Identification and Characterization of a G Protein-binding Cluster in α7 Nicotinic Acetylcholine Receptors*

Background: α7 nicotinic acetylcholine receptors are ligand-gated ion channels that bind intracellular signaling proteins. Results: A mutation in the M3-M4 loop of the α7 nicotinic acetylcholine receptor abolishes Gαq activation of intracellular calcium stores. Conclusion: α7 nicotinic acetylcholine receptors functionally couple G proteins. Significance: Nicotinic acetylcholine receptors engage metabotropic-signaling responses. α7 nicotinic acetylcholine receptors (nAChRs) play an important role in synaptic transmission and inflammation. In response to ligands, this receptor channel opens to conduct cations into the cell but desensitizes rapidly. In recent studies we show that α7 nAChRs bind signaling proteins such as heterotrimeric GTP-binding proteins (G proteins). Here, we demonstrate that direct coupling of α7 nAChRs to G proteins enables a downstream calcium signaling response that can persist beyond the expected time course of channel activation. This process depends on a G protein-binding cluster (GPBC) in the M3-M4 loop of the receptor. A mutation of the GPBC in the α7 nAChR (α7345–348A) abolishes interaction with Gαq as well as Gβγ while having no effect on receptor synthesis, cell-surface trafficking, or α-bungarotoxin binding. Expression of α7345–348A, however, did significantly attenuate the α7 nAChR-induced Gαq calcium signaling response as evidenced by a decrease in PLC-β activation and IP3R-mediated calcium store release in the presence of the α7 selective agonist choline. Taken together, the data provides new evidence for the existence of a GPBC in nAChRs serving to promote intracellular signaling.

␣7 nicotinic acetylcholine receptors (nAChRs) play an important role in synaptic transmission and inflammation. In response to ligands, this receptor channel opens to conduct cations into the cell but desensitizes rapidly. In recent studies we show that ␣7 nAChRs bind signaling proteins such as heterotrimeric GTP-binding proteins (G proteins). Here, we demonstrate that direct coupling of ␣7 nAChRs to G proteins enables a downstream calcium signaling response that can persist beyond the expected time course of channel activation. This process depends on a G protein-binding cluster (GPBC) in the M3-M4 loop of the receptor. A mutation of the GPBC in the ␣7 nAChR (␣7 345-348A ) abolishes interaction with G␣ q as well as G␤␥ while having no effect on receptor synthesis, cell-surface trafficking, or ␣-bungarotoxin binding. Expression of ␣7 345-348A , however, did significantly attenuate the ␣7 nAChR-induced G␣ q calcium signaling response as evidenced by a decrease in PLC-␤ activation and IP 3 R-mediated calcium store release in the presence of the ␣7 selective agonist choline. Taken together, the data provides new evidence for the existence of a GPBC in nAChRs serving to promote intracellular signaling.
Nicotinic acetylcholine receptors (nAChRs) 2 comprise an important class of the Cys-loop ligand-gated ion channel superfamily, which mediate communication between neurons by conversion of chemical neurotransmitter signals into a transmembrane flux of ions (1). In most cell types, co-expression of ionotropic nAChRs as well as metabotropic muscarinic receptors ensures a fast and slow acetylcholine signaling response, respectively. At least nine different nAChR subunits are expressed in the mammalian brain. In the hippocampus and cortex homomeric ␣7 and heteromeric ␣4␤2 nAChRs have been shown to contribute to neurotransmitter release and dendritic plasticity (2). Upon ligand activation, ␣7 nAChRs conduct cations into the cell (1)(2)(3)(4). However, because the receptor channel desensitizes within milliseconds (5) it is possible that ligand binding can also set into motion longer lived downstream signaling events.
Nicotinic receptors couple to a myriad of signaling, scaffolding, and trafficking proteins in neural and immune cells (4,6). All nAChRs maintain an M3-M4 loop, which varies in length and sequence identity between different subunits (7)(8)(9). The M3-M4 loop contributes to the trafficking and clustering of the nAChR via an association with the cellular cytoskeleton (10). nAChRs are also found to associate with several types of G proteins, which can contribute to cross-talk between nAChRs and G protein-coupled receptors (4,8,11,12).
Studies on the mechanisms of G protein binding to the glycine receptor 1 (GlyR1) reveal the existence of a G proteinbinding cluster (GPBC) within the M3-M4 loop of Cys-loop receptors (8). A mutation at the GPBC in the human GlyR1 is sufficient to abolish G protein binding in HEK 293 cells (13). A protein sequence alignment of the M3-M4 loop reveals a conservation of the GPBC in the ␣7 nAChR. A mutation of these residues in ␣7 nAChRs attenuates interaction with G proteins and reduces the capacity of the receptor to activate phospholipase C (PLC) and inositol triphosphate (IP 3 ) calcium release in response to choline. These findings show a role for G protein coupling in ␣7 nAChR signaling.
Mutagenesis of the GPBC-The Structure-Genetic Matrix was determined based on functional similarity comparisons among amino acid residues within homologus regions of the human ␣1 GlyR subunit (GlyR1) and nAChRs. A percentage similarity to the GlyR1 core GPBC (14) sequence was determined for each nAChR residue in the alignment (13). Percentage similarity was derived from the sum of derived frequencies of amino acid replacements as previously described (17).
For site-directed mutagenesis at amino acid residues 345-348 of the human ␣7 nAChR (RMKR to AAAA) a forward primer 5Ј-ctcctgaactggtgtgcatggtttctggccgcggcggcgcccggagaggacaagg-3Ј and reverse primer 5Ј-ccttgtcctctccgggcgccgccgcggccagaaaccatgcacaccagttcaggag-3Ј were used in a mutagenesis experiment using the QuikChange II XL site-directed mutagenesis kit from Stratagene (CA). Briefly, 10 ng of the DNA template were used in the PCR condition at the following settings: 1 min at 95°C; 50 s at 95°C; 50 s at 62°C; 7 min at 68°C for 18 cycles; and 68°C for 7 min. The product was sequenced using the ABI 310 genetic analyzer (Applied Biosystems).
Preparation of the Soluble Membrane Fraction-Solubilized proteins were obtained from cultured cells using a non-denaturing Lysis Buffer (1% Triton X-100, 137 mM NaCl, 2 mM EDTA, and 20 mM Tris-HCl, pH 8) supplemented with protease and phosphatase inhibitors (Roche, Penzeberg, Germany) as described (4,18). Soluble membrane fractions were harvested from the prefrontal cortex, hippocampus, and striatum of adult C57BL6 mice (4). Mice were anesthetized using 5% isoflurane and all tissue was dissected in an ice-cold dissection buffer (Hanks' balanced salt solution supplemented with 10 mM HEPES) and then placed into 10 volumes of a cold homogenization buffer (0.32 M sucrose, 10 mM HEPES, 2 mM EDTA, pH 7.4) supplemented with protease and phosphatase inhibitor mixtures (Roche). Tissue was homogenized with a glass Dounce homogonizer on ice and then centrifuged at 1000 ϫ g at 4°C to collect the supernatant, which was subsequently centrifuged for 60 min at 200,000 ϫ g. Protein concentrations were determined via a Bradford assay (Thermo Scientific).
Immunoprecipitation, Bungarotoxin Complex (Bgtx C ) Isolation, and Proteomics-Immunoprecipitation (IP) of protein complexes was performed as described (23). Protein G Dynabeads (Life Technologies) were preconjugated with 5 g of the following antibodies (Abs): ␣7 nAChR (C-20) polyclonal (Santa Cruz); anti-G␣ q (New East Bioscience); and anti-G␤ (T-20) (Santa Cruz). 500 g of cell lysate or 250 g of brain tissue was used in each IP experiment. Proteins were eluted in LDS buffer (Life Technologies), separated onto BisTris gradient gels, and transferred onto nitrocellulose membranes (Invitrogen). HRP secondary antibodies were purchased from Jackson Immu-noResearch. Bands were visualized using SuperSignal West Pico Chemiluminescent substrate (Thermo Scientific) via the Gel Doc Imaging system (Bio-Rad). SeeBlue Ladder (Life Technologies) was used as a protein standard. Presented values are based on averages from three independent experiments. Bgtx C isolation was performed using a cyanogen bromideactivated ␣-bungarotoxin-Sepharose matrix as described (24). 1500 g of solubilized membrane proteins was mixed with 100 l of a Sepharose bead matrix overnight at 4°C. Beads were centrifuged at 2,000 ϫ g and unbound proteins were removed by washing in ice-cold TBS. Bgtx C was eluted by a 1 M carbamylcholine chloride solution followed by precipitation with acetone. Proteins were separated by a SDS-PAGE gel and select bands were excised from the Coomassie-stained gel for mass spectrometry (MS) protein analysis using liquid chromatography-electrospray ionization (LC-ESI) (4). Tandem MS collected by Xcalibur (version 2.0.2) was searched against the NCBI protein database using SEQUEST (Bioworks software from ThermoFisher, version 3.3.1). SEQUEST search results were filtered as: minimum X correlation of 1.9, 2.2, and 3.5 for 1ϩ, 2ϩ, and 3ϩ ions, respectively, and ⌬C n Ͼ 0.1. Protein Score (PS) exclusion was set at X correlation Ͻ 0.1.
Calcium Imaging-Cells were transfected with GCaMP5G (25) for detection of rapid calcium dynamics using a Zeiss Axio Observer Z.1 with attached mRM camera using an acquisition rate of 1 frame per 70 ms for 75 s at 2 ϫ 2 binning. Phototoxicity and bleaching were minimized using low-wavelength and neutral density light filters (26). Drugs were applied to the recording chamber via a gravity fed perfusion system at a flow rate of 1 ml/s after 70 frames of baseline recordings. For replacement experiments 1.26 mM barium was added to a calcium-free Hanks' balanced salt solution (Life Technologies). Regions of interest (ROIs) were normalized as ⌬F/F and analyzed using ImageJ (NIH). Area under the curve (AUC) was determined using the formula 1 ⁄ 2(X 1 ϩ X 2 )(time step difference between frames). ROIs were averaged over each condition as shown (18). Eight cells were imaged per experimental condition and all experiments were performed in triplicate to obtain group averages (n ϭ 8).
PLC-␤ Translocation-PIP 2 breakdown is detected quantitatively by imaging the translocation of a pleckstrin homology (PH) domain of PLC (␦) tagged with mCherry (18,27). A PH-mCherry sensor was co-transfected with GCaMP5G into PC12 cells. PH-mCherry fluorescence (excitation ϭ 555 nm) was visualized at an acquisition rate of 1 frame per 10 s for 1 min with 2 ϫ 2 binning. Drugs were applied 10 s after a 1-frame (10 s) baseline recording. PH-mCherry translocation was determined using a 2-point ROI analysis (ROI 1, fluorescence at the plasma membrane (F m ); ROI 2, cytosol fluorescence (F c )). Translocation of PH-mCherry was determined using the equa- (18,27). Fluorescent values were normalized for the area (m 2 ) measured with Image J (NIH). Experiments were repeated in triplicate to obtain group averages (n ϭ 8).
Statistical Analysis-Data were compared by one-way ANOVA and Tukey's HSD post hoc or Student's t test (p Ͻ 0.05). Data are presented as mean Ϯ S.E. of at least three independent experiments.

Results
Proteomic Evidence on the Existence of ␣7 nAChR/G Protein Complexes in the Brain-Studies have revealed associations between nAChRs and G proteins in neural and non-neural cells (4,9,18,23,26,28). We surveyed ␣7 nAChR associations with G proteins in the prefrontal cortex, hippocampus, and striatum of the mouse brain. ␣7 nAChR-specific antibodies were used to co-IP ␣7 along with G proteins from neural membranes fractions. A reverse co-IP experiment was performed using anti-G protein antibodies to detect ␣7 nAChRs in the same membrane fraction. Co-IP experiments were also performed from ␣7 knock-out mice (␣7 Ϫ/Ϫ ) as a negative control (29). As shown in Fig. 1A, ␣7 nAChRs were found in association with G␣ s , G␣ q , and G␣ i proteins within the hippocampus and prefrontal cortex. Striatal fractions suggest ␣7 interaction with G␣ q and G␣ i subunits (Fig. 1A). Comparable levels of G␣ s , G␣ q , and G␣ i protein expression was seen on a Western blot in the threebrain structures (data not shown). An association between ␣7 nAChRs and G␤ was observed in the prefrontal cortex and hippocampus, whereas little to no detection of G␣ o was seen in the adult brain (Fig. 1A). Complementary experiments in which the anti-␣7 antibody C-20 was used to co-IP G␣ q and G␤ from the hippocampus confirmed association between the ␣7 nAChR and these G proteins (Fig. 1A).
␣-Bgtx is a selective antagonist of the ␣7 nAChR. We determined subtypes of G proteins that associated with a Bgtx C matrix from the mouse brain using tissue from ␣7 Ϫ/Ϫ as a negative control. This purification method enabled us to examine G proteins that interact with nAChRs (24). Western blotting confirmed expression of ␣7 subunits within the Bgtx C fraction (Fig. 1B). As shown in Fig. 1C, SDS-PAGE fractionation of the Bgtx C indicates greater specificity and stronger yield when the nicotinic agonist, carbachol, is used to elute from the Bgtx matrix. Gel lanes of the Bgtx C were divided into 3 molecular mass fractions: F1, 190-90 kDa; F2, 89-45 kDa; and F3, 44-15 kDa (Fig. 1C). Proteomic analysis was performed on the individual fractions using tryptic in-gel digestion followed by highperformance LC-ESI. A cohort of G proteins was identified in the F2 and F3 fractions (Fig. 1D). These G proteins were not detected in Bgtx C from ␣7 Ϫ/Ϫ mice (data not shown).
Identification of a GPBC within the M3-M4 Loop of nAChRs-We examined the existence of a G protein-binding motif within nAChRs. Mutagenesis experiments in the structurally related GlyR1 have demonstrated that amino acids RFRR at position 360 -363 within the M3-M4 loop enable G protein coupling (13). This binding sequence is similar to a G␤␥ binding motif (RFFP) within the inward rectifying potassium channel (GIRK1) (30). Sequence alignment of mammalian nAChRs revealed a conservation of residues at the GPBC site in the M3-M4 loop ( Fig. 2A). Based on amino acid similarity we found that ␣7 nAChRs possess 88% homology with the GyR1 at the GPBC ( Fig. 2A). Other nAChRs such as the ␣4, which have been shown to functionally couple to G proteins in immune cells (28), also demonstrate sequence similarity at the GPBC (63%). Because sequence similarity at the putative GPBC varies FIGURE 1. Proteomic detection of ␣7 receptor and G-protein complexes in the brain. A, IP using anti-G␣ s /G␣ q /G␣ i /G␣ o /G␤ or anti-␣7 antibodies to isolate ␣7/G protein complexes from 250 g of mouse hippocampus (1), prefrontal cortex (2), and striatum (3) membrane fractions. Blots were immunoblotted with mAb 306 or anti-G␣ q /␤ antibodies. The blots are representative of 3 separate experiments. B and C, ␣-Bgtx-binding proteins (Bgtx C ) were isolated from 1500 g of a membrane protein fraction of the adult mouse brain. Bgtx C was eluted with 1 or 0.5 M carbamylcholine (Carb). B, Western blot detection of ␣7 within the Bgtx C using mAb306. Experiments in ␣7 Ϫ/Ϫ tissue were used as a negative control. C, Bgtx C was divided into 3 molecular weight fractions for proteomic analysis: between nAChR subunits, it is plausible that nAChRs could differentially couple G proteins.
Generation of a G Protein Binding Dominant Negative ␣7 nAChR-To test if the GPBC in ␣7 nAChRs directs G protein binding, we mutated amino acids at position 345 to 348 in the human ␣7 subunit. All four amino acids were changed to alanine thereby generating a ␣7 345-348A nAChR proposed to lack a G protein coupling capacity. A similar mutation at the GPBC (RFRR, residues 360 -364) in the GlyR1 was found to abolish association with G proteins in HEK293 cells (13). Constructs encoding ␣7 nAChRs with an amino terminus YFP tag (YFP-␣7ϩ and YFP-␣7 345-348A ) were transfected into PC12 cells, which endogenously express the ␣7 receptor (4), YFP labeling of the nAChR at the amino terminus has been shown to not interfere with the trafficking or function of the nAChR (31). As shown in Fig. 2B, YFP-␣7ϩ and YFP-␣7 345-348A cells presented similar receptor expression patterns in the soma, neurite, and at the growth cone (GC) suggesting that the mutation in ␣7 345-348A does not impair receptor synthesis or trafficking (Fig. 2B).
Direct visualization of assembled ␣7 nAChRs using fBgtx in non-permeablized cells was used to determine the spatial distribution of cell surface receptors in ␣7, ␣7 ϩ , and ␣7 345-348A expressing PC12 cells (Fig. 2C). Recently we have shown a role for ␣7 nAChRs in the development of hippocampal neurons and PC12 cells via receptor localization at the GC (26). Strong fBgtx labeling was seen at the GC in ␣7, ␣7ϩ, and ␣7 345-348A nAChR expressing cells (Fig. 2C). Distribution analysis of the fBgtx signal indicates that overexpression of either receptor variant does not alter fBgtx labeling at the cell surface (Fig. 2C). The topographic distribution of the fBgtx signal was comparable between ␣7ϩ, ␣7 345-348A , and ␣7 controls (Fig. 2C). Statistical analysis indicates no significant difference in fBgtx distribution between the experimental groups.
We examined the expression of the ␣7 345-348A nAChR in an N2a cell line that does not express endogenous ␣7 nAChRs. N2a cells were co-transfected with Ric-3, which has been shown to promote the trafficking of ␣7 nAChRs to the cell surface (32)(33)(34). Visualization of fBgtx labeling in non-permeablized N2a cells transfected with ␣7 or ␣7 345-348A suggest that a mutation of the GPBC does not interfere with ␣7 nAChR cell surface expression in a heterologous system (Fig. 2D). Statistical analysis of the fBgtx signal indicates no significant difference in the level of ␣7 nAChR expression in the soma or GC of N2a cells transfected with either the wild-type or mutant construct (Fig. 2D). The findings confirm that the mutation in ␣7 345-348A does not interfere with the trafficking or expression of the nAChR.
A Loss in G Protein Coupling in ␣7 345-348A Expressing Cells-We determined the expression of ␣7 nAChRs in transfected cells. Western blot analysis confirms that transfection of ␣7 nAChRs augments total ␣7 subunit expression in PC12 cells, which express endogenous ␣7 nAChRs. Transfection with ␣7 (␣7ϩ) increased the immunoreactive ␣7 signal by over 60% from endogenous mock-transfected control cells, whereas transfection with the mutant ␣7 345-348A increased the total ␣7 signal by 48% over the endogenous ␣7 from control cells (Fig.  3A). Transfection of PC12 cells with ␣7 siRNA significantly reduces ␣7 nAChR protein expression and confirms the specificity of mAb306 on the Western blot ( Fig. 3A) (16,26). Studies in N2a cells indicate that transfection with ␣7 345-348A yields a similar ␣7 nAChR expression as the wild-type (␣7) and supports the finding that mutation at the GPBC does not impact the synthesis of the nAChR in cells.
Recently, we have shown that coupling to G␣ q enables ␣7-mediated calcium release from the endoplasmic reticulum (ER) leading to changes in neurite growth (18). We tested the ability of transfected ␣7 345-348A nAChRs to function as dominant negative regulators of G protein binding in PC12 cells, which endogenously express ␣7 nAChRs. Using an anti-G␣ q antibody we observed interactions between ␣7 nAChRs and G proteins in an IP experiment from transfected PC12 cells. As shown in Fig. 3A, coupling between G␣ q and ␣7 nAChRs was virtually abolished by expression of ␣7 345-348A . A noticeable loss in G␣ q (Ϫ62.18%) and G␤␥ (Ϫ20.03%) expression within the ␣7 nAChR complex IP was seen in cells transfected with ␣7 345-348A (Fig. 3A). An increase in G protein association within the ␣7 complex was observed in ␣7ϩ cells (G␣ q ϩ 16.71%; G␤␥ ϩ19.90%) (Fig. 3A). Similar findings in transfected N2a cells indicate a loss in G protein association within the ␣7 complex when ␣7 345-348A nAChRs are expressed (Fig. 3B). In particular, when the ␣7 345-348A nAChR was expressed by itself in N2a cells, little to no G␣ q /G␤␥ was detected within the ␣7 complex IP (Fig. 3B). Compared with N2a cells transfected with the wild-type ␣7 nAChR, ␣7 345-348A expressing N2a cells show little to no nAChR/G protein association (Fig. 3B). The data complements earlier findings on the ability of ␣7 345-348A to function as a dominant negative blocker of G protein coupling in PC12 cells, and suggests that the GPBC directs nAChR association with G␣ q and G␤␥.
An Attenuation of the Calcium Transient Response to Choline in ␣7 345-348A Cells-Upon binding ligands, ␣7 nAChRs conduct both calcium and sodium into the cell but rapidly desensitize resulting in a short-lived ionotropic response (21,35). Recent studies indicate that activation of ␣7 nAChRs can promote calcium release from the ER (18). In hippocampal slices this ␣7 nAChR-mediated calcium response persists for several minutes thereby modulating synaptic plasticity (6). To determine the role of G␣ q activation in ␣7 nAChR calcium signaling, we examined responses to choline, a selective ␣7 nAChR agonist using GCaMP5G imaging in PC12 cells (18,25). Cholinemediated calcium transients were seen in the neurite and GC (Fig. 4, A-C) consistent with fBgtx labeling at these sites (Fig. 2) (18). Calcium transients in the neurite peaked at 620% ⌬F/F (Ϯ88.7%). In ␣7ϩ cells, choline application resulted in a calcium signal that peaked at 1050% ⌬F/F (Ϯ176.4%) in the neurite. Calcium transients in the GC were found to last for ϳ1.6 s in both ␣7 and ␣7ϩ cells peaking at 1396% (Ϯ154.4%) and 1316% (Ϯ146.9%) ⌬F/F , relatively (Fig. 4, A and B).
Choline-induced Calcium Transients Depend on G␣ q and Calcium Channel Activity-We tested the role of a proposed G␣ q inhibitor SP in ␣7 calcium signaling in PC12 cells (18). As shown in Fig. 5, A and B, choline peak responses were significantly diminished (Ϫ53.87%) in cells pretreated with SP compared with vehicle controls (p ϭ 0.005). SP pretreatment did not significantly alter calcium peaks in ␣7 345-348A nAChR expressing cells, showing a small (Ϫ31.24%) reduction in calcium responses relative to the ␣7 345-348A baseline measure (p Ͼ 0.05). Similar results were seen when examining the AUC values for ␣7 (291.00% ⌬F/F 2 ϫ s Ϯ 113.78%) or ␣7 345-348A (253.75% ⌬F/F 2 ϫ s Ϯ 138.65%, p ϭ 0.003) nAChR expressing cells pretreated with SP (Fig. 5C). Because ␣7 345-348A -transfected cells did not show any responsiveness to SP, these findings suggest this receptor mutant is not functionally coupled to G␣ q .
A one-way ANOVA revealed a significant difference in both peak calcium responses (F(3,31) ϭ 8.048, p ϭ 0.001) and total AUC calcium transients (F ϭ (3,31) ϭ 8.191, p Ͻ 0.005) of SP, barium, and nifedipine on choline-induced calcium transients in ␣7 nAChR expressing PC12 cells. This effect was not observed in PC12 cells transfected with the ␣7 345-348A mutant. ␣7 345-348A nAChRs Do Not Activate PLC-Activation of G␣ q sets into motion PLC-associated signaling leading to the formation of IP 3 and the mobilization of calcium release from the ER (23). This second messenger pathway can be examined via the genetically encoded PLC(␦) sensor pleckstrin homology mCherry probe (PH-mCherry) (27). A translocation of PH-mCherry from the cell surface to the cytosol has been found to quantitatively relate to PIP 2 breakdown and IP 3 production (18,27). Treatment of PC12 cells with 10 mM choline was associated with a translocation of PH-mCherry from the cell surface as determined by the presence of the fluorescence signal within 1 m of the edge of the cell into the cytosol of the GC (Fig. 6, A   and B). Pre-treatment of cells with SP abolished this translocation (Fig. 6B). In PC12 cells expressing ␣7 345-348A nAChRs choline had a weak effect on PH-mCherry translocation relative to empty plasmid-transfected controls. Expression ␣7 345-348A nAChRs was surprisingly associated with strong levels of PH-mCherry at the cell surface in the absence of drug treatment (Fig. 6B).
Sequential imaging of PH-mCherry and GCaMP5G confirms that choline promotes a rise in intracellular calcium and PH-mCherry translocation in the same cellular compartment (Fig.  6, B and C). Cytoplasmic translocation of PH-mCherry occurred on a slower time scale (40 s after choline application) than peak calcium responses (ϳ1 s after choline application). These kinetics are consistent with the translocation of the PH domain sensor in the cell (20,29).
Ligand Stimulation of the ␣7 Receptor Reduces Its Interaction with G Proteins-We examined the dynamics of ␣7 nAChR/G protein coupling in response to ligand stimulation. PC12 cells were treated with the selective ␣7 nAChR agonist choline (10 mM) for 2 min prior to analysis. An IP assay was utilized to examine changes in nAChR/G protein coupling in drug-treated versus vehicle-treated control cells. As shown in Fig. 8, A and B, IP of the ␣7 using the C-20 antibody suggests that choline application attenuates G protein binding with the nAChR. Choline treatment resulted in a 56% reduction in G␣ q and 47% reduction in G␤ association within the ␣7 nAChR complex (Fig. 8B). In cells transfected with the ␣7 345-348A nAChR a surprising 38.81% increase in G␤ detection was seen in the ␣7 nAChR IP experiment following choline treatment (Fig. 8, A and B). The data suggest a role for ligands in the modulation of ␣7 nAChR interaction with G proteins.

A Motif for G Protein Binding in Nicotinic Receptors-Cys-
loop receptors such as the GlyR1 have been shown to directly bind G proteins via specific amino acids within the M3-M4 intracellular loop (13). In this article, we show a similar GPBC in ␣7 nAChRs and propose the existence of a GPBC in other nAChR subunits such as ␣4 and ␣1 based on sequence homology at this site. This is consistent with recent studies that indicate that nicotine stimulation of ␣4 nAChRs activates G␣ i/o signaling in CD4ϩ T-cells leading to cytokine release in mice (28). A mutation of the GPBC within ␣7 nAChRs was found sufficient to eliminate association with G␣ q as well as G␤ subunits in both PC12 and N2a cells. In preliminary studies, we found that individual mutations at the GPBC did not signifi-cantly impact G protein binding, 3 confirming that all four residues contribute to a G protein binding core within the ␣7 nAChR. In light of the similarity of this site with the G protein binding residues (RFFP) within the GIRK1 channel, it is plausible that a minimal consensus sequence consisting of RXYR 3 J. R. King and N. Kabbani, unpublished data.  (where X is any amino acid and Y is a basic amino acid) facilitates G protein binding in various proteins (30).
Coupling to G␣ q appears essential for ␣7 nAChR-mediated calcium release from the ER; however, ␣7 nAChRs can also couple to other G proteins (24). This may be determined by factors such as the localization of the nAChR within subcellular domains such as lipid rafts which, may enable proximity to specific G proteins (39), or the interaction between the nAChR and adaptor molecules, such as G protein-regulated inducer of neurite outgrowth 1, which may facilitate coupling to specific G proteins (4,26). Receptor associations with G␤␥ can also influence coupling to G␣ (40). This is in agreement with our proteomic findings that indicate that ␣7 nAChRs bind both G␣ as well as G␤␥ proteins and that this association is reduced in the presence of a ligand. The existence of a nAChR/G protein trimer complex is also consistent with FRET and chemical cross-linking evidence on functional associations between trimeric G proteins and their targets (41).
The M3-M4 loop is presumed to constitute an important intracellular protein-protein interaction interface within nAChRs controlling functional features of the receptor such as its targeting to axons or dendrites (31,42,43). Our findings expand this theory and indicate that binding of G proteins enables a downstream signaling response by ␣7 nAChRs. Based on this study we can infer a dominant negative function to ␣7 345-348A nAChR subunit expression in cells. Whether 1 or more ␣7 345-348A subunits are incorporated into the homomeric receptor is sufficient to entirely disrupt G protein signaling is an important focus of future studies. Based on our findings in heterologous N2a cells, expression of the ␣7 345-348A subunit alone enables the formation of fBgtx binding nAChRs at the cell surface. Compared with cells that expressed the wild-type ␣7, ␣7 345-348A expressing N2a cells were found to respond to choline with a noticeably reduced calcium transient. A similar reduction (Ͼ50%) in calcium responsiveness to choline is also found in PC12 cells that were transfected with ␣7 345-348A subunits and suggests that expression of this dominant negative disrupts the G protein coupling of endogenous ␣7 nAChRs. Similarly, co-IP experiments demonstrate that expression of ␣7 345-348A in PC12 and N2a cells is associated with a prominent reduction in G protein interaction with ␣7 nAChRs.
A G protein signaling response may account for some of the non-conducting channel functions of new compounds termed "silent agonists" such as NS6740, which may be able to engage intracellular calcium stores (44). Although it will be of interest to test the effects of such compounds on the G protein signaling response of the ␣7 nAChR, proteomic findings presented in this study suggest that G protein association is detectable in the Bgtx bound state. It will be important to determine whether agonist currents through the ␣7 345-348A nAChR are altered relative to the wild-type. Based on previous findings on the impact of mutations in the M3-M4 loop, it is expected that currents are unaffected in the ␣7 345-348A nAChR (45).
G Protein Coupling Activates a Local Rise in Intracellular Calcium-Ligand activation of the nAChR leads to a depolarization of the neuronal membrane and changes in neuronal firing (38). Activation of nAChRs can also mediate longer-lived cellular signals including the modulation of survival and plasticity genes (15). Positive allosteric modulation of the ␣7 nAChR promotes cytotoxicity in SH-SY5Y cells via the release of calcium from the ER (46). In adult hippocampal neurons, ␣7 nAChR activation of ER calcium release promotes calcium transients that modulate glutamate release at synapses (6). Our earlier studies indicate a role of ␣7/G protein signaling in regulating the release of calcium from the ER during axon development via their interactions at the GC (18). Functional coupling to G proteins, such as G␣ q , provide an important mechanistic explanation on how ␣7 nAChRs regulate intracellular calcium release from the ER and may play a role in receptor distribution.
Based on the current state of evidence, we propose a model by which direct G protein interactions enable metabotropic signaling via the ␣7 nAChR receptor (Fig. 9). In this model, nAChR signaling via the G protein pathway may involve crosstalk with nearby G protein-coupled receptors, which can mediate GTP exchange on the G␣ q subunit. An analysis of the peak response and duration of the PLC sensor (PH-mCherry) clearly shows that choline activation of ␣7 nAChRs enhances the activity of PLC in the same cellular compartment as measured calcium responses. Here and previously, we have shown that Bgtx, a GPBC mutation, as well as the G␣ q blocking peptide SP, attenuates this process in cells (18). Moreover, in cells expressing the ␣7 345-348A mutant, a loss in PLC activation is consistent with the role of G␣ q coupling in nAChR-associated PLC signaling and IP 3 -mediated calcium store release. However, it is not clear why expression of the ␣7 345-348A mutant increases the PH-mCherry signal at the cell surface.
Pharmacological inhibition of IP 3 Rs by xestospongin C appears sufficient to attenuate the ␣7 calcium signal at the GC. Varied responses to xestospongin C exist throughout the cell and suggest a role for compartment specific mechanisms. At pharmacological levels consistent with IP 3 R inhibition (47), xestospongin C attenuates the choline-induced calcium response to the same extent as ␣7 345-348A nAChR expression. The evidence suggests that ␣7 nAChRs in concert with G proteins promote a local elevation in cellular calcium that can persist for seconds beyond the expected time frame of ␣7 channel activation. ␣7-Mediated calcium transients are strongest near the ER consistent with the role of IP 3 R activation downstream of nAChR signaling (35,38). Interestingly, this time period of intracellular calcium signaling may persist during ␣7 nAChR desensitization (35). Replacement of extracellular calcium with barium, which does not pass through the ␣7 nAChR, did not alter the calcium transient response to choline. Because barium is permeable via VGCCs, however, our findings underscore a role of calcium channels in the G protein-associated nAChR signaling response (37,38). This is evidenced by the finding that choline-mediated calcium transients are noticeably attenuated in the presence of the L-type channel blocker nifedipine in ␣7 but not ␣7 345-348 nAChR expressing cells. Because VGCC are regulated by G␤␥, the findings suggest a mechanism of VGCC regulation by nAChRs via G proteins (48). This process may contribute to a local rise in intracellular calcium via cholineinduced calcium release (46).