6

Click here to load reader

Intracellular complexes of the 2 subunit of the nicotinic ... complexes of the 2 subunit of the nicotinic acetylcholine receptor in brain identified by proteomics Nadine Kabbani*†,

  • Upload
    dobao

  • View
    221

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Intracellular complexes of the 2 subunit of the nicotinic ... complexes of the 2 subunit of the nicotinic acetylcholine receptor in brain identified by proteomics Nadine Kabbani*†,

Intracellular complexes of the �2 subunit of thenicotinic acetylcholine receptor in brainidentified by proteomicsNadine Kabbani*†, Matthew P. Woll‡, Robert Levenson‡, Jon M. Lindstrom§, and Jean-Pierre Changeux*†

*Recepteurs et Cognition, Centre National de la Recherche Scientifique Unite de Recherche Associee 2182, Institut Pasteur, 75724 Paris, France;‡Department of Pharmacology, Pennsylvania State College of Medicine, Hershey, PA 17033; and §Department of Neuroscience,University of Pennsylvania Medical School, Philadelphia, PA 19104

Contributed by Jean-Pierre Changeux, October 31, 2007 (sent for review August 30, 2007)

Nicotine acetylcholine receptors (nAChRs) comprise a family ofligand-gated channels widely expressed in the mammalian brain.The �2 subunit is an abundant protein subunit critically involved inthe cognitive and behavioral properties of nicotine as well as in themechanisms of nicotine addiction. In this work, we used matrix-assisted laser desorption ionization time-of-flight tandem massspectrometry (MALDI-TOF-TOF MS/MS) to uncover protein inter-actions of the intracellular loop of the �2 subunit and componentsof immunoprecipitated �2–nAChR complexes from mouse brain.Using the �2-knockout mouse to exclude nonspecific binding tothe �2 antibody, we identify 21 nAChR-interacting proteins (NIPs)expressed in brain. Western blot analysis confirmed the associationbetween the �2 subunit and candidate NIPs. Based on theirfunctional profiles, the hypothesis is suggested that the identifiedNIPs can regulate the trafficking and signaling of the �2–nAChR.Interactions of the �2 subunit with NIPs such as G protein �, Gprotein-regulated inducer of neurite outgrowth 1, and G protein-activated K� channel 1 suggest a link between nAChRs and cellularG protein pathways. These findings reveal intracellular interactionsof the �2 subunit and may contribute to the understanding of themechanisms of nAChR signaling and trafficking in neurons.

mass spectrometry � receptor complex

N icotine is a common drug of addiction and a leading causeof preventable deaths in developed countries (1, 2). The

target of nicotine is a class of nicotinic acetylcholine receptors(nAChRs) existing as pentameric channels made up of variousreceptor subunits and distributed throughout the brain (3). Todate, 11 neuronal subunits have been identified: �2–�9 and�2–�4. The specific combination of these subunits confers thepharmacological and physiological properties of the nAChR (4).Studies show that �90% of the high-affinity nicotine receptorsites within the brain consist of the �4�2–nAChR (5, 6). Topo-logical analysis of nAChR subunits indicates the presence of alarge, highly divergent intracellular loop (M3–M4 loop) that isimplicated in trafficking and targeting properties of nAChRs (7,8). Recent analysis of a prokaryotic homolog of the nAChRfamily reveals that the M3–M4 loop is unique to the evolution ofthe nAChR in eukaryotes (9).

The generation of subunit-specific knockout (KO) mice hasproven valuable in defining the role of individual nAChRsubunits in brain physiology and animal behavior (10). Experi-ments in mice lacking the �2 subunit (�2�/�) demonstrate animportant role for this receptor subunit in neuronal developmentand plasticity (10), protection from excitotoxicity (11), and themechanisms underlying cognitive and social behaviors (12).�2�/� mice also exhibit a loss of nicotine self-administration andnicotine-elicited firing and dopamine release from dopaminergicneurons of the ventral tegmental area (13, 14). Because reex-pression of the �2 subunit in �2�/� mice was found sufficient torestore the electrophysiological and behavioral effects of nico-

tine (5), this subunit is likely a central mediator in the mechanismof nicotine addiction.

The major functional properties of receptor systems have beenfound to depend on their association with various intracellularmolecules (15). Receptor proteomics has thus emerged as aprimary approach in the molecular analysis of receptor signalingand regulation (16–18). Multiprotein complexes (termed signal-ing complexes or signalplexes) have been found to associate withreceptors such as glutamate N-methyl-D-aspartate (NMDA) andmGluR5 (19, 20), glucocorticoid (21), serotonin 5-hydroxytryp-tamine type 2A and 2C (22), dopamine D1 and D2 (23), and theATP-gated channel (P2X7) (24). Earlier studies based on yeasttwo-hybrid screens indicate that the �4-nAChR binds the scaf-fold molecule 14-3-3� (8) and the calcium sensor visinin-likeprotein 1 (25). In this work, we explore interactions of the �2subunit by using matrix-assisted laser desorption ionizationtime-of-f light tandem mass spectrometry (MALDI-TOF-TOFMS/MS). Defining the constituents of the �2–nAChR signalplexrepresents an essential step in understanding the mechanisms ofnAChR-mediated signaling and regulation in brain.

ResultsProteins Interacting with the M3–M4 Loop of the �2 Subunit. TheM3–M4 loop of nAChRs mediates important receptor propertiessuch as export from the endoplasmic reticulum (ER) (25, 26),targeting to axonal or dendritic compartments (27), and inter-actions with scaffold and signaling molecules (8, 28). To identifyproteins that associate with the �2–M3–M4 loop, we generateda fusion protein encoding the M3–M4 loop (amino acid residues326–454 of mouse �2) in association with GST [GST-�2 (M3–M4)], and we used a pulldown strategy to isolate the interactionsof GST-�2 (M3–M4) from mouse brain [supporting information(SI) Fig. 4A]. We used a Sepharose matrix to purify GST-�2(M3–M4), and the purified product was verified by Coomassiestaining and Western blotting with an anti-GST antibody (SI Fig.4 B and C). Western blot analysis of purified GST-�2 (M3–M4)indicates expression of a prominent band at 42 kDa, the expectedmolecular mass of GST-�2 (M3–M4) on an SDS/polyacrylamidegel (SI Fig. 4C). We also detected a minor band with reactivityto anti-GST at �38 kDa (SI Fig. 4C). Because this band did notstain with an antibody directed against the �2 subunit (data notshown), we conclude that it represents a nonspecific interactionof the GST antibody on the blot.

Author contributions: N.K. and J.-P.C. designed research; N.K., M.P.W., and R.L. performedresearch; J.M.L. contributed new reagents/analytic tools; N.K. analyzed data; and N.K.wrote the paper.

The authors declare no conflict of interest.

†To whom correspondence may be addressed. E-mail: [email protected] [email protected].

This article contains supporting information online at www.pnas.org/cgi/content/full/0710314104/DC1.

© 2007 by The National Academy of Sciences of the USA

20570–20575 � PNAS � December 18, 2007 � vol. 104 � no. 51 www.pnas.org�cgi�doi�10.1073�pnas.0710314104

Page 2: Intracellular complexes of the 2 subunit of the nicotinic ... complexes of the 2 subunit of the nicotinic acetylcholine receptor in brain identified by proteomics Nadine Kabbani*†,

For pulldown experiments we immobilized GST-�2 (M3–M4)or GST alone (as a control) onto the Sepharose matrix. Beadswere then incubated with solubilized preparations of mousebrain to isolate proteins that associate with GST-�2 (M3–M4) orGST alone. Results of the pulldown assay are presented in Fig.1. Coomassie blue staining of the gel shows a difference in theprotein interaction profile of GST-�2 (M3–M4) and GST alone(Fig. 1B). In total, seven unique bands were visualized in theGST-�2 (M3–M4) pulldown lane (annotated bands in Fig. 1B),and these bands were excised for MALDI-TOF-TOF MS/MSanalysis [the pulldown experiment was performed four times(n � 4), and the results were similar in each case (data notshown)]. A sample MS spectrum (for synaptotagmin 1) is shownin Fig. 1C. The mass of the parent peptides used to identify theinteracting proteins is listed in SI Table 3). A complete list ofnAChR-interacting proteins (NIPs) derived from the GST-�2(M3–M4) pulldown experiment is presented in Table 1.

We tested interaction between GST-�2 (M3–M4) and NIPs byWestern blotting. As shown in Fig. 1 A, immunoblot analysis ofprotein complexes derived from the pulldown assay confirmedthe interaction of dynamin 1, clathrin heavy chain (HC), synap-totagmin 1, and N-ethylmaleimide-sensitive factor (NSF) withGST-�2 (M3–M4).

Interactions of the �2–nAChR in the Brain Defined by the Immuno-precipitation Method. We used a monoclonal antibody (mAb 270)directed against the N terminus of the �2 subunit (29) to isolate�2 subunit-containing nAChRs (�2–nAChRs) from the brain ofmouse (SI Fig. 5). Previous studies indicate that mAb 270immunoprecipitates the �2 subunit from membrane prepara-tions of cultured cells and native brain tissue (29, 30). Weassessed the ability of mAb 270 to immunoprecipitate the �2subunit from the mouse brain. Western blot experiments, withpolyclonal anti-�2 (H-92) as an immunoprobe, indicate thatmAb 270 successfully immunoprecipitated the �2 subunit fromtransfected human embryonic kidney (HEK) 293 cells as well asfrom brain tissue of wild type (WT) mice (Fig. 2 A and B). Insamples of �2�/� mice, subject to immunoprecipitation withmAb 270 under equivalent conditions, anti-�2 reactivity was notdetected on the Western blot (Fig. 2 A). Probing the blot with apolyclonal antibody directed against the �4 subunit (H-133),however, revealed the presence of �4 within the immunopre-cipitated complex of WT and �2�/� mouse brain (Fig. 2 A).These findings indicate reactivity between mAb 270 and themouse �4 subunit within the immunoprecipitation complexderived from brain, and they suggest reactivity between mAb 270and non-�2–nAChR subunits under these conditions (31).

We compared the constituents of the immunoprecipitated �2subunit complex between WT and �2�/� mice to eliminate thenonspecific binding of mAb 270 in brain tissue (32). A repre-sentative Coomassie blue-stained gel shows the results of animmunoprecipitation experiment conducted in WT and �2�/�

mice (Fig. 2C). Protein bands that appeared common to the twosample lanes were omitted from the MALDI-TOF-TOF MS/MSanalysis, and only the bands unique to the WT (annotated bandsin Fig. 2C) were analyzed. A total of 23 unique bands were foundin the WT sample lane, and proteomic analysis of these bandsrevealed the identity of 16 NIPs from NCBI (Fig. 2C and Table2). In addition, 6 other putative NIPs matched ‘‘hypothetical’’protein products of mouse genomic DNA sequences listed in theNational Center for Biotechnology Information (NCBI) (anno-tated bands lacking numbers in Fig. 2C).

The MS spectrum for a representative NIP [G protein regu-lated inducer of neurite outgrowth (GPRIN) 1] is shown in Fig.3B, and the mass of the parent peptides used to identify NIPs is

Fig. 1. Identification of NIPs from GST-�2 (M3–M4) pulldown experiments.(A) Western blots showing immunoreactivity for anti-dynamin 1, anti-clathrinHC, anti-synatotagmin 1, and anti-NSF in the pulldown assay of GST-�2(M3–M4). (B) Coomassie blue-stained gel showing a spectrum of proteininteractions for GST-�2 (M3–M4) and GST alone (control) in pulldown exper-iments. Arrowheads indicate the positions of bands (listed in Table 1) analyzedby MALDI-TOF-TOF MS/MS. The asterisk band shows the position of GST-�2(M3–M4) on the gel. (C) MS spectrum for synaptotagmin 1 (T � trypsinautolysis product) isolated by the pulldown experiment.

Table 1. Intracellular proteins that associate with GST-�2 (M3–M4)

Gelband Protein

NCBIaccession no.

Mass,kDa

No. of uniquepeptides

MASCOTscore* CI†

1 Tubulin (�1) gi 4990313 50.8 14 290 1001 Tubulin (�6) gi 13435777 50.6 13 290 1002 Clathrin HC‡ gi 51259242 193.2 23 89 1003 NSF‡ gi 31543349 83.13 22 105 99.94 Dynamin 1‡ gi 21961254 98.14 19 199 1005 Heat shock protein 90 (� subunit) gi 14714615 92.71 22 100 99.96 P lysozyme, structural gi 7305247 17.24 10 87 99.97 Synaptotagmin 1‡ gi 94730428 47.71 22 268 100

NIPs corresponding to gel band numbers in Fig. 1B are identified by protein accession number in NCBI, molecular mass (Mass), thenumber of unique sequenced peptides, a MASCOT protein score, and a protein confidence index (CI). Complete information on thesequence and mass of the digested peptides used to identify each protein is presented in SI Table 3.*MASCOT minimal protein score � 70.†Derived from GPS Explorer.‡Protein interactions of GST-�2 (M3–M4) validated by Western blot analysis.

Kabbani et al. PNAS � December 18, 2007 � vol. 104 � no. 51 � 20571

NEU

ROSC

IEN

CE

Page 3: Intracellular complexes of the 2 subunit of the nicotinic ... complexes of the 2 subunit of the nicotinic acetylcholine receptor in brain identified by proteomics Nadine Kabbani*†,

presented in SI Table 4. A complete list of NIPs identified byimmunoprecipitation is presented in Table 2.

Western blot analysis confirmed the presence of dynamin 1, Gprotein-coupled inwardly rectifying potassium channel (GIRK)1, and guanine nucleotide-binding protein O (Go�) within com-plexes of the �2 subunit (Fig. 3A). These NIPs were not detectedwithin immunoprecipitated complexes of the �2�/� brain underthe same experimental conditions (Fig. 3A). To test whetherNIPs identified by pulldown were present within the immuno-precipitated �2 subunit complex, we probed the blot withantibodies against NSF and synaptotagmin 1 (Fig. 3C). Anti-bodies directed against these two proteins also reacted weaklywith the immunoprecipitated complexes from �2�/� mice (Fig.3C). Multiple repeats (n � 3) of this experiment confirmed thedifference in band intensity between the WT and �2�/� samplelane for these proteins (data not shown), suggesting that theyhave been excluded from the proteomic analysis because of theirpresence in the �2�/� lane. The presence of NIPs within theimmunoprecipitated complex of the �2�/� suggests that theymay interact with additional nAChR subunits, such as �4,immunoprecipitated by the antibody.

DiscussionWe have examined constituents of the �2 subunit complex frombrain by using MALDI-TOF-TOF MS/MS analysis of multipro-tein complexes isolated by affinity purification with pulldownand immunoprecipitation methods. The isolation of proteincomplexes from native tissue provides a significant advantageover traditional protein interaction screens, such as the yeasttwo-hybrid, because it allows for the analysis of interactionswithin the normal cellular milieu (33). We conclude that the �2

subunit associates with well �20 cellular proteins termed NIPs.NIPs represent components of diverse �2–nAChR complexesexisting within various brain regions, cell types, and subcellularcompartments. At present, we do not know the specific profileand distribution of the various �2–nAChR complexes that existin the brain. Moreover, because many protein interactions aretransient, our proteomic analysis more likely identifies stableinteractions of the �2 subunit. This is underscored by thediscovery of cytoskeletal and scaffold molecules, such as clathrinand tubulin, and a dearth in the identification of signalingmolecules such as kinases among the list of NIPs. Our findingsare consistent with those of others on protein interactions forreceptor and channel molecules and highlight a limitation of thecurrent high-throughput proteomic approach (33).

The discovery of NIPs was achieved by pulldown and immu-noprecipitation assays, and several NIPs (such as clathrin, dy-namin, and tubulin) were found in both approaches. Some of theidentified proteins (such as lysozyme, creatine kinase, andGAPDH) may represent interaction artifacts that arise fromnonspecific binding to the antibody or the fusion protein or evenpossibly contaminants within the sample preparation. A proteinsuch as myelin basic protein, for example, may associate with thereceptor complex by means of (nonspecific) hydrophobic inter-actions. However, it is important to consider that myelin basicprotein is known to bind various channel and receptor molecules(34) and that nAChRs are expressed within glial cells such asoligodendrocytes (35), suggesting that interaction between thesetwo proteins cannot be excluded. Because the results of thepulldown study are based on interactions of the M3–M4 loopfusion protein with proteins expressed in brain, we propose thatthe NIPs identified by the pulldown strategy (such as synapto-tagmin 1 and NSF) associate with the �2 subunit by coupling tothe intracellular region of the receptor either directly or indi-rectly (as by an intermediate protein). Moreover, because inter-actions between proteins depend on conditions defined by theexperimental paradigm, such as those imposing constraints onthe stability of protein bonds as well as the proximity ofinteraction motifs (36), it is expected that NIPs identified bypulldown will differ from NIPs identified by immunoprecipita-tion (33, 37). Clearly, the combined use of multiple techniquesprovides the best strategy for identifying the full spectrum ofinteractions for any given protein (38). This fact is underscoredby a report on the proteomic analysis of the NMDA receptorfrom brain, showing �77 binding partners for this receptor byuse of immunoprecipitation, pulldown, and yeast two-hybridprocedures (19).

Antibody cross-reactivity can lead to the identification offalse-positive interactions, thereby confounding the proteomicresults. To overcome problems associated with cross-reactivityof the �2 antibody, we have conducted control experiments using�2�/� mice. This strategy aims to ensure selection of interactionsthat are specific to the �2 subunit expressed in the WT brain. Itwas somewhat perplexing that we did not detect associationbetween �2 and other nAChRs (such as �4, �6, �5, and �3) asexpected for mature �2–nAChRs (39). Because Western blottingexperiments revealed that nAChRs, such as �4, are present inimmunoprecipitated complexes of the �2�/�, we conclude thatadditional nAChR subunits have been excluded from the MSanalysis because of their presence in the �2�/� control lane.These findings indicate a secondary reactivity of the anti-�2antibody (mAb 270) to non-�2–nAChRs during the immuno-precipitation procedure. This reactivity is consistent with pre-vious observations on cross-reactivity among nAChR antibodiesin tissue (31) and may underlie the presence of some of thepolypeptides within the immunoprecipitated complexes of the�2�/� mouse.

The presence of chaperone and transport proteins, such aseukaryotic elongation factor (eEF) and heat shock protein 90,

Fig. 2. Immunoprecipitation of the �2 subunit from brain. (A) Western blotsshowing the expression of �2 subunits within immunoprecipitated complexesfrom WT (�/�) mouse, using mAb 270. Expression of �4 was also detectedwithin immunoprecipitated complexes of WT (�/�) and �2�/� mouse brain.(B) Immunoblot confirming comigration of the �2 subunit from the WT (�/�)mouse brain with the �2 subunit expressed in HEK 293 cells. (C) Coomassieblue-stained gel showing proteins that coimmunoprecipitate with the �2subunit in WT (�/�) and �2�/� brain preparations. Arrowheads point to theposition of bands analyzed by MALDI-TOF-TOF MS/MS (listed in Table 2).Arrowheads without a number indicate the positions of bands that corre-spond to unknown proteins within NCBI. Asterisks point to the light- andheavy-chain Ig subunits of the monoclonal antibody.

20572 � www.pnas.org�cgi�doi�10.1073�pnas.0710314104 Kabbani et al.

Page 4: Intracellular complexes of the 2 subunit of the nicotinic ... complexes of the 2 subunit of the nicotinic acetylcholine receptor in brain identified by proteomics Nadine Kabbani*†,

suggests that our samples contain a fraction of unassembled �2subunits from cells. Studies in cultured cells demonstrate thepresence of large pools of partially assembled �2 subunits thatare in association with Golgi and ER membranes. These subunitswere found to assemble into mature �2–nAChRs in response tonicotine (40). Ultrastructural findings in rodent brain alsosupport the notion that the majority of �2–nAChRs reside withinthe cell and in association with organelles such as the ER as wellas with transport vesicles (41). Because studies show that chronicnicotine enhances the maturation of unassembled �2 subunitswithout altering �2 expression (30), it is tempting to speculatethat molecules such as NIPs can contribute to the maturation andtrafficking of �2–nAChRs in response to nicotine treatment.

Interestingly, alterations in the expression of NSF and dynamin1 have recently been documented in the brains of rodentssubjected to chronic nicotine (42).

Protein Complexes for Trafficking the �2 Subunit in Cells. Themovement of receptors within cells requires the activity ofmolecules such as dynamin, a GTPase that mediates vesiclebudding, and clathrin, a major constituent of the vesicle coat (8,25). Several ligand-gated ion channels, such as GABAA, arefound to traffic in a dynamin-dependent fashion (43). Ourexperiments indicate association of the �2 subunit with theneuron-specific dynamin 1 isoform and with the HC of theclathrin complex. Although it is not clear whether these inter-actions represent direct contacts between the nAChR and theNIP, our work supports the idea that �2–nAChRs are traffickedin a dynamin- and clathrin-dependent manner within neurons.Such a model of nAChR trafficking is in agreement withultrastructural evidence of �2 subunit localization within clath-rin-coated vesicles in neurons of the cortex (41).

The family of eEF proteins mediate protein synthesis andtrafficking in cells and recently have been shown to contribute toplasticity within the brain (44). Studies indicate that eEF familyproteins also bind to neurotransmitter receptors, such as the D3dopamine and M4 muscarinic receptors, and mediate theirtrafficking to the plasma membrane (45, 46). We have uncoveredan association between the �2 subunit and eEF2. It is possiblethat eEF2 interacts with �2–nAChRs within the ER, and thisinteraction contributes to subunit maturation and folding. Al-ternatively, it is possible that eEF2–�2 subunit associationsmediate the assembly and targeting of �2–nAChRs to the plasmamembrane.

NSF is an ATPase involved in the movement of transportvesicles within cells (47). NSF has been found to interact withneurotransmitter receptors such as the glutamate AMPA recep-tor and the �2-adrenergic receptor and to regulate their trans-port to synapses (48, 49). NSF has also been found to mediate

Fig. 3. Confirmation of NIP–�2 subunit interactions. (A) Representative blotsshowing immunoreactivity to dynamin 1, GIRK1, and Go� subunits within theimmunoprecipitated complexes from the brain of WT (�/�) and �2�/� mice.(B) MS spectrum for GPRIN1 (T � trypsin autolysis product). (C) Western blotsconfirming the presence of synatotagmin 1 and NSF within immunoprecipi-tated complexes of the �2 subunit from mouse brain. Note that a weakreactivity to both antibodies was also detected in lanes of the �2�/�.

Table 2. Intracellular proteins that coimmunoprecipitate with the �2 subunit in brain

Gelband Protein NCBI accession no.

Mass,kDa

No. of uniquepeptides

MASCOTscore* CI†

1 Actin � gi 49868 39.5 19 220 1001 Actin � gi 809561 41.33 21 214 1002 GAPDH gi 55153885 36.09 14 79 99.83 Myelin basic protein 1 gi 69885032 21.5 15 121 1003 Myelin basic protein 3 gi 69885049 18.5 10 149 1003 Myelin basic protein 5 gi 69885065 17.23 13 88 99.94 eEF-2 gi 13938072 96.2 20 165 1005 Pyruvate kinase 3 gi 16741633 58.42 12 134 1006 GIRK1 (Kir3)‡ gi 54037432 50.8 9 70 96.77 Clathrin HC‡ gi 51259242 193.6 74 399 99.998 Dynamin 1‡ gi 18093102 96.21 22 193 1009 Tubulin �2 gi 80478451 50.8 24 281 100

10 Cyclic nucleotide phophodiesterase 1 gi 14193680 45.02 18 177 10011 Voltage-dependent anion channel 1 gi 38051979 32.06 11 72 97.312 Spectrin � chain gi 94366114 286.7 57 217 10013 G-protein (Go�)

‡ gi 6754012 40.63 11 80 99.514 Mitochondrial ATP synthase H � pump, � subunit gi 89574015 48.05 17 193 10015 Neuronal nAChR, �2 subunit‡ gi 31542397 57.87 14 101 99.916 Creatine kinase gi 6753428 47.37 22 264 10017 GPRIN1 gi 81892663 96.06 18 89 98.5

NIPs corresponding to gel band numbers in Fig. 2C are identified by protein accession number in NCBI, molecular mass (Mass), the number of unique sequencedpeptides, a MASCOT protein score, and a protein confidence index (CI). Complete information on the sequence and mass of the digested peptides used to identifyeach protein is presented in SI Table 4.*MASCOT minimal protein score � 70.†Derived from GPS Explorer.‡Protein interactions validated by Western blotting.

Kabbani et al. PNAS � December 18, 2007 � vol. 104 � no. 51 � 20573

NEU

ROSC

IEN

CE

Page 5: Intracellular complexes of the 2 subunit of the nicotinic ... complexes of the 2 subunit of the nicotinic acetylcholine receptor in brain identified by proteomics Nadine Kabbani*†,

the trafficking of the nAChR �7 subunit within postsynapticregions of ciliary ganglion neuron (50). We find that the M3–M4loop of the �2 subunit binds to NSF, and we confirm thisinteraction by using coimmunoprecipitation and Western blot-ting methods. Based on these findings, we hypothesize thatinteraction with NSF directs the trafficking of �2–nAChRs inneurons.

Potential Signaling Properties of the �2 Subunit. nAChRs have beendetected within the cell body, dendrites, and at synapses of neurons(39, 51). Presynaptically localized nAChRs, such as the �4�2, areknown to play an important role in regulating neurotransmitterrelease in the brain (3). To date, however, little is known about themechanisms by which nAChRs regulate neurotransmitter release.Our data indicate an interaction between the �2 subunit andsynaptotagmin 1, a calcium sensor involved in vesicle docking andfusion with the synaptic plasma membrane (52). In previous studies,synaptotagmin 1 has also been found to interact with voltage-gatedcalcium channels (53) and with muscarinic receptors (54). Our worksuggests that synaptotagmin 1 associates with the �2–nAChRwithin the M3–M4 loop and that this interaction may represent apossible mechanism whereby �2–nAChRs can regulate neurotrans-mitter release from synapses.

Our results indicate an interaction between the �2 subunit andcellular signaling proteins such as Go�, GIRK1, and GPRIN1.These findings suggest an unexpected association between �2–nAChRs and cellular G protein signaling pathways. Althoughthis idea is unique for the nAChR, it is supported by reportsshowing an effect of G protein inhibitors on the gating andfunctional response of nAChRs in various types of cells (55, 56).Recently, an interaction between �2–nAChRs and the G pro-tein-coupled dopamine D2 receptor in rat striatum has beenreported (57). Our immunoprecipitation experiments confirminteraction between �2–nAChRs, GIRK1 channels, and Go�

subunits in the brain. However, the MASCOT score for GIRK1appears low based on immunoprecipitation studies, suggestingthat GIRK1 interacts with the �2 subunit indirectly. In onemodel, GIRK1 may interact with the �2–nAChR through amutual association of the Go� subunit (58). This interaction issupported by Western blot findings that suggest that Go� is moreabundant then GIRK1 within the immunoprecipitated �2–nAChR complex (compare band intensity in Fig. 3A). Thismodel of indirect coupling between �2–nAChRs and GIRK1channels is consistent with other reports on the nature ofassociations between receptors and channels in cells (59). Lastly,because GPRIN1 is reported to bind activated Go� subunits (60),GPRIN1 may also play a role in the coupling among �2–nAChRs, Go� subunits, and GIRK channel molecules. Theseproteomic findings represent a model for the signaling andregulation of the �2–nAChR in the brain by their association tothe cellular G protein pathway.

MethodsDissection and Protein Preparation. Brain tissue was pooled from adult C57BL/6WT or �2�/� mice (61). The dissection was conducted in a cold buffer solution(4 mM Hepes, 1 mM EDTA, and 0.32 M sucrose, pH 7.4), and then the tissue wastransferred to a cold cutting buffer [10 mM Tris (pH 7.4), 320 mM sucrose, 1 mMPMSF, and 1� a protease inhibitor mixture (Roche)] for homogenization at amedium speed. The homogenate was centrifuged at 700 � g for 10 min at 4°C,and the supernatant fraction was collected while the pellet was homogenizedand recentrifuged. Supernatants were subject to ultracentrifugation at40,000 � g for 60 min at 4°C, and the pellet was solubilized overnight at 4°Cin a solution of 100 mM Tris (pH 7.4), 1% Triton X-100, and 1� the proteaseinhibitor mixture. Protein concentrations were determined with the Bradfordreagent kit (Bio-Rad).

Preparation of the GST Fusion Protein and Pulldown Procedures. GST fusionprotein constructs encoding the M3–M4 loop of the �2 subunit [GST-�2(M3–M4)] were generated corresponding to amino acid residues 326–454 of

the mouse �2 subunit. Bacterial transformations and protein inductions wereperformed as described in ref. 62. Bacterial BL21 (DE3) cells were grownovernight in 2� YTA medium using ampicillin (Sigma) selection. Proteinproduction was induced with 5 mM isopropyl-�-D-thiogalactopyranoside(Sigma) for 4 h. The bacteria were harvested by centrifugation at 7,500 � g for15 min at 4°C, and the pellet was suspended in cold HNG buffer [20 mM Hepes(pH 7.4), 1 mM EDTA, 1 mM EGTA, 1 mM MgCl2, 150 mM NaCl, 10% (vol/vol)glycerol, 1% Triton X-100, and 1� the protease inhibitor mixture]. The cellswere lysed on ice by ultrasonication, and the lysate was centrifuged at16,000 � g for 10 min at 4°C.

Purification of fusion proteins was performed by using a Sepharose beadmatrix (Amersham Pharmacia), and protein bound matrix was washed threetimes with HNG buffer before pulldown experiments. Pulldown experimentswere performed by incubating the Sepharose matrix-immobilized GST/GST-�2(M3–M4) with the solubilized protein preparations overnight at 4°C [50 �g ofimmobilized GST/GST-�2 (M3–M4) per 1,500 �g of solubilized protein prepa-rations]. The beads were then washed three times with a cold wash buffer (1�PBS, 0.1% Triton X-100, and 1� protease inhibitor mixture) before proteinswere eluted by boiling in 1% SDS for 5 min. Proteins were separated by anSDS/polyacrylamide (4–20%) gradient gel (Invitrogen), and the gel wasstained with 0.1% (wt/vol) Coomassie blue R350 solution (Sigma) [20% (vol/vol) methanol, and 10% (vol/vol) acetic acid] for protein visualization ortransferred onto a nitrocellulose membrane for Western blot analysis.

Immunoprecipitation and Western Blot Analysis. HEK 293 cells were maintainedin Dulbecco’s modified Eagle’s medium and 10% FBS (Gibco). Cellular trans-fection of the nAChR was done by using plasmids encoding human �2 and �4subunits as described in ref. 63. Immunoprecipitation experiments were per-formed by using monoclonal antibody mAb 270, which has been characterized(29). Solubilized proteins (1,500 �g) were mixed with 2 �g of mAb 270overnight at 4°C. Immunocomplexes were then captured with protein A/GDynabeads (Invitrogen) for 1 h at 4°C before being washed three times withthe wash buffer. Proteins were eluted with LDS buffer (Invitrogen) at 70°C for15 min and then separated by SDS/PAGE. Gels were stained with Coomassie ortransferred onto a nitrocellulose membrane.

Western blot analysis was performed with the following primary antibod-ies: anti-�2 (H-92) (Santa Cruz Biotechnology); anti-�4 (H-133) (Santa CruzBiotechnology); anti-clathrin HC (Santa Cruz Biotechnology); anti-NSF (Up-state Biotechnology); anti-dynamin 1 (PharMingen–Becton Dickinson); anti-synaptotagmin 1 (Chemicon); anti-GTP-binding protein (G protein) Go� (SantaCruz Biotechnology), anti-GIRK1 (Alomone Laboratories); and anti-GST (Am-ersham Pharmacia). Primary antibodies were complexed with species-specifichorseradish peroxidase-conjugated secondary antibodies (Jackson Immu-noResearch) and then visualized by enhanced chemiluminescence (ECL) usingthe ECL plus kit (Amersham Pharmacia).

Protein Identification with MS and MS/MS Analysis. Selected bands wereexcised into 96-well plates. Destaining, reduction, alkylation, and trypsindigestion of the proteins were followed by peptide extraction and carried outby using the Progest Investigator (Genomic Solutions). After desalting (C18-�ZipTip; Millipore), peptides were eluted directly by using ProMS Investigator(Genomic Solutions) onto a 96-well stainless steel MALDI target plate (AppliedBiosystems) by a 0.5-�l CHCA matrix (10 mg/ml in 70% CAN, 30% H2O, 0.1%trifluoroacetic acid).

MS and MS/MS data for protein identification were obtained by using aMALDI-TOF-TOF instrument (4800 proteomics analyzer; Applied Biosystems).For positive-ion reflector mode spectra, 3,000 laser shots were averaged. ForMS calibration, autolysis peaks of trypsin ([M�H]� � 842.5100 and 2,211.1046)were used as internal calibrates. Monoisotopic peak masses were determinedwithin the mass range of 800–4,000 Da, and up to 12 of the most intense ionsignals were selected as precursors for MS/MS acquisition, excluding thetrypsin autolysis peaks and the matrix ion signals.

In MS/MS positive ion mode, 4,000 spectra were averaged, collision energywas 2 kV, collision gas was air, and default calibration was set by using theGlu1-Fibrino-peptide B ([M�H]� � 1,570.6696) spotted onto 14 positions ofthe MALDI target. Combined peptide mass fingerprinting PMF and MS/MSqueries were performed by using the MASCOT search engine 2.1 (MatrixScience, Ltd.) embedded into GPS-Explorer Software 3.5 (Applied Biosystems)on the NCBI database with the following parameter settings: 50 ppm massaccuracy, trypsin cleavage, one missed cleavage allowed, carbamidomethyla-tion set as fixed modification, oxidation of methionine was allowed as vari-able modification, MS/MS fragment tolerance was set to 0.3 Da. Protein hitswith MASCOT protein score �70 and a GPS Explorer protein confidence index�95% were used for further manual validation.

20574 � www.pnas.org�cgi�doi�10.1073�pnas.0710314104 Kabbani et al.

Page 6: Intracellular complexes of the 2 subunit of the nicotinic ... complexes of the 2 subunit of the nicotinic acetylcholine receptor in brain identified by proteomics Nadine Kabbani*†,

ACKNOWLEDGMENTS. We thank Drs. Pierre-Jean Corringer, Thomas Grutter,and Abdelkader Namane for helpful discussions. This work was supported bythe Institut Pasteur and the Centre National de la Recherche Scientifique

(J.-P.C. and N.K.) and in part by an award from Philip Morris, Inc., and PhilipMorris International (to N.K.) and National Institutes of Health Grants MH068789 (to R.L.) and NS11323 (to J.M.L.).

1. Lerman C (2006) LDI Issue Brief 11:1–4.2. Viveros MP, Marco EM, File SE (2006) Neurosci Biobehav Rev 30:1161–1181.3. Changeux JP, Bessis A, Bourgeois JP, Corringer PJ, Devillers-Thiery A, Eisele JL,

Kerszberg M, Lena C, Le Novere N, Picciotto M, et al. (1996) Cold Spring Harb SympQuant Biol 61:343–362.

4. Champtiaux N, Gotti C, Cordero-Erausquin M, David DJ, Przybylski C, Lena C, ClementiF, Moretti M, Rossi FM, Le Novere N, et al. (2003) J Neurosci 23:7820–7829.

5. Maskos U, Molles BE, Pons S, Besson M, Guiard BP, Guilloux JP, Evrard A, Cazala P,Cormier A, Mameli-Engvall M, et al. (2005) Nature 436:103–107.

6. Whiting P, Esch F, Shimasaki S, Lindstrom J (1987) FEBS Lett 219:459–463.7. Le Novere N, Corringer PJ, Changeux JP (2002) J Neurobiol 53:447–456.8. Ren XQ, Cheng SB, Treuil MW, Mukherjee J, Rao J, Braunewell KH, Lindstrom JM,

Anand R (2005) J Neurosci 25:6676–6686.9. Bocquet N, Prado de Carvalho L, Cartaud J, Neyton J, Le Poupon C, Taly A, Grutter T,

Changeux JP, Corringer PJ (2007) Nature 445:116–119.10. Champtiaux N, Changeux JP (2002) Curr Drug Targets CNS Neurol Disord 1:319–330.11. Stevens TR, Krueger SR, Fitzsimonds RM, Picciotto MR (2003) J Neurosci 23:10093–

10099.12. Granon S, Faure P, Changeux JP (2003) Proc Natl Acad Sci USA 100:9596–9601.13. Mameli-Engvall M, Evrard A, Pons S, Maskos U, Svensson TH, Changeux JP, Faure P

(2006) Neuron 50:911–921.14. Picciotto MR, Zoli M, Zachariou V, Changeux JP (1997) Biochem Soc Trans 25:824–829.15. Ben-Shlomo I, Yu Hsu S, Rauch R, Kowalski HW, Hsueh AJ (2003) Sci STKE 2003, RE9.16. Abul-Husn NS, Devi LA (2006) J Pharmacol Exp Ther 318:461–468.17. Bockaert J, Fagni L, Dumuis A, Marin P (2004) Pharmacol Ther 103:203–221.18. Bray D (1998) Annu Rev Biophys Biomol Struct 27:59–75.19. Husi H, Ward MA, Choudhary JS, Blackstock WP, Grant SG (2000) Nat Neurosci 3:661–

669.20. Farr CD, Gafken PR, Norbeck AD, Doneanu CE, Stapels MD, Barofsky DF, Minami M,

Saugstad JA (2004) J Neurochem 91:438–450.21. Hedman E, Widen C, Asadi A, Dinnetz I, Schroder WP, Gustafsson JA, Wikstrom AC

(2006) Proteomics 6:3114–3126.22. Becamel C, Gavarini S, Chanrion B, Alonso G, Galeotti N, Dumuis A, Bockaert J, Marin

P (2004) J Biol Chem 279:20257–20266.23. Kabbani N, Levenson R (2007) Eur J Pharmacol 572:83–93.24. Kim M, Jiang LH, Wilson HL, North RA, Surprenant A (2001) EMBO J 20:6347–6358.25. Lin L, Jeanclos EM, Treuil M, Braunewell KH, Gundelfinger ED, Anand R (2002) J Biol

Chem 277:41872–41878.26. Nashmi R, Dickinson ME, McKinney S, Jareb M, Labarca C, Fraser SE, Lester HA (2003)

J Neurosci 23:11554–11567.27. Xu J, Zhu Y, Heinemann SF (2006) J Neurosci 26:9780–9793.28. Conroy WG, Liu Z, Nai Q, Coggan JS, Berg DK (2003) Neuron 38:759–771.29. Whiting P, Lindstrom J (1987) Proc Natl Acad Sci USA 84:595–599.30. Sallette J, Pons S, Devillers-Thiery A, Soudant M, Prado de Carvalho L, Changeux JP,

Corringer PJ (2005) Neuron 46:595–607.31. Moser N, Mechawar N, Jones I, Gochberg-Sarver A, Orr-Urtreger A, Plomann M, Salas

R, Molles B, Marubio L, Roth U, et al. (2007) J Neurochem 102:479–492.

32. Whiteaker P, Cooper JF, Salminen O, Marks MJ, McClure-Begley TD, Brown RW, CollinsAC, Lindstrom JM (2006) J Comp Neurol 499:1016–1038.

33. Cho S, Park SG, Lee DH, Park BC (2004) J Biochem Mol Biol 37:45–52.34. Kim H, Jo S, Song HJ, Park ZY, Park CS (2007) Proteomics 7:2591–2602.35. Rogers SW, Gregori NZ, Carlson N, Gahring LC, Noble M (2001) Glia 33:306–313.36. Chaudhuri A, Chant J (2005) BioEssays 27:958–969.37. Bader JS, Chaudhuri A, Rothberg JM, Chant J (2004) Nat Biotechnol 22:78–85.38. Armstrong JD, Pocklington AJ, Cumiskey MA, Grant SG (2006) Proteomics 6:4724–

4731.39. Gotti C, Zoli M, Clementi F (2006) Trends Pharmacol Sci 27:482–491.40. Kuryatov A, Luo J, Cooper J, Lindstrom J (2005) Mol Pharmacol 68:1839–1851.41. Hill JA, Jr, Zoli M, Bourgeois JP, Changeux JP (1993) J Neurosci 13:1551–

1568.42. Hwang YY, Li MD (2006) Proteomics 6:3138–3153.43. Barnes EM, Jr (2000) Life Sci 66:1063–1070.44. Ryazanov AG, Rudkin BB, Spirin AS (1991) FEBS Lett 285:170–175.45. Cho DI, Oak MH, Yang HJ, Choi HK, Janssen GM, Kim KM (2003) Life Sci 73:2991–3004.46. McClatchy DB, Knudsen CR, Clark BF, Kahn RA, Hall RA, Levey AI (2002) J Biol Chem

277:29268–29274.47. Ungermann C, Langosch D (2005) J Cell Sci 118:3819–3828.48. Cong M, Perry SJ, Hu LA, Hanson PI, Claing A, Lefkowitz RJ (2001) J Biol Chem

276:45145–45152.49. Zou S, Li L, Pei L, Vukusic B, Van Tol HH, Lee FJ, Wan Q, Liu F (2005) J Neurosci

25:4385–4395.50. Liu Z, Tearle AW, Nai Q, Berg DK (2005) J Neurosci 25:1159–1168.51. Moretti M, Vailati S, Zoli M, Lippi G, Riganti L, Longhi R, Viegi A, Clementi F, Gotti C

(2004) Mol Pharmacol 66:85–96.52. Yoshihara M, Montana ES (2004) Neuroscientist 10:566–574.53. Catterall WA (1999) Ann NY Acad Sci 868:144–159.54. Linial M, Ilouz N, Parnas H (1997) J Physiol (London) 504:251–258.55. Fischer H, Liu DM, Lee A, Harries JC, Adams DJ (2005) J Neurosci 25:3571–3577.56. Oyaizu M, Narahashi T (1999) Brain Res 822:72–79.57. Quarta D, Ciruela F, Patkar K, Borycz J, Solinas M, Lluis C, Franco R, Wise RA,

Goldberg SR, Hope BT, et al. (2007) Neuropsychopharmacology 32:35– 42.

58. Wickman K, Karschin C, Karschin A, Picciotto MR, Clapham DE (2000) J Neurosci20:5608–5615.

59. Catterall WA, Hulme JT, Jiang X, Few WP (2006) J Recept Signal Transduct Res26:577–598.

60. Coronas V, Durand M, Chabot JG, Jourdan F, Quirion R (2000) Neuroscience 98:213–219.

61. Picciotto MR, Zoli M, Lena C, Bessis A, Lallemand Y, Le Novere N, Vincent P, Pich EM,Brulet P, Changeux JP (1995) Nature 374:65–67.

62. Korf U, Kohl T, van der Zandt H, Zahn R, Schleeger S, Ueberle B, Wandschneider S,Bechtel S, Schnolzer M, Ottleben H, et al. (2005) Proteomics 5:3571–3580.

63. Kabbani N, Negyessy L, Lin R, Goldman-Rakic P, Levenson R (2002) J Neurosci 22:8476–8486.

Kabbani et al. PNAS � December 18, 2007 � vol. 104 � no. 51 � 20575

NEU

ROSC

IEN

CE