点击显示 收起
Departments of Molecular and Medical Pharmacology (Z.-W.C., C.-S.S.C., T.A.L., R.W.O.) and Anesthesiology (R.O., R.W.O.), Division of Molecular Medicine, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, California
Abstract
GABAA receptor-associated protein (GABARAP) was isolated previously in a yeast two-hybrid screen using the intracellular loop of the 2 subunit of the GABAA receptor as bait. GABARAP has been shown to participate in the membrane-clustering and intracellular-trafficking of GABAA receptors, including a stimulation of the surface expression of GABAA receptors. To assess this quantitatively, we used Xenopus laevis oocytes expressing 122S-containing GABAA receptors to demonstrate that coexpression of GABARAP increased net surface levels of GABAA receptors as shown by both increased GABA currents and surface-expressed protein. This GABARAP stimulation of GABA currents required the receptor 2 subunit and full-length GABARAP: deletion of the microtubule-binding domain (amino acids 1eC22) or disrupting the polymerization of microtubules abolished the enhancement, indicating that the effect of GABARAP was derived from the interaction with microtubules. GABARAP coexpression did not alter the general properties of GABAA receptors such as sensitivity to GABA or benzodiazepines, but it increased surface levels of receptor protein in oocytes. Rather, it seems to supplement inadequate amounts of endogenous GABARAP to support optimum trafficking and/or stabilization of surface GABAA receptors.
GABA, an important inhibitory neurotransmitter in both vertebrates and invertebrates, acts on GABAA and GABAB receptors that are expressed ubiquitously in the central nervous system. GABAA receptors also represent a major site of action of clinically relevant drugs, such as benzodiazepines, barbiturates, and general anesthetics. In mammals, 19 related subunits have been detected: 1 to 6, 1 to 3, 1 to 3, , , , , and 1 to 3. The subunits are combined into approximately 20 different native heteropentameric isoforms. The combination of 1, 2, and 2 in the ratio of 2:2:1 is the most abundant form in the adult central nervous system (Macdonald and Olsen, 1994). Each subunit possesses a long extracellular N terminus, which in some cases carries the neurotransmitter-binding site, four membrane-spanning domains, including the ion channel wall M2, and a large variable-sequence intracellular loop between M3 and M4. Because both the N and C termini of GABAA receptor subunits extend outside the cell membrane, the intracellular M3eCM4 loop becomes the most important domain interacting with the intracellular environment. As a consequence, studies of the function of receptor intracellular loop-associated proteins become important for understanding the mechanisms of regulating receptor activity.
GABAA receptor-associated protein (GABARAP) was cloned in a yeast two-hybrid system using the intracellular loop of 2 as bait. GST pull-down and coimmunoprecipitation results confirmed this interaction in vivo (Wang et al., 1999). The interaction is specifically limited to 1, 2S, 2L, and 3 but not other GABAA receptor subunits (Wang et al., 1999; Nymann-Andersen et al., 2002b). GABARAP binds with 2 subunits through a.a. 36 to 52 in GABARAP (Wang et al., 1999; Nymann-Andersen et al., 2002b), especially a.a. 36 to 40 (Leil et al., 2004). This is followed by or overlaps with a self-dimerization domain at a.a. 41 to 51 (Wang et al., 1999; Nymann-Andersen et al., 2002b,c; Leil, 2004). The sequence of GABARAP has 31% identity and 64% similarity to light chain 3 of microtubule-associated protein, suggesting that GABARAP might link GABAA receptors to the cytoskeleton. GABARAP interacts directly with both soluble tubulin and microtubules, as demonstrated by GST pull-down and coimmunoprecipitation, and is also associated with microfilaments in intact cells. The cellular distribution of GABARAP was altered by cytoskeleton-disrupting drugs, such as nocodazole, taxol, and cytochalasin D (Wang and Olsen, 2000). Studies derived from GST pull-down using truncated GABARAP (Wang and Olsen, 2000) and crystal structure(Coyle et al., 2002) confirmed that the -helix in the N terminus (a.a. 1eC22) of GABARAP is responsible for its interaction with microtubules. This tubulin-binding domain is required for GABARAP clustering of GABAA receptors in a quail fibroblast recombinant-expression system (Chen et al., 2000). The clustered receptors demonstrate slightly altered channel kinetics: lower apparent affinity for GABA, faster deactivation, and slower desensitization (Chen et al., 2000). In L929 cells, coexpression of GABARAP promotes the formation of GABAA receptor clusters that are capable of showing large single-channel conductance, which is observed in neurons but not in recombinant cells lacking GABARAP (Everitt et al., 2004). This suggests that GABARAP might play an important role in the membrane, possibly synaptic, organization of GABAA receptors.
GABARAP also binds with gephyrin, a postsynaptic marker of GABAA receptors (Sassoe-Pognetto et al., 1995; Essrich et al., 1998), in both the yeast two-hybrid system and GST pull-down experiments (Kneussel et al., 2000). Coexpression of both GABARAP and gephyrin results in a recruitment of cytoplasmic gephyrin to GABARAP-rich membrane-associated loci in PC12 cells. However, little GABARAP was colocalized with GABAA receptors at postsynaptic sites, and GABARAP localization remains normal in the gephyrin knockout mouse, whereas postsynaptic GABAA receptors are disrupted, suggesting that GABARAP functions before or independently of gephyrin. This further indicates that GABARAP might participate in intracellular receptor trafficking rather than anchoring (Kneussel et al., 2000; Kittler et al., 2001). We have shown that GABAA receptor 2 subunits can interact with each other, providing a possible mechanism for aggregation at synapses, after delivery to the cell surface, possibly near synapses, by GABARAP (Nymann-Andersen et al., 2002a). In addition, GST pull-down and immunofluorescence microscopy demonstrated that GABARAP interacts with the ATPase, N-ethylmaleimideeCsensitive factor, a protein implicated in intracellular vesicle fusion (Kittler et al., 2001). From these results, we predict that GABARAP might regulate the cell-surface number of GABAA receptors through transportation along the cytoskeleton. However, there is no direct evidence supporting this prediction. To address this issue, we used the Xenopus laevis oocyte expression system, which allowed us to measure quantitatively the influence of GABARAP on the levels of surface-expressed GABAA receptors using both a functional assay of current amplitude with the two-electrode voltage clamp and the amount of receptor protein expressed at the surface using biotinylation of intact cells. The results showed that coexpressed exogenous GABARAP increases the surface levels of GABAA receptors by interacting with 2 subunits, and microtubules are essential for this enhancement. This might involve increased delivery to the cell surface or reduced removal of receptors from the cell surface.
Materials and Methods
Molecular Cloning. Rat 1, 2, and 2S GABA receptor subunits and GABARAP cDNAs were subcloned into expression vector pBlueScript II SK after digestion by PstI and HindIII. Deletion of a.a. 1 to 22 from GABARAP was generated by polymerase chain reaction using primers 5'-CTG CAG ATT ATT AAG AAA TAC CCG GAC CGG-3' and 5'-AAG CTT TCA CAG ACC GTA GAC ACT-3'. It was then subcloned into pBlueScript II SK as for the other clones.
Expression of Rat GABAA Receptors in X. laevis Oocytes. Capped mRNA (cRNA) was synthesized by in vitro transcription from ApaI-linearized cDNA constructs using the mMessage mMachine kit (Ambion, Austin, TX) as described previously (Chang et al., 2003). cRNA concentrations were measured by both UV absorption and RNA gel electrophoresis. X. laevis oocytes were prepared and injected with a total volume of 50 nl of cRNA mixture. After injection, oocytes were maintained in six-well plates at 17 to 19°C in 100 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, and 5 mM HEPES supplemented with 50 e蘥/ml gentamycin and 100 e蘥/ml streptomycin and penicillin.
Two-Electrode Voltage-Clamp Analysis. Three to five days after injection of cRNA, oocytes under a two-electrode voltage clamp (voltage held at eC70 mV) were gravity-perfused continuously with ND96 recording solution (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, and 5 mM HEPES, pH 7.4) at an approximate rate of 5 ml/min as described previously (Chang et al., 2003). The perfusion was stopped during the administration and incubation period to keep a constant drug concentration. Electrodes were filled with 3 M KCl and had a resistance of 0.5 to 1.5 M. The oocytes were exposed to various concentrations of GABA ranging from 0.1 e to 1 mM for 30 s. The recording was carried out using an Axoclamp-2A amplifier (Axon Instruments Inc., Union City, CA) interfaced to a computer with a DigiData 1322-A device (Axon Instruments). The data were processed with two-way ANOVA, and the dose-response curves were fitted by nonlinear regression using Prism software (GraphPad Software Inc., San Diego, CA). The equation for sigmoidal dose-response (variable slope) is Y = bottom + (top eC bottom)/(1 + 10(logEC50 eC X) x nH). X is the logarithm of concentration, Y is the response, and nH is the Hill slope. Y starts at bottom and goes to top with a sigmoid shape. The EC50 value for each curve was calculated individually and analyzed by t test.
Cut-Open Oocyte Voltage Clamp. The same day after recording the GABA current, potassium currents were recorded by the cut-open Vaseline gap voltage clamp (Stefani and Bezanilla, 1998). The external solution contains 110 mM sodium methanesulfonate, 2 mM Ca(MES)2, and 10 mM HEPES. The internal solution contains 120 mM potassium glutamate and 10 mM HEPES. Intracellular micropipettes were filled with 2700 mM sodium methanesulfonate and 10 mM NaCl. All solutions were buffered to pH 7.0. Low-access resistance to the oocyte interior was obtained by permeabilizing the oocyte bottom with 0.1% saponin.
Cell-Surface Biotinylation. Surface-expressed proteins were biotinylated by the membrane-impermeable EZ-link Sulfo-NHS-SS-biotin (Pierce Chemical, Rockford, IL). Four days after the cRNA injection, oocytes (15 oocytes/group) were washed with ND96 solution three times and then incubated with 1 mg/ml Sulfo-NHS-SS-biotin at room temperature for 30 min. Then cells were washed with 25 mM Tris, pH 8.0, followed by two washes with phosphate-buffered saline. Oocytes were homogenized in 200 e蘬 of RIPA buffer (150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, and 50 mM Tris, pH 7.5). The yolk and cellular debris were removed after centrifugation at 3600g for 10 min three times. A 20-e蘬 aliquot was taken and mixed with 2x SDS loading buffer to detect the total protein. The remaining supernatant was incubated with streptavidin beads (Pierce) at 4°C overnight.
Western Blot. The biotinylated proteins were eluted from the beads by SDS loading buffer. The eluted proteins and the whole proteins were loaded on SDS-PAGE and transferred to polyvinylidene difluoride membrane by a semidry method. The membrane was then incubated in phosphate-buffered saline buffer and 0.05% Tween with 5% nonfat milk and anti-2 at room temperature for 1 h. After the incubation with secondary antibody for 1 h, the membrane was visualized by enhanced chemiluminescence (Amersham Biosciences Inc., Piscataway, NJ). The blots were scanned and analyzed with Quantity One software (Bio-Rad, Hercules, CA), compared by oneway ANOVA.
Results
GABARAP Increases GABA Currents in Oocytes Expressing 122S GABAA Receptor Subunits, and a Control Recombinant Channel, the Shaker K+ Channel, Is Not Changed. The cRNA mixture of 1, 2, and 2S subunits together with GABARAP or H2O in the ratio of 1:1:2:4 was injected into oocytes. Three to five days after the injection, GABA currents were measured with two-electrode voltage clamp with the holding potential at eC70 mV. The oocytes were exposed to different doses of GABA for 30 s. Application of GABA ranging from 1.0 e to 1 mM to oocytes produced a concentration-dependent increase in current. Figure 1 shows typical GABA currents elicited by different doses of GABA, without (a) and with (b) GABARAP. The dose-response relationship was obtained by normalizing each current to the maximum average effect of the "no GABARAP" group in the same batch to minimize the variance between different sets of oocytes. Over the concentrations ranging from 10 e GABA, a significant inward current was observed that reached a plateau at 1 mM. GABA currents recorded from oocytes coexpressing GABARAP with 122S (n = 16) subunits were significantly bigger than those expressing only 122S subunits (, p < 0.01 from two-way ANOVA, n = 10). The enhancement by GABARAP was 2- to 3-fold. The enhancement was only caused by the larger maximum response to GABA, not the apparent affinity, because the EC50 value for GABA was not altered. Despite different levels of expression in different batches of injected oocytes, within each given batch, the increased GABA current with GABARAP was robust and reproducible over numerous observations. To show that this enhancement is specific to GABAA receptors, Shaker K+ channel was included in our experiment as a control. In contrast, the current produced by expression of the Shaker K+ channel was not changed in oocytes expressing exogenous GABARAP compared with controls (Fig. 1, bottom).
GABARAP Has No Effect on Flurazepam Enhancement of GABA Current. Oocytes treated the same as above were recorded under two-electrode voltage clamp with the holding potential at eC70 mV. To detect the enhancement of flurazepam on GABA currents, the EC20 value was calculated from the dose-response curve shown in Fig. 1b. The EC20 value for GABA in both curves was approximately 2.5 e. Then, different concentrations of GABA were applied to obtain the dose-response curve of each oocyte. Three of 17 oocytes with EC20 values greater than 3 e or less than 2 e were removed from further flurazepam study. The procedure includes two phases. In both phases, the concentration of flurazepam remains constant. In the first phase, flurazepam was administrated alone. Fifteen seconds later, GABA (2.5 e) was added to the flurazepam for another 30-s incubation. Flurazepam itself at a concentration from 0.01 to 10 e did not induce any current. When combined with 2.5 e GABA, 0.1 e flurazepam enhanced the GABA current approximately 1.5-fold in oocytes expressing only 122S subunits (as shown in Fig. 2). The enhancement reached a maximum by increasing the GABA current 2.5-fold with 10 e flurazepam. The same pattern of enhancement of flurazepam was obtained in oocytes coexpressing 122S subunits together with GABARAP. Comparing these two groups, neither the maximum effect nor the EC50 value of flurazepam had any significant difference (p > 0.05 from two-way ANOVA, n = 14).
Enhancement by GABARAP Is Acting through 2 Subunits. Because 2 subunits are the major interactor of GABARAP, we tested whether 2 subunits were required for the enhancement by GABARAP. Oocytes expressing 12 subunits or 12 subunits with GABRAP were recorded by two-electrode voltage clamp with the holding potential at eC70 mV. As shown in Fig. 3, in general, the GABA currents induced from oocytes expressing only 1 and 2 subunits were lower than those with 122S subunits. The currents measured from oocytes expressing 12 subunits and GABARAP had no significant difference from those expressing only 12 subunits. The dose-response curves of these two groups overlapped with each other (Fig. 3).
The Enhancement Effect of GABARAP on GABA Currents Is Mediated by Polymerized Microtubules. GABARAP has been shown to interact with both soluble and polymerized microtubules (Wang and Olsen, 2000). This interaction was regarded as responsible for facilitating intracellular trafficking of GABAA receptors. In previous studies (Chen et al., 2000), we have shown that interfering with polymerization of microtubules by nocodazole decreased the clustering of GABAA receptors. Here, we used the same drug to determine whether the enhancement effect of GABARAP was functioning through microtubules. Three to five days after the injection of cRNA, oocytes expressing 122S subunits with or without GABARAP were incubated with nocodazole (1 e蘥/ml) overnight. Then, the GABA currents were recorded under the same situation as described above. In agreement with Fig. 1b, without nocodazole treatment, the GABA currents induced from oocytes coexpressing GABARAP with 122S subunit were greater than that expressing the same subunits alone. In general, after treatment by nocodazole, the induced GABA currents were smaller than those from normal oocytes. This phenomenon was more obvious in oocytes expressing both 122S subunits and GABARAP. The GABA currents decreased dramatically after nocodazole incubation, even down to the same level as no GABARAP coexpression (, p < 0.01 from two-way ANOVA). There was no difference between oocytes with or without GABARAP coexpression after the treatment. Nocodazole did not affect the apparent affinity of GABA, because the EC50 value remained the same in these groups (Fig. 4, top).
The N Terminus (a.a. 1eC22) Is Essential for the Enhancement of GABARAP. The results above suggest that polymerized microtubules are important for the function of GABARAP. We used a truncated GABARAP23eC117 to verify whether the interaction between GABARAP and microtubules was essential. We chose this truncated GABARAP on the basis of GST pull-down and crystal structure studies indicating that a.a. 1 to 22 in the N terminus of GABARAP are responsible for microtubule binding (Wang and Olsen, 2000; Coyle et al., 2002). Three groups of oocytes were injected with 122S subunits with H2O, GABARAP, or GABARAP23eC117. Three to five days after the injection of cRNA, the GABA-evoked currents on oocytes were recorded. In agreement with the above results, the currents from the GABARAP group were greater than the water group (, p < 0.05 from two-way ANOVA). The currents from the GABARAP group were also greater than GABARAP23eC117 (p < 0.05 from two-way ANOVA). GABARAP23eC117 lost the enhancement effect on induced GABA currents, because its dose-response curve overlapped that of the non-GABARAP group (Fig. 4, bottom).
Enhanced GABA Currents Are Caused by the Increased Surface Expression of GABAA Receptors. To determine whether the enhanced GABA currents were caused by increased GABAA receptor number on the cell surface, we measured the surface-expressed GABAA receptor by the cell-surface biotinylation assay. Because the functional GABAA receptors observed in these oocytes are well-established to require the and subunits and the demonstrated benzodiazepine sensitivity requires the subunit plus and , the 2 subunit was chosen as a marker to measure the level of pentameric receptor. Four days after injection of 1, 2, and 2S subunit cRNA with H2O, GABARAP, or GABARAP23eC117, oocytes were recorded to confirm the enhancement of GABARAP as shown in Figs. 1 and 4. The remaining oocytes without recording (15 oocytes/group) were incubated with biotin and harvested by RIPA buffer. Aliquots of 20 e蘬 were taken from each group before incubation with streptavidin-agarose to detect the whole protein, including the surface proteins. Both surface and whole proteins were loaded onto SDS-PAGE and were detected by anti-2 (for the M3eCM4 intracellular loop). A 54-kDa band corresponding to 2 was detected in every lane loading either biotinylated proteins or total proteins. In oocytes coinjected with GABARAP and 122S subunits, the surface level of GABAA receptor (Fig. 5) demonstrated a significantly higher level than that without GABARAP (1.97 ± 0.23-fold, p < 0.01, n = 3). In agreement with the results from electrophysiology, deletion of the microtubule-binding domain of GABARAP (a.a. 1eC22) showed the same surface-expression level of GABAA receptors (0.91 ± 0.08-fold, not significant) as that with 122S subunits expressed alone. However, coexpression of GABARAP, either wild-type (1.01 ± 0.03-fold) or truncated (1.16 ± 0.13-fold), did not change the steady-state expression of GABAA receptors, because the total level of GABAA receptor protein was not changed.
Discussion
Our results, for the first time, indicate that coexpression of GABARAP increased the steady-state expression of GABAA receptors at the cell surface. Coexpression of GABARAP increased GABA currents by 2- to 3-fold. The enhancement was limited to an increase in the maximum effect because the EC50 value of GABA was not significantly changed. Neither did the dose-dependence of benzodiazepine enhancement of GABA currents show any significant change in the GABARAP-enhanced GABA currents.
This GABARAP-dependent increase in GABA currents could be caused by increased surface level of GABAA receptors, or it could be caused by the formation of a different sort of GABAA receptors at the surface. To determine whether the GABARAP-dependent increase in GABA currents in oocytes could be caused by the increased surface GABAA receptors, we measured the surface content by biotinylation of surface proteins, collecting them by avidin binding and then identifying GABAA receptors 2 subunit polypeptide by Western blot, a marker for the pentameric receptors. As mentioned in the Results section, the electrophysiological data confirmed that , , and subunits were included as an intact pentameric receptor because of the observed GABA currents and the flurazepam enhancement. In addition, our previous data demonstrated that when GABARAP binds the 2 subunit in brain or cells, it binds the native pentameric protein, as indicated by the presence of ligand-binding in the GABARAP-immunoprecipitated GABAA receptors (Nymann-Andersen et al., 2002b). Our surface-biotinylation result clearly showed that GABARAP increased the surface-expressed GABAA receptors. GABARAP had no effect on the steady-state levels of GABAA receptor, because the total amount of GABAA receptor remained the same.
As mentioned in the Introduction, GABARAP in some cells promotes clustering of GABAA receptors with different channel properties such as single-channel conductance and decay rates (Chen et al., 2000; Everitt et al., 2004). A demonstration of increased surface-receptor protein does not exclude the possibility that increased conductance states could also contribute to greater GABA currents.
Each figure represents a result typical of three repetitions. The absolute GABA currents of oocytes vary 2-fold or even more in different batches because of variable eggs, injection, expression rate, degradation, and time of expression. Within a given batch of oocytes, there is a significantly bigger current with GABARAP coexpression than without; this is also true for the average of many batches. The error bars show that variation within a batch is 5 to 20% at individual concentrations and that GABARAP produces a highly significant 2-fold increase at every point and in every curve as calculated by ANOVA. The shape of the currents may show some slight differences between batches; for example, coexpression of GABARAP might affect the deactivation rate or desensitization rate of GABA currents and drug modulation, as we observed previously in QT-6 cells (Chen et al., 2000). However, two-electrode voltage clamp of oocytes is not the best system for measuring kinetics accurately, and this will be studied in other cells in our future work. Our working hypothesis is that when low levels of endogenous GABARAP are present in a cell, one can see a stimulation of receptor-surface levels; in addition, there may be a change in properties of the existing surface receptors (e.g., caused by clustering), and this can be detected with proper tools. Such GABARAP-dependent effects on receptor properties may be seen more readily when the levels of GABARAP are higher than the minimum requirement for facilitating intracellular trafficking.
At the same time as the work described here, we also demonstrated that overexpression of GABARAP increases the surface numbers of GABAA receptor in both cultured hippocampal neurons and COS7 cells in immunofluorescence and flow-cytometry studies (Leil et al., 2004). Mutagenesis of the key amino acids of GABARAP in the 2-binding domain prevent the stimulation of GABARAP (Leil et al., 2004). However, no electrophysiology or protein quantification was done in that study as was done in the current work. In addition, a peptide derived from the intracellular loop of GABAA receptor 2 that blocks the binding of GABARAP prevents GABARAP effects on GABAR trafficking and clustering in QT-6 cells (Nymann-Andersen et al., 2002b). Endogenous GABARAP has been detected by immunostaining in primary cultured neurons, even in QT-6 cells and L929 cells, although in low amounts (Chen et al., 2000; Everitt et al., 2004; Leil et al., 2004). In our Western blot experiments, we did not detect any endogenous GABARAP in oocytes (data not shown).
The 2 subunits play an essential role for translocation of other subunits to the cell surface and synaptic localization of GABAA receptors. Taking advantage of the 2 subunit knockout mouse, Essrich et al. (1998) reported that the receptor is functionally impaired as indicated by reduced channel conductance and a perinatally lethal phenotype in the majority of mutant mice, although the absence of 2 subunits did not cause a change in expression levels, regional distribution, or membrane localization of the other subunits. Our results demonstrate that first, the GABA currents induced in oocytes expressing 12 subunits are dramatically lower than those with 122S subunit expression. This phenomenon was observed by other groups (Connolly et al., 1999; Haas and Macdonald, 1999). Second, without the 2 subunit, the enhancement of GABARAP was abolished. This suggests that the 2 subunit is required for the enhancement by GABARAP.
Besides GABARAP, other newly cloned proteins also demonstrate trafficking effects on GABAA receptors through interaction with the M3eCM4 intracellular loop. These include the Golgi-specific zinc-finger protein GODZ with 2 subunits (Keller et al., 2004); the ubiquitin-like protein Plic-1 with 1 to 3, 6, and 1 to 3 subunits of GABAA receptors but not 2L or subunits (Bedford et al., 2001); GABAA receptor interacting factor-1 with the 2 (Beck et al., 2002; Kittler et al., 2004); and phospholipase C-related inactive protein type 1 with subunits, possibly involving GABARAP (Terunuma et al., 2004).
In this report, a two-electrode voltage-clamp recording gave the first functional assay verifying that the microtubule-GABARAP interaction plays an important role in facilitating GABAA receptor function. We observed that the GABARAP-increased GABA currents were prevented by nocodazole treatment. A dramatic decrease was observed only in the GABARAP coexpression group in which the enhancement of GABARAP was totally abolished. It is feasible that oocytes express sufficient quantities of machinery, including microtubules for formation of receptor clusters, because disruption of microtubules has been shown to alter clustering and properties of GABAA receptors expressed in oocytes (Whatley et al., 1994). Accumulated evidence indicates that the cytoskeleton is essential for clustering GABAA receptors (Chen et al., 2000; Ho et al., 2001; Petrini et al., 2003). It reminds us that a reduction in clustering state or misdistribution of GABAA receptors by nocodazole or colchicine, drugs disrupting the polymerization of microtubules, might work by interrupting the interaction between polymerized microtubules and receptor-associated proteins, such as GABARAP and gephyrin. Similar to our results, in primary cultured neurons, the disruption of microtubules has been noted by Petrini et al. (2003) to decrease GABAA receptor clustering and alter the channel-opening rate. The crystal structure and GST pull-down studies using truncated GABARAP demonstrate the N terminus (a.a. 1eC22) is responsible for microtubule interaction (Nymann-Andersen et al., 2002b,c). Our data further confirmed that interaction with microtubules is essential for GABARAP function. Deletion of the microtubule-binding domain on GABARAP abolished both the enhancement of GABA currents and the increased surface level of GABAA receptors in oocytes. This indicates that this GABARAP-mediated increase of GABAA receptor cell-surface level involves transportation along microtubules.
There are several candidate mechanisms underlying this enhancement. Given that GABARAP also interacts with the vesicle fusion protein N-ethylmaleimideeCsensitive factor and the known functions of GABARAP's homologs (e.g., Golgi-associated ATPase enhancer of 16 kDa is believed to participate in intra-Golgi transportation machinery) (Sagiv et al., 2000; Muller et al., 2002), we predict that GABARAP might increase the trafficking of GABAA receptors to the cell surface. GABARAP might also stabilize the cell-surface level of GABAA receptors by decreasing endocytosis or degradation. Future study identifying the trafficking rate or half-life of surface-expressed GABAA receptors in the presence of GABARAP will help to define possible effects on endocytosis and the mechanisms of regulation of GABARAP's role in receptor trafficking and plasticity.
Acknowledgements
We thank Dr. Diane Papazian for reagents.
Supported by National Institutes of Health grants NS28772 and NS35985 (to R.W.O.).
doi:10.1124/mol.104.009878.
References
Beck M, Brickley K, Wilkinson HL, Sharma S, Smith M, Chazot PL, Pollard S, and Stephenson FA (2002) Identification, molecular cloning and characterization of a novel GABAA receptor-associated protein, GRIF-1. J Biol Chem 277: 30079eC30090.
Bedford FK, Kittler JT, Muller E, Thomas P, Uren JM, Merlo D, Wisden W, Triller A, Smart TG, and Moss SJ (2001) GABAA receptor cell surface number and subunit stability are regulated by the ubiquitin-like protein Plic-1. Nat Neurosci 4: 908eC916.
Chang CS, Olcese R, and Olsen RW (2003) A single M1 residue in the 2 subunit alters channel gating of GABAA receptor in anesthetic modulation and direct activation. J Biol Chem 278: 42821eC42828.
Chen L, Wang H, Vicini S, and Olsen RW (2000) The gamma-aminobutyric acid type A (GABAA) receptor-associated protein (GABARAP) promotes GABAA receptor clustering and modulates the channel kinetics. Proc Natl Acad Sci USA 97: 11557eC11562.
Connolly CN, Kittler JT, Thomas P, Uren JM, Brandon NJ, Smart TG, and Moss SJ (1999) Cell surface stability of -aminobutyric acid type A receptors. Dependence on protein kinase C activity and subunit composition. J Biol Chem 274: 36565eC36572.
Coyle JE, Qamar S, Rajashankar KR, and Nikolov DB (2002) Structure of GABARAP in two conformations: implications for GABAA receptor localization and tubulin binding. Neuron 33: 63eC74.
Essrich C, Lorez M, Benson JA, Fritschy JM, and Luscher B (1998) Postsynaptic clustering of major GABAA receptor subtypes requires the gamma 2 subunit and gephyrin. Nat Neurosci 1: 563eC571.
Everitt AB, Luu T, Cromer B, Tierney ML, Birnir B, Olsen RW, and Gage PW (2004) Conductance of recombinant GABAA channels is increased in cells co-expressing GABAA receptor-associated protein. J Biol Chem 279: 21701eC21706.
Haas KF and Macdonald RL (1999) GABAA receptor subunit gamma2 and delta subtypes confer unique kinetic properties on recombinant GABAA receptor currents in mouse fibroblasts. J Physiol 514 (Pt 1): 27eC45.
Ho WH, Wang SM, and Yin HS (2001) Regulation of the subcellular distribution and gene expression of GABAA receptor by microtubules and microfilaments in cultured brain neurons. J Cell Biochem 83: 291eC303.
Keller CA, Yuan X, Panzanelli P, Martin ML, Alldred M, Sassoe-Pognetto M, and Luscher B (2004) The gamma2 subunit of GABAA receptors is a substrate for palmitoylation by GODZ. J Neurosci 24: 5881eC5891.
Kittler JT, Rostaing P, Schiavo G, Fritschy JM, Olsen R, Triller A, and Moss SJ (2001) The subcellular distribution of GABARAP and its ability to interact with NSF suggest a role for this protein in the intracellular transport of GABAA receptors. Mol Cell Neurosci 18: 13eC25.
Kittler JT, Thomas P, Tretter V, Bogdanov YD, Haucke V, Smart TG, and Moss SJ (2004) Huntingtin-associated protein 1 regulates inhibitory synaptic transmission by modulating -aminobutyric acid type A receptor membrane trafficking. Proc Natl Acad Sci USA 101: 12736eC12741.
Kneussel M, Haverkamp S, Fuhrmann JC, Wang H, Wassle H, Olsen RW, and Betz H (2000) The -aminobutyric acid type A receptor (GABAA R)-associated protein GABARAP interacts with gephyrin but is not involved in receptor anchoring at the synapse. Proc Natl Acad Sci USA 97: 8594eC8599.
Leil TA, Chen ZW, Chang CS, and Olsen RW (2004) GABAA receptor-associated protein traffics GABAA receptors to the plasma membrane in neurons. J Neurosci 24: 11429eC11438.
Macdonald RL and Olsen RW (1994) GABAA receptor channels. Annu Rev Neurosci 17: 569eC602.
Muller JM, Shorter J, Newman R, Deinhardt K, Sagiv Y, Elazar Z, Warren G, and Shima DT (2002) Sequential SNARE disassembly and GATE-16-GOS-28 complex assembly mediated by distinct NSF activities drives Golgi membrane fusion. J Cell Biol 157: 1161eC1173.
Nymann-Andersen J, Sawyer GW, and Olsen RW (2002a) Interaction between GABAA receptor subunit intracellular loops: implications for higher order complex formation. J Neurochem 83: 1164eC1171.
Nymann-Andersen J, Wang H, Chen L, Kittler JT, Moss SJ, and Olsen RW (2002b) Subunit specificity and interaction domain between GABAA receptor-associated protein (GABARAP) and GABA(A) receptors. J Neurochem 80: 815eC823.
Nymann-Andersen J, Wang H, and Olsen RW (2002c) Biochemical identification of the binding domain in the GABAA receptor-associated protein (GABARAP) mediating dimer formation. Neuropharmacology 43: 476eC481.
Petrini EM, Zacchi P, Barberis A, Mozrzymas JW, and Cherubini E (2003) Declusterization of GABAA receptors affects the kinetic properties of GABAergic currents in cultured hippocampal neurons. J Biol Chem 278: 16271eC16279.
Sagiv Y, Legesse-Miller A, Porat A, and Elazar Z (2000) GATE-16, a membrane transport modulator, interacts with NSF and the Golgi v-SNARE GOS-28. EMBO (Eur Mol Biol Organ) J 19: 1494eC1504.
Sassoe-Pognetto M, Kirsch J, Grunert U, Greferath U, Fritschy JM, Mohler H, Betz H, and Wassle H (1995) Colocalization of gephyrin and GABAA-receptor subunits in the rat retina. J Comp Neurol 357: 1eC14.
Stefani E and Bezanilla F (1998) Cut-open oocyte voltage-clamp technique. Methods Enzymol 293: 300eC318.
Terunuma M, Jang IS, Ha SH, Kittler JT, Kanematsu T, Jovanovic JN, Nakayama KI, Akaike N, Ryu SH, Moss SJ, et al. (2004) GABAA receptor phospho-dependent modulation is regulated by phospholipase C-related inactive protein type 1, a novel protein phosphatase 1 anchoring protein. J Neurosci 24: 7074eC7084.
Wang H, Bedford FK, Brandon NJ, Moss SJ, and Olsen RW (1999) GABA(A)-receptor-associated protein links GABAA receptors and the cytoskeleton. Nature (Lond) 397: 69eC72.
Wang H and Olsen RW (2000) Binding of the GABAA receptor-associated protein (GABARAP) to microtubules and microfilaments suggests involvement of the cytoskeleton in GABARAP-GABAA receptor interaction. J Neurochem 75: 644eC655.
Whatley VJ, Mihic SJ, Allan AM, McQuilkin SJ, and Harris RA (1994) -Aminobutyric acid A receptor function is inhibited by microtubule depolymerization. J Biol Chem 269: 19546eC19552.