Identification of Molecular Determinants That Are Important in the Assembly of N-Methyl-d-aspartate Receptors
2001; Elsevier BV; Volume: 276; Issue: 22 Linguagem: Inglês
10.1074/jbc.m101382200
ISSN1083-351X
AutoresElisabeth Meddows, B. Le Bourdellès, Sarah Grimwood, Keith A. Wafford, Satpal Sandhu, Paul J. Whiting, R. A. Jeffrey McIlhinney,
Tópico(s)Chemical Synthesis and Analysis
ResumoTo determine which domains of theN-methyl-d-aspartate (NMDA) receptor are important for the assembly of functional receptors, a number of N- and C-terminal truncations of the NR1a subunit have been produced. Truncations containing a complete ligand binding domain bound glycine antagonist and gave binding constants similar to those of the native subunit, suggesting they were folding to form antagonist binding sites. Since NR2A is not transported to the cell surface unless it is associated with NR1 (McIlhinney, R. A. J., Le Bourdellès, B., Tricuad, N., Molnar, E., Streit, P., and Whiting, P. J. (1998) Neuropharmacology 37, 1355–1367), surface expression of NR2A can be used to monitor the association of the subunits. There was progressive loss of NR2A cell surface expression as the N terminus of NR1a was shortened, with complete loss when truncated beyond residue 380. Removal of the C terminus and/or the last transmembrane domain did not affect NR2A surface expression. Similar results were obtained in co-immunoprecipitation experiments. The oligomerization status of the co-expressed NR1a constructs and NR2A subunits was investigated using a non-denaturing gel electrophoresis system (blue native-polyacrylamide gel electrophoresis) and sucrose density gradient centrifugation. The blue native-polyacrylamide gel electrophoresis system also showed that the NR1a subunits could form a homodimer, which was confirmed using soluble constructs of the NR1a subunit. Together these results suggest the residues N-terminal of residue 380 are important for the association of NR2A with NR1a and that the complete N-terminal domain of the NR1a subunit is required for oligomerization with NR2A. To determine which domains of theN-methyl-d-aspartate (NMDA) receptor are important for the assembly of functional receptors, a number of N- and C-terminal truncations of the NR1a subunit have been produced. Truncations containing a complete ligand binding domain bound glycine antagonist and gave binding constants similar to those of the native subunit, suggesting they were folding to form antagonist binding sites. Since NR2A is not transported to the cell surface unless it is associated with NR1 (McIlhinney, R. A. J., Le Bourdellès, B., Tricuad, N., Molnar, E., Streit, P., and Whiting, P. J. (1998) Neuropharmacology 37, 1355–1367), surface expression of NR2A can be used to monitor the association of the subunits. There was progressive loss of NR2A cell surface expression as the N terminus of NR1a was shortened, with complete loss when truncated beyond residue 380. Removal of the C terminus and/or the last transmembrane domain did not affect NR2A surface expression. Similar results were obtained in co-immunoprecipitation experiments. The oligomerization status of the co-expressed NR1a constructs and NR2A subunits was investigated using a non-denaturing gel electrophoresis system (blue native-polyacrylamide gel electrophoresis) and sucrose density gradient centrifugation. The blue native-polyacrylamide gel electrophoresis system also showed that the NR1a subunits could form a homodimer, which was confirmed using soluble constructs of the NR1a subunit. Together these results suggest the residues N-terminal of residue 380 are important for the association of NR2A with NR1a and that the complete N-terminal domain of the NR1a subunit is required for oligomerization with NR2A. N-methyl-d-aspartate transmembrane domain nicotinic acetylcholine receptor human embryonic kidney polyacrylamide gel electrophoresis blue native α-amino-3-hydroxy-5-methyl-4-isoxazole proprionate bis(2-hydroxyethyl)iminotris(hydroxymethyl)methane γ-aminobutyric acid subunit A The N-methyl-d-aspartate (NMDA)1 subtype of the glutamate receptor family is a hetero-oligomeric protein composed of two classes of NMDA receptor subunits: NR1 and NR2. The NR1 subunit is encoded by a single gene, which undergoes extensive splicing to generate eight different splice variants that differ in regional distribution and functional properties (2Dingledine R. Borges K. Bowie D. Traynelis S.F. Pharmacol. Rev. 1999; 51: 7-61PubMed Google Scholar). The NR2 subunit class consists of four different subunits, NR2A–NR2D, encoded by four separate but closely related genes (2Dingledine R. Borges K. Bowie D. Traynelis S.F. Pharmacol. Rev. 1999; 51: 7-61PubMed Google Scholar). A number of studies of mammalian cell lines either permanently or transiently transfected with NR1 alone have indicated that the NR1 subunit does not form glycine-glutamate-responsive channels and requires the presence of NR2 to do so (3Cik M. Chazot P.L. Stephenson F.A. Biochem. J. 1993; 296: 877-883Crossref PubMed Scopus (104) Google Scholar, 4Grimwood S. Le Bourdelles B. Whiting P.J. J. Neurochem. 1995; 64: 525-530Crossref PubMed Scopus (73) Google Scholar, 5Varney M.A. Jachec C. Deal C. J. Pharmacol. Exp. Ther. 1996; 279: 367-378PubMed Google Scholar). Other studies have shown that the NR1 and NR2 subunits contribute differently to the binding sites of a functional NMDA receptor. The NR1 subunit forms the glycine binding site (6Kuryatov A. Laube B. Betz H. Kuhse J. Neuron. 1994; 12: 1291-1300Abstract Full Text PDF PubMed Scopus (337) Google Scholar, 7Wafford K.A. Kathoria M. Bain C.J. Marshall G. Le Bourdelles B. Kemp J. Whiting P.J. Mol. Pharmacol. 1995; 47: 374-380PubMed Google Scholar, 8Hirai H. Kirsch J. Laube B. Betz H. Kuhse J. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 6031-6036Crossref PubMed Scopus (195) Google Scholar), and the NR2 subunit provides part of the glutamate binding site (9Laube B. Hirai H. Sturgess M. Betz H. Kushe J. Neuron. 1997; 18: 493-503Abstract Full Text Full Text PDF PubMed Scopus (411) Google Scholar, 10Anson L.C. Chen P.E. Wyllie D.J.A. Colquhoun D. Schoepfer R. J. Neurosci. 1998; 18: 581-589Crossref PubMed Google Scholar). Thus, different combinations of both subunits co-assemble to form functionally distinct NMDA receptors. However, the biochemical and functional studies reported to date are ambiguous with regard to NMDA receptor subunit stoichiometry. Functional studies indicate that binding of at least two molecules of both glutamate and glycine is required for NMDA receptor activation, suggesting that at least four subunits must co-assemble (11Patneau D.K. Mayer M.L. J. Neurosci. 1990; 10: 2385-2399Crossref PubMed Google Scholar, 12Benveniste M. Mayer M.L. Br. J. Pharmacol. 1991; 104: 207-221Crossref PubMed Scopus (76) Google Scholar, 13Clements J.D. Westbrook G.L. Neuron. 1991; 7: 605-613Abstract Full Text PDF PubMed Scopus (273) Google Scholar). The molecular size of native NMDA receptors, as determined by both gel filtration and native polyacrylamide gel electrophoresis, is in the range of 605–850 kDa, which is consistent with the co-assembly of between four to five subunits (14Blahos J. Wenthold R.J. J. Biol. Chem. 1996; 271: 15669-15674Abstract Full Text Full Text PDF PubMed Scopus (140) Google Scholar, 15Brose N. Gasic G.P. Vetter D.E. Sullivan J.M. Heinemann S.F. J. Biol. Chem. 1993; 268: 22663-22671Abstract Full Text PDF PubMed Google Scholar, 16Chazot P.L. Coleman S.K. Cik M. Stephenson F.A. J. Biol. Chem. 1994; 269: 24403-24409Abstract Full Text PDF PubMed Google Scholar). A recent biochemical study has suggested that there are three NR2 subunits per NMDA receptor complex, indicating that the NMDA receptor is at least a pentamer (17Hawkins L. Chazot P. Stephenson F.A. J. Biol. Chem. 1999; 274: 27211-27218Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar). Electrophysiological studies on the subunit stoichiometry using co-expressed wild-type and mutant forms of either the NR1 or NR2 subunits have also been inconclusive suggesting two or three NR1 subunits or two or three NR2 subunits per functional NMDA receptor complex (18Behe P. Stern P. Wylie D.J.A. Nassar M. Schoepfer R. Colquhoun D. Proc. R. Soc. Lond. Ser. B Biol. Sci. 1995; 262: 205-213Crossref PubMed Scopus (118) Google Scholar, 19Laube B. Kuhse J. Betz H. J. Neurosci. 1998; 18: 2954-2961Crossref PubMed Google Scholar, 20Premkumar L.S. Auerbach A. J. Gen. Physiol. 1997; 110: 485-502Crossref PubMed Scopus (101) Google Scholar). The regions of NMDA receptor subunits mediating the assembly of hetero-oligomeric NMDA receptors have, to date, not been identified. All glutamate receptor subunits are thought to share a common transmembrane topology and domain structure with three transmembrane domains (TMI, -III, and -IV), a second membrane domain forming a re-entrant loop that partly lines the ion channel pore, an extracellular N terminus, and an intracellular C terminus (21Bennett J.A. Dingledine R. Neuron. 1995; 14: 373-384Abstract Full Text PDF PubMed Scopus (248) Google Scholar, 22Hollmann M. The Ionotropic Glutamate Receptors. Humana Press, New York1997: 3-98Google Scholar, 23Wo Z.G. Oswald R.E. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 7154-7158Crossref PubMed Scopus (179) Google Scholar). The ligand binding domain is thought to be formed between part of the N terminus (the S1 domain) just before TMI and the extracellular loop between TMIII and TMIV (the S2 domain) (6Kuryatov A. Laube B. Betz H. Kuhse J. Neuron. 1994; 12: 1291-1300Abstract Full Text PDF PubMed Scopus (337) Google Scholar, 24Lampinen M. Pentikainen O. Johnson M.S. Keinanen K. EMBO J. 1998; 17: 4704-4711Crossref PubMed Scopus (60) Google Scholar, 25Sternbach Y. Bettler B. Hartley M. Sheppard P.O. O'Hara P.J. Heinemann S.F. Neuron. 1994; 13: 1345-1357Abstract Full Text PDF PubMed Scopus (397) Google Scholar). Approximately the first 400 amino acids of the N terminus share sequence homology to the bacterial periplasmic leucine-isoleucine-valine-binding protein (LIVBP domain) (26Ferns M. Hoch W. Campanelli J.T. Rupp F. Hall Z.W. Scheller R.H. Neuron. 1992; 8: 1079-1086Abstract Full Text PDF PubMed Scopus (191) Google Scholar). The proximal N-terminal domain has recently been suggested to be important in the assembly of AMPA receptor subunits in mammalian cells (27Leuschner W.D. Hoch W. J. Biol. Chem. 1999; 274: 16907-16916Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar). In this study we addressed the question of which domains of the NR1a NMDA receptor subunit are important for their assembly into oligomeric receptor complexes. Human embryonic kidney (HEK) 293 cells were transfected with mutated NR1a constructs containing deletions to the N and C termini, NR1a chimeras, and soluble secreted forms of the NR1a subunit and wild-type NR2A subunits. Since we have shown previously that NR2A is not expressed at the cell surface unless it is co-expressed with NR1 (1McIlhinney R.A.J. Le Bourdellès B. Tricuad N. Molnar E. Streit P. Whiting P.J. Neuropharmacology. 1998; 37: 1355-1367Crossref PubMed Scopus (134) Google Scholar), cell surface expression of the NR2A subunit has been used to monitor subunit association, together with co-immunoprecipitation of the different subunits. The formation of oligomeric complexes has been monitored using a novel nondenaturing gel electrophoresis system and sucrose gradient sedimentation. All the truncated NR1a subunits are derived from a human NR1a cDNA and were generated by using standard mutagenesis techniques previously described in detail (28Sandhu S. Grimwood S. Mortshire-Smith R.J. Whiting P.J. Le Bourdelles B. J. Neurochem. 1999; 72: 1694-1698Crossref PubMed Scopus (2) Google Scholar). An octapeptide FLAG epitope tag (KDYKDDDDK) was introduced into the N terminus between Asp23 and Lys25, just C-terminal to the putative signal cleavage point. The constructs were verified by sequencing and by in vitro translation (TNT T7 Transcription/Translation System; Promega; Fig. 1). Radioligand binding assays using [3H]L-689,560 were performed as previously described in detail (4Grimwood S. Le Bourdelles B. Whiting P.J. J. Neurochem. 1995; 64: 525-530Crossref PubMed Scopus (73) Google Scholar, 28Sandhu S. Grimwood S. Mortshire-Smith R.J. Whiting P.J. Le Bourdelles B. J. Neurochem. 1999; 72: 1694-1698Crossref PubMed Scopus (2) Google Scholar). HEK-293 cells were cultured and transfected using the calcium phosphate method. The NMDA receptor NR1a and NR2A subunits in the expression vectors pcDNAI/Amp and pCDM8 were transfected at a ratio of 1:3. After transfection, the cells were grown in the presence of 0.5 mm ketamine (Sigma) and harvested 24 h later. Membranes were prepared from the cells using hypotonic lysis, shearing, and centrifugation as described previously (29Robbins M.J. Ciruela F. Rhodes A. McIlhinney R.A.J. J. Neurochem. 1999; 72: 2539-2547Crossref PubMed Scopus (72) Google Scholar) except that 20 mm iodoacetamide was added to the lysis medium prior to cell lysis. Adult female Xenopus laevis were anesthetized by immersion in a 0.1% solution of 3-aminobenzoic acid ethyl ester, with the pH adjusted with 1 mNaHCO3 to that of the water in which the toad was housed, for 30–45 min, and stage V and stage VI oocytes were surgically removed. After mild collagenase treatment to remove follicle cells (type IA, 0.5 mg/ml, for 6 min), the oocyte nuclei were directly injected with 10–20 nl of injection buffer (88 mm NaCl, 1 mm KCl, 15 mm HEPES, at pH 7, filtered through nitrocellulose) containing different combinations of human NMDA subunit cDNAs (20 ng/μl) engineered into the expression vector pCDM8 or pcDNAI/Amp. NR1a truncations and NR2A cDNAs were injected at a ratio of 1:3. Oocytes were maintained at 19–20 °C in modified Barth's medium consisting of 88 mm NaCl, 1 mmKCl, 10 mm HEPES, 0.82 mmMgSO4, 0.33 mmCa(NO3)2, 0.91 mmCaCl2, 2.4 mm NaHCO3, at pH 7.5 supplemented with 50 μg/ml gentamycin, 10 μg/ml streptomycin, 10 units/ml penicillin, and 2 mm sodium pyruvate) for up to 6 days. For electrophysiological recordings, oocytes were placed in a 50-μl bath and continually perfused at 4–6 ml/min with Barium Ringer's solution (115 mm NaCl, 2.5 mm KCl, 10 mm HEPES, 1.8 mm BaCl2, pH 7.2). Cells were impaled with two 1–3-megohm electrodes containing 2m KCl and voltage-clamped at −70 mV. In all experiments drugs were applied in the perfusate until the peak of the response was observed. Transiently transfected HEK-293 cells were overlaid with borate buffer (10 mm boric acid, 150 mm NaCl, pH 8.8) containing 50 μg/ml of the non-permeant reactive ester sulfo-NHS-biotin (Pierce; dissolved at 10 mg/ml in N,N-dimethylformamide). Unreacted ester was removed by incubating the cells for 5 min with 1 m ammonium chloride. The cells were washed with Tris-saline (pH 7.4) then lysed on ice in RIPA buffer (50 mm Tris-HCl, pH 7.5, 1% (w/v) Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 100 mmNaCl, 1 mm EDTA) plus a protease inhibitor mixture (Roche Molecular Biochemicals) and 20 mm iodoacetamide. The lysate was centrifuged at 10,000 × g, and a tenth of the supernatant removed for total cell lysate analysis. The biotin-labeled surface proteins in the remaining supernatants were affinity isolated with 100 μl of streptavidin-agarose beads (Sigma) rotating for 3.5 h at 4 °C. The resulting pellets were washed twice with RIPA buffer and twice with 50 mm Tris-HCl, pH 8.0, and proteins eluted from the streptavidin-agarose beads by addition of 2× reducing sample buffer (20 mmdithiothreitol, 2% (w/v) SDS, 10% (v/v) glycerol, 100 mmTris-HCl). The samples were then analyzed using SDS-PAGE and the Western immunoblot protocol. The possibility that the sulfo-NHS-biotin penetrates the cells during the biotinylation procedure was controlled for by probing the total cell lysates and streptavidin isolates for the intracellular protein β-tubulin. Only those experiments in which this control was negative were used for analysis. Transiently transfected HEK-293 cells were lysed on ice with RIPA buffer containing 20 mmiodoacetamide and protease inhibitors (Roche Molecular Biochemicals). The lysate was centrifuged at 10,000 × g, and an aliquot was removed for total cell lysate analysis. A fifth of the lysate was made up to 1 ml with lysis buffer and then rotated overnight at 4 °C with 1–2 μg/ml precipitating antibody. A 100-μl suspension of Protein G-Sepharose Fast Flow beads (Amersham Pharmacia Biotech) was added to the precipitates and mixed by rotation at 4 °C for 2 h. The immunoprecipitates were pelleted by centrifugation, and the resulting pellet was washed twice in RIPA buffer and then twice with 50 mm Tris-HCl, pH 8.0. Under these conditions, quantitative immunoprecipitation of the subunits was achieved. The immunocomplexes were eluted from the Protein G beads by mixing with 2× reducing sample buffer and boiling the samples. Both the precipitates and total cell lysates were resolved by SDS-PAGE and Western immunoblot. The antibodies used for immunoprecipitation were a sheep anti-NR1a antibody (previously characterized; Ref. 1McIlhinney R.A.J. Le Bourdellès B. Tricuad N. Molnar E. Streit P. Whiting P.J. Neuropharmacology. 1998; 37: 1355-1367Crossref PubMed Scopus (134) Google Scholar) and the anti-FLAG antibody M2 from Sigma. The NR1-GluR1 chimera was precipitated using an anti-GluR1 antibody (30Molnar E. McIlhinney R.A.J. Baude A. Nusser Z. Somogyi P. J. Neurochem. 1994; 63: 683-693Crossref PubMed Scopus (39) Google Scholar). Proteins were separated by SDS-PAGE using 7.5% polyacrylamide gels under reducing conditions. After transfer to nitrocellulose membrane using a Transblot semidry transfer cell (Bio-Rad), the membranes were blocked in 5% (w/v) nonfat dried milk in phosphate-buffered saline plus 0.05% Tween for 1 h. The primary antibodies, applied to the immunoblots overnight at 4 °C, were a rabbit anti-NR2A antibody and the 23.F6 antibody, which recognizes the N terminus of NR2A (1McIlhinney R.A.J. Le Bourdellès B. Tricuad N. Molnar E. Streit P. Whiting P.J. Neuropharmacology. 1998; 37: 1355-1367Crossref PubMed Scopus (134) Google Scholar). The NR1 subunit was detected using antibodies directed to residues 600–800 (NMDAR1; PharMingen) or to residues 1–564 of the subunit (Calbiochem) or the C-terminal tail. The primary antibodies were detected using the donkey anti-sheep (Sigma; 1/2000), goat anti-rabbit, or goat anti-mouse (Promega; 1/5000) antibodies conjugated to horseradish peroxidase in conjunction with the chemiluminescence SuperSignal kit (Pierce). Analysis of the oligomeric structure of native protein complexes was achieved by using a modification of the method of BN-PAGE as described previously (31Schagger H. von Jagow G. Anal. Biochem. 1991; 199: 220-231Crossref Scopus (1918) Google Scholar, 32Schagger H. Cramer W.A. von Jagow G. Anal. Biochem. 1994; 217: 220-230Crossref PubMed Scopus (1043) Google Scholar). Membrane samples were prepared by mixing with 1 mg/ml DNase (+10 mm CaCl2, 10 mm MgCl2) and leaving at room temperature for 15 min. An equal volume of 2× BN-PAGE sample buffer (200 mm BisTris, 150 mm6-aminocaproic acid, 2% Triton X-100, pH 7) was added to each sample and left on ice for 15 min. The membrane samples were centrifuged at 100,000 × g and mixed with 5% Serva blue dye. The markers thyroglobulin, bovine serum albumin, apoferritin, and β-amylase were mixed with 5% Serva blue dye and run with the samples on a 5–18% gel containing no detergent. The samples and markers were stacked at 100 V and then run at a constant 500 V (15 mA; 4 °C). The gel was then subjected to a revised Western immunoblot protocol. Excess dye from the top membrane was removed with destain (34% methanol, 10% acetic acid, 2% glycerol in H2O). The marker lanes were stained with Coomassie Blue dye (50% methanol, 10% acetic acid, 0.2% (w/v) Coomassie Blue) and then destained to visualize the protein bands. The section of the membrane containing the proteins was washed with phosphate-buffered saline plus 0.05% Tween and processed in the same manner as the SDS-PAGE immunoblots. In order to calibrate the gel, membranes from cells expressing GABAA receptors of composition α3β3γ2 and purified Torpedo nACh receptors (nAChR) were used. The sera used for identification were anti-α3 for the GABAAreceptor and monoclonal antibody 210 (mAb210) for the nAChR, which recognizes an epitope on the α1 subunit. These receptors and antibodies were generously provided by Merck Sharp & Dohme and Professor Lindstrom, University of Pennsylvania, respectively. HEK-293 cell lysates were prepared by lysis in 1% (v/v) Triton X-100 in Tris-saline lysis buffer containing protease inhibitors and 20 mmiodoacetamide. Sucrose density centrifugation of HEK-293 cell lysates was performed using cell lysates layered on 4-ml continuous 10–40% (w/v) sucrose gradients centrifuged in a SW60 Sorvall rotor for 16 h at 100,000 × g, 4 °C. Twenty fractions were collected from each gradient and analyzed by SDS-PAGE. Five replicate 25-cm2 flasks for each secreted protein were transfected with 10 μg of DNA. Fresh AimV medium (Life Technologies, Inc.), containing no serum and supplemented with 2 mml-glutamine and 100 units/ml penicillin, was added 24 h after transfection, and the cells were left for another 3 days of growth. The medium, with added protease inhibitors, was centrifuged at 1000 × g for 10 min, the supernatant removed and concentrated down 5-fold using Vivaspin 4-ml concentration tubes. The concentrated medium was filtered and analyzed by chromatography on a Superose 6 column (Amersham Pharmacia Biotech) using a Liquid Chromatography Controller LCC-500. The fractions were analyzed by dot-blot, using nitrocellulose membrane (Schleicher & Schuell), and SDS-PAGE and both membranes processed using the Western immunoblot protocol. The structures of the different NR1a constructs used in this study are illustrated in Fig. 1. All of the NR1a truncated subunits, with the exception of the NR1aΔ7, NR1aΔ8, the NR1a N-terminal truncation, and the NR1-GluR1 chimera, formed high affinity binding sites for the glycine antagonist [3H]l-689,560 when expressed in HEK-293 cells (28Sandhu S. Grimwood S. Mortshire-Smith R.J. Whiting P.J. Le Bourdelles B. J. Neurochem. 1999; 72: 1694-1698Crossref PubMed Scopus (2) Google Scholar). Since the NR1 truncations that did not bind the glycine antagonist contain deletions that remove parts of the S1 domain important for the formation of the glycine binding site, these results are in agreement with our current understanding of the structure of other glutamate receptor subunits. HEK-293 cells co-expressing the NR1a constructs containing deletions up to residue 380 in the N terminus (NR1aΔ2-NR1aΔ4) and NR2A were analyzed for the expression of functional channels inXenopus oocytes (Table I). Although unmodified NR1a and the NR1aΔ2 and NR1aΔ3 constructs formed functional channels with NR2A at the cell surface, deletion up to residue 380 within the N-terminal domain of NR1a (NR1aΔ4) abolished NMDA channel function (Table I). Similar data were also obtained in HEK-293 cells (data not shown). The currents found in the NR1aΔ2 and NR1aΔ3 constructs were reduced in amplitude compared with those seen with the wild-type NR1a and NR2A subunits, suggesting that these truncated subunits form smaller numbers of functional channels at the cell surface. Deletion of the C terminus (NR1aΔ1) and TMIV (NR1aΔ5) also produced no detectable functional channels with NR2A. Since NR1aΔ4 forms a functional glycine binding site, these results show that the amino acid sequence before residue 380 is important for the formation of functional NMDA receptors.Table IElectrophysiology data from Xenopus oocytes and HEK-293 cells co-expressing the NR1a truncations with NR2ACurrent inXenopus oocytespANR1a + NR2A2107 ± 460 (n = 13)NR1aΔ1 + NR2A598 ± 272 (n = 6)NR1aΔ2 + NR2A209 ± 37 (n = 7)NR1aΔ3 + NR2A472 ± 149 (n = 6)NR1aΔ4 + NR2A0 (n = 9)NR1aΔ5 + NR2A0 (n = 9)Electrophysiological assays using 10 μm glutamate + 10 μm glycine were carried out in Xenopusoocytes injected with the indicated NR1a constructs and wild-type NR2A. Open table in a new tab Electrophysiological assays using 10 μm glutamate + 10 μm glycine were carried out in Xenopusoocytes injected with the indicated NR1a constructs and wild-type NR2A. Cell surface expression of the NR2A subunit was determined by cell surface biotinylation following co-expression with the NR1a truncations. As reported previously (1McIlhinney R.A.J. Le Bourdellès B. Tricuad N. Molnar E. Streit P. Whiting P.J. Neuropharmacology. 1998; 37: 1355-1367Crossref PubMed Scopus (134) Google Scholar), co-expression of NR1a and NR2A resulted in a streptavidin-isolated 180-kDa immunoreactive band, which was absent when NR2A was expressed alone (Fig.2 A). Progressive deletions to the N-terminal domain of NR1a led to a progressive decrease in the cell surface expression of NR2A until, with NR1aΔ4, there was no detectable surface NR2A (Fig. 2 B). Similarly, no detectable levels of cell surface NR2A were found when the subunit was co-expressed with NR1aΔ7 and NR1aΔ8 (data not shown). Strikingly, the presence of a complete N terminus but deleted C terminus and TMIV of NR1a (NR1aΔ1 and NR1aΔ5, respectively) did not affect cell surface expression of NR2A (Fig. 2 C). It should be noted that generally the levels of immunoreactive NR1a, the NR1a truncations, and NR2A were comparable in the cell lysates (Fig. 2, B andC, lysates). It is unlikely, therefore, that the differences in the surface expression of the constructs reflects differing levels of expression of the proteins. With the exception of NR1aΔ5, the cell surface expression data are in good agreement with the functional channel data, which suggests that the subunits can assemble to form functional channels provided the deletions in the N terminus of NR1a occur before residue 380. The lack of ion channel formation following expression of NR1aΔ5 with NR2A must therefore reflect some other effect of the loss of TMIV from the NR1a subunit, since this truncation does give rise to surface expression of NR2A. Since cell surface expression of NR2A might reflect both subunit association and oligomerization, co-immunoprecipitation studies were performed to determine the level of association of NR2A with the different NR1a truncations. The results showed that truncation of NR1a at the C terminus had no effect on the association of the subunits as illustrated for NR1aΔ5 (Fig.3 A). However, truncation of the N terminus resulted in a progressive loss of co-immunoprecipitating NR2A, which was barely detectable with NR1aΔ7 or NR1aΔ8 (Fig.3 B). Thus, there appears to be some residual association of NR1aΔ4 with NR2A, although this subunit does not give rise to either functional channels or surface expression of NR2A when the subunits are expressed together. The results above could be interpreted as suggesting that the N-terminal region of NR1a from residues 1–380 are critical for subunit association. To test if they are sufficient for this, we have expressed NR2A with NR1a truncated just after the putative re-entrant loop of the second membrane domain and a chimera of NR1 where the N terminus of NR1a replaces that of GluR1 (Fig. 1). Co-expression of the NR1 truncation and NR1-GluR1 chimera with NR2A in HEK-293 cells did not give rise to cell surface expression (Fig.4 A), nor did they co-immunoprecipitate with NR2A (Fig. 4 B). This suggests that residual precipitation seen when NR1aΔ4 and NR2A are co-expressed (Fig. 3 B) is unlikely to be nonspecific and could suggest that the NR1a N-terminal alone may not be sufficient for association with NR2A. The BN-PAGE described by Schagger et al. (31Schagger H. von Jagow G. Anal. Biochem. 1991; 199: 220-231Crossref Scopus (1918) Google Scholar, 32Schagger H. Cramer W.A. von Jagow G. Anal. Biochem. 1994; 217: 220-230Crossref PubMed Scopus (1043) Google Scholar) was adapted to provide a method for the determination of NMDA receptor subunit assembly. In the course of these studies, the BN-PAGE system has also been used to investigate nAChR oligomerization and the domains that are important in glycine receptor assembly (33Nicke A. Rettinger J. Mutschler E. Schmalzing G. J. Recept. Signal Transduct. Res. 1999; 19: 493-507Crossref PubMed Scopus (31) Google Scholar, 34Griffon N. Buttner C. Nicke A. Kuhse J. Schmalzing G. Betz H. EMBO J. 1999; 18: 4711-4721Crossref PubMed Scopus (105) Google Scholar). When BN-PAGE is used to analyze membranes derived from HEK-293 cells expressing the NR1a subunit alone, immunoreactive bands with apparent molecular masses of 200 and 420 kDa can be detected (Fig.5 A). The amount of the 200-kDa NR1a-immunoreactive band detected was variable and could reflect the different expression levels of the subunits in the different membrane preparations. When NR2A was co-expressed with NR1a, the molecular mass of the major NR1a immunoreactive species always shifted to give a broad band with a mean molecular mass of 860 kDa, with two additional immunoreactive bands at 420 and 200 kDa. Consistently, NR2A immunoreactivity could also be detected in the 860-kDa band but was not detected in the 200-kDa NR1a immunoreactive band (Fig. 5 A). However, NR2A could also be detected in an immunoreactive band with a molecular mass of ∼420 kDa when the 23.F6 antibody is used for detection, probably reflecting the greater sensitivity of this antibody compared with the anti-FLAG antibody. The intensity of this band, like that of the NR1a 200-kDa band, could be variable. The expression of the NR2A subunit alone in HEK-293 membranes resulted in streaking of the immunoreactive material, suggesting that the NR2A subunit might not be folding properly in the absence of NR1a (Fig. 5 A,right panel). Because the molecular sizes of the other receptor complexes analyzed by BN-PAGE have been reported to be larger than expected, the modified gel system used here for the NMDA receptors was used to analyze GABAA receptors and Torpedo nAChR of known stoichiometry (32Schagger H. Cramer W.A. von Jagow G. Anal. Biochem. 1994; 217: 220-230Crossref PubMed Scopus (1043) Google Scholar, 35Nicke A
Referência(s)