FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2045637142> ?p ?o ?g. }

• W2045637142 endingPage "4711" @default.
• W2045637142 startingPage "4704" @default.
• W2045637142 abstract "Article17 August 1998free access AMPA receptors and bacterial periplasmic amino acid-binding proteins share the ionic mechanism of ligand recognition Milla Lampinen Milla Lampinen Department of Biosciences (Division of Biochemistry) and Institute of Biotechnology, P.O.Box 56, FIN-00014 University of Helsinki, Helsinki, Finland VTT Biotechnology and Food Research, FIN-02044 VTT Espoo, Finland Search for more papers by this author Olli Pentikäinen Olli Pentikäinen Department of Biochemistry and Pharmacy, Åbo Akademi University, FIN-20521 Turku, Finland Turku Center for Biotechnology, University of Turku, FIN-20521 Turku, Finland Search for more papers by this author Mark S. Johnson Mark S. Johnson Department of Biochemistry and Pharmacy, Åbo Akademi University, FIN-20521 Turku, Finland Turku Center for Biotechnology, University of Turku, FIN-20521 Turku, Finland Search for more papers by this author Kari Keinänen Corresponding Author Kari Keinänen Department of Biosciences (Division of Biochemistry) and Institute of Biotechnology, P.O.Box 56, FIN-00014 University of Helsinki, Helsinki, Finland VTT Biotechnology and Food Research, FIN-02044 VTT Espoo, Finland Search for more papers by this author Milla Lampinen Milla Lampinen Department of Biosciences (Division of Biochemistry) and Institute of Biotechnology, P.O.Box 56, FIN-00014 University of Helsinki, Helsinki, Finland VTT Biotechnology and Food Research, FIN-02044 VTT Espoo, Finland Search for more papers by this author Olli Pentikäinen Olli Pentikäinen Department of Biochemistry and Pharmacy, Åbo Akademi University, FIN-20521 Turku, Finland Turku Center for Biotechnology, University of Turku, FIN-20521 Turku, Finland Search for more papers by this author Mark S. Johnson Mark S. Johnson Department of Biochemistry and Pharmacy, Åbo Akademi University, FIN-20521 Turku, Finland Turku Center for Biotechnology, University of Turku, FIN-20521 Turku, Finland Search for more papers by this author Kari Keinänen Corresponding Author Kari Keinänen Department of Biosciences (Division of Biochemistry) and Institute of Biotechnology, P.O.Box 56, FIN-00014 University of Helsinki, Helsinki, Finland VTT Biotechnology and Food Research, FIN-02044 VTT Espoo, Finland Search for more papers by this author Author Information Milla Lampinen1,2, Olli Pentikäinen3,4, Mark S. Johnson3,4 and Kari Keinänen 1,2 1Department of Biosciences (Division of Biochemistry) and Institute of Biotechnology, P.O.Box 56, FIN-00014 University of Helsinki, Helsinki, Finland 2VTT Biotechnology and Food Research, FIN-02044 VTT Espoo, Finland 3Department of Biochemistry and Pharmacy, Åbo Akademi University, FIN-20521 Turku, Finland 4Turku Center for Biotechnology, University of Turku, FIN-20521 Turku, Finland *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:4704-4711https://doi.org/10.1093/emboj/17.16.4704 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info In order to identify key structural determinants for ligand recognition, we subjected the ligand-binding domain of the α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA)-selective glutamate receptor GluR-D subunit to site-directed mutagenesis. Based on the analysis of the [3H]AMPA-binding properties of the mutated binding sites, we constructed a revised three-dimensional model of the ligand-binding site, different in many respects from previously published models. In particular, our results indicate that the residues Arg507 and Glu727 represent the structural and functional correlates of Arg77 and Asp161 in the homologous bacterial lysine/ornithine/arginine-binding protein and histidine-binding protein, and directly interact with the α-carboxyl and α-amino group of the bound ligand, respectively. In contrast, Glu424, implicated previously in ionic interactions with the α-amino group of the agonist, is unlikely to have such a role in ligand binding. Our results indicate that glutamate receptors share with the bacterial polar amino acid-binding proteins the fundamental mechanism of amino acid recognition. Introduction Ionotropic glutamate receptors (iGluRs) mediate excitatory neurotransmission in the mammalian central nervous system, and are involved in the control of synaptic strength and in the pathological mechanisms leading to excitotoxic neuronal damage. Like other ligand-gated channels, iGluRs are oligomeric complexes composed of homologous subunits. The subunits encompass 900–1300 residues and, according to the current model for the transmembrane topology, the subunits harbor an extracellular N-terminus, an intracellular C-terminus, three transmembrane segments (M1, M3 and M4) and one pore loop (M2), which inserts into the membrane from the intracellular side and lines the interior of the channel pore (Hollmann et al., 1994; Bennett and Dingledine, 1995; Wo and Oswald, 1995). According to the model, the extracellular part of the receptor comprises the ∼550 residues preceding M1 and the segment between M3 and M4, including two regions S1 and S2 which are sequence-related to bacterial polar amino acid-binding proteins (Nakanishi et al., 1990; O'Hara et al., 1993; Stern-Bach et al., 1994). This relationship between glutamate receptors and the bacterial proteins, considered together with the similarity in the chemical nature of their cognate ligands (amino acids), suggests that (i) the S1 and S2 segments form the neurotransmitter-binding site in iGluR subunits, and (ii) the mechanism of amino acid binding is similar in these two classes of proteins. Importantly, three-dimensional structures of several bacterial amino acid-binding proteins have been determined to high resolution, providing detailed information on their ligand-binding mechanism (Kang et al., 1991; Oh et al., 1993, 1994a, b; Yao et al., 1994; Hsiao et al., 1996). In the absence of direct structural information on glutamate receptors, the bacterial proteins may thus serve as templates for modeling the ligand-binding site. Furthermore, in bacterial proteins, ligand binding is accompanied by closing of the two-lobed binding domain around the ligand, a mechanism which in iGluRs may be exploited to couple ligand binding to the transmembrane movements leading to channel opening (Wo and Oswald, 1995; Laube et al., 1997; Paas, 1998). The postulated structural relationship between the ligand-binding sites of iGluRs and the bacterial proteins is supported by some experimental evidence. Studies on chimeric and mutated receptors indicate that the S1 and S2 segments indeed determine the ligand pharmacology of iGluRs (Kuryatov et al., 1994; Stern-Bach et al., 1994; Kuusinen et al., 1995). Site-directed mutagenesis studies on α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) (Uchino et al., 1992; Li et al., 1995; Mano et al., 1996; Kawamoto et al., 1997) and N-methyl-D-aspartate (NMDA) receptors (Kuryatov et al., 1994; Wafford et al., 1995; Hirai et al., 1996; Laube et al., 1997; Wood et al., 1997; Anson et al., 1998), and the kainate-binding proteins of chick (Paas et al., 1996b) and goldfish (Wo and Oswald, 1996) have identified a number of residues in the S1 and S2 segments which contribute to ligand pharmacology. In homology-based structural models, these residues largely localize along the cleft between the predicted two lobes of the ligand-binding domain, supporting similarity between the binding sites. (Paas et al., 1996b; Sutcliffe et al., 1996; Laube et al., 1997). In the bacterial lysine/arginine/ornithine-binding protein (LAOBP; Oh et al., 1993, 1994a) and histidine-binding protein (HBP; Oh et al., 1994b; Yao et al., 1994), ligand recognition is mediated by multiple hydrogen bonds and by salt bridges (Oh et al., 1994a, b; Yao et al., 1994). In particular, the ionized side chains of conserved arginine (Arg77) and aspartate (Asp161) residues neutralize the charged α-carboxylate and α-amino groups of the bound amino acid ligand, respectively. Interestingly, bacterial leucine/isoleucine/valine-binding protein (LIVBP; Sack et al., 1989), which has a structural fold similar to that of LAOBP and HBP, uses exclusively hydrogen bonds and hydrophobic stacking for ligand binding, indicating that ionic interactions are not a universal feature of amino acid binding. It may be noteworthy, however, that no clear sequence similarity exists between LIVBP and the polar amino acid-binding proteins (including the S1 and S2 segments of iGluRs). The possible contribution of ionic interactions in ligand recognition by glutamate receptors is still unclear. An arginine residue in the S1 segment of iGluR subunits (equivalent to Arg507 in the GluR-D AMPA receptor) which corresponds to Arg77 has been proposed to interact either with the α-carboxylate (Kawamoto et al., 1997; Laube et al., 1997) or with the γ-carboxylate (Sutcliffe et al., 1996) of glutamate. In contrast, the identity and even the existence in the S2 segment of iGluRs of an anionic residue corresponding to Asp161 has remained unclear, and the molecular models presented for iGluR–ligand interaction differ from the bacterial binding mode in this respect. As the interactions between the receptor and α-carboxyl and α-amino groups are expected to be the most highly conserved in amino acid-binding proteins, the above-mentioned results raise serious doubts on the validity of the structural models of iGluRs reported so far. In the present study, we provide evidence that Arg507 and Glu727 in the AMPA receptor subunit GluR-D are functionally and structurally equivalent to Arg77 and Asp161 in the bacterial proteins, indicating that the fundamental mechanism of amino acid binding is conserved. In addition, our results demonstrate that structural determinants important for the binding of the competitive antagonists are present in both lobes of the predicted two-lobed binding site. Results As the invariant (charged) α-amino and α-carboxyl groups represent a structural feature which is common to iGluR agonists and the ligands of the bacterial polar amino acid-binding proteins, we examined whether the interactions between the receptor and these functional groups in the ligand are mediated by homologous charged residues in these two classes of proteins. In a multiple alignment of iGluR subunits and bacterial polar amino acid-binding proteins (Figure 1A), Arg507 (numbering herein refers to the GluR-D subunit and starts from the first translated residue; Keinänen et al., 1990) and Glu727 show up as the potential correlates of Arg77 and Asp161 of LAOBP and HBP, which interact with the α-carboxyl and α-amino group of the bound amino acid. We examined the role of these and other, mostly charged residues in ligand binding by introducing site-directed mutations into the ligand-binding domain of the AMPA-selective GluR-D subunit, produced as a soluble, secreted S1–S2 fusion protein (Kuusinen et al., 1995). By selecting this model system, which reproduces the ligand-binding characteristics of the parental membrane-bound receptor, we wished to minimize the indirect contributions of the mutations to the assembly of the receptor, transit to the cell surface, coupling of ligand binding with channel gating, and desensitization. Figure 1.Sequence alignment of three GluR subunits with four bacterial proteins illustrating the regions around LAOBP residues Arg77 (A) and Asp161 (B) involved in binding of the α-aminocarboxyl group of the amino acid ligand. The shading indicates sequence identity (bold type) or isofunctionality. The sequences used in the alignment: rat iGluR subunits GluR-D (DDBJ/EMBL/GenBank accession No. M36421), GluR-6 (Z11548) and NR2B (M91562); Escherichia coli glutamine-binding protein (QBP; X14180), Salmonella typhimurium histidine-binding protein (HBP; V01373) and lysine/arginine/ornithine-binding protein (LAOBP; V01368), and Corynebacterium glutamicum glutamate-binding protein (GLUB; X81191). Download figure Download PowerPoint Role of Arg507 The functional importance of the arginine residue corresponding to Arg507 in GluR-D has been demonstrated in NMDA (Kuryatov et al., 1994; Wafford et al., 1995) and AMPA receptors (Uchino et al., 1992; Kawamoto et al., 1997) and in the chick kainate-binding protein (Paas et al., 1996b). This residue is conserved in all iGluR subunits and aligns with Arg77 of LAOBP and HBP (Figure 1A), which forms a salt bridge with the negatively charged α-carboxylate of the bound amino acid and participates in a hydrogen bonding network within the binding pocket (Oh et al., 1993). Expression of the GluR-D S1–S2-binding site carrying the conservative mutation R507K resulted in the secretion into the culture medium of a polypeptide with the expected size (Figure 2A), suggesting that the mutant protein was not grossly misfolded. The culture supernatants containing the mutant polypeptide were dialyzed extensively and tested for [3H]AMPA binding (Figure 2A). In contrast to the wild-type S1–S2 polypeptide, which bound the radioligand with a high affinity (Kd 37 nM), no specific binding was observed with the R507K mutant even at high radioligand concentrations (Figure 2A). Figure 2.Site-directed mutagenesis of charged residues. (A) Cationic residues Arg507, Lys431, Lys432 and Lys471. (B) Anionic residues Glu710, Glu727 and Glu424. Expression and [3H]AMPA-binding activity of the mutant binding sites are shown. The upper panel shows an anti-Flag Western blot of the insect cell supernatants. The dashes on the left indicate the positions of the mass markers: 55.6, 42.7, 36.5 and 26.6 kDa. The lower panel indicates the specific binding of 5 nM [3H]AMPA as ranked qualitatively in four categories: −, no radioligand binding above background (500 d.p.m.); +, 1000–4000 d.p.m./25 μl sample; ++, 5000–14 000 d.p.m.; +++, > 15 000 d.p.m./25 μl. Download figure Download PowerPoint Role of other cationic residues As glutamate agonists share, in addition to the α-carboxylate, another distal anionic group (γ-carboxylate of L-Glu), the role of additional positively charged residues in the predicted ionic interactions was examined by subjecting three lysine residues in the S1 segment, Lys431, Lys432 and Lys471, that could be located in the predicted ligand-binding pocket depending upon the alignment, to mutagenesis. Substitutions of lysine residues at positions 431 and 432 by serine did not result in any significant alterations in the [3H]AMPA-binding affinity or pharmacological profile (Figure 2A; Table I). The role of the lysine residue at position 471, previously shown to contribute to ligand pharmacology of AMPA (Uchino et al., 1992; Li et al., 1995) and NMDA receptors (Wafford et al., 1995), and of chick kainate-binding protein (Paas et al., 1996a, b), was investigated by analyzing the ligand-binding properties of site-directed mutants K471R, K471H and K471A. All three were secreted into the culture medium at levels comparable with that of the wild-type binding domain, and bound [3H]AMPA with essentially unchanged properties except for the K471H mutant which showed decreased affinity both for the radioligand and for unlabeled L-glutamate (Figure 2A; Table I). Table 1. Ligand-binding properties of the mutated binding sites Mutation Kd (nM) Ki (μM) L-Glutamate Kainate DNQX Wild-type 37 ± 13 0.19 ± 0.09 1.9 ± 0.47 0.49 ± 0.18 E424D 29 ± 2.2 1.3 ± 0.53a 3.3 ± 0.83 1.0 ± 0.11 E424Q 28 ± 7.0 1.1 ± 0.16b 7.0 ± 3.6 0.48 ± 0.13 E424A 35 ± 6.2 0.25 ± 0.04 11 ± 1.6b 1.3 ± 0.10a K431S 60 ± 17 0.30 ± 0.13 2.2 ± 0.23 0.35 ± 0.14 K432S 33 ± 4.8 0.38 ± 0.20 2.8 ± 0.69 ND K471R 29 ± 7.9 0.29 ± 0.07 3.7 ± 2.1 0.66 ± 0.31 K471H 90 ± 22a 0.77 ± 0.15a 1.9 ± 0.38 0.48 ± 0.12 K471A 37 ± 14 0.31 ± 0.04 2.9 ± 0.58 0.55 ± 0.26 Y472F 380 ± 80b 0.20 ± 0.07 2.2 ± 0.33 2.01 ± 0.47a R507K NB – – – E710D 24 ± 7.2 0.23 ± 0.07 1.6 ± 0.30 0.33 ± 0.15 E710A 50 ± 21 ND 2.6 ± 0.49 ND E727D 4.8 ± 0.7b 1.7 ± 0.39b 1600 ± 740a 0.013 ± 0.006a E727Q NB – – – E727S NB – – – E727A NB – – – [H]AMPA-binding affinity and the relative affinities (Ki values) of unlabeled L-glutamate, kainate and DNQX are indicated. The numbers represent average values ± SD from 3–10 independent experiments. ND not determined; NB, no measurable binding. The mutants which differ significantly from the wild-type are indicated by superscripts: a P <0.02, b P <0.002. Role of Glu727 A conserved anionic residue in the S2 segment of iGluR subunits, Glu727 in GluR-D, aligns with Asp161 (Figure 1B). Although the overall sequence conservation between iGluRs and the polar amino acid-binding proteins is very low in this region, contributing to the differences in many earlier published alignments, this anionic residue invariably is preceded by a non-polar aliphatic or aromatic residue at the −3 and an alanine residue at the −4 position. Asp161 is involved in charge neutralization of the positively charged primary amino group of the bound amino acid in both LAOBP and HBP (Oh et al., 1993, 1994b; Yao et al., 1994). Accordingly, we chose Glu727 as a target for site-directed mutagenesis. As a safeguard against ambiguities in the alignment, another nearly conserved negatively charged residue, Glu710, was also mutagenized. In order to test the importance of the polarity and charge of the side chain, mutations E727D, E727Q, E727S and E727A were studied. At position 710, two mutations (E710D and E710A) were studied. The level of expression and secretion into the culture medium were similar to the wild-type control for each of the mutants (Figure 2B). In the [3H]AMPA-binding assay, however, mutations at position 727 resulted in dramatic effects on the ligand-binding properties whereas mutations at position 710 had no effect (Figure 2B; Table I). The presence of an uncharged residue (aliphatic or polar) at position 727 always resulted in the total loss of measurable [3H]AMPA binding. In contrast, the E727D mutant which preserved the negative charge of the side chain (and, thus, the possibility to participate in ionic interactions), bound [3H]AMPA and displayed a highly interesting ligand-binding phenotype (Figure 3; Table I). The radioligand, [3H]AMPA, bound to the E727D mutant with a 5- to-10-fold increased affinity (Kd 5 nM), but in the displacement assay, unlabeled agonists L-glutamate and kainate bound with at least 10- and 1000-fold decreased affinities, respectively. Figure 3.Binding of [3H]AMPA to GluR-D/S1S2 E727D. (A) Saturation binding isotherm; (B) displacement of 5 nM radioligand by unlabeled L-Glu, AMPA, kainate and DNQX. Download figure Download PowerPoint Role of Glu424 Glu424 is located in the N-terminal ‘hot spot’ region of the S1 segment and aligns with the Tyr14 in LAOBP and HBP, a residue which contacts the distal end of the side chain of the bound amino acid. Although in the bacterial proteins this residue is not involved in recognition of the α-amino group of the ligand, a recent molecular model was used to suggest that Glu424 forms an ionic bond with the α-ammonium group of L-glutamate in AMPA and kainate receptors (Sutcliffe et al., 1996), with glutamate in the opposite orientation opposite to that seen in defined three-dimensional structures of both the polar amino acid-binding proteins, the model proposed by Paas and co-workers (Paas et al., 1996b) and in our model (see below). Furthermore, mutations at the corresponding residue have been reported to exert effects on the ligand pharmacology of glutamate receptors (Uchino et al., 1992; Kuryatov et al., 1994; Mano et al., 1996; Wo and Oswald, 1996; Laube et al., 1997). Therefore, we prepared and characterized three different Glu424 mutants, E424D, E424Q and E424A. These were all expressed and secreted at a level comparable with that of the wild-type ligand-binding domain (Figure 2B; Table I). In the radioligand-binding assay, all three mutants displayed [3H]AMPA binding with an affinity at the wild-type level, which, considering the loss of charge with the E424Q and E424A mutants, argues against a role for a coulombic interaction with the ligand α-amino group. In a ligand displacement analysis, the E424D and E424Q mutants showed a slightly (∼5-fold) decreased affinity for L-glutamate, whereas the E424A mutant displayed wild-type glutamate affinity (Table I). Interaction with quinoxalinedione antagonist We have shown previously that the structural determinants responsible for binding of the quinoxalinedione antagonists are also present within the S1–S2 ligand-binding domain (Kuusinen et al., 1995). Therefore, it was interesting to test the potency of DNQX to inhibit [3H]AMPA binding to the mutated binding sites. Of all the mutations described above, only two, E424A and E727D, showed alterations: DNQX interacted with the E424A mutants with a slightly decreased affinity, and with the E727D mutant with an increased affinity (Table I). As the chick kainate-binding protein residue Tyr71, corresponding to Tyr472 in GluR-D, has been implicated in antagonist binding, we characterized the ligand-binding properties of an additional mutant Y472F (Paas et al., 1996b). In agreement with the results of Paas and co-workers (Paas et al., 1996b), mutation Y472F resulted in a decrease in the antagonist interaction. At the same time, however, binding of the radioligand was even more severely affected (Table I). Molecular model of the binding site A three-dimensional model of the structure of the S1–S2 polypeptide was constructed on the basis of homology to the bacterial polar amino acid-binding proteins (Figure 4). This model structure was used to visualize and interpret interactions of GluR-D with ligand (L-glutamate) as indicated by mutagenesis of the receptor. Arg507 and Glu727 occupy the corresponding positions of Arg77 and Asp161 of the HBP and LAOBP, illustrated in Figure 4B and C, showing the striking similarity between the ligand complexes of the HBP and the proposed model of GluR-D, especially as regards the α-amino and α-carboxylate groups of the bound amino acid. Key stabilizing interactions in the GluR-D model include hydrogen and ionic bonds (all <3.0 Å) between the side chain of Arg507 and the α-carboxylate oxygens of the ligand; the side chain of Glu727 and the α-amino group of the ligand; hydrogen bonds between the main-chain NH groups of Thr502 and Thr677 and one oxygen each of the α-carboxylate of the ligand; and a hydrogen bond between the main-chain carbonyl of Thr502 and the side chain of Arg507. Mutagenesis of the semi-conserved cationic residues Lys431, Lys432 and Lys471 in the S1 segment did not drastically alter the ligand-binding behavior of the receptor, excluding any substantial contribution of these residues to ionic stabilization of the proximal (‘α’) or distal (‘γ’) anionic group of the ligand. The structural basis for the subclass-selective agonist pharmacologies of AMPA, NMDA and kainate receptors is likely to be related to the interaction of the ligand's distal anionic group with the receptor. To address this problem in detail, a more extensive mutagenesis study, currently in progress, is needed. Figure 4.(A) Model of the closed-form of the S1–S2 ligand-binding domain of GluR-D based on the three-dimensional structures of two bacterial periplasmic binding proteins: LAOBP with bound lysine (Protein Data Bank structure code: 1lst) and HBP with bound histidine (1hsl). N and C refer to the N- and C-terminal ends, respectively, of the portion of the receptor encompassing the S1 and S2 domains; I and III indicate the N- and C-terminal ends of the membrane-spanning units, I–III, not a part of the model and located between domains S1 and S2. These regions are replaced by a linker peptide in the recombinant S1–S2 construct. Bound glutamate is shown as a ball and stick figure; Arg507, blue sphere; Glu727, green sphere. α-Helices are shown in red, and β-strands in cyan, the S1 connecting regions in white, and the S2 connecting regions in yellow. (B) Detailed view of interactions between the α-amino and α-carboxylate groups of bound glutamate (Glutamate) and main-chain and side-chain atoms of the GluR-D model. Only the side chains of Arg507 and Glu727 are shown. Predicted hydrogen bonds are indicated by dashes. Oxygen atoms (red), nitrogen atoms (blue), carbon (gray). (C) Corresponding interactions between bound histidine in the X-ray structure of HBP. Only the side chains of Arg77 and Asp161 are shown. The experimental evidence and the model reported here show that the analogs of the HBP residues Arg77 and Asp161 are GluR-D residues Arg507 and Glu727, respectively. This figure was prepared using MOLSCRIPT (Kraulis, 1991), RASTER3D (Bacon and Anderson, 1988) and RENDER (Merritt and Murphy, 1994). Download figure Download PowerPoint Discussion Our results indicate that GluR-D residues Arg507 and Glu727 are functionally equivalent to Arg77 and Asp161 of LAOBP and HBP, interacting with the oppositely charged α-carboxyl and α-amino groups, respectively, of the bound amino acid. The importance of Arg507 to ligand binding agrees with previous studies showing that substitutions of the corresponding arginine residue in GluRs invariably lead to loss of function (Uchino et al., 1992; Wafford et al., 1995; Hirai et al., 1996; Kawamoto et al., 1997; Laube et al., 1997). In our model, Arg507 interacts with the agonist's α-carboxylate as observed in the crystal structures of LAOBP and HBP and in the model presented for the glutamate site of the NR2B subunit (Laube et al., 1997), and not with the γ-carboxylate of the ligand as suggested by Sutcliffe and co-workers (1996). We consider the interaction with the α-carboxylate, rather than with the γ-carboxylate, more likely for two reasons. First, the invariant structure (carboxylate) of the α-anionic group contrasts with the diversity of distal anionic structures of natural and synthetic iGluR agonists and, therefore, is more in agreement with the absolute conservation of Arg507. Secondly, the orientation of glutamate in the binding site model presented by Sutcliffe and co-workers (1996) differs from amino acids bound to LAOBP and HBP, in contrast with what would be expected from the similarities in the sequences and chemical structures of the ligands of these two classes of proteins (i.e. iGluRs and the bacterial proteins). The major new finding of the present study is the functional similarity between Glu727 (in GluR-D) and Asp161 in bacterial LAOBP and HBP in neutralizing the positive charge of the ligand's α-ammonium group. Previous findings on inactivation of the receptor due to non-conservative mutations introduced at this position (Mano et al., 1996; Paas et al., 1996b) have been interpreted as indirect structural alterations rather than direct effects on ligand recognition. Williams and co-workers (1996) reported that the replacement of Asp732 (equivalent to Glu727) of the NR1 subunit receptor by a glutamate residue resulted in a drastic (>4000-fold) decrease in the glycine sensitivity of the NR1/NR2A NMDA receptor expressed in Xenopus oocytes, and speculated that possible mechanisms might be direct ligand interaction and indirect effects on channel gating. In the present study, the mutation of Glu727 to an aspartate, in contrast to three mutations that neutralized the negative charge (Glu→Gln, Glu→Ser and Glu→Ala), not only retained high-affinity AMPA binding but produced a ligand-binding profile which differs from the wild-type receptor for all agonist and antagonist ligands tested. The binding affinity for AMPA was increased 5- to 10-fold, whereas the affinities for glutamate and quisqualate were decreased by one order of magnitude. For kainate, the drop in the binding affinity was more dramatic, ∼1000-fold, corresponding to a standard free energy change of 17 kJ/mol. The fact that glutamine and serine at position 727 do not, in contrast to glutamate and aspartate, support high-affinity agonist binding indicates the importance of the negative charge. We therefore predict that the ionized γ-carboxylate of Glu727 is in the vicinity (∼2.8 Å) of the positively charged α-amino group of the agonist and, in addition to coulombic attraction, forms a hydrogen bond with it. The opposite effects of the E727D mutation on binding affinities of the three agonists may illustrate the different stereochemical interactions between the ligand and the slightly altered binding site. Previously, Glu424 or a corresponding residue has been suggested to interact directly via coulombic attraction and hydrogen bonds with the α-amino group of the ligand in the NR1 and NR2B subunits of the NMDA receptor (Laube et al., 1997), chick kainate-binding protein (Paas et al., 1996b) and in AMPA and kainate receptors (Sutcliffe et al., 1996). In support of this, substitutions of the glutamate at this position in the NR2B subunit of the NMDA receptor and in the GluR-A AMPA receptors resulted in decreased agonist sensitivities (Uchino et al., 1992; Mano et al., 1996; Laube et al., 1997). In a ligand-binding study of chick kainate receptor, a charge-neutralizing mutation of the glutamate at this position to valine led to a 100-fold decrease in the affinity for kainate but did not influence affinity for glutamate (Paas et al., 1996b). The three different substitutions at residue 424 characterized in this study, including one (Glu→Ala) with an uncharged aliphatic side chain, all displayed high-affinity AMPA binding, and in two cases, slightly altered binding of unlabeled glutamate. This indicates that ionic interactions or hydrogen bonds between the ligand and Glu424" @default.
• W2045637142 created "2016-06-24" @default.
• W2045637142 creator A5036290531 @default.
• W2045637142 creator A5056949200 @default.
• W2045637142 creator A5089062316 @default.
• W2045637142 creator A5090456130 @default.
• W2045637142 date "1998-08-17" @default.
• W2045637142 modified "2023-10-16" @default.
• W2045637142 title "AMPA receptors and bacterial periplasmic amino acid-binding proteins share the ionic mechanism of ligand recognition" @default.
• W2045637142 cites W11163000 @default.
• W2045637142 cites W1485790693 @default.
• W2045637142 cites W1516646376 @default.
• W2045637142 cites W1591736283 @default.
• W2045637142 cites W1608735505 @default.
• W2045637142 cites W1686297271 @default.
• W2045637142 cites W1744070583 @default.
• W2045637142 cites W1805473917 @default.
• W2045637142 cites W1845040160 @default.
• W2045637142 cites W1871333004 @default.
• W2045637142 cites W1966041739 @default.
• W2045637142 cites W1991740949 @default.
• W2045637142 cites W2005913521 @default.
• W2045637142 cites W2006406301 @default.
• W2045637142 cites W2007760520 @default.
• W2045637142 cites W2019019017 @default.
• W2045637142 cites W2028133023 @default.
• W2045637142 cites W2035189018 @default.
• W2045637142 cites W2040850636 @default.
• W2045637142 cites W2042732542 @default.
• W2045637142 cites W2050050663 @default.
• W2045637142 cites W2053040211 @default.
• W2045637142 cites W2053303814 @default.
• W2045637142 cites W2057537207 @default.
• W2045637142 cites W2058870251 @default.
• W2045637142 cites W2063461600 @default.
• W2045637142 cites W2065283382 @default.
• W2045637142 cites W2078248419 @default.
• W2045637142 cites W2086050619 @default.
• W2045637142 cites W2091765514 @default.
• W2045637142 cites W2107349404 @default.
• W2045637142 cites W2108778496 @default.
• W2045637142 cites W2117177933 @default.
• W2045637142 cites W2134083569 @default.
• W2045637142 cites W2161252555 @default.
• W2045637142 cites W2168311980 @default.
• W2045637142 cites W2308472860 @default.
• W2045637142 cites W2336145557 @default.
• W2045637142 cites W2407495012 @default.
• W2045637142 doi "https://doi.org/10.1093/emboj/17.16.4704" @default.
• W2045637142 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/1170799" @default.
• W2045637142 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/9707429" @default.
• W2045637142 hasPublicationYear "1998" @default.
• W2045637142 type Work @default.
• W2045637142 sameAs 2045637142 @default.
• W2045637142 citedByCount "64" @default.
• W2045637142 countsByYear W20456371422013 @default.
• W2045637142 countsByYear W20456371422014 @default.
• W2045637142 countsByYear W20456371422015 @default.
• W2045637142 countsByYear W20456371422016 @default.
• W2045637142 countsByYear W20456371422019 @default.
• W2045637142 countsByYear W20456371422020 @default.
• W2045637142 crossrefType "journal-article" @default.
• W2045637142 hasAuthorship W2045637142A5036290531 @default.
• W2045637142 hasAuthorship W2045637142A5056949200 @default.
• W2045637142 hasAuthorship W2045637142A5089062316 @default.
• W2045637142 hasAuthorship W2045637142A5090456130 @default.
• W2045637142 hasBestOaLocation W20456371421 @default.
• W2045637142 hasConcept C104317684 @default.
• W2045637142 hasConcept C111472728 @default.
• W2045637142 hasConcept C116569031 @default.
• W2045637142 hasConcept C12554922 @default.
• W2045637142 hasConcept C138885662 @default.
• W2045637142 hasConcept C160268369 @default.
• W2045637142 hasConcept C170493617 @default.
• W2045637142 hasConcept C201663137 @default.
• W2045637142 hasConcept C515207424 @default.
• W2045637142 hasConcept C547475151 @default.
• W2045637142 hasConcept C55493867 @default.
• W2045637142 hasConcept C61174792 @default.
• W2045637142 hasConcept C86803240 @default.
• W2045637142 hasConcept C89611455 @default.
• W2045637142 hasConcept C95444343 @default.
• W2045637142 hasConceptScore W2045637142C104317684 @default.
• W2045637142 hasConceptScore W2045637142C111472728 @default.
• W2045637142 hasConceptScore W2045637142C116569031 @default.
• W2045637142 hasConceptScore W2045637142C12554922 @default.
• W2045637142 hasConceptScore W2045637142C138885662 @default.
• W2045637142 hasConceptScore W2045637142C160268369 @default.
• W2045637142 hasConceptScore W2045637142C170493617 @default.
• W2045637142 hasConceptScore W2045637142C201663137 @default.
• W2045637142 hasConceptScore W2045637142C515207424 @default.
• W2045637142 hasConceptScore W2045637142C547475151 @default.
• W2045637142 hasConceptScore W2045637142C55493867 @default.
• W2045637142 hasConceptScore W2045637142C61174792 @default.
• W2045637142 hasConceptScore W2045637142C86803240 @default.
• W2045637142 hasConceptScore W2045637142C89611455 @default.
• W2045637142 hasConceptScore W2045637142C95444343 @default.
• W2045637142 hasIssue "16" @default.

• next