Endoplasmic Reticulum-associated Degradation Controls Cell Surface Expression of γ-Aminobutyric Acid, Type B Receptors*

Background: The amount of cell surface GABAB receptors determines the strength of GABAB-mediated inhibition of neuronal excitability. Results: GABAB receptors are Lys48-linked polyubiquitinated and degraded by proteasomes via ERAD. Conclusion: ERAD constitutively degrades GABAB receptors and thereby determines the number of functional receptors available for signaling. Significance: Modulation of ERAD activity may be a mechanism to adjust the level of functional GABAB receptors. Metabotropic GABAB receptors are crucial for controlling the excitability of neurons by mediating slow inhibition in the CNS. The strength of receptor signaling depends on the number of cell surface receptors, which is thought to be regulated by trafficking and degradation mechanisms. Although the mechanisms of GABAB receptor trafficking are studied to some extent, it is currently unclear whether receptor degradation actively controls the number of GABAB receptors available for signaling. Here we tested the hypothesis that proteasomal degradation contributes to the regulation of GABAB receptor expression levels. Blocking proteasomal activity in cultured cortical neurons considerably enhanced total and cell surface expression of GABAB receptors, indicating the constitutive degradation of the receptors by proteasomes. Proteasomal degradation required Lys48-linked polyubiquitination of lysines 767/771 in the C-terminal domain of the GABAB2 subunit. Inactivation of these ubiquitination sites increased receptor levels and GABAB receptor signaling in neurons. Proteasomal degradation was mediated by endoplasmic reticulum-associated degradation (ERAD) as shown by the accumulation of receptors in the endoplasmic reticulum upon inhibition of proteasomes, by the increase of receptor levels, as well as receptor signaling upon blocking ERAD function, and by the interaction of GABAB receptors with the essential ERAD components Hrd1 and p97. In conclusion, the data support a model in which the fraction of GABAB receptors available for plasma membrane trafficking is regulated by degradation via the ERAD machinery. Thus, modulation of ERAD activity by changes in physiological conditions may represent a mechanism to adjust receptor numbers and thereby signaling strength.


Metabotropic GABA B receptors are crucial for controlling the excitability of neurons by mediating slow inhibition in the CNS.
The strength of receptor signaling depends on the number of cell surface receptors, which is thought to be regulated by trafficking and degradation mechanisms. Although the mechanisms of GABA B receptor trafficking are studied to some extent, it is currently unclear whether receptor degradation actively controls the number of GABA B receptors available for signaling. Here we tested the hypothesis that proteasomal degradation contributes to the regulation of GABA B receptor expression levels. Blocking proteasomal activity in cultured cortical neurons considerably enhanced total and cell surface expression of GABA B receptors, indicating the constitutive degradation of the receptors by proteasomes. Proteasomal degradation required Lys 48 -linked polyubiquitination of lysines 767/771 in the C-terminal domain of the GABA B2 subunit. Inactivation of these ubiquitination sites increased receptor levels and GABA B receptor signaling in neurons. Proteasomal degradation was mediated by endoplasmic reticulum-associated degradation (ERAD) as shown by the accumulation of receptors in the endoplasmic reticulum upon inhibition of proteasomes, by the increase of receptor levels, as well as receptor signaling upon blocking ERAD function, and by the interaction of GABA B receptors with the essential ERAD components Hrd1 and p97. In conclusion, the data support a model in which the fraction of GABA B receptors available for plasma membrane trafficking is regulated by degradation via the ERAD machinery. Thus, modulation of ERAD activity by changes in physiological conditions may represent a mechanism to adjust receptor numbers and thereby signaling strength.
The signaling strength of neurotransmitter receptors is significantly controlled by the number of receptors in the plasma membrane. Protein synthesis, cell surface trafficking, endocytotic removal from the plasma membrane, and degradation of the receptors need to be precisely balanced to maintain an appropriate level of cell surface receptors. These mechanisms thus provide means for adapting receptor numbers in response to plastic changes in neurons. There is accumulating evidence that regulated protein degradation via the ubiquitin-proteasome system plays an important integrative role in synaptic plasticity (1)(2)(3). Proteasomal degradation at the endoplasmic reticulum (ER) 3 is crucial for the quality control of newly synthesized receptors. Incorrectly folded and misassembled receptor proteins are efficiently eliminated from the endoplasmic reticulum via the ER-associated degradation (ERAD) (4). Defective receptor proteins are polyubiquitinated, exported from the ER membrane and degraded by proteasomes in the cytoplasm. There is evidence that ERAD may also be involved in the regulation of the number of functional receptors in response to physiological stimuli. Prolonged activation of IP 3 receptors, which release Ca 2ϩ from the ER, down-regulates the expression of the receptors in ER membranes via ERAD-dependent proteasomal degradation (5). This is thought to be a homeostatic response to counterbalance excessive accumulation of Ca 2ϩ in the cytoplasm. However, it is currently unclear whether the ERAD machinery contributes to the regulation of the cell surface density of neurotransmitter receptors. GABA B receptors are G protein-coupled receptors assembled from the two subunits GABA B1 and GABA B2 . They mediate slow inhibitory neurotransmission in the CNS and are thought to be involved in a variety of neurological disorders (6). It is meanwhile well established that GABA B receptors are * This work was supported by Swiss National Science Foundation Grants endocytosed from the plasma membrane via the classical dynamin-and clathrin-dependent pathway and are eventually degraded in lysosomes (7). Lysosomal targeting appears to be mediated by the ESCRT (endosomal sorting complex required for transport) machinery (8) that sorts mono-and Lys 63 -linked polyubiquitinated proteins to lysosomes (9). It is currently unclear whether proteasomal degradation contributes to the regulation of GABA B receptors available for signal transduction. Therefore, we tested in this study the hypothesis that cell surface levels of GABA B receptors might be controlled by proteasomal degradation.
Mutation of GABA B2 -Lysines 767 and 771 in GABA B2 were mutated to arginines using the QuikChange II XL site-directed mutagenesis kit from Stratagene according to the manufacturer's instructions.
Culture and Transfection of Cortical Neurons-Primary neuronal cultures of cerebral cortex were prepared from day 18 embryos of time-pregnant Wistar rats as described previously (10,11). Neurons were kept in culture for 12-17 days before being used. Neurons were transfected with plasmid DNA using magnetofection as detailed by Buerli et al. (16).
Proteasome Activity Assay-Neurons cultured in 96-well plates were incubated for 12 h with either 10 M MG132, 50 M lactacystin, or 20 M betulinic acid followed by determination of proteasome activity using the Proteasome Glo Chymotrypsin-like cell-based assay (Promega) according to the manufacturer's instructions.
Immunoprecipitation and Western Blotting-Immunoprecipitation of GABA B receptors from deoxycholate extracts of rat brain membranes and Western blotting for the detection of GABA B2 and ubiquitin was done as described previously (10,17).
Immunocytochemistry and Confocal Laser Scanning Microscopy-Double labeling immunocytochemistry was performed with cortical neurons cultured on coverslips as described previously (10,11,17). Neurons were analyzed by confocal laser scanning microscopy (LSM510 Meta; Zeiss, 100ϫ plan apochromat oil differential interference contrast objective, 1.4 NA) at a resolution of 1,024 ϫ 1,024 pixels in the sequential mode. Quantification of fluorescence signals and image processing was done as detailed in Ref. 11. Images shown represent a single optical layer.
In-cell Western Assay-The in-cell Western assay was exactly done as in Ref. 11. Neurons cultured in 96-well plates were treated with the drug to be tested for the indicated time at 37°C and 5% CO 2 . After fixation and permeabilization, the neurons were incubated simultaneously with GABA B receptor and actin antibodies. Nonspecific GABA B receptor antibody binding was determined in parallel cultures by competition using the respective peptide-antigen (10 g/ml). After incubation with the appropriate secondary antibodies, the fluorescence was measured with the Odyssey infrared imaging system (LI-COR Biosciences). Specific GABA B signals were normalized to the actin signal determined in parallel.
In Situ Proximity Ligation Assay (PLA)-The in situ PLA technology is a highly sensitive antibody-based method for the microscopic detection of protein-protein interactions and post-translational protein modifications in cultured cells and tissue section (18,19). For in situ PLA, we used Duolink PLA probes and detection reagents according to the manufacturer's instructions (Olink Bioscience). The specificity of the PLA signal was validated for each pair of antibodies in HEK 293 cell expressing or not expressing GABA B receptors. In addition, in neurons, omitting one of the primary antibodies did not generate PLA signals.
For signal quantification, cells were imaged for GABA B receptor expression and PLA signals by confocal microscopy (LSM510 Meta; Zeiss, 100ϫ plan apochromat oil differential interference contrast objective, 1.4 NA, resolution 1,024 ϫ 1,024 pixels, sequential mode). GABA B receptor fluorescence intensities, PLA spots, and the cell area were quantified using ImageJ. PLA signals were normalized to the GABA B receptor signal and the cell area.
Electrophysiology-Cortical neurons at 13-15 days in vitro were recorded in the whole cell voltage clamp configuration at room temperature. Total spontaneous postsynaptic currents (sPSCs) were recorded at a holding potential of Ϫ60 mV. Baclofen-evoked potassium currents were elicited using a 10-s pulse of 50 M baclofen at Ϫ90 mV. Patch electrodes were filled with 120 mM CsCl/KCl, 10 mM EGTA, 10 mM HEPES (pH 7.4), 4 mM MgCl 2 , 0.5 mM GTP, and 2 mM ATP. Spontaneous PSCs recordings were performed using intracellular CsCl, whereas the potassium currents were recorded using an intracellular solution containing KCl. The external solution contained 140 mM NaCl, 10 mM KCl, 2 mM CaCl 2 , 1 mM MgCl 2 , 10 mM HEPES (pH 7.4), and 10 mM glucose. Potassium currents were recorded in the presence of tetrodotoxin (0.5 M), 7-nitro-2,3-dioxo-1,4dihydroquinoxaline-6-carbonitrile (CNQX, 2 M), and bicuculline (4 M). To enhance the amplitude of the baclofen-evoked currents, the potassium concentration of the extracellular solution was increased to 30 mM, and the sodium concentration was reduced to 120 mM (to keep osmolarity constant) before the application of the GABA B agonist. All the synaptic events displaying amplitudes above the background noise (5-12 pA) were identified and analyzed off-line using MiniAnalysis 6.0.7 software (Synaptosoft). Mean amplitudes and frequency values were obtained from 1 min of epoch recordings on each experimental condition and normalized to the control condition of the individual neuron.
Statistical Analysis-The data are presented as means Ϯ S.E. The statistical analysis of data were performed with the GraphPad Prism 5 software. Unpaired t test was used for comparing two conditions and one-way ANOVA followed by Dunnett's post hoc test for analysis of multiple conditions. The level of significance and the n values are indicated in the figure legends. Differences were considered statistically significant when p Ͻ 0.05.

The Expression Level of GABA B Receptors Is Controlled by
Proteasomal Degradation-It is currently unknown whether the ubiquitin-proteasome system contributes to the regulation of GABA B receptor expression levels in neurons. To gain evidence for a potential degradation of GABA B receptors by proteasomes, we treated cultured cortical neurons for 12 h with the proteasome inhibitors MG132 or lactacystin and determined the GABA B1 and GABA B2 protein expression levels. Under these conditions, MG132 and lactacystin decreased proteasomal activity to 31 Ϯ 2 and 17 Ϯ 1% of untreated controls, respectively (Fig. 1A). Both drug treatments increased total GABA B receptor expression levels (MG132: GABA B1 , 131 Ϯ 2%; GABA B2 , 143 Ϯ 5%; lactacystin: GABA B1 , 142 Ϯ 4%; GABA B2 , 147 Ϯ 2% of control; Fig. 1B), suggesting that under basal conditions GABA B receptors were constitutively degraded to a certain extent by proteasomes.
Prolonged inhibition of proteasomes depletes the pool of free ubiquitin (20,21), which might also affect ubiquitin-dependent processes unrelated to proteasomal degradation. There is some evidence that GABA B receptors are sorted to lysosomes via the ubiquitin-dependent ESCRT (endosomal sorting complex required for transport) machinery (8). Hence, prolonged inhibition of proteasomes might indirectly compromise lysosomal degradation of the receptors. However, an indirect contribution of lysosomal degradation could be ruled out. Pharmacologically increasing proteasome activity by treating cortical neurons for 12 h with the proteasome activator betulinic acid (22), which enhanced proteasomal activity to 143 Ϯ 17% of control (Fig. 1A), significantly decreased GABA B receptor levels (GABA B1 , 69 Ϯ 2%; GABA B2 , 64 Ϯ 3% of control; Fig. 1C).
It has recently been shown that inhibition of the proteasomeassociated deubiquitinating enzyme USP14 enhanced the degradation of proteasome substrates (23). Inhibition of USP14 by incubation of cortical neurons with SMI-USP14 (small molecule inhibitor of USP14) strongly reduced GABA B receptor levels (GABA B1 , 49 Ϯ 4%; GABA B2 , 29 Ϯ 2% of control; Fig. 1D), further supporting the view that GABA B receptors are degraded by proteasomes.
Finally, we assessed the functional consequences of decreased GABA B receptor levels after enhancing proteasomal activity with betulinic acid by measuring spontaneous synaptic activity in electrophysiological experiments. Activation of GABA B receptors with the selective agonist baclofen considerably decreased the amplitude as well as the frequency of sPSCs to 43 Ϯ 4 and 56 Ϯ 7%, respectively (Fig. 1E). Treatment of cultures for 12 h with betulinic acid diminished baclofen-induced inhibition of sPSCs (amplitude, from 43 Ϯ 4 to 90 Ϯ 12% of control; frequency, from 56 Ϯ 7 to 94 Ϯ 15% of control; Fig.  1E), supporting the hypothesis that enhanced proteasomal activity leads to reduced levels of functional GABA B receptors available for neuronal inhibition.
GABA B Receptors Undergo Lys 48 -linked Polyubiquitination-Lys 48 -linked polyubiquitination of proteins serves as a signal for proteasomal degradation. Consistent with polyubiquitination, GABA B receptors immunoprecipitated from deoxycholate extracts of crude rat brain membranes exhibited on Western blots ubiquitin immunoreactivity in the high molecular range ( Fig. 2A). This suggests that GABA B receptors are ubiquitinated under basal conditions to a certain extent. Likewise, using the in situ PLA, we found that GABA B receptors in cultured cortical neurons display Lys 48 -linked polyubiquitination, which was considerably increased upon inhibition of proteasomal activity with MG 132 (172 Ϯ 11% of control; Fig. 2B). This indicates the accumulation of Lys 48 -linked polyubiquitinated GABA B receptors destined for proteasomal degradation.
The C-terminal Domain of GABA B2 Contains a Major Lys 48linked Polyubiquitination Site-An in silico analysis predicted two lysines in the C-terminal domain of GABA B2 at positions 767 and 771 as likely candidates for ubiquitination. We inactivated these potential ubiquitination sites by exchanging both lysines for arginines (GABA B2 (RR)) (Fig. 3A). Upon transfection into HEK 293 cells, GABA B2 (RR) displayed reduced Lys 48 -linked polyubiquitination (61 Ϯ 6% of wild type; Fig. 3B), indicating that Lys 767/771 is a main site for Lys 48 -linked polyubiquitination in GABA B2 .
Overexpressing GABA B2 (RR) in cultured neurons increased GABA B receptor levels to a similar level as observed after chronic proteasome inhibition (GABA B1 , 152 Ϯ 15%; GABA B2 , 146 Ϯ 9% of control; Fig. 3D). This suggests that Lys 767/771 in GABA B2 is the major Lys 48 -linked polyubiquitination site required for proteasomal degradation of GABA B receptors.
The functional consequence of the increased GABA B2 cell surface density after transfecting neurons with GABA B2 (RR) was analyzed by measuring baclofen-induced K ϩ currents using whole cell patch clamp recordings. Transfection of GABA B2 (RR) in neurons resulted in 2.8 Ϯ 0.6-fold increased K ϩ channel current amplitudes after activation of GABA B receptors with baclofen as compared with neurons transfected with wild type GABA B2 (Fig. 3E). Thus, preventing proteasomal degradation of GABA B2 by overexpression of GABA B2 (RR) increased the number of functional cell surface GABA B receptors available for signaling.
Cell Surface Expression of GABA B Receptors Is Regulated by ERAD-The most likely mechanism for proteasomal degradation of GABA B receptors is the ERAD. If GABA B receptors are degraded by ERAD, inhibition of proteasomal activity should result in an accumulation of GABA B receptors in the ER. Indeed, blocking proteasomal activity in neurons for 12 h with MG132 increased the number of GABA B2 clusters (136 Ϯ 6% of control) as well as the clusters co-localizing with a marker protein for the ER (protein-disulfide isomerase (PDI), 133 Ϯ 7% of control; Fig. 4A).
To further establish the role of ERAD in regulating cellular GABA B receptor levels, we tested the effect of directly inhibit- were immunoprecipitated from deoxycholate extracts of rat brain membranes using either GABA B1a,b or GABA B2N antibodies. The immunoprecipitate was subjected to Western blotting for detection of GABA B2 and ubiquitin. The high molecular smear detected with the ubiquitin antibody is typical for polyubiquitinated proteins. The ubiquitin signal in the GABA B1 immunoprecipitate was considerably weaker than in the GABA B2 immunoprecipitate because the GABA B1a,b antibody beads were less efficient in precipitating GABA B receptors than the GABA B2 antibody beads. Specificity of the immunoprecipitation was verified with nonimmune antibodies (control). IP, immunoprecipitate; Ab, antibody. B, inhibition of proteasomes enhanced Lys 48 -linked polyubiquitination of GABA B receptors. Neurons were incubated for 12 h in the absence (control) or presence of MG132 and processed for in situ PLA using antibodies directed against GABA B2 and Lys 48linked polyubiquitin to detect Lys 48 -linked polyubiquitinated GABA B receptors (white dots in images, left panel). Right panel, quantification of in situ PLA signals (n ϭ 30 neurons). ***, p Ͻ 0.0001, t test. Scale bar, 5 m. C, overexpression in neurons of a Lys 48 chain elongation defective ubiquitin mutant upregulated GABA B receptors. Neurons were transfected with plasmids containing HA-tagged ubiquitin (Ub) or HA-tagged mutant ubiquitin (Ub(K48R)). Neurons were stained for GABA B1 or GABA B2 (red) and Ub (green). Top panels, representative images. Bottom panels, quantification of total GABA B1 and GABA B2 levels in neurons expressing Ub or Ub(K48R) (n ϭ 28 -40 neurons). ***, p Ͻ 0.0001, t test. Scale bar, 10 m. ing ERAD. Treatment of neurons for 12 h with the ERAD inhibitor Eeyarestatin I (EerI) (24,25) increased both total GABA B2 (183 Ϯ 15% of control; Fig. 4B) and cell surface levels of GABA B2 (204 Ϯ 32% of control; Fig. 4C). Overexpression of GABA B2 (RR), which lack the main Lys 48 -linked polyubiquitination sites, did not further increase total (112 Ϯ 7% of control; Fig. 4D) or cell surface GABA B receptor levels (93 Ϯ 15% of control; Fig. 4E). These observations indicate that Lys 48 -linked polyubiquitinated GABA B receptors are degraded by ERAD.
GABA B Receptors Interact with the ERAD E3 Ubiquitin Ligase Hrd1-Hrd1 is one prototypical ERAD E3 ubiquitin ligase responsible for Lys 48 -linked polyubiquitination of ERAD substrates (26). Using in situ PLA, we further confirmed the potential degradation of GABA B receptors via ERAD by showing that GABA B receptors interact with Hrd1 (Fig. 5A). Inhibition of ERAD for 12 h with EerI increased the number of interactions (GABA B2 /Hrd1, 490 Ϯ 45%; GABA B1 /Hrd1, 305 Ϯ 18% of control; Fig. 5A), indicating the accumulation of GABA B receptors at this central ERAD multiprotein complex. In line with this observation, blocking ERAD function for 12 h with  EerI considerably increased the level of Lys 48 -linked polyubiquitinated GABA B receptors (242 Ϯ 21% of control; Fig. 5B).
GABA B Receptors Interact with the Essential ERAD Component p97-The AAA-ATPase p97 is a central constituent of the ERAD machinery involved in the retrotranslocation of proteins to the cytoplasm for proteasomal degradation (27). Using in situ PLA, we found that GABA B receptors interact with p97 (Fig. 6A). This finding further demonstrates the ERAD-mediated degradation of GABA B receptors. Inhibition of p97 by EerI decreased the interaction of GABA B2 with p97 (40 Ϯ 8% of control; Fig. 6A), suggesting that the association is activity-dependent.
Whole cell patch clamp recordings finally verified that inhibition of ERAD function by overexpression of p97(DKO) increased the level of functional cell surface GABA B receptors (Fig. 6F). Neurons overexpressing p97(DKO) displayed considerably increased amplitudes of baclofen-induced K ϩ currents (control, 72 Ϯ 14 pA; p97(DKO), 139 Ϯ 14 pA; Fig. 6F). These experiments show that GABA B receptors are degraded by ERAD, which affects the levels of total and cell surface GABA B receptors.

DISCUSSION
Mechanisms controlling the cell surface density of GABA B receptors are of pivotal importance for determining the level of GABA B receptor-mediated neuronal inhibition. Because GABA B receptors control glutamatergic neurotransmission (28), modulation of their cell surface density is presumed to significantly contribute to synaptic plasticity. However, the mechanisms that control cell surface expression of GABA B receptors are largely unknown. In the present study, we identified proteosomal degradation via the ER-resident ERAD machinery as a mechanism that determines cell surface expression of GABA B receptors.
Our data indicate that a fraction of GABA B receptors in the ER is constitutively Lys 48 -linked polyubiqutinated and . Neurons were transfected with plasmids containing HA-tagged p97 or its dominant-negative mutant HA-tagged p97(DKO) and stained for GABA B2 (red) and HA (green). Scale bars, 10 m. Lower panels, quantification of GABA B2 fluorescence signals (n ϭ 40 -50 neurons). ***, p Ͻ 0.0001, t test. D and E, overexpression of GABA B2 (RR) in neurons transfected with wild type p97 or p97(DKO) did not result in an additional increase of total (D) or cell surface (E) GABA B2 levels. Neurons were transfected with plasmids containing GABA B2 (RR) and either HA-tagged p97 or its dominant-negative mutant HA-tagged p97(DKO) and stained for GABA B2 (red) and HA (green). Scale bars, 10 m. Lower panels, quantification of GABA B2 fluorescence signals (n ϭ 28 -30 neurons). n.s., p Ͼ 0.05, t test. F, disruption of ERAD by overexpression of p97(DKO) in neurons increased GABA B receptor-mediated K ϩ currents. Neurons were transfected either with plasmids containing EGFP (control) or with plasmids containing p97(DKO). Left panels, representative traces of baclofen-induced K ϩ currents. Right panel, K ϩ current amplitudes (n ϭ 10 for control and n ϭ 8 for p97(DKO)). *, p Ͻ 0.05, t test. degraded by the ERAD machinery. This conclusion is based on the observation that blocking proteasomal activity, inhibiting ERAD function, or interfering with GABA B receptor Lys 48linked polyubiquitination increased the expression levels of GABA B receptors in neurons. Lysines 767/771 in the C-terminal domain of GABA B2 appear to represent the main Lys 48linked polyubiquitination sites required for proteasomal degradation because their mutational inactivation rendered GABA B receptors largely immune to degradation. It is currently unclear whether Lys 48 -linked polyubiquitination of both lysines or only of Lys 767 or Lys 771 serves as a tag for proteasomal degradation. A recent proteomic study analyzing the ubiquitination state of rat brain synaptic proteins identified Lys 771 in GABA B2 as being ubiquitinated (29). This observation favors Lys 771 as the main Lys 48 -linked polyubiquitination site in GABA B2 .
There are several lines of evidence that in particular GABA B receptors residing in the ER are degraded by proteasomes via ERAD. First, upon blocking proteasomal activity, the receptors accumulated in the ER. Second, blocking ERAD function pharmacologically or by overexpressing a dominant-negative mutant of the AAA-ATPase p97, which mediates the retrotranslocation of proteins to the cytoplasm for proteasomal degradation (27), increased GABA B receptor levels. Third, GABA B receptors interacted with the ERAD proteins p97 and Hrd1. Hrd1 is the prototypical ERAD E3 ligase (26) and most likely one of the ubiquitin ligases that mediate ubiquitination of GABA B receptors because stalling proteasomal degradation considerably increased its interaction with GABA B receptors and the level of Lys 48 -linked polyubiquitinated GABA B receptors.
In all cases tested, GABA B1 and GABA B2 were concomitantly up-or down-regulated to a similar extent, suggesting that assembled GABA B receptor complexes are degraded by ERAD. This notion is further strengthened by the finding that 1) inactivation of the ubiquitination sites in GABA B2 increased the expression levels of GABA B1 and GABA B2 as well as GABA B receptor-activated K ϩ current amplitudes, 2) that interfering with ERAD function increased GABA B receptor function (baclofen-induced K ϩ currents), and 3) that both GABA B1 and GABA B2 generated in situ PLA signals with the ERAD E3 ubiquitin ligase Hrd1, although only Lys 48 -linked polyubiquitination of Lys 767/771 in GABA B2 appears to be required for proteasomal degradation of the receptors.
What might be the physiological implications of this mechanism? The most firmly established function of ERAD is the degradation of aberrant proteins in the ER (30). In addition, ERAD has been shown to rapidly degrade activated IP 3 receptors in the ER to prevent excessive elevation of cytosolic Ca 2ϩ concentrations (5), indicating that ERAD may also contribute to the regulation of functional receptors. Because blocking ERAD increased the level of functional GABA B receptors, and ERAD appears to degrade assembled heterodimeric receptors, it is rather unlikely that the role of ERAD is simply the degradation of un-or misfolded GABA B receptor subunits. The constitutive degradation of GABA B receptors suggests that ERAD controls the amount of receptors available for cell surface trafficking. This view is supported by recent studies on the regulation of cell surface GABA A receptors. Chronic suppression of neuronal activity or inhibition of L-type voltage-gated calcium channels decreased the level of functional GABA A receptors in the neuronal plasma membrane by a mechanism dependent on the ubiquitination of the GABA A receptor ␤3-subunit and proteasome activity, most likely via the ERAD pathway (31,32). These findings imply that regulation of ERAD activity is a potential mechanism to adjust the level of functional GABA B receptors to changing physiological condition. Our finding that modulation of proteasomal activity up-or down-regulates the level of functional GABA B receptors supports this view. Interestingly, the level of proteasomal activity correlates with the activity state of neurons (33). We therefore hypothesize that the amount of functional GABA B receptors inserted into the plasma membrane is regulated by neuronal activity via ERAD.