Subunit Dependencies of N-Methyl-D-aspartate (NMDA) Receptor-Induced -Amino-3-hydroxy-5-methyl-4- isoxazolepropionic Acid (AMPA) Receptor Internalization

N-Methyl-D-aspartate (NMDA) receptor (NMDAR) activity regulates the net number of -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors (AMPAR) at the cell surface by modulating the balance between AMPAR membrane insertion and endocytosis. In this study, we addressed the role of NMDAR subtypes and of NMDAR-mediated Ca influx in the NMDAR-induced endocytosis of GluR2-containing AMPARs in primary murine hippocampal neurons. We found that NMDAR activation enhanced the endocytosis of AMPARs containing the GluR2 splice variants with short, but not long, cytoplasmic tails. NMDA-induced GluR2 endocytosis was completely inhibited by pharmacological block of NR2B-containing NMDARs. In turn, preferential block of NR2A-containing NMDARs did not affect NMDA-induced AMPAR endocytosis, indicating that AMPAR internalization is controlled by a restricted set of NMDARs. The NMDA-induced GluR2 internalization was also observed in the absence of extracellular Na ions, suggesting that membrane depolarization is not a prerequisite for this effect. Furthermore, the activation of Ca -impermeable NMDARs containing the mutant NR1(N598R) subunit failed to enhance AMPAR endocytosis, indicating a requirement of Ca influx directly through the NMDAR channels. In summary, our findings suggest that the NMDAR-induced selective internalization of short C-terminal GluR2-containing AMPARs requires a Ca signal that originates from NMDAR channels and is processed in an NMDAR subtype-restricted manner. The regulation of AMPAR membrane density by NMDAR activity is a key mechanism for the expression of NMDARdependent plasticity via the membrane insertion or endocytosis of AMPARs (Malenka and Bear, 2004; Kim et al., 2005; Zhu et al., 2005). The early determinants in the NMDAR signaling leading to AMPAR internalization are largely unknown. AMPAR channels are formed by heterotetrameric combinations of GluR1–4 subunits that determine the electrophysiological properties of AMPARs and govern their trafficking via C-terminal cytoplasmic domains (Sheng and Kim, 2002; Collingridge et al., 2004). The GluR2 subunit is particularly relevant in that it renders AMPAR channels impermeable to Ca (Dingledine et al., 1999) and is present in the predominant AMPAR heteromers of pyramidal hippocampal neurons (Wenthold et al., 1996). The cytoplasmic tail of the short GluR2 C-terminal splice variant “drives” the constitutive recycling and the NMDAR-induced internalization of AMPARs (Shi et al., 2001; Lee et al., 2004). In contrast, the long GluR2 C-terminal splice variant contributes to This work was funded by a Wellcome Trust Senior Fellowship and a Wellcome grant to R.S., and a Biotechnology and Biological Sciences Research Council grant to M.S. and R.S. C.M.T. was supported in part by a Wellcome Trust Traveling Fellowship, and G.F.R. was the recipient of a Deutsche Forschungsgemeinschaft Emma Noether Fellowship. 1 Current affiliation: Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA. □S The online version of this article (available at http://molpharm. aspetjournals.org) contains supplemental material. Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org. doi:10.1124/mol.105.018580. ABBREVIATIONS: AMPAR, -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; AMPA, -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; NMDAR, N-methyl-D-aspartate receptor; NMDA, N-methyl-D-aspartate; LTP, long-term potentiation; LTD, long-term depression; UTR, untranslated region; (E)GFP, (enhanced) green fluorescent protein; DIV, days in vitro; aCSF, artificial cerebrospinal fluid; KA, kainate; IFP, ifenprodil; NVP, NVP-AAM077; NVP-AAM077, the active diastereomer of (1RS,1S)-PEAQX, which is [(R)-[(S)-1-(4-bromophenyl)-ethylamino](2,3-dioxo-1,2,3,4-tetrahydroquinoxalin-5-yl)-methyl]-phosphonic acid; HS, horse serum; APV, 2-amino-5-phosphonopentanoic acid; Ro, Ro256981; Ro25-6981, ( )-(R*,S*)-(4-hydroxyphenyl)-methyl-4-(phenylmethyl)-1-piperidine propanol; TRITC, tetramethylrhodamine isothiocyanate; E-GluR2-S, EGFP-tagged GluR2, short C terminus; E-GluR2-L, EGFP-tagged GluR2, long C terminus; I-V, current-voltage. 0026-895X/06/6904-1251–1259$20.00 MOLECULAR PHARMACOLOGY Vol. 69, No. 4 Copyright © 2006 The American Society for Pharmacology and Experimental Therapeutics 18580/3102947 Mol Pharmacol 69:1251–1259, 2006 Printed in U.S.A. 1251 http://molpharm.aspetjournals.org/content/suppl/2006/03/16/mol.105.018580.DC1 Supplemental material to this article can be found at: at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from at A PE T Jornals on Jne 2, 2017 m oharm .aspeurnals.org D ow nladed from AMPAR membrane insertion (Kolleker et al., 2003), as do GluR1 and the long C-terminal splice variant of GluR4 (Hayashi et al., 2000; Zhu et al., 2000), or to AMPAR removal during depotentiation (Zhu et al., 2005). The contribution of the long GluR2 C-terminal splice variant to AMPAR endocytosis induced by acute NMDAR activation, however, remains unknown. NMDAR subtypes are heterotetrameric complexes typically formed by NR1 and NR2A-D subunits (Wenthold et al., 2003). In the hippocampus and the neocortex, NR2B predominates in early postnatal life, whereas NR2A is gradually expressed during the first two postnatal weeks, resulting in diheteromeric (NR1/NR2A, NR1/NR2B) and triheteromeric (NR1/NR2A/NR2B) receptor subtypes (Kew et al., 1998). This developmental profile is parallelled in vitro (Li et al., 2002). Both NMDAR-dependent synaptic long-term potentiation (LTP) and depression (LTD), and the NMDAR-induced membrane insertion or removal of AMPARs require a postsynaptic Ca transient elicited by NMDAR activation (Lisman, 1989; Sheng and Kim, 2002). Beyond the model in which the Ca signal kinetics directs the balance of this duality (Lisman, 1989), recent observations suggest that the NR2 subunit composition of NMDAR complexes plays a decisive role in the effects downstream of NMDAR activation (Liu et al., 2004; Massey et al., 2004; Kim et al., 2005). Here, we investigated the role of NMDAR subtypes in the NMDA-induced endocytosis of surface AMPARs containing N-terminally EGFP-tagged GluR2 expressed in hippocampal cultures using a Sindbis virus-derived system. NMDAR activation enhanced the intracellular accumulation of AMPARs containing the short, but not the long C-terminal splice variant of GluR2. This effect was initiated by NR2Bbut not NR2A-containing NMDARs and required Ca influx through the activated NMDAR channels. We propose that NR2B-specific processing of the Ca signals carried by NMDARs constitute an early step in the regulated AMPAR endocytosis. Materials and Methods Materials. All chemicals and drugs were obtained from Sigma (Poole, Dorset, UK), unless specified otherwise. Animal Procedures. Procedures were performed according to the Animals Scientific Procedures Act of 1986, under license of the Home Office. Animals were kept at the Biological Services Unit of the University College London. Sindbis Virus Expression of Recombinant E-GluR2 in Cultured Neurons. We constructed an N-terminally EGFP-tagged GluR2 subunit following a strategy similar to that described by Shi et al. (1999). A multistep cloning procedure resulted in a construct containing from the 5 to the 3 end: the 5 -UTR from Xenopus laevis hemoglobin, rat GluR1 signal peptide, and four amino acid residues beyond the signal-peptide-cleavage site, EGFP coding sequence from vector pEGFP-N1 (Clontech, Mountain View, CA), rat GluR2 flop coding sequence from codon 21 onward, and the Sindbis-specific 3 -UTR and poly(A) tail from the SINrep(nsP2S726) plasmid (Kim et al., 2004). The relevant DNA sequences were confirmed by sequencing and submitted to the European Molecular Biology Laboratory as pGFP-GluR2-Sin (accession number AJ870439). For the generation of short and long C-terminal GluR2 splice variants, the respective sequences were amplified by RT-PCR from total brain mouse E17 RNA, encoding the amino acid residues from Y837KSR (TACAAGTCAAGGG) onward through a sequence common to both splice variants (CCTTGAATTGTAAG, 130 base pairs beyond the STOP codon in the 3 -UTR of the long splice variant), and shuttled into pGFPGluR2-Sin. The splice variant-specific sense primers were 5 -TACAAGTCAAGGGCCGAGGCGAAACGAATGAAGGTGGCAAAGAAT (for GluR2 short) and 5 -TACAAGTCAAGGGCCGAGGCGAAACGAATGAAGATGACCTTGAGC (for GluR2 long). The common antisense primer was 5 -CCTTGAATTGTAAGGAAAGATTACTCGAGG-