonspecific binding) for 1 h at 4°C and then with 30 l of protein A/G Plus-agarose beads overnight at 4°C. The beads were collected by centrifugation at 8000 ϫ g for 30 s and washed three times with lysis buffer. The ␣1 subunit complex was eluted by incubation with 30 l of SDS loading buffer in the presence of DTT. The immunopurified eluents were separated in an 8% SDS-PAGE gel, and Western blot analysis was performed.
siRNA Transfection-HEK293 cells were seeded at ϳ2.5 ϫ 10 5 cells/well in 6-well plates for siRNA treatment. Cells were allowed to reach ϳ70% confluency before transfection. The following siRNA duplexes were obtained from Dharmacon: VCP (catalog nos. J-008727-09 and J-008727-12) and non-targeting siRNA (D-001810-01-20) as a negative control. Cells were transfected with corresponding 50 nM siRNA using HiPerfect transfection reagent (Qiagen) according to the transfection protocol of the manufacturer prior to protein analysis. Fortyeight hours post-transfection, cells were harvested, lysed, and subjected to SDS-PAGE and Western blot analysis.
Cycloheximide Chase Assay-HEK293 cells stably overexpressing ␣1(A322D)␤2␥2 GABA A receptors were seeded at 2.5 ϫ 10 5 cells/well in 6-well plates and incubated at 37°C overnight. Cells were then treated with the chemical compounds or dimethyl sulfoxide vehicle control prior to cycloheximide chase. To stop protein translation, cells were treated with 100 g/ml cycloheximide (Ameresco) for the indicated time before being lysed for protein analysis.
Biotinylation of Cell Surface Proteins-HEK293 and SH-SY5Y cells stably overexpressing ␣1(A322D)␤2␥2 receptors were plated in 10-cm dishes for surface biotinylation experiments according to published procedures (31). Intact cells were washed twice with ice-cold PBS and incubated with the membrane-impermeable biotinylation reagent Sulfo-NHS SS-Biotin (0.5 mg/ml, Pierce) in PBS containing 0.1 mM CaCl 2 and 1 mM MgCl 2 (PBS ϩ CM) for 30 min at 4°C to label surface membrane proteins. To quench the reaction, cells were incubated with 10 mM glycine in ice-cold PBS ϩ CM twice for 5 min at 4°C. Sulfhydryl groups were blocked by incubating the cells with 5 nM N-ethylmaleimide in PBS for 15 min at room temperature. Cells were solubilized for 1 h at 4°C in lysis buffer (1% Triton X-100, 50 mM Tris-HCl, 150 mM NaCl, and 5 mM EDTA (pH 7.5)) supplemented with Roche complete protease inhibitor mixture and 5 mM N-ethylmaleimide. The lysates were cleared by centrifugation (16,000 ϫ g, 10 min at 4°C) to pellet cellular debris. The supernatant contained the biotinylated surface proteins. The concentration of the supernatant was measured using a microBCA assay (Pierce). Biotinylated surface proteins were affinity-purified from the above supernatant by incubating for 1 h at 4°C with 100 l of immobilized neutravidin-conjugated agarose bead slurry (Pierce) and being subjected to centrifugation (16,000 ϫ g, 10 min, at 4°C). The beads . Endo H-resistant ␣1 subunit bands represent properly folded, post-ER ␣1 subunit glycoforms that traffic at least to the Golgi compartment, whereas endo H-sensitive ␣1 subunit bands represent immature ␣1 subunit glycoforms that are retained in the ER. The peptide-N-glycosidase F enzyme cleaves between the innermost N-acetyl-D-glucosamine and asparagine residues from N-linked glycoproteins, serving as a control for unglycosylated ␣1 subunits (lane 3). After endo H digestion, subunits with a molecular weight equal to unglycosylated ␣1 subunits were considered endo H-sensitive (bottom arrow), whereas those with a higher molecular weight were considered endo H-resistant (top arrow) (lanes 2 and 5). The curly bracket indicates the top endo H-resistant bands in lanes 2, 5, 7, and 9 (no endo H-resistant bands were visible in lanes 7 and 9). ␤-actin served as a protein loading control. Ctl, no enzyme digestion control; Eh, endo H; Pf, peptide-N-glycosidase F; IB, immunoblot. C and D, the cellular interaction (direct or indirect) between the ␣1 subunit and VCP was verified by immunoprecipitation and Western blot analyses. HEK293 cells expressing WT ␣1␤2␥2 or ␣1(A322D)␤2␥2 GABA A receptors were lysed and immunoprecipitated with mouse anti-␣1 subunit antibody or normal mouse IgG for negative, nonspecific binding control before being subjected to SDS-PAGE and Western blot analysis (C) (n ϭ 3). The ratio of the VCP/␣1 subunit post-immunoprecipitation, as a measure of the interaction between VCP and ␣1 subunit, was quantified, normalized to that of the WT, and is shown in D. EV, empty vector; IP, immunoprecipitation. Each data point in D is reported as mean Ϯ S.E. **, p Ͻ 0.01.
were washed six times with buffer (0.5% Triton X-100, 50 mM Tris-HCl, 150 mM NaCl, and 5 mM EDTA (pH 7.5)). Surface proteins were eluted from beads by boiling for 5 min with 200 l of Laemmli sample buffer/urea buffer (2ϫ Laemmli sample buffer with 100 mM DTT and 6 M urea (pH 6.8)) for SDS-PAGE and Western blot analysis.

A Combination Strategy to Enhance GABA A R Proteostasis
Whole-cell Patch Clamp Electrophysiology Recording-Whole-cell currents were recorded 48 h post-transfection using HEK293 cells. The glass electrodes were pulled from thinwalled borosilicate capillary glass (Kimble-Chase) and fire-polished on a DMZ universal puller (Zeitz Instruments), having a tip resistance of 3-5 M⍀. The internal solution contained 153 mM KCl, 1 mM MgCl 2 , 5 mM EGTA, 10 mM HEPES, and 2 mM MgATP (pH 7.3). The external solution contained 142 mM NaCl, 8 mM KCl, 6 mM MgCl 2 , 1 mM CaCl 2 , 10 mM glucose, 10 mM HEPES, and 120 nM fenvalerate (pH 7.4). Coverslips containing HEK293 cells were placed in an RC-25 recording chamber (Warner Instruments) on the stage of an Olympus IX-71 inverted fluorescence microscope and perfused with external solution. Fast GABA application was accomplished with a pressure-controlled perfusion system (Warner Instruments) positioned within 50 m of the cell utilizing a Quartz MicroManifold with 100-m inner diameter inlet tubes (ALA Scientific). The whole-cell GABA-induced currents were recorded at a holding potential of Ϫ20 mV in voltage clamp mode using an Axopatch 200B amplifier (Molecular Devices). The signals were filtered at 2 kHz and detected at 10 kHz using pClamp10 acquisition software.
Quantitative RT-PCR-The relative expression levels of target genes were analyzed using quantitative RT-PCR according to a published procedure (34). Cells were incubated with the chemical compounds at 37°C for the indicated amount of time before total RNA was extracted from the cells using an RNeasy mini kit (Qiagen, catalog no. 74104). cDNA was synthesized from 500 ng of total RNA using a QuantiTect reverse transcription kit (Qiagen, catalog no. 205311). Quantitative PCR reactions (45 cycles of 15 s at 94°C, 30 s at 57°C, and 30 s at 72°C) were performed using cDNA, a QuantiTect SYBR Green PCR kit (Qiagen, catalog no. 204143), and corresponding primers in the StepOnePlus system (Applied Biosystems) and analyzed using StepOne v2.2 software (Applied Biosystems). The forward and reverse primers for CHOP and GAPDH (housekeeping gene control) were as follows: CHOP forward (5Ј Ͼ 3Ј) GGAAACAGAGTGGTCATTCCC and reverse (5Ј Ͼ 3Ј) CTGCTTGAGCCGTTCATTCTC and GAPDH forward (5Ј Ͼ 3Ј) GTCGGAGTCAACGGATT and reverse (5Ј Ͼ 3Ј) AAGCTTCCCGTTCTCAG. The threshold cycle (C T ) was extracted from the PCR amplification plot, and the ⌬C T value was defined as ⌬C T ϭ C T (target gene) Ϫ C T (housekeeping gene). The relative mRNA expression level of target genes of the chemical compound-treated cells was normalized to that of untreated cells as follows: relative mRNA expression level ϭ 2 exp [Ϫ (⌬C T (treated cells) Ϫ ⌬C T (untreated cells))]. Each data point was evaluated in triplicate and measured three times.
Statistical Analysis-All data are presented as mean Ϯ S.E., and any statistical significance was calculated using two-tailed Student's t test.

VCP Interacts Stronger with Misfolding-prone ␣1(A322D)
Subunits Than with WT ␣1 Subunits-It has been reported that the A322D mutation in the ␣1 subunit resulted in rapid degradation of the ␣1(A322D) subunit mainly by ERAD (29). We confirmed this result by adding lactacystin (5 M, 24 h), a potent proteasome inhibitor, to HEK293 cells overexpressing WT or ␣1(A322D)␤2␥2 GABA A receptors. Lactacystin treatment clearly increased the total protein level for the WT ␣1 subunit (Fig. 1B, cf. lane 4 with lane 1) and the ␣1(A322D) subunit (Fig.  1B, cf. lane 8 with lane 6), indicating that proteasome inhibition led to the accumulation of total ␣1 subunits in cells. To determine whether lactacystin enhanced the maturation of the ␣1(A322D) subunit, we used an endo H enzyme digestion assay to monitor its folding and trafficking. The endo H enzyme selectively cleaves after N-acetylglucosamine (GlcNAc) in the N-linked glycans incorporated on the ␣1 subunit in the ER. After the high-mannose form is remodeled in the Golgi, endo H is not able to remove the oligosaccharide chain. Therefore, endo H-resistant ␣1 subunits represent properly folded, post-ER ␣1 subunit glycoforms that traffic at least to the Golgi compartment. The peptide-N-glycosidase F enzyme cleaves between the innermost GlcNAc and asparagine residues from N-linked glycoproteins, serving as a control for unglycosylated ␣1 subunits (Fig. 1B, lane 3). Lactacystin treatment only increased the immature endo H-sensitive ER glycoforms of the ␣1(A322D) subunit (Fig. 1B, cf. lane 9 with lane 7), indicating that proteasome inhibition itself did not enhance the folding and trafficking of the ␣1(A322D) subunit.  (n ϭ 3). The intensity of ␣1 subunit was quantified, normalized to that of the WT, and is shown in the right panels. IB, immunoblot. C and D, knockdown of VCP using siRNA1 (C) or siRNA2 (D) increases the endo H-resistant post-ER glycoform of the ␣1(A322D) subunit. Forty-eight hours post-siRNA transfection, cells were harvested, lysed, digested by endo H enzyme, and subjected to SDS-PAGE and Western blot analysis (n ϭ 3). After endo H digestion, subunits with a molecular weight equal to unglycosylated ␣1 subunits were considered endo H-sensitive (bottom arrow), whereas those with a higher molecular weight were considered endo H-resistant (top arrow) (lanes 2 and 4). The curly brackets indicate the top endo H-resistant bands in lanes 2 and 4. There are two endo H-resistant bands because the ␣1 subunit has two glycosylation sites (Asn-38 and Asn-138) in the ER. The ratio of endo H-resistant to total ␣1(A322D) serves as a measure of trafficking efficiency of the ␣1(A322D) subunit. Quantification of the ratio of endo H-resistant/total ␣1(A322D) subunit bands is shown in the right panels. E, silencing VCP using siRNA1 increases the BiP protein level. Forty-eight hours post-siRNA transfection, cells were harvested, lysed, and subjected to SDS-PAGE and immunoblotting (n ϭ 2). *, p Ͻ 0.05; **, p Ͻ 0.01.

A Combination Strategy to Enhance GABA A R Proteostasis
JANUARY 2, 2015 • VOLUME 290 • NUMBER 1
Silencing VCP Increases the Total Protein Level and Trafficking Efficiency of the ␣1(A322D) Subunit-Because of the critical role of VCP in extracting misfolded ␣1 subunits from the ER to the cytosol for degradation, we hypothesized that silencing VCP increased the total ␣1(A322D) protein level. Knockdown of VCP using siRNA (69% knockdown efficiency) increased the total protein level of ␣1(A322D) subunits 3.5-fold in HEK293 cells expressing ␣1(A322D)␤2␥2 GABA A receptors, whereas knockdown of VCP using siRNA (60% knockdown efficiency) increased the total protein level of ␣1 subunits 1.6-fold in HEK293 cells expressing WT GABA A receptors ( Fig. 2A, quantification is shown in the right panel). In addition, we tested a second VCP siRNA to avoid a potential siRNA off-target effect. As shown in Fig. 2B, VCP depletion using this siRNA also significantly increased both WT and mutant ␣1 subunit total protein levels, although less increase was observed than with the application of the first VCP siRNA. In both VCP siRNA treatment cases, a more dramatic increase of the total protein was found in cells expressing ␣1(A322D) subunits than in those expressing WT ␣1 subunits.
To determine whether the increased total ␣1(A322D) subunit protein after VCP depletion folded properly in the ER, we used an endo H enzyme digestion assay. Knockdown of VCP using siRNA (69% knockdown efficiency) clearly increased the upper two endo H-resistant ␣1(A322D) subunit bands in HEK293 cells (Fig. 2C, cf. lane 4 with lane 2), indicating that silencing VCP increased properly folded, post-ER glycoforms of the ␣1(A322D) subunit. The ratio of endo H-resistant/total ␣1(A322D) serves as a measure of trafficking efficiency of the ␣1(A322D) subunit. VCP knockdown significantly increased the ratio of endo H-resistant/total ␣1(A322D) (Fig. 2C, cf. lane 4 with lane 2, quantification is shown in the right panel), indicating that VCP knockdown increased the trafficking efficiency of the ␣1(A322D) subunit. Consistently, the application of a second VCP siRNA also yielded a significant increase in ␣1(A322D) subunit trafficking efficiency (Fig. 2D).
Silencing VCP increased the protein level of BiP but not Hsp70, Hsp90, or Grp94 in HEK293 cells expressing ␣1(A322D)␤2␥2 receptors (Fig. 2E). We showed previously that BiP overexpression promotes the folding and trafficking of the ␣1(A322D) subunit (33). Therefore, an increase in BiP protein level after VCP depletion could partially contribute to the enhanced trafficking of the ␣1(A322D) subunit.
However, VCP depletion (69% knockdown efficiency) caused cell toxicity (data not shown). Therefore, in the following experiments, we used a VCP inhibitor under non-toxic conditions and determined its effect on ␣1(A322D) subunit maturation. Fig. 3A for its chemical structure) (35). No apparent cell toxicity was observed for 5 M EerI treatment up to 6 h in HEK293 cells expressing ␣1(A322D)␤2␥2 GABA A receptors according to trypan blue staining (data not shown). Therefore, we used this condition for our following experiments.

Inhibiting VCP Using Eeyarestatin I Increases the Total Protein Level and Trafficking Efficiency of the ␣1(A322D) Subunit, and Coapplication of SAHA Enhances This Effect without an Induction of the Unfolded Protein Response-EerI is a potent VCP inhibitor (see
EerI treatment (5 M, 6 h) significantly increased the total protein level of the ␣1(A322D) subunit, but not the ␤2 or ␥2 subunits, in HEK293 cells expressing ␣1(A322D)␤2␥2 GABA A receptors (Fig. 3B, cf. lane 2 with lane 1). Coapplication of SAHA (2.5 M, 24 h) with EerI (5 M, 6 h) further increased the ␣1(A322D) subunit protein level compared with EerI treatment alone (Fig. 3B, cf. lane 3 with lane 2), and this combination increased the ␤2 or ␥2 subunit protein level because of the effect of SAHA (Fig. 3B) (33).   (Fig. 3C, cf. lane 6 with lane 4, see the right panel for quantification). Interestingly, this combination also increased the trafficking efficiency of WT ␣1 subunits (Fig.  3D, cf. lane 4 with lane 2, see the right panel for quantification), presumably because certain amounts of WT subunits are also targeted to ERAD (Fig. 1B). However, the effect of this coapplication to increase the trafficking efficiency is more dramatic for ␣1(A322D) subunits than WT ␣1 subunits (cf. Fig. 3C with Fig.  3D).
We next evaluated whether such treatments induced the unfolded protein response (UPR). The UPR regulates ER proteostasis by enhancing the ER folding capacity through transcriptional up-regulation of ER chaperones as well as decreasing the ER folding load (36 -38). If ER proteostasis cannot be reestablished during prolonged ER stress, then the UPR can FIGURE 4. EerI treatment reduces the ERAD of the ␣1(A322D) subunit in HEK293 cells, and coapplication of SAHA further attenuates its ERAD without an apparent effect on its endocytosis. A, effect of EerI and SAHA on the degradation of the ␣1(A322D) subunit using CHX chase analysis. HEK293 cells stably expressing ␣1(A322D)␤2␥2 receptors were treated with EerI (5 M, 6 h) in the absence or presence of SAHA (2.5 M, 24 h), followed by a chase with CHX (100 g/ml), which inhibits protein synthesis, for the indicated time. Cells were harvested, lysed, and subjected to 8% SDS-PAGE gel and immunoblot (IB) analysis (n ϭ 3). The dosage regimen is shown at the top. Degradation kinetics were plotted by quantifying ␣1(A322D) intensity against time after CHX addition. Data were normalized to the ␣1(A322D) intensity at t ϭ 0 after CHX addition (right panel). DMSO, dimethyl sulfoxide. B, effect of EerI and SAHA on ubiquitinated ␣1(A322D) subunits. HEK293 cells stably expressing ␣1(A322D)␤2␥2 receptors were treated with EerI (5 M, 6 h) in the absence or presence of SAHA (2.5 M, 24 h). Then the cells were lysed, immunoprecipitated using an anti-␣1 antibody, and blotted for ubiquitin. The dosage regimen is shown at the top. The ubiquitin band intensity relative to ␣1(A322D) subunits post-immunoprecipitation (IP) was quantified and is shown in the right panel (n ϭ 2). C, effect of EerI (5 M, 6 h) and SAHA (2.5 M, 24 h) on the cell surface ␣1(A322D) subunits in the absence or presence of a potent dynamin-1 inhibitor, dynole 34-2 (2.5 M, 24 h). HEK293 cells stably expressing ␣1(A322D)␤2␥2 receptors were treated with chemical compounds as in the dosage regimen. Then the surface ␣1(A322D) subunit was measured using a cell surface biotinylation assay. Quantification of the surface ␣1(A322D) subunit intensity is shown in the right panel (n ϭ 3). Each data point in A, B, and C is reported as mean Ϯ S.E. *, p Ͻ 0.05; **, p Ͻ 0.01. lead to apoptosis. In mammals, three ER stress sensor proteins, IRE1, ATF6, and PERK, have been established. PERK activation eventually leads to apoptosis during sustained ER stress. We tested whether our treatments activated the IRE1 or PERK pathways. IRE1 responds to ER stress by oligomerization, resulting in trans-autophosphorylation that activates its cytosolic endonuclease function, which precisely splices the mRNA that encodes the transcription factor XBP1. RT-PCR analysis was used to detect the spliced form of XBP-1 (XBP1-s) in HEK293 cells expressing ␣1(A322D)␤2␥2 receptors. Thapsigargin (1 M, 6 h), a potent UPR inducer, was used as a positive control (Fig. 3E, lane 5). EerI treatment (5 M, 6 h) with or without SAHA (2.5 M, 24 h) did not lead to the detection of XBP1-s, indicating that the IRE1 arm was not activated (Fig. 3E,  lanes 3 and 4). The transcription factor CCAAT/enhancerbinding protein homologous protein (CHOP, also known as DDIT3/GADD153) operates as a downstream component of PERK activation. Quantitative RT-PCR analysis showed that EerI treatment (5 M, 6 h) with or without SAHA (2.5 M, 24 h) only slightly increased CHOP mRNA levels, but not significantly (Fig. 3F), indicating that such treatments did not result in substantial CHOP induction or proapoptotic PERK activation.
Combining EerI and SAHA Treatment Additively Increases GABA-induced Currents-To study the functional consequence of increased cell surface levels of ␣1(A322D) subunits, we used whole-cell patch clamp electrophysiological experiments to record GABA-induced chloride currents. Compared with the WT, the A322D mutation in the ␣1 subunit resulted in significantly lowered GABA-induced currents. The peak current was 1.0 Ϯ 0.4 picoampere (pA) in response to GABA (3 mM) in HEK293 cells expressing ␣1(A322D)␤2␥2 receptors, whereas it was 138 Ϯ 12 pA in response to GABA (1 mM) in HEK293 cells expressing WT ␣1␤2␥2 receptors (Fig. 6A). EerI (5 M, 6 h) treatment slightly increased the peak current to 4.3 Ϯ 0.3 pA in HEK293 cells expressing ␣1(A322D)␤2␥2 receptors (Fig. 6A, see Fig. 6B for quantification). Acute application of EerI (5 M, 1 min) in the external perfusion-recording medium to HEK293 cells expressing GABA A receptors during the whole-cell patch clamp experiment did not significantly change the GABA-induced peak current (data not shown), indicating that EerI did not act as an antagonist of GABA A receptors. Furthermore, coapplication of EerI (5 M, 6 h) and SAHA (2.5 M, 24 h) increased the peak current to 33.6 Ϯ 3.0 pA in HEK293 cells expressing ␣1(A322D)␤2␥2 receptors, amounting to 24% of that in cells expressing WT receptors (Fig.  6, A and B). This combination significantly rescued function of the misfolding-prone GABA A receptors.

DISCUSSION
Here we showed that VCP inhibition reduces the ERAD of ␣1(A322D) subunits without an apparent effect on their dynamin-1-dependent endocytosis. The ␣1(A322D) subunits cannot efficiently form the TM3 ␣ helix in the ER membrane, resulting in its misfolding (40). Misfolded ␣1(A322D) subunits are recognized by the ER quality control machinery, polyubiquitinated, extracted from the ER membrane to the cytosol, and targeted to the proteasome for degradation. The major players in each ERAD step (15) for GABA A receptors have not yet been established. Our findings strongly support the critical role of VCP in extracting misfolded ␣1(A322D) subunits from the ER membrane to the cytosol for degradation. As a result, VCP inhibition stabilizes ␣1(A322D) subunits in the ER so that more proteins can engage the trafficking machinery for transport to the plasma membrane.
Furthermore, we demonstrated an additive rescue of mutant ␣1(A322D) subunit function in patch clamp experiments using both EerI and SAHA. EerI treatment inhibits the VCP-medi-ated ERAD and accumulates available mutant ␣1 subunits in the ER. Our prior study revealed that SAHA enhances the folding of ␣1(A322D) subunits posttranslationally by promoting calnexin and BiP-assisted folding. Therefore, increased mutant ␣1 subunits in the ER by EerI treatment can fold properly after SAHA treatment, yielding an additive rescue of mutant ␣1 subunit folding, trafficking, and, therefore, function (Fig. 7).
It is desirable to test the effect of SAHA and EerI in vivo. There are at least 19 GABA A receptor subunit isoforms in humans: six ␣ subunits (␣1-6), three ␤ subunits (␤1-3), three ␥ subunits (␥1-3), one ␦ subunit, one ⑀ subunit, one subunit, one subunit, and three subunits (1-3). Many knockout or knockin mice have been generated for various subunits of GABA A receptors (43), including ␣1 knockout mice (44 -46). Recent studies showed that deleting ␣1 subunits can produce absence-like epilepsy (44). Because the A322D mutation leads to exceedingly low surface expression of ␣1 subunits and has a dominant negative effect (47), it is highly possible that the ␣1(A322D) knockin mice recapitulate epilepsy phenotypes. Therefore, it is of great interest to generate ␣1(A322D) knockin mice and examine the in vivo effect of SAHA and EerI in future studies. Because epilepsy is a threshold event, the significant functional rescue using this combination strategy might be clinical relevant and, therefore, holds great promise to be further developed for the treatment of idiopathic epilepsy resulting from GABA A receptor misfolding. 7. Proposed model for an additive rescue of the function of ␣1(A322D) subunits using both EerI and SAHA. VCP extracts misfolded ␣1(A322D) subunits from the ER membrane to the cytosol for degradation through ERAD. SAHA treatment promotes the folding and trafficking of mutant subunits posttranslationally by enhancing calnexin and BiP-assisted folding. EerI treatment inhibits VCP-mediated ERAD and accumulates more mutant subunits in the ER, which can be further rescued by SAHA treatment. Therefore, coapplication of SAHA and EerI yields an additive restoration of the function of mutant subunits. Red asterisk, A322D mutation in the ␣1 subunit. The folded subunit schematics are constructed using the crystal structure of GABA A receptor ␤3 subunits (PDB code 4COF) (48).