Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Activity-dependent regulation of -̆amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors… Lanius, Ruth Anne 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1995-983051.pdf [ 3.1MB ]
JSON: 831-1.0088854.json
JSON-LD: 831-1.0088854-ld.json
RDF/XML (Pretty): 831-1.0088854-rdf.xml
RDF/JSON: 831-1.0088854-rdf.json
Turtle: 831-1.0088854-turtle.txt
N-Triples: 831-1.0088854-rdf-ntriples.txt
Original Record: 831-1.0088854-source.json
Full Text

Full Text

ACTIVITY-DEPENDENT REGULATION OF oc-AMINO-3-HYDROXY-5-METHYL-4-ISOXAZOLE PROPIONIC ACID (AMPA) RECEPTORS IN RAT NEOCORTEXbyRUTH ANNE LANIUSB.Sc., University of Victoria, 1990A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Neuroscience Programme)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAAugust 1994© Ruth Anne Lanius, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of_________________The University of British ColumbiaVancouver, CanadaDate / aY / 19 SDE-6 (2/88)‘IABSTRACTThe study of x-amino-3-hydroxy-5-methyl-isoxazole propionic acid (AMPA)receptors is of great interest given that these receptors mediate most fastexcitatory synaptic neurotransmission in brain and are involved in neuroplasticphenomena such as long-term potentiation and long-term depression.Understanding the molecular mechanisms involved in the regulation of AMPAreceptors may therefore provide insight into many aspects of neuronal function,including normal synaptic transmission and synaptic neuroplasticity. To assessactivity-dependent regulation of cortical AMPA receptors, radioligand bindingmethods employing the competitive AMPA receptor antagonist[3H]-6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) were used to study the effects of variousregulatory stimuli on the agonist binding site of AMPA receptors in rat corticalslices.AMPA receptor regulation was studied in response to a variety of stimuli,including agonist (AM PA), pharmacological depolarization (veratridine +glutamate), as well as the phosphorylating enzymes calcium (Ca2+)/calmodulin-dependent kinase II (CaMKH) and protein kinase A (PKA).Treatment with AMPA or veratridine led to approximately 20% decreases in[3H]-CNQX binding. Similar decreases in[3H]-CNQX binding were seenfollowing treatment with CaMKII (—35%) and PKA (—30%).The effects of AMPA and veratridine could be blocked by inhibitors of CaMKlland PKA, suggesting that phosphorylation reactions are involved in AMPAreceptor regulation by AMPA and veratridine. Moreover, loperamide, a nonspecific inhibitor of voltage-gated Ca2 channels was able to inhibit the AMPA‘l(and veratridine-induced regulation of AMPA receptors. These resultssuggested that AMPA and veratridine may result in the activation of voltagegated Ca2 channels, in turn leading to changes in[3H]-CNQX binding throughthe activation of CaMKII and/or PKA. Ca2 alone was able to decrease[3H]-CNQX binding over a concentration range of 0.1 to 1 mM, an effect which couldbe blocked by specific inhibitors of CaMKII or PKA.These data indicate that AMPA and veratridine, agents intended to mimicaspects of synaptic transmission, lead to the regulation of AMPA receptors via aCa2 influx through voltage-gated Ca2 channels and the activation of specificphosphorylating enzymes. These results provide for a novel mechanism otAMPA receptor regulation and may establish the framework for a clearerunderstanding of the modulation of synaptic activity in normal conditions,following modifications of synaptic strength, and in some forms ofneuropathology.ivTABLE OF CONTENTSAbstract iiTable of contents ivList of Figures ViiiList of Tables XList of Abbreviations XiGENERAL INTRODUCTION 1Neocortex: Structure and Function 1Synaptic Transmission 2Receptors: The First Postsynaptic Stage 3lonotropic Receptors 4lonotropic Receptor Regulation: Definitions 6lonotropic Receptor Regulation by Phosphorylation 8AMPA Receptors 8Phosphorylation of other lonotropic Receptors 9Kainate Receptors 9NMDA Receptors 10Nicotinic Achetyicholine Receptors 11GABAA Receptors 11Glycine Receptors 12The Role of AMPA Receptors in Neuroplasticity 13LTP 13VLTD 15The Role of AMPA Receptors in NeurodegenerativeDiseases 17Thesis Objectives 18GENERAL METHODS 19The Cortical Slice as a Model System 19“Living” Slice Preparation: Advantages and Disadvantages 19Animals 21Preparation of Brain Slices 21Evidence for Slice Viability 22Radioligand Methods 23Radioligand 24[3H]-CNQX Binding Assays 24Statistical Analysis 25Chapter 1 Characterization of a[3H]-CNQX Binding Site 26Introduction 26Materials and Methods 27Results 30Discussion 38Chapter 2 AMPA Receptor Regulation by Agonist and DepolarizingStimuli 42Introduction 42Neurotransmitter Receptor Binding Response to AgonistStimulation 43lonotropic Receptors 43G-Protein Coupled Receptors 44Neurotransmitter Receptor Response to CellularDepolarization 46lonotropic Receptors 46G-protein Coupled Receptors 46Materials and Methods 48A Brief Overview of Methods 48Experiments Designed to Study the Effects of Agonistand Depolarizing Stimuli on[3H]-CNQX Binding 48Control Experiments to Distinguish between Competitionand Regulation Effects of Veratridine and AMPA 49Neurotoxicity of AMPA and Glutamate 512-Deoxy-D-[14CJGlucose Uptake Studies 52Results 53Regulation Experiments 532-Deoxy-D-[14C] Glucose Uptake Studies 55Discussion 65Chapter 3 AMPA Receptor Regulation by Phosphorylation 71Introduction 71Materials and Methods 73Results 76Discussion 83Chapter 4 AMPA Receptor Regulation by Agonist and DepolarizingStimuli requires CaMKII and PKA 90Introduction 90Materials and Methods 91V iiResults 92Discussion 98Chapter 5 Ca2-Dependence of AMPA Receptor Regulation 100Introduction 100Materials and Methods 101The Effects of Ca2,Na, and K on[3H]-CNQX Binding 101The Effects of Loperamide on AMPA Receptor Regulationby Agonist and Depolarizing Stimuli 102Results 103The Effects of Ca2,Na, and K on[3H]-CNQX Binding 103The Effects of Loperamide on AMPA Receptor Regulationby Agonist and Depolarizing Stimuli 104Discussion 110GENERAL DISCUSSION 117Summary of Data 117A Model of AMPA Receptor Regulation 118AMPA Receptor Regulation: Implications for NeuralFunction 119AMPA Receptor Regulation in Normal SynapticNeurotransmission 120AMPA Receptor Regulation in Synaptic Neuroplasticity 120Receptor Regulation and Neuropathology 121Concluding Remarks 122FUTURE DIRECTIONS 123REFERENCES 125viiiLIST OF FIGURESFigure 1: Association and Dissociation Time Course of[3H]-CNQX Binding 33Figures 2A &B: Competition of[3H]-CNQX Binding with SpecificAMPA Analogues 35Figure 3: Saturation Binding Isotherm for Specific[3H]-CNQXBinding 36Figure 4: Effects of Freeze/Thaw Treatment or Incubation withTris-Acetate on[3H]-CNQX Binding 37Figure 5: Concentration-Response of AMPA 56Figure 6: Concentration-Response of v+g 57Figure 7: Time course of AMPA-Induced Regulation of[3H]-CNQX Binding 58Figure 8: Reversibility of AMPA-Induced Regulation of[3H]-CNQX Binding 59Figure 9: Time course of v+g-Induced Regulation of[3H]-CNQX Binding 60Figure 10: Reversibility of v+g-lnduced Regulation of[3H]-CNQX Binding 61Figure 11: Rinse Out Times for AMPA and v-i-g 62Figure 12: AMPA Receptor Regulation in Response to AMPAand v-i-g 63Figure 13: 14C-2-Deoxyglucose Uptake of Cortical Slices AfterTreatment with AMPA or Glutamate 64Figure 14: Time Course of CaMKII-lnducedRegulation of[3H]-CNQX Binding 78Figure 15: Reversibility of CaMKII-Induced Regulation of[3H]-CNQX Binding 79Figure 16: Time Course of PKA-lnduced Regulation of [3HJ-CNQX Binding 80Figure 17: Reversibility of PKA-lnduced Regulation of[3H]-CNQX Binding 81Figure 18: Regulation of[3H]-CNQX Binding by CaMKII, PKAandTPA 82Figure 19: Protein Kinase Dependence of AMPA-InducedRegulation of[3H]-CNQX Binding 94Figure 20: Protein Kinase Dependence of AMPA-InducedRegulation of[3H]-CNQX Binding 95Figure 21: Protein Kinase Dependence of v+g-lnducedRegulation of[3H]-CNQX Binding 96Figure 22: Protein Kinase Dependence of v-i-g-lnducedRegulation of[3H]-CNQX Binding 97Figure 23: Effects of Ca 2+ on[3H]-CNQX Binding 105Figure 24: Effects of Ca 2+ on[3H]-CNQX Binding 106Figure 25: Effects of Na and K on[3H]-CNQX Binding 107Figure 26: Effects of Loperamide on AMPA-Induced Regulationof[3H]-CNQX Binding 108Figure 27: Effects of Loperamide on v-t-g-lnduced Regulationof[3H]-CNQX Binding 109LIST OF TABLESTable 1: Pharmacological Agents Used to Analyze[3H]-CNQXBinding to AMPA Receptors in Rat Cortical Slices 29Table 2: Compounds Used to Study AMPA Receptor Regulation 48Table 3: Protein Kinase Inhibitors 75Table 4: Protein Kinase Inhibtors Used to Inhibit Regulation byAMPA and v+g 92xxiLIST OF ABBREVIATIONSAMPA cc-amino-3-hydroxy-5-methyl-4-isoxazole propionateAP5 DL-2-am ino-5-phosphonovaleric acidBark 13-adrenergic kinaseBOAA 13-N-Oxalylamino-L-alanineCa2 CalciumCaM Kit a2/caImodulin-dependent protein kinase I IC1 ChlorideCNQX [3H]-6-cyano-7-nitroquinoxaline-2,3-dionecAM P Cyclic 3’, 5’ -adenosine monophosphate2-DG 2-deoxy-D-giucoseDul+ Dulbecco’s+GABAA y-amino-butyric acidAG-Protein Guanyl-nucleotide-binding proteinIP Inhibiting peptidePotassiumKN-62 1 -[N,O-bis(1 ,5-isoquinolinesulfonyl)-N-methyl-L-tyrosyl]-4-phenylpiperazineLTD Long-term depressionLTP Long-term potentiationmACh Muscarinic acetylcholineMg2 MagnesiumMgCI2 Magnesium chlorideNa SodiumXiinACH Nicotinic acetyicholineNBQX 6-nitro-7-sulfamoylbenzo-(f)quinoxalin-213 dioneNMDA N-methyl-D-aspartateNMS [3H]-N-methyl scopolaminePKA Protein kinase APKC Protein kinase CQNB Quinuclidinyl benzilateR p-CAM PS R p-adenosine 3’, 5’-cyclic monophosphothioateTBST [35S]-butyl bicyclophosphorothionateTPA 1 2-O-tetradecanoylphorbol 13-acetateTTX Tetrodotoxinv-i-g veratridine + glutamateTCP N-(1(2-thienyl)cyclohexyl)-3,4-piperidine1GENERAL INTRODUCTIONOverviewNeocortex: Structure and FunctionThe cerebral cortex forms the layer of gray matter which covers the entiresurface of the cerebral hemispheres in mammals. It is characterized by alaminar organization of its cellular components. In humans, the cortex has anarea of approximately 1 square meter, makes up two-thirds of the neuronalbrain mass, and contains approximately three-quarters of all brain synapses(Carpenter, 1988; Rakic, 1988).Cortical neurons develop from segments of the telencephalic vesicle. Cells inthe germinal zone which surround the lumen migrate peripherally to form thecortical sheath. Early during fetal development, cortical neurons begin to formsix characteristic horizontal layers. All cells formed simultaneously migrate tothe same cortical lamina; cells migrating later pass through deep layers to formthe more superficial laminae. In primates and humans, the first cortical neuronsare generated early during embryonic development with the full complement ofcortical neurons being reached during the first half of gestation (Carpenter,1988; Rakic, 1988).The highest level of information processing in the brain occurs in the cortex. Inmammals, three basic types of neocortex can be distinguished: sensory, motor,and association. Sensory cortex (visual, auditory or somatosensory) receivesdirect sensory input from specific thalamic nuclei. It receives messages from the2sense organs as well as messages of touch and temperature from throughoutthe body. Motor cortex receives indirect input from cerebellum, dorsal columnnuclei, and other cortical regions and is involved in the control and coordinationof motor output. Association cortex, in contrast to sensory cortex, mainlyreceives a variety of input from other cortical areas. Association cortexbecomes dominant in amount to sensory cortex in primates and humans andappears to be responsible for the majority of cognitive function (Carpenter,1988).A key feature of cortical neurons is their ability to alter their responsecharacteristics to changing input conditions. Such activity-dependentmodifications can be of many forms, including long-term potentiation (LTP)(Kirkwood et at., 1993) and long-term depression (LTD) (reviewed by Linden,1994) as well as modifications in sensory experience (Wiesel and Hubel, 1963).Such effects are often referred to as examples of ‘neuroplasticity’ and arethought to be qualitatively similar to those processes underlying learning andmemory. While cortex is not unique in expressing neuroplasticity, the variety ofneuroplastic phenomena in cortex appears to be greater than for any otherbrain region.Synaptic TransmissionIn the brain, neuron-to-neuron communication occurs at chemical or electricalsynapses. Chemical synapses, representing the bulk of synapses in. the brain,involve the release of a chemical mediator (neurotransmitter) from thepresynaptic neuron and initiate current flow in the postsynaptic cell (Eccles,1976).3The initial stage in neurotransmission begins with a presynaptic action potentialreaching the axon terminal. Depolarization of the terminal is followed by anincrease in calcium (Ca2+) influx via voltage-gated Ca2+ channels leading tothe release of neurotransmitter by exocytosis of synaptic vesicles into thesynaptic cleft. The Lt postsynaptic step of neurotransmission then occurs withbinding of the neurotransmitter to postsynaptic receptor proteins. The binding ofa neurotransmitter to its target receptor leads to the opening of specific ionchannels, ionic flow across the neural membrane, and a change in the cell’smembrane potential. Cation flow into the neuron gives a depolarization whichmay result in a postsynaptic action potential. In contrast, anion flow into theneuron leads to a hyperpolarization of the target neuron. Neurotransmitterbinding to target receptors is thus the first crucial step in postsynaptic signaltransduction (Edelman et al., 1987).Receptors: The First Postsynaptic StaaeSynaptic receptors can be divided into two major classes, ionotropic andguanyl-nucleotide-binding protein (G-protein) coupled. lonotropic receptors areligand-gated ion channels, which, upon neurotransmitter/receptor interactions,transiently open an associated ion channel leading to an ionic current flux(reviewed by Raymond et al., 1993a). In contrast, G-protein coupled receptorscan indirectly activate ion channels. For these receptors, the interactionbetween the receptor and the ion channel is mediated by a G-protein.Neurotransmitter binding to G-protein coupled receptors usually leads to theactivation of effector enzymes that produce intracellular second messengers,such as cyclic 3’,5’-adenosine monophosphate (CAMP). The secondmessenger, in turn, acts on the channel to modulate channel function through4the activation of enzymes such as protein kinases. In some cases, G-proteinscan also interact with ion channels directly. Synaptic transmission mediated byG-protein coupled receptors is slower compared to that of ionotropic receptorsand usually leads to modulatory changes by altering the threshold of cells(Edelman et al., 1987b).lonotroDic ReceDtorslonotropic receptors (ligand-gated ion channels) are pentameric structuresconsisting of homologous subunits which surround a central aqueous pore.Each subunit consists of four transmembrane spanning regions labeled Mlthrough M4. An intracellular cytoplasmic loop located between M3 and M4 hasbeen shown to contain phosphorylation sites for a variety of protein kinases inall ionotropic receptor subunits cloned so far. An extracellular amino terminalsequence contains the agonist binding site for the receptor (reviewed byRaymond et al., 1993a; Swope et al., 1992; Huganir and Greengard, 1990).Examples of such receptors are the nicotinic acetylcholine (nACh), ‘‘-aminobutyric acidA (GABAA), glycine, and some glutamate receptors.The primary function of ionotropic receptors is to mediate fast synaptictransmission in the brain (reviewed by Raymond et al., 1993a). The binding ofspecific neurotransmitters to target ionotropic receptors on the postsynapticmembrane leads to rapid alterations in the membrane permeability to particularions, providing the basis for neural inhibition and excitation. For example, theamino acid, GABA, is the primary inhibitory neurotransmitter in brain. WhenGABA binds to GABAA receptors, associated chloride (Clj channels open andCl flows into the cell giving rise to a hyperpolarizing transmembrane potential5(reviewed by Stephenson, 1988). Glutamate provides the basis for rapid neuralexcitation in the brain (Monaghan et al., 1989). When glutamate binds toglutamatergic ionotropic receptors, it opens associated cation channels,allowing some ions to enter the cell and giving rise to a depolarizingtransmembrane potential.lonotropic glutamate receptors can be divided into three distinct subtypes,adopting the name of their preferred agonist: i) N-methyl-D-aspartate (NMDA),ii) kainate, and iii) -amino-3-hydroxy-5-methyl-4-isoxazole propionate (AM PA)(Monaghan et at., 1989). In addition to these three ionotropic subtypes, a Gprotein-coupled (metabotropic) glutamate receptor has been identified(Monaghan et al., 1989). Recent molecular cloning studies have identifiedhomologous subunits for the NMDA (NMDA Ri, NMDA R2A-D), kainate (GluR5-7, Ka 1, KA 2), and AMPA (GIuR 1-4) ionotropic receptor subtypes. NMDAreceptors are characterized by a voltage-dependent magnesium (Mg+) block aswell as by a high permeability to Ca2. In contrast, kainate and AMPAreceptors mainly flux Na (Seeburg, 1993).AMPA receptors have been the focus of many studies. They appear to mediatethe majority of fast excitatory synaptic currents in the central nervous system(Seeburg, 1993). In addition, they seem to play key roles in several forms ofneuroplasticity, including LTP (Maren et al., 1993: Tocco et al., 1992) and LTD(reviewed by Linden, 1994; Linden and Conner, 1993). The mechanisminvolved in the regulation of these receptors has thus become an importantissue in the study of nervous system function.6Protein phosphorylation catalyzed by protein kinases has been recognized asone of the primary mechanisms for regulating the activities for a wide variety ofproteins (reviewed by Edelman et al., 1987), including neurotransmitterreceptors (reviewed by Swope et al., 1992; Huganir and Greengard, 1990).Phosphorylation of all ionotropic receptors identified so far has been shown tolead to the modification of receptor function (reviewed by Raymond et al.,1993a; Swope et al., 1992; Huganir and Greengard, 1990).The recent sequencing of the AMPA receptor has allowed the identification ofconsensus sequences for phosphorylation by serine/threonine kinases such asCa2/calmodulin-dependent protein kinase (CaMKII) and protein kinase C(PKC) on the major intracellular loop of all four AMPA receptor subunits (GluR1-4) (Boulter et al., 1990; Keinanen et al., 1990). Moreover, low affinityphosphorylation sites for protein kinase A (PKA) may be present on GluR 1-4(Kennelly and Krebs, 1991). Since these enzymes are known to be highlyexpressed in the CNS (Hanson and Schulman, 1992; Nairn et al., 1985;Walaas et al., 1983 a&b), and the role of phosphorylation reactions in theregulation of receptor function has been widely documented, studies havecentered on the potential roles of these enzymes in the regulation of AMPA andother ionotropic receptors.lonotroic R eceDtor Regulation: Definitions‘Regulation’, ‘sequestration’, ‘up/down-regulation’ and ‘desensitization’ areterms which are often used to describe the functional modification of receptors.Regulation is often used to describe alterations in receptor characteristics, suchas receptor number or affinity. ‘Sequestration’ refers to a process of7internalization of cell surface receptors leaving them unresponsive toextracellular signals. Sequestration has been suggested to play a role in theregulation of G-protein coupled receptors (for review see Hausdorff et al., 1990).The extent to which sequestration operates to regulate ionotropic receptors isnot known. Receptor-mediated responses may also be diminished by thedegradation of existing receptors thereby reducing functional receptor number.This is termed ‘down-regulation’ and may be defined as a decrease in overallreceptor number from a ‘receptor pool’ which includes both cell surface andinternal receptors (reviewed by Hausdorff et al., 1990). ‘Up-regulation’ refers toan increase in overall receptor number, sometimes involving synthesis of newreceptor proteins. For events such as sequestration or up/down-regulation thetime course is usually relatively long, with a time frame of many hours to days(Maloteaux et al., 1987; Klein et al., 1979), whereas receptor regulation oftenoccurs within minutes to hours (Yang et al., 1994; Kitamura et al., 1993;Tabuteau et aL, 1993; Lanius and Shaw, 1993; Shaw and Scarth, 1991; Shawet al., 1989; Luqmani et al., 1979; Siman and Klein, 1979). Studies examiningreceptor regulation, sequestration, and up/down-regulation typically useradioligand binding assays to assess the characteristics of the neurotransmitterbinding site. In the following chapters the term ‘regulation’ will be taken tomean any change in receptor binding characteristics, Qy.In contrast to receptor regulation, receptor desensitization is a term which refersto a decrease in cellular response to agonist in the continued presence ofagonist. Receptor desensitization is usually studied by employingelectrophysiological techniques to measure post-synaptic receptor-evokedcurrents. Alterations in such currents may be due to various aspects of ionchannel properties, including the probability of channel opening, the number of8active channels, as well as channel unitary conductance (Raymond et al.,1993b; Greengard et al., 1991; Wang et at., 1991). A crucial point here is thatalterations in receptor-mediated currents may reflect either changes in bindingat the agonist binding site and/or alterations in the associated ion channel.In the section below, the effects of phosphorylation on receptor regulation aswell as receptor-mediated currents will be discussed for AMPA and otherionotropic receptors.lonotroDic ReceDtor Regulation by PhosDhorylationAMPA ReceDtorsModification of AMPA receptor function by CaMKII, PKC and PKA- mediatedphosphorylation has been the focus of several studies (Soderling et al., 1994;Tan et al., 1994; McGlade-McCulloh et at., 1993; Keller et at., 1992; Shaw et at.,1992b; Greengard et al., 1991). A recent study has shown that activation ofCaMKII present endogenously in synaptosomes resulted in strongly enhancedphosphorylation of GluRl. Activation of PKC slightly enhanced phosphorylationof GluRl, white activation of PKA resulted in little, if any, phosphorylation ofGluR 1. Similar results were reported when postsynaptic densities were used asthe source of glutamate receptors and endogenous protein kinases. In thispreparation only CaMKII strongly enhanced phosphorylation of GluRl; PKC andPKA had no significant effect. Activation of CaMKII also led to an enhancementof AMPA receptor-mediated currents in cultured hippocampal neurons,suggesting that AMPA receptors are highly modulated by their phosphorylationstate (Tan et at., 1994; McGlade-McCulloh et at., 1993).9In contrast to the above studies, PKA has been demonstrated to increase theopening frequency and the mean open time of the AMPA receptor channels inhippocampal pyramidal neurons (Greengard et aL, 1991). Furthermore, Kelleret al. (1992) reported that bath application of a membrane permeable analogueof cAMP, an activator of PKA, potentiated currents through AMPA receptorchannels comprised of GluRl and GIuR3 subunits expressed in Xenopusoocytes. Other results have shown that exposure of adult cortical rat slices tothe catalytic subunit of PKA resulted in a 30% decrease in[3H]-6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) binding. This effect could be completelyblocked using a specific peptide inhibitor of PKA (Shaw et a!., 1992). Thuswhile the effects of PKA differed in the studies cited above, a role for PKA inAMPA receptor regulation seems warranted.PhosDhorvlation of other lonotroDic ReceDtorsReceptor regulation by phosphorylation is not unique to the AMPA receptorpopulation. Consensus sequences for various protein kinases have beenidentified for many other ionotropic receptors, including the excitatory kainateand NMDA receptors as well as the nACh, GABAA, and glycine receptors(reviewed by Swope et al., 1992). In the following sections, a brief review of therole of phosphorylation in the functional modification of these receptors will bepresented.Kainate ReceDtorsThe amino acid sequence of the high affinity kainate GIuR 6 subunit reveals astrong consensus sequence for phosphorylation by PKA (Egebjerg et a!., 1991).10Raymond et al. (1993b) have recently reported that the GIuR 6 glutamatereceptor expressed in mammalian kidney cells was directly phosphorylated byPKA, and that intracellularly applied PKA was able to increase the amplitude ofthe glutamate mediated response. Site specific mutagenesis of the serineresidue 684 abolished PKA-mediated phosphorylation of this site andeliminated the potentiation of the glutamate evoked currents. PKA has alsobeen shown to result in an increase in kainate mediated currents in otherpreparations, including retinal horizontal cells (Liman et al., 1989) and culturedhippocampal neurons (Wang et al., 1991). Furthermore, Ortega and Teichberg(1990) reported that the in vitro phosphorylation of a putative subunit of thechick cerebellar kainate receptor by the catalytic subunit of PKA appeared to bedecreased in the presence of the agonists kainate and domoate, thussuggesting a possible interaction between agonist binding and receptorphosphorylation.NMDA ReceDtorsThe excitatory NMDA receptor population has also been shown to bemodulated by phosphorylation. In isolated trigeminal neurons, PKC potentiatedNMDA-mediated currents by reducing the voltage-dependent Mg2 block ofNMDA-receptor channels (Chen and Huang, 1992). These results aresupported by various studies which also reported a PKC-mediatedenhancement of the amplitude of NMDA-receptor mediated currents in oocytes(Kelso et al., 1992), hippocampal CAl neurons (Aniksztejn et al., 1992), andspinal cord neurons (Gerber et aL, 1989). In addition, radioligand bindingstudies have shown that phorbol ester-induced activation of PKC decreased theKd but not the Bmax of[3H]-MK-801 (non-competitive NMDA receptor11antagonist) binding. In contrast, phorbol ester treatment had no effect on [3H]-CGS-19755 (competitive NMDA antagonist) binding (Kitamura et at., 1993).These findings suggested that activation of PKC specifically affects the NMDAreceptor channel without having an effect on the NMDA agonist binding site.Nicotinic Acetyicholine ReceDtorsThe nACh receptors are similar to the excitatory amino acid receptors in thattheir activation leads to depolarizing transmembrane potentials by increasingthe membrane permeability to cations. The amino acid sequence of variousnACh receptor subunits reveals consensus sequences for phosphorylation byPKA, PKC, as well as a tyrosine-specific protein kinase (reviewed by Swope etal., 1992; Huganir and Greengard, 1990). Electrophysiological studies inTorpedo californica have shown that phosphorytation of nACh receptors byPKA, PKC, and an unidentified protein tyrosine kinase can result in thefunctional modification of these receptors, leading to an increase in the rate ofdesensitization (Hopfield et at., 1988; Huganir et at., 1986).GABAA ReceDtorsScrutiny of the sequence of GABAA receptor subunits also revealed theexistence of consensus sequences for PKA, PKC, and a protein tyrosine kinase(reviewed by Swope et al., 1992; Huganir and Greengard, 1990). Purifiedpreparations of GABAA receptors could be phosphorylated directly by both PKCand PKA (Browning et al., 1990; Kirkness et at, 1989), as well as by anunidentified kinase present in partially purified preparations of GABAAreceptors (Sweetnam et aL, 1988). Physiological studies examining the effects12of PKA on GABAA receptor function have led to contradictory results. Exposureof cells to the catalytic subunit of PKA or activators of PKA has been shown toattenuate (Porter et at., 1990; Tehrani et at., 1988) or, conversely, potentiate(Kano and Konnerth, 1992) GABAA receptor-mediated currents. Recent studieshave demonstrated that the attenuation of GABAA receptor-mediated currentsby PKA is associated with increased phosphorylation of the GABAA receptor(Leidenheimer et al., 1991; Moss et al., 1992). Moreover, PKA-mediatedphosphorylation of the GABAA receptor appeared to decrease the extent ofrapid desensitization to agonist (Moss et at., 1992). Radioligand bindingstudies have reported a decrease in[3H]-SR 95531 (competitive GABAAantagonist) binding after treatment with the catalytic subunit of PKA. This effectcould be blocked by a specific PKA inhibiting peptide (Shaw et al., 1992b).Glycine RecertorsGlycine receptors are another inhibitory receptor population which have beenshown to be regulated by protein phosphorylation. Song and Huang (1990)have reported that intracellular application of cAMP to cultured trigeminalneurons leads to a 70% increase in the amplitude of glycine evoked currents.In contrast, Ruiz-Gomez et al. (1991) have shown that glycine receptors purifiedfrom rat spinal cord are phosphorylated by PKC, but not PKA on a specificserine residue. It is therefore possible that PKA was exerting its effectsindirectly on glycine receptors. Vaello et al. (1992) subsequently showed thatthe rate of phosphorylation of the glycine receptor by PKC was higher in thepresence of agonist than in the presence of antagonists. Moreover, theyreported a 3-fold decrease in[3H]-strychnine binding when this protein was13phosphorylated by PKC, thus providing evidence that phosphorylation can alteragonist binding.In summary, receptor phosphorylation appears to be a common way ofmodulating ionotropic receptor function and is by no means unique to the AMPAreceptor population. Phosphorylation by a variety of protein kinases has beenshown to regulate the receptor associated ion channel as well as the agonistbinding site for several receptor populations studied thus far. The functionalconsequences of such phosphorylation, however, can be diverse depending onthe specific receptor or kinase involved.The Role of AMPA ReceDtors in NeuroDlasticityThe term neuroplasticity broadly refers to the ability of neurons to alter somefunctional property in response to alterations in input. Such alterations areoften long-lasting, if not permanent, and are thought to be qualitatively similar tothe processes providing the basis of learning and memory (reviewed by Blissand Collingridge, 1993). AMPA receptors have been suggested to play a rolein two important model systems of neuroplasticity, namely LTP (Maren et al.,1993; Tocco et al., 1992) and LTD (reviewed by Linden, 1994).LTPIn the hippocampus and cortex, periods of intense electrical stimulation (tetani)(usually a train of 50-100 stimuli at 100 Hz or more) in specific neural circuitsresult in an enhancement of synaptic transmission. This phenomenon, referredto as LTP, can persist for many hours in an in vitro slice preparation or in the14anaesthetized animal, and for days when induced in the freely moving animal.LTP has been studied most extensively in pyramidal cells of the CAl region inhippocampus following stimulation of the Schaffer collateral pathway.However, it can also be elicited in area CA3 of the hippocam pus, in theneocortex, as well as in various other areas in the central nervous system (seeBliss and Collingridge, 1993 for review).NMDA receptor activation and subsequent influx of Ca2+ has often been shownto be necessary for the induction of some forms of LTP (for review see Bliss andCollingridge, 1993). However, increasing evidence points to a crucial role ofAMPA receptors in the maintenance of LTP. Tocco et al. (1992) have reportedan increase in[3H]-AM PA but not 3 H] N-( 1 -(2-thienyl)cyclohexyl)-3,4-piperidine([3H]-TCP) (non-competitive NMDA receptor antagonist) binding inhippocampal sections of animals that exhibited LTP after stimulation of thepertorant pathway. This increase in binding could be blocked by administrationof NMDA receptor antagonists prior to tetanic stimulation. Further studies byMaren et al. (1993) showed an increase in[3H]-AMPA binding in the dentategyrus that was highly correlated with the magnitude of LTP recorded in thisstructure. These changes in[3H]-AMPA receptor binding were attributable tochanges in AMPA receptor number rather than affinity as determined bysaturation binding analyses.Studies have also shown a gradual increase in sensitivity to AMPA followingthe induction of LTP. This effect, however, took at least one hour to reach itsmaximum and could be blocked by DL-2-amino-5-phosphonovaleric acid(AP5), a NMDA antagonist, as well as by K-252b, a potent PKC inhibitor. Theselatter results together with the increase in[3H]-AMPA binding following NMDA15receptor-dependent LTP described above suggest that a NMDA-dependentmodification of AMPA receptors may be responsible for the maintenance ofsome forms of LTP (Reymann et al, 1990).Recent reports have suggested that AMPA receptors can exist in one of twoforms termed ‘flip’ and ‘flop’. The ‘flip’ form produces a higher conductancechannel and desensitizes less rapidly than the ‘flop’ form (Sommer et al., 1990).Collingridge and Singer (1991) have speculated that the activation of kinasesvia NMDA receptor activation during the induction of LTP may result in anincrease in the ‘flip’ form of AMPA receptors, leading to an increase in synapticefficacy through higher conductance, less rapidly desensitizing AMPAreceptors.LTDLTD is a form of synaptic plasticity in which prolonged periods of low-frequencystimulation (usually 1-3 Hz) elicit a synaptic depression that is input-specific andpersists for many hours. Although this phenomenon has not been studied inchronic preparations, studies using acute intact animals or cell cultures haveshown the synaptic depression to persist for many hours (reviewed by Linden,1994). LTD in cerebellum appears to manifest itself as a decrease in AMPAmediated currents as a result of protein kinase activity (Linden et al., 1993,1991; Ito and Karachot, 1992). Several processes are necessary and sufficientfor the induction of cerebellar LTD: Na influx through AMPA receptor ionchannels, Ca2+ influx via voltage-gated Ca2+ channels, and metabotropicglutamate receptor activation and subsequent activation of PKC leading toalterations in AMPA receptor function (Linden, 1994; Linden et al., 1993; 1991).16Although Na influx through voltage-gated Na channels is sufficient to induceLTD in some cases, activation of AMPA receptors appears to be much moreeffective (Linden et al., 1993). In contrast to the induction of LTD, there is strongevidence that the expression of this phenomenon is mediated entirely byalterations in the number or sensitivity of postsynaptic AMPA receptors as aresult of protein kinase C activity (Linden, 1994; Linden et al., 1993; 1991). Itoand Karachot (1992) further reported that activation of PKC by phorbol esterscould mimic LTD, thus suggesting that LTD may result from phosphorylationand concomitant regulation of AMPA receptors by PKC.It is interesting to note that the expression of cerebellar LTD is thought to bemediated by the modulation of AMPA receptors as a result of PKC activation.McGlade-McCulloh et al. (1993) did not report a change in the AMPA receptor-mediated currents in rat hippocampal pyramidal cells by PKC. Thesedifferences in the action of PKC, however, may be explained by differences inGIuR subunits among the different brain regions studied. Cerebellar Purkinjecells are known not to possess GluRl and GluR4, whereas both of thesesubunits are known to be present in cortex and hippocampus (Martin et al.,1993). In support of this view, it has been reported that receptorphosphorylation has differential effects depending on the subunit compositionof the receptor (Krishek et al., 1994; Moss et al., 1992).In summary, AMPA receptors clearly play a significant role in some forms ofneuroplasticity such as LTP and LTD. Protein kinase-mediated modulation ofthe AMPA receptor population has been suggested to occur after the inductionof both LTP and LTD and may be involved in the maintenance of both of thesephenomena.17The Role of AMPA ReceDtors in Neurodecienerative DiseasesNeuronal cell death may be caused by increased intracellular Ca2+ as a resultof an overstimulation of excitatory amino acid receptors by glutamate oraspartate (reviewed by Lipton and Rosenberg, 1994). Such ‘excitotoxic’ actionshave been reported to contribute significantly to neuronal loss occurring afteracute and chronic forms of neuronal degeneration (Meidrum and Garthwaite,1991). lschemia, an acute form of neural degeneration, has been shown to beassociated with AMPA receptor-mediated excitotoxicity. The main evidence forthis comes from pharmacological studies which showed that the competitiveAM PA receptor antagonist 6-nitro-7-su lfamoylbenzo(f)q uinoxalin-2, 3-d lone(NBQX) appeared to protect against cortical damage resulting from completeglobal ischemia even when there was significant post-ischemic delay beforedrug administration (Sheardown et al., 1990). These latter findings suggestedan important role for AMPA receptors in the mechanisms underlying ischemiaand suggested that overstimulation of AMPA receptors may have effects onother processes leading to neuronal cell death. For example, AMPA receptorshave been shown to play a key role in neurolathyrism, a chronic spasticneurodegenerative syndrome found predominantly in East Africa and SouthernAsia. Neurolathyrism is associated with dietary consumption of the chick peaLathyrus sativus, a plant containing the toxin 13-N-Oxalylamino-L-alanine(BOAA). BOAA acts as an AMPA receptor agonist (Bridges et al., 1989) onspinal cord neurons and eventually leads to selective destruction of upper andlower motor neurons (Meidrum and Garthwaite, 1991). Although it remains tobe determined why a toxin given systemically leads to the lesioning of only veryselective neuronal populations, a possible explanation is that only AMPA18receptors present in upper and lower motor neurons contain binding sites forthis particular toxin.Thesis ObiectivesThe literature clearly reveals important roles for AMPA receptors in normalexcitatory neural transmission in the central nervous system and links thesereceptors to some forms of neuroplasticity and neuropathology. A clearerunderstanding of the molecular mechanism involved in activity-dependentregulation of AMPA receptors may provide new insights into these events.Studies of the characteristics and regulation of these receptors are thusimportant goals and the major focus of this thesis.Using radioligand binding assays employing the competitive AMPA receptorantagonist[3H]-6-cyano-7-nitroqu inoxaline-2,3-dione (CNQX), activity-dependent AMPA receptor regulation in rat neocortex will be examined.Chapter 1 will describe characterization studies of a[3H]-CNQX binding site.Chapter 2 will use[3H]-CNQX binding to examine the regulation of the AMPAreceptor population to agonist and depolarizing stimuli, stimuli which arebelieved to mimic aspects of synaptic neurotransmission seen in viva Chapter3 will examine the effects of phosphorylating enzymes (i.e. protein kinases) on[3H]-CNQX binding. Chapter 4 will establish that agonist and depolarizingstimuli exert their effects through the activation of particular protein kinases.Finally, Chapter 5 will examine the role of specific ionic species in the cascadeof events leading to kinase-mediated regulation of the AMPA receptor agonistbinding site.19GENERAL METHODSThe Cortical Slice as a Model SystemModels are simplified representations of biological processes. They are used insituations when it is not possible to study the phenomenon of interest in vivo, asis often the case in the biological sciences (Zbinden, 1992). At a neuralsystems level, a model preparation might employ an isolated neural circuitrather than the whole structure, e.g. the study of LTP in cortical or hippocampalslices rather than in the intact brain. In the latter case, a large number ofvariables remain uncontrolled. Such variables include the concentration ofadministered drugs, temperature, as well as synaptic connections with otherbrain structures. Many such variables, however, can be better controlled inslice or other isolated preparations. The work described in this thesis uses acortical slice preparation in which most cells are alive and in relatively intactneural circuits in order to study activity-dependent AMPA receptor regulation.“Living’ Slice PreDaration: Advantages and DisadvantagesThe study of neurotransmitter receptors in the central nervous system has oftenemployed techniques in which the cells are completely or partially disrupted.Examples of such techniques include the widely used homogenatepreparations (for review see Bylund and Yamamura, 1990; Yamamura et al.,1990) or those employing thin in vitro sections (for review see Young andKuhar, 1987; Unnerstall et al., 1982). Although these techniques have providedsubstantial information about receptor characteristics as well as the regionaldistribution of many receptors, they are limited in terms of studies of receptor20function in living (intact) cells under physiological conditions (for review seeShaw and Wilkinson, 1994; Wilkinson et al., 1986). A “living” tissue preparationtherefore offers the following advantages over homogenate or in vitroautoradiographic techniques:(1) Since the receptors are on living cells which are still in relatively intactneural circuits, assays can be performed under conditions which are as close tophysiological conditions as possible. Neurons are thus affected by processeswhich occur in living cells such as depolarization and neurotransmitter release.(2) Homogenate and in vitro autoradiographic techniques may conceal theeffects of experimentally-induced receptor regulation. For example, anexperimental paradigm that may lead to the sequestrationlinternalization ofreceptors as part of the regulation process may fail to reveal changes inreceptor binding because of the inability to distinguish internal from membranebound receptors. In a “living” slice preparation in which the majority of plasmamembranes are intact, however, membrane and internal receptors can bedistinguished. This is usually accomplished by the use of hydrophilic versuslipophilic radioligands. Due to the inability of hydrophilic ligands to crossplasma membranes, these radioligands label membrane bound receptors only.Lipophilic radioligands, on the other hand, are able to cross plasmamembranes and therefore label both membrane and internal receptors. (3) Theuse of relatively intact tissue may diminish the release of proteolytic enzymes forreceptors and/or neurotransmitters whose activity might lower binding levels.Although the above mentioned points are distinct advantages of the “living”slice technique, there are also several disadvantages to using this technique.(1) Receptor regulation studies using radioligand binding cannot determine on21which cell types receptor regulation is occurring. It is therefore impossible todistinguish neuronal from glial cell receptor regulation using this preparation.(2) Naturally occurring agonists, e.g. glutamate (see Chapters 1 and 2; Laniusand Shaw, 1992), GABA (Shaw and Scarth, 1991), and ACh (Van Huizen et al.,1989) often fail to show an effect in both competition and regulationexperiments. This may, in part, be due to the presence of functional uptakesites as well as degradative enzymes which are still present in intact tissue (forreview see Shaw and Wilkinson, 1994). It is therefore difficult to examine theroles of putative neurotransmitters in receptor regulation using a “living” slicepreparation. However, this problem can be overcome by employing agonistanalogues. (3) Only radioligands whose dissociation rate constants are slowerthan the time required to rinse out the free ligand can be used for radioligandbinding in intact slices. It is for that reason that antagonists generally appear towork better than agonists.AnimalsAll experiments were performed with adult male Sprague-Dawley rats (>60days). The rats were raised under controlled lighting conditions (12:12 h lightdark cycle) and had free access to food and water.PreDaration of Brain SlicesAnimals were anaesthetized with halothane and sacrificed by decapitation. Thewhole brain was then removed in less than 1 minute from the time ofdecapitation and immersed in a modified ice-cold Dulbecco’s phosphatebuffered saline solution (Gibco, Grand Island, NY). The modified medium22‘Dulbecco’s plus’ (Dul+), contained glucose (1 mg/mL), Hepes (25 mM), andhydrogen peroxide (0.003%), the latter added as a source of molecular oxygen(Walton and Fulton, 1983).After removal of the brain, blocks of cortex approximately 5 mm in length by2mm in width were rapidly dissected out and placed in cold Dul (4°C). Theblocks contained mainly sensory areas including parietal cortex areas 1,2,3 and7 and occipital cortex area 17 and 18 (Krieg, 1946) as well as some adheringwhite matter. Coronal tissue slices of approximately 400 pm thickness wereobtained with a tissue slicer (Bennett et al., 1983). The 25 slices that werenormally obtained from each rat cortex were separated and placed at random inthe wells of tissue culture plates, each well containing 0.5 mL of cold Dul+.Random placement of slices in the wells allowed for the control of variations inslice size.Evidence for Slice ViabilityThe brain slice preparation described above is essentially identical to the brainslice preparation used for electrophysiological studies by many laboratories(Reid et al., 1988; Teyler, 1980; Mcllwain et al., 1951). Cortical slices arethought to contain mostly living cells for the following reasons:(1) Trypan blue, a dye that cannot be taken up by cells with intact plasmamembranes (Tennant, 1964), labeled many cells at the cut edges of the slice,but relatively few towards the center of the slice. Quantification of thepercentage of dead cells was estimated to be approximately 15% at 0 h and30% after 6 h at 30°C (Van Huizen et el., 1989). It is worth noting that some23cells in the interior of the slice were labeled showing that the dye has access toall areas of the slice.(2) Electron microscopic examination revealed that cells at the edges of theslice appeared to be mostly dead cells, with the center of the slice containinghealthy tissue for up to 6 h at 30°C (Van Huizen et al., 1989). At the edges ofthe slice, a large number of swollen dendrites without their cytoplasmicorganelles were seen. However, dendrites in the center of the slice stillcontained cytoplasmic organelles, their membranes were intact, and axonterminals were making synapses on somata as well as on dendritic spines (VanHuizen et al., 1989).(3) Electrophysiological experiments showed that field action potentials couldbe obtained in layers 213 of visual cortex after stimulation of the optic radiation(Van Huizen et al., 1989).(4) Cells showed 2-deoxy-D-[14C]glucose uptake, a marker of cellular viability(Sokoloff et al., 1977), for up to 6 hours (see Chapter 2 and Van Huizen et al.,1989).Radioliciand MethodsA variety of techniques can be used to study neurotransmitter/receptorinteractions. Some of these are indirect, e.g. electrophysiological methods, inthat they do not allow the measurement of receptor density. Directmeasurements include fluorescent- and radioligand techniques. Of the latter,24radioligand binding is a well established and quantitative method for studyingreceptor characteristics, regulation, and distributions.RadioliQand[3H]-CNQX (specific activity 15.6-26.7 Cilmmol), a competitive AMPA receptorantagonist, was purchased from New England Nuclear (NEN). Although CNQXhas been shown to bind to both AMPA and kainate receptors (Young and Fagg,1991), CNQX has been shown to bind to AMPA receptors with approximatelyfive times greater affinity (Honore et al., 1988). In the present preparation,kainate did not compete for 5-10 nM[3H]-CNQX, the concentration rangeemployed in the experiments described below (see Chapter 1 and Lanius andShaw, 1992). The use of 5-10 nM concentrations of[3H]-CNQX shouldtherefore provide a relatively specific measure of AMPA receptor binding.The choice of a competitive antagonist (CNQX) instead of an agonist (AM PA) isoften preferable in binding experiments in intact cells. Usually, competitiveantagonists have slower dissociation rate constants than agonists. This factor iscritical for radioligand binding experiments in slices because the rinse-out timefor free radioligand must be shorter than the dissociation rate constant.[3H]-CNQX Binding AssaysImmediately prior to the experiment,[3H]-CNQX was diluted to the desiredconcentration (5-10 nM) in DuI. Five hundred pL of this solution were added tothe slices after the original DuI was removed. Incubation with[3H]-CNQX wasallowed to proceed for 3 h (see Chapter 1 Results). Three slices were normally25used to determine total (receptor and non-receptor) binding, while two sliceswere used for a determination of non-specific (non-receptor) binding. Nonspecific binding was determined by the addition of 5 pL of 10-2 M unlabeledCNQX as competitor to give a final bath concentration of M. Proteincontent in these slices was determined by a modification of the method of Lowryet al. (1951), and specific binding could thus be expressed as femtomolesbound per mg protein.Following the incubation step with[3H]-CNQX, a 20 pL sample of the buffer inthe well was removed to determine iree’ ligand concentration. The rest of thebuffer was removed with a Pasteur pipette and the slices were rinsed twice for 5mm with 0.5 mL Dul at 4°C (see Chapter 1 Results). After the final rinse, thebuffer was removed, the slices were picked up with a small circle of glassmicrofibre filter paper (Whatman GF/B), and were placed directly in scintillationcounting vials containing 4 mL of Formula 963 counting cocktail (NEN). Theamount of bound ligand was determined in a LS6000 IC Beckman scintillationcounter (efficiency for 3H, approximately 55%) after a minimum interval of 12 h.Statistical AnalysisStatistical analysis of binding was performed using a Student’s t-test (twotailed) or a one-way analysis of variance (ANOVA) (p<0.05).26CHAPTER 1CHARACTERIZATION OF A[3H]-CNQX BINDING SITEIntroductionIn order for a binding site to be identified as a receptor a number of criteria mustbe fulfilled. The binding site must show (1) saturability, (2) competition byappropriate pharmacological agents, (3) steady-state binding, as well as (4) theappropriate distribution (Boulton et al., 1985). A bare minimum requirement forbinding is that it be saturable. Saturation of specific binding should occur at, orbelow, concentrations of ligand comparable to those required to produce abiological effect (typically in the nanomolar or micromolar range) (Burt, 1980).Second, pharmacological specificity is of vital importance in recognizing abinding site as a receptor. The binding of a ligand should be displaceable bycompounds, including agonists and antagonists, which have been shown tomimic or inhibit the biological response of the ligand (Burt, 1980). Moreover,the ability of drugs to compete for the radioactively labeled ligand shouldcorrelate with the potencies of these drugs as either agonists or antagonists inbiological systems (Burt, 1980). Moreover, the binding site must show steadystate binding after a certain time period, depending on the receptor (Boulton etaI., 1985). Finally, the distribution of binding sites should be present in tissuespreviously demonstrated to possess the neurotransmitter of the binding site inquestion (Boulton et al., 1985), and the relative distribution of receptors shouldroughly correlate with the density of synaptic junctions utilizing thatneurotransmitter (Boulton et al., 1985).27In addition to the above discussed criteria, it is imperative to show that theactivation of a binding site is associated with a physiological response such asa change in membrane potential or stimulation of second messenger activity. Ifall the above criteria have been fulfilled, one may assume that the binding sitesin question represent actual neurotransm itter receptors.In the following chapter, characterization studies, including saturation binding,competition, and steady-state binding of a[3H]-CNQX binding site in ratneocortex will be described. These experiments will test the hypothesis that[3H]-CNQX is binding to an AMPA receptor population.Materials and MethodsBrain slices were prepared as described in the General Methods.All[3H]-CNQX characterization experiments were performed at 4°C to preventreceptor regulation which may confound the results (Van Huizen et al., 1989)and to decrease binding to potential uptake sites (Liron et al., 1988).Characterization experiments included association/ dissociation time courses,competition, as well as saturation binding experiments. Time courseexperiments allow the calculation of the association and dissociation rateconstants. The equilibrium dissociation constant (I<d) can be determined fromtime course experiments in which the association (k+i) and dissociation (k-i)rate constants are calculated. The Kd is the ratio k-i Ikj and should, withinexperimental error, be similar to the Kd determined from saturation bindingexperiments (Boulton et al., 1985). The dissociation rate constant, k-i, can becalculated by multiplying the slope of log [B] (specific binding) against time by e28(-2.303) (Bennet, 1978). The association rate constant, k+i, can be calculatedfrom the amount of the bound ligand at various time intervals (Bt) untilequilibrium (Beq) has been reached from a plot of In (Beq - Bt) versus time. Theslope of this line (kb) relates to both the k+i and k-i and free ligandconcentration (L) in the following way:kj = kbs - k-i[LICompetition studies used the following compounds at a concentration rangefrom M to 10-12 M (see Table 1 below for a description of thesesubstances): glutamate, AMPA, CNQX, NBQX, quisqualate, NMDA and kainate.These studies allow the calculation of lC values, the concentration ofunlabeled ligand required to inhibit 50% of the radioactive ligand binding(Boulton et al., 1985), and show if[3H]-CNQX is indeed binding to a specificAMPA receptor.29Compound Site of Actionglutamate natural agonistinteracts with AMPA, kainate, NMDA,and metabotropic glutamate receptorsAMPA potent agonist which specificallyinteracts with AMPA receptorsCNQX corn petitive AM PA/kainate antagonist(5 times higher affinity for AMPA thankainate receptors (Honore et al.,1988))weak inhibitor of NMDA receptorsbecause of its ability to compete withglycine for its binding siteNBQX corn petitive AM PA/kainate antagonist,little affinity for the NMDA glycine sitequisqualate AMPA/kainate and metabotropicreceptor agonistNMDA NMDA receptor agonistkainate kainate/AM PA receptor agonistTable 1: Pharmacological agents used to analyze[3H]-CNQX binding to AMPAreceptors in rat cortical slices (see Young and Fagg, 1991).For saturation binding experiments a concentration range of 2 to 909 nM [3H]-CNQX was used. From such experiments the dissociation equilibrium constant,30Kd, as well as the maximum number of binding sites, Bmax, can be calculated.Bmax and Kd values in this case were computed using nonlinear regressionanalyses (GraphPad PR I SM program me).In experiments designed to examine the distribution of surface versus possibleinternal AMPA receptors, some slices were slowly frozen to -30°C and thenthawed in order to disrupt the cellular membranes prior to attempting thebinding assays. A consideration of the structure of CNQX suggests that it ispolar and should, therefore, have restricted passage through intact plasmamembranes. The use of frozen/thawed versus intact cells labeled with [3H]-CNQX should thus allow experiments to distinguish the percentage of AMPAreceptors present in the plasma membrane and in intracellular compartments.In addition, to examine possible[3H]-CNQX binding to uptake sites controlslices incubated in Dul+ were compared to those incubated in 50mM Trisacetate, a Na+ free buffer. Since uptake sites are temperature as well assodium dependent (Kanai et al., 1993), assays performed at 4°C in Tris-acetateshould eliminate the binding to potential glutamate uptake sites.ResultsTo determine whether the[3HJ-CNQX binding site shows steady-state binding,time course experiments examining the association and dissociation of specific[3H]-CNQX binding were carried out (Figure 1). At 4°C,[3H]-CNQX bindingreached steady-state by 150 mm, remaining stable for up to 5 h. Dissociation ofspecific[3H]-CNQX binding was determined once steady-state binding hadbeen reached at 150 mm.[3H]-CNQX did not easily dissociate at 4°C, showing31approximately 40% decreases in binding after 40 mm rinse times. From theseexperiments a 3 h incubation time for[3H]-CNQX was chosen. Two 5 mmrinses were adequate to achieve maximum specific binding and reduce freeligand concentration.To evaluate the specificity of[3H]-CNQX binding, the ability of variouscompounds to compete for[3H]-CNQX binding was examined (Figures 2A & B).AMPA, CNQX and NBQX were the most effective compounds showing lC50values near io6 M and final non-specific binding levels of about 30%.Quisqualate was less effective having an lC5O value near 1 M and final nonspecific binding levels of approximately 58% (Figure 2A). In Fig. 2B,competition curves for the agonists glutamate, kainate and NMDA are shown.Neither glutamate, kainate, nor NMDA were effective in competing for[3H]-CNQX binding.To determine whether[3H]-CNQX binding is saturable, saturation bindingstudies using a concentration range of 2 to 909 nM[3H]-CNQX were carried out(Figure 3).[3H]-CNQX labeled binding sites in a saturable manner. Nonlinearregression analyses of saturation binding data indicated that[3H]-CNQX boundto a single population of receptor sites with a Kd=565 and a Bmaxl 1.4pmol/mg protein. Non-specific binding ranged from 36 % of total binding to 77% of total binding at 2 and 909 nM[3H]-CNQX, respectively.Figure 4 illustrates[3H]-CNQX binding for slices incubated under differentconditions. These conditions included slices in Dul, slices in Dul followingfreeze/thaw treatment, and slices incubated in 50 mM Tris-acetate buffer. Thesedata show that the level of binding was not significantly altered from controlvalues at any of the tested concentrations of ligand for either the freeze/thaw orTris-acetate treated slices (p>O.05, Student’s t-test) (see Discussion).3233120 -100-CCx_____az9Figure 1: Association and dissociation measurements of[3H]-CNQX binding inadult rat neocortical slices at 4°C. Ligand concentration was 5-10 nM. Specificbinding is plotted against time for increasing incubation times (.) and a fixedrinse time (2x5 mm) or for a fixed incubation time (3h) and variable rinse times(o). Rinse onset is indicated by the arrow and is shown in the inset time scale..20C)0EE‘C•1.4’/ 0.7kHi .1.10 20 4060..806040-20 -00• Association0 Dissociation111•150 100 150 200 250Time (mm)300 350 40034Figure 2: Competition of[3H]-CNQX binding wfth specific AMPA analogues. A:competition with the AMPA agonists AMPA and quisqualate as well as with theAMPA antagonists CNQX and NBQX. B: competition with glutamate, NMDA,and kainate. In all experiments, the ligand concentration was 5-10 nM. Allincubations with[3H]-CNQX lasted 3 h at 4°C. Rinses were 2x5 mm.350)CCCu0I00)C•0CCu0I0A•A•AA A• Glutamate0 NMDAA KainateI I IA0I-’• CNQXNBQX0 Quisqualate• AMPA0000..1201008060-40-20120 -10080-60-40-20-‘II. • I • I • I • I-10 -8 -6 -4 -2 0-Log Competitor Concentration (M)B.Ao// I-10 -8 -6 -4 -2- Log Competitor Concentration (M)36300000-0)::::::r:r20 250 500 750 1000Concentration (nM)Figure 3: Saturation binding isotherm for specific[3H]-CNQX binding. Theincubation time in varying concentrations of[3H]-CNQX (2-909 nM) was 3 h at4°C. Rinse time was 2x5 mm. Specific binding is expressed as dpm.3740002C)C 3000C><o 2000z9()1000C)‘4-C)00Figure 4: Effects of freeze/thaw treatment or incubation with 50 mM Tris-acetateon[3H]-CNQX binding. Ligand concentration was 5-10 nM. All incubationswith[3H]-CNQX lasted 3 h at 4°C. Rinses were 2 x 5 mm. Error bars giveS.E.M. (n=3).38DiscussionUsing in vitro living rat cortical slices, a[3H]-CNQX binding site wascharacterized. Time course experiments showed that 5-10 nM[3H]-CNQXreached steady-state after 150 minutes at 4°C and remained stable for up to 5hours. Nielsen et al. (1990) reported steady-state binding of 50 nM[3H]-CNQXin 20 pm sections of rat cerebellum after 15 minutes at 2°C. The discrepancybetween the time needed to reach steady-state binding may be explained bythe differences in the preparations and radioligand concentrations used, Theuse of 400 pm compared to 20 pm thick tissue may lead to a slower rate ofdiffusion of the radioligand into the slice, thus resulting in an apparent increasein the association rate constant. Moreover, 50 nM[3H]-CNQX labels AMPA aswell as kainate receptors (Nielsen et al., 1990). It is therefore possible thatthese combined sites possess a different association rate constant.In the present preparation, 5-10 nM[3H]-CNQX did not easily dissociate at 4°C,showing approximately 40% decreases in binding after 40 mm rinse times. Incontrast, Nielsen et aI. (1990) showed that 50 nM[3H]-CNQX binding in 20 pmthick sections of rat cerebellum was fully reversible within 30 minutes at 2°C.Although it remains unknown why[3H]-CNQX binding was not fully reversible inthe present preparation, differences in the preparation may be responsible forthe observed differences. Since[3H]-CNQX did not easily dissociate, thedissociation rate constant (k-i) and Kd could not be calculated from the timecourse experiments.Competition studies showed that AMPA, CNQX, NBQX, and quisqualate wereeffective as competitors for 5-10 nM[3H]-CNQX binding. Glutamate, kainate,39and NMDA showed no competition for 5-10 nM[3H]-CNQX binding. AMPA(AMPA agonist), CNQX (AMPA/kainate antagonist), and NBQX (AMPA/kainateantagonist) all had lC5O values near 10-6 M; quisqualate (AMPA/kainate andmetabotropic receptor agonist) was less effective, showing an IC5Q value ofi- M. NMDA was not able to compete for[3H]-CNQX binding. Nielsen et al.(1990) reported inhibition of 50 nM[3H]-CNQX binding in the molecular layer ofcerebellum by AMPA (K 0.015 pM), quisqualate (Kj 0.77 pM), L-glutamate (K5.5 pM), and kainate (K 20 pM), with no competition by NMDA. Althoughglutamate as the natural agonist of the AMPA receptor was moderately effectivein the competition of 50 nM[3H]-CNQX binding using in vitro autoradiography(Nielsen et al., 1990), it was ineffective in the present preparation. As discussedin the General Methods, one disadvantage of the °living” slice technique is thatnaturally occurring agonists, e.g., glutamate (Lanius and Shaw, 1993), GABA(Shaw and Scarth, 1991) and acetylcholine (Van Huizen et al., 1989), often failto show an effect in both competition and regulation experiments. This may, inpart, be due to the presence of functional uptake sites as well as degradativeenzymes which are still present in intact tissue (reviewed by Shaw andWilkinson, 1994). It is therefore difficult to examine the roles of putativeneurotransmitters in receptor regulation using the present slice preparation.Although kainate resulted in some competition of 50 nM[3H]-CNQX (Nielsen etal., 1990), it was ineffective in inhibiting 5-10 nM[3H]-CNQX. Since CNQX hasbeen reported to have a five times higher affinity for AMPA than for kainatereceptors (Honore et al., 1988), the use of 50 nM versus 5-10 nM[3H]-CNQXconcentrations could result in increased labeling of kainate receptors by [3H]-CNQX. This may explain the ability of kainate to compete for 50 nM[3H]-CNQXbinding but not for 5-10 nM[3H]-CNQX binding. Nevertheless, the lack of effect40of kainate in the present preparation suggests that at 5-10 nM,[3H]-CNQXlabels an AMPA receptor to which kainate does not bind.We have previously reported the presence of an apparent high- and low-affinity[3H]-CNQX binding site using ligand concentrations ranging from 1 to 75 nM.The high-affinity[3H]-CNQX binding site showed a Kd=1 1 nM and a Bmax=470fmol/mg protein and appeared to saturate between 20 and 35 nM[3H]-CNQX(Lanius and Shaw, 1992). The saturation binding isotherms for theseexperiments were fit by eye. Subsequent nonlinear regression analyses ofsaturation binding isotherms obtained by using ligand concentrations from 2 to909 nM[3H]-CNQX indicated that[3HJ-CNQX bound to a single population ofreceptor sites with a Kcj565 nM and a Bmaxl 1.4 pmol/mg protein. Theseresults are in agreement with previous studies (Maren et at, 1993; Honore etat, 1989; Nielsen et al., 1990) which have reported that[3H]-CNQX binds withequal affinity to two states (high and low affinity) of the AMPA receptor (Honoreet al., 1989; Nielsen et al., 1990). The Kd obtained from the present data ishigher than the Kd previously reported by Maren et al. (1993) (131 nM), Honoreet al. (1989) (39 nM), and Nielsen et al. (1990) (67 nM). However, thisdiscrepancy may be due to the fact that nonlinear regression analyses wereused to generate the present Kd value. The Kd values obtained by Maren et al.(1993), Honore et al. (1989), and Nielsen et al. (1990) were generated usingconventional Scatchard analyses (linear regression).The 6rnax value obtained in the present preparation was found to be higherthan that obtained previously in the molecular layer of cerebellum (11.4pmol/mg protein versus 3.54 pmol/mg protein) (Nielsen et al., 1990). Thisdisparity, however, is likely attributable to the fact that higher densities of AMPA41receptor binding have been observed in some areas of rat cortex compared tothe molecular layer of cerebellum (Nielsen et al., 1990; Olsen et al., 1987). Thepresent Bmax value was found to be similar to the Bmax value attained in thedentate gyrus (11.2 pmol/mg protein) (Maren et al., 1993).The experiments comparing slices incubated in Dul+ to those which werefreeze/thawed or incubated in Tris-acetate showed no significant differences in[3H]-CNQX binding among the different groups. However, standard errors inthese experiments were very big and are likely due to rather large variations inslice size which were unavoidable when the technique was first employed. Inthe first case, freeze/thawing ruptures cellular membranes, thereby exposingpotential internal AMPA receptors to[3H]-CNQX which, because of its polarity,does not cross intact cellular membranes. These results may be interpreted assupporting the idea that most functional AMPA receptors are on cell surfaces. Inthe second case, the use of a Tris-acetate buffer at 4°C was intended toexamine whether[3H]-CNQX was binding to potential uptake sites. Binding touptake sites has been reported to be Na (Kanai et al., 1993) as well astemperature-dependent (Zaczek et al., 1987).[3H]-CNQX binding carried out ina buffer which does not contain Na+ ions at 4°C should therefore eliminate anypotential binding to uptake sites. Since brain slices incubated in Tris-actetatedid not show decreased levels of[3H]-CNQX binding from slices incubated inDul, the present data support the view that most[3H]-CNQX binding observedis to receptors rather than to uptake sites. Moreover, binding to uptake sites isnon-saturable and cannot be displaced by non-radioactive ligand (Bylund,1992). The finding that[3H]-CNQX binding was saturable and could bedisplaced by non-radioactive competitor therefore further suggests that[3H]-CNQX is binding to receptors and not to uptake sites.42CHAPTER 2AMPA RECEPTOR REGULATION BY AGONIST AND DEPOLARIZING STIMULIIntroductionOne mechanism by which neuronal function can be regulated is by modifications inthe elements of synaptic transmission. Modifications of synaptic transmission can takeplace at the pre- or postsynaptic elements of the synapse. Presynaptically, suchregulation can occur through the regulation of neurotransmitter synthesis (Wurtmanand Fernstrom, 1976), release (Greengard et al., 1993), or uptake (Kanai et al., 1993).Postsynaptically, modification may occur by regulation of neurotransmitter inactivation(Bird and Aghajanian, 1975) or alterations in neurotransmitter receptor characteristics(Shaw et al., 1994). As neurotransmitter binding to target receptors represents the firstcrucial step in postsynaptic signal transduction, this chapter will focus on alterations inAMPA receptor binding properties following treatment with stimuli designed to mimicsome aspects of synaptic transmission. These stimuli are (1) application of the agonistAMPA to mimic neurotransmitter evoked receptor modification and (2) cellulardepolarization using veratridine intended to produce an independent means ofassessing current/voltage effects in the absence of direct receptor stimulation.Receptor regulation has often been observed in response to two key features ofsynaptic transmission, including agonist binding at receptors and neurotransmitterinduced conductance changes for specific ions. Much ongoing research hasemployed experiments designed to mimic some aspects of synapticneurotransm ission, using agonist and chemically-induced increases in cellulardepolarization. Various receptor populations, including the GABAA, kainate, and43muscarinic ACh (mACh) receptors have been shown to respond to agonist anddepolarizing stimuli by a change in the number of functional receptors.Neurotransmitter ReceDtor Response to Agonist Stimulationlonotropic ReceDtorsBoth GABAA and kainate receptor populations have been reported to show functionalmodifications in response to treatment with agonists (Lanius and Shaw, 1993; Mehtaand Ticku, 1992; Shaw and Scarth, 1991; Hablitz et al., 1989; Tehrani and Barnes,1988; Maloteaux et al., 1987). Hablitz et al. (1989) and Tehrani and Barnes (1988)demonstrated that treatment with 100 pM GABA for seven days in cultures of cerebralneurons prepared from chick embryos reduced[3HJ-flunitrazepam binding byapproximately 70% and led to a significant reduction in the GABA-gated C1 uptake.This reduction was due to a decrease in receptor number rather than affinity and couldbe abolished by the concomitant exposure to the GABA receptor antagonist R 5135,suggesting that GABAA receptor occupancy by agonist was required for suchregulation. Similar results were reported by Maloteaux et al. (1987) who found areversible decrease in the Bmax value of[3H]-flunitrazepam binding in rat culturedforebrain neurons after 48 hour incubations with 1 mM GABA or 0.1 mM muscimol(GABAA agonist). Forty-eight hour incubations with GABA also led to 31 % and 23%decreases in the Bmax values of[3H]-muscimol and[35S]-butylbicyclophosphorothionate (TBST) (GABAA receptor-associated ion channelantagonist), respectively. Since these decreases in the number of benzodiazepineand GABAA receptors were measured after homogenization, the GABA or muscimolinduced decreases of specific binding appeared to represent a true receptor lossrather than an internalization or sequestration of the receptors (Maloteaux et al., 1987).44Such a receptor loss could be explained by agonist-induced internalization of thereceptor followed by receptor degradation intracellularly.The above results are supported by another study which reported a significantdecrease in GABAA receptor number in cultured cortical neurons after treatment with500 pM GABA for five days (Mehta and Ticku, 1992). The level of[3H]-GABA bindingdeclined as a function of the duration of GABA exposure with a maximal decrease of40% occurring at 5 days of treatment. Twenty percent decreases were observed asearly as after 24 hours of treatment. Moreover, a 35% decrease in TBST binding wasseen after treatment with 500 pM GABA for five days.On a faster time scale, Shaw and Scarth (1991) reported a 12% decrease in GABAAreceptor number in adult rat cortical slices following 2 hour incubations with 10 pMmuscimol (GABAA agonist) as measured by[3H]-SR 95531 (GABAA antagonist)binding. Similar to the inhibitory GABAA receptors, the excitatory kainate receptorpopulation in adult rat cortical slices showed 26% decreases in receptor numberfollowing 2 hour incubations with 10 pM kainate as assessed by[3H]-kainate binding(Lanius and Shaw, 1993).G-rotein CouDled ReceptorsThe G-protein-coupled mACh receptors have also been reported to decrease innumber following treatment with agonist for several hours to days (Shaw et al., 1989;Klein et al., 1979; Siman and Klein, 1979). Activation of this receptor population with100 pM ACh for 12 hours on neuron-like NG1O8-15 hybrid cells reduced the numberof muscarinic ACh receptors by 88% as assessed by[3H]-quinuclidinyl benzilate(QNB) binding (Klein et al., 1979). This decrease in receptor number remained stable45for up to three days. Withdrawal of ACh resulted in a slow increase in mACh receptornumber that could be blocked by concomitant exposure to cycloheximide, a proteinsynthesis inhibitor. These findings suggested that receptor breakdown/down-regulation was the mechanism underlying the observed receptor regulation (Klein etal., 1979).On a faster time scale, Siman and Klein (1979) reported a 35% decrease in[3H]-QNBbinding in cultured embryonic chicken cerebrum cultures following 90 minuteexposures with 1 mM carabachol, a muscarinic receptor agonist. This decrease inbinding further decreased to 60% after 9 hours and remained stable for 4 days. Thedecreases in[3H]-QNB binding observed in response to treatment with carbacholcould be partly inhibited by cytochalasin B, a microfilament disrupter. Sincecytochalasin B is known to disrupt filaments thereby blocking endocytosis, thedecreases in[3HJ-QNB binding appeared to partially result from internalization of thereceptors (Siman and Klein, 1979). In adult rat cortical slices, 2 hour exposures to 10pM carbachol led to 26% and 36% decreases in surface and total mACh receptornumber as determined by[H]-N-methyl scopolamine (NMS) and[3H]-QNB binding,respectively (Shaw et al., 1989). Such regulation was highly dependent ontemperature with no regulation occurring at temperatures below 30°C.It was recently reported that agonist-induced receptor regulation of human m3 mAChcould only occur in the presence of carboxyl-terminal threonine residues (Yang et al.,1994). Exposure of the human m3 mACh receptor expressed in Chinese hamsterovary cells to 1 mM carabachol for 4 hours resulted in a 20% decrease in [3H]-scopolamine binding. This carbachol-induced decrease in[3H]-scopolamine bindingcould be abolished by site-directed mutagenesis of the threonine residues (Thr550’553, 554) present in the receptor cytoplasmic carboxyl terminal. Since these46threonine sites represent possible phosphorylation sites for PKC, it is possible thatreceptor phosphorylation may play a role in mACh receptor regulation.Neurotransmitter Receptor Response to Cellular DepolarizationIonotroic ReceptorsBoth kainate and GABAA receptor populations have been reported to show functionalmodifications in response to treatment with veratridine, an agent which leads todepolarization by blocking sodium channel inactivation (Catterall, 1980). For thekainate receptor population the response to veratridine was similar to that observed inresponse to agonist stimulation (Lanius and Shaw, 1993). In adult rat cortical slices,treatment with 10 pM veratridine for up to two hours led to 55% decreases in kainatereceptor number as assessed by[3H]-kainate binding. In contrast, the inhibitoryGABAA receptor population in rat cortical slices responded to veratridine treatment fortwo hours by showing a 58% increase in receptor number as measured by[3H]-SR5531 binding (Shaw and Scarth, 1991). These results are supported by Tabuteau etal. (1993) who reported increases in[3H]-SR 95531 binding in response to 3 hourtreatments with a variety of depolarizing agents in rat hippocam pal slices.Depolarization induced by treatment with K+, veratridine, v+g, and ouabain led to 72,113, 127, and 215% increases in[3H]-SR 95531 binding, respectively.G-protein Coupled ReceptorsLiles and Nathanson (1987) reported that incubation with 50 pM veratridine for 24hours induced a 200% increase in muscarinic receptor number in neuroblastoma cellsas determined by[3H]-QNB binding. Increases in mACh number seen as a result of47this treatment could be blocked by tetrodotoxin (TTX) and could be returned to controllevels within 20 hours upon withdrawal of veratridine. Chronic membranedepolarization induced by incubation in a medium containing 60 mM potassiumchloride led to a TTX-insensitive 50% increase in mACh receptor number after 24hours. These results are supported by a recent study which demonstrated an increasein BODIPY® FL (muscarinic Mi receptor-selective antagonist) binding following 7 daytreatments with 40 mM potassium chloride in cultured rat visual cortex neurons (Wanget al., 1994). BODIPY® FL binding showed 58% and 40% increases in dendriticprocesses and cell bodies, respectively.On a faster time scale, Luqmani et al. (1979) reported a 17% decrease in[3H]-N-methylatropine (NMA) (mACh antagonist) binding in response to 30 minutestimulations with 76 pM veratridine in synaptosomes prepared from rat cerebral cortexthrough an unknown mechanism. The decrease in[3H]-NMA was not a result ofneurotransmitter-receptor interactions as the phenomenon was not achieved bytreatment of synaptosomes with the agonist carbachol (0.1-100 pM). Furthermore, Kdepolarization did not decrease[3H]-NMA binding, suggesting that the decrease in[3H]-NMA binding observed was specifically associated with the opening of Nachannels. Similarly, treatment of rat cortical slices with 10 pM veratridine for 2 hoursresulted in 26% and 11% decreases in[3H]-NMS and[3H]-QNB binding, respectively(Shaw et al., 1989).In summary, the above studies have shown that both ionotropic and G-protein coupledreceptors can be regulated in response to agonist- and pharmacologically-induceddepolarization. The consequences of such regulation, however, can be diversedepending on the specific receptor or stimulus involved.48Materials and MethodsA Brief Overview of MethodsBrain slices were prepared as described previously in the General Methods. Thecortical slices were exposed to agonist (AMPA) or veratridine + glutamate (v+g)treatment at 37°C. Following incubation with these substances, the slices were rinsedand incubated with radioligand at 4°C as described previously.Experiments Designed to Study the Effects ofAgonist and DeDolarizing Stimuli on fH1-CNQX BindingIn order to study regulation of the AMPA receptor population by agonist or depolarizingstimuli, AMPA and v+g were used.Compound Concentration (M)AMPA i05veratridine 1glutamateTable 2: Compounds used to study AMPA receptor regulationAMPA was used to achieve direct activation of the AMPA receptor. In areas of thebrain where the GIuR2 subunit is not present, activation of AMPA receptors leads to aninflux of both Na+ and Ca2+. However, in brain regions which express GIuR2,activation of AMPA receptors leads to an influx of Na currents through AMPAreceptor-associated ion channels, only (Hollmann et al., 1991). Since GIuR2 is highly49expressed in cortex (Martin et at., 1993; Petralia and Wenthold, 1992), activation ofAMPA receptors in cortex only results in an influx of Na through AMPA receptor-associated ion channels. However, membrane depolarization as a result of AMPAreceptor-mediated Na+ influx has been shown to lead to the activation of voltagegated Ca2 channels with a concomitant influx of Ca2 (Church et al., 1994).A combination of veratridine and glutamate (v+g) was utilized to produce cellulardepolarization. Glutamate increases inward sodium currents via AM PA, kainate, orNMDA receptors, resulting in membrane depolarization and the opening of voltagegated Na and Ca2 channels; veratridine acts to keep the voltage-gated Nachannels open by blocking their inactivation (Catterall, 1980).Concentrations of M were chosen for AMPA, veratridine, and glutamate. Dose-response curves had shown these concentrations to be sufficient to obtain maximumeffects over the time period examined (see Results). Incubation with both compoundstook place for a minimum of 25 mm. Time course experiments had shown theseincubation times to be most effective to obtain maximum effects (see Results). Allregulation experiments were carried out at 37°C, since it has previously been shownthat maximum regulatory effects are obtained at physiological temperatures (VanHuizen et al., 1989).Control ExDeriments to Distinguish between Corn jetition andRegulation Effects of Veratridine and AMPAFor experiments designed to measure receptor regulation by agonists or veratridine, akey control is to establish the length of rinse necessary to remove all of the agonist orveratridine present. Veratridine has been shown previously to compete for muscarinic50ACh (Van Huizen et al., 1988; Shaw et al., 1989), GABAA (Shaw and Scarth, 1991),and AMPA receptor (Lanius and Shaw, 1992) binding by unknown mechanisms. Inaddition, AMPA has been shown to be able to compete for[3H]-CNQX binding (seeChapter 1). Failure to determine this rinse time can lead to confusion betweenagonist/veratridine-induced regulation and residual agonist/veratridine corn petitionwith the radioligand. Further, all stages in such determination must be made at lowtemperatures (4°C) to prevent regulation during the application of the agonist orveratridine and to further prevent any possible ‘reregulation’ during the rinse outphase (Van Huizen et al., 1989). In order to distinguish between possible competitionand regulation effects, post-AM PA or veratridine rinse times required to return bindingto control levels were measured at 4°C. Rinse times which were adequate to eliminateany competitive effects of these compounds were subsequently used.To determine whether the changes in[3H]-CNQX binding observed as a result ofAMPA or v+g treatment were reversible, adult rat cortical slices were incubated withi5 M AMPA or M v-i-g for a minimum of 25 mm at 37°C. The slices were thenrinsed twice for 30 mm with cold Dul after which fresh Dul was added for 5 to 45 mmat 37°C. Radioligand binding was then carried out as described previously.Experiments designed to measure possible internalization of AMPA receptors afterregulation with AMPA or v÷g compared[3H]-CNQX binding in slices which werefreeze/thawed to ones which remained intact after regulation with AMPA or v+g. Since[3H]-CNQX only labels surface receptors in an intact slice preparation (see Chapter 1),the comparison of[3H]-CNQX binding in intact to freeze/thawed cortical slices shoulddemonstrate whether AMPA and v-t-g result in the internalization or sequestration ofthe receptors. If the receptors were indeed internalized as a result of AMPA or v+gtreatment,[3H]-CNQX binding in the freeze/thawed slices should not be different fromcontrol. For this experiment, cortical slices were incubated with i05 M AMPA or i0551M v+g for 25 mm at 37 DC. Half of the slices were then rinsed twice for 30 mm with coldDuI, frozen to -30°C, and rapidly thawed. The other half was left intact for theradioligand binding step. Radioligand binding was then carried out as describedpreviously.Neurotoxicity of AMPA and GlutamateAlthough it has long been known that the neurotransmitter glutamate and itsanalogues kainic acid, NMDA, and AMPA can act as neurotoxins (Brorson et al., 1994;Koh et al., 1990; Siman and Card, 1988; Frandsen and Schousboe, 1987),controversy in the literature exists about which conditions are necessary and/orsufficient to produce neuronal damage or cell death. Frandsen and Schousboe(1987) have reported neurotoxic effects of AMPA (ED5O 10 pM) in cultured cerebralcortex neurons after exposure times of 1-10 mm as assessed by the release of thecytoplasmic enzyme lactate dehydrogenase. Moreover, Brorson et al. (1994) showedthat 20 mm exposures of cerebellar cultures to AMPA (30 pM) resulted in 25% celldeath compared to control cultures as assessed by a fluorescent cytotoxicity assay. Incontrast, Koh et al. (1990) suggested that AMPA concentrations as high as 1mM led toonly very slow neuronal degeneration in cultured embryonic cerebral cortex neurons.Continuous exposures for up to 24 h were required to elicit substantial cell loss.Glucose and magnesium have been shown to act as strong neuroprotective agentsagainst excitotoxic compounds (Cox et al., 1989; Lysko et al., 1989; Finkbeiner andStevens, 1988; Hahn et al., 1988). Cox et al. (1989) reported that cerebellar neuronsin primary culture were resistant to a 30 mm exposure to glutamate concentrations ashigh as 5 mM in the presence of 5 x i05 M glucose and 1 x i03 M magnesium.Lysko et al. (1989) suggested that cultured cerebellar granule cells were resistant to5240 mm exposures of glutamate at concentrations as high as 5 mM in the presence ofM glucose and 5.6 x M MgCI2. Moreover, Finkbeiner and Stevens (1988)found 10 pM concentrations of glutamate to be non-toxic in the presence of at least 1.8x M MgCI2 in cultured hippocampal neurons of the CAl region. Finally, Hahn etal. (1987) reported that 16 h exposures of 0.5 xi03- 5 x i3 M glutamate in thepresence of 1.8 x M Ca2 and 8 x M MgCI2 did not result in glutamateinduced cell death of rat retinal ganglion cells. Dul, the medium used in the presentexperiments, contained 4.7 x 10-2 M MgCI2 as well as 5.6 x M glucose. As theconcentration of both substances was higher than that used by the studies citedabove, the treatments employed in the present experiments should not have resultedin extensive neural death.2-Deoxy-D-[i4ClGlucose UDtake StudiesIn order to determine whether treatments with AMPA or glutamate resulted in celldeath in the present preparation, viability studies examining the uptake of 2-deoxy-D-[14C]glucose (2-DG) were perlormed (Sokoloff et al., 1977). 2-DG is an analogue ofglucose which differs from glucose in the replacement of the hydroxyl group on thesecond carbon atom by a hydrogen atom. 2-DG competes with glucose for theglucose carrier that transports glucose into the tissue. Once 2-DG has entered thetissue, its metabolism is identical to that of glucose until a point in the glycolyticpathway is reached where its irregular structure prevents its further metabolism(Sokoloff et al., 1977).Cortical slices were incubated with AMPA (10-5 M) or glutamate (10-5 M) for 25 mm to2h at 37°C. After this incubation, slices were rinsed twice for 10 mm with Dulbecco’sphosphate buffered saline. Five hundred pL of Dulbecco’s phosphate buffered saline53containing 5 pCi/mL[14C]deoxyglucose-6-phosphate (57.2 mCi/mmol) were thenadded to each cortical slice (Keler and Smith, 1989), and the incubation was allowedto proceed for 30 mm at 37°C. The cortical slices were then washed twice for 5 mmwith Dulbecco’s phosphate buffered saline and counted in a LS 6000 IC Beckmanscintillation counter.ResultsRegulation ExDerimentsTo determine the most effective concentration of AMPA on[3H]-CNQX binding, aconcentration-response curve for AMPA over a concentration range of M to io-4M was carried out (Figure 5). Concentrations ranging from 1 M to 1 M AMPAhad little effect on[3H]-CNQX binding. Concentrations ranging from 10-6 M to i0 MAMPA resulted in 16% to 23% decreases in[3H]-CNQX binding. Maximum effectswere observed with 1 M and 1 cr4 M AMPA. A concentration of 1 5 M wastherefore employed in regulation experiments.To evaluate the most effective concentration of v+g on[3H]-CNQX binding, aconcentration-response curve for v+g over a concentration range of 1 M to 1 0 M(for both compounds) was carried out (Figure 6). Maximum effects were observed withi- M and i04 M v-i-g (16% and 18% decreases in[3H1-CNQX binding,respectively). A concentration of 1 M was therefore employed in regulationexperiments.Time course experiments examining specific[3H1-CNQX binding as a function ofincreasing incubation times with i05 M AMPA at 37°C were conducted to establish54the incubation time most effective in inducing AMPA receptor regulation by AMPA(Figure 7).[3H1-CNQX binding was decreased by 28% following a 15 mm incubationtime; this effect remained relatively stable for up to 2 h. A minimum incubation time of15 mm was therefore employed.Time course experiments examining specific[3H]-CNQX binding following thewithdrawal of AMPA were carried out to determine whether the decrease in [3H]-CNQX binding observed as a result of treatment with i05 M AMPA was reversible(Figure 8). The decrease in[3H}-CNQX binding observed as a result of AMPAtreatment was fully reversible upon withdrawal of AM PA. Following washout of AM PA,[3H]-CNQX binding gradually increased, reaching control values by approximately 20mm.Time course experiments examining specific[3H]-CNQX binding as a function ofincreasing incubation times with 10 ‘ M v+g at 37°C were carried out to establish theincubation time most effective in inducing AMPA receptor regulation by v÷g (Figure 9).[3H]-CNQX binding was decreased by 18% following a 25 mm incubation time; thiseffect remained relatively stable for up to 2 h. A minimum incubation time of 25 mmwas therefore employed.Time course experiments examining specific[3H]-CNQX binding following thewithdrawal of v+g were conducted to determine whether the decrease in[3H]-CNQXbinding observed as a result of treatment with i05 M v+g was reversible (Figure 10).The decrease in[3H]-CNQX binding observed as a result of v÷g treatment was fullyreversible upon withdrawal of v+g. Following washout of v+g,[3H]-CNQX bindinggradually increased, reaching control values by approximately 20 mm.55Experiments designed to distinguish between possible competition and regulationeffects of AMPA and v-i-g showed that by 60 mm (2 x 30 mm rinses) at 4°C,[3H]-CNQXbinding had returned to levels not significantly different from control under either ofthese conditions (Figure 11). In subsequent regulation experiments, a rinse time ofthis duration was therefore routinely employed.Figure 12 shows the effects of iO-S M AMPA or v+g on[3H]-.CNQX binding. BothAMPA and v+g significantly decreased binding by an average of 25% and 20%,respectively (p<0.05, Student’s t-test). Similar decreases in binding were observedwhen the cortical slices were freeze/thawed after regulation with AMPA or v-i-g,suggesting that the mechanism underlying the regulation of these substances was notoccurring through internalization of the receptors.2-Deoxv-D-[i4ClGlucose Urtake StudiesResults of 2-deoxy-D-[l4C]glucose uptake studies showed that 2-deoxyglucoseuptake levels in control slices remained stable for 2 h and decreased by approximately25% after 6 h. Slices which were frozen to -30°C and then rapidly thawed, a processwhich kills cells by rupturing their cellular membranes, showed significantly decreased2-deoxyglucose uptake levels as compared to control (p<0.05, Student’s t-test).Exposure of cortical slices to 10 pM AMPA or glutamate for 25 mm or 2h did not resultin statistically significant different 2-deoxyglucose uptake levels from control slices(p.<0.05, Student’s t-test) (see Figure 13). One may have expected an increase in 2-deoxyglucose uptake levels after increases in synaptic activity due to stimulation withAMPA or glutamate. However, the slices were exposed to 2-deoxyglucose after theywere rinsed for 2 x 10 mm following exposure to AMPA or glutamate. It is thereforelikely that synaptic activity returned to control levels by this time.563500- T I0) iT DC3250-02 3000-oC0.0 2750- T T2500--10-8-6 -4-2 0Log Concentration (M)Figure 5: Concentration response of AMPA. Specific[3H]-CNQX binding is plotted asa function of increasing AMPA concentrations. Ligand concentration was 5-10 nM. Allincubations with[3H]-CNQX were for 3 h at 4°C. Rinse times were 2x5 mm. Error barsgive S.E.M. (n=3). Specific binding is expressed as dpm.57,_wT[10) 3500- jCCVC IC3250 ToC)0.Ci) 30002750-Log Concentration (M)Figure 6: Concentration response of v÷g. Specific[3H]-CNQX binding is plotted as afunction of increasing v+g concentrations. Assays were carried out as described inFigure 5. Error bars give S.E.M. (n=3). Specific binding is expressed as dpm.580--5 -+o -10..15.0W0-20-C.)WOa iW’-25DI-’ I-300’45. • • • • • •0 20 40 60 80 100 120 140Time (mm)Figure 7: Time course of AMPA-induceci regulation of[3H]-CNQX binding. Ligandconcentration was 5-10 nM. All incubations with[3H]-CNQX were for 3 h at 4°C.Rinse times were 2x5 mm. Error bars give S.E.M. (n=3).592800T2600 C0) T jC240O IT2 2200o TIC) C0. 1.(I)18000 10 20 30 40 ContTime (mm)Figure 8: Reversibility of AMPA-induced regulation of[3H]-CNQX binding followingwithdrawal of AM PA. Cortical slices were incubated with i05 M AMPA for 25 mm at37°C. The slices were then rinsed twice for 30 mm with cold Dul after which freshDul was added for 5 to 45 mm at 37°C. All incubations with[3H]-CNQX were for 3 hat 4°C. Rinse times were 2x5 mm. Error bars give S.E.M. (n=3). Specific binding isexpressed as dpm.600---5-L.C)-10o-15><00-20-25--30 • • • • i • •0 20 40 60 80 100 120 140Time (mm)Figure 9: Time course of v-i-g-induced regulation of[3H]-CNQX binding. Ligandconcentration was 5-10 nM. All incubations with[3H]-CNQX were for 3 h at 4°C.Rinse times were 2x5 mm. Error bars give S.E.M. (n=3).61UwTCC6OOOC J_TDCth soo4!500045000 10 20 30 40 ContTime (mm)Figure 10: Reversibility of v-i-g-induced regulation of[3H]-CNQX binding. Corticalslices were incubated with 1 cr5 M v+g for 25 mm at 37°C. The slices were then rinsedtwice for 30 mm with cold Dul after which fresh Dul was added for 5 to 45 mm at37°C. All incubations with[3H]-CNQX were for 3 h at 4°C. Rinse times were 2x5 mm.Error bars give S.E.M. (n=3). Specific binding is expressed as dpm.621206060OZ- Ic)0C.) 200100Figure 11: Rinse out times for AMPA or v+g treatments. Control cortical slices werecompared to slices treated with 105M v+g or AMPA for a minimum of 25 mm at 4°C,followed by two 4°C rinses of equal length (10 mm to 80 mm total time). Following therinses, the slices were incubated with 5-10 nM[3H]-CNQX for 3 h at 4°C. Slices wererinsed for 2 x 5 mm. Error bars give S.E.M. (n=3).0 20 40 60 80Time (mm)6350000) 4000t•0• 3000C.)t 2oooC.)00.ioooFigure 12: AMPA receptor regulation in response to AMPA or v-i-g. Rat cortical sliceswere incubated with AMPA (10-5 M) or a combination of v-i-g (10-5 M) for a minimum of25 mm at 37°C. Following this incubation, slices were rinsed for 60 mm (2 x 30 mm).Slices were then incubated at 4°C with 5-10 nM[3H]-CNQX for 3 h and rinsed for 2 x 5mm. Error bars give S.E.M. (n=3). Asterisks indicate significant differences fromcontrol binding to the p<0.05 level (Students t-test). Specific binding is expressed asdpm.Control QAMPA ElAMPA freezevgIJvg freeze •643000200011000Figure 13:14C-2-deoxyglucose uptake of cortical slices after treatment with AMPA orglutamate. Cortical slices were incubated with AMPA (10-5 M) or glutamate (10-5 M)for 25 mm or 2h at 37°C. Slices were then rinsed with Dulbecco’s phosphate bufferedsaline for 2 x 30 mm. Five hundred pL of Dulbecco’s phosphate buffered salinecontaining 5 pCiImL[14C]deoxyglucose-6-phosphate were then added to eachcortical slice, and the incubation was allowed to proceed for 30 mm at 37°C. Theslices were then washed twice for 5 mm with Dulbecco’s phosphate buffered saline.Error bars give S.E.M. (n=5).0mm 0 25 60 120360 25 120 25 120 Freeze/ThawControl AMPA Glutamate65DiscussionThe preceding regulation experiments have demonstrated reversible decreases in[3H]-CNQX binding in response to M AMPA treatment. These effects were seenafter 15 minutes of AMPA treatment and remained stable for up to two hours.The decreases in binding observed in response to a short duration (minutes to hours)of agonist stimulation were similar in direction to those reported previously for otherreceptor populations for their own agonists in this and other preparations. Agonisttreatment of the lonotropic GABAA and kainate receptors for 2 hours led to 12% and26% decreases in GABAA (Shaw and Scarth, 1991) and kainate (Lanius and Shaw,1993) receptor number, respectively. Similarly, agonist treatment of the G-proteincoupled mACh receptors for 1.5 and 4 hours has been reported to result in 20% (Yanget al., 1994) and 36% (Siman and Klein, 1979) decreases in mACh receptor number,respectively. Although the mechanism underlying such regulation has remainedlargely unknown, evidence has suggested that both receptor phosphorylation (Yang etal., 1994) as well as receptor internalization (Siman and Klein, 1979) may be partiallyresponsible (see Introduction).Agonist treatment for many hours to days resulted in decreases in receptor bindingwhich were, on average, greater in magnitude than the ones observed after a shorterduration of agonist treatment. Agonist treatment of GABAA receptors for 5 and 7 daysresulted in 40% (Mehta and Ticku, 1992) and 70% (Hablitz et al., 1989; Tehrani andBarnes, 1988) decreases in GABAA receptor number, respectively. Agonist treatmentof mACh receptors for 12 hours resulted in 88% decreases in mACh receptor number(Klein et al., 1979).66Treatment with v-i-g (10-5 M) led to reversible 17% decreases in[3H]-CNQX binding ina similar manner to the effects of AMPA. These effects were seen after 25 minutes andremained stable for up to two hours. The decreases in binding observed in responseto short durations (minutes to hours) of veratridine treatment were akin to thosereported for kainate- (Lanius arid Shaw, 1993) and muscarinic ACh receptors (Shawet al., 1989; Luqmani et al., 1979) in the same and other preparations. Exposures ofrat cortical slices to 10 pM veratridine led to a 55% decrease in kainate receptornumber (Lanius and Shaw, 1993). Similarly, 30 minute and 2 hour exposures toveratridine led to 17% (Shaw et al., 1989) and 26% (Luqmani et al., 1979) decreasesin mACh receptor binding, respectively. In contrast, the inhibitory GABAA receptorpopulation in rat cortical slices showed a 58% increase in receptor number inresponse to 10 pM veratridine treatment for 2 hours (Shaw and Scarth, 1991).Similarly, treatment of rat hippocampal slices with 10 pM veratridine for 3 hoursresulted in 127% increases in GABAA receptor number as assessed by[3H]SR95531 binding (Tabuteau et al., 1993).It is conceivable that the direction of regulation to, increases in cellular depolarizationmay be linked to the type of activity induced by agonist stimulation. GABAA agonistsincrease chloride currents to hyperpolarize neurons, an opposite effect to that inducedby v÷g. For AMPA and kainate receptors, both agonists and veratridine and glutamateact to increase sodium currents, and both treatments led to a decrease in receptorbinding.We have previously reported that quisqualate- and veratridine-induced changes in[3H]-CNQX binding are likely due to changes in high-affinity AMPA receptor number(Lanius and Shaw, 1992; Shaw and Lanius, 1992). However, this interpretation ofresults has to be reevaluated.[3H]-CNQX appears to bind to both the high- and low-67affinity component of the AMPA receptor (Honore et al., 1989). It therefore remainsunknown whether AMPA receptor regulation induced by AMPA and veratridine is dueto the regulation of the high- and/or low-affinity component of[3H]-CNQX binding.Honore et al. (1989) reported that at a concentration of 2 nM, the high-affinity sitecorresponded to 28% of[3H]-CNQX binding and the low affinity site corresponded to72% of the binding. Moreover, it remains unknown whether a change in receptornumber or affinity is responsible for the observed changes in[3H]-CNQX binding.However, several possibilities seem plausible. (1) It is possible that the decreases in[3H]-CNQX binding observed after treatment with AMPA and veratridine reflect a shiftfrom the high-affinity state to the low-affinity state of the AMPA receptor. (2) Analternative possibility is that AMPA and veratridine lead to a change in receptornumber of the high- and/or low-affinity state of the AMPA receptor. (3) A thirdpossibility is that AMPA and veratridine lead to a change in receptor affinity of the high-and/or low affinity state of the AMPA receptor. (4) Lastly, it is conceivable that AMPAand veratridine result in changes in receptor number J2 affinity of the high- and/orlow-affinity state of the AMPA receptor. Since agonist stimulation and increases incellular depolarization have often been reported to lead to changes in receptornumber (see Chapter 2 Introduction), however, it is likely that the changes in [3H]-CNQX binding seen in response to treatment with AMPA and veratridine are due tochanges in the number of high- and/or low-affinity AMPA receptors.Is it possible that the observed AMPA receptor regulation is an artifact of potentialcompetitive effects of AMPA or v+g for[3H]-CNQX binding or, alternatively, could it bean artifact of cell death due to possible neurotoxic actions of these agents?Competitive effects of AMPA or v+g can be excluded, since experiments designed todistinguish between possible competition and regulation effects showed that by 60minutes at 4°C,[3H1-CNQX binding had returned to levels not significantly different68from control. Moreover, it is unlikely that the decreases in[3H1-CNQX binding inresponse to AMPA or v+g treatment were the result of cell death. First, viability studiesemploying 2-deoxy-D-[14C]glucose uptake showed that exposure of cortical slices to10 pM AMPA or glutamate for up to two hours did not result in significantly different 2-deoxyglucose uptake levels from control slices. Second, the regulatory effects ofAMPA and v-t-g were completely reversible within approximately 20 minutes andtherefore are unlikely to be an artifact of cell death.Although the mechanism underlying the decreases in AMPA receptor number as aresult of agonist and depolarizing treatment is not clear at present, several possibilitiescan be considered: (1) receptor sequestration! internalization, (2) receptor down-regulation, (3) repression of receptor gene expression with concomitant receptordown-regulation, and/or (4) a change in the phosphorylation state of the receptorwhich has been previously shown to alter receptor function, includingagonist/antagonist binding, for a variety of ionotropic receptors (Kitamura et al., 1993;Krieger et al., 1993; Shaw et al., 1992; Vaello et al., 1992). Receptorsequestrationlinternalization appears unlikely since there was no difference in [3H]-CNQX binding between intact and freeze/thawed slices. Receptor-down regulationalso does not seem to occur, since the decrease in[3H]-CNQX binding was fullyreversible within approximately 20 minutes, and de novo receptor synthesis is unlikelyto occur in such a short period (Collins et al., 1991; Campbell et al., 1991).Furthermore, repression of receptor gene expression with concomitant receptor downregulation is improbable to account for the decrease in[3H]-CNQX binding, becausethe decrease in binding was fully reversible within 20 minutes. The possibility thatchanges in the phosphorylation state of the AMPA receptor are responsible for thedecrease in[3H]-CNQX binding observed seems more likely. Phosphorylating agentshave been shown to alter agonist/antagonist binding for a variety of receptor69populations, including the GABAA (Shaw et al., 1992), glycine (Ruiz-Gomez et al.,1991), and NMDA (Kitamura et al., 1993; Krieger et al., 1993) receptors. The role ofphosphorylation reactions in AMPA receptor regulation as assessed by[3H]-CNQXbinding will therefore be examined in the chapters to follow.An additional issue which needs to be addressed is whether the regulation induced byAMPA or v+g occurs on neurons and/or glial cells. Studies examining theim m unocytochemical localization and expression of AM PA-selective glutamatereceptors in cortex have led to contradictory results (Patneau et al., 1994; Martin et al.,1993; Petralia and Wenthold, 1992). Patneau et al. have reported rapidlydesensitizing responses to AMPA in rat cortical oligodendrocytes. Moreover, mRNAfor GluR 2-4 appeared to be expressed in rat cortical oligodendrocytes (Patneau et al.,1994). In contrast, Petralia and Wenthold have suggested only limited evidence forGIuR 1-4 immunoreactivity in cortical glial cells, although cortical astrocytes haverecently been shown to exhibit GluR 4 immunoreactivity (Martin et al., 1993).However, since native AMPA receptors in rat brain usually exist as hetero-oligomerscomposed of two or more of the four GluR subunits (Wenthold et al., 1992), it remainsquestionable whether GIuR 4 subunits alone are able to form functional AMPAreceptors. Moreover, studies using cultured rat cortical astrocytes have found [3H]-CNQX binding not to be present on these cells (Lanius et al., unpublishedobservations). It is, therefore, unlikely that the regulation observed in response toAMPA or v+g treatment is a result of astrocytic AMPA receptor regulation in the presentpreparation. Nevertheless, future experiments could further resolve this issue bystudying AMPA receptor regulation to AMPA or v+g in pure neuronal cell cultures.In summary, data have shown that the AMPA receptor population in adult rat cortexcan be regulated by treatment with agonists or v+g, agents that mimic aspects ofsynaptic neurotransmission in vivo. The regulation of these and other receptors (Yang70et al., 1994; Wang et aL, 1994; Lanius and Shaw, 1993; Tabuteau et al., 1993 Mehtaand Ticku, 1992; Shaw and Scarth, 1991; Hablitz et aL, 1989; Shaw et al., 1989;Tehrani and Barnes, 1988; Maloteaux et al., 1987; Liles and Nathanson, 1987; Klein etal., 1979; Luqmani et al., 1979) in neural tissue lend support to the view that receptorregulation may play a key role in the control of neural function. The regulation ofreceptor binding following agonist or depolarizing stimuli may serve to alter theresponse to additional neurotransmitter release. This may not only constitute apossible homeostatic mechanism in the control of interneural communication, but mayalso provide an essential means for controlling future input activity.71CHAPTER 3AMPA RECEPTOR REGULATION BY PHOSPHORYLATIONIntroductionThe previous chapter has shown that stimulation of rat cortical slices by AMPAor v+g decreased[3H]-CNQX binding. However, the mechanism by which suchregulation occurs remains unknown. One possibility is that depolarization inresponse to AMPA or v-i-g leads to protein phosphorylation and perhaps specificreceptor phosphorylation. In support of this view, it has been shown thatdepolarization induced by veratridine or high external K stimulated theincorporation of 32P into two specific proteins in a rat synaptosomal preparation(Sieghart et aL, 1979; Forn and Greengard, 1978). More recent studies havedocumented the role of phosphorylation reactions in the regulation of manylonotropic receptor properties (for review see Swope et al., 1992; Huganir andGreengard, 1991). Phosphorylating agents have also been shown to alteragonist/antagonist binding for a variety of ionotropic receptor populations,including the GABAA (Shaw et al., 1992), glycine (Vaello et al., 1992), andNMDA (Kitamura et at., 1993; Krieger et at., 1993) receptors.Experiments designed to examine the effects of multifunctional protein kinaseson[3H]-CNQX binding will be the focus of this chapter. These experiments willtest the hypothesis that phosphorylation reactions are involved in the regulationof the agonist binding site of the AMPA receptor population.72As stated in the General Introduction, the functional modification of AMPAreceptors by CaMKII, PKC, as well as PKA has been the focus of several studies(McGlade-McCulloh et al., 1993; Keller et al., 1992; Greengard et al., 1991).CaMKII (Tan et al., 1994; McGlade McCuIIoh et al., 1993), PKC (Tan et al.,1994; McGlade McCulloh et al., 1993) as well as PKA (Keller et al., 1992; Shawet al., 1992; Greengard et al., 1991) have been reported to be involved in theregulation of the AMPA receptor population in various neural preparations.Native AMPA receptors have been shown to be phosphorylated by CaMKII in avariety of preparations (McGlade-McCulloh et al., 1993). Activation ofendogenous stores of CaMKII led to the phosphorylation of GIuR1 inhippocampal postsyn aptic densities, synaptosomes and cultured hi ppocampalpyramidal neurons. PKC only resulted in a slight enhancement ofphosphorylation of GluRl, whereas treatment with the catalytic subunit of PKAdid not result in the phosphorylation of the GIuR1 subunit in any of thesepreparations. Moreover, activation of endogenous CaMKII led a three- tofourfold enhancement of AMPA receptor-mediated currents in culturedhippocampal neurons. In a further study, Tan et al. (1994) reported thatactivation of NMDA receptors by treatment of hippocampal neurons withglutamate and glycine increased 32P labeling of immunoprecipitated AMPAreceptors by 145% of control values. This increased phosphorylation of theAMPA receptor population was primarily 32P-serine with little32P-threonineand no detectable32P-tyrosine, and could be blocked by a NMDA receptorantagonist (AP-5) or by 1 -[N, O-bis( 1 ,5-isoquinolinesulfonyl)-N-methyl-L-tyrosyl]-4-phenylpiperazine (KN-62), a membrane-permeable inhibitor of CaMKII.These results suggested that Ca2 influx through the NMDA receptor73population could result in the activation of CaMKII and concomitantphosphorylation and regulation of the AMPA receptor.In contrast to the above, Keller et al. (1992) reported that bath application of amembrane permeable analogue of cAMP, an activator of PKA, potentiated thecurrents and also the flux of calcium through AMPA receptor channelscomprised of GIuR1 and GIuR3 subunits expressed in Xenopus oocytes. Theaverage current increases were significantly smaller when oocytes were loadedwith the PKA inhibitor Rp-adenosine 3’,5’-cyclic monophosphothioate (RpcAMPS). In addition to the preceding results, PKA had been demonstrated toincrease the opening frequency and the mean open time of the AMPA receptorchannels in hippocampal pyramidal neurons, an effect which could becompletely blocked by a specific PKA inhibiting peptide (Greengard et at.,1991). Other results suggested that exposure of adult cortical rat slices to thecatalytic subunit of PKA led to a significant decrease in[3H]-CNQX binding.This effect could be blocked by a specific peptide inhibitor of PKA (Shaw et at.,1992).Materials and MethodsBrain slices were prepared as described in the General Methods.Unlike the usual method of slice preparation where every effort was made tokeep the slices alive, the slices in the present experiments were frozen to -30°Cand then rapidly thawed. Freeze!thaw treatment promotes ice-crystal formationin slices thereby disrupting cellular membranes and permeabilizing cells,thereby allowing phosphorylating enzymes to reach intracellular74phosphorylation sites. Such treatment had no effect on[3H}-CNQX binding(see Chapter 1 and Lanius and Shaw, 1992). Once the slices were thawed,500 pL Dulbecco’s medium containing the following enzymes or enzymeactivators was added to each slice:i) CaM Kll activating cocktail (0.5 mM CaCl2, 2 pM calmodulin, 5 pM PKAinhibiting peptide (IP), 2 pM PKC-lP as described by Tan et al., 1994;McGlade-McCulloh et al., 1993)ii) catalytic subunit of PKA (Sigma, Type Ill, 20 ng/mL, Stelzer et al., 1988)iii) 1 2-O-tetradecanoylphorbol 13-acetate (TPA) (potent PKC activator, 200 nM;Pelech et al., 1991; Cochet et al., 1986; Pelech et al., 1986; Tapley andMurray, 1985; Kraft et aL, 1982)Dose response curves for the various enzymes or enzyme activators were notcarried out, since optimal concentrations have been previously determined (seereferences above), Incubation with the CaMKII activating cocktail or thecatalytic subunit of PKA was allowed to proceed for a minimum of 25 mm at37°C. These incubation times were shown by time course experiments toachieve maximal effects (Figures 13 & 14). Incubation with TPA was allowed tocontinue for 15 mm (Tan et al., 1994; Cochet et al., 1986; Pelech et al., 1986) at37°C. The slices were then rinsed twice for 30 mm at 4°C, and radioligandbinding was carried out as described as in the General Methods.To determine whether the changes in[3H]-CNQX binding observed as a resultof CaMKII or PKA treatment were reversible, adult rat cortical slices wereincubated with CaMKII activating cocktail or the catalytic subunit of PKA for 25mm at 37°C. The slices were then rinsed twice for 30 mm with cold Dul after75which fresh Dul was added for 5 to 45 mm at 37°C. Radioligand binding wasthen carried out as described in the General Methods.An additional set of experiments examined whether the effects of CaM Kil orPKA on AMPA receptor regulation could be inhibited using selective inhibitors(see Table 3 below). Since TPA had no effect on[3H]-CNQX binding, PKCinhibitors were not used. Dose-response curves for the various kinaseinhibitors were not carried out, since optimal concentrations have beenpreviously determined for all these inhibitors in a variety of differentpreparations (see Table 3 for individual references).Inhibitor Specificity Concentration ReferencesPKA-IP PKA 5 pM 1) Tan etal., 19942) McGlade-McCullohetal., 19933) Roth et aL, 19914) Smith et al., 1990KN-62 CaMKII 50 pM 1) Tansey et al., 19922) lshii et al., 19913) Tokumitsu et al., 1990Table 3: Protein kinase inhibitorsSlices were co-incubated with either the enzyme or enzyme activator as well asthe specific inhibitor for that particular enzyme for a minimum of 25 mm at 37°C.76After this incubation, the slices were rinsed twice for 30 mm at 4°C, after whichreceptor binding assays were carried out as described in the General Methods.ResultsTime course experiments examining[3H]-CNQX binding as a function ofincreasing incubation times with a CaMKII activating cocktail at 37°C werecarried out to establish the incubation time most effective in inducing AMPAreceptor regulation by CaMKII (Figure 14).[3H]-CNQX binding was decreasedby 35% following a 20 mm incubation time; this effect remained stable for up to2 h. A minimum incubation time of 20 mm was, therefore, employed insubsequent experiments.Time course experiments examining[3H]-CNQX binding following thewithdrawal of a CaMKII activating cocktail were carried out to determinewhether the decrease in[3H]-CNQX binding observed as a result of treatmentwith CaM Kil activating cocktail was reversible (Figure 15). The decrease in[3H]-CNQX binding observed as a result of treatment with a CaM KIl activatingcocktail was reversible upon withdrawal of the CaMKII activating cocktail.Following washout of the CaMKII activating cocktail, AMPA receptor numbergradually increased, reaching control values by approximately 30 mm.Time course experiments examining specific[3H]-CNQX binding as a functionof increasing incubation times with the catalytic subunit of PKA at 37°C werecarried out to establish the incubation time most effective in inducing AMPAreceptor regulation by PKA (Figure 16).[3H]-CNQX binding was decreased by30% following a 25 mm incubation time; this effect remained stable for up to 2 h.77A minimum incubation time of 25 mm was therefore employed in subsequentexperiments.Time course experiments examining specific[3H]-CNQX binding following thewithdrawal of PKA were carried out to determine whether the decrease in [3H]-CNQX binding observed as a result of treatment with the catalytic subunit ofPKA was reversible (Figure 17). The decrease in[3H]-CNQX binding observedas a result of treatment with PKA was reversible upon withdrawal of PKA.Following washout of PKA, AMPA receptor number gradually increased,reaching control values by approximately 15 mm.The effects of CaM KIl, PKA or PKC treatment on the AMPA receptor populationare shown in Figure 18. Activation of endogenous CaMKII led to a significant33% decrease in[3H]-CNQX binding (p.<O.05, Student’s t-test) which could beblocked by KN-62 (n=4). Similarly, PKA led to a significant 30% decrease in[3H]-CNQX binding (p<O.05, Student’s t-test) which could be blocked by aspecific PKA inhibiting peptide (PKAIP) (n=4). Treatment of cortical slices withTPA had no effect on[3H]-CNQX binding (n=3). KN-62 and PKAIP had nosignificant independent effect on[3H]-CNQX binding (data not shown).780•C -10,oCo00c) -30••Time (mm)Figure 14: Time course experiments showing[3H]-CNQX binding as a functionof increasing incubation times with a CaMKII activating cocktail. Ligandconcentration was 5-10 nM. All incubations with[3H]-CNQX were for 3 h at 4°C.Rinse times were 2x5 mm. Error bars give S.E.M. (n=3).793500-— T TCD_l_ J3000-C.92500- - -‘-0 -II0w2000TCl)J1500-0 10 20 30 40 ContTime (mm)Figure 15: Reversibility of CaMKII-induced regulation of[3H]-CNQX binding.Cortical slices were incubated with CaMKII activating cocktail for 25 mm at37°C. The slices were then rinsed twice for 30 mm with cold Dul after whichfresh Dul was added for 5 to 45 mm at 37°C. All incubations with[3H]-CNQXwere for 3 h at 4°C. Rinse times were 2x5 mm. Error bars give S.E.M.. Specificbinding is expressed as dpm.800-0- -20-4)C)29-30-I-I0-40 i • • • • •0 20 40 60 80 100 120 140Time (mm)Figure 16: Time course experiments showing[3H]-CNQX binding as a functionof increasing incubation times with the catalytic subunit of PKA. Ligandconcentration was 5-10 nM. All incubations with[3H]-CNQX were for 3 h at 4°C.Rinse times were 2x5 mm. Error bars give S.E.M. (n=3).810)CC00C)C.(1)45004000 11TTI C-I--IC-I-3000TIDI0 10 20 30 40Time (mm)Figure 17: Reversibility of PKA-induced regulation of[3H]-CNQX binding.Cortical slices were incubated with PKA for 25 mm at 37°C. The slices werethen rinsed twice for 30 mm with cold Dul after which fresh Dul was added for5 to 45 mm at 37°C. All incubations with[3H]-CNQX were for 3 h at 4°C. Rinsetimes were 2x5 mm. Error bars give S.E.M.. Specific binding is expressed asdpm.A 82a)CCC)U0.0B0)C0________Control Q.2 TPAC.)C)0.Cl)Figure 18: Enzymatic control of the regulation of AMPA receptors labeled with[3H]-CNQX. AMPA receptor binding values were determined after CaMKII,PKA, or TPA treatment alone and in conjunction with selective inhibitors (seeTable 3). Error bars show S.E.M. (see Results for n values), and asterisksindicate significance to the p<O.05 level (Student’s t-test). Specific binding isexpressed as dpm.Control QCaMKII E!JCaMKII÷KN-62 WIPKA IIIPKA+PKAIP83DiscussionThe preceding regulation experiments have demonstrated reversible decreasesin[3H]-CNQX binding in response to a CaMKII activating cocktail and treatmentwith the catalytic subunit of PKA. These effects were seen within approximately25 minutes of treatment and remained stable for up to two hours. The actions ofthese enzymes were selective and could be blocked by the CaMKII inhibitorKN-62 and a specific PKA inhibiting peptide, respectively. Treatment of corticalslices with TPA, an activator of PKC, had no effect on[3H]-CNQX binding.We have previously reported that PKA-induced changes in[3H]-CNQX bindingare likely due to changes in high-affinity AMPA receptor number (Shaw et at.,1992). However, the interpretation of these results has to be reevaluated. Itremains unknown whether AMPA receptor regulation induced by CaMKII andPKA is due to the regulation of the high- or low-affinity component of[3H]-CNQXbinding. Moreover, it remains unknown whether a change in receptor numberand/or affinity is responsible for the observed changes in[3H]-CNQX binding.However, as discussed, in Chapter 2, several possibilities seem plausible. (1) Itis possible that the decreases in[3H]-CNQX binding observed after treatmentwith CaM KIl and PKA reflect a shift from the high-affinity state to the low-affinitystate of the AMPA receptor. (2) An alternative possibility is that CaM Kit and PKAlead to a change in receptor number of the high- and/or low-affinity state of theAMPA receptor. (3) A third possibility is that CaM KIl and PKA lead to a changein receptor affinity of the high- and/or low affinity state of the AMPA receptor. (4)Lastly, it is conceivable that CaM KIl and PKA result in changes in receptornumber affinity of the high- and/or low-affinity state of the AMPA receptor.Experiments using newly developed competitive antagonists which84preferentially bind to the high- or low-affinity component of the AMPA receptor(Ebert et al., 1994) could be used to distinguish whether CaMKII, PKA, AMPA, orveratridine are regulating the high and/or low-affinity component of the AMPAreceptor.Changes in ligand binding as a result of treatment with phosphorylating agentshave been reported previously for a variety of receptor populations, includingthe glycine (Vaello et al. 1992), GABAA (Shaw et al., 1992) and NMDA(Kitamura et al., 1993; Krieger et al., 1993) receptors. Vaello et al. (1992)examined the effects of PKC-mediated phosphorylation of glycine receptorspurified from rat spinal cord on the subsequent interaction of glycine with itsreceptor. Under control conditions glycine displaced[3H]-strychnine bindingwith an 1C50 of 58 pM, whereas for the phosphorylated glycine receptors thelC50 was 200 pM. These results indicated a three-fold decrease in the ability ofglycine to interact with phosphorylated glycine receptors.A decrease in[3H]-SR 95531 (GABAA antagonist) binding was reportedfollowing treatment with the catalytic subunit of PKA in rat cortical slices. Theseeffects were attributable to changes in receptor number as indicated bysaturation binding analyses and could be blocked by a specific PKA inhibitingpeptide (Shaw et al., 1992). For the excitatory NMDA receptor channel,increases in[3H]-MK-801 binding have been reported following TPA treatment(Kitamura et al., 1993) in postsynaptic densities of rat brain and sections ofhuman spinal cord (Krieger et al., 1993). Kitamura et al. (1993) attributed thesechanges in[3H]-MK-801 binding to changes in receptor affinity rather thannumber. TPA appeared to have no effect on radioligand binding to the agonist85binding site of the NMDA receptor complex, suggesting that phosphorylationonly modulates channel function of NMDA receptors.The time course required to observe changes in the amplitude of receptor-mediated currents following receptor phosphorylation appears to resemble thetime course needed to detect CaMKII and PKA-induced alterations in [3H]-CNQX binding. GABAA receptors expressed in human embryonic kidney 293cells showed a 30%-40% depression of GABA currents after 15 to 20 minutetreatment with 300 pM cAMP. Site-specific mutagenesis of the serine residuephosphorylated by PKA completely eliminated the depression of the GABAAreceptor-mediated currents (Moss et al., 1992). In contrast, 100% increases inthe amplitude of kainate receptor-mediated currents were observed afterintracellular application of the catalytic subunit of PKA in human embryonickidney 293 cells expressing GIuR 6. This potentiation was complete after 35minutes and could be blocked by a specific PKA inhibiting peptide (Raymond etal., 1993). A similar potentiation of AMPA receptor-mediated currents wasreported in hippocampal neurons following treatment with activated CaMKII.This three- to four-fold increase in AMPA receptor-mediated currents wascomplete after approximately 12 to 15 minutes (McGlade-McCulloh et al., 1993).It is therefore possible that changes in[3H]-CNQX binding reflectphosphorylation-induced changes in AMPA receptor-mediated currents. In thisview, phosphorylation leads to a change in the agonist binding sites with anaccompanying alteration in the receptor-mediated current response.Although the mechanism underlying the decreases in AMPA receptor numberas a result of CaMKll or PKA treatment is not clear, several possibilities can beconsidered: (1) receptor sequestration! internalization, (2) receptor down-86regulation and/or (3) a change in the conformational state of the receptorresulting in decreased[3H]-CNQX binding. Receptor internalization/sequestration can be excluded, since a decrease in AMPA receptor numberdue to internalization of the receptors cannot be detected in freeze/thawedslices. Receptor-down regulation also did not seem to occur, since thedecrease in[3H]-CNQX binding was fully reversible within approximately 30minutes, and de novo receptor synthesis is unlikely to occur in such a shorttime frame (Collins et al., 1991; Campbell et al., 1991). However, onepossibility is that CaM KIl and PKA lead to changes in the conformational state ofthe AMPA receptor agonist binding site thereby preventing binding of thelabeled probe. In this case, total receptor number need not necessarily change,but functional receptor number will be altered. In other words, receptor proteinsmay still be present in the membrane, but will be maintained in a non-functionalstate.Protein kinases, including CaMKll and PKA, are known to be involved incontrolling the amount of neurotransmitter released presynaptically through thephosphorylation of presynaptic proteins synapsin I and II (Browning and Dudek,1992; Greengard et al., 1993). Is it therefore possible that the observed effectsof these enzymes are not attributable to receptor phosphorylation but rather toaltered neurotransmitter release at the presynaptic terminal resulting in receptorregulation? This possibility seems unlikely since the present experiments wereperformed using a frozen/thawed preparation. This procedure should lead to adepletion of neurotransmitter due to lost membrane integrity and diffusion ofneurotransm lifer.87The present data reporting changes in[3H]-CNQX binding in response toCaM Ku activating cocktail and the catalytic subunit of PKA are in generalagreement with the literature suggesting regulation of the AMPA receptorpopulation by CaMKll (McGlade-McCulloh et al., 1993) and PKA (Keller et al.,1992; Shaw et al., 1992; Greengard et al., 1991). However, several differencesremain to be resolved. McGlade-McCulloh et al. (1993) failed to see an effect ofPKA on AMPA receptor phosphorylation, whereas other groups have shownAMPA receptor modification by this enzyme (Keller et al., 1992; Shaw et al.,1992; Greengard et al., 1991). How can these opposing results be reconciled?Several possible explanations can be considered. First, it has been reportedthat the consensus sequence for phosphorylation by CaMKII is identical to a lowaffinity consensus sequence for phosphorylation by PKA (Kennely and Krebs,1991). In vitro, PKA has been shown to phosphorylate CaMKII phosphorylationsites as frequently, and perhaps as efficiently, as PKA phosphorylation sites(Kennely and Krebs, 1991). The observed effects of PKA may therefore be dueto the phosphorylation of CaMKll phosphorylation sites by PKA. Second, PKAmay be part of a cascade of phosphorylation reactions ultimately leading to thedirect phosphorylation of AMPA receptors by CaMKll. An alternativeexplanation may involve differences in receptor subunit expression in thedifferent preparations which may affect the ability of PKA to phosphorylate theAMPA receptor population. In support of this view, it has recently been reportedthat phosphorylation of the GABAA receptor has differential effects dependingon the receptor subunit composition and the number of phosphorylation sitescontained within these subunits (Krishek et al., 1994; Moss et al., 1992).Another controversy which remains to be resolved concerns the direction ofregulation of AMPA receptors observed after activation of CaMKII or treatment88with the catalytic subunit of PKA. Both CaM Ku and PKA have been shown toincrease AMPA receptor-mediated currents (McGlade-McCuIIoh et al., 1993;Keller et al. 1992; Greengard et al., 1991). However, the present results showthat both of these enzymes decrease[3H]-CNQX binding. In this regard, wehave recently reported that the effects of PKA are strictly dependent on thepostnatal age of the animals used (reviewed by Lanius et al., 1993; Shaw andLanius, 1992). In animals less than 40 days postnatal age, treatment with thecatalytic subunit of PKA resulted in a significant increase in[3H]-CNQX binding.In contrast, in animals greater than 60 days postnatal age, PKA resulted in asignificant decrease in[3H]-CNQX binding. Similar results have also beenreported for the GABAA receptor (reviewed by Lanius et al., 1993; Shaw andLanius, 1992). Although the reasons for these age-dependent differencesremain unknown, these results may be due to the differential expression ofprotein kinases (Lai and Lemke, 1991) which have been demonstrated to occurin the nervous system during development. Alternatively, differences in theexpression of AMPA receptor subunits during development (Martin et aL, 1993;Wenthold et al., 1992) may be contributing to the observed differences inreceptor regulation. Since McGlade-McCulloh et al. (1993) and Greengard etal. (1991) used hippocampal cultures from newborn rat pups, it is conceivablethat the above controversy may arise from differences in postnatal age.Although the current data have shown that phosphorylating enzymes can alter[3H1-CNQX binding, it remains unknown whether the AMPA receptor populationis directly phosphorylated by these enzymes. Future studies using antibodiesdirected against GIuR 1-4, phosphoserine, phosphothreonine, andphosphotyrosine residues to probe Western blots of treated and untreated ratcortical slices could resolve this issue.89In summary, the present chapter has shown that activation of endogenousCaM Ku as well as treatment with the catalytic subunit of PKA can alter[3H]-CNQX binding. Both treatments resulted in a decrease in[3H]-CNQX bindingsimilar to the decrease in binding observed in response to treatment with AMPAand veratridine. Moreover, the time course of regulation by CaMKII and PKAclosely resembled that of AMPA and veratridine. Based on these similarities, itis possible that the regulatory action of AMPA and veratridine may beattributable to endogenous kinase activity. This issue will be explored in thechapters to follow.50CHAPTER 4AMPA RECEPTOR REGULATION BY AGONIST ANDDEPOLARIZING STIMULI REQUIRES CaMKII AND/OR PKAIntroductionAgonist (AMPA) stimulation and pharmacologically-induced increases incellular depolarization have been shown to lead to the regulation of corticalAMPA receptors. In cortical slices agonist stimulation as well as increases incellular depolarization led to approximately 20% decreases in[3H]-CNQXbinding (Chapter 2). Using a similar in vitro preparation, the ability of CaMKlland PKA to induce regulation of[3H]-CNQX binding was shown. Activation ofCaMKII and treatment with the catalytic subunit of PKA both led toapproximately 35% and 30% decreases in[3H]-CNQX binding, respectively(Chapter 3).AMPA receptor regulation by CaMKII and PKA appeared to be qualitativelysimilar to regulation by agonist and depolarizing stimuli. Moreover, the timecourse of regulation by CaMKll and PKA closely resembled the time course ofAMPA- and veratridine-induced regulation. All four of these agents were able toinduce maximal AMPA receptor regulation within approximately 25 mm. Basedon these similarities, it was hypothesized that the regulatory action of thesestimuli may be attributable to endogenous kinase activity. The experiments inthis chapter are therefore designed to examine whether regulation achieved byagonist and depolarizing stimuli can be blocked by selective protein kinaseinhibitors. Since AMPA receptors have been shown to be regulated by CaMKII91and PKA, the effects of inhibitors of these enzymes on agonist and depolarizingstimuli will be examined.Materials and MethodsIntact brain slices, prepared as described previously in the General Methods,were incubated for a minimum of 25 mm at 37°C with a combination of thefollowing compounds (see below) to determine their effects on AMPA receptorregulation:AI) AMPA (10-5 M)ii) AMPA + CaMKII inhibitor KN-62 (10pM)iii) AMPA + PKA inhibitor Rp-adenosine 3’,5’-cyclic monophosphothioate(Rp-cAMPS) (100 pM)iv) AMPA + PKC inhibitor bisindolylmaleimide (0.5 pM)Bi) Veratridine + Glutamate (10-5 M)i) Veratridine + Glutamate + CaMKII inhibitor KN-62 (10 pM)ii) Veratridine + Glutamate + PKA inhibitor Rp-cAMPS (100 pM)iii) Veratridine + Glutamate + PKC inhibitor bisindolylmaleimide (0.5 pM)Following incubation with the above mentioned substances, the slices wererinsed for 2 x 30 mm with 4°C Dul. Radioligand binding was then carried outas described previously.92Inhibitor Specificity Concentration ReferenceKN-62 CaMKII 10 pM 1) Tansey et al., 19922) Ishii et aL, 19913) Tokumitsu et al., 1990Rp-CAMPS PKA 100 pM 1) Wang et al., 19912) Van Haastert et al.,1984Bisindolyl- PKC 0.5 pM Toullec et a!., 1991maleimideTable 4: Protein kinase inhibitors used to inhibit regulation by AMPA and v+gResultsTo evaluate whether the regulatory effects of AMPA on[3H]-CNQX bindingcould be attributed to endogenous kinase activity, the effects of inhibitors ofCaMKll, PKA, or PKC on decreases in[3H]-CNQX binding seen as a result ofAMPA receptor stimulation were examined (Figures 19&20). Treatment ofcortical slices with AMPA (105 M) resulted in a significant decrease in [3H]-CNQX binding to AMPA receptors (-19% to -29%) compared to control (p<0.05,Student’s t-test) (Figure 19A n=3; Figure 19B n=4; Figure 20A n=3; Figure 20Bn=3). Experiments in Figures 19A&B were not pooled since the concentrationof[3H]-CNQX used varied between the two experiments. The effect of AMPAcould be blocked by the CaMKll inhibitor KN-62 (Figure 19A n=4; Figure 19Bn=3) as well as the PKA inhibitor Rp-cAMPS (n=4) (Figure 20). The PKC93inhibitor bisindolylmaleimide was not able to prevent the decreases in [3H]-CNQX binding seen as a result of AMPA receptor stimulation (n=4) (Figure 20).KN-62, bisindolylmaleimide, and Rp-cAMPS had no independent effect on [3H]-CNQX binding (data not shown).To determine whether the regulatory effects of v+g on[3H]-CNQX binding couldbe attributed to endogenous kinase activity, the effects of inhibitors of CaMKII,PKA, or PKC on decreases in[3H]-CNQX binding seen as a result of v+gtreatment were examined (Figures 21 &22). Treatment of cortical slices with v-i-gresulted in significant decreases (1 8%-22%) in[3H]-CNQX binding compared tocontrol (Figure 21A n=2; Figure 22A n=3; Figure 22B n=3) (p<0.05, Student’s ttest) (Figures 21A & 22A&B). In Figure 21 B, a one-way analysis of variance(ANOVA) showed a significant difference among the groups (p<0.05).Experiments in Figures 21 A&B were not pooled since the concentration of[3H]-CNQX used varied between the two experiments. The effects of v+g wereabolished by the CaMKII inhibitor KN-62 (Fig. 21 A n=4; Fig. 21 B n=3) as well asthe PKA inhibitor Rp-cAMPS (n=4). Bisindolylmaleimide was not able to inhibitthe decreases in[3H]-CNQX binding observed as a result of v+g treatment(n=4).94A0C.)0C)0.C,)B0)CC00_____________0)a(1)Figure 19: Regulation of AMPA receptors labeled with[3H]-CNQX.[3H]-CNQXbinding was determined after agonist stimulation alone and in conjunction withselective kinase inhibitors (see Table 4). Error bars show S.E.M. (see Resultsfor n values), and asterisks indicate significance to the p<O.05 level (Student’s ttest). Experiments were not pooled since the concentration of[3H]-CNQX usedvaried between the two experiments.Control QAMPAAMPA + KN-62 •Control QAMPAAMPA + KN-62 I—-.CD0sa’ C CD U) 0. Co CD CO (I) D 0 C) CD U) CD -h C) 0) C) CD 0 CD.0 0 (11 CD CD C,) 0. CD D (I)U)—.CD<CDCD0)5cojj0.CDCDCDCD——.CD.-‘•D•CD0-IL4.-hU)0)>CDCD-IUCD0)>—lCD-‘0)0CDCDU)0.-S (1)-S—o.CD‘0)U)D 0)1D0)0.6fl15ZoO•0><Co—.CDC—DC.’)E:ii—-CD06CoDC)+>0-U-SpecificBindingSpecificBinding-C,)0§§§§-U +-UC)o>0U,96A4O00C______:5• 300002000-C.)a)0.ooo0— IB3000-T2500- £0) Tiooo_____0.U)500-U-— I—.Figure 21: Regulation of AMPA receptors labeled with[3H]-CNQX.[3H]-CNQXbinding was determined after v-i-g treatment alone and in conjunction withselective kinase inhibitors (see Table 4). Error bars show S.E.M. (see Resultsfor n values), and asterisks indicate significance to the p<O.05 level (Student’s ttest). Experiments were not pooled since the concentration of[3H]-CNQX usedvaried between the two experiments.* tsl‘/7/77/‘/7/7/ Wh.S.‘//// %%%%.,1‘/7/“7Control Dvgvg + KN-62Control Qvgvg + KN-62 I97ACU0wC(1)B5000C)4000C3000020000C.C,)1000Figure 22: Regulation of AMPA receptors labeled with[3H]-CNQX.[3H]-CNQXbinding was determined after v-t-g treatment alone and in conjunction withselective kinase inhibitors (see Table 4). Error bars show S.E.M. (see Resultsfor n values), and asterisks indicate significance to the p<O.05 level (Student’s ttest).Control Qvgvg + RpcAMP QUControl Dvgvg+BIS98DiscussionThe preceding data have shown that the decreases in[3H]-CNQX bindingobserved as a result of AMPA or v+g treatment could be blocked by inhibitors ofCaMKll and PKA (see Chapter 5 for discussion; Pasqualotto et al., 1994). ThePKC inhibitor bisindolylmaleimide had no effect on AMPA receptor regulationinduced by AMPA or v+g. These results suggested that protein kinases,particularly CaMKll and/or PKA, are an essential element in the regulatoryresponse of AMPA receptors to depolarizing stimuli such as AMPA and v+g.A common feature of both AMPA and v-i-g is that they cause cellulardepolarization. It has previously been shown that cellular depolarization canlead to the phosphorylation of intracellular proteins. Sieghart et al. (1978) haveshown that depolarization induced by veratridine and high K stimulated theincorporation of 32P into two specific proteins in a rat synaptosomalpreparation. Furthermore, it is now well established that depolarization andconcomitant Ca2+ influx into the presynaptic terminal is required for controllingthe amount of neurotransmitter released through CaMK I and II- as well as PKAmediated phosphorylation of the presynaptic proteins synapsin I and II(Greengard, 1993; Browning and Dudek, 1992). It is therefore possible thatcellular depolarization induced by AMPA or v+g promotes the activation ofCaMKII and/or PKA, thus resulting in decreases in[3H]-CNQX binding.It is well recognized that alterations in membrane potential result from changesin the ionic currents across the cellular membrane. Both the agonist anddepolarizing stimuli employed to study activity-dependent AMPA receptorregulation are known to result in alterations in the ionic currents across the99cellular membrane. Veratridine results in increased intracellular Na+ due to itsability to prevent the inactivation of Na channels (Catterall, 1980). Indirecteffects of veratridine also include increased intracellular Ca2+ concentrationsthrough the depolarization-induced activation of voltage-gated Ca2 channels.Activation of cortical AMPA receptorsleads to cellular depolarization due toincreased Na fluxes through AMPA receptors (Seeburg, 1993) withconcomitant increases in intracellular Ca2+due to the depolarization-inducedactivation of voltage-gated Ca2 channels (Church et a!, 1994). The role ofthese ions in mediating decreases in[3H]-CNQX binding will therefore be thefocus of the following chapter.100CHAPTER 5Ca2-DEPENDENCE OF AMPA RECEPTOR REGULATIONIntroductionA common feature of both agonist and depolarizing stimuli employed tomodulate AMPA receptor characteristics is that they cause transient, butprofound changes in postsynaptic ionic concentrations within the cell. Changesin the postsynaptic intracellular ionic environment as a result of agonist ordepolarizing stimuli may provide the trigger for the control of AMPA receptorregulation by phosphorylation. In this view, the ionic currents which crossneural membranes in response to various depolarizing stimuli will activatespecific kinases leading to receptor regulation. For AMPA receptors, whoseactivation by agonists leads to a direct influx of Na+ as well as a secondaryinflux of Ca2+through voltage-gated Ca2+ channels, a role for one or both ofthese ions in the activation of kinase-mediated decreases in[3H]-CNQX bindingwill be investigated. Experiments will examine the effects of Ca2 and Na on[3H]-CNQX binding. Moreover, to examine a potential role for voltage-gatedCa2 channels in AMPA receptor regulation by agonist and depolarizingstimuli, the effects of loperamide, a non-specific blocker of voltage-gated Ca2channels, will be investigated. Electrophysiological studies have recentlyshown that Ca2+ influxes through voltage-gated Ca2+channels as a result ofAMPA receptor stimulation could be blocked by the antidiarrheal agentloperamide (Church et aL, 1994).101Although the effects of Ca2+ as a trigger for certain kinases such as Ca2+phospholipid-dependent PKC (Nishizuka, 1988) and Ca2/c lmod ul ifldependent kinase (Hanson and Schulman, 1992) have been well documented,much earlier studies had shown that the phosphorylation of certain membraneproteins of the electric organs of Torpedo Cailfornica and Torpedo mamoratawere controlled by specific ionic species (Na+ and K+) not usually associatedwith the activation of protein kinases (Gordon et al., 1977; Saitoh andChangeux, 1980). Saitoh and Changeux (1980) reported that an increase inthe concentration of Na+ from 25 to 250 mM led to a decrease of thephosphorylation of approximately 80% of membrane proteins in Torpedomamorata ; increasing concentrations of K+ enhanced the phosphorylation ofsome polypeptide chains with a molecular weight equal to or higher than85 000. Similarly, Gordon et al. (1977) showed that phosphorylation of severalpolypeptides from the electric organ of Torpedo californica was stimulated bythe addition of K. These results show that although Ca2 appears to be themost likely candidate for triggering phosphorylation reactions, a role for otherions (Na+/K+) cannot be ruled out.Materials and MethodsThe Effects of Ca?±j±, and K± on [H1-CNQX BindingFull details of the cortical slice preparation and receptor assays are discussedin the General Methods. The brains of Sprague-Dawley rats were rapidlyremoved and dissected in a 50 mM Tris-acetate buffer (pH 7.4). The slicesobtained from each brain were placed into tissue culture wells, slowly frozen to-30°C, and then rapidly thawed.102Slices were then pre-incubated in different concentrations of sodium acetate (6-200 mM), potassium acetate (6-200 mM), or calcium acetate (0.1-1 mM) for aminimum of 25 mm at 37°C before incubation with[3H]-CNQX. Similar Ca2concentrations have been previously employed to activate CaMKII (Tan et al.,1994; McGlade-McCulloh et al., 1993). The buffers were adjusted tophysiological pH (7.4) using 99% glacial acetic acid in all cases.In order to determine whether the effects observed were due to ions acting onpotential endogenous kinases, slices were co-incubated with the specific ionsin addition to selective kinase inhibitors: 10 pM KN-62, a CaMKII specificinhibitor (Tokumitsu et al., 1990), 5pM protein kinase A inhibiting peptide(Smith et at., 1990), or 0.5 pM bisindolylmaleimide, a PKC inhibitor (Toutlec etal., 1991).Following the preincubation step, slices were rinsed for 2 x 30 mm with 100 mMTris-acetate buffer (4°C) and incubated with 5-10 nM[3H]-CNQX in Tris-acetatebuffer for 3 h at 4°C. After the radioligand binding step, slices were rinsed for 2x 5 mm with 100 mM Tris-acetate buffer (pH 7.4) at 4°C and counted in aBeckman 6000 IC scintillation counter.The Effects of LoDeramide on AMPA ReceDtor Regulation by Agonist andDepolarizing StimuliIntact brain slices were prepared as described in the General Methods. Sliceswere co-incubated with AMPA (105 M) or v÷g (10-5 M) and 25 pM or 100 pMloperamide (Church et al., 1994; Pasqualotto et al., 1994) in Dul for a minimumof 25 mm at 37°C. Following this incubation, slices were rinsed for 2 x 30 mm103with DuI at 4°C. Radioligand binding was then carried out as described in theGeneral Methods.ResultsThe Effects of Ca2±J±. and K± on [aH]-CNQX BindingTo determine the effects of Ca2on[3H]-CNQX binding, the effects of Ca2(0.1-1 mM) alone and in combination with KN-62, PKAIP, orbisindolylmaleimide were examined (Figures 23 & 24). Ca2 significantlydecreased[3H]-CNQX binding over a concentration range of 0.1 to 1 mM(p<0.05, Student’s t-test). This decrease could be partially blocked by PKAIP(n=3). KN-62 was able to completely inhibitCa2-induced decreases in [3H]-CNQX binding (n=3). Bisindolylmaleimide had no effect on the decrease in[3H]-CNQX binding observed after incubation with Ca2 (n=3). Neither KN-62,PKAIP, nor bisindolylmaleimide had any independent effect on[3H]-CNQXbinding (data not shown).The effects of Na and K (6-200 mM) on[3H]-CNQX binding were examined(Figure 25). Neither K nor Na in this concentration range had a significanteffect on[3H]-CNQX binding.In order to control for possible effects of acetate alone, the effect of calciumacetate was compared to that of sodium and potassium acetate. Calciumacetate resulted in a decrease in[3H]-CNQX binding, whereas neither sodiumnor potassium showed an effect.104The Effects of Loperamide on AMPA Receptor Regulation by Agonist andDepolarizing StimuliTo establish whether AMPA-induced regulation of AMPA receptors requires aninflux of Ca2 through voltage-gated Ca2 channels, the effects of loperamide(25 p.M or 100 p.M) on decreases in[3H]-CNQX binding seen as a result ofAMPA treatment were examined (Figure 26). Treatment with 1 M AMPAresulted in a significant decrease in[3H]-CNQX binding (p.<0.05, Student’s ttest) which could only be partially blocked by 25 p.M loperamide (n=5) (Figure26A). However, the decrease in[3H]-CNQX binding as a result of AMPAtreatment could be completely blocked by 100 p.M loperamide (n=5) (Figure26B). A one-way analysis of variance (ANOVA) showed a significant differenceamong the groups (p<0.05). Loperamide did not have any significantindependent effects on[3H]-CNQX binding (data not shown).To determine whether veratridine-induced regulation of AMPA receptorsrequires an influx of Ca2+via voltage-gated Ca2+ channels, the effects ofloperamide (100 p.M) on decreases in[3H]-CNQX binding seen as a result ofv+g treatment were examined (Figure 27). Treatment with veratridine resultedin significant 23% decreases in[3H]-CNQX binding (p<0.05, Student’s t-test)which could be completely blocked by 100 p.M loperamide (n=3).105I I I I0 0.25 0.5 0.75 1 1.25Ionic Concentration (mM)Figure 23: Effects of Ca2on[3H]-CNQX Binding. Significance to the p<O.05level (Student’s t-test) is indicated by an asterisk. Error bars show S. E. M. (seeResults for n values). Specific binding is expressed as dpm.DT0W•3000-2000-1000-C0‘4-0a)a.(I)-IC* TCC* I*Ca2’ CCa2 + KN-62 12A0)•0CC.)I.C.)0)C,)Ca2 aCa2 + BISFigure 24: Effects of Ca2 on[3H]-CNQX Binding. Significance to the p<O.05level (Student’s t-test) is indicated by an asterisk. Error bars show S.E.M. (seeResults for n values). Specific binding is expressed as dpm.c:i 4I50003000-2000-1000-106o*-,-a*0)C•0Cm0‘4-00)0.CoBT-I-*a*Ca2 aCa2 + PKI ‘T0I*a*,J I I I I0 0.25 0.5 0.75 1 1.25Ionic Concentration (mM)-Ii;i* **2000 -1500-1000-500-00 0.25IonicI I I0.5 0.75 1 1.25Concentration (mM)107A30002500-IT T TDIDJI1500-011000K0- I I I0 50 100 150 200Ionic Concentration (mM)BT1D TTDDDTJ0) 3000 D DC I J_•0C00a)Q) 1000-Na+0- I I I0 50 100 150 200Ionic Concentration (mM)Figure 25: Effects of K and Na on[3H]-CNQX Binding. Error bars showS.E.M. (see Results for n values). Specific binding is expressed as dpm.CC00w0.Cl)108300025002000150010005000A0)C•0C0‘I0G)0.Cl)BControl QAMPAAMPA + LOP (low) IJControl QAMPA 0AMPA + LOP (high) EEDFigure 26: The effects of loperamide on AMPA receptor regulation by AMPA.[3H]-CNQX binding was determined after AMPA treatment alone and inconjunction with loperamide (25 pM or 100 pM). Error bars show S.E.M. (seeResults for n values), and asterisks indicate significance to the p<0.05 level(Student’s t-test). Specific binding is expressed as dpm.1090)•0C0I.0a,__________0.Cl)Figure 27: The effects of loperamide on AMPA receptor regulation by v+g. [3H]-CNQX binding was determined after v-t-g treatment alone and in conjunctionwith loperamide (100 pM). Error bars show S.E.M. (see Results for n values),and asterisks indicate significance to the p<O.05 level (Student’s t-test).Specific binding is expressed as dpm.Control Qvgvg + LOP :110DiscussionA decrease in[3H]-CNQX binding was observed in response to treatment with0.1-1 mM calcium acetate. These decreases in binding could be completelyblocked by the CaMKll inhibitor KN-62 and partially blocked by a PKA inhibitingpeptide. Nielsen et al. (1990) have previously examined the effects of 2.5 mMcalcium acetate on[3H]-CNQX and[3H]-AMPA binding in cerebellar rat brainsections. Although they reported a significant decrease in[3H]-AMPA bindingin response to treatment with 2.5 mM calcium acetate, the effect of calciumacetate on[3H]-CNQX (50 nM) binding was not statistically significant.However, since 50 nM[3H]-CNQX labels AMPA- as well as kainate receptors, itis possible thatCa2-medi ted effects on the AMPA receptor population aremasked by[3H]-CNQX binding to kainate receptors.Much earlier studies had shown the ability of Ca2 to modify[3H]-glutamatebinding in rat hippocampus through an unknown mechanism (Baudry andLynch, 1979). Ca2was found to be very potent in enhancing[3H]-glutamatebinding over a concentration range of 25 pM to 50 mM. However, maximaleffects (100%) were seen at a concentration of approximately 1 mM. Scatchardanalyses of[3H]-glutamate binding in the presence of 1 mM Ca2 suggestedthat Ca2 altered the maximum number of[3H]-glutamate binding sites withoutchanging their affinity.Since Ca2 led to decreases in[3H]-CNQX binding which could be blocked bythe CaMKII inhibitor KN-62, the role of Ca2 in AMPA receptor regulation byAMPA and veratridine were further examined. Loperamide, a non-specificblocker of voltage-gated Ca2 channels, was able to inhibit the decrease in111[3H]-CNQX binding observed in response to treatment with AMPA or v-t-g(Pasqualotto et al., 1994). Loperamide has previously been shown to blockCa2 influxes through voltage-gated Ca2 channels as a result of AMPAreceptor stimulation in hippocampal pyramidal neurons (Church et al., 1994).These results suggested that Ca2 influx through voltage-gated Ca2 channelsis necessary for AMPA- and veratridine-induced AMPA receptor regulation viaCaMKII- and/or PKA-mediated phosphorylation reactions. However,loperamide has also been shown to display weak calmodulin antagonist activity(lC5O values ranging from 75-100 pM) (Kachur et al., 1986). It is thereforeplausible that the inhibitory effects of this compound are due to calmodulinantagonism. Experiments employing other Ca2+ channel blockers, includingnifedipine (L-type channel blocker) and w-conotoxin (N-type channel blocker)could distinguish between calmodulin- and Ca2 channel-mediatedantagonism.It remains unclear why only the higher concentration of loperamide (100 pM)was able to block the effects of AM PA. However, it is possible that thisconcentration was more effective since it blocks both voltage-gated Ca2+channels and calmodulin.How do AMPA and veratridine fulfill the requirements for the activation ofCaMKII and PKA? The factors leading to the activation of CaMKII, PKA and PKChave been well documented. CaMKII relies on the presence of both Ca2+andcalmodulin for its activation. In the presence of micromolar concentrations ofCa2+, calmodulin undergoes a conformational change exposing hydrophobicbinding sites which interact with a calmodulin-binding domain on the enzyme.Binding ofCa2/calmodulin is thought to unfold the molecule, thereby exposing112and activating the catalytic region (for review see Fujisawa, 1992; Hanson andSchulman, 1992).In contrast to CaM KIl, the activation of PKA is dependent on the production ofcAMP which usually results from the stimulation of adenylate cyclase followingthe activation of certain G-protein coupled receptors (for review Scott, 1991;Nairn et al., 1985). In the absence of cAMP, the enzyme consists of tworegulatory subunits bound to two catalytic subunits. The binding of cAMP to theregulatory subunits leads to the dissociation of the catalytic subunits, which arenow able to express phosphotransferase activity.Since agonist (AMPA) stimulation as well as chemically-induced increases indepolarization can lead to an increase in intracellular Ca2+ concentrations,such treatments could lead to the activation of CaMKII with concomitantchanges in[3H]-CNQX binding. An alternative mechanism of CaMKll activationhas been suggested by Fukanaga et al. (1992) and Tan et al. (1994) whoreported activation of CaMKll in response to increases in intracellular Ca2+ as aresult of NMDA receptor activation in cultured hippocampal neurons. Glutamateelevated theCa2-independent activity of CaMKII, an effect which could beblocked by the NMDA receptor antagonist AP-5.Although increases in intracellular Ca2 (reviewed by Nishizuka, 1988) as aresult of agonist stimulation or increases in cellular depolarization induced byveratridine could lead to the activation of PKC, the PKC inhibitorbisindolylmaleimide had no effect on the agonist- or veratridine-induceddecreases in[3H]-CNQX binding. These results are not surprising consideringthe lack of effect of TPA on3H]-CNQX binding (Chapter 3).113It remains unclear why the inhibiting peptide of a cAMP-dependent enzyme(PKA) and RpcAMPS would blockCa2-medi ted decreases in[3H]-CNQXbinding. These results are especially unexpected since Ca2+can lead to theactivation of phosphodiesterase, an enzyme which breaks down cAMP (Stryer,1988). However, several possibilities seem plausible:(1) It has been reported that the consensus sequence for phosphorylation byCaMKII is identical to a low affinity consensus sequence for phosphorylation byPKA (Kennely and Krebs, 1991). It is therefore possible that the effects of thePKA inhibiting peptide, an inhibitor whose mechanism is based on theconsensus sequence of PKA, are due to the partial inhibition of CaM Kil.Moreover, RpCAMPS at the concentration used in the present experiments (0.1mM) has been shown to bind competitively to the ATP binding site of proteinkinases (Pelech, personal communication) and may therefore act to inhibitCaMKll activity and associated decreases in[3H]-CNQX binding.(2) Agonist and depolarizing stimuli activate both PKA and CaM Ku through anunknown mechanism, leading to AMPA receptor regulation throughphosphorylation by both PKA and CaMKII. This mechanism, however, isunlikely to occur since both PKA and Cam Ku inhibitors are able toindeDendently block the effects of agonist and depolarizing stimuli.(3) Activation of AMPA receptors results in a Ca2 influx via voltage-gated Ca2channels leading to the activation of CaMKII. Activation of the latter representsthe initial step in a cascade of phosphorylation reactions ultimately leading tothe direct phosphorylation of AMPA receptors by PKA. However, this possibilityseems unlikely for two reasons. First, AMPA receptors (GIuR 1 - GIuR 4) do notappear to have consensus sequences for phosphorylation by PKA (Keinanen et114al., 1990; Boulter et al., 1990). Second, PKA does not appear to directlyphosphorylate AMPA receptors (Tan et al., 1994; McGlade-McCuIIoh et al.,1993).(4) Activation of AMPA receptors results in a Ca2 influx through voltage-gatedCa2 channels, leading to the activation of PKA through the activation ofadenylyl cyclase. In regard to the latter, it has recently been shown thatactivation of NMDA receptors results in an increase in cAMP viaCa2/calmodulin stimulation of adenylyl cyclase (Chetkovich and Sweatt,1993). The activation of PKA, in turn, results in a cascade of phosphorylationreactions, ultimately leading to the direct phosphorylation of AMPA receptors byCaMKII. This option appears more likely at present due to the presence ofstrong consensus sequences for phosphorylation by CaMKII on GIuR 1-4 aswell as the ability of CaMKll to directly phosphorylate the AMPA receptor(McGlade-McCulloh et al., 1993).(5) An alternative explanation would be that activation of AMPA receptorsresults in an influx of Ca2+via voltage-gated Ca2+ channels resulting in theactivation of both PKA and CaMKll through an unknown mechanism. The latterare both required for the activation of an unidentified AMPA receptor specifickinase, leading to the direct phosphorylation of the AMPA receptor population.A receptor specific kinase proposed above would be similar to the specificreceptor kinase described for the adrenergic receptors (Benovic et al., 1987)and postulated for the GABAA receptor population (Sweetnam et al., 1988).The 13-adrenergic kinase (Bark) is a cAMP-independent kinase that specificallyphosphorylates the agonist-occupied forms of the I3-adrenergic receptor.115Receptor phosphorylation by Bark promotes the binding of another protein, (3-arrestin, and this interaction appears to result in uncoupling of receptors and theassociated G-protein (Benovic et al., 1987). Evidence for a novel GABAAreceptor-associated kinase comes from results which showed GABAA receptorphosphorylation by a receptor-associated protein kinase in partially purifiedpreparations of GABAA receptors (Sweetnam et al., 1988). This novel, second-messenger-independent protein kinase may be involved in the functionalmodification of GABAA receptors.Although the exact order and constituents of the cascade of events leading toAMPA receptor regulation by agonist and depolarizing stimuli remain unknown,PKA and/or CaMKll seem to be involved in achieving such regulation. Theinvolvement of multiple kinases in AMPA receptor regulation does not appear tobe unique to this receptor population. For example, as discussed in theGeneral Introduction, an examination of the amino acid sequence of GABAAand nACh receptor subunits revealed the existence of consensus sequencestor a variety of protein kinases, including PKA, PKC, and a protein tyrosinekinase (Swope et al., 1992). For GABAA receptors, phosphorylation by any oneof these kinases resulted in a decrease in GABAA receptor-mediated currents(Moss et al., 1992). Similarly, for nACh receptors, phosphorylation by PKA,PKC, or a protein tyrosine kinase led to an increase in the rapid rate ofdesensitization (Hopfield et al., 1988; Huganir et al., 1986). Phosphorylation ofGABAA and nACh receptors can thus be regulated by several secondmessenger systems, allowing for the heterologous regulation of these receptorpopulations. Although AMPA receptors also show consensus sequences for avariety of protein kinases, the regulation of this receptor population by agonist116and depolarizing stimuli may be different in that it requires more than oneprotein kinase.There are two major implications of the present results. First, the data providefurther evidence for the involvement of Ca2 in the regulation of AMPAreceptors by phosphorylation. Although it remains unknown whether Ca2directly activates the kinase(s) involved in AMPA receptor regulation, such aprocess may reveal a crucial step in the cascade of events leading to AMPAreceptor regulation. Such a cascade may begin with neurotransmitter bindingto AMPA receptors leading to an alteration in ionic current through channelsassociated with the receptors and voltage-gated Ca2+ channels withsubsequent changes in intracellular Ca2+ concentrations. Changes inintracellular Ca2+ concentrations may then be the factors triggering specifickinase(s), ultimately leading to the regulation of the receptors byphosphorylation.At a practical level, the above results suggest that the medium in which receptorbinding studies are carried out can alter radioligand binding by stimulation ofendogenous kinases. These findings make it imperative to strictly control theconditions under which such binding studies are performed. Attempts to modifythe media by the additions of various ions might inadvertently lead to ionmediated enzyme action and result in changes in binding levels. The likelihoodthat such variations will occur necessitates a careful examination of ionicdependence in each separate case.117GENERAL DISCUSSIONSummary of Data(1) A[3H]-CNQX binding site in rat cortex was studied and found to satisfy thebasic criteria for receptor characterization (Boulton et al., 1985), includingsteady-state binding, competition by specific AMPA analogues, as well assaturable binding. These data allowed the description of this[3H]-CNQXbinding site as an AMPA receptor (Chapter 1).(2) AMPA or v+g treatment of cortical slices resulted in reversible decreases in[3H]-CNQX binding (Chapter 2).(3) Activation of endogenous CaMKII and treatment with the catalytic subunit ofPKA also led to reversible decreases in[3H]-CNQX binding. These enzyme-induced alterations in[3H]-CNQX binding resembled the decreases in [3H]-CNQX binding observed in response to AMPA and v+g treatment, bothqualitatively and quantitatively (Chapter 3).(4) AMPA receptor regulation induced by AMPA and v+g could be completelyblocked by CaMKll and PKA inhibitors. Loperamide, a non-specific inhibitor ofvoltage-gated Ca2+ channels, was also able to abolish the regulatory effects ofthese stimuli. These results suggested that CaMKII, PKA, as well as an influx ofCa2through voltage-gated Ca2 channels are involved in the regulation ofAMPA receptors by agonist and depolarizing stimuli (Chapter 4).118(5) Calcium ions alone were sufficient to produce significant, concentration-dependent decreases in[3H]-CNQX binding over a concentration range of 0.1-1 mM. These decreases in binding could be blocked by CaMKII and PKAinhibitors. Although it remains unclear why the inhibiting peptide of a cAMP-dependent enzyme (PKA) would blockCa2-medi ted decreases in [3H]-CNQX binding, several possibilities are discussed in Chapter 5.These data have provided direct measurements of alterations in the agonistbinding site of the AMPA receptor. Although cellular current responsesfollowing activity-dependent regulation of AMPA receptors have not beenevaluated, it is probable that a decrease in[3H]-CNQX binding reflects adiminution of AMPA receptor-mediated currents. Decreases in AMPA receptornumber and/or affinity should offer fewer targets to endogenousneurotransmitter, leading to a decrease in postsynaptic currents to the samelevel of presynaptic stimulation. With regard to the latter, Shahi and Baudry(1992) have reported that increasing binding affinity of agonists to AMPAreceptors increases synaptic responses at glutamatergic synapses.A Model of AMPA ReceDtor RegulationThe results described above suggest a cascade of events leading to AMPAreceptor regulation. Such a cascade may begin with cellular depolarizationthrough ion channels associated with the AMPA receptor population and/orvoltage-gated ion channels and subsequent changes in intracellular Ca2+concentrations. Changes in subsynaptic Ca2+ concentrations may then be thefactors triggering specific kinases, ultimately leading to AMPA receptorregulation by phosphorylation. Alterations in receptor binding may, in turn, alter119the neural response to subsequent neurotransmitter release. A cascade of thistype proposed here may prove to be a general mechanism underlyingionotropic receptor regulation.A heterologous mechanism may also be proposed in which regulation of AMPAreceptors occurs through the activation of a G-protein coupled or anotherionotropic receptor population. Activation of a G-protein coupled receptorpopulation may lead to the production of cAMP in turn resulting in the activationof PKA and concomitant regulation of AMPA receptors. Alternatively, activationof another ionotropic receptor population (e.g. NMDA) may lead to an alterationin ionic current through receptor associated ion channels, or, alternatively,through the activation of voltage-gated ion channels.AMPA Receptor Regulation: lmDlications for Neural FunctionThe consequences of AMPA receptor regulation may have importantimplications for many aspects of neural function. First, alterations in functionalAMPA receptor levels may play an integral role in the control of normal synapticneurotransmission. Second, AMPA receptor regulation may be a keycomponent for the maintenance of some forms of synaptic neuroplasticity.Finally, AMPA receptor regulation in response to a neurotoxic environment mayreduce the neurotoxic damage and cell death associated with certainneurodegenerative disorders.120AMPA ReceDtor Regulation in Normal Svnatic NeurotransmissionThe regulation of AM PA, and other ionotropic receptors, in response todepolarizing stimuli may be a feature in the control of normal synaptictransmission. Alterations in functional receptor levels and responses maydetermine the level of response to subsequent neurotransmitter release. This,in turn, may account for transient changes in the efficacy of synapticneurotransmission and could constitute a possible homeostatic mechanism inthe control of interneuronal communication.AMPA ReceDtor Regulation in Svnatic NeuroplasticityDepolarizing stimuli may, under certain conditions, result in long-lastingalterations in functional AMPA receptor levels, thus resulting in a permanentalteration of the level of response to subsequent neurotransmitter release.Changes in AMPA receptor number have been postulated to provide the basisfor some forms of synaptic plasticity, including LTP (Maren et al., 1993; Tocco etal., 1992) and cerebellar LTD (Linden, 1994). The stages in AMPA receptorregulation may provide more details concerning such phenomena. Thus thepresent model would predict that for plastic modifications of function to occurAMPA receptors controlling particular ionic currents must be present. Second,regulatory enzymes, e.g. kinases, must be present. The action of such enzymeswill, in some cases, be controlled by specific ions and ionic concentrations.Ionic currents following neurotransmitter activation of receptors will, in turn, actto turn on regulatory enzymes which will then lead to receptor regulation byphosphorylation. The modification in[3H]-CNQX binding in such cases mayprovide for a LTD-like phenomenon similar to the one observed between121parallel fibres and Purkinje cells in the cerebellum (reviewed by Linden et al.,1994) (see General Introduction). LTD is thought to be mediated entirely by adecrease in the number or sensitivity of postsynaptic AMPA receptors as aresult of protein kinase activity (Linden, 1994; Linden et al., 1993; 1991).ReceDtor Regulation and NeuroøathologyIn many neurologic disorders, including ischemia, epilepsy, as well as morechronic neurodegenerative states such as amyotrophic lateral sclerosis,neurolathyrism, and Huntington’s disease, neuronal damage has beenpostulated to arise through an overstimulation of excitatory amino acidreceptors by glutamate or aspartate (reviewed by Lipton and Rosenberg, 1994).Overstimulation of glutamate receptors results in increased intracellular Ca2+levels which, in turn, have been shown to lead to cell death through theactivation of proteases, nitric oxide synthase,as well as PKC and CaMKII(reviewed by Lipton and Rosenberg, 1994).A decrease in[3H]-CNQX binding in response to an overstimulation of AMPA orother ionotropic receptors may be viewed as a neuroprotective mechanism.Decreases in functional AMPA receptor binding should potentially lead todecreased Ca2+ influx and associated CaM Kll activity. Moreover, it may provepossible to deliberately manipulate elements in the AMPA receptor regulationcascade in order to control increased Ca2+ influx and CaMKII activation, thusreducing neurotoxic damage and cell death associated with certainneurodegenerative disorders.Concluding RemarksIn conclusion, AMPA receptor regulation may be involved in many aspects ofneuronal function, including normal synaptic neurotransm ission, synapticneuroplasticity, and some forms of neuropathology. A better understanding ofthe molecular mechanisms involved in the regulation of AMPA receptors andionotropic receptors in general may provide a framework for studies of synapticfunction and dysfunction.122123FUTURE DIRECTIONS(1) Electrophysiological experiments examining AMPA receptor-mediatedcurrents in response to treatment with AMPA, v+g, or activators of CaMKII andPKA could be carried out to determine functional correlates of decreases in[3H]-CNQX binding. These experiments would allow a direct comparisonbetween changes in[3H]-CNQX binding and alterations in AMPA receptor-mediated current responses.(2) To determine whether treatment with AMPA, v+g, an activating cocktail ofCaMKll, or the catalytic subunit of PKA leads to a change in the phosphorylationstate of the AMPA receptor population, studies using antibodies directedagainst GIuR 1-4, phosphoserine, phosphothreonine, and phosphotyrosineresidues to probe Western blots of treated and untreated rat cortical slices couldbe carried out. These experiments would determine whether changes in thephosphorylation state of AMPA receptors correlate with changes in[3HJ-CNQXbinding.(3) CaM Kil and PKA activity assays could be carried out to allow thedetermination of CaMKll and PKA phosphotransferase activity as a result ofAMPA and veratridine treatment. Furthermore, CaM Ku and PKAphosphotransferase activity could be measured following treatment with AMPAor v+g inhibitors of CaM KIl, PKA, and voltage-gated Ca2 channels.Addition of these inhibitors should abolish any increase in CaM Kil and PKAphosphotransferase activity induced by AMPA or v+g.124(4) To further investigate the role of Ca2 and voltage-gated Ca2 channels inAMPA receptor regulation, AMPA- and v-i-g-induced receptor regulation couldbe studied in a Ca2-free medium. The inability of AMPA and v+g to induceAMPA receptor regulation in such a medium would provide further evidencethat the influx of Ca2 is an essential step leading to AMPA receptor regulation.Furthermore, more specific Ca2+ channel blockers could be used to inhibitAMPA receptor regulation. This would allow the determination of the specificvoltage-gated Ca2 channel(s) involved in AMPA receptor regulation by AMPAand v+g.125REFERENCESAniksztejn, L., Kleschevnikov, A., Ben-An, Y., Represa, A. (1992). Carbacholpotentiates NMDA current via protein kinase C in rat hippocampalslices. Soc. Neurosci. Abstr. 18, 256.Baudry, M. and Lynch, G. (1979). Regulation of glutamate receptors by cations.Nature 279, 748-750.Bennet, J.P. (1978). Methods in binding studies in neurotransmitter receptorbinding. Yamamura, H.l., Enn, S.J., and Kuhar, M.J., eds. New York:Raven Press, 57-90.Bennett, G.W., Sharp, T., Marsden, C.A. and Parker, T.L. (1983). A manually-operated brain tissue slicer suitable for neurotransmitter release studies.J. Neurosci. Methods 7, 107-115.Benovic, J.L., Kuhn, H., Weyand, I., Codina, J., Cron, M.G. and Letkowitz, R.J.(1987). Functional desensitization of the isolated 13-adrenergic receptorby the 13-adrenergic receptor kinase: potential role of an analog for theretinal protein arrestin (48-kDa protein). Proc. Natl. Acad. Sci. USA 84,8879-8882.Bird, S.J. and Aghajanian, G.K. (1975). Denervation supersensitivity in thecholinergic septohippocampal pathway: a microiontophoretic study.Brain Res. 100, 355-370.Bliss, T.V.P. and Collingridge, G.L. (1993). Synaptic model of memory: longterm potentiation in the hippocampus. Nature 361, 31-39.126Boulter, J., Hollmann, M., O’Shea-Greenfield, A., Hartley, M., Deneris, E., Maron,C. and Heinemann, S. (1990). Molecular cloning and functionalexpression of glutamate receptor subunit genes. Science 249, 1033-1037.Boulton, A.A., Baker, G.B., Hrdina, P.D., editors. (1985). General principles ofreceptor binding. Neuromethods 4. Clifton, NJ: Humanan Press, 2-22.Bridges, R.J., Stevens D.R., Kahle, J.S., Nunn, P.B., Kadri, M. and Cotman, C.W.(1989). Structure-function studies on N-oxalyl-diamino-dicarboxylicacids and excitatory amino-acid receptors-evidence that Beta-L-ODAP isa selective non-NMDA agonist. J. Neurosci. 9 (6), 2073-2079.Brorson, J.R., Manzolillo, P.A. and Miller, R. J. (1994). Ca2entry viaAMPA/KA receptors and excitotoxicity in cultured cerebellar Purkinjecells. J. Neurosci. 14, 187-197.Browning, M.D., Bureau, M., Dudek, E.M. and Olsen, R.W. (1990). Proteinkinase C and cAMP-dependent protein kinase phosphorylate the 13subunit of the purified y-aminobutyric acid-A receptor. Proc. Nati. Acad.Sd. USA 87, 1315-131 8.Burt, D. (1980). Basic receptor methods II. Problems of interpretation in bindingstudies. In: Short course syllabus: receptor binding techniques.Cincinnati, Ohio: Soc. Neurosci. , 53-69.127Bylund, D.B. (1980). Analysis of receptor binding data. In: Short CourseSyllabus: Receptor Binding Techniques. Cincinnati, Ohio: Soc.Neurosci., 70-99.Bylund, D.B. (1992). A practical guide for receptor binding. Biotech Update 5,188.Bylund, D.B. and Yamamura, H.l. (1990). Methods for receptor binding. In:Methods in neurotransmitter receptor analysis. New York: Raven Press,1-35.Campbell, P.T., Hnatowich, M., O’Dowd, B.F., Caron, M.G., Lefkowitz, R.J. andHausdorff, W.P. (1991). Mutations of the human 132-adrenergic receptorthat impair coupling to G interfere with receptor down-regulation but notsequestration. Molec. Pharmacol. 39, 192-198.Carpenter, M.B. Core text of Neuroanatomy. (1988). Baltimore: Williams &Wilkins, 390-433.Catterall, W.A. (1980). Neurotoxins that act on voltage-sensitive sodiumchannels in excitable membranes. Annu. Rev. Pharmacol. Toxicol. 20,15-43.Chen, L. and Huang, L.Y. (1992). Protein kinase C reduces Mg2+ block ofNMDA-receptor channel as a mechanism of modulation. Nature 356,521 -523.128Chetkovich, D.M. and Sweatt, J.D. (1993). NMDA receptor activation increasescyclic AMP in area CAl of the hippocampus via calcium/calmodulinstimulation of adenylyl cyclase. J. Neurochem. 61(5), 1933-42.Church, J., Abel-Hamid, E.J. and MacDonald, J.F. (1994). Loperamide blockshigh voltage-activated calcium channels and N-methyl- D-aspartateevoked responses in rat and mouse cultured hippocampal pyramidalneurons. Molec. PharmacoL, 45, 747-757.Cochet, C., Souvignet, C., Keramidas, M. and Chambaz, E.M. (1986). Alteredcatalytic properties of protein kinase C in phorbol ester treated cells.Biochem. Biophys. Res. Comm. 134, 1031-1037.Collingridge, G.L. and Singer, W. (1991). Excitatory amino acid receptors andsynaptic plasticity. Trends in Pharmacol. Sd. Special Report, 42-48.Collins, S., Caron, M.G. and Lefkowitz, R.J. (1991). Regulation of adrenergicreceptor responsiveness through modulation of receptor geneexpression. Annu. Rev. Physiol. 53, 497-508.Cox, J.A., Lysko, P.G. and Henneberg, R.C. (1989). Amino acid neurotoxicity atthe N-methyl-D-aspartate receptor in cultured neurons: role of voltagedependent magnesium block. Brain Res. 499, 267-272.Ebert, B., Madsen, U., Lund, T.M., Lenz, S.M., and Krogsgaard-Larson, P.(1994). Molecular pharmacology of the AMPA agonist, (S)-2-amino-3-(3-hydroxy-5-phenyl-4-isoxazolyl) propionic acid [(S)-APPA] and the AM PAantagonist, (R)-APPA. Neurochem. lnt. 24, 507-51 5.129Eccles, J.C. (1976). From electrical to chemical transmission in the centralnervous system. The closing address of the Sir Henry Dale CentennialSymposium. Notes Rec. R. Soc. Lond. 30, 21 9-230.Edelman, A.M., Blumenthal, D.K. and Krebs, E.G. (1987a). Proteinserine/threonine kinases. Annu. Rev. Biochem. 56, 567-613.Edelman, G.M., Gall, W.E. and Cowan, W.M., eds. (1987b). Synaptic Function.New York: Wiley.Egebjerg, J., Bettler, B., Hermans-Borgmeyer, I. and Heinemann, S. (1991).Cloning of a cDNA for a glutamate receptor subunit activated bykainate but not AM PA. Nature 351, 745-748.Finkbeiner, S. and Stevens, C.F. (1988). Applications of quantitativemeasurements for assessing glutamate neurotoxicity. Neurobiology 85,4071-4074.Forn, J. and Greengard, P. (1978). Depolarizing agents and cyclic nucleotidesregulate the phosphorylation of specific neuronal proteins in rat cerebralcortex slices. Proc. Natl. Acad. Sci. USA 75 (10), 51 95-51 99.Frandsen, A. and Schousboe, A. (1987). Time and concentration dependencyof the toxicity of excitatory amino acids on cerebral neurons in primaryculture. Neurochem. lnt. 10, 583-591.Fujisawa, H. (1990). Calmodulin-dependent protein kinase II. BioEssays 12,27-29.130Fukanaga, K., Soderling, T.R. and Miyamoto, E. (1992). Activation ofCa2/calmoduIin-dependent protein kinase II and protein kinase C byglutamate in cultured rat hippocampal neurons. J. Biol. Chem. 31,22527-22533.Gerber, G., Kangrga, I., Ryu, P.D., Larew, J.S.A. and Randic, M. (1989).Multiple effects of phorbol esters in the rat spinal dorsal horn. J.Neurosci. 9, 3606-61 7.Gordon, A.S., Davis, C.G. and Diamond, I. (1977). Phosphorylation ofmembrane proteins at a cholinergic synapse. Proc. NatI. Acad. Sci.USA 74, 263-267.Greengard, P., Valtorta,F., Czernik, A.J. and Benfenati, F. (1993). Synapticvesicle phosphoproteins and regulation of synaptic function. Science259, 780-785.Greengard, P., Jen, J., Nairn, A.C.and Stevens, C.F. (1991). Enhancement ofthe glutamate response by cAMP-dependent protein kinase inhippocampal neurons. Science 253, 1135-1138.Guiffrida, R. and Rustoni, A. (1988). Glutamate and aspartate immunoreactivityin corticothalamic neurons of rats. In: Cellular Thalamic Mechanisms. M.Bentivlio and R. Spreafico, eds. Amsterdam: Elsevier, 311-320.Gulifrida, R. and Rustoni, A. (1989). Glutamate and aspartate immunoreactivityin corticospinal neurons of rat. J. Comp. Neurol. 288, 154-164.131Hablitz, J.J., Tehrani, M.H.J. and Barnes, Jr., E.M. (1989). Chronic exposure ofdeveloping cortical neurons to GABA down-regulatesGABAlbenzodiazepine receptors and GABA-gated chloride currents.Brain Res. 501, 332-338.Hahn, J.S., Aizenman, E. and Lipton, S.A. (1988). Central mammalian neuronsnormally resistant to glutamate toxicity are made sensitive by elevatedextracellular Ca2: toxicity is blocked by the N-methyl-D-aspartateantagonist MK-801. Neurobiology 85, 6556-6560.Hall, R.A., Kessler, M. and Lynch, G. (1992). Evidence that high- and low-affinityDL- -amino-3-hyd roxy-5-methylisoxazole-4-propionic acid (AM PA)binding sites reflect membrane-dependent states of a single receptor. J.Neurochem. 59, 1997-2004.Hanson, P.1. and Schulman, H. (1992). NeuronalCa2/calmodulin-dependentprotein kinases. Annu. Rev. Biochem. 61, 559-601.Hausdorff, W.P., Caron, M.G. and Lefkowitz, R.J. (1990). Turning off the signal:desensitization of 13-adrenergic receptor function. FASEB J. 4, 2881-2889.Hollmann, M., Hartley, M. and Heinemann, S. (1991). Ca2 permeability of KAAMPA-gated glutamate receptor channels depends on subunitcomposition. Science 252, 851 -853.Honore, T., Drejer, J., Nielsen, E.O., and Nielsen, M. (1989). Non-NMDAglutamate receptor antagonist3H-CNQX binds with equal affinity to twoagonist states of quisqualate receptors. Biochem. Pharmacol. 38, 3207-3212.132Honore, T., Davies, S.N., Drejer, J., Fletcher, E.J., Jacobsen, P., Lodge, D. andNielsen, F.E. (1988). Quinoxalinediones: potent competitive non-NMDAglutamate receptor antagonists. Science 241, 701-703.Honore, T., Lauridsen, J. and Krogsgaard-Larsen, P. (1982). The binding of[3H]AMPA, a structural analogue of glutamic acid, to rat brainmembranes. J. Neurochem. 38, 173-178.Hopfield, J.F., Tank, D.W., Greengard, P. and Huganir, R.L. (1988). Functionalmodulation of the nicotinic acetyicholine receptor by tyrosinephosphorylation. Nature 336, 677-680.Huganir, R.L., Delcour, A.H., Greengard, P. and Hess, G.P. (1986).Phosphorylation of the nicotinic acetylcholine receptor regulates its rateof desensitization. Nature 321, 774-776.Huganir, R.L. and Greengard, P. (1990). Regulation of neurotransmitterreceptor desensitization by protein phosphorylation. Neuron 5, 55-67.Hunter, C., Wheaton, K.D. and Wenthold, R.J. (1990). Solubilization and partialpurification of -amino-3-hydroxy-5-methyl-4-isoxazolepropionic acidbinding sites from rat brain. J. Neurochem. 54, 118-125.Hunter, C. and Wenthold, R. (1992). Solubilization and purification of an-amino-3-hydroxy-5-methylisoxazole-4-propionic acid binding proteinfrom bovine brain. J. Neurosci. 4, 1379-1385.133lshii, A., Kiuchi, K., Kobayashi, R., Sumi, M., Hidaka, H. and Nagatsu, T. (1991).A selectiveCa2/caImoduIin-dependent protein kinase II inhibitor, KN62, inhibits the enhanced phosphorylation and the activation of tyrosinehydroxylase by 56 mM K in rat pheochromocytoma PC12 cells.Biochem. 176, 1051-1056.Ito, M. and Karachot, L. (1992). Protein kinases and phosphatase inhibitorsmediating long-term desensitization of glutamate receptors in cerebellarPurkinje cells. Neurosci. Res. 14, 27-38.Kachur, J.F., Morgan, D.W., and Gaginella, T.S. (1986). Effects ofdextromethorphan on guinea pig ileal contractility in vitro: comparisonwith levomethorphan, loperamide and codeine. J. Pharmacol. Exp. Ther.239, 661 -667.Kano, M. and Konnerth, A. (1992). Potentiation of GABA-mediated currents bycAMP-dependent protein kinase. Neuroreport 3, 563-566.Keinanen, K., Wisden, W., Sommer, B., Werner, P., Herb, A., Verdoorn, T.A.,Sakmann, B. and Seeburg, RH. (1990). A family of AMPA-selectiveglutamate receptors. Science 249, 556-560.Keler, T. and Smith, C.A.D. (1989). Effects on growth and energy metabolism inuntransformed and transformed BALB/c mouse fibroblasts by a novelcytotoxic compound. Cancer Research 49, 7093-7097.Keller, B.U., Hollmann, M., Heinemann, S. and Konnerth, A. (1992). Calciuminflux through subunits GluRl/GIuR3 of kainate/AMPA receptor channelsis regulated by cAMP dependent protein kinase. EMBO J. 11, 891-896.134Kelso, S.R., Nelson, T.E. and Leonard, J.P. (1992). Protein kinase C-mediatedenhancement of NMDA currents by metabotropic glutamate receptors inXenopus oocytes. J. Physiol. 449, 705-718.Kennelly, P.J. and Krebs, E.G. (1991). Consensus sequences as substratespecificity determinants for protein kinases and protein phosphatases. J.Biol. Chem. 266, 15555-15558.Kirkness, E.F., Bovenkerk, C.F., Ueda, T. and Turner, A.J. (1989).Phosphorylation of y-aminobutyrate (GABA)/benzodiazepine receptorsby cyclic AMP-dependent protein kinase. Biochem. J. 29, 613-61 6.Kirkwood, A., Dudek, S.M., Gold, J.T., Aizenman, C.D. and Bear, M.F. (1993).Common forms of synaptic plasticity in the hippocampus and neocortexin vitro. Science 260, 151 8-1521.Kitamura, Y., Miyazaki, A., Yamanaka, Y. and Nomura, Y. (1993).Stimulatory effects of protein kinase C and calmodulin kinase II on Nmethyl-D-aspartate receptor/channels in the postsynaptic density of ratbrain. J. Neurochem. 61, 100-109.Klein, W.L., Nathanson, N. and Nirenberg, M. (1979). Muscarinic acetylcholinereceptor regulation by accelerated rate of receptor loss. Biochem. &Biophys. Res. Comm. 90, 506-51 2.Koh, J-Y., Goldberg, M.P., Hartley, D.M. and Choi, D.W. (1990). Non-NMDAreceptor-mediated neurotoxicity in cortical culture. J. Neurosci. 10 (2),693-705.135Kraft, A.S., Anderson, W.B. Cooper, H.L. and Sando, J.J. (1982). Decrease incytosolic calcium/phospholipid-dependent protein kinase activityfollowing phorbol ester treatment of EL4 thymoma cells. J. Biol. Chem.257, 13193-13196.Krieg, W.J.S. (1946). Connections of the cerebral cortex. The albino rat. Atopography of the cortical areas. J. Comp. Neurol. 84, 221 -275.Krieger, C., Wagey, R., Lanius, R.A., and Shaw, C.A. (1993). Activation of PKCreverses apparent NMDA receptor reduction in ALS. NeuroReport 4,931-934.Krishek, B.J. , Blackstone, C. and Huganir, R.L. (1994). Regulation of GABAAreceptor function by protein kinase C phosphorylation. Neuron 12 (5),1081-1096.Lai, C. and Lemke, G. (1991). An extended family of protein-tyrosine kinasegenes differentially expressed in the vertebrate nervous system. Neuron6, 691-704.Lanius, R.A. and Shaw, C. (1992). Characterization and regulation of a highaffinity[3H]-CNQX labeled AMPA receptor in rat neocortex. Molec. BrainRes. 15, 256-262.Lanius, R.A., Pasqualotto, B.A. and Shaw, C.A. (1993a). Age-dependentexpression, phosphorylation and function of neurotransm itter receptors:pharmacological implications. Trends in Pharmacol. Sci. 14(1 1), 403-408.136Lanius, R.A., Pasqualotto, B.A. and Shaw, C.A. (1993b). A novel mechanism ofAMPA receptor regulation: lonically triggered kinases andphosphatases. NeuroReport 4, 795-798.Lanius, R.A. and Shaw, C.A. (1993). High-affinity kainate binding sites inliving slices of rat neocortex: Characterization and regulation. Neurosci.55 (1), 139-145.Leidenheimer, N.J., Browning, M.D. and Harris, R.A. (1991). GABAA receptorphosphorylation: multiple sites actions and artifacts. TrendsPharmacol. Sci. 12, 84-86.Liles, W.C. and Nathanson, N.M. (1987). Regulation of muscarinicacetyicholine receptor number in cultured neuronal cells by chronicmembrane depolarization. J. Neurosci. 7(8), 2556-2563.Liman, E.R., Knapp, A.G. and Dowling, J.E. (1989). Enhancement of kainategated currents in retinal horizontal cells by cyclic AMP-dependent proteinkinase. Brain Res. 481, 399-402.Linden, D.J. (1994). Long-term synaptic depression in the mammalian brain.Neuron 12, 457-472.Linden, D.J., Dickinson, M.H., Smeyne, M. and Conner, J.A. (1991). A longterm depression of AMPA currents in cultured cerebellar Purkinjeneurons. Neuron 7, 81-89.137Linden, D.J., Smeyne, M. and Connor, J.A. (1993). Induction of cerebellarlong-term depression in culture requires postsynaptic action of sodiumions. Neuron 11, 1093-1100.Linden, S.J. and Connor, J.A. (1993). Cellular mechanisms of long-termdepression in the cerebellum Curr. Opin. Neurobiol. 3, 401 -406.Lipton, S.A. and Rosenberg,P.A. (1994). Excitatory amino acids as a finalcommon pathway for neurologic disorders. New Eng. J. of Med. 330,613-622.Liron, Z., Wong, E. and Roberts, E. (1988). Studies on uptake of y-aminobutyricacid by mouse brain particles; toward the development of a model. BrainRes. 444, 119-132.Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951). Proteinmeasurement with the Folin phenol reagent. J. Biol Chem. 193, 265-275.Luqmani, Y.A., Bradford, H.F., Birdsall, N.J. and Hulme, E.C. (1979).Depolarization-induced changes in muscarinic cholinergic receptors insynaptosomes. Nature 277, 481-482.Lysko, P.G., Cox, J.A., Vigano, A. and Henneberry, R.C. (1989). Excitatoryamino acid neurotoxicity at the N-methyl-D-aspartate receptor in culturedneurons: pharmacological characterization. Brain Res. 499, 258-266.138Maloteaux, J.-M. , Octave, J.-N., Gossuin, A., Laterre, C. and Trouet, A. (1987).GABA induces down-regulation of the benzodiazepine-GABA receptorcomplex in the rat cultured neurons. Eur. J. Pharmacol. 144, 173-1 83.Maren, S., Tocco, G., Standley, S., Baudry, M. and Thompson, R.F. (1993).Postsynaptic factors in the expression of long-term potentiation (LTP):increased glutamate receptor binding following LTP induction in vivo.Proc. Nati. Acad. Sd. USA 90, 9654-9658.Martin, L.J., Blackstone, C.D., Levey, A.l., Huganir, R.L. and Price, D.L. (1993).AMPA Glutamate receptor subunits are differentially distributed in ratbrain. Neurosci. 53, 327-358.McGlade-McCulloh, E., Yamamoto, H., Tan, S.E. and Brickey, D.A. (1993).Phosphorylation and regulation of glutamate receptors bycalciumlcalmoldulin-dependent protein kinase II. Nature 362, 640-642.Mcllwain, H., Buchel, L. and Chesire, J.D. (1951). The inorganic phosphate andphosphocreatine of brain especially during metabolism in vitro.Biochem. J. 48, 12-20.Mehta, A.K. and Ticku, M.K. (1992). Chronic GABA exposure down-regulatesGABA-benzodiazepine receptor-lonophore corn plex in cultured cerebralcortical neurons. Mol. Brain Res. 16, 29-36.Meldrum, B. and Garthwaite, J. (1991). Excitatory amino acid neurotoxicity andneurodegenerative disease. Trends in Pharmacol. Sci. Special Report,54-62.139Monaghan, D.T., Bridges, R.J. and Cotman, C.W. (1989). The excitatory aminoacid receptors: their classes, pharmacology, and distinct properties inthe function of the central nervous system. Annu. Rev. Pharmacol.Toxicol. 29, 365-402.Moss, S.J., Smart, T.G., Blackstone, C.D. and Huganir, R.L. (1992). Functionalmodulation of GABAA receptors by cAMP-dependent proteinphosphorylation. Science 257, 661-665.Nairn, A.C., Hemmings, Jr., H.C. and Greengard, P. (1985). Protein Kinases inthe brain. Ann. Rev. Biochem. 54, 931 -76.Nielsen, E.O., Drejer, J., Cha, J-H J., Young, A.B. and Honore, T. (1990).Autoradiographic characterization and localization of quisqualatebinding sites in rat brain using the antagonist[3H]6-7-nitroquinoxaline-2, 3-dione: comparison with (R , S)-[3H] -am ino-3-hydroxy-5-methyl-4-isoxazolepropionic acid binding sites. J.Neurochem. 54, 686-694.Nishizuka, Y. (1988). The molecular heterogeneity of protein kinase C and itsimplications for cellular regulation. Nature 334, 661 -665.Olsen, R.W., Szamraj, 0. and Houser, C.R. (1987).[3H]AMPA binding toglutamate receptor subpopulations in rat brain. Brain Res. 402, 243-254.Ortega, A. and Teichberg, V.1. (1990). Phosphorylation of the 49-kDa putativesubunit of the chick cerebellar kainate receptor and its regulation bykainatergic ligands. J. Bid. Chem. 265, 21404-21406.140Pasqualotto, B.A., Lanius, R.A. and Shaw, C.A. (1994). Heterologousregulation of cortical AMPA receptors through voltage-gated Ca2channels. Soc. Neurosci. Abstract, 210.11.Patneau, D.K., Wright, P.W., Winters, C., Mayer, M.L., and Gallo, V. (1994).Glial cells of the oligodendrocyte lineage express both kainate- andAMPA -preferring subtypes of glutamate receptor. Neuron 12, 357-371.Petralia, R.S. and Wenthold, R.J. (1992). Light and electronimmunocytochemical localization of AMPA-selective glutamate receptorsin the rat brain. J. Comp. Neurol. 318, 329-354.Pelech, S.L., Meier, K.E. and Krebs, E.G. (1986). Rapid microassay for proteinkinase C translocation in Swiss 3T3 cells. J. Biol. Chem. 25, 8348-8353.Pelech, S.L., Samiei, M., Charest, D.L., Howard, S.L. and Salari, H. (1991).Characterization of calcium-independent forms of protein kinase C inphorbol ester-treated rabbit platelets. J. Biol. Chem. 266, 8696-8705.Porter, N.M., Twyman, R.E., Uhler, M.D. and MacDonald, R.L. (1990). CyclicAMP-dependent protein kinase decreases GABAA receptor current inmouse spinal neurons. Neuron 5, 789-796.Rakic, P. (1988). Specification of cerebral cortical areas. Science 241, 170-176.Raymond, L.A., Blackstone, C.D. and Huganir, R.L. (1993a). Phosphorylation ofamino acid neurotransmitter receptors in synaptic plasticity. Trends inNeurosci. 16, 147-153.141Raymond, L.A., Blackstone, C.D. and Huganir, R.L. (1993b). Phosphorylationand modulation of recombinant GIuR6 glutamate receptors by a cAMP-dependent protein kinases. Nature 361, 637-641.Reid, K.H., Edmonds, Jr., H.L., Schurr, A., Tseng, M.T. and West, C.A. (1988).Pitfalls in the use of brain slices. Progress in Neurobiology 31, 1-18.Reymann, K.G., Davies, S.N., Matthies, H., Kase, H. and Collingridge, G.L.(1990). Activation of a K-252b-sensitive protein kinase is necessary for apost-synaptic phase of long-term potentiation in area CAl of rathippocampus. European J. Neurosci. 2, 481 -486.Ruiz-Gomez, A., Vaello, M.L., Valdivieso, F. and Mayor, F. (1991).Phosphorylation of the 48k Da subunit of the glycine receptor by proteinkinase C. J. Biol. Chem. 266, 559-565.Saitoh, T. and Changeux, J.P. (1980). Phosphorylation in vitro of membranefragments of Torpedo mamorata electric organ. Eur. J. Biochem. 105,51-62.Scott, J.D. (1991). Cyclic nucleotide-dependent protein kinases. Pharmacol.Therapeutics 50, 123-145.Seeburg, P.H. (1993). The TINS/TiPS Lecture: The molecular biology ofmammalian glutamate receptor channels. Trends in Neurosci. 16, 359-364.142Shaw, C.A. and Lanius, R.A. (1992). Reversible kinase and phosphataseregulation of brain amino acid receptors in postnatal development. Dev.Brain Res. 70, 153-156.Shaw, C., Lanius, R.A., Scarth, B.A. and Wilkinson, M. (1992a). The “living”brain slice technique: utility in radioligand receptor characterization andregulation studies. Biotech Update 7, 165-170.Shaw, C., Pasqualotto, B.A. and Lanius, R.A. (1992b). A role forkinase/phosphatase action in the regulation of brain amino acidreceptors. Molec. Neuropharmacol. 2, 297-302.Shaw, C., Lanius, R.A. and Van den Doel, K. (1994). The origin of synapticneuroplasticity: crucial molecules or a dynamical cascade? Brain Res.Rev., in press.Shaw, C. and Scarth, B.A. (1991). Characterization and differential regulationor GABAA and benzodiazepine receptors in rat neocortex. Mol. BrainRes. 11, 273-282.Shaw, C.A. and Wilkinson, M. (1994). Receptor characterization and regulationin intact tissue preparations: pharmacological implications. Biochem.Pharmacol., in press.Shaw, C., Van Huizen, F., Cynader, M.S., Wilkinson, M. (1989). A role forpotassium channels in the regulation of cortical muscarinic acetylcholinereceptors in an in vitro slice preparation. Mol. Brain Res. 5, 71-83.143Sheardown, M.J., Nielsen, E.O., Hansen, A.J., Jacobsen, P. and Honore, T.(1990). 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo (F) quinoxaline: aneuroprotectant for cerebral ischemia. Science 247, 571 -574.Sieghart, W., Forn, J. and Greengard, P. (1979). Ca2 and cyclic AMP regulatephosphorylation of same two membrane-associated proteins specific tonerve tissue. Proc. Nati. Acad. Sci. 76 (5), 2475-2479.Siman, R.G. and Card, J.P. (1988). Excitatory amino acid neurotoxicity in thehippocampal slice preparation. Neurosci. 26 (2), 433-447.Siman, R.G. and Klein, W.L. (1979). Cholinergic activity regulates muscarinicreceptors in central nervous cultures. Proc. Natl. Acad. Sci. USA 76,4141-4145.Smith, M.K., Coibran, R.J. and Soderling, T.R. (1990). Specificities ofautoinhibitory domain peptides for four protein kinases. Implications forintact cell studies of protein kinase function. J. Biol. Chem. 265, 1837-1840.Soderling, T.R., Tan, S.E., McGlade-McCulloh, E., Yamamoto, H. andFukanaga, K. (1994). Excitatory interactions between glutamatereceptors and protein kinases. J. Neurobiol. 3, 304-311.Sokoloff, L., Reivich, M., Kennedy, C., Des Rosiers, M.H., Patlak, C.S., Pettigrew,K.D., Sakurdada, 0. and Shinohara, M. K. (1977). The[14C]Deoxyglucose method for the measurement of local cerebralglucose utilization: theory, procedure and normal values in theconscious and anesthetized albino rat. J. Neurochem. 28, 897-91 6.144Sommer, B., Keinanen, K., Verdoorn, T.A., Wisden, W., Burnashev, N., Herb, A.,Kohier, M., Takagi, T., Sakmann, B. and Seeburg, P.H. (1990). Flip andflop: a cell-specific functional switch in glutamate-operated channels ofthe CNS. Science 249, 1580-1 585.Song, Y. and Huang, L.-Y. M. (1990). Modulation of glycine receptor chloridechannels by cAMP-dependent protein kinase in spinal trigeminalneurons. Nature 384, 242-245.Staubli, U., Ambros-Ingerson, J., and Lynch, G. (1992). Receptor changes andLTP: an analysis using aniracetam, a drug that reversibly modifiesglutamate (AMPA) receptors. Hippocampus 2, 49-58.Stelzer, A., Kay, A.R. and Wong, R:S.K. (1988). GABAA-receptor function inhippocampal cells is maintained by phosphorylation factors. Science241, 339-341.Stephenson, F.A. (1988). Understanding the GABAA receptor: a chemicallygated ion channel. Biochem. J. 249, 21-31.Stryer, L. (1988). Biochemistry. New York: Freeman.Sweetnam, P.M., Lloyd, J., Gallombardo, P., Malison, R.T., Gallager, D.W.,Tailman, J.F. and Nestler, E.J. (1988). Phosphorylation of the GABAAbenzodiazepine receptor a subunit by a receptor-associated proteinkinase. J. Neurochem. 51, 1274-1284.145Swope, S.L., Blackstone, C.D. and Huganir, R.L. (1992). Phosphorylation ofligand gated ion channels: a possible mode of synaptic plasticity.FASEB J. 6, 2514-2523.Tabuteau, H., de Laneroll, N. and Brines, M. (1993). Depolarizationupregulates GABAA receptor number in living hippocampal slices. Soc.Neurosci. Abstracts 19, 1183.Tan, Soon-Eng, Wenthold, R.J. and Soderling, T.R. (1994). Phosphorylation ofAM PA-type glutamate receptors by calci um/calmod ulin-dependentprotein kinase II and protein kinase C in cultured hippocampalneurons. J. Neurosci. 14(3), 1123-1129.Tang, C.-M., Shi, Q.-Y., Katchman, A., and Lynch, G. (1991). Modulation of thetime course of fast EPSCs and glutamate channel kinetics by aniracetam.Science 254, 288-290.Tansey, M.G., Word, R.A., Hidaka, H., Singer, H.A., Schworer, C.M., Kamm, K.E.and Stull, J.T. (1992). Phosphorylation of myosin light chain kinase bythe multifunctional calmodulin-dependent protein kinase Il in smoothmuscle cells. J. Biochem. 267, 12511-12516.Tapley, P.M. and Murray, A.W. (1985). Evidence that treatment of platelets withphorbol ester causes proteolytic activation ofCa2+activated,phospholipid-dependent protein kinase. Eur. J. Biochém. 151, 419-423.Tehrani, M.H.J. and Barnes, Jr., E.M. (1988). GABA down-regulates theGABA/benzodiazepine receptor complex in developing cerebralneurons. Neurosci. Lett. 87, 288-292.146Tennant, J.R. (1964). Evaluation of the trypan blue technique fOr determinationof cell viability. Transplantation 2 (6), 685-694.Teyler, T.J. (1980). Brain slice preparation: hippocampus. Brain Res. Bull. 5,391-403.Tocco,G., Maren, S., Shors, T.J., Baudry, M. and Thompson, R.F. (1992). Long-term potentiation is associated with increased[3H]AMPA binding in rathippocampus. Brain Res. 573, 228-234.Tokumitsu, H., Chijiwa, T., Hagiwara, M., Mizutani, A., Terasawa, M.and Hidaka,H. (1990). KN-62, 1-[N,O- bis (5-isoquinolinesulfonyl) -N-methyl-L-tyrosyl]-4-phenylpiperazine, a specific inhibitor ofCa2/calmodulin-dependent protein kinase II. J. Biol. Chem. 265, 4315-4320.Toullec, D., Pianetti, P. Coste, H., Bellevergue, P., Gran-Perret, T., Ajakane, M.,Baudet, V., Boissin, P., Boursier, E. and Loriolle, F. (1991). Thebisindolylmaleimide GE 109203x is a potent and selective inhibitor ofprotein kinse C. J. Biol. Chem. 266, 15771-15781.Unnerstall, J.R., Niehoff, D.L., Kuhar, M.J. and Palacios, J.M. (1982).Quantitative receptor autoradiography using [3H ] labeled ultrafilm:application on multiple benzodiazepine receptors. J. Neurosci. Methods6 (1-2), 59-73.Vaello, M-L., Ruiz-Gomez, A. and Mayor, Jr., F. (1992). Glycinergic ligandsmodulate the rate of phosphorylation of the glycine receptor by proteinkinase C. Biochem. Biophys. Res. Comm. 188 (2), 81 3-81 9.147Van Haastert, P.J.M., Van Driel, R., Jastoruf, B., Baraniak, J., Stec, W.J. and DeWit, R.J. (1984). Competitive cAMP antagonists for cAMP-receptorproteins. J. Biol. Chem. 259, 10020-10024.Van Huizen, F., Shaw, C., Wilkinson, M. and Cynader, M. (1989).Characterization of muscarinic acetylcholine receptors in rat cerebralcortex with concomitant morphological and physiological assessmentof tissue viability. Mol. Brain Res. 5, 59-69.Van Huizen, F., Wilkinson, M., Cynader, M. and Shaw, C. (1988). Sodiumchannel toxins veratrine and veratridine modify opioid and muscarinicbut not B-adrenergic binding sites in brain slices. Brain Res. Bull. 21,129-132.Walaas, S.I., Nairn, A.C. and Greengard, P. (1983a). DARP-32, a dopamineand adenosine 3’,5’ -monophosphate-regulated phosphoproteinenriched in dopamine-innervated brain region. I. regional and cellulardistribution in the rat brain. J. Neurosci. 3, 302-311.Walaas, S.l., Nairn, A.C. and Greengard, P. (1983b). Regional distribution ofcalcium- and cyclic adenosine 3’, 5’ -monophosphate-regulated proteinphosphorylation systems in mammalian brain. J. Neurosci. 3, 291 -301.Walton, K. and Fulton, B. (1983). Hydrogen peroxide as a source of molecularoxygen for in vitro mammalian CNS preparations. Brain Res. 278, 387-393.Wang, L.Y., Salter, M.W. and MacDonald, J.F. (1991). Regulation of kainatereceptors by cAMP-dependent protein kinase and phosphatases.Science 253, 1132-1 138.148Wang, Y., Qiang, G., Fei., M., Haugland, R. P. and Cynader, M.S. (1994).Activity-dependent expression and distribution of Ml, muscarinic AChreceptors in visual cortex neuronal cultures. J. Neurosci. 14 (7), 4147-4158.Wenthold, R.J., Yokotani, N., Doi, K. and Wada, P. (1992). Immunochemicalcharacterization of the non-NMDA glutamate receptor using subunit-specific antibodies. Evidence for a hetero-oligomeric structure in ratbrain. J. Biol. Chem. 267, 501 -507.Wiesel, T.N. and Hubel, D.H. (1963). Single-cell responses in striate cortex ofkittens deprived of vision in one eye. J. Neurophysiol. 26, 1003-1017.Wilkinson, M., Jacobson, W. and Watson-Wright, W. (1986). Minireview: tissueslices in radioligand binding assays: studies in brain, pineal, andmuscle. Life Sci. 39, 2037-2048.Wurtman, R.J. and Fernstrom, J.D. (1976). Control of brain neurotransmittersyntheses by precursor availability and nutritional state. Biochem.Pharmacol. 25, 1691-1696.Yamamura, H.l., Enna, S.J. and Kuhar M.J. (1990). Methods inneurotransmitter receptor analysis. New York: Raven Press.Yang, J., Logsdon, C.D., Johansen, T.E. and Williams, J.A. (1994). Human m3muscarinic acetyicholine receptor carboxyl-terminal threonine residuesare required for agonist-induced receptor down-regulation. Molec.Pharmacol. 44, 1158-1164.149Young, A.B. and Fagg, G.E. (1991). Excitatory amino acid receptors in thebrain: membrane binding and receptor autoradiographic approaches.Trends in Pharmacol. Sci. Special Report, 18-24.Young, S. and Kuhar, M.J. (1987). A new method for receptor autoradiography:[3H] -opiold receptors in rat brain. Brain Res. 179, 255-270.Zaczek, R., Arlis, S., Marki, A., Murphy, T., Drucker, H. and Coyle, J.T. (1987).Characteristics of chloride-dependent incorporation of glutamate intobrain membranes argue against a receptor binding site.Neuropharmacol. 26, 281-287.Zbinden, G., ed. (1992). The brain in bits and pieces. In vitro techniques inneurobiology, neuropharmacology and neurotoxicology. Zollikon,Switzerland: MCT Verlag., 13-15


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items