Arrestin Is Required for Agonist-induced Trafficking of Voltage-dependent Calcium Channels*

Many metabotropic receptors in the nervous system act through signaling pathways that result in the inhibition of voltage-dependent calcium channels. Our previous findings showed that activation of seven-transmembrane receptors results in the internalization of calcium channels. This internalization takes place within a few seconds, raising the question of whether the endocytic machinery is in close proximity to the calcium channel to cause such rapid internalization. Here we show that voltage-dependent calcium channels are pre-associated with arrestin, a protein known to play a role in receptor trafficking. Upon GABAB receptor activation, receptors are recruited to the arrestin-channel complex and internalized. β-Arrestin 1 selectively binds to the SNARE-binding region of the calcium channel. Peptides containing the arrestin-binding site of the channel disrupt agonist-induced channel internalization. Taken together these data suggest a novel neuronal role for arrestin.

Inhibition of voltage-dependent calcium channels by seventransmembrane receptors (7TMR) 2 is one of the primary means of regulation of calcium-dependent physiological processes such as synaptic transmission, muscle contraction, and membrane excitability. In neurons, the Ca v 2.2 (N-type) channel is a prominent target for G protein-mediated modulation (1,2). Inhibition of Ca v 2.2 channels can be voltage-dependent, and mediated by direct interactions with G protein ␤-␥ subunits (3,4). In addition, kinases such as protein kinase C and tyrosine kinases have been shown to inhibit Ca v 2.2 channels in a voltage-independent manner (5,6). Additional mechanisms may exist by which Ca 2ϩ influx is regulated. Dunlap and Fischbach (7) have suggested that transmitter-mediated shortening of the duration of the action potential could be due to a decrease in the number of voltage-dependent calcium channels at the membrane. Recently we have reported an additional mechanism by which 7TMRs can regulate neuronal calcium levels that involves a rapid internalization of voltage-dependent calcium channels into clathrin-coated vesicles upon receptor activation (8). Here we demonstrate that ␤-arrestin 1 is associated with Ca v 2.2 channels and that activation of 7TMRs results in the formation of an arrestin-receptor-channel complex. This interaction is required for internalization of calcium channels and plays a role in the modulation of calcium current.
Peptides-Sequences of the fluoresceinated 894 -929 and 920 -944 peptides used in this study were based on Ca v 2.2 ␣ 1 sequence from chick dorsal root ganglion (DRG) neuron (CDB1, GenBank TM AAD51815). Peptides were synthesized by FastMoc chemistry at the Tufts University Core Facility (Boston, MA) and purified by high performance liquid chromatography with Ͼ97% purity as determined by mass spectrometry. The N terminus included the sequence of the penetratin domain of the Drosophila protein Antennapedia. Peptides were dissolved in 5 mM acetic acid at 1 mg/ml and diluted into HEPES-buffered saline (0.01 M HEPES, pH 7.4, and 0.15 M NaCl) for biochemical experiments.
The fluoresceinated peptides that show no significant homology with other proteins was detected as tested by BLAST search. Control experiments were performed and no differences were detected between cells loaded with the peptide and unloaded cells. For each peptide used in our studies we performed time course and concentration-response experiments to determine optimal experimental conditions. Pilot studies were conducted in which fluoresceinated peptides were used to assess peptide entry into the cells.
Cell Culture-Embryonic chick sensory neurons were grown in culture as previously described (6).
Transmitter Application-Agonist was prepared fresh in HBS Ca 2ϩ external buffer (2.5 mM KCl) at 100 mM concentration (ϫ1000) (Ϯ) baclofen (4-amino-3-(4-chlorophenyl)butanoic acid) (Sigma). Stock solution was diluted in the appropriate HBS Ca 2ϩ external buffer immediately prior to experiments. Cells were washed once with HBS Ca 2ϩ external buffer (2.5 mM KCl) at room temperature followed by the addition of 2 ml of HBS Ca 2ϩ external buffer (60 mM KCl), with or without a final concentration of 100 M baclofen for 20 s or 5 min at room temperature.
Immunohistochemistry-Cultures grown on poly-L-lysine glass coverslips were fixed and permeabilized in methanol at Ϫ20°C for 15 min followed by three 5-min washes in PHEM buffer (60 mM PIPES, 25 mM HEPES, 10 mM EGTA, 2 mM MgCl 2 , pH 6.9). Blocking was performed using 5% bovine serum albumin in PHEM buffer for 1-h at 4°C, and incubation with primary antibody in 1% bovine serum albumin and 1% normal goat sera in PHEM buffer was performed overnight at 4°C. After washes with PHEM buffer, coverslips were incubated with fluorophore-conjugated secondary antibodies in 1% bovine serum albumin PHEM buffer for 1.5 h at room temperature in the dark. Glass coverslips were washed 4 times (5 min each) in 1% bovine serum albumin PHEM buffer and mounted on glass slides with one drop of Vectashield anti-fade reagent (Vector Laboratories, Burlingame, CA) and sealed.
Confocal Imaging-Confocal laser scanning microscopy was performed at the MSSM Microscopy Shared Resource Facility, using a Zeiss Meta510 (UV) microscope with an inverted Axiovert. Images of fixed cells were obtained with a pinhole setting of 1.0 using a UV ϫ63 1.4NA oil objective lens at slow acquisition speed with ϫ4 frame averaging accumulation. The number of sections were calculated by the software based on acquisition of sections at 240-nm intervals in the Z-plane. The confocal microscope settings were kept the same for all scans. All morphometric measurements were done using Metamorph image analysis software (Universal Imaging Corporation, West Chester, PA). Neurons were selected and carefully manually traced for maximum accuracy. The average intensity of fluorescence signal was measured in the traced regions and background staining (determined over neuron-free areas of the culture) was subtracted. Intensity measurements are expressed in arbitrary units of fluorescence per square area.
The integrated density of each optical slice was measured and the total surface and cytoplasmic intensity per pixel were calculated. Membrane and cytoplasmic staining were assessed by integrated density morphometric analysis using Metamorph. We used regions of interest and for every optical slice the whole area was defined as total fluorescence and the interior of the cell as the cytosolic fluorescence. The membrane fluorescence was defined as the difference of total cytosolic. The integrated values were determined by measuring the fluorescence values as a function of area.
The plasma membrane was stained with 1 M FM4-64X, a form of the FM4-64 dye that can be used in fixed cells (Invitrogen). Line scans of intensity profiles across the cells were generated with Metamorph (Universal Imaging). We measured the fluorescence intensity over a distance covering the membrane and the cytosol. Three line profiles, avoiding the nucleus, were performed to obtain an average profile of fluorescence intensity for each cell.
For co-localization of two different proteins of interest, pictures, usually green in one case and red in the other were merged and co-localized puncta, which appear yellow, were counted for each cell. For the measurement of the degree of co-localization, the correlation coefficient (Pearson coefficient) between the two different signals was calculated using Metamorph. For each experiment random groups of cells were scored for individual puncta and overlapping puncta of two proteins of interest in matched pairs per cell with a minimum of 25 cells scored per experiment and conditions for manual counting and 10 cells per experiment and conditions for automated counting. Imaging analysis was performed in a doubleblind fashion. Cells were given a code and analyzed in a random manner.
Statistical analysis was performed using Student's t-tests, or one-way analysis of variance as appropriate. Statistical differences of p Ͻ 0.05 were considered significant.
Living Neuron Experiments-Biotinylated -conotoxin GVIA (Bachem, Torrance, CA) was incubated with Quantum dot 655-conjugated streptavidin (Invitrogen) for 5 min, then 200 l of complete medium was added and the solution was incubated at 30°C for 20 min. DRG neurons grown on glass bottom culture dishes were preincubated with 120 nM Quantum dot 655-labeled -conotoxin GVIA for 1 h in DRG medium at 30°C in a CO 2 incubator to minimize ligand-bound channel re-uptake. Cells were washed twice with 1 mM Ca 2ϩ external buffer (1 mM CaCl 2 , 133 mM NaCl, 0.8 mM MgCl 2 , 10 mM tetraethylammonium-Cl Ϫ , 25 mM HEPES, 12.5 mM NaOH, 5 mM dextrose, 0.3 mM tetrodotoxin (Calbiochem, La Jolla, CA)) to remove unbound conotoxin. Specificity of biotinylated conotoxin GVIA-Quantum dot-streptavidin conjugate labeling of Ca v 2.2 channels was determined by preincubating cells with unlabeled -conotoxin GVIA (750 nM) and biotinylated -conotoxin GVIA-Quantum dot 655-streptavidin conjugate (120 nM) for 3 h at 30°C in a CO 2 incubator prior to confocal imaging. Regions of interest were scanned at high acquisition speed at 2-s intervals in one X-Y focal plane with the appropriate stage head configured for the Bioptechs glass bottom culture dishes fitted with a perfusion pump.
For analysis, regions of interest were selected and carefully manually traced for maximum accuracy. The average intensity of fluorescence signal in the traced regions was measured using Physiology version 3.2 software (Zeiss) and background staining (determined over neuron-free areas of the culture) was subtracted. Intensity measurements are expressed in arbitrary units of fluorescence per square area.

R E T R A C T E D R E T R A C T E D
the antibodies to allow subsequent identification using epifluorescence optics. Control experiments were performed by injecting vehicle, preimmune serum or rabbit IgG-containing internal solution. Confocal microscopy was performed 30 min after injection.
Electrophysiology-Whole cell recordings were performed as described in Ref. 6. For extracellular application, agents were diluted into standard extracellular saline and applied via a widebore pipette. For the experiments presented in this report, calcium current has been corrected for rundown by measuring calcium current as a function of time in control cells without transmitter. Cells used for experiments exhibited a rundown of the current of less than 1%/min.
Pulldown-2 mg of rat brain lysate was incubated with 100 g of His 6 -tagged recombinant protein bound to nickel beads for 4 h at 4°C. The beads were spun down and washed 3 times. Beads were mixed with 25 l of Laemmli sample buffer and boiled for 5 min. After spinning down the beads, the supernatant sample was resolved by 7.5% SDS-acrylamide gel. Immunodetection of arrestin was carried out using anti-arrestin (Chemicon, 1:1000).
To test interactions of recombinant arrestin with the calcium channel, 2 g of His 6 -tagged recombinant protein containing the channel sequence was mixed and incubated with 2 g of arrestin for 30 min at room temperature. Nickel beads were used to pull down; immunodetection of ␤-arrestin 1 or 2 was carried out using antibodies provided by the Lefkowitz laboratory (1:1000).

Pre-association of the Calcium Channel with Arrestin-Im-
aging experiments in live embryonic chick DRG neurons have shown that, within seconds of receptor activation, calcium channels are cleared from the membrane and sequestered in clathrin-coated vesicles (8). The fast kinetics of internalization of calcium channels raises the question of whether components of the exocytic machinery exist in close proximity or pre-associated with the calcium channel. Whereas there is a high degree of co-localization between calcium channels and the endosomal markers Rab5 and clathrin heavy chain in cells exposed to agonist, there is a very low level of co-localization prior to receptor activation (8).
By analogy to 7TMR trafficking, we sought to determine whether arrestin associates with internalized calcium channels. Both ␤-arrestin 1 and 2 are expressed in DRG neurons as determined in immunoblotting experiments (Fig. 1a). We next determined whether calcium channels were spatially distrib-uted in close proximity to arrestin. In the presence of saline, most of the calcium channels are localized in the plasma membrane as the channels co-localize with the plasma membrane marker FM4-64 (Fig. 1b). Optical slices from saline-treated neurons show that both calcium channels and arrestin co-localize as indicated by the yellow signal indicating overlap of green fluorescent signal (Ca v 2.2 channels) and red fluorescent signal (arrestin). Both proteins are associated with the top slices; in the middle slices the fluorescence signal forms a ring around the periphery of the cell suggesting association with the membrane (Fig. 1, c and e, and supplementary materials Fig.  S1). Upon exposure to agonist for 20 s, the fluorescence signal becomes more intense in the middle slices and a decrease is observed in the levels of membrane-associated arrestin and Ca v 2.2 channel (Fig. 1d and e, and supplementary materials Fig.  S2). The Pearson correlation coefficient between calcium channels and arrestin is 0.73 Ϯ 0.09 in saline-treated cells and 0.8 Ϯ 0.06 in baclofen-treated cells (n ϭ 25). Together these results suggest that arrestin and Ca v 2.2 channels are preassociated in the cell surface and are internalized upon 7TMR activation.
Arrestin co-precipitates with Ca v 2.2 channel protein from chick DRG neurons treated with saline or baclofen (Fig. 2a) giving further support to the observation that arrestin is preassociated with the calcium channel in an agonist-independent fashion. We have previously used anti-pan-␣ 1 antibody to precipitate Ca v 2.2 channels from DRG neurons (9,25). Arrestin immunoprecipitates from cultured embryonic chick DRG neurons and rat brain lysates contained Ca v 2.2 channel protein (Fig. 2b). Moreover, the detection of arrestin/Ca v 2.2 channel association in rat brain lysates suggests that this interaction takes place in central nervous system neurons.
Association of arrestin with calcium channels might occur proximal to release sites as suggested by the high degree of correlation between the localization of arrestin and that of synapsin I, a synaptic vesicle protein (Fig. 2, c and d). 83% of the arrestin co-localize with synapsin (n ϭ 20, Fig. 2d).
Ca v 2.2 Channel-Arrestin-Receptor Complexes Are Formed during Calcium Channel Internalization-Because arrestin is known to interact with phosphorylated 7TMRs (10, 11), we determined whether arrestin and the Ca v 2.2 channel are part of a complex that contains the receptor. Indirect immunofluorescence using confocal laser microscopy was used to visualize the GABABR1 subunit of the GABA B receptor (blue signal), Ca v 2.2 channel (green signal), and arrestin (red signal). X-Y optical slices were taken from the top to bottom of DRG neurons treated with saline or baclofen. Very low co-localization was detected between the receptor and arrestin or Ca v 2.2 channel in saline-treated neurons ( Fig. 3a and supplementary materials  Fig. S3). The Ca v 2.2 channel-arrestin correlation coefficient is 0.70; the other values for co-localization are below 0.10 (n ϭ 20, Fig. 3e).
Upon a 20-s exposure to baclofen, DRG neurons exhibited an increase in the degree of co-localization between Ca v 2.2 channels, arrestin, and receptors as indicated by the white fluorescence signal (Fig. 3, b and e, and supplementary materials Fig.  S4). Most of the co-localization takes place in the middle slices suggesting that the signal is cytoplasmic and that these proteins are internalized together. The co-localization of the Ca v 2.2 OCTOBER , second panel) and calcium channels were detected by indirect immunofluorescence using anti-pan-␣ 1 antibody followed by Oregon Green-conjugated anti-rabbit IgG (green signal, third panel). Right panel shows a merged image (yellow signal). X-Y optical slices were acquired at 0.2-m intervals from the top to bottom of each cell. A middle optical slice is shown. Scale bar represents 10 m. Data are representative of three independent experiments. Line scans of intensity profiles across the cells were generated with Metamorph (Universal Imaging). We measured the fluorescence intensity over a distance covering the membrane and cytosol. Three line profiles, avoiding the nucleus, were performed to obtain an average profile of fluorescence intensity for each cell. Green lines represent the intensity profile for the calcium channel staining and the red lines represent the intensity profile for the FM4-64 signal. c and d, calcium channels and ␤-arrestin were detected by indirect immunofluorescence using anti-pan-␣ 1 antibody followed by Oregon Green-conjugated anti-rabbit IgG and anti-␤-arrestin 1 followed by Cy3 anti-mouse IgG, respectively. A series of merged images (Ca v 2.2 channel/arrestin) of X-Y optical slices acquired at 0.2-m intervals from the top to bottom of each cell is shown. Cells were treated with saline (c) or 100 M baclofen (d) for 20 s. Scale bar represents 10 m. Data are representative of seven independent experiments. e, histogram showing integrated fluorescence values for the merged signal for the calcium channel and arrestin associated in the membrane (M) and cytoplasm (C ). Error bars represent mean Ϯ S.E. and analysis between independent matched groups (matched by saline alone or agonist) was significant at: *, p Ͻ 0.001 and **, p Ͻ 0.0001.

R E T R A C T E D R E T R A C T E D M a y 1 3 , 2 0 1 1 M a y 1 3 , 2 0 1 1
channel-arrestin-receptor is transient, as the analysis of neurons exposed to agonist for 1 or 5 min revealed a very low degree of co-localization (Fig. 3, c, d, and e). Our results show that co-localization of the receptor with arrestin and the calcium channel is highest at 20 s, a time point in which inhibition of calcium channels by 7TMRs is still at its maximum (12). The degree of co-localization between the proteins is lower at 1 and 5 min, which parallels the desensitization of transmitter-mediated inhibition of calcium current.
Activation of GABA B receptors increases the amount of receptor that co-precipitates with Ca v 2.2 channel (Fig. 4). When the presence of the GABABR2 was probed by immuno-blotting of calcium channel precipitates a band in the 130-kDa region was observed in agreement to published reports (13). Exposure of neurons to agonist resulted in a 4-fold increase in the amount of receptor detected in the immunoprecipitates (Fig. 4). When membranes were probed for arrestin, no change in the amount of arrestin associated with the channel was observed, which is in agreement with the results shown in Fig.  2a. The biochemical and imaging data suggest that receptorarrestin-channel complexes are formed upon 7TMR activation and that these complexes are internalized.
Mapping of Arrestin-Ca v 2.2 Channel Interaction-The Ca v 2.2 channel/arrestin interaction was further supported by FIGURE 2. Calcium channels associate with arrestin. a, calcium channels from primary cultures of DRG neurons treated with saline or baclofen for 20 s were precipitated using anti-pan-␣ 1 antibody and immunoblotted for arrestin. b, arrestin was precipitated from DRG neuron or rat brain lysate using anti-arrestin and immunoblotted for calcium channel using anti-␤1B antibody. c, arrestin co-localizes with synaptic proteins. Co-distribution of arrestin (blue) with synapsin (green) was detected by indirect immunofluorescence in DRG neurons under saline conditions. Confocal images of top X-Y optical slices show arrestin, synapsin, and merged arrestin/synapsin distribution. Scale bar represents 10 m. Data are representative of three independent experiments. d, quantitation of co-localization between arrestin, synapsin, and calcium channels in a population of 20 independent cells. For the measurement of the degree of co-localization, the correlation coefficients (Pearson coefficients) between the arrestin (blue signal) and synapsin (green signal) and calcium channels (red signal, not shown) were calculated using Metamorph. Error bars represent S.D. OCTOBER

R E T R A C T E D R E T R A C T E D
results from experiments in which His 6 -tagged recombinant proteins containing sequences from the cytoplasmic regions of the ␣ 1 subunit of Ca v 2.2 channels were used in arrestin pulldown assays from rat brain lysate. Recombinant protein containing the SNARE-binding or synprint region from this loop II-III region (amino acids (aa) 726 -984) bound to arrestin (Fig.  5a), whereas no binding was detected in samples incubated with the remainder of loop II-III or the C terminus. Truncation of the synprint domain protein showed that aa 894 -944 are required for binding to arrestin (Fig. 5b).

R E T R A C T E D R E T R A C T E D
Recombinant ␤-arrestin 1 binds to recombinant protein containing the sequence of the synprint region of the Ca v 2.2 channel suggesting a direct interaction between the two proteins ( Fig. 5c). This interaction is selective for ␤-arrestin 1, as no binding to the synprint region of the calcium channels was detected when recombinant ␤-arrestin 2 was used in the experiments (Fig. 5d).
␤-Arrestin 1 Is Required for Agonist-mediated Internalization of Calcium Channels-We tested whether arrestin plays a role in agonist-induced internalization of voltage-dependent calcium channels. Calcium channels at the cell surface were labeled using biotinylated -conotoxin GVIA bound to streptavidin-conjugated quantum dots. Chick DRG neurons express only one type of calcium channel, Ca v 2.2 channel (N type) (14), which is located both at the terminals and the soma. Channels at both locations are coupled to the exocytic machinery (15,16). Ca v 2.2 channels exhibited a punctate distribution in the top plane of the membrane (Fig. 6, a and  b). Preincubation of DRG neurons with unlabeled toxin prevents the binding of Qdot 655-streptavidin-biotinylated -conotoxin GVIA demonstrating that the probe binds selectively (supplementary materials Fig. S5). Incubation of DRG neurons with the Qdot 655-conjugated streptavidin alone did not result in significant labeling of the surface of DRG neurons (supplementary materials Fig. S5). These results are not qualitatively different from those that we have previously obtained using rhodamine-conjugated -conotoxin GVIA (8).
In live cell imaging experiments, exposure of DRG neurons to baclofen, a GABA B receptor agonist (100 M) and well established modulator of these channels, produced a decrease in fluorescence signal in the top surface of the cell within 2 s (Ref. 8, Calcium channels were precipitated (IP) from DRG neurons treated with saline or baclofen for 20 s using anti-pan-␣ 1 antibody and immunoblotted (IB) for GABA B receptor (GABABR2). The membrane was stripped and probed for arrestin. Histogram shows quantitation of the density of the GABABR2 band from 5 independent experiments. Error bars represent mean Ϯ S.E. and analysis between independent matched groups (matched by saline alone or agonist) was significant at: *, p Ͻ 0.05. FIGURE 5. Arrestin binds to the SNARE-binding region of the calcium channel. a, rat brain lysate was incubated with His 6 -tagged recombinant proteins containing sequences from the cytoplasmic regions of the ␣ 1 subunit of the Ca v 2.2 channel. Precipitation was performed using nickel beads. Immunoblotting was performed for arrestin. Data are representative of four independent experiments. b, interaction of arrestin with loop II-III of calcium channel. Rat brain lysate was incubated with His 6 -tagged recombinant proteins with sequences spanning regions of loop II-III. Precipitation was performed using nickel beads. Immunoblotting was performed for arrestin. Data are representative of four independent experiments. c, direct interaction of arrestin with aa 894 -944 of the calcium channel. Recombinant ␤-arrestin 1 was incubated with His 6 -tagged recombinant proteins containing aa 728 -795 or 894 -944 of the calcium channel. Precipitation was performed using nickel beads. Immunoblotting was performed for arrestin. d, precipitation was performed using recombinant ␤-arrestin 1 or ␤-arrestin 2. OCTOBER Fig. 6a) concurrent with an increase in fluorescence in the middle optical slices (Fig. 6, a and d). The decrease in fluorescence in the top surface was transient as the signal reappeared at the cell surface within several minutes of continued agonist presence (Fig. 6a). After channels reappeared, the agonist was washed away and cells were exposed to saline for 1 min. A second application of agonist failed to elicit any change suggesting that the response had desensitized (Fig. 6c). No change in fluorescence was observed in cells exposed to saline (Fig. 6b).

R E T R A C T E D R E T R A C T E D
We designed cell-permeant peptides based on the sequence of the arrestin-binding site in the Ca v 2.2 channels (Gallus gallus CDB1, aa 894 -944) to interfere with arrestin/channel interaction in DRG neurons. Two peptides were designed: one containing aa 894 -929 and a second one containing aa 920 -944. The N termini of the peptides included the sequence of the penetratin domain of the Drosophila protein Antennapedia and the C termini contained fluoresceinated amino acids to visualize the peptides. Entry into the cells of the cell-permeant peptides was monitored by fluorescence and 80% of the cells contained peptide following a 5-min incubation. The 894 -929 peptide prevented agonist-induced Ca v 2.2 channel internalization without altering their basal distribution (Fig. 6e). The scrambled peptide (data not shown) and 920 -944 peptide were without effect (Fig. 6e).
To determine whether there is selectivity in which ␤-arrestin is involved in channel trafficking, we injected DRG neurons with antibodies that preferentially recognize ␤-arrestin 1 or ␤-arrestin 2. The antibodies were introduced by microinjection into the cell body and fluorescein-dextran was included in the solution to visualize the injected cells. Calcium channels were visualized with Qdot 655-streptavidin and biotinylated -conotoxin GVIA. Neurons injected with anti-␤-arrestin 1 showed a significant decrease in baclofen-induced internalization of Ca v 2.2 channels; changes observed in the presence of anti-␤-arrestin 2 were not significant (Fig. 6f ). Taken together, these results suggest that ␤-arrestin 1 is required for agonistinduced internalization of calcium channels. Cells injected with anti-␤-arrestin 1 showed a decrease in agonist-mediated voltage-independent inhibition (Fig. 6g).

DISCUSSION
We have shown that, in both embryonic chick DRG neurons and rat brain, ␤-arrestin 1 is bound to Ca v 2.2 channels in the absence of receptor activation. This is in contrast to previous reports of arrestin-binding proteins that showed a requirement for receptor activation for the nucleation of signaling complexes. Unlike the interaction of arrestin with receptors (17) and NHE5 (18), the interaction of calcium channels with arrestin does not require phosphorylation, as channel phosphorylation is undetectable prior to activation of GABA B receptors (9). Furthermore, the arrestin-binding site in the calcium channels does not contain consensus sites for phosphorylation.
The arrestin-binding site in Ca v 2.2 channels is located within loop II-III in the SNARE-binding region, a region important for the regulation of both channel activity and secretion (19). Binding of syntaxin to this channel region plays a role in voltagedependent inhibition by stabilizing the binding of G protein ␤-␥ subunits to the channel (20). Synaptotagmin also binds to the channel in this region upon depolarization, in a calcium-dependent manner (21). Deletion of the SNARE-binding region (rat 726 -984) from loop II-III results in mistargetting of Ca v 2.1 channels (22); future studies should address whether this is due to long-term disruption of arrestin/calcium channel interaction. It should be noted that the arrestin-binding region, aa 894 -929, is highly conserved (63% homology) among voltagedependent calcium channels further suggesting that arrestin/ channel interaction will be found in different regions of the brain throughout different species.
Arrestin-channel-receptor complexes were detected at early time points (20 s) in the time course of the agonist-induced response. Our previous electrophysiological studies of calcium channel modulation have shown that, at this time point, calcium channel inhibition is maximal and desensitization is not observable (12). Channel internalization was not observed in experimental paradigms that favored desensitization such as long-term exposure to agonist (Ͼ40 s, Fig. 6, a and d) or repetitive exposure to agonist (Fig. 6c). Furthermore, in experiments in which neurons were exposed to agonist for 1 min, a time point in which calcium channels return to the surface and desensitization of 7TMR-induced calcium channel inhibition have been observed, there is no significant receptor/arrestin co-localization and receptors have returned to the surface. The formation of these complexes seems to be important for the suppression of neuronal excitability through the inhibition of FIGURE 6. ␤-Arrestin 1/calcium channel interaction is required for channel trafficking. a and b, time-lapse confocal images of live DRG neurons. Images from the top and middle planes of the cell were acquired at 2-s intervals (10-s interval images are shown here for clarity); calcium channels were visualized with Quantum dot 655-conjugated streptavidin bound to biotinylated -conotoxin GVIA. Neurons were exposed to 100 M baclofen, a GABA B receptor agonist (a) or saline (b). Negative time values represent time points prior to addition of agonist. Images are representative of 20 independent experiments; each experiment consisting of a minimum of 4 cells. c, confocal images from the top surface of DRG shown in panel a after a second application of baclofen. d, histogram plot of agonist-induced changes in fluorescence in the top and middle optical slices. X-Y optical slices were taken before (t ϭ Ϫ10) and during the application of 100 M baclofen (t ϭ ϩ10). The integrated density of each optical slice was measured and the total surface and cytoplasmic intensity per pixel was calculated. Membrane (M) and cytoplasmic (C ) staining was assessed by integrated morphometric analysis using Metamorph (Universal Imaging) of peripheral staining relative to total cells staining. Data values are the mean from 20 cells and error bars represent mean Ϯ S.E.. Data analysis between independent matched groups (matched by saline alone or agonist) was significant at: *, p Ͻ 0.001. e, cells were incubated with fluoresceinated peptides containing aa 894 -929 or 920 -944. Agonist was added when indicated by the bar and fluorescence in the top slice was measured as a function of time. Values plotted represent the mean from 10 cells. f, ␤-arrestin-1 selectively mediates agonist-induced sequestration of calcium channels. Effect of antibodies raised against ␤-arrestin 1 and ␤-arrestin 2 on baclofen-induced calcium channel sequestration. Antibodies (1:1000) were introduced into the cell bodies by microinjection. The antibodies were injected in the presence of fluoresceinated-dextran to allow subsequent visualization using epifluorescence. Experiments were performed 1 h after injection. Fluorescence as a function of time was measured for the region of interest using Physiology software version 3.2. Each time point represents the mean of 5 cells. g, effect of antibodies raised against ␤-arrestin 1 on agonist-induced voltage-independent inhibition. Antibodies (1:1000) were introduced into the cell bodies by microinjection. The antibodies were injected in the presence of fluoresceinated-dextran to allow subsequent visualization using epifluorescence. Experiments were performed 1 h after injection. Analysis between independent matched groups (matched by saline alone or agonist) was significant at: *, p Ͻ 0.005. OCTOBER