Biochem. J. (2011) 438, 255–263 (Printed in Great Britain) doi:10.1042/BJ20110801 255

Crystal structure of the glutamate receptor GluA1 N-terminal domainGuorui YAO*†1, Yinong ZONG†1, Shenyan GU†, Jie ZHOU†, Huaxi XU*†, Irimpan I. MATHEWS‡ and Rongsheng JIN†2

*Fujian Provincial Key Laboratory of Neurodegenerative Disease and Aging Research, College of Medicine, Xiamen University, Xiamen, Fujian 361005, China, †Center forNeuroscience, Aging and Stem Cell Research, Sanford-Burnham Medical Research Institute, 10901 North Torrey Pines Road, La Jolla, CA 92037, U.S.A., and ‡Stanford SynchrotronRadiation Lightsource, 2575 Sand Hill Road, Menlo Park, CA 94025, U.S.A.

The AMPA (α-amino-3-hydroxy-5-methylisoxazole-4-propionicacid) subfamily of iGluRs (ionotropic glutamate receptors) isessential for fast excitatory neurotransmission in the centralnervous system. The malfunction of AMPARs (AMPA receptors)has been implicated in many neurological diseases, includingAlzheimer’s disease, Parkinson’s disease and amyotrophic lateralsclerosis. The active channels of AMPARs and other iGluRsubfamilies are tetramers formed exclusively by assembly ofsubunits within the same subfamily. It has been proposed thatthe assembly process is controlled mainly by the extracellularATD (N-terminal domain) of iGluR. In addition, ATD hasalso been implicated in synaptogenesis, iGluR trafficking andtrans-synaptic signalling, through unknown mechanisms. We

report in the present study a 2.5 Å (1 Å = 0.1 nm) resolutioncrystal structure of the ATD of GluA1. Comparative analyses ofthe structure of GluA1-ATD and other subunits sheds light on ourunderstanding of how ATD drives subfamily-specific assembly ofAMPARs. In addition, analysis of the crystal lattice of GluA1-ATD suggests a novel mechanism by which the ATD mightparticipate in inter-tetramer AMPAR clustering, as well as intrans-synaptic protein–protein interactions.

Key words: α-amino-3-hydroxy-5-methylisoxazole-4-propionicacid receptor (AMPAR), glutamate receptor, ion channel,N-terminal domain (ATD), structural biology.

INTRODUCTION

iGluRs (ionotropic glutamate receptors) are ligand-gated ionchannels that form transmembrane cation-permeable channels.iGluRs exist as three distinct sub-families according to theiragonist specificity and amino acid sequence: AMPARs (α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptors), kain-ate receptors and NMDARs (N-methyl-D-aspartate receptors).Activation of iGluRs is critical for the induction of some formsof LTP (long-term potentiation), a type of synaptic plasticityassociated with learning and memory [1,2]. iGluRs play crucialroles in normal neuronal function and in neurological disease,and thus are being actively pursued as therapeutic targets forthe treatment of amyotrophic lateral sclerosis, neuropathic pain,major depression, Alzheimer’s disease and Parkinson’s disease.

Despite having divergent functional properties, iGluR familymembers adopt a common modular architecture. A typicaliGluR subunit contains four distinct domains: an extracellularATD (N-terminal domain; ∼400 residues), an extracellular LBD(ligand-binding domain; ∼300 residues), three TMDs [TM(transmembrane) domains] (TM1–TM3) plus a re-entrant poreloop (P) and an intracellular C-terminal domain (Figure 1a).Active iGluR channels are tetramers formed exclusively byassembly of subunits within the same subfamily. This process isthought to be mediated in part by the ATD through a mechanismthat is not fully understood.

Historically, structural biology has made significantcontributions to our understanding of iGluR function. The firstcrystal structure of an iGluR, the LBD of GluA2, was determinedin 1998 [3]. Since then, more than a hundred high-resolutioncrystal structures of iGluR LBD have been reported, includingsubunits from all three subfamilies of iGluR either in the apo

form or in complex with a variety of agonists, antagonistsand modulators. Collectively, these structures have establishedthe detailed molecular mechanisms underlying iGluR channelactivation, inhibition and desensitization [4,5]. Nevertheless,the first crystal structure of a full-length iGluR ion channel, thehomomeric GluA2, was not accomplished until 2009 [6]. This ex-traordinary advance has yielded unprecedented information on themolecular architecture and symmetry of iGluR. Structural studieson full-length iGluRs are exceptionally difficult to perform, and todate have yielded only a single snapshot of a homomeric GluA2bound with an antagonist. It is therefore not surprising that ourunderstanding of iGluR structure and function continues to bederived from studies with isolated recombinant ATDs and LBDs.

In contrast to the well-characterized LBD, much less is knownabout the molecular properties of the ATD, even though thisdomain is crucial for the physiological function of iGluRs. Forexample, the spontaneous mouse mutation hotfoot is a recessivemutation characterized by cerebellar ataxia and jerky movementof the hind limbs, and is often caused by in-frame deletions ofvarious regions of the ATD of GluRδ2 [7]. The functionalactivities of ATD fall into three general categories. First, it iswidely accepted that the iGluR-ATD guides receptor subfamily-specific assembly, ensuring that only subunits within the samesubfamily assemble with one another [8–11]. Interestingly, thesubfamily-specific assembly of tetrameric voltage-gated ShakerK+ channels is also determined by the N-terminal T1 domain[12,13]. Secondly, the NMDAR-ATD modulates receptor functionby providing binding sites for allosteric modulators that includeprotons, zinc ions, polyamines and small organic molecules suchas ifenprodil [1,14–16]. However, no small molecules have beenshown to bind the AMPAR-ATD or the ATD of kainate receptors.Thirdly, the ATD resides within the synaptic cleft, and thus

Abbreviations used: ATD, N-terminal domain; AMPAR, α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor; iGluR, ionotropic glutamatereceptor; IP, immunoprecipitation; LIVBP, leucine/isoleucine/valine-binding protein; LBD, ligand-binding domain; mGluR, metabotropic glutamate receptor;Ni-NTA, Ni2 + -nitrilotriacetate; NMDAR, N-methyl-D-aspartate receptor; PEG, poly(ethylene glycol); S-loop, specificity loop; TMD, transmembrane domain.

1 These authors contributed equally to this work2 To whom correspondence should be addressed (email rjin@sanfordburnham.org).The structural co-ordinates and diffraction data for GluA1-ATD reported will appear in the PDB under accession code 3SAJ.

c© The Authors Journal compilation c© 2011 Biochemical Society

256 G. Yao and others

Figure 1 The structure of GluA1-ATD

(a) iGluRs have a modular architecture that includes the ATD, LBD, TMD (transmembrane ion channel domain) and the C-terminal intracellular region. The ligand-binding pocket on LBD is definedas a red pentagon. Shown is one subunit on the basis of crystal structure of the full-length tetrameric GluA2 [6]. (b) Ribbon diagram of the structure of GluA1-ATD (PDB code 3SAJ). The secondarystructures are numbered in a sequential order and the α-helix, β-strand and loop are coloured in yellow, blue and green respectively. The missing loop between helices α9 and α10 is shown as abroken line. (c) View of the structure of GluA1-ATD from the side following a rotation of ∼90◦ around a vertical axis.

may be involved in protein–protein interactions. By doing so,the ATD could regulate both dendritic spine morphogenesis andpresynaptic stability [17,18]. Some ATD-binding partners havebeen identified, including neuronal pentraxins (Narp, NP1 andNPR) [19,20] and N-cadherin [21] that bind to AMPAR-ATD,and an ephrin receptor that binds to NMDAR [22].

Our limited knowledge of the structure and function of the ATDis in part due to the challenge associated with recombinant proteinproduction. Large-scale protein expression of the ATD usinginsect and mammalian cells has been successful only recently,and has resulted in a number of ATD structures, including theATDs of GluA2, GluK2, GluK3, GluK5 and GluN2B [23–28].Structural analyses have revealed several unique features of ATDswithin each subfamily of iGluRs with respect to the protomerstructure, subfamily-specific subunit assembly and competencefor ligand binding. These findings have clarified how the ATDprevents subunits from different subfamilies from assemblinginto a tetramer. However, the role played by the ATD in drivingassembly of different subunits within a subfamily remains unclear.This is a critical process, as most native AMPAR channels areheteromers that co-assemble with the GluA2 subunit, whichserves to decrease channel permeability to calcium ions. Thisproblem has motivated us to perform structural studies on othersubunits of AMPARs to complement the structure of the GluA2-ATD, which we reported in 2009 [23]. In the present study wereport the structure of the GluA1-ATD at 2.5 Å (1 Å = 0.1 nm) res-olution. The crystal structures of GluA3 and GluN1-ATD werereported while this manuscript was in preparation [29,30].

EXPERIMENTAL

Protein expression and purification

The rat GluA1-ATD (Pro4–Ala374) (amino acid numberingcorresponds to the mature polypeptide after cleavage of the

endogenous signal peptide) was cloned into a modified pFastBacvector (Invitrogen). The human placental alkaline phosphatasesignal peptide was added to the N-terminus of GluA1-ATD.A C-terminal Myc-tag and His8-tag following a PreScissionprotease cleavage site were introduced to facilitate purificationand characterization. The protein was expressed and secreted fromSpodoptera frugiperda (Sf9) insect cells, and purified from mediaby Ni-NTA (Ni2 + -nitrilotriacetate) affinity chromatography(Qiagen). The C-terminal tags were removed by PreScissionprotease treatment after buffer exchange into 20 mM Tris/HCl,pH 8.0, 150 mM NaCl and 1 mM EDTA. The cleaved proteinwas then applied to an ion-exchange column (Mono STM 5/50 GL;GE Healthcare) after a 5-fold dilution to pH 6.0, and subsequentlyeluted with a NaCl gradient. The peak fractions containing GluA1-ATD were exchanged into a buffer containing 20 mM Tris/HCl,pH 8.0, 150 mM NaCl and 1 mM EDTA, and concentrated to∼10 mg/ml for crystallization.

For co-IP (co-immunoprecipitation) studies, GluA1-ATD wasexpressed and purified as described above, except that the Myc-tagand His-tag were not excised. Cloning, expression and purificationof GluA2-ATD were carried out as described previously [23].As a negative control, a ∼40 kDa bacterial protein with aC-terminal Myc-tag was expressed in Escherichia coli andpurified to homogeneity. All proteins were exchanged into thesame buffer (25 mM Tris/HCl, pH 8.0, 200 mM NaCl and 0.2%Nonidet P40) before the assay was performed.

Crystallization and diffraction data collection

Initial crystallization screens were carried out using aPhoenix crystallization robot (Art Robbins Instrument) andcommercial high-throughput crystallization screen kits. Afterextensive manual optimization, the best GluA1-ATD crystalswere grown by hanging-drop vapor diffusion at 18 ◦C, inwhich the protein (10 mg/ml) was mixed in 1:1 ratio with

c© The Authors Journal compilation c© 2011 Biochemical Society

Structure of the GluA1 N-terminal domain 257

a reservoir solution containing 100 mM sodium acetate,pH 5.0, 24% PEG [poly(ethylene glycol)] 3350 and 200 mMMgCl2. Single crystals were obtained only by micro-seeding.The crystals were cryoprotected in the reservoir solutionsupplemented with 15% glycerol. The X-ray diffraction datawere collected at − 173 ◦C at the microfocus beam line12–2, Stanford Synchrotron Radiation Lightsource (SSRL),using a Dectris Pilatus 6M Pixel detector. All data setswere processed and scaled by using iMOSFLM [31]. Datacollection statistics are summarized in Supplementary Table S1(at http://www.BiochemJ.org/bj/438/bj4380255add.htm).

Structure determination and refinement

The structure of GluA1-ATD was determined by molecularreplacement using Phaser [32]. One protomer of GluA2-ATD(PDB code 3H5V) was used as the search model to locatethe four molecules of GluA1-ATD in one asymmetric unit.The preliminary structural model was subsequently refined withPhenix 1.5 [33] and re-built with COOT [34] in an iterativemanner. Refinement progress was monitored with the free Rvalue using a 5% randomly selected test set [35]. The GluA1-ATD structure was refined to 2.5 Å with Rwork/Rfree = 0.22/0.28.The structural refinement statistics are listed in SupplementaryTable S1. The co-ordinates and diffraction data for GluR1-ATDwere deposited in the Protein Data Bank (PDB code 3SAJ).Figures were prepared with PyMOL (http://www.pymol.org).

Co-IP

Co-IP assays were carried out in a buffer containing 25 mMTris/HCl, pH 8.0, 200 mM NaCl and 0.2 % Nonidet P-40. Theantibodies used for co-IP and Western blotting were a mouse anti-GluA2-ATD antibody (MAB397, Millipore) and a mouseanti-Myc antibody (9B11, Cell Signaling Technology). The sameamounts of proteins, antibodies and Protein A resins wereused in all experiments. The IP reactions were performed at4 ◦C overnight. After extensive washing, the immunocomplexeswere separated by SDS/PAGE, transferred on to a nitrocellulosemembrane and immunoblotted with anti-GluA2 or anti-Mycantibody. Western blots were visualized with an alkalinephosphatase-conjugated rabbit anti-mouse secondary antibody(Bio-Rad Laboratories).

RESULTS AND DISCUSSION

Purification, crystallization and structure determination

Of all the modular domains of iGluR, the ATD has thelargest sequence diversity across different families, showing only0.2% sequence identity. The ATD also has the lowest sequenceidentity within each subfamily; for example, AMPARs have∼35% identity in the ATD, compared with 80 % identity inthe LBD and 87 % in the transmembrane ion channel region[5]. Interestingly, a structure similarity search showed that theATD adopts a structural fold similar to that of the bacterialLIVBP (leucine/isoleucine/valine-binding protein; a member ofthe periplasmic ligand-binding protein family) [36], the LBDof mGluR (metabotropic glutamate receptor) [37] and the LBD ofNPR (natriuretic peptide receptor) [38]. Nevertheless, the overallsequence identity between the ATD and these proteins is quitelow (below 15 %).

We expressed the GluA1-ATD (residues Pro4–Ala374;numbering based on the mature GluA1) as a secreted form inSf9 insect cells. The secreted protein was directly purified from

the media by Ni-NTA affinity chromatography followed by ion-exchange chromatography. The final yield was approximately0.3 mg of purified protein per litre of cell culture.

Multiple crystal forms of GluA1-ATD were identifiedby robotic high-throughput screening followed by manualoptimization. Most of the crystals diffracted weakly to ∼6Å resolution, consistent with our previous finding [23]. Afterextensive optimization of crystallization conditions, we focusedon a condition using PEG 3350 as a precipitant, which yieldedcrystals that routinely diffracted to ∼4 Å. Interestingly, theplate-like crystals grew from this condition only after micro-seeding. The original seed crystals were obtained from a dropset up by robot. This drop showed severe precipitation aftersetup; tiny crystals grew after a month at 18 ◦C. Micro-seedingusing these crystals yielded plate-like crystals that were furtheroptimized. Single crystals grown under these conditions showedgreat variation in their thickness, and easily stacked on top of eachother. This property significantly slowed the progress of structuredetermination. A large number of crystals have been screened andthe best data set diffracted to approximately 2.5 Å (SupplementaryTable S1). The structure of GluA1-ATD was solved by molecularreplacement using the structure of GluA2-ATD (PDB code 3H5V)as the search model. The structure was refined at 2.5 Å resolution(Rwork/Rfree = 0.22/0.28) and shows excellent crystallographic andstereochemical statistics (Supplementary Table S1).

Overall architecture of GluA1-ATD

The crystal of GluA1-ATD shows four molecules in oneasymmetric unit forming two pairs of dimers. Excellent electrondensities were observed for 365 residues, but residues His260–Trp265 were not observed. GluA1-ATD has a two-domain flytrap-like structure (Figure 1b). The N-terminal lobe (L1) and theC-terminal lobe (L2) each have an α/β topology with the centralβ-sheets surrounded by α-helices, and are connected bythree short inter-domain loops. All four crystallographicallyindependent protomers have a similar conformation. Closeinspection reveals that the overall structure of GluA1-ATD issimilar to that of GluA2 and GluA3 [23–25,29]. Pair-wisecomparisons of Cα atoms yielded rmsd (root mean squaredeviation) values of 0.8 Å (GluA1 compared with GluA2) and 0.9Å (GluA1 compared with GluA3) in the core of the L1 domain.More diversity was observed in the L2 domain; ∼1.3 Å betweenGluA1 and A2, and ∼1.5 Å between GluA1 and A3.

Despite the conserved structural cores, there are five prominentdifferences among GluA1–GluA3. First, a loop that connectshelices α9 and α10 (His260–Trp265) has no visible electron densityon GluA1-ATD, suggesting a highly flexible conformation in thisarea (Figure 1b). As a result, the helix α9 on GluA1-ATD swingsaway from the core of L1 by as far as 4.5 Å at the C-terminus of α9,compared with GluA2. Secondly, GluA1 has a shorter helix α6(Glu165–Phe171) on L2 than does GluA2 (Glu169–Glu179). Further-more, in comparison with GluA2, GluA1 has a shorter loop linkingβ7 and α6 (Ile159–Glu164 on GluA1 compared with Val159–Asp168

on GluA2) and a longer linker between α6 and β8 (Gln172–Glu179

on GluA1 compared with Leu180–Arg184 on GluA2). In contrast,GluA2 and GluA3 adopt very similar conformations in this region.

The third structural difference among GluA1–GluA3 is in theS-loop (specificity loop) that links helices α10 and α11 and isattached to the core structure on L1 through a disulfide bond(Figure 1b). It is termed the S-loop because it plays a key role ininteractions at the ATD dimer interface, and because its sequenceand length is more conserved within than between receptorsubfamilies [23,26]. This loop adopts a different conformation

c© The Authors Journal compilation c© 2011 Biochemical Society

258 G. Yao and others

in GluA1 (Ile294–Trp313) than in GluA2 (Ile298–Trp317) or GluA3(Val301–Trp320), despite an almost identical amino acid sequence[23]. Unexpectedly, the S-loop in GluA1 participates in extensiveinter-dimer interactions. The potential physiological relevance ofthis unique feature will be discussed further below. Fourthly,a loop connecting helices α7 and β9 on L2 adopts a differentconformation to the equivalent loop in GluA2. It is worth notingthat in GluA2 this loop is mainly responsible for the dimer-of-dimers interaction in the ATD in the context of a tetramericchannel [6]. Such assembly in the ATD will not form if this loopadopts the conformation as observed in GluA1-ATD.

Finally, we observed extensive trans-domain interactions in theclamshell cleft on GluA1-ATD (see Supplementary Figure S1 athttp://www.BiochemJ.org/bj/438/bj4380255add.htm). Four pairsof hydrogen bonds and salt bridges stabilize a partially closedconformation involving residues Arg267 and Tyr270 on L1 andSer188, Asp219 and Gln236 on L2. In addition, Phe95 on L1 and Arg135

on L2 form a cation π interaction at the front edge of theclamshell (Supplementary Figure S1). Among these interactions,only one hydrogen bond formed between Asp219 and Tyr270 isconserved on GluA2; none of these interactions are observed inGluA3. Nevertheless, the ATDs of GluA1–GluA3 all adopt similarpartially closed conformations, with only a moderate domaintwisting movement of approximately 3 to 6 degrees between L1and L2, which suggests that the trans-domain interactions are notthe major force stabilizing the closed clamshell conformation.In sharp contrast, the homologous structures, e.g. mGluR-LBDor LIVBP, could have a domain closure up to 50 degrees uponligand binding in the cleft. A similar mechanism is proposed forNMDAR, in which binding of a modulator in the ATD cleft couldtrigger the close-open movement of the ATD and subsequentlyinduce structural rearrangement at the ATD–ATD and ATD–LBDinterfaces [39,40]. However, direct evidence for such a mechanismhas not yet been obtained. Interestingly, an unassigned electrondensity was observed in the ATD cleft of GluA2-ATD usingthe 1.75 Å resolution data, which could potentially be a ligand[29]. However, no similar density was observed in another highresolution GluA2-ATD structure (1.8 Å) or in any of the knownstructures of the kainate receptor ATDs at resolutions up to 1.4 Å(GluK5) [24,27].

Dimeric organization of the ATD is not rigid

Both the GluA1 and GluA2 ATDs form tight dimers in solution[23,24]. In the crystal structure, the GluA1-ATD forms two pairsof indistinguishable dimers, with each dimer possessing a non-crystallographic 2-fold symmetry. The L1–L1 dimer interface isformed primarily by helices α2 and α3 in combination with theS-loop lying on the top (Figure 2a). The aromatic ringof Phe50 of the α2 helix of one protomer inserts into ahydrophobic pocket formed by residues Phe82/Ala85/Leu86 onthe α2 helix and Leu306/Ala310 of the S-loop on the otherprotomer. Three hydrogen bonds are observed on the L1–L1 interface, involving Ser81 on α3 and Asp48/Ser49/Phe50 onα2. Although the L1–L1 interface of GluA1-ATD is almostidentical to that of GluA2, two key differences stand out.First, GluA2 Asn54 on α2 that forms a hydrogen bond withthe main chain carbonyl group of Leu310 on the S-loop isreplaced with Tyr54 on GluA1. The large side-chain of Tyr54 notonly abolishes this hydrogen bond, but also pushes back theS-loop and thus negatively affects intra-dimer interaction.Secondly, GluA2 Thr78 forms water-mediated hydrogen bondbetween L1 domains, but this polar interaction is abolishedwhen Thr78 is replaced with Met78 on GluA1. On the GluA1dimer interface, the two Met78 residues are in close contact with

a distance of 6.6 Å between their Cα atoms. The relativelylarge hydrophobic side chain of a methionine residue insteadof threonine at this position is not optimal for a tight L1–L1interaction. Collectively, these observations suggest that Tyr54

and Met78 on GluA1 contribute to the ∼2-fold reduction in dimerassociation affinity of GluA1 compared with GluA2 [23].

Interactions at the L2 dimer interface are primarily hydro-phobic, and are mediated by residues Leu137/Leu140/Leu144/Ala148

on helix α5, all of which protrude into the L2–L2 interface, to-gether with Ala156/Val157 on β7 (Figure 2d). This pattern is almostidentical to that of the GluA2 structure, with one subtle differencebeing that GluA2 Ile157 is replaced with GluA1 Val157. The highlyconserved intra-dimer interactions between GluA1 and GluA2suggest that the predominant hetero-assembly of GluA1 andGluA2 into the same tetrameric channel in vivo is unlikely to bedetermined solely at the ATD level. Other domains, e.g. LBD andTMD, are likely to be required to fulfill the complex assembly [8].

Surprisingly, the L2–L2 interface on GluA3-ATD is quitedifferent to that of GluA1 or GluA2. The key interactingresidues on GluA1, Leu137 and Val157, are replaced with Phe143

and Arg163 respectively on GluA3. In this case, the closelypacked L2–L2 interface that is conserved on GluA1 andGluA2 cannot be maintained because the hydrophobic patchis disrupted by insertion of the large side-chain of Phe143 andby a positively charged Arg163. As a result, the GluA3 L2–L2interface breaks up, and the dimer is maintained primarily bythe conserved L1–L1 interface (see Supplementary Figure S2c athttp://www.BiochemJ.org/bj/438/bj4380255add.htm). This couldexplain the fact that GluA3-ATD has the lowest affinity forhomodimerization among all AMPARs (its Kd value is more than600-fold lower than that of GluA2-ATD) [29].

To better understand the flexibility of dimer organ-ization in the AMPAR family, we superimposed thedimeric ATDs of GluA1–GluA3 on the basis of the Cαatoms of one protomer (Mol-A) and examined theconformational changes of the other protomer (Mol-B)(Figure 3). The protomer structures of GluA1-GluA3 areall very similar (Mol-A, Figures 3a and 3b). The Mol-Bof GluA1 and A2 in a dimer are almost identical, as couldbe predicted from their highly conserved L1–L1 and L2–L2 interfaces. Significant dimer rearrangement is observed onGluA3-ATD where the Mol-B of GluA3 is rotated clockwisealong an axis perpendicular to the dimer interface (Figures 3cand 3d). Rotation and translation up to ∼8 ◦ and ∼7.4 Å areobserved on GluA3-ATD, using as references helices α3 and α5that mediate key interactions on L1 and L2 respectively.

When we compared all six available structures of the non-NMDARs, GluA1, GluA2, GluA3, GluK2, GluK3 and GluK5,we observed a correlation between the stability of the ATD dimerinterface and the receptor’s ability to form a functional homomericion channel. Receptors that have large and stable dimer interfaceson both L1 and L2 domains, such as GluA1, GluA2, GluK2and GluK3, all form functional channels. In contrast,GluA3 and GluK5 both have twisted and significantly weakeneddimer interfaces, and, interestingly, GluA3 has a high propensityto form heteromeric assemblies with other AMPARs, whereasGluK5 requires obligate co-assembly with GluK1–GluK3 to formfunctional channels [27,29]. This suggests that formation of astable ATD dimer through homo- or hetero-dimerization is a keydeterminant for the assembly of a functional tetrameric channel.

We have shown previously that the ATD of GluA1 and GluA2form homodimers in solution, with Kd values of 270 nM and152 nM respectively [23]. In the present study, we observed thatGluA1 and GluA2 ATD could directly interact with each other(Figure 2c). A robust interaction was observed between the ATD

c© The Authors Journal compilation c© 2011 Biochemical Society

Structure of the GluA1 N-terminal domain 259

Figure 2 GluA1-ATD forms a homodimer that is stabilized by extensive intra-dimer interactions

(a) Ribbon diagram of the structure of a GluA1-ATD homodimer viewed perpendicular to the molecular two-fold axis. The two protomers are coloured blue and orange respectively. The dimer interfacecan be divided into two sections that are outlined in red and blue boxes respectively. (b and d) Close view of the dimer interface located in L1 (red frame) and L2 (blue frame). Key residues thatparticipate in intra-dimer interactions are shown as sticks. Broken lines indicate hydrogen bonds. (c) ATD mediates heterodimerization between GluA1 and GluA2. Purified recombinant GluA1-ATDwith a C-terminal Myc-tag and GluA2-ATD were used for IP. A specific anti-GluA2-ATD antibody (MAB397, Millipore) and an anti-Myc antibody (9B11, Cell Signaling Technology) were used for IPand Western blotting (IB). A non-related Myc-tagged protein (∼40 kDa) was used as a negative control (labeled as #). Please note that the ATD bands were smeared due to glycosylation. Signalsoutside the blue boxes are non-specific due to antibody heavy chains (*) or an unknown protein that came from the anti-GluA2 antibody (**). ‘Input’ represents the starting materials for IP.

of GluA1 and GluA2 by co-IP when relatively high concentrationsof recombinant proteins were used (∼4 μM for each). It is likelythat GluA1 and GluA2 ATD form heterodimers, because notetrameric species was observed when a mixture of GluA1 andGluA2 ATDs was analysed by analytical ultracentrifugation ata similar concentration (results not shown). The interaction wasmuch weaker when co-IP was performed at relatively low proteinconcentration, suggesting only a weak interaction between GluA1and GluA2 at the ATD region. These findings are consistentwith a model that heteromerization of AMPAR is mediated co-operatively by interactions at multiple regions, including the ATD,the LBD and the transmembrane channel [8]. Interestingly, ithas been reported recently that GluA1 and GluA2 ATDs formheterodimers with a very high affinity (Kd ∼0.4 nM) [25], and thesame group estimated that GluA2-ATD forms homodimers witha similar Kd (∼1.8 nM). It raises a question as to how these twohigh-affinity binding events, the homodimerization of GluA2 andthe heterodimerization between GluA1 and GluA2, are fine-tuned.

L2 domain has a high degree of flexibility

To investigate further the structural flexibility of the ATD, weanalysed the B-factor distribution in each of the three membersof AMPAR with known structures. The B-factors in a proteincrystal structure reflect the fluctuation of each atom about itsaverage position, and are important indicators of the flexibilityand dynamics of a protein. Since measured B-factors in differentstructures are affected by differences in crystal handling, datacollection and structure refinement, we examined only thedistribution of B-factors within each structure.

The B-factor analysis shows that across GluA1–GluA3, thecore of the L1 domain is the most stable, whereas the L2 domain

has greater flexibility (see Supplementary Figure S2). This isconsistent with the dynamic analysis of GluA3-ATD showingthat its L2 domain is very mobile [29]. As expected, the peripheralregions of GluA1-ATD that show large structural differences withGluA2 and GluA3, including α6, α9 and the S-loop, all showhigh B-factors. Interestingly, GluA1-ATD seems to have higherintrinsic flexibility on the L2 domain compared with GluA2,despite their highly conserved intra-dimer interactions. Althoughthe precise explanation for this difference awaits further study, thehigh degree of flexibility of the L2 domain of GluA1 may explainits preference to form heteromeric channels with GluA2 ratherthan homomeric channels. The flexible structure of GluA1-ATDcould also explain the difficulty in crystallizing the proteins thatwe have encountered.

The combination of a stable L1 domain and a mobile L2 domainon the ATD is reminiscent of the ligand-binding property of NPR,which has an LBD closely related to the ATD. NPR-LBD sharesa similar dimer assembly to the ATD, except that the L2 domainsare separated in a manner similar to that observed on GluA3-ATD (Supplementary Figure S2c) [38]. The natriuretic peptideligand binds in the centre of the L2–L2 dimer interface on NPR-LBD, and subsequently pulls the L2 lobe in the dimer into closerproximity. Since the dimer of NPR is fixed by the stable L1–L1interface, the movement of the L2 domain induces a 13.5 ◦ moreopen clamshell for each protomer [38].

Interestingly, the binding site for spermine, an allostericmodulator for NMDAR, is predicted to locate near the L2–L2dimer interface of the ATD [14,41,42]. In the context of thefull-length tetrameric GluA2, the L2 of ATD is in direct contactwith the D1 domain of LBD, and the stability of the LBD D1–D1 interface is a key determinant of the kinetics of channelactivation and desensitization [4,43]. It is intriguing to propose

c© The Authors Journal compilation c© 2011 Biochemical Society

260 G. Yao and others

Figure 3 Dimeric organization of the ATD is not rigid

The structures of the ATD are shown as cartoons with α-helices shown as cylinders. The ATD of GluA1, GluA2 and GluA3 are coloured in orange, grey and green respectively. (a and b) One protomer(Mol-A) of each of GluA1–GluA3 is superimposed on the basis of the Cα atoms. The two views differ by a rotation of ∼90◦ around a vertical axis. (c and d) The dimeric ATDs of GluA1–GluA3 aresuperimposed on the basis of the Cα atoms on one protomer (Mol-A), whereas the other protomer (Mol-B) is left free to move. For clarity, only Mol-A of GluA1 is shown in (c) and it is omitted in(d). Helices α3 and α5 are selected as references for comparison, since they mediate key interactions on L1 and L2 respectively.

that modulator binding on the L2–L2 interface of AMPAR mightchange the mobility of the L2 domain, which in turn would affectthe D1–D1 interface of the LBD and allosterically modulatechannel activity. Thus the L2–L2 dimer interface of AMPARcould provide a good target for development of therapeuticallyactive allosteric modulators.

Crystal lattice suggests a novel mode of ATD-mediated AMPARassociation

Remarkably, the isolated ATDs of three different iGluR subunits(GluA2, GluK2 and GluK3) all crystallize from various crystalforms in a similar dimer-of-dimers assembly, replicating thestructure found in the full-length GluA2 [6,23,24,26,27]. Thisis another example showing that physiological protein–proteininteracting interfaces are frequently used for crystal packing. Asshown in Figure 4(d), only the two ‘proximal’ protomers withina tetramer (Mol-A and Mol-A’) bind to each other via the L2domain. The two ‘distal’ protomers (Mol-B and Mol-B’) stayfurther away. The tetramerization of the isolated ATD is consistentwith the suggestion that the ATD plays a key role in tetramerassembly of iGluR in vivo [8–11].

Interestingly, another bona fide protein–protein interactinginterface on iGluR, the LBD dimer interface, was originallyidentified as a crystal-packing ‘artefact’. For example, AMPAR-LBD forms a functionally authentic dimer in the crystal lattice butis predominantly monomeric in solution [6,43–45]. The LBDs ofGluN1 and GluN2A do not oligomerize in solution, but crystallizeas a physiologically relevant heterodimer [46].

Motivated by these observations, we inspected the crystal latticeof GluA1-ATD and found a novel inter-dimer interface on the ATD(Figure 4). This interface shows several significant differencesfrom the known dimer-of-dimers ATD interface. First, the knowninterface is mediated by the L2 domain, whereas the new interfaceis mediated by the L1 domain. For simplicity, we refer to theknown interface as the tail-to-tail interface and the novel packingas the head-to-head interface. Secondly, both ATD protomersin a dimer are involved in the head-to-head interface, whereasonly one of the two protomers makes contact in the tail-to-tailinterface. Thirdly, a large solvent-accessible surface of ∼1900 Å2

is buried in the head-to-head packing on GluA1-ATD, which iseven larger than the intra-dimer interface (∼1500 Å2). Itis generally accepted that a buried interface larger than 700 Å2 islikely to have physiological relevance [47]. In contrast, the tail-to-tail inter-dimer interface in the GluA2 crystal structure is only∼330 Å2. Finally, the head-to-head interface is mainly formed bythe S-loop (Figure 4b). Asp304 and Asn308 on the S-loop form fivepairs of hydrogen bonds with Asn6, Pro39 and Ile41 on its bindingpartner, and Pro309 and Val311 on the S-loop form hydrophobicinteractions with Leu28, Pro39 and Ile41 (Figure 4b). Since thecomposition of the S-loop is subfamily-specific, such head-to-head assembly will probably lead to association of iGluRs withineach subfamily.

Structural modeling suggests that the novel head-to-headpacking of the ATD is unlikely to exist within a tetramer, due toconstraints from the downstream LBD [6]. However, the head-to-head interface could mediate the inter-tetramer interaction upona slight conformational change on the ATD through a flexiblelinker (∼17 residues) connecting the ATD and LBD (Figure 4d).

c© The Authors Journal compilation c© 2011 Biochemical Society

Structure of the GluA1 N-terminal domain 261

Figure 4 Molecular packing in GluA1-ATD crystal reveals a novel head-to-head arrangement of the dimeric ATD.

(a) Two pairs of ATD dimer are shown as cartoons with α-helices shown as cylinders. The dimers are composed of molecule-A (Mol-A, green) and B (Mol-B, grey) and molecule-A’ (Mol-A’, blue)and B’ (Mol-B’, orange) respectively. The S-loop is coloured in red and labelled as a, b, a’ and b’. The blue parallelogram, using helices α9 and α10 as references, outlines a molecular surface thatlocates on the top of the dimer-of-dimers assembly. Two different views of this head-to-head dimer-of-dimers arrangement of the ATD are shown following a rotation of ∼90◦ around a horizontalaxis (a and c). A close view on the inter-dimer interface, highlighted in the red box, is shown in (b). Key interacting residues are shown as sticks, and broken lines represent hydrogen bonding.(d) The crystal structure of the full-length GluA2 is shown on the left-hand side, where the tail-to-tail assembly on the ATD is mediated by Mol-A and Mol-A’. In the plausible head-to-head ATDarrangement, all four protomers are involved in binding that could lead to inter-tetramer association of AMPAR on the synaptic membrane. The S-loops are depicted by red arrows.

Indeed, a similar Y-shape conformation of a tetrameric AMPARwas observed by single particle electron microscopy, where thetwo ATD dimers swing away from each other while the remainingtetramer structure remains the same [48,49].

The ATD-mediated inter-tetramer aggregation of AMPARraises the intriguing possibility that ATD might play arole in clustering of the highly concentrated AMPAR atthe synaptic membrane. This would be consistent with theobservation that neuronal pentraxins (Narp, NP1 and NPR)interact with the ATD of AMPAR and promote clusteringof the receptors [19,20,50]. Furthermore, the head-to-headassembly of the ATD generates a large molecular surface ontop of the ATD in the shape of a parallelogram (diagonals107 Å×56 Å), which stands ∼130 Å away from the post-synaptic membrane, making it an ideal surface to interact withpresynaptic proteins (Figure 4 and Supplementary Figure S3 athttp://www.BiochemJ.org/bj/438/bj4380255add.htm). In keepingwith this hypothesis, it has been shown that AMPARs play a

crucial structural role in regulating the stability of presynapticinputs, which is mediated by the ATD and is independent ofreceptor-mediated channel activity [18].

CONCLUSION

The present study reports the first crystal structure of the ATDof the GluA1 subunit of iGluRs. Detailed structural analysescomparing GluA1–GluA3 suggest that homodimerization of ATDis mediated mainly by interactions between the L1 domain, whichare highly conserved within the AMPAR family. The L2 domainhas greater relative mobility and is probably responsible for thedifferent affinities for homodimerization among the AMPARATDs. It is notable that the intra-dimer interface is highlyconserved between GluA1 and GluA2, and we observe weakheterodimerization between them. Thus the isolated ATD isprobably insufficient to drive the preferred hetero-assembly of

c© The Authors Journal compilation c© 2011 Biochemical Society

262 G. Yao and others

AMPARs as observed in vivo. This conclusion is consistent witha model in which the LBD and the transmembrane channel alsoplay important roles in heteromerization of AMPARs [8].

Our results from the present study suggest a plausible head-to-head arrangement of the dimeric ATD that is mostly mediatedby the specificity loop. This unique inter-tetramer associationof AMPAR provides a novel suggestion as to how highlyconcentrated AMPARs form clusters on the synaptic membrane,and how the ATD might be involved in trans-synaptic signaltransduction. Several synaptically localized molecules functionacross the synaptic cleft to reciprocally co-ordinate differentiationon both sides of the synapse, including neurexin–neuroligin,SynCAM, cadherin, EphrinB–EphB and Liprin-α/LAR. TheATD crystal structure reported in the present study will allowinvestigations of the role played by ATD in trans-synapticsignalling and in regulating presynaptic stability, and will adda new dimension to our understanding of iGluR function.

AUTHOR CONTRIBUTION

Jie Zhou cloned GluA1-ATD and optimized the expression conditions. Guorui Yaoperformed the protein expression, purification, characterization and crystallization assays.Irimpan Mathews collected the diffraction data. Yinong Zong and Shenyan Gu performedstructure determination and analysis. Rongsheng Jin supervised the project. All authorswere involved in manuscript preparation.

ACKNOWLEDGEMENTS

Portions of this research were performed at the Stanford Synchrotron Radiation Lightsource(SSRL), a national user facility operated by Stanford University on behalf of the U.S.Department of Energy, Office of Basic Energy Sciences. The SSRL Structural MolecularBiology Program is supported by the Department of Energy, Office of Biological andEnvironmental Research, and by the National Institutes of Health, National Center forResearch Resources, Biomedical Technology Program and the National Institute of GeneralMedical Sciences.

FUNDING

This work was supported by the National Institutes of Health [grant numbers R21AG033813and R01GM090023], by the Alfred P. Sloan Research Fellowship and by the start-upresearch fund from the Sanford-Burnham Medical Research Institute.

REFERENCES

1 Dingledine, R., Borges, K., Bowie, D. and Traynelis, S. F. (1999) The glutamate receptorion channels. Pharmacol. Rev. 51, 7–61

2 Malenka, R. C. and Nicoll, R. A. (1999) Long-term potentiation – a decade of progress?Science 285, 1870–1874

3 Armstrong, N., Sun, Y., Chen, G. Q. and Gouaux, E. (1998) Structure of a glutamate-receptor ligand-binding core in complex with kainate. Nature 395, 913–917

4 Mayer, M. L. (2006) Glutamate receptors at atomic resolution. Nature 440, 456–4625 Traynelis, S. F., Wollmuth, L. P., McBain, C. J., Menniti, F. S., Vance, K. M., Ogden, K. K.,

Hansen, K. B., Yuan, H., Myers, S. J. and Dingledine, R. (2010) Glutamate receptor ionchannels: structure, regulation, and function. Pharmacol. Rev. 62, 405–496

6 Sobolevsky, A. I., Rosconi, M. P. and Gouaux, E. (2009) X-ray structure, symmetry andmechanism of an AMPA-subtype glutamate receptor. Nature 462, 745–756

7 Wang, Y., Matsuda, S., Drews, V., Torashima, T., Meisler, M. H. and Yuzaki, M. (2003) Ahot spot for hotfoot mutations in the gene encoding the δ2 glutamate receptor. Eur. J.Neurosci. 17, 1581–1590

8 Ayalon, G. and Stern-Bach, Y. (2001) Functional assembly of AMPA and kainate receptorsis mediated by several discrete protein-protein interactions. Neuron 31, 103–113

9 Ayalon, G., Segev, E., Elgavish, S. and Stern-Bach, Y. (2005) Two regions in the N-terminaldomain of ionotropic glutamate receptor 3 form the subunit oligomerization interfaces thatcontrol subtype-specific receptor assembly. J. Biol. Chem. 280, 15053–15060

10 Stern-Bach, Y., Bettler, B., Hartley, M., Sheppard, P. O., O’Hara, P. J. and Heinemann, S. F.(1994) Agonist selectivity of glutamate receptors is specified by two domains structurallyrelated to bacterial amino acid-binding proteins. Neuron 13, 1345–1357

11 Leuschner, W. D. and Hoch, W. (1999) Subtype-specific assembly of α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor subunits is mediated by theirN-terminal domains. J. Biol. Chem. 274, 16907–16916

12 Kreusch, A., Pfaffinger, P. J., Stevens, C. F. and Choe, S. (1998) Crystal structure of thetetramerization domain of the Shaker potassium channel. Nature 392, 945–948

13 Zerangue, N., Jan, Y. N. and Jan, L. Y. (2000) An artificial tetramerization domain restoresefficient assembly of functional Shaker channels lacking T1. Proc. Natl. Acad. Sci. U.S.A.97, 3591–3595

14 Masuko, T., Kashiwagi, K., Kuno, T., Nguyen, N. D., Pahk, A. J., Fukuchi, J., Igarashi, K.and Williams, K. (1999) A regulatory domain (R1-R2) in the amino terminus of theN-methyl-D-aspartate receptor: effects of spermine, protons, and ifenprodil, andstructural similarity to bacterial leucine/isoleucine/valine binding protein. Mol.Pharmacol. 55, 957–969

15 Zheng, F., Erreger, K., Low, C. M., Banke, T., Lee, C. J., Conn, P. J. and Traynelis, S. F.(2001) Allosteric interaction between the amino terminal domain and the ligand bindingdomain of NR2A. Nat. Neurosci. 4, 894–901

16 Perin-Dureau, F., Rachline, J., Neyton, J. and Paoletti, P. (2002) Mapping the binding siteof the neuroprotectant ifenprodil on NMDA receptors. J. Neurosci. 22, 5955–5965

17 Passafaro, M., Nakagawa, T., Sala, C. and Sheng, M. (2003) Induction of dendritic spinesby an extracellular domain of AMPA receptor subunit GluR2. Nature 424, 677–681

18 Ripley, B., Otto, S., Tiglio, K., Williams, M. E. and Ghosh, A. (2011) Regulation of synapticstability by AMPA receptor reverse signaling. Proc. Natl. Acad. Sci. U.S.A. 108, 367–372

19 O’Brien, R. J., Xu, D., Petralia, R. S., Steward, O., Huganir, R. L. and Worley, P. (1999)Synaptic clustering of AMPA receptors by the extracellular immediate-early gene productNarp. Neuron 23, 309–323

20 Sia, G. M., Beique, J. C., Rumbaugh, G., Cho, R., Worley, P. F. and Huganir, R. L. (2007)Interaction of the N-terminal domain of the AMPA receptor GluR4 subunit with theneuronal pentraxin NP1 mediates GluR4 synaptic recruitment. Neuron 55, 87–102

21 Saglietti, L., Dequidt, C., Kamieniarz, K., Rousset, M. C., Valnegri, P., Thoumine, O.,Beretta, F., Fagni, L., Choquet, D., Sala, C. et al. (2007) Extracellular interactions betweenGluR2 and N-cadherin in spine regulation. Neuron 54, 461–477

22 Dalva, M. B., Takasu, M. A., Lin, M. Z., Shamah, S. M., Hu, L., Gale, N. W. and Greenberg,M. E. (2000) EphB receptors interact with NMDA receptors and regulate excitatorysynapse formation. Cell 103, 945–956

23 Jin, R., Singh, S. K., Gu, S., Furukawa, H., Sobolevsky, A. I., Zhou, J., Jin, Y. and Gouaux,E. (2009) Crystal structure and association behaviour of the GluR2 amino-terminaldomain. EMBO. J. 28, 1812–1823

24 Clayton, A., Siebold, C., Gilbert, R. J., Sutton, G. C., Harlos, K., McIlhinney, R. A., Jones,E. Y. and Aricescu, A. R. (2009) Crystal structure of the GluR2 amino-terminal domainprovides insights into the architecture and assembly of ionotropic glutamate receptors.J. Mol. Biol. 392, 1125–1132

25 Rossmann, M., Sukumaran, M., Penn, A. C., Veprintsev, D. B., Babu, M. M. and Greger, I.H. (2011) Subunit-selective N-terminal domain associations organize the formation ofAMPA receptor heteromers. EMBO J. 30, 959–971

26 Kumar, J., Schuck, P., Jin, R. and Mayer, M. L. (2009) The N-terminal domain ofGluR6-subtype glutamate receptor ion channels. Nat. Struct. Mol. Biol. 16, 631–638

27 Kumar, J. and Mayer, M. L. (2010) Crystal structures of the glutamate receptor ionchannel GluK3 and GluK5 amino-terminal domains. J. Mol. Biol. 404, 680–696

28 Karakas, E., Simorowski, N. and Furukawa, H. (2009) Structure of the zinc-boundamino-terminal domain of the NMDA receptor NR2B subunit. EMBO J. 28, 3910–3920

29 Sukumaran, M., Rossmann, M., Shrivastava, I., Dutta, A., Bahar, I. and Greger, I. H.(2011) Dynamics and allosteric potential of the AMPA receptor N-terminal domain. EMBOJ. 30, 972–982

30 Farina, A. N., Blain, K. Y., Maruo, T., Kwiatkowski, W., Choe, S. and Nakagawa, T. (2011)Separation of domain contacts is required for heterotetrameric assembly of functionalNMDA receptors. J. Neurosci. 31, 3565–3579

31 Battye, T. G., Kontogiannis, L., Johnson, O., Powell, H. R. and Leslie, A. G. (2011)iMOSFLM: a new graphical interface for diffraction-image processing with MOSFLM.Acta Crystallogr. Sect. D Biol. Crystallogr. 67, 271–281

32 McCoy, A. J., Grosse-Kunstleve, R. W., Adams, P. D., Winn, M. D., Storoni, L. C. andRead, R. J. (2007) Phaser crystallographic software. J. Appl. Cryst. 40, 658–674

33 Adams, P. D., Afonine, P. V., Bunkoczi, G., Chen, V. B., Davis, I. W., Echols, N., Headd, J.J., Hung, L. W., Kapral, G. J., Grosse-Kunstleve, R. W. et al. (2010) PHENIX: acomprehensive Python-based system for macromolecular structure solution. ActaCrystallogr. Sect. D Biol. Crystallogr. 66, 213–221

34 Emsley, P. and Cowtan, K. (2004) COOT: model-building tools for molecular graphics.Acta Crystallogr. Sect. D Biol. Crystallogr. 60, 2126–2132

35 Brunger, A. T. (1992) Free R value: a novel statistical quantity for assessing the accuracyof crystal structures. Nature 355, 472–475

36 Trakhanov, S., Vyas, N. K., Luecke, H., Kristensen, D. M., Ma, J. and Quiocho, F. A. (2005)Ligand-free and -bound structures of the binding protein (LivJ) of the Escherichia coliABC leucine/isoleucine/valine transport system: trajectory and dynamics of theinterdomain rotation and ligand specificity. Biochemistry 44, 6597–6608

c© The Authors Journal compilation c© 2011 Biochemical Society

Structure of the GluA1 N-terminal domain 263

37 Jingami, H., Nakanishi, S. and Morikawa, K. (2003) Structure of the metabotropicglutamate receptor. Curr. Opin. Neurobiol. 13, 271–278

38 He, X., Chow, D., Martick, M. M. and Garcia, K. C. (2001) Allosteric activation of aspring-loaded natriuretic peptide receptor dimer by hormone. Science 293, 1657–1662

39 Gielen, M., Le Goff, A., Stroebel, D., Johnson, J. W., Neyton, J. and Paoletti, P. (2008)Structural rearrangements of NR1/NR2A NMDA receptors during allosteric inhibition.Neuron 57, 80–93

40 Gielen, M., Siegler Retchless, B., Mony, L., Johnson, J. W. and Paoletti, P. (2009)Mechanism of differential control of NMDA receptor activity by NR2 subunits. Nature459, 703–707

41 Traynelis, S. F., Hartley, M. and Heinemann, S. F. (1995) Control of proton sensitivity ofthe NMDA receptor by RNA splicing and polyamines. Science 268, 873–876

42 Huggins, D. J. and Grant, G. H. (2005) The function of the amino terminal domain inNMDA receptor modulation. J. Mol. Graph. Model. 23, 381–388

43 Sun, Y., Olson, R., Horning, M., Armstrong, N., Mayer, M. and Gouaux, E. (2002)Mechanism of glutamate receptor desensitization. Nature 417, 245–253

44 Armstrong, N. and Gouaux, E. (2000) Mechanisms for activation and antagonism of anAMPA-sensitive glutamate receptor: crystal structures of the GluR2 ligand binding core.Neuron 28, 165–181

45 Jin, R., Banke, T. G., Mayer, M. L., Traynelis, S. F. and Gouaux, E. (2003) Structural basisfor partial agonist action at ionotropic glutamate receptors. Nat. Neurosci. 6,803–810

46 Furukawa, H., Singh, S. K., Mancusso, R. and Gouaux, E. (2005) Subunit arrangementand function in NMDA receptors. Nature 438, 185–192

47 Janin, J. (1997) Specific versus non-specific contacts in protein crystals. Nat. Struct.Biol. 4, 973–974

48 Nakagawa, T., Cheng, Y., Ramm, E., Sheng, M. and Walz, T. (2005) Structure and differentconformational states of native AMPA receptor complexes. Nature 433, 545–549

49 Nakagawa, T. (2010) The biochemistry, ultrastructure, and subunit assembly mechanismof AMPA receptors. Mol. Neurobiol. 42, 161–184

50 Song, I. and Huganir, R. L. (2002) Regulation of AMPA receptors during synapticplasticity. Trends Neurosci. 25, 578–588

Received 5 May 2011/3 June 2011; accepted 6 June 2011Published as BJ Immediate Publication 6 June 2011, doi:10.1042/BJ20110801

c© The Authors Journal compilation c© 2011 Biochemical Society

Biochem. J. (2011) 438, 255–263 (Printed in Great Britain) doi:10.1042/BJ20110801

SUPPLEMENTARY ONLINE DATACrystal structure of the glutamate receptor GluA1 N-terminal domainGuorui YAO*†1 , Yinong ZONG†1 , Shenyan GU†, Jie ZHOU†, Huaxi XU*†, Irimpan I. MATHEWS‡ and Rongsheng JIN†2

*Fujian Provincial Key Laboratory of Neurodegenerative Disease and Aging Research, College of Medicine, Xiamen University, Xiamen, Fujian 361005, China, †Center forNeuroscience, Aging and Stem Cell Research, Sanford-Burnham Medical Research Institute, 10901 North Torrey Pines Road, La Jolla, CA 92037, U.S.A., and ‡Stanford SynchrotronRadiation Lightsource, 2575 Sand Hill Road, Menlo Park, CA 94025, U.S.A.

Figure S1 Inter-domain interactions stabilize the closed-cleft conformationof GluA1-ATD

The L1 and L2 domains of GluA1-ATD are coloured in blue and green respectively. Key residuesthat participate in inter-domain interactions are shown as sticks. Dashed lines indicate hydrogenbonds. (b) Close-up of the outlined elliptical area in (a).

Figure S2 B-factor distribution on the ATD

The structures of the dimeric GluA1 (a), GluA2 (b) and GluA3 (c) ATDs are shown in sausage-stylecartoon representation, with the tube diameter and colour scaled by values of B-factors. B-factorsare scaled within each structure and coloured in a gradient varying from blue (the lowest values)to red (the highest values).

1 These authors contributed equally to this work2 To whom correspondence should be addressed (email rjin@sanfordburnham.org).The structural co-ordinates and diffraction data for GluA1-ATD reported will appear in the PDB under accession code 3SAJ.

c© The Authors Journal compilation c© 2011 Biochemical Society

G. Yao and others

Figure S3 A novel head-to-head arrangement of the dimeric ATD

Two pairs of ATD dimers are shown as surface representation. They are composed of molecule-A (Mol-A, green) and B (Mol-B, grey) and molecule-A’ (Mol-A’, blue) and B’ (Mol-B’, orange)respectively. The S-loop is coloured in red. Two different views of this head-to-head dimer-of-dimers arrangement of the ATD are shown following a rotation of ∼90◦ around a horizontal axis.

Table S1 Data collection and refinement statistics

Values in parentheses represent the highest resolution shell. R merge =∑

h∑

j|I hj −〈I h〉|/∑

h∑

j| I h,j|. R work = (∑‖F o|−|F c‖)/

∑|F o|. For calculation of R free, 5 % of the reflections were set aside.rmsd, root mean square deviation.

Data collection GluA1-ATD

Wavelength (A) 1.0000Space group P212121

Unit cell (a, b, c; A) 92.58, 94.42, 186.36Resolution (A) 50.00–2.50 (2.55–2.50)Total observations 330249Unique observations 57222Average redundancy 5.8 (4.8)Data completeness (%) 99.7 (97.5)I/σ I 10.3 (3.5)Rmerge (%) 6.6 (40.6)Refinement

Number of protein atoms 11928Number of ligand (sugar) atoms 466Number of water atoms 143Rwork (%) 22.2 (23.8)R free (%) 28.1 (33.3)rmsd bond lengths (A) 0.015rmsd bond angles (◦) 1.791

Averaged B factorProtein main chain 30.89Protein side chain 33.00Ligands 105.70Water 36.55

Ramachandran statisticsMost favoured (%) 96.3Additional allowed (%) 3.7Disallowed (%) 0

Received 5 May 2011/3 June 2011; accepted 6 June 2011Published as BJ Immediate Publication 6 June 2011, doi:10.1042/BJ20110801

c© The Authors Journal compilation c© 2011 Biochemical Society