Kainate Receptors and Synaptic Transmission James E

Home , Kainate receptor, NS102, Saclofen

Progress in Neurobiology 70 (2003) 387–407

Kainate receptors and synaptic transmission James E. Huettner∗ Department of Cell Biology and Physiology, Washington University School of Medicine, 660 South Euclid Avenue, St. Louis, MO 63110, USA Received 20 February 2003; accepted 25 July 2003

Abstract Excitatory glutamatergic transmission involves a variety of different receptor types, each with distinct properties and functions. Physiolog- ical studies have identified both post- and presynaptic roles for kainate receptors, which are a subtype of the ionotropic glutamate receptors. Kainate receptors contribute to excitatory postsynaptic currents in many regions of the central nervous system including hippocampus, cortex, spinal cord and retina. In some cases, postsynaptic kainate receptors are co-distributed with ␣-amino-3-hydroxy-5-methyl-4- isoxazolepropionic acid (AMPA) and N-methyl-d-aspartate (NMDA) receptors, but there are also synapses where transmission is mediated exclusively by postsynaptic kainate receptors: for example, in the retina at connections made by cones onto off bipolar cells. Modulation of transmitter release by presynaptic kainate receptors can occur at both excitatory and inhibitory synapses. The depolarization of nerve terminals by current flow through ionotropic kainate receptors appears sufficient to account for most examples of presynaptic regulation; however, a number of studies have provided evidence for metabotropic effects on transmitter release that can be initiated by activation of kainate receptors. Recent analysis of knockout mice lacking one or more of the subunits that contribute to kainate receptors, as well as studies with subunit-selective agonists and antagonists, have revealed the important roles that kainate receptors play in short- and long-term synaptic plasticity. This review briefly addresses the properties of kainate receptors and considers in greater detail the physiological analysis of their contributions to synaptic transmission. © 2003 Elsevier Ltd. All rights reserved.

Contents

1. Introduction ...... 388 2. Kainate receptor properties ...... 388 3. Kainate receptor distribution and function ...... 390 3.1. Hippocampus ...... 390 3.1.1. Presynaptic receptors ...... 390 3.1.2. Postsynaptic receptors ...... 395 3.1.3. Transgenic mice ...... 396 3.1.4. Synaptic plasticity...... 398 3.2. Cortex ...... 398 3.3. Amygdala...... 399 3.4. Retina ...... 399 3.5. Striatum ...... 400 3.6. Hypothalamus ...... 400

Abbreviations: GYKI53655, 1-(4-aminophenyl)-3-methylcarbamyl-4-methyl-7,8-methylenedioxy-3,4-dihydro-5H-2,3-benzodiazepine; SYM2081, 2S, 4R-4-methylglutamate; SYM2206, (±)-4-(4-aminophenyl)-1,2-dihydro-1-methyl-2-propylcarbamoyl-6,7-methylenedioxyphthalazine; CNQX, 6-cyano- 7-nitroquinoxaline-2,3-dione; APV, 2-amino-5-phosphono-valerate; NMDA, N-methyl-d-aspartate; CPCCOEt, 7-(hydroxyimino)cyclopropa[β]-chromen- 1␣-carboxylate ethylester; AMPA, ␣-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; ATPA, (RS)-2-amino-3-(3-hydroxy-5-tertbutylisoxazol-4- yl)propanoic acid; DRG, dorsal root ganglion; trans-PDC, trans-pyrrolidine-2,4-carboxylic acid; TTX, tetrodotoxin ∗ Tel.: +1-314-362-6624; fax: +1-314-362-7463. E-mail address: [email protected] (J.E. Huettner). URL: http://www.cellbio.wustl.edu/faculty/huettner/.

3.7. Cerebellum ...... 400 3.8. Spinal cord ...... 401 3.9. Dorsal root ganglia ...... 401 4. Perspectives ...... 402 Acknowledgements ...... 402 References ...... 402

1. Introduction rents (Kiskin et al., 1986; Keinänen et al., 1990; Patneau and Mayer, 1991), and AMPA can activate at least some types Kainate receptors are one of three subtypes of ionotropic of kainate receptor (Herb et al., 1992). receptors for the excitatory transmitter l-glutamate There are five different subunits that contribute to kainate (Dingledine et al., 1999). The other two subtypes, which receptors (Hollmann and Heinemann, 1994). They fall into are named for the synthetic agonists N-methyl-d-aspartate two families, based on sequence homology and agonist bind- 2 (NMDA) and ␣-amino-3-hydroxy-5-methyl-4-isoxazolepro- ing properties. GLUK5, GLUK6 and GLUK7 are approxi- pionic acid (AMPA), are known to mediate postsynaptic mately 70% identical (Bettler et al., 1990, 1992; Egebjerg currents at excitatory synapses throughout the brain and et al., 1991; Sommer et al., 1992). The GLUK1 and GLUK2 spinal cord (Mayer and Westbrook, 1987). The physiologi- subunits also are 70% identical (Werner et al., 1991; Herb cal properties of kainate receptors (Chittajallu et al., 1999; et al., 1992; Sakimura et al., 1992), but share only 40% Lerma et al., 2001), and their roles in synaptic transmission identity with GLUK5, GLUK6 and GLUK7. Both families (Frerking and Nicoll, 2000; Kullmann, 2001; Lerma, 2003), of kainate receptor subunits also display weaker identity have been discerned only recently, following the discovery with subunits of AMPA (30–35%) and NMDA receptors of selective antagonists that allow for isolation of kainate (10–20%). In addition, all of the glutamate receptor sub- receptor-mediated currents (Paternain et al., 1995; Wilding units are thought to adopt the same membrane topology. The and Huettner, 1995; Bleakman et al., 1996a). Additional amino terminal half of each subunit is extracellular. There interest in kainate receptors has been raised by the cloning are four hydrophobic segments: three membrane spanning and characterization of their subunit cDNAs (Hollmann domains and a “p-loop” that dips into the membrane from and Heinemann, 1994), and by the recognition that kainate the cytoplasmic face to form the pore (Hollmann et al., 1994; receptor subunits are distinct from subunits that contribute Roche et al., 1994; Bennett and Dingledine, 1995). to AMPA receptors (Boulter et al., 1990; Keinänen et al., The GLUK5 and GLUK6 subunits, but not the other 1990) and to NMDA receptors (Kutsuwada et al., 1992; kainate receptor subunits, can undergo mRNA editing that Monyer et al., 1992; Moriyoshi et al., 1991). changes an amino acid in the channel pore and regulates permeation properties (Sommer et al., 1991). For both GLUK5 and GLUK6, as well as the GLUA2 subunit of AMPA recep- 2. Kainate receptor properties tors, the genomic sequence encodes a glutamine residue in the edited location in the p-loop that is converted by editing Kainate receptors were originally defined by Watkins and to code for an arginine (Sommer et al., 1991). In all three coworkers (Davies et al., 1979; Watkins and Evans, 1981) cases, mature receptors comprised of unedited subunits dis- based on the pharmacology of neuronal responses to excita- play inwardly rectifying current–voltage (I–V) relations ow- tory amino acids. In particular, the selective depolarization ing to block of outward current by intracellular polyamines, of isolated dorsal root fibers by kainate led them to propose whereas receptors including edited subunits resist polyamine a unique receptor for kainate that was distinct from the bind- block and have linear I–V relations (Bowie and Mayer, ing sites activated by NMDA and AMPA.1 Subsequent work 1995; Kamboj et al., 1995; Isa et al., 1995; Donevan and has confirmed the existence of three different receptor sub- Rogawski, 1995; Koh et al., 1995; Bähring et al., 1997). types (Hollmann and Heinemann, 1994; Dingledine et al., Editing at the Q/R site also determines single channel 1999), although it also has been recognized that many ex- conductance and calcium permeability. Fully unedited citatory amino acids, including kainate and AMPA, are not receptors exhibit a higher relative calcium permeability entirely selective for only one receptor class. Thus, kainate (Egebjerg and Heinemann, 1993; Burnashev et al., 1995, activates AMPA receptors to produce large sustained cur- 1996) and a higher unitary conductance (Howe, 1996; Swanson et al., 1996) as compared to receptors that include one or more edited subunits. In addition to the Q/R site, the 1 The original classification proposed by Watkins and coworkers identified NMDA, kainate and quisqualate receptors; however, AMPA was subsequently recognized as a more selective agonist than quisqualate and 2 IUPHAR nomenclature (Lodge and Dingledine, 2000) used throughout the classification was revised (see Monaghan et al., 1989; Watkins et al., this review. Previous designations: GLUK1, KA1; GLUK2, KA2; GLUK5, 1990). GluR5; GLUK6, GluR6; GLUK7, GluR7. J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 389

GLUK6 subunit also displays two additional sites of RNA work has been done on heteromeric channels that include editing in the first transmembrane domain (Köhler et al., the GLUK1 subunit. 1993). Binding studies have documented both high and low 3 Agonist binding to glutamate receptor subunits involves affinity sites (KD ∼ 5 and 50 nM for [ H]kainate, respec- regions in the N-terminal domain and the extracellular tively) in rat brain membranes (London and Coyle, 1979; domain between the last two transmembrane segments Unnerstall and Wamsley, 1983; Hampson et al., 1987). It (Kuusinen et al., 1995; Stern-Bach et al., 1994; Armstrong has been proposed that high affinity receptors may corre- et al., 1998; Armstrong and Gouaux, 2000). Many of spond to heteromeric assemblies that included the GLUK1 the glutamate receptor subunit mRNAs can be alternately or GLUK2 subunits, whereas receptors lacking GLUK1 or spliced, resulting in structural diversity among mature re- GLUK2 would exhibit lower affinity (Contractor et al., ceptors (Hollmann and Heinemann, 1994). For example, 2003). On the other hand, recombinant heteromeric recep- the GLUK5 subunit N-terminal domain can either include tors that include GLUK1 or GLUK2 typically display much or lack a 15 amino acids segment encoded by an alternately lower agonist affinity when expressed in vitro than the ho- spliced exon (Bettler et al., 1990). In addition, the carboxy momeric binding sites formed by expression of GLUK1 or terminal domains of GLUK5, GLUK6 and GLUK7 display GLUK2 in isolation (Herb et al., 1992; Pemberton et al., 3 alternate splicing that alters their length and, in some cases, 1998). For example, H-kainate binds to GLUK2/GLUK5 their terminal residues (Sommer et al., 1992; Gregor et al., heteromers with a KD of 90 nM, as compared to 15 nM for 1993; Schiffer et al., 1997). For several of the glutamate GLUK2 homomers or 73 nM for GLUK5 homomers (Herb receptor subunits, the cytoplasmic carboxy terminus is et al., 1992). Further work is needed to resolve the subunit known to contain sites for phosphorylation (Dingledine composition of high affinity binding sites in native mem- et al., 1999) that may serve to regulate channel function branes, as it is not clear what function the high affinity (Raymond et al., 1993; Wang et al., 1993; Dildy-Mayfield sites would serve if they represent homomeric assemblies and Harris, 1994; Traynelis and Wahl, 1997; Ghetti and of GLUK1 or GLUK2. Heinemann, 2000). In addition, the GLUK6 subunit can be More recently, several groups have presented evidence palmitoylated at two cysteine residues located in the carboxy that GLUK5 and GLUK6 can co-assemble into heteromeric terminal domain (Pickering et al., 1995). Specific interac- channels (Bortolotto et al., 1999; Cui and Mayer, 1999; tions with modulatory proteins also have been described for Paternain et al., 2000). These studies took advantage of the carboxy terminal domains of individual glutamate receptor fact that inclusion of a subunit with an R at the Q/R editing subunits. For example, several of the subunits, including site will significantly reduce polyamine block of outward GLUK6, can interact with proteins that contain specific PDZ current. Thus, co-expression of GLUK5(Q) with GLUK6(R) domains (Garcia et al., 1998; Hirbec et al., 2003). Such yields receptors that display little or no inward rectifica- interactions are thought to coordinate the joint synaptic lo- tion, but which are sensitive to the GLUK5-selective com- calization of receptors with other PDZ domain-containing pounds ATPA and LY382884 (Clarke et al., 1997; Lauridsen proteins into functional complexes (Savinainen et al., 2001; et al., 1985; O’Neill et al., 1998). Similar results were ob- Sheng and Pak, 2000; Mehta et al., 2001; Coussen et al., tained upon co-expression of GLUK5(R) with GLUK6(Q). 2002). In addition, Cui and Mayer (1999) demonstrated functional A great deal has been learned from studies in which co-assembly of GLUK7 with either GLUK5 or GLUK6. Thus, kainate receptor subunits were expressed in Xenopus a large number of distinct kainate receptor subtypes could oocytes or in transfected HEK 293 cells. GLUK5, GLUK6 be assembled based on the combinatorial mixing of five dif- and GLUK7 are capable of forming functional homomeric ferent subunits. ligand-gated channels when expressed in isolation (Bettler The distribution of cells expressing kainate receptor mR- et al., 1990; Egebjerg et al., 1991; Sommer et al., 1992; NAs has been mapped by in situ hybridization (Wisden Schiffer et al., 1997). Binding of radioactive ligands to and Seeburg, 1993; Bahn et al., 1994; Tölle et al., 1993; these homomeric receptors indicate dissociation constants Paternain et al., 2000; Bureau et al., 1999). In addition, for kainate in the range of 50–100 nM (Bettler et al., 1992; several laboratories have used RT-PCR to determine sub- Lomeli et al., 1992; Sommer et al., 1992; Schiffer et al., unit expression in individual neurons (Ruano et al., 1995; 1997). When the GLUK1 and GLUK2 subunits are expressed Ghasemzadeh et al., 1996; Porter et al., 1998; Cauli et al., in isolation they form high affinity binding sites for kainate, 2000; Dai et al., 2002) or small populations of cells isolated with dissociation constants in the range of 5–15 nM (Werner from specific regions (Belcher and Howe, 1997; Sahara et al., et al., 1991; Herb et al., 1992). In contrast to GLUK5, 1997; Paarmann et al., 2000). By in situ hybridization, cells GLUK6 and GLUK7, however, GLUK1 and GLUK2 do not displaying prominent expression of the GLUK5, GLUK6, form detectable homomeric channels when expressed alone GLUK7, and GLUK2 kainate receptor subunits are distributed in oocytes or in mammalian cells. Co-expression studies throughout the CNS including cortex, striatum, hippocam- (Herb et al., 1992; Sakimura et al., 1992) have demon- pus and cerebellum. Notable expression of the GLUK1 sub- strated the formation of heteromeric channels that include unit is observed primarily in hippocampal CA3 and dentate the GLUK2 subunit together with GLUK5 or GLUK6. Less granule neurons, whereas message for GLUK2 appears to be 390 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 more abundant and more widespread than for GLUK1 or the 3. Kainate receptor distribution and function other subunits. Interested readers should refer to the original literature for detailed descriptions of expression in specific 3.1. Hippocampus cell types (Bettler et al., 1990; Egebjerg et al., 1991; Werner et al., 1991; Lomeli et al., 1992; Wisden and Seeburg, 1993; Detailed study of the role of kainate receptors in synap- Bahn et al., 1994; Tölle et al., 1993; Paternain et al., 2000; tic transmission has only been possible since the discovery Bureau et al., 1999). of selective AMPA receptor antagonists in 1995 (Paternain Analysis of receptors expressed by neurons in culture or et al., 1995; Wilding and Huettner, 1995; Bleakman et al., in situ confirms that native receptors are, in fact, heteroge- 1996a). Before that time, a number of reports described in- neous. Cells from different regions of the nervous system teresting effects on excitability (Robinson and Deadwyler, express kainate receptors with distinct physiological and 1981; Westbrook and Lothman, 1983) and synaptic trans- pharmacological properties (Wilding and Huettner, 2001). mission (Collingridge et al., 1983; Kehl et al., 1984; Fisher Features common among all neuronal kainate receptors in- and Alger, 1984) following exposure to low doses of kainate. clude activation by micromolar concentrations of domoate It seems likely that kainate receptors mediated many of these and kainate (Lerma et al., 1993; Paternain et al., 1998; effects, particularly in experiments that used submicromolar Wilding and Huettner, 1997), activation and potent desen- concentrations of kainate; however, the possibility remains sitization by 4-methylglutamate, SYM2081 (Jones et al., that both AMPA and kainate receptors were activated during 1997; Donevan et al., 1998), and blockade by lanthanum exposure to the agonist in these early studies. With the ad- (Huettner et al., 1998) and by the competitive antagonist vent of GYKI53655, and other selective AMPA and kainate CNQX (Wilding and Huettner, 1995; Paternain et al., 1996). receptor antagonists, it became possible to define the sepa- Differences exist from one cell type to the next in the rate contributions of kainate, AMPA and NMDA receptors exact potency of specific agonists for receptor activation (see Table 1 and Bleakman and Lodge, 1998). Many of the or desensitization, and in the effect of GLUK5-selective initial studies focused on hippocampal neurons. agonists and antagonists (Kerchner et al., 2001a; Wilding and Huettner, 2001). In nearly all cases, it remains to be 3.1.1. Presynaptic receptors determined whether these differences in receptor properties The first direct evidence for involvement of kainate re- between cell types arise from differential subunit expres- ceptors in synaptic transmission demonstrated a presynaptic sion, alternate splicing, interaction with cell type specific reduction in transmitter release by Schaffer collaterals onto accessory proteins or some combination of these possibili- CA1 neurons upon the activation of kainate receptors in ties. Recently, progress in defining the subunit composition acute hippocampal slices (Chittajallu et al., 1996). Subse- of kainate receptors in specific cell types has been made quent studies reported a similar reduction in transmission using subunit-selective drugs and knockout mice defi- evoked from hippocampal GABAergic interneurons (Clarke cient in expression of particular kainate receptor subunits et al., 1997; Rodriguez-Moreno et al., 1997). In addition, (discussed below), but much more work is needed to de- evidence was obtained that postsynaptic kainate receptors fine the structural basis for heterogeneity of endogenous contribute to the EPSC at mossy fiber synapses (Vignes and receptors. Collingridge, 1997; Castillo et al., 1997; Mulle et al., 1998)

Table 1 Selected pharmacological agents for kainate receptor analysis APV Competitive antagonist of NMDA receptors ATPA Agonist selective for GLUK5, but also able to activate AMPA receptors and GLUK6/GLUK2 heteromers with lower potency CNQX Competitive antagonist of AMPA and kainate receptors Concanavalin A Lectin that inhibits kainate receptor desensitization Cyclothiazide Allosteric regulator that potentiates AMPA receptors Domoate Kainate receptor agonist producing less desensitization than kainate Glutamate Endogenous transmitter and agonist for kainate, AMPA, NMDA and mGluRs GYKI52466 Non-competitive antagonist selective for AMPA receptors, but not as potent or as selective as GYKI53655 GYKI53655 Non-competitive antagonist selective for AMPA receptors Kainate Kainate receptor agonist, but also activates AMPA receptors LY293558 Competitive antagonist selective for GLUK5, but also an AMPA receptor antagonist LY294486 Competitive antagonist selective for GLUK5, but also a weak AMPA receptor antagonist LY377770 Active isomer of LY294486 LY382884 Competitive antagonist selective for GLUK5 NBQX Competitive antagonist of AMPA and kainate receptors NS102 Antagonist selective for kainate receptors, but with limited solubility in physiological solutions SYM2081 Agonist selective for kainate receptors that produces strong, persistent desensitization SYM2206 Non-competitive antagonist for AMPA receptors comparable to GYKI53655 Selectivity of additional compounds provided in the text. J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 391 and at excitatory synapses onto interneurons (Cossart et al., pre-loaded rat hippocampal synaptosomes. The IC50 for 1998; Frerking et al., 1998). kainate was 27 ␮M, and there was no effect of kainate on In order to study the presynaptic actions of kainate, basal release in normal K+. NS102 (10 ␮M) and CNQX slices or cultures are equilibrated with a selective AMPA (30 ␮M) partially overcame the inhibition of release by antagonist. GYKI53655 or SYM2206 at 100 ␮M pro- kainate; however, NBQX (10 ␮M) GYKI52466 (100 ␮M) vide complete non-competitive blockade of AMPA recep- and cyclothiazide (100 ␮M) had little or no effect on this tors (Paternain et al., 1995; Wilding and Huettner, 1995; action of kainate. In physiological recordings from CA1 Pelletier et al., 1996; Bleakman et al., 1996a), while caus- neurons in rat hippocampus, Chittajallu et al. (1996) mon- ing only slight reduction in currents mediated by kainate itored NMDA receptor-mediated EPSCs in the presence receptors (Wilding and Huettner, 2001). Driven by the lim- of GYKI52466 (100 ␮M). Exposure to 1–30 ␮M kainate ited availability of GYKI53655 some investigators have elicited an initial burst of spontaneous EPSCs but then pro- employed lower doses of GYKI53655 (40–50 ␮M), or used duced a sustained reduction in EPSCs evoked by stimulation the less selective, but commercially available, compound of Schaffer collateral fibers. Domoate had a similar action at GYKI52466. Under these conditions, currents mediated by 100 nM, whereas 300 nM kainate only caused a modest en- AMPA receptors will be significantly reduced but may not hancement of evoked transmission, and 100 nM kainate had be totally abolished (Donevan et al., 1994; Paternain et al., no effect. These effects of kainate were not blocked by an- 1995; Wilding and Huettner, 1995). To study transmission tagonists of GABAA (50 ␮M picrotoxin), GABAB (10 ␮M at excitatory synapses with AMPA receptors blocked, mag- CGP55845) or adenosine A1 (0.1 ␮M DPCPX) receptors. nesium can be removed from the medium revealing the Similar results were obtained by Kamiya and Ozawa NMDA receptor-mediated component of excitatory post- (1998), who also studied synapses formed by Schaffer synaptic current (Davies and Collingridge, 1989; Hestrin collaterals onto CA1 neurons in 14–21-day-old rats. In et al., 1990). For study of evoked IPSCs, both AMPA and this study, presynaptic calcium transients were monitored NMDA receptors can be blocked. Using these conditions, by labeling the Schaffer collateral fibers with rhod-2AM. superfusion of kainate, domoate or other kainate recep- Bath applications of 1 ␮M kainate reduced the postsynaptic tor agonists produces a reduction in postsynaptic current, EPSC and the presynaptic calcium signal in parallel. In which recovers when the agonist is removed. This effect addition, paired pulse facilitation was enhanced during exon transmission is sensitive to block by CNQX, indicat- posure to kainate. Such changes in paired pulse modulation ing a dependence on kainate receptor activation. However, are believed to be a reliable indicator of a presynaptic locus because kainate might stimulate release of endogenous of regulation. Kamiya and Ozawa (1998) repeated these ob- transmitters, including glutamate, well-controlled studies servations in the presence of picrotoxin to block ionotropic use antagonists for a number of additional receptors such as GABAA receptors and thus reduce the possibility of indirect GABAB, adenosine, and metabotropic glutamate receptors effects mediated by an action of kainate on interneurons to limit the possibility of indirect effects. (but see below). They also showed that kainate did not As discussed throughout this review, exposure to kainate affect the presynaptic fiber volley and that the effects of receptor agonists can produce significantly different results kainate were reduced by NS102 (Verdoorn et al., 1994). depending on the agonist concentration. This observation Subsequent work by Frerking et al. (2001) confirmed and has led some authors to question the physiological relevance extended the evidence for presynaptic modulation of release of effects observed with kainate concentrations above 1 ␮M by demonstrating a reduction in quantal content during (Ben-Ari and Cossart, 2000). To allow readers to draw their exposure to 200 nM domoate. To determine whether depo- own conclusions, the concentrations of both agonists and larization of presynaptic fibers was required for this form of antagonists are provided for each specific experimental re- modulation, Frerking et al. (2001) compared the action of sult discussed. Most of the studies cited have included con- domoate with increases in extracellular KCl. Concentrations trol experiments to mitigate possible non-selective actions of KCl that were sufficient to inhibit evoked release caused of kainate. In addition, steady-state concentration-response a comparable reduction in the presynaptic fiber volley and analysis indicates that 20–30 ␮M kainate is needed to pro- an increase in the frequency of mEPSCs, whereas domoate duce half-maximal activation of whole-cell currents in iso- had neither effect. Thus, although a high concentration lated neurons (Lerma et al., 1993; Paternain et al., 1998; of domoate (20 ␮M) was shown to be capable of causing Wilding and Huettner, 1997), suggesting that doses greater significant depolarization, depolarization of presynaptic than 1 ␮M might be required to achieve significant receptor terminals was not essential for modulation of release by activation during slow perfusion of intact tissue slices. 200 nM domoate. Instead, the effect of domoate on evoked release was suppressed by inhibitors normally associated 3.1.1.1. Excitatory synapses. Chittajallu et al. (1996) with metabotropic receptors, including pertussis toxin and were the first to provide direct evidence that presynaptic the protein-modifying reagent N-ethylmaleimide. Evidence kainate receptors may modulate release of glutamate. They for metabotropic actions via kainate receptors has also been showed that exposure to kainate produced a dose-dependent obtained for inhibitory nerve terminals in hippocampal area reduction in K+-stimulated release of [3H]glutamate from CA1 (Rodriguez-Moreno and Lerma, 1998; and see below) 392 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 and for postsynaptic regulation of potassium current that transmission and to enhance the fiber volley, presumably by underlies the slow afterhyperpolarization (IsAHP) in CA1 activation of presynaptic kainate receptors (Schmitz et al., pyramidal cells (Melyan et al., 2002). Although the exact 2000). mechanisms whereby kainate receptors elicit metabotropic More recent work (Schmitz et al., 2001a,b; Contractor effects are not clear, Frerking et al. (2001) argue that the et al., 2001; Lauri et al., 2001a,b; Kamiya et al., 2002), how- presynaptic action of kainate and domoate on Schaffer col- ever, has modified and extended the initial interpretation of lateral fiber transmission is likely to be direct, rather than an the role that presynaptic kainate receptors play at mossy indirect action of metabotropic neuromodulators released fiber synapses. Whereas initial studies focused on the inhi- from interneurons. Their evidence for a direct action on bition produced by exogenous kainate at 200 nM and above, excitatory fibers included the inability to prevent regula- subsequent work has revealed evidence for potentiation of tion by a cocktail of antagonists to known neuromodula- mossy fiber transmission (see also Kehl et al., 1984), either tors as well as the demonstration that solutions containing by low concentrations of kainate (50 nM) or by release of en- high divalent ion concentrations (8 mM Ca2+ and 17 mM dogenous glutamate during short-term frequency-dependent Mg2+) were able to block interneuron spiking elicited by plasticity. Schmitz et al. (2001a) found that 50 nM kainate kainate receptor agonists, but did not prevent regulation caused a 50–100% enhancement in mossy fiber transmis- of Schaffer collateral transmission. On the other hand, in sion, an effect that was mimicked by mild depolarization some (Mulle et al., 2000; Cossart et al., 2001), but not all with 4 mM KCl. Full dose response curves for kainate and (Frerking et al., 1999; Jiang et al., 2001; Rodriguez-Moreno KCl demonstrated a striking parallel in their effects on EPSC et al., 1997), studies of hippocampal area CA1, exposure to amplitude and afferent fiber volleys. Low effector concen- kainate was found to stimulate quantal release from some trations enhanced both of these parameters. EPSC ampli- interneurons during action potential blockade by TTX (see tude increased to a peak at 50 nM kainate or 4 mM KCl, but below), raising the possibility that kainate-evoked release of was reduced relative to baseline for higher concentrations a metabotropic neuromodulator may not have been entirely of either agent. Effects on fiber volley amplitude showed a eliminated by elevating the divalent ion concentrations. similar trend, but were displaced to higher concentrations, There is clearly a need for more work to explore the po- reaching maximal enhancement at 500 nM kainate or 8 mM tential coupling mechanism between kainate receptors and KCl and declining below baseline at higher doses. NEM- or pertussis toxin-sensitive effectors. Several groups (Schmitz et al., 2001a; Contractor et al., In addition to the Schaffer collateral-commissural 2001; Lauri et al., 2001a) have now demonstrated a role synapses onto CA1 neurons, the role of presynaptic kainate for presynaptic kainate receptors in frequency-dependent receptors has also been examined at mossy fiber synapses facilitation at mossy fiber synapses. Schmitz et al. (2001a) formed by dentate granule cells onto CA3 pyramidal neu- monitored the amplitude of NMDA receptor-mediated rons (Vignes et al., 1998; Kamiya and Ozawa, 2000; Schmitz EPSCs in the context of AMPA receptor blockade with et al., 2000). In this case, incubation with 200 nM kainate 20 ␮M GYKI53655. Application of 10 ␮M CNQX or alters the presynaptic fiber volley, increasing the excitabil- 50 ␮M NBQX significantly blunted facilitation of EPSCs ity of presynaptic fibers but reducing the calcium influx elicited by repetitive stimulation at 25 or 100 Hz. In ad- elicited by orthodromic action potentials. Schmitz et al. dition, this study (Schmitz et al., 2001a) demonstrated (2000) examined the effect of kainate on fiber excitability heterosynaptic facilitation of mossy fiber synapses follow- in detail. Low doses of kainate were found to enhance the ing a short tetanus (3 pulses/200 Hz) delivered to asso- orthodromic fiber volley as well as the firing of spikes by ciational/commissural fibers. Schmitz et al. (2001a) also granule cells in response to antidromic stimulation of mossy confirmed, however, the occurrence of heterosynaptic de- fibers. A higher dose of kainate (10 ␮M) caused an initial pression for longer (10 pulses/200 Hz) or higher intensity increase in the fiber volley followed by a prolonged period stimulation (Schmitz et al., 2000). Lauri et al. (2001a) of fiber volley suppression that significantly outlasted the observed a similar reduction in frequency dependent facil- period of exposure to kainate. Elevation of extracellular itation at mossy fiber synapses with the GLUK5-selective K+, from endogenous sources, during exposure to kainate antagonist LY382884 (10 ␮M). This reduction was not af- was not sufficient to account for these changes in fiber fected by 1 ␮M CGP55845, a GABAB receptor antagonist, excitability (Schmitz et al., 2000). Interestingly, these ini- and was not mimicked by the protein kinase C (PKC) in- tial studies found that even conditions that elevated fiber hibitor calphostin C, which attenuates effects ascribed to excitability nevertheless caused a reduction in the strength the metabotropic action of kainate (Rodriguez-Moreno and of transmission (see below). The amplitude of field EPSPs Lerma, 1998). Frequency-dependent facilitation at mossy (Kamiya and Ozawa, 2000), or whole-cell EPSCs (Vignes fiber synapses has also been examined in subunit defi- et al., 1998), recorded in the absence of NMDA or AMPA cient mice (Contractor et al., 2001). Knocking out GLUK5 receptor antagonists was reduced by ∼60% during expo- caused little or no change in short-term plasticity, whereas sure to 0.2–1 ␮M kainate. In addition, release of glutamate deletion of GLUK6 reduced paired pulse facilitation (PPF) upon stimulation of associational/commissural fibers was for inter-pulse intervals less than 40 ms and significantly shown to cause heterosynaptic inhibition of mossy fiber reduced facilitation during brief tetani at 0.5–5 Hz. Optical J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 393 recordings of calcium elevation and membrane potential fibers by ATPA or an indirect effect mediated by some agent also support a role for presynaptic kainate receptors in other than, or in addition to, GABA. As mentioned above, paired pulse facilitation (Kamiya et al., 2002). In this study the ability of LY382884 to block modulation by exogenous (Kamiya et al., 2002), 10 ␮M CNQX reduced PPF of presy- ATPA and to inhibit both short-term (Lauri et al., 2001a) naptic calcium elevation and presynaptic afterdepolariza- and long-term (Bortolotto et al., 1999) plasticity at mossy tions to a much greater extent than did 100 ␮M GYKI52466. fiber synapses argues that GLUK5 contributes to receptors Moreover, there was no significant effect on PPF by expo- involved in presynaptic modulation at excitatory synapses. sure to the GABAB receptor agonist baclofen (10 ␮M), the On the other hand, compelling evidence that the GLUK5 sub- A1 adenosine receptor agonist 2-chloroadenosine (10 ␮M), unit is not absolutely required for presynaptic modulation or the group II mGluR agonist DCG-IV (1 ␮M). of mossy fibers, or associational/commissural fibers, comes The pharmacological profile and subunit composition of from studies demonstrating a persistence of modulation in presynaptic receptors that affect excitatory transmission in slices from GLUK5-deficient mice (Contractor et al., 2000, the hippocampus remain controversial, both for mossy fibers 2001; see Section 3.1.3). arising from dentate granule cells and for the excitatory axons of CA3 pyramidal cells. In both cases, Collingridge 3.1.1.2. Inhibitory synapses. In addition to the presynap- and coworkers (Vignes et al., 1998; Bortolotto et al., 1999; tic actions of kainate just described, kainate receptors also Lauri et al., 2001a,b; Clarke and Collingridge, 2002)have have been implicated in regulating the evoked release of suggested that the presynaptic receptors include a GLUK5 GABA from inhibitory interneurons. Early studies (Fisher subunit. Their initial results (Vignes et al., 1998; Bortolotto and Alger, 1984; Kehl et al., 1984) showed that 0.3–1 ␮M et al., 1999) demonstrated that the GLUK5-selective ag- kainate reduced the amplitude of spontaneous and evoked onist ATPA was equally effective as kainate at reducing IPSPs recorded in both CA1 and CA2/CA3 pyramidal cells transmission by mossy fibers and associational/commissural (see also Sloviter and Damiano, 1981). More recent work, fibers onto CA3 pyramidal neurons, or by Schaffer collat- using AMPA-selective antagonists, has confirmed this ef- eral/commissural fibers onto pyramidal cells in CA1 (Vignes fect and demonstrated that it involves kainate receptors. All et al., 1998). In subsequent work (Clarke and Collingridge, of the groups that have studied the action of kainate on in- 2002), inhibition by ATPA of excitatory transmission in area hibitory transmission in hippocampus agree that exposure CA1 was found to be independent of GABAA, GABAB, to micromolar concentrations of kainate causes inhibition of muscarinic or adenosine A1 receptor activation, suggesting evoked release (but see Mulle et al., 2000). There is disagree- that ATPA may act directly on excitatory terminals. On the ment, however, concerning the mechanism(s) that underlie other hand, a number of differences were noted between the this inhibition. In addition, similar to the work on excitatory effects of ATPA and kainate on transmission in area CA1 transmission, more recent studies have provided evidence (Clarke and Collingridge, 2002): first, the GLUK5 selective for enhancement of inhibitory transmission by submicromo- antagonist LY382884 was effective at blocking the action lar doses of kainate (Jiang et al., 2001; Cossart et al., 2001). of ATPA, but was weaker against kainate. Second, elevated Clarke et al. (1997) recorded from CA1 neurons and Ca2+ depressed the effect of ATPA, but not kainate. Third, showed that evoked IPSPs were suppressed by 5 ␮M kainate kainate, but not ATPA, enhanced excitability of the CA1 or by the GLUK5-selective agonist ATPA (10 ␮M). More- neurons. Together, these findings raise the possibility that over, the GLUK5-selective antagonist LY294486 prevented ATPA and kainate influence excitatory transmission via dis- regulation of IPSPs by both kainate and ATPA. They also tinct mechanisms. Additional evidence for differences be- showed that GABAA and GABAB receptor-mediated com- tween the actions of ATPA and kainate also comes from ponents of the IPSP were blocked to the same extent, and work on mossy fiber synapses. Schmitz et al. (2000) found that feedback mediated by GABAB autoreceptors was not that, in contrast to kainate, ATPA did not alter the affer- required for kainate and ATPA to reduce inhibitory trans- ent fiber volley at mossy fiber synapses (see also Vignes mission. Lerma and coworkers (Rodriguez-Moreno et al., et al., 1998). Moreover, the reduction in mossy fiber trans- 1997) also have studied inhibitory inputs to CA1 neurons and mission produced by ATPA, but not kainate, was blocked by demonstrated a reduction in evoked IPSPs by superfusion the GABAB receptor antagonist SCH50911 (20 ␮M), sug- with kainate. The concentration-response relation for this gesting that ATPA may exert its effect indirectly by stimu- effect was “bell-shaped”, suggesting that activated, but not lating interneurons to fire and release GABA, which would desensitized, receptors were responsible. The apparent EC50 then act on presynaptic GABAB receptors on the mossy for kainate was 20 ␮M, in agreement with work on kainate fiber terminals (Schmitz et al., 2000). This interpretation has receptor activation in cultured hippocampal neurons (Lerma been contested by Lauri et al. (2001b), however, who per- et al., 1993; Wilding and Huettner, 1997). Modulation of IP- formed similar experiments and observed no effect of 20 ␮M SCs declined at higher kainate doses with an apparent IC50 of SCH50911 on the ability of ATPA to suppress transmission 370 ␮M (cf. Paternain et al., 1998). Lerma’s group showed onto CA3 neurons either by mossy fibers or by associa- that kainate increased the proportion of failures of evoked tional/commissural fibers. These results (Lauri et al., 2001b) IPSCs, which suggests a presynaptic change in transmitter would be consistent either with a direct effect on mossy release probability. Moreover, exposure to kainate produced 394 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 changes in the IPSC’s coefficient of variation that are consis- in the axons of interneurons. In this study, spontaneous tent with a presynaptic locus of change. Finally, they found action currents were recorded in interneurons under volt- that during treatment with kainate spontaneous miniature age clamp during exposure to 1 ␮M kainate. Frequency of IPSCs were reduced in frequency and somewhat reduced in spontaneous action currents was largely insensitive to the amplitude, although this effect on miniature IPSCs has not holding potential, which suggests that they initiated in the been confirmed by most subsequent studies (e.g. Frerking axon at a location that was electrically remote from the et al., 1998, 1999; but see Contractor et al., 2000; Behr et al., voltage-clamped somato-dendritic compartment. 2002). Whether kainate receptors are specifically localized to Rodriguez-Moreno et al. (1997) have suggested that the presynaptic terminals of inhibitory neurons that synapse kainate receptors on presynaptic GABAergic terminals on pyramidal cells remains controversial (Ben-Ari and reduce transmitter release by a G-protein-mediated activa- Cossart, 2000; Kullmann, 2001). Nicoll and coworkers tion of phospholipase C and PKC (see also Ziegra et al., (Frerking et al., 1999) have proposed that the reduction in 1992; Cunha et al., 1997, 1999, 2000). They showed that evoked IPSCs recorded in pyramidal cells during exposure pre-incubation with pertussis toxin blocked the effect of to kainate can be explained without invoking presynap- kainate on evoked IPSCs, as did exposure to the PKC in- tic kainate receptors. They suggest that GABA released hibitors staurosporine and calphostin C. It remains to be by the increase in spontaneous action potential firing acts determined whether these metabotropic actions of kainate on postsynaptic GABAA receptors to shunt postsynap- involve a direct coupling of kainate receptors to specific tic current. Frerking et al. (1999) further proposed that G-proteins or whether they result from an indirect mech- spontaneous release of GABA would activate presynaptic anism such as kainate-stimulated release of some other GABAB receptors to reduce subsequent evoked release (see endogenous compound which may then activate its own also Kerchner et al., 2001b), although Clarke et al. (1997) G-protein-coupled receptor. Evidence for such an indirect have reported that antagonism of GABAB receptors did not mechanism involving adenosine has recently been obtained prevent the reduction in evoked IPSCs in area CA1. Ad- in the striatum (see below). Rodriguez-Moreno et al. (2000) ditional postsynaptic shunting may be provided by kainate have tested for indirect modulation by other transmitters receptors expressed by the pyramidal neurons (Bureau via blockade with a cocktail of inhibitors to metabotropic et al., 1999). On the other hand, Lerma and coworkers glutamate receptors (MCPG and MPPG at 1.5 mM), opi- (Rodriguez-Moreno et al., 2000) have noted the lack of oid receptors (naloxone, 100 ␮M), GABAB receptors correlation between the frequency of spontaneous IPSCs (2-hydroxysaclofen, 50–150 ␮M), and adenosine receptors elicited by different agonists and the extent of reduction in (DPCPX, 0.1 ␮M). Their results argue against involvement evoked IPSC amplitude. In particular, they observed that of endogenous agonists for these receptors, but do not rule low doses of glutamate (3 and 10 ␮M) suppressed evoked out the possibility that some other endogenous compound transmission without any effect on the frequency of spon- may produce indirect inhibition of evoked IPSCs. taneous release events, whereas AMPA (50 ␮M) greatly Several studies have implicated kainate receptors on increased the spontaneous IPSC frequency, but had no ef- the soma and dendrites of GABAergic interneurons in the fect on the amplitude of evoked IPSCs. Kainate (0.3 and regulation of evoked IPSC amplitude. Papers by Cossart 3 ␮M) and ATPA (1 and 10 ␮M) both enhanced sIPSC fre- et al. (1998) and by Frerking et al. (1998) examined in- quency and reduced eIPSC amplitude (Rodriguez-Moreno terneurons in area CA1 of rat hippocampal slices. Both et al., 2000). Local synaptic release of glutamate by a con- studies demonstrated that activation of kainate receptors ditioning train to excitatory fibers is sufficient to inhibit caused an increase in interneuronal action potential firing. evoked IPSCs by a kainate receptor-dependent mechanism In the absence of TTX, exposure to kainate (0.25–10 ␮M) (Min et al., 1999), which would be consistent with a local increased the rate of spontaneous IPSCs recorded in pyra- effect on inhibitory terminals. Such conditioning trains also midal cells, but at the same time decreased the amplitude enhance axonal excitability in interneurons (Semyanov and of IPSCs evoked by electrical stimulation (see also Bureau Kullmann, 2001), however, so that a mechanism involving et al., 1999). ATPA (1 ␮M) was as effective as kainate in elevated spontaneous release cannot be ruled out. triggering action potentials (Cossart et al., 1998), which As noted above, several recent studies have reported an is consistent with evidence from studies of knockout mice enhancement of inhibitory transmission by exposure to low for GLUK5 expression by CA1 interneurons (Mulle et al., doses of kainate (Jiang et al., 2001; Cossart et al., 2001). 2000). Both Frerking et al. (1998) and Cossart et al. (1998) Cossart et al. (2001) recorded an increase in mIPSC fre- demonstrated a contribution by kainate receptors to post- quency in CA1 interneurons during exposure to 250 nM synaptic excitatory inputs to interneurons and argued that kainate. This enhancement of spontaneous release was not these somatodendritic kainate receptors were responsible blocked by 100 ␮M cadmium, suggesting that it did not re- for the effects of kainate on evoked transmission. In sub- quire activation of voltage-gated calcium currents, and it was sequent work on guinea pig slices, however, Semyanov and not mimicked by ATPA, suggesting that the GLUK5 subunit Kullmann (2001) have suggested that spike initiation during did not contribute (see also Mulle et al., 2000). The action exposure to kainate receptor agonists may occur primarily of kainate was blocked by CNQX (50 ␮M) but was not J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 395 affected by GYKI53655 (30 ␮M) or by a combination of the NMDA receptors, both groups observed a residual EPSC mGluR antagonists MCPG (500 ␮M) plus MSOP (100 ␮M). when mossy fibers were stimulated repetitively, but not when Cossart et al. (2001) used minimal extracellular stimulation similar stimuli were delivered to associational/commissural to evoke IPSCs in interneurons and observed that kainate inputs to CA3 cells. The residual EPSC was attributed to (250 nM) application produced a significant reduction in kainate receptors (Vignes and Collingridge, 1997; Castillo the proportion of failures in every cell tested. In contrast, et al., 1997) because it was blocked by 10–50 ␮M CNQX IPSCs evoked in CA1 pyramidal neurons displayed much but was unaffected by cyclothiazide (100 ␮M) or by antago- greater heterogeneity in their response to kainate (250 nM) nists of metabotropic glutamate receptors (1.5 mM MCPG), superfusion, with some cells (2/8) showing a significant nicotinic cholinergic receptors (10 nM methyllycaconitine), increase, some cells (2/8) showing a decrease in IPSC fail- or the ATP receptor antagonist suramin (30 ␮M). Additional ures and some cells (4/8) showing no change. Jiang et al. pharmacological analysis (Vignes et al., 1997; Bortolotto (2001) used paired recordings from interneurons and CA1 et al., 1999) indicated that three different GLUK5-selective pyramidal cells to study the effect of kainate and endoge- antagonists, LY293558, LY294486 and LY382884, reduced nous glutamate on the success rate of unitary evoked IPSCs. the amplitude of the kainate receptor-mediated EPSCs Probability of release was low in approximately half of the recorded in CA3 neurons. These results are consistent with pairs and exposure to 300 nM kainate, or release of endoge- the possibility that the postsynaptic receptors at mossy fiber nous glutamate via conditioning stimulus trains to excitatory synapses include a GLUK5 subunit (Vignes et al., 1997); fibers, caused a significant increase in transmission for these however, this interpretation does not fit well with the ex- pairs. Kainate had little effect on pairs with a high initial pression pattern for GLUK5 mRNA (Paternain et al., 2000) probability of transmission; however, exposure to 10 ␮M or with recent results in GLUK5-deficient mice, which are CNQX, but not 50 ␮M GYKI53655 or SYM2206, caused discussed in Section 3.1.3. In addition, more recent pharma- a reduction in success rate for these pairs, suggesting that cological studies with the most selective GLUK5 antagonist, activation of kainate receptors by ambient glutamate was LY382884, indicate that at mossy fiber synapses this com- partly responsible for the high success rate at these synapses pound only affects presynaptic kainate receptors on mossy (Jiang et al., 2001). Finally, Semyanov and Kullmann (2001) fiber terminals and does not block postsynaptic receptors recorded from interneurons and observed an increase in the on the CA3 pyramidal cells (Lauri et al., 2001a). The lack amplitude of spontaneous IPSCs during exposure to 1 ␮M of postsynaptic blockade by LY382884 argues against a kainate, which they attributed to multiquantal release. contribution of GLUK5 to these receptors. In addition to the extensive work that has been done on The kainate receptor-mediated EPSC at mossy fiber CA1 interneurons, Behr et al. (2002) have recently com- synapses had a relatively linear current–voltage relation, pared presynaptic effects of kainate receptor activation on suggesting that the (Q/R) sites in the channel pore were inhibitory transmission in the dentate gyrus of control and edited. Both the rise and decay times for the kainate kindled rats. In slices from control rats, application of glu- receptor-mediated EPSCs were slower than for the AMPA tamate (100 ␮M) or kainate (10 ␮M) in the presence of receptor-mediated component of transmission. This feature SYM2206 (100 ␮M) and D-APV (60 ␮M) increased the has been observed at other excitatory synapses (Kidd and frequency of TTX-sensitive spontaneous IPSCs recorded Isaac, 1999; Li et al., 1999; Bureau et al., 2000) but is in dentate granule neurons but reduced the amplitude of not well explained. Superfusion with blockers of glutamate evoked IPSCs and the frequency of TTX-resistant mIPSCs. uptake, such as 100–500 ␮M trans-PDC, did not alter the In tissue from kindled rats, which displayed a higher rest- kinetics of the kainate receptor-mediated EPSCs (Vignes ing frequency of spontaneous IPSCs, activation of kainate and Collingridge, 1997; Castillo et al., 1997), which sug- receptors decreased all three parameters: sIPSC frequency, gests that their slow kinetics does not depend on glutamate eIPSC amplitude and mIPSC frequency. The GABAB re- diffusion to an extrasynaptic location. ceptor antagonist CGP55845A (2 ␮M) had no effect on As mentioned above, postsynaptic kainate receptors depression of eIPSCs by kainate, suggesting that feedback also contribute to the EPSCs received by CA1 interneu- of spontaneously released GABA onto presynaptic GABAB rons (Frerking et al., 1998; Cossart et al., 1998). Here as receptors was not required (Behr et al., 2002). well, the rise and decay times were slower and the peak amplitude was smaller for the kainate receptor-mediated 3.1.2. Postsynaptic receptors component of the EPSC than for the component medi- In addition to presynaptic kainate receptors that may ated by AMPA receptors. Total charge transfer via the two modulate transmitter release, some cells also express post- components was similar, however, owing to the slower de- synaptic kainate receptors that can directly mediate excita- cay of kainate receptor-mediated EPSCs. Frerking et al. tory transmission. Two studies published together in 1997 (1998) observed no significant difference between these (Vignes and Collingridge, 1997; Castillo et al., 1997) de- two components of transmission in the quantity 1/CV2, sug- scribed the activation of postsynaptic kainate receptors at gesting that a similar number of release events contribute to mossy fiber synapses onto CA3 neurons in rat and guinea each component, and predicting that mEPSCs mediated by pig. After blockade of transmission mediated by AMPA and kainate receptors would be slower and smaller in amplitude 396 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 than for AMPA receptor mEPSCs. Subsequent work has by blockers for a variety of ionotropic and metabotropic confirmed this difference in hippocampus (Cossart et al., receptors, suggesting a direct metabotropic action medi- 2002) and cortex (see below, Kidd and Isaac, 1999)by ated by kainate receptors. Blockers that were tested include analysis of spontaneous miniature synaptic currents. In antagonists of conventional mGluRs (1 mM MCPG and hippocampal CA3 pyramidal neurons and CA1 interneu- 250 ␮M MSOP), GABA (100 ␮M picrotoxin and 200 ␮M rons, Cossart et al. (2002) recorded mEPSCs mediated by 2-OH-saclofen), muscarinic (1 ␮M atropine), opioid (10 ␮M AMPA receptors only, and mEPSCs involving only kainate naloxone), cannabinoid (2 ␮M AM 251) and adenosine receptors, as well as mixed mEPSCs that included pharma- (0.1 ␮M DPCPX) receptors (Melyan et al., 2002). cologically separable components mediated by AMPA and kainate receptors. In both CA3 pyramidal cells and CA1 in- 3.1.3. Transgenic mice terneurons, the mEPSCs mediated by kainate receptors had To address the function of specific kainate receptor sub- a smaller peak amplitude and slower kinetics than AMPA units, Heinemann and coworkers (Mulle et al., 1998; Mulle receptor mEPSCs, but the overall charge transfer was com- et al., 2000; Contractor et al., 2003) have generated lines of parable for the two components. Comparison of pyramidal mice in which genes for individual subunits have been dis- cells with interneurons indicated faster kinetics in the in- rupted by homologous recombination. All of the mice lack- terneurons for both AMPA and kainate receptor-mediated ing individual subunits are viable. Although the properties components of transmission (Cossart et al., 2002). In the of these mice have not yet been exhaustively assessed, even case of interneurons (Cossart et al., 2002), the decay time the initial characterization of their phenotypes has yielded constant for kainate receptor-mediated EPSCs (∼10 ms) interesting insights into the makeup of CNS kainate recep- was comparable to deactivation rates for agonist-gated cur- tors. Interpretation of experiments on knockout animals is rents, whereas in pyramidal cells the time constant of decay complicated by the possibility of developmental compensa- was much slower (∼90 ms). tion; however, studies published to date provide little evi- Frerking and Ohliger-Frerking (2002) have considered dence for compensatory changes in expression levels of the the functional implications resulting from the slower kinet- remaining kainate receptor subunits. ics for kainate receptor-mediated synaptic currents relative The first knockout study (Mulle et al., 1998) described to EPSCs mediated by AMPA receptors. Not surprisingly, the properties of mice lacking GLUK6. Autoradiography for their analysis and subsequent modeling of EPSCs in CA1 [3H]kainate binding (Mulle et al., 1998) revealed that the interneurons (Frerking and Ohliger-Frerking, 2002) suggest GLUK6-deficient animals lacked the high affinity labeling that the kainate receptor-mediated component of synaptic in hippocampal area CA3 that is prominent in wild-type current contributes a substantial tonic depolarization over animals (Monaghan and Cotman, 1982). Physiological a broad range of physiological firing frequencies, whereas recordings in slices from knockout animals showed that AMPA receptors subserve phasic or transient excitation. the potency of kainate for activation of whole-cell current Such considerations are likely to apply at other synapses in CA3 neurons was reduced by at least 10-fold relative where kainate receptors mediate components of synaptic to wild-type (change in threshold dose from 1 to 10 ␮M). current that are relatively slower in their rise and decay ki- Moreover, CA3 neurons from GLUK6-deficient animals did netics than the EPSCs mediated by AMPA receptors (Kidd not exhibit kainate receptor-mediated EPSCs when mossy and Isaac, 1999; Li et al., 1999; Bureau et al., 2000; Castillo fibers were stimulated in GYKI53655. Subsequent analysis et al., 1997; Vignes and Collingridge, 1997; Cossart et al., of area CA1 in GLUK6-deficient mice revealed a loss of 2001). whole-cell currents mediated by kainate receptors in pyrami- In contrast to CA1 interneurons, CA1 pyramidal cells do dal cells (Bureau et al., 1999). Currents mediated by kainate not exhibit a detectable component of excitatory postsynap- receptors in area CA1 interneurons were not eliminated, tic current that is mediated by kainate receptors, although however, which supports a role for GLUK5 in interneuronal application of exogenous kainate receptor agonists does kainate receptors. Depolarization of interneurons mediated evoke kainate receptor-mediated current in CA1 pyrami- by kainate receptors evoked spontaneous firing of action dal cells (Chittajallu et al., 1996; Bureau et al., 1999). In potentials and produced spontaneous inhibitory synaptic addition, exposure to low doses of kainate was shown to potentials that were recorded in pyramidal cells in both inhibit postspike potassium current that underlies the slow the wild-type and GLUK6 knockout mice (Bureau et al., afterhyperpolarization (IsAHP) in CA1 pyramids (Melyan 1999). A more recent study (Mulle et al., 2000)onGLUK5 et al., 2002). This action of kainate was mimicked by knockout mice and GLUK5/GLUK6 double knockouts has domoate (200 nM) but not ATPA (2 ␮M) or the AMPA shown that both GLUK5 and GLUK6 contribute to kainate receptor-selective agonist (S)-5-fluorowillardine (300 nM); receptors in some CA1 interneurons. it was blocked by CNQX (20 ␮M) but not by GYKI52466 Overall, the GLUK6 knockout mice showed relatively (100 ␮M). Further analysis suggested a metabotropic ba- normal behavior, including comparable learning ability to sis for this modulation. Treatment with N-ethylmaleimide wild-type animals in a water maze test (Mulle et al., 1998). or the PKC inhibitor calphostin C prevented the action of When challenged with interperitoneal kainate injections, kainate on IsAHP. The effect of kainate was not prevented however, the knockout mice displayed a higher seizure J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 397 threshold and significantly less histological evidence of ex- knockouts or GLUK5/GLUK6 double knockouts. Kainate citotoxic damage to CA3 neurons, which are known to be also had a complex effect on transmission by perforant particularly vulnerable to kainate toxicity in normal animals path inputs (Contractor et al., 2000). In wild-type mice (Nadler et al., 1978). Collectively, these results suggest that kainate potentiated these inputs whereas in slices from production of functional kainate receptors in CA3 pyrami- GLUK6-deficient, and some GLUK5-deficient, animals the dal cells, as well as CA1 pyramidal neurons, is absolutely amplitude of evoked synaptic currents was reduced by dependent on expression of the GLUK6 subunit. kainate. A slight potentiation by kainate was preserved in More recent work (Contractor et al., 2000, 2001) has some of the GLUK5-deficient animals, but kainate had no ef- used knockout mice to evaluate the roles of GLUK5 fect on perforant path transmission in the double knockouts. and GLUK6 in presynaptic modulation of inputs to CA3 Knockout mice have also been used to analyze the role hippocampal neurons. In agreement with earlier studies of the GLUK2 subunit at mossy fiber synapses (Contractor (Chittajallu et al., 1996; Kamiya and Ozawa, 1998; Schmitz et al., 2003). In contrast to GLUK6, which is required for et al., 2000), Contractor et al. (2000, 2001) observed a both pre and postsynaptic receptors at this synapse (Mulle reduction of synaptic transmission by mossy fibers and et al., 1998; Contractor et al., 2000), deletion of GLUK2 associational/commissural fibers during superfusion with left presynaptic and postsynaptic kainate receptors in place 3 ␮M kainate. This inhibition by kainate was preserved in but altered their function (Contractor et al., 2003). Postsy- GLUK5-deficient animals but was absent in GLUK6 knock- naptic kainate receptors mediated a component of evoked out mice or in GLUK5/GLUK6 double knockouts. These mossy fiber EPSCs in GLUK2 knockout slices, but with results are in agreement with anatomical studies (Bureau a faster rate of decay than was observed in cells from et al., 1999; Paternain et al., 2000) showing low expression wild-type animals. Reductions in transmission attributed to of GLUK5 in dentate granule cells and CA3 pyramidal cells activation of presynaptic kainate receptors persisted in the by in situ hybridization, but they are at odds with the work GLUK2-deficient mice; however the potentiation of trans- on GLUK5-selective drugs, described above (Vignes et al., mission by low doses of kainate or by the heterosynaptic 1998; Bortolotto et al., 1999; Lauri et al., 2001a), which sug- release of glutamate was absent from the GLUK2 knockout gested that the presynaptic kainate receptor on mossy fiber slices. Contractor et al. (2003) suggest that all of these terminals included a GLUK5 subunit. Either the granule cells changes would be consistent with a reduction in the affinity produce heteromeric receptors that lack the GLUK5 subunit, of kainate receptors upon deletion of GLUK2. Interestingly, but which are nevertheless sensitive to GLUK5-selective the GLUK2 knockout slices showed no significant deficits decahydroisoquinoline antagonists, or the neurons actually in mossy fiber homosynaptic frequency facilitation or LTP produce sufficient GLUK5 subunits to render their receptors (Contractor et al., 2003), both of which involve activation sensitive to these drugs, but the cells fail to produce func- of presynaptic kainate receptors (Lauri et al., 2001a,b; tional homomeric GLUK5 receptors in animals that lack Contractor et al., 2001; Schmitz et al., 2001a). GLUK6 (Huettner, 2001; Lauri et al., 2003). For example, In addition to knockout lines, Heinemann and cowork- it could be imagined that GLUK5 splice variants which ers have also used homologous recombination to generate function poorly on their own (Sommer et al., 1992) might animals with altered editing at the Q/R sites of GLUK5 still contribute to heteromeric receptors in combination (Sailer et al., 1999) and GLUK6 (Vissel et al., 2001). In with GLUK6. Alternatively, GLUK6 might be required for adult wild-type animals approximately 50% of mRNAs for the correct subcellular delivery of heteromeric receptors to the GLUK5 subunit have been edited to encode an arginine presynaptic terminals (Huettner, 2001). Direct experiments residue at the Q/R site. For the GLUK6 subunit approx- with GLUK5-selective compounds on knockout animals imately 75% of mRNA is edited. For both subunits, the should eventually resolve these different possibilities. level of editing increases during development (Bernard and Contractor et al. (2000) also examined the effect of 3 ␮M Khrestchatisky, 1994; Paschen et al., 1997). Sailer et al. kainate on mini frequency in CA3 pyramidal neurons and (1999) created homozygous mice encoding an R at the Q/R on transmission to these cells by perforant path inputs. Ear- position of GLUK5. These mice showed reduced kainate lier work by Castillo et al. (1997) in guinea pig CA3 cells receptor-mediated current amplitudes in DRG cells but reported no change in mEPSC frequency with kainate ap- were surprisingly normal in their overall behavior and in plications (0.3 to 3 ␮M); however, Contractor et al. (2000) their susceptibility to kainate-induced seizures. In contrast, observed a significant increase in the frequency of mEPSCs Vissel et al. (2001) created homozygous mice in which in both wild-type and GLUK6-deficient mice. Interestingly, editing of the GLUK6 subunit was prevented. These mice such an increase was not obtained in GLUK5-deficient appeared normal in a battery of behavioral tests; however, animals. Moreover, superfusion of 200 ␮M cadmium was their susceptibility to kainate-induced seizures was signifi- able to block the increase produced by kainate in wild-type cantly increased relative to wild-type. Studies of hippocam- mice, suggesting a possible role for voltage-gated calcium pal neurons in vitro demonstrated more robust elevation of channels in the effect on mini frequency. With Cd present, cytosolic calcium upon exposure to kainate in cells lack- 3 ␮M kainate actually reduced the frequency of minis in ing GLUK6(R). In addition, NMDA receptor-independent wild-type and GLUK5-deficient animals, but not in GLUK6 LTP could be induced at the medial perforant path synapse 398 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 onto dentate granule cells of GLUK6(Q) mice but not in in GLUK6-deficient tissue as compared to wild-type. To wild-type. explain the ability of GLUK5-selective antagonists to block mossy fiber LTP, Lauri et al. (2003) propose that dentate 3.1.4. Synaptic plasticity granule cells normally express heteromeric kainate recep- Studies of transmission onto CA3 neurons provided the tors that include both the GLUK5 and GLUK6 subunits; first direct evidence that kainate receptors may contribute to functional receptors continue to be expressed in animals synaptic plasticity (Bortolotto et al., 1999). Previous work lacking GLUK5, but not when GLUK6 is deleted. (Harris and Cotman, 1986; Zalutsky and Nicoll, 1990) had shown that associational/commissural fiber inputs to CA3 3.2. Cortex neurons exhibit conventional NMDA receptor-dependent LTP, whereas LTP of mossy fiber inputs is not sensitive Kidd and Isaac (1999) have demonstrated a contribu- to NMDA antagonists. Additional studies (Castillo et al., tion by postsynaptic kainate receptors to thalamocortical 1994; Ito and Sugiyama, 1991) suggested that blockade transmission. Their results, obtained in slices from 3- to of non-NMDA receptors with the broad-spectrum antago- 8-day-old rats, indicate that kainate receptors are present nists CNQX or kynurenic acid did not prevent induction of at these synapses early in development. They recorded mossy fiber LTP; however, Bortolotto et al. (1999) found from layer IV neurons in primary somatosensory cortex. that high doses of these antagonists (10 mM kynurenic acid In most cases, evoked EPSCs displayed a component me- or 10 ␮M CNQX), as well as the GLUK5-selective antag- diated by AMPA and by kainate receptors, although some onist LY382884 (10 ␮M), were capable of blocking LTP examples of AMPA- and kainate-only evoked EPSCs were induction at mossy fiber synapses under their recording observed. Analysis of spontaneous synaptic events revealed conditions. In contrast, LY382884 had no effect on LTP two distinct populations: one with a fast decay typical of induction at associational/commissural synapses onto CA3 AMPA receptors and the other with a much slower de- cells (Bortolotto et al., 1999). There has been disagree- cay rate, which matched the slow decay of evoked kainate ment over the dose-dependence of LTP-induction blockade receptor-mediated synaptic currents. The slow kinetics of by non-NMDA receptors antagonists (Nicoll et al., 2000; events mediated by kainate receptors apparently does not Bortolotto et al., 2000; Schmitz et al., 2001b),butitis arise from the slow diffusion of transmitter from sites of generally agreed (Schmitz et al., 2001a; Contractor et al., release. Manipulations designed to alter glutamate uptake, 2001; Lauri et al., 2001a,b) that presynaptic kainate re- including temperature elevation and incubation with trans- ceptors mediate a significant component of short-term porter antagonists, produced equivalent effects on AMPA frequency-dependent facilitation of mossy fiber transmis- and kainate receptor-mediated components (Kidd and Isaac, sion (see above, Section 3.1.1.1). Lauri et al. (2001a) pro- 2001). Collectively, these results suggest that kainate re- vided evidence that activation of these presynaptic kainate ceptors and AMPA receptors may be segregated to distinct receptors was required for the induction of mossy fiber LTP, synaptic contacts and that the AMPA receptors progressively and that establishment of LTP occluded the facilitation that replace kainate receptors during the course of development. was sensitive to kainate receptor antagonists (Lauri et al., Induction of LTP by a pairing protocol led to a reduction 2001a; Ji and Staubli, 2002). in the component of transmission mediated by kainate re- More recently, Lauri et al. (2003) have implicated ceptors, coupled with a significant increase in the AMPA calcium-induced calcium release from intracellular stores receptor-mediated component (Kidd and Isaac, 1999). as an important component of mossy fiber LTP. Their re- In addition to postsynaptic kainate receptors, some tha- sults (Lauri et al., 2003) suggest that the necessary trigger lamocortical synapses may exhibit presynaptic kainate re- for release from internal stores can be provided by calcium ceptors that modulate release. Kidd et al. (2002) tested for entry either through calcium-permeable kainate receptors or presynaptic modulation by kainate receptor agonists at tha- through voltage-gated calcium channels, with entry through lamic inputs to barrel cortex. In slices from early postnatal voltage-gated channels being the dominant factor when rats (P3 to P5) repeated stimulation of thalamocortical inputs extracellular calcium levels are high (∼4 mM). This study at frequencies ranging from 10 to 100 Hz caused significant helps to resolve earlier disagreements over the ability of depression of EPSCs recorded in layer IV neurons. Exposure kainate receptor antagonists to block LTP induction at this to ATPA produced a similar reduction in EPSC amplitude. synapse (Nicoll et al., 2000; Bortolotto et al., 2000), be- Application of LY382884 (10 ␮M) reduced the depression cause experiments that appeared to be in conflict were per- observed at high frequencies (50 and 100 Hz), but caused formed under different ambient calcium levels. Additional minimal change at lower frequencies, leading Kidd et al. evidence supporting a role for kainate receptors in mossy (2002) to propose two components of depression: a rapidly fiber LTP induction has come from analysis of knock- decaying component, mediated by kainate receptors, that out mice (Contractor et al., 2001). This work (Contractor was only apparent with high frequency stimulation, and a et al., 2001), however, indicated a requirement for the slower component of depression that did not involve kainate GLUK6 subunit. Mossy fiber LTP was normal in slices receptors. Interestingly, the kainate receptor-mediated presy- taken from GLUK5 knockout mice but significantly reduced naptic modulation was not detected in slices from slightly J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 399 older animals (P7 to P8), suggesting a developmental change tion period (homosynaptic potentiation), but instead was in sensory processing that involves a change in kainate re- generalized to other inputs (heterosynaptic potentiation), ceptor expression or properties (Kidd et al., 2002). including inputs from the basal amygdala (Li et al., 2001). Evidence for presynaptic kainate receptors has also been Inhibitory transmission in the amygdala is also subject to obtained (Ali et al., 2001) at synapses formed by inhibitory regulation by kainate receptors. Braga et al. (2003) recorded interneurons onto layer V pyramidal cells in rat motor cor- evoked IPSCs from pyramidal cells in the basolateral amyg- tex (P17 to P22). Paired recordings demonstrated a reduc- dala, observing a reduction in failures with exposure to tion in evoked IPSCs during exposure to ATPA (1 ␮M) or low doses of ATPA (300 nM) or glutamate (5 ␮M), and an glutamate (10 ␮M). This action was inhibited by 30 ␮M increase in failures with higher agonist doses (1–10 ␮M CNQX, but was not blocked by GYKI53655 (50 ␮M) or by ATPA, 30–200 ␮M glutamate). These effects were recorded inhibitors of mGluRs (1 mM MCPG and 100 ␮M CPPG), in the chronic presence of APV (50 ␮M), GYKI53655 GABAB (100 ␮M CPG 55845), opioid (100 ␮M naloxone), (50 ␮M), SCH50911 (20 ␮M) and CPCCOEt (30 ␮M) to muscarinic (50 ␮M atropine), and adenosine (1 ␮M DPCPX) block NMDA, AMPA, GABAB and group I mGluRs. A receptors (Ali et al., 2001). Changes in the failure rate and similar bi-directional modulation was also observed for coefficient of variation produced by kainate receptor ac- mIPSCs recorded in the presence of TTX; 300 nM ATPA tivation were consistent with a presynaptic change in re- enhanced mIPSC frequency, whereas higher doses reduced lease. Interestingly, depolarization of the postsynaptic cell the frequency (Braga et al., 2003). Effects of ATPA and to −40 mV for 1.5 s produced a CNQX-sensitive reduction glutamate were blocked by LY293558. Moreover, exposure in evoked IPSCs that Ali et al. (2001) attributed to dendritic to this antagonist reduced evoked IPSC amplitude and in- release of glutamate and subsequent activation of presynap- creased failures, suggesting that endogenous glutamate may tic kainate receptors. provide tonic activation of kainate receptors in the slice under control conditions (Braga et al., 2003). 3.3. Amygdala 3.4. Retina Li and Rogawski (1998) have reported on synaptic transmission mediated by kainate receptors in the basolateral In the ground squirrel retina (DeVries and Schwartz, amygdala. They used intracellular electrodes to record from 1999; DeVries, 2000), kainate receptors mediate transmis- cells in acute slices. EPSPs were evoked by stimulation of sion between cone photoreceptors and specific classes of either the external capsule or the basal amygdala. Superfu- “off” bipolar cells. Using paired recordings from cones and sion with 50 ␮M GYKI52466 or 53655 completely blocked bipolar cells, DeVries and Schwartz (1999) observed that responses to basal amygdala stimulation, but caused only step depolarization of the cone from −70 to 0 mV evoked a partial blockade of EPSPs evoked by external capsule stim- transient excitatory current in the postsynaptic bipolar cell. ulation. The residual EPSP displayed temporal summation This synaptic response was blocked by CNQX (30 ␮M), during brief 50 Hz trains. It was blocked almost completely but was insensitive to GYKI53655 (25 ␮M), cyclothiazide by addition of 20 ␮M CNQX, 10 ␮M LY293558, or 20 ␮M (200 ␮M) and APV. During prolonged depolarizations, the LY377770 to the bath (Li and Rogawski, 1998; Li et al., synaptic current decayed rapidly (τ ∼ 5 ms) to reach a 2001). Subsequent work by Li et al. (2001) has demonstrated steady state level that was less than 5% of the initial peak involvement of kainate receptors in homosynaptic and het- amplitude. Similar desensitizing currents were observed erosynaptic potentiation of transmission in the amygdala. In when glutamate was applied rapidly to isolated bipolar cells. contrast to studies of LTP in the hippocampus, which nor- Strong desensitization to glutamate is typical of kainate mally elicit potentiation with brief high-frequency stimulus receptors, and may help to reduce the cost and maintain trains, Li et al. (2001) observed progressive enhancement the sensitivity of tonic signaling at this synapse. Both of transmission during prolonged (15 min) low frequency rod and cone photoreceptors are depolarized in the dark, (1 Hz) stimulation. Potentiation was blocked by antagonists causing ongoing release of transmitter onto second-order with selectivity for the GLUK5 subunit of kainate receptors cells. Exposure to light produces a graded hyperpolariza- (LY377770 and LY382884) but not by antagonists to NMDA tion of photoreceptors and a decrease in transmitter re- (100 ␮M APV), AMPA (50 ␮M GYKI53655) or group lease. Because of the high input resistance of bipolar cells I metabotropic (20 ␮M CPCCOEt) glutamate receptors. (∼1G), DeVries and Schwartz (1999) note that even the Potentiation was mimicked by brief (10min) superfusion relatively small synaptic currents that are evoked through with the GLUK5-selective agonist ATPA (20 ␮M) and was desensitized kainate receptors (approximately −10 pA), blocked by calcium-free medium or by pre-equilibration when summed over inputs from several cones, will be with the cell-permeable chelator BAPTA-AM. In addition, sufficient to drive the bipolar cell membrane potential both the NMDA and AMPA receptor-mediated components through its normal 25 mV range of operation. Moreover, of transmission were potentiated, suggesting a presynap- the properties of kainate receptor desensitization will en- tic change in glutamate release. Interestingly, potentiation sure that postsynaptic responses in off-bipolar cells will was not restricted to the fibers stimulated during the induc- be greatest for abrupt reductions in light intensity follow- 400 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 ing a prolonged exposure to light (DeVries and Schwartz, is consistent with a presynaptic decrease in release probabil- 1999). ity. In this case, the effects of kainate were not prevented by In subsequent work, DeVries (2000) showed that two dif- metabotropic glutamate receptor antagonists (1 mM MCPG ferent morphologically identified subtypes of off bipolars plus 100 ␮M CPPG), or by GABAB (20 ␮M SCH50911) cells, b3 and b7 cells, exhibited kainate receptors with dis- or adenosine (200 ␮M theophylline) receptor antagonists tinct physiological properties. In both cell types kainate re- (Crowder and Weiner, 2002). Modulation of excitatory ceptors were activated by ATPA, suggesting a contribution transmission was preserved in single knockouts of either by the GLUK5 subunit; however, the b3 class of off bipolars GLUK5 or GLUK6, but was eliminated in double knockouts expressed kainate receptors with slow recovery from desen- of both GLUK5 and GLUK6 (Casassus and Mulle, 2002). sitization to glutamate (t1/2 ∼ 1 s), whereas b7 bipolar cells recovered from desensitization more quickly (t1/2 ∼ 0.2 s). 3.6. Hypothalamus Transmission from cones onto a third type of off bipolar cell (b2) was found to be mediated by AMPA receptors (DeVries, In a study of rat hypothalamic neurons Liu et al. (1999) 2000). observed an increase in spontaneous IPSC frequency during superfusion with kainate. Recordings from hypothala- 3.5. Striatum mic neurons in acute tissue slices revealed a 50% increase in IPSC frequency during perfusion with 1 ␮M kainate plus Kainate receptor subunit expression (Bischoff et al., 1997; APV (50 ␮M) and GYKI52466 (100 ␮M). In cultured hy- Bahn et al., 1994) and kainate binding (Monaghan and pothalamic neurons 10 ␮M kainate caused a similar 50% in- Cotman, 1982) are prominent in the striatum. The majority crease in the frequency, but no change in the amplitude, of of projection cells express GLUK6 (Chergui et al., 2000). mIPSCs recorded in the presence of TTX and GYKI52466, An early study (Calabresi et al., 1990) observed responses suggesting a presynaptic site of action. Kainate also caused to kainate at submicromolar concentrations that are likely to a 7–40 pA increase in the holding current at −70 mV and have been selective for kainate receptors. More recent work enhanced the frequency of spontaneous evoked IPSCs in (Chergui et al., 2000) with AMPA-selective antagonists the absence of TTX, indicating the likely presence of soma- and GLUK6-deficient mice has confirmed the expression of todendritic kainate receptors in these neurons as well (Liu functional kainate receptors by cells in slices of striatum. No et al., 1999). evidence has been obtained for kainate receptor-mediated excitatory postsynaptic currents in striatal cells (Chergui 3.7. Cerebellum et al., 2000); however, modulation of GABAergic transmission was observed upon activation of kainate receptors A variety of cell types within the cerebellum express with exogenous agonist. During superfusion with kainate kainate receptors, with each cell type showing a distinct pat- receptor agonists, IPSCs evoked by local stimulation were tern of subunit expression. Purkinje cells express the GLUK5 reduced in amplitude (Chergui et al., 2000). Although and GLUK1 subunits (Wisden and Seeburg, 1993). Kainate kainate receptor agonists elicited depolarization and firing receptor-mediated currents have been recorded in Purkinje of action potentials in striatal cells, these effects did not cells (Renard et al., 1995; Brickley et al., 1999), although the appear to directly underlie the change in evoked GABAer- function of these receptors remains obscure. Granule cells gic transmission. Instead, much of the reduction in evoked express GLUK5, GLUK6 and GLUK2 even before they mi- transmission could be prevented by adenosine A2A receptor grate to their final destination in the internal granule layer antagonists, suggesting that exposure to kainate stimulated between postnatal days 7–14 (Belcher and Howe, 1997; the release of an endogenous agonist for these receptors Pemberton et al., 1998). Analysis of granule cell mRNA from within the slice (Chergui et al., 2000). by RT-PCR (Belcher and Howe, 1997) indicates that at all Two studies have examined kainate receptors in the ven- stages the Q/R site of GLUK5 is largely unedited, whereas tral striatum or nucleus accumbens (Crowder and Weiner, that of GLUK6 is predominantly in the edited form. Overall, 2002; Casassus and Mulle, 2002). As discussed above for the level of editing increases during postnatal development, dorsal striatum, most neurons in the nucleus accumbens as does expression of the major enzymes responsible for possessed functional kainate receptors that depended on ex- editing. Premigratory granule cells in the external granule pression of the GLUK6 subunit (Casassus and Mulle, 2002). layer exhibit kainate receptor-mediated currents following No evidence was obtained for a contribution by postsynap- exposure to Con A, but currents mediated by AMPA recep- tic kainate receptors at excitatory synapses (Crowder and tors were small or absent at this stage (Smith et al., 1999). Weiner, 2002; Casassus and Mulle, 2002); however, excita- More mature cells in the internal granule layer express both tory inputs to accumbens neurons were inhibited by exposure kainate and AMPA receptors. to low concentrations of kainate (0.3–1 ␮M). This decline A recent paper by Delaney and Jahr (2002) provides in excitatory transmission was associated with a reduction evidence for presynaptic kainate receptors at granule cell in 1/CV2 (Casassus and Mulle, 2002) and an increase in synapses onto Purkinje cells and stellate cells. In both paired-pulse facilitation (Crowder and Weiner, 2002), which cases, activation of these presynaptic receptors by low doses J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 401 of exogenous agonist (5 nM domoate), or low frequency and partially prevented by ␻-conotoxin GVIA (0.5 ␮M) and stimulation (10–20 Hz), resulted in facilitation of release. MVIIC (0.5 ␮M), which are inhibitors of N and P/Q type Higher agonist doses (10–500 nM domoate) or higher calcium channels. These properties suggest a mechanism frequency stimulation (100 Hz) produced synaptic depres- involving depolarization of presynaptic terminals leading sion at synapses onto stellate cells. In contrast, kainate to calcium entry through voltage-gated channels. Although receptor-mediated facilitation of parallel fiber to Purkinje kainate increased mIPSC frequency, under most conditions cell synapses was observed at all stimulus frequencies tested exposure to kainate caused a net reduction of evoked trans- (up to 100 Hz) and for exposure to 50 nM domoate; 500 nM mitter release. This effect on evoked inhibitory transmission domoate produced depression of granule cell to Purkinje was mimicked and occluded by the GABAB receptor agonist cell synapses. baclofen (5 ␮M) and was blocked by the GABAB antago- Both GLUK5 and GLUK6, but not GLUK7 or GLUK1 or nist CGP55845 (10 ␮M), suggesting that local elevation of GLUK2, have been detected in cerebellar Golgi cells (Bureau GABA by an increase in spontaneous quantal release was et al., 2000), which are interneurons with cell bodies in sufficient to activate presynaptic GABAB receptors and sup- the granule cell layer. Kainate application evokes current in press evoked transmission. Interestingly, the activation of these cells and kainate receptors mediate a component of GABAB receptors had a selective action on evoked release, excitatory postsynaptic current produced by stimulation of causing no decrease in mIPSC frequency under resting con- parallel fibers. Both the agonist-evoked and synaptic currents ditions, or when the frequency was elevated by exposure are absent in GLUK6 knockout mice (Bureau et al., 2000). to kainate or KCl (Kerchner et al., 2001b). Recordings in acute spinal slices demonstrated that endogenous glutamate, 3.8. Spinal cord released upon stimulation of primary afferent fibers, caused suppression of eIPSCs via activation of kainate receptors. Kainate receptor expression is not particularly prominent This effect of primary afferent stimulation was blocked by in the spinal cord, as judged by in situ hybridization (Tölle CNQX (20 ␮M) but not SYM2206 (100 ␮M). In addition, et al., 1993) or immunocytochemistry (Petralia et al., 1994); it was prevented by the GABAB antagonist CGP55845 however, several studies have provided physiological evi- (10 ␮M), which is consistent with the feedback mechanism dence for functional kainate receptors in spinal neurons (Li delineated in cell culture (Kerchner et al., 2001b). et al., 1999; Kerchner et al., 2001a; Wilding and Huettner, 2001). As in the hippocampus, primary afferent synapses 3.9. Dorsal root ganglia in the dorsal horn of the spinal cord may represent an example where both pre and postsynaptic kainate receptors Kainate receptors were first described on sensory axons in may play a role in transmission at individual contacts. Evi- peripheral nerve. Studies by Evans and coworkers (Davies dence for postsynaptic kainate receptors was obtained by Li et al., 1979; Agrawal and Evans, 1986) demonstrated depo- et al. (1999) in studies of rat spinal cord slices. Whole-cell larization of dorsal root fibers by exposure to kainate, which recording in layer II of the dorsal horn revealed a slow com- displayed a different pharmacology from the depolarization ponent of primary afferent transmission that was insensi- of spinal neurons by excitatory amino acids (see also Lee tive to 100 ␮M GYKI53655 or SYM2206, but was inhibited et al., 2002). These results led Watkins and Evans (1981) by 10 ␮M CNQX and by selective desensitization with 1 to propose the existence of a unique class of receptors se- to 5 ␮M SYM2081. This kainate receptor-mediated compo- lective for kainate. Further analysis in this system showed nent of transmission was most pronounced for stimulation that treatment with kainate also produced selective block- intensities that were sufficient to activate high threshold A ade of action potential conduction along C-fibers (Agrawal delta and C fibers. Lower intensity stimulation evoked EP- and Evans, 1986), which are the small diameter, high thresh- SCs that were entirely mediated by AMPA receptors. In con- old afferents that convey nociceptive and thermoreceptive trast to the hippocampus and amygdala, however, kainate information into the CNS. Subsequent patch-clamp stud- receptor-mediated EPSCs in the dorsal horn showed depres- ies (Huettner, 1990; Wong and Mayer, 1993; Lee et al., sion during brief stimulation trains. Exposure to Con A re- 2001) on cells isolated from dorsal root ganglia (DRG) duced the extent of depression, which suggests that it may confirmed the expression of functional kainate receptors involve the desensitization of postsynaptic kainate receptors. on the small diameter sensory neurons that give rise to More recent work (Kerchner et al., 2001b) has provided C-fibers. evidence for expression of presynaptic kainate receptors Several studies have attempted to address the role that by spinal cord inhibitory interneurons. In the presence of DRG cell kainate receptors serve. It has been proposed that TTX, application of kainate (10 ␮M) or glutamate (30 ␮M) kainate receptors function as presynaptic autoreceptors at to cultured rat dorsal horn neurons increased the frequency primary afferent synapses in the dorsal horn (Agrawal and of mIPSCs by 8–10-fold, suggesting a direct action on Evans, 1986; Hwang et al., 2001). There also is evidence for presynaptic nerve terminals. This elevation of release re- routing of kainate receptors to the peripheral axon branch quired extracellular Na+ and Ca2+. It was blocked by the (Agrawal and Evans, 1986; Ault and Hildebrand, 1993; non-selective calcium channel antagonist Cd2+ (50 ␮M) Coggeshall and Carlton, 1998) where they might function 402 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 as sensory receptors, detecting glutamate released following death (Nadler et al., 1978). Recent experiments by Smolders tissue damage. et al. (2002) suggest that GLUK5-selective antagonists can In a study of cultured DRG cells, Lee et al. (1999) ob- block the induction of seizures by pilocarpine or electri- served that application of 100 ␮M kainate evoked action cal stimulation, as well as suppress pre-established seizure potentials that were recorded in the cell body. In record- activity. Another recent study of seizure propagation from ings from neurons in lamina II of spinal cord slices, this one hippocampus to the contralateral hippocampus found study showed an increase in spontaneous postsynaptic cur- that ATPA (1 ␮M) produced a paradoxical anti-epileptic rents during superfusion with 10 ␮M kainate. More recently, effect, which was attributed to preferential activation of in- Kerchner et al. (2001a) have shown both in culture and in terneurons (Khalilov et al., 2001). In addition, several stud- slices that evoked transmission from DRG cells to spinal dor- ies have provided evidence for antinociceptive effects by sal horn neurons was inhibited by exposure to 10 ␮M kainate GLUK5-selective compounds (Procter et al., 1998; Simmons or to the GLUK5-selective agonist ATPA (2 ␮M). Kainate et al., 1998; Stanfa and Dickenson, 1999; Sutton et al., 1999; receptors expressed by rat DRG neurons were shown to be Mascias et al., 2002), supporting a role for kainate receptors selectively desensitized by ATPA (Kerchner et al., 2001a; in pain sensation (Ruscheweyh and Sankuhler, 2002). Wilding and Huettner, 2001), which had little effect on neu- Much remains to be learned about the structural basis rons from rat dorsal horn. Further analysis of knockout mice for kainate receptor heterogeneity and the mechanisms that (Kerchner et al., 2002) confirmed this differential sensitivity underlie the metabotropic actions that have been attributed to ATPA, but also provided evidence that both GLUK5 and to kainate receptors. In addition, the slow time course of GLUK6 contribute to the receptors expressed by DRG and EPSCs mediated by postsynaptic kainate receptors remains spinal neurons. poorly explained. In a few cases, the time course of decay is comparable to receptor deactivation (Cossart et al., 2002; DeVries and Schwartz, 1999), but in other cells, the EP- 4. Perspectives SCs decay at a much slower rate. All of the ongoing work on kainate receptors would be aided by the development In many regions of the nervous system, the involvement of of more potent and more selective antagonists. Compounds presynaptic and/or postsynaptic kainate receptors in synaptic selective for GLUK5 have yielded a great deal of informa- transmission is now firmly established. Although controver- tion (Bleakman et al., 1996b; O’Neill et al., 1998), but the sies and uncertainties remain, a number of general themes field still lacks a selective antagonist that blocks GLUK6 or have emerged. First, activation of kainate receptors produces the other kainate receptor subunits (Bleakman and Lodge, bi-directional modulation of both excitatory and inhibitory 1998). Ever since the pioneering studies by Watkins and transmission. Low to moderate activation enhances trans- coworkers (Watkins and Evans, 1981), progress in elucidat- mission, whereas stronger activation inhibits transmission. ing the function of glutamate receptors has hinged on the Second, synaptic currents mediated by postsynaptic kainate discovery of new pharmacological tools. This trend appears receptors exhibit lower peak amplitude and slower decay certain to continue in the case of kainate receptors. kinetics than AMPA receptor-mediated EPSCs recorded in the same cell. Total charge delivery by AMPA and kainate receptor-mediated EPSCs may be comparable. Acknowledgements Third, synaptic kainate receptors may play a strictly ionotropic role, or in some cases may trigger a metabotropic I am grateful to the NIH for supporting my research cascade. The mechanistic links between known kainate (NS30888) and to Geoff Swanson, Tim Wilding and Geoff receptor subunits and metabotropic effectors have not yet Kerchner for careful reading of the manuscript. been completely defined. Fourth, knockout mice have provided valuable information about the composition of native receptors and the function of individual subunits, although References interpretation of these studies is complicated by the possibility of compensation. Future experiments will likely Agrawal, S.G., Evans, R.H., 1986. The primary afferent depolarizing extend the roster of synapses where kainate receptors play action of kainate in the rat. Br. J. Pharmacol. 87, 345–355. a role. In addition, future work should broaden our un- Ali, A.B., Rossier, J., Staiger, J.F., Audinat, E., 2001. Kainate receptors regulate unitary IPSCs elicited in pyramidal cells by fast-spiking derstanding of the natural activity patterns that best elicit interneurons in the neocortex. J. Neurosci. 21, 2992–2999. kainate receptor activation by endogenous glutamate and Armstrong, N., Gouaux, E., 2000. Mechanisms for activation and shed more light on the consequences of such activation on antagonism of an AMPA-sensitive glutamate receptor: crystal structures the operation of individual synapses and neuronal networks. of the GluR2 ligand binding core. Neuron 28, 165–181. Clinical interest in kainate receptors seems certain to in- Armstrong, N., Sun, Y., Chen, G.Q., Gouaux, E., 1998. Structure of a glutamate-receptor ligand-binding core in complex with kainate. Nature crease. There is substantial evidence that these receptors con- 395, 913–917. tribute to the well-known ability of kainate to elicit seizures Ault, B., Hildebrand, L.M., 1993. Activation of nociceptive reflexes by (Ben-Ari and Cossart, 2000) and cause excitotoxic cell peripheral kainate receptors. J. Pharmacol. Exp. Ther. 265, 927–932. J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 403

Bahn, S., Volk, B., Wisden, W., 1994. Kainate receptor gene expression Bureau, I., Dieudonne, S., Coussen, F., Mulle, C., 2000. Kainate in the developing rat brain. J. Neurosci. 14, 5525–5547. receptor-mediated synaptic currents in cerebellar Golgi cells are not Bähring, R., Bowie, D., Benveniste, M., Mayer, M.L., 1997. Permeation shaped by diffusion of glutamate. Proc. Natl. Acad. Sci. U.S.A. 97, and block of rat GluR6 glutamate receptor channels by internal and 6838–6843. external polyamines. J. Physiol. (Lond.) 502, 575–589. Burnashev, N., Zhou, Z., Neher, E., Sakmann, B., 1995. Fractional calcium Behr, J., Gebhardt, C., Heinemann, U., Mody, I., 2002. Kindling enhances currents through recombinant GluR channels of the NMDA, AMPA kainate receptor-mediated depression of GABAergic inhibition in rat and kainate receptor subtypes. J. Physiol. 485, 403–418. granule cells. Eur. J. Neurosci. 16, 861–867. Burnashev, N., Villarroel, A., Sakmann, B., 1996. Dimensions and ion Belcher, S.M., Howe, J.R., 1997. Characterization of RNA editing of the selectivity of recombinant AMPA and kainate receptor channels and glutamate-receptor subunits GluR5 and GluR6 in granule cells during their dependence on Q/R site residues. J. Physiol. 496, 165–173. cerebellar development. Mol. Brain Res. 52, 130–138. Calabresi, P., De Murtas, M., Mercuri, N.B., Bernardi, G., 1990. Ben-Ari, Y., Cossart, R., 2000. Kainate, a double agent that generates Kainic acid on neostriatal neurons intracellularly recorded in vitro: seizures: two decades of progress. Trends Neurosci. 23, 580–587. electrophysiological evidence for differential neuronal sensitivity. J. Bennett, J.A., Dingledine, R., 1995. Topology profile for a glutamate Neurosci. 10, 3960–3969. receptor: three transmembrane domains and a channel-lining reentrant Casassus, G., Mulle, C., 2002. Functional characterization of kainate membrane loop. Neuron 14, 373–384. receptors in the mouse nucleus accumbens. Neuropharmacology 42, Bernard, A., Khrestchatisky, M., 1994. Assessing the extent of RNA 603–611. editing in the TMII regions of GluR5 and GluR6 kainate receptors Castillo, P.E., Weisskopf, M.G., Nicoll, R.A., 1994. The role of Ca2+ during rat brain development. J. Neurochem. 62, 2057–2060. channels in hippocampal mossy fiber synaptic transmission and Bettler, B., Boulter, J., Hermans-Borgmeyer, I., O’Shea-Greenfield, A., long-term potentiation. Neuron 12, 261–269. Deneris, E., Moll, C., Borgmeyer, U., Hollmann, M., Heinemann, S., Castillo, P.E., Malenka, R.C., Nicoll, R.A., 1997. Kainate receptors 1990. Cloning of a novel glutamate receptor subunit, GluR5: expression mediate a slow postsynaptic current in hippocampal CA3 neurons. in the nervous system during development. Neuron 5, 583–595. Nature 388, 182–186. Bettler, B., Egebjerg, J., Sharma, G., Pecht, G., Hermans-Borgmeyer, I., Cauli, B., Porter, J.T., Tsuzuki, K., Lambolez, B., Rossier, J., Quenet, B., Moll, C., Stevens, C.F., Heinemann, S., 1992. Cloning of a putative Audinat, E., 2000. Classification of fusiform neocortical interneurons glutamate receptor: a low affinity kainate-binding subunit. Neuron 8, based on unsupervised clustering. Proc. Natl. Acad. Sci. U.S.A. 97, 257–265. 6144–6149. Bischoff, S., Barhanin, J., Bettler, B., Mulle, C., Heinemann, S., 1997. Chergui, K., Bouron, A., Normand, E., Mulle, C., 2000. Functional GluR6 Spatial distribution of kainate receptor subunit mRNA in the mouse kainate receptors in the striatum: indirect downregulation of synaptic basal ganglia and ventral mesencephalon. J. Comp. Neurol. 379, 541– transmission. J. Neurosci. 20, 2175–2182. 562. Chittajallu, R., Vignes, M., Dev, K.K., Barnes, J.M., Collingridge, G.L., Bleakman, D., Lodge, D., 1998. Neuropharmacology of AMPA and Henley, J.M., 1996. Regulation of glutamate release by presynaptic kainate receptors. Neuropharmacology 37, 1187–1204. kainate receptors in the hippocampus. Nature 379, 78–81. Bleakman, D., Ballyk, B.A., Schoepp, D.D., Palmer, A.J., Bath, C.P., Chittajallu, R., Braithwaite, S.P., Clarke, V.R., Henley, J.M., 1999. Sharpe, E.F., Woolley, M.L., Bufton, H.R., Kamboj, R.K., Tarnawa, Kainate receptors: subunits, synaptic localization and function. Trends I., Lodge, D., 1996a. Activity of 2,3-benzodiazepines at native rat and Pharmacol. Sci. 20, 26–35. recombinant human glutamate receptors in vitro: stereospecificity and Clarke, V.R., Collingridge, G.L., 2002. Characterization of the effects selectivity profiles. Neuropharmacology 35, 1689–1702. of ATPA, a GLU(K5) receptor-selective agonist, on excitatory Bleakman, D., Schoepp, D.D., Ballyk, B., Bufton, H., Sharpe, E.F., synaptic transmission in area CA1 of rat hippocampal slices. Thomas, K., Ornstein, P.L., Kamboj, R.K., 1996b. Pharmacological Neuropharmacology 42, 889–902. discrimination of GluR5 and GluR6 kainate receptor subtypes by Clarke, V.R., Ballyk, B.A., Hoo, K.H., Mandelzys, A., Pellizzari, A., Bath, (3S,4aR,6R,8aR)-6-[2-(1(2)H-tetrazole-5-yl)ethyl]decahydroisdoquino- C.P., Thomas, J., Sharpe, E.F., Davies, C.H., Ornstein, P.L., Schoepp, line-3-carboxylic-acid. Mol. Pharmacol. 49, 581–585. D.D., Kamboj, R.K., Collingridge, G.L., Lodge, D., Bleakman, D., Bortolotto, Z.A., Clarke, V.R., Delany, C.M., Parry, M.C., Smolders, I., 1997. A hippocampal GluR5 kainate receptor regulating inhibitory Vignes, M., Ho, K.H., Miu, P., Brinton, B.T., Fantaske, R., Ogden, synaptic transmission. Nature 389, 599–603. A., Gates, M., Ornstein, P.L., Lodge, D., Bleakman, D., Collingridge, Coggeshall, R.E., Carlton, S.M., 1998. Ultrastructural analysis of NMDA, G.L., 1999. Kainate receptors are involved in synaptic plasticity. Nature AMPA, and kainate receptors on unmyelinated and myelinated axons 402, 297–301. in the periphery. J. Comp. Neurol. 391, 78–86. Bortolotto, Z.A., Clarke, V.R.J., Delany, C.M., Vignes, M., Collingridge, Collingridge, G.L., Kehl, S.J., McLennan, H., 1983. Excitatory amino G.L., 2000. Kainate receptors and synaptic plasticity. Nature 406, 957. acids in synaptic transmission in the Schaffer collateral-commissural Boulter, J., Hollmann, M., O’Shea-Greenfield, A., Hartley, M., Deneris, pathway of the rat hippocampus. J. Physiol. 334, 33–46. E., Maron, C., Heinemann, S., 1990. Molecular cloning and functional Contractor, A., Swanson, G.T., Sailer, A., O’Gorman, S., Heinemann, expression of glutamate receptor subunit genes. Science 249, 1033– S.F., 2000. Identification of the kainate receptor subunits underlying 1037. modulation of excitatory synaptic transmission in the CA3 region of Bowie, D., Mayer, M.L., 1995. Inward rectification of both AMPA and the hippocampus. J. Neurosci. 20, 8269–8278. kainate subtype glutamate receptors generated by polyamine-mediated Contractor, A., Swanson, G., Heinemann, S.F., 2001. Kainate receptors ion channel block. Neuron 15, 453–462. are involved in short- and long-term plasticity at mossy fiber synapses Braga, M.F., Aroniadou-Anderjaska, V., Xie, J., Li, H., 2003. Bidirectional in the hippocampus. Neuron 29, 209–216. modulation of GABA release by presynaptic glutamate receptor 5 Contractor, A., Sailer, A.W., Darstein, M., Maron, C., Xu, J., Swanson, kainate receptors in the basolateral amygdala. J. Neurosci. 23, 442–452. G.T., Heinemann, S.F., 2003. Loss of kainate receptor-mediated Brickley, S., Swanson, G.T., Contractor, A., Farrant, M., Sailer, A., heterosynaptic facilitation of mossy-fiber synapses in KA2−/− mice. Cull-Candy, S.G., Heinemann, S.F., 1999. Functional GluR5-containing J. Neurosci. 23, 422–429. kainate receptors are restricted to extrasynaptic sites in Purkinje cells Cossart, R., Esclapez, M., Hirsch, J.C., Bernard, C., Ben-Ari, Y., 1998. of the mouse cerebellum. J. Physiol. 521P, 90P. GluR5 kainate receptor activation in interneurons increases tonic Bureau, I., Bischoff, S., Heinemann, S.F., Mulle, C., 1999. Kainate inhibition of pyramidal cells. Nat. Neurosci. 1, 470–478. receptor-mediated responses in the CA1 field of wild-type and Cossart, R., Tyzio, R., Dinocourt, C., Esclapez, M., Hirsch, J.C., Ben-Ari, GluR6-deficient mice. J. Neurosci. 19, 653–663. Y., Bernard, C., 2001. Presynaptic kainate receptors that enhance the 404 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407

release of GABA on CA1 hippocampal interneurons. Neuron 29, 497– Frerking, M., Ohliger-Frerking, P., 2002. AMPA receptors and kainate 508. receptors encode different features of afferent activity. J. Neurosci. 22, Cossart, R., Epsztein, J., Tyzio, R., Becq, H., Hirsch, J., Ben-Ari, Y., 7434–7443. Crepel, V., 2002. Quantal release of glutamate generates pure kainate Frerking, M., Malenka, R.C., Nicoll, R.A., 1998. Synaptic activation of and mixed AMPA/kainate EPSCs in hippocampal neurons. Neuron 35, kainate receptors on hippocampal interneurons. Nat. Neurosci. 1, 479– 147–159. 486. Coussen, F., Normand, E., Marchal, C., Costet, P., Choquet, D., Lambert, Frerking, M., Petersen, C.C., Nicoll, R.A., 1999. Mechanisms underlying M., Mege, R.M., Mulle, C., 2002. Recruitment of the kainate kainate receptor-mediated disinhibition in the hippocampus. Proc. Natl. receptor subunit glutamate receptor 6 by cadherin/catenin complexes. Acad. Sci. U.S.A. 96, 12917–12922. J. Neurosci. 22, 6426–6436. Frerking, M., Schmitz, D., Zhou, Q., Johansen, J., Nicoll, R.A., 2001. Crowder, T.L., Weiner, J.L., 2002. Functional characterization of kainate Kainate receptors depress excitatory synaptic transmission at CA3 → receptors in the rat nucleus accumbens core region. J. Neurophysiol. CA1 synapses in the hippocampus via a direct presynaptic action. J. 88, 41–48. Neurosci. 21, 2958–2966. Cui, C., Mayer, M.L., 1999. Heteromeric kainate receptors formed by the Garcia, E.P., Mehta, S., Blair, L.A., Wells, D.G., Shang, J., Fukushima, co-assembly of GluR5, GluR6, and GluR7. J. Neurosci. 19, 8281–8291. T., Fallon, J.R., Garner, C.C., Marshall, J., 1998. SAP90 binds and Cunha, R.A., Constantino, M.D., Ribeiro, J.A., 1997. Inhibition of [3H] clusters kainate receptors causing incomplete desensitization. Neuron gamma-aminobatyric acid release by kainate receptor activation in rat 21, 727–739. hippocampal synaptosomes. Eur. J. Pharmacol. 323, 167–172. Ghasemzadeh, M.B., Sharma, S., Surmeier, D.J., Eberwine, J.H., Cunha, R.A., Malva, J.O., Ribeiro, J.A., 1999. Kainate receptors coupled Chesselet, M.F., 1996. Multiplicity of glutamate receptor subunits in to G(i)/G(o) proteins in the rat hippocampus. Mol. Pharmacol. 56, single striatal neurons: an RNA amplification study. Mol. Pharmacol. 429–433. 49, 852–859. Cunha, R.A., Malva, J.O., Ribeiro, J.A., 2000. Pertussis toxin prevents presynaptic inhibition by kainate receptors of rat hippocampal Ghetti, A., Heinemann, S.F., 2000. NMDA-dependent modulation of + [(3)H]GABA release. FEBS Lett. 469, 159–162. hippocampal kainate receptors by calcineurin and Ca(2 )/calmodulin- Dai, W.M., Christensen, K.V., Egebjerg, J., Ebert, B., Lambert, J.D., 2002. dependent protein kinase. J. Neurosci. 20, 2766–2773. Correlation of the expression of kainate receptor subtypes to responses Gregor, P., O’Hara, B.F., Yang, X., Uhl, G.R., 1993. Expression and evoked in cultured cortical and spinal cord neurons. Brain Res. 926, novel subunit isoforms of glutamate receptor genes GluR5 and GluR6. 94–107. NeuroReport 4, 1343–1346. Davies, S.N., Collingridge, G.L., 1989. Role of excitatory amino acid Hampson, D.R., Huie, D., Wenthold, R.J., 1987. Solubilization of kainic receptors in synaptic transmission in area CA1 of rat hippocampus. acid binding sites from rat brain. J. Neurochem. 49, 1209–1215. Proc. R. Soc. Lond. B 236, 373–384. Harris, E.W., Cotman, C.W., 1986. Long-term potentiation of guinea Davies, J., Evans, R.H., Francis, A.A., Watkins, J.C., 1979. Excitatory pig mossy fiber responses is not blocked by N-methyl-d-aspartate amino acid receptors and synaptic excitation in the mammalian central antagonists. Neurosci. Lett. 70, 132–137. nervous system. J. Physiol. (Paris) 75, 641–654. Herb, A., Burnashev, N., Werner, P., Sakmann, B., Wisden, W., Seeburg, Delaney, A.J., Jahr, C.E., 2002. Kainate receptors differentially regulate P.H., 1992. The KA2 subunit of excitatory amino acid receptors shows release at two parallel fiber synapses. Neuron 36, 475–482. widespread expression in brain and forms ion channels with distantly DeVries, S.H., 2000. Bipolar cells use kainate and AMPA receptors to related subunits. Neuron 8, 775–785. filter visual information into separate channels. Neuron 28, 847–856. Hestrin, S., Nicoll, R.A., Perkel, D.J., Sah, P., 1990. Analysis of excitatory DeVries, S.H., Schwartz, E.A., 1999. Kainate receptors mediate synaptic synaptic action in pyramidal cells using whole-cell recording from rat transmission between cones and “off” bipolar cells in a mammalian hippocampal slices. J. Physiol. 422, 203–225. retina. Nature 397, 157–160. Hirbec, H., Francis, J.C., Lauri, S.E., Braithwaite, S.P., Coussen, F., Mulle, Dildy-Mayfield, J.E., Harris, R.A., 1994. Activation of protein kinase C., Dev, K.K., Couthino, V., Meyer, G., Isaac, J.T., Collingridge, G.L., C inhibits Kainate-induced currents in oocytes expressing glutamate Henley, J.M., 2003. Rapid and differential regulation of AMPA and receptor subunit. J. Neurochem. 62, 1639–1642. kainate receptors at hippocampal mossy fibre synapses by PICK1 and Dingledine, R., Borges, K., Bowie, D., Traynelis, S.F., 1999. The glutamate GRIP. Neuron 37, 625–638. receptor ion channels. Pharmacol. Rev. 51, 7–61. Hollmann, M., Heinemann, S., 1994. Cloned glutamate receptors. Annu. Donevan, S.D., Rogawski, M.A., 1995. Intracellular polyamines mediate Rev. Neurosci. 17, 31–108. inward rectification of Ca(2+)-permeable ␣-amino-3-hydroxy-5- Hollmann, M., Maron, C., Heinemann, S., 1994. N-Glycosylation site methyl-4-isoxazolepropionic acid receptors. Proc. Natl. Acad. Sci. tagging suggests a three transmembrane domain topology for the U.S.A. 92, 9298–9302. glutamate receptor GluR1. Neuron 13, 1331–1343. Donevan, S.D., Yamaguchi, S., Rogawski, M.A., 1994. Non-N-methyl- Howe, J.R., 1996. Homomeric and heteromeric ion channels formed d-aspartate receptor antagonism by 3-N-substituted 2,3-benzodiaz- from the kainate-type subunits GluR6 and KA2 have very small, but epines: relationship to anticonvulsant activity. J. Pharmacol. Exp. Ther. different, unitary conductances. J. Neurophysiol. 76, 510–519. 271, 25–29. Donevan, S.D., Beg, A., Gunther, J.M., Twyman, R.E., 1998. The Huettner, J.E., 1990. Glutamate receptor channels in rat DRG neurons: methylglutamate, SYM2081, is a potent and highly selective agonist activation by kainate and quisqualate blockade of desensitization by at kainate receptors. J. Pharmacol. Exp. Ther. 285, 539–545. Con A. Neuron 5, 255–266. Egebjerg, J., Heinemann, S.F., 1993. Ca2+ permeability of unedited and Huettner, J.E., 2001. Kainate receptors: knocking out plasticity. Trends edited versions of the kainate selective glutamate receptor GluR6. Proc. Neurosci. 24, 365–366. Natl. Acad. Sci. U.S.A. 90, 755–759. Huettner, J.E., Stack, E., Wilding, T.J., 1998. Antagonism of neuronal Egebjerg, J., Bettler, B., Hermans-Borgmeyer, I., Heinemann, S., 1991. kainate receptors by lanthanum and gadolinium. Neuropharmacology Cloning of a cDNA for a glutamate receptor subunit activated by 37, 1239–1247. kainate but not AMPA. Nature 351, 745–748. Hwang, S.J., Pagliardini, S., Rustioni, A., Valtschanoff, J.G., 2001. Fisher, R.S., Alger, B.E., 1984. Electrophysiological mechanisms of kainic Presynaptic kainate receptors in primary afferents to the superficial acid-induced epileptiform activity in the rat hippocampal slice. J. laminae of the rat spinal cord. J. Comp. Neurol. 436, 275–289. Neurosci. 4, 1312–1323. Isa, T., Iino, M., Itazawa, S., Ozawa, S., 1995. Spermine mediates Frerking, M., Nicoll, R.A., 2000. Synaptic kainate receptors. Curr. Opin. inward rectification of Ca(2+)-permeable AMPA receptor channels. Neurobiol. 10, 342–351. NeuroReport 6, 2045–2048. J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 405

Ito, I., Sugiyama, H., 1991. Roles of glutamate receptors in long-term Mishina, M., 1992. Molecular diversity of the NMDA receptor channel. potentiation at hippocampal mossy fiber synapses. NeuroReport 2, Nature 358, 36–41. 333–336. Kuusinen, A., Arvola, M., Keinanen, K., 1995. Molecular dissection of the Ji, Z., Staubli, U., 2002. Presynaptic kainate receptors play different agonist binding site of an AMPA receptor. EMBO J. 14, 6327–6332. physiological roles in mossy fiber and associational–commissural Lauridsen, J., Honore, T., Krogsgaard-Larsen, P., 1985. Ibotenic acid synapses in CA3 of hippocampus from adult rats. Neurosci. Lett. 331, analogues. Synthesis, molecular flexibility, and in vitro activity of 71–74. agonists and antagonists at central glutamic acid receptors. J. Med. Jiang, L., Xu, J., Nedergaard, M., Kang, J., 2001. A kainate receptor Chem. 28, 668–672. increases the efficacy of GABAergic synapses. Neuron 30, 503–513. Lauri, S.E., Bortolotto, Z.A., Bleakman, D., Ornstein, P.L., Lodge, D., Jones, K.A., Wilding, T.J., Huettner, J.E., Costa, A.M., 1997. Isaac, J.T., Collingridge, G.L., 2001a. A critical role of a facilitatory Desensitization of kainate receptors by kainate, glutamate and presynaptic kainate receptor in mossy fiber LTP. Neuron 32, 697–709. diastereomers of 4-methylglutamate. Neuropharmacology 36, 853–863. Lauri, S.E., Delany, C., Clarke V.R.J., Bortolotto, Z.A., Ornstein, Kamboj, S.K., Swanson, G.T., Cull-Candy, S.G., 1995. Intracellular P.L., Isaac, J.T.R., Collingridge, G.L., 2001b. Synaptic activation of spermine confers rectification on rat calcium-permeable AMPA and a presynaptic kainate receptor facilitates AMPA receptor-mediated kainate receptors. J. Physiol. (Lond.) 486, 297–303. synaptic transmission at hippocampal mossy fibre synapses. Kamiya, H., Ozawa, S., 1998. Kainate receptor-mediated inhibition of Neuropharmacology 41, 907–915. presynaptic Ca2+ influx and EPSP in area CA1 of the rat hippocampus. Lauri, S.E., Bortolotto, Z.A., Nistico, R., Bleakman, D., Ornstein, P.L., J. Physiol. (Lond.) 509, 833–845. Lodge, D., Isaac, J.T., Collingridge, G.L., 2003. A role for Ca2+ stores Kamiya, H., Ozawa, S., 2000. Kainate receptor-mediated presynaptic in kainate receptor-dependent synaptic facilitation and LTP at mossy inhibition at the mouse hippocampal mossy fibre synapse. J. Physiol. fiber synapses in the hippocampus. Neuron 39, 327–341. (Lond.) 523, 653–665. Lee, C.J., Engelman, H.S., MacDermott, A.B., 1999. Activation of kainate Kamiya, H., Ozawa, S., Manabe, T., 2002. Kainate receptor-dependent receptors on rat sensory neurons evokes action potential firing and may short-term plasticity of presynaptic Ca2+ influx at the hippocampal modulate transmitter release. Ann. N.Y. Acad. Sci. 868, 546–549. mossy fiber synapses. J. Neurosci. 22, 9237–9243. Lee, C.J., Kong, H., Manzini, M.C., Albuquerque, C., Chao, M.V., Kehl, S.J., McLennan, H., Collingridge, G.L., 1984. Effects of folic MacDermott, A.B., 2001. Kainate receptors expressed by a and kainic acids on synaptic responses of hippocampal neurones. subpopulation of developing nociceptors rapidly switch from high to Neuroscience 11, 111–124. low Ca2+ permeability. J. Neurosci. 21, 4572–4581. Keinänen, K., Wisden, W., Sommer, B., Werner, P., Herb, A., Verdoorn, Lee, C.J., Bardoni, R., Tong, C.K., Engelman, H.S., Joseph, D.J., T.A., Sakmann, B., Seeburg, P.H., 1990. A family of AMPA-selective Magherini, P.C., MacDermott, A.B., 2002. Functional expression of glutamate receptors. Science 249, 556–560. AMPA receptors on central terminals of rat dorsal root ganglion Kerchner, G.A., Wilding, T.J., Li, P., Wang, G.D., Zhuo, M., Huettner, neurons and presynaptic inhibition of glutamate release. Neuron 35, J.E., 2001a. Presynaptic kainate receptors regulate spinal sensory 135–146. transmission. J. Neurosci. 21, 59–66. Lerma, J., 2003. Roles and rules of kainate receptors in synaptic Kerchner, G.A., Wang, G.D., Qiu, C.S., Huettner, J.E., Zhuo, M., 2001b. transmission. Nat. Rev. Neurosci. 4, 481–495. Direct presynaptic regulation of GABA/glycine release by kainate Lerma, J., Paternain, A.V., Naranjo, J.R., Mellstrom, B., 1993. Functional receptors in the dorsal horn: an ionotropic mechanism. Neuron 32, kainate-selective glutamate receptors in cultured hippocampal neurons. 477–488. Proc. Natl. Acad. Sci. U.S.A. 90, 11688–11692. Kerchner, G.A., Wilding, T.J., Huettner, J.E., Zhuo, M., 2002. Kainate Lerma, J., Paternain, A.V., Rodriguez-Moreno, A., Lopez-Garcia, J.C., receptor subunits underlying presynaptic regulation of transmitter 2001. Molecular physiology of kainate receptors. Physiol. Rev. 81, release in the dorsal horn. J. Neurosci. 22, 8010–8017. 971–998. Khalilov, I., Hirsch, J., Cossart, R., Ben-Ari, Y., 2001. Paradoxical Li, H., Rogawski, M.A., 1998. GluR5 kainate receptor-mediated synaptic anti-epileptic effects of a GLUK5 agonist of kainate receptors. J. transmission in rat basolateral amygdala in vitro. Neuropharmacology Neurophysiol. 88, 523–527. 37, 1279–1286. Kidd, F.L., Isaac, J.T., 1999. Developmental and activity-dependent Li, P., Wilding, T.J., Kim, S.J., Calejesan, A.A., Huettner, J.E., Zhuo, regulation of kainate receptors at thalamocortical synapses. Nature 400, M., 1999. Kainate-receptor-mediated sensory synaptic transmission in 569–573. mammalian spinal cord. Nature 397, 161–164. Kidd, F.L., Isaac, J.T., 2001. Kinetics and activation of postsynaptic kainate Li, H., Chen, A., Xing, G., Wei, M.L., Rogawski, M.A., 2001. Kainate receptors at thalamocortical synapses: role of glutamate clearance. J. receptor-mediated heterosynaptic facilitation in the amygdala. Nat. Neurophysiol. 86, 1139–1148. Neurosci. 4, 612–620. Kidd, F.L., Coumis, U., Collingridge, G.L., Crabtree, J.W., Isaac, J.T., Liu, Q.S., Patrylo, P.R., Gao, X.B., van den Pol, A.N., 1999. Kainate acts 2002. A presynaptic kainate receptor is involved in regulating the at presynaptic receptors to increase GABA release from hypothalamic dynamic properties of thalamocortical synapses during development. neurons. J. Neurophysiol. 82, 1059–1062. Neuron 34, 635–646. Lodge, D., Dingledine, R., 2000. Ionotropic glutamate receptors. In: The Kiskin, N.I., Krishtal, O.A., Tsyndrenko, A.Ya., 1986. Excitatory amino IUPHAR Compendium of Receptor Characterization and Classification, acid receptors in hippocampal neurons: kainate fails to desensitize second ed. IUPHAR Media Ltd., London, pp. 189–194. them. Neurosci. Lett. 63, 225–230. London, E.D., Coyle, J.T., 1979. Specific binding of [3H]kainic acid to Koh, D.S., Burnashev, N., Jonas, P., 1995. Block of native receptor sites in rat brain. Mol. Pharmacol. 15, 492–505. Ca(2+)-permeable AMPA receptors in rat brain by intracellular Lomeli, H., Wisden, W., Köhler, M., Keinänen, K., Sommer, B., Seeburg, polyamines generates double rectification. J. Physiol. 486, 305–312. P.H., 1992. High-affinity kainate and domoate receptors in rat brain. Köhler, M., Burnashev, N., Sakmann, B., Seeburg, P.H., 1993. FEBS Lett. 307, 139–143. Determinants of Ca2+ permeability in both TM1 and TM2 of high Mascias, P., Scheede, M., Bloms-Funke, P., Chizh, B., 2002. Modulation affinity kainate receptor channels: diversity by RNA editing. Neuron of spinal nociception by GluR5 kainate receptor ligands in acute 10, 491–500. and hyperalgesic states and the role of gabaergic mechanisms. Kullmann, D.M., 2001. Presynaptic kainate receptors in the hippocampus: Neuropharmacology 43, 327–339. slowly emerging from obscurity. Neuron 32, 561–564. Mayer, M.L., Westbrook, G.L., 1987. The physiology of excitatory amino Kutsuwada, T., Kashiwabuchi, N., Mori, H., Sakimura, K., Kushiya, E., aids in the vertebrate central nervous system. Prog. Neurobiol. 28, Araki, K., Meguro, H., Masaki, H., Kumanishi, T., Arakawa, M., 197–276. 406 J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407

Mehta, S., Wu, H., Garner, C.C., Marshall, J., 2001. Molecular Pemberton, K.E., Belcher, S.M., Ripellino, J.A., Howe, J.R., 1998. mechanisms regulating the differential association of kainate receptor High-affinity kainate-type ion channels in rat cerebellar granule cells. subunits with SAP90/PSD-95 and SAP97. J. Biol. Chem. 276, 16092– J. Physiol. 510, 401–420. 16099. Petralia, R.S., Wang, Y.-X., Wenthold, R.J., 1994. Histological and Melyan, Z., Wheal, H.V., Lancaster, B., 2002. Metabotropic-mediated ultrastructural localization of the kainate receptor subunits, KA2 kainate receptor regulation of IsAHP and excitability in pyramidal and GluR6/7, in the rat nervous system using selective antipeptide cells. Neuron 34, 107–114. antibodies. J. Comp. Neurol. 349, 85–110. Min, M.Y., Melyan, Z., Kullmann, D.M., 1999. Synaptically released Pickering, D.S., Taverna, F.A., Salter, M.W., Hampson, D.R., 1995. glutamate reduces gamma-aminobutyric acid (GABA)ergic inhibition Palmitoylation of the GluR6 kainate receptor. Proc. Natl. Acad. Sci. in the hippocampus via kainate receptors. Proc. Natl. Acad. Sci. U.S.A. U.S.A. 92, 12090–12094. 96, 9932–9937. Porter, J.T., Cauli, B., Staiger, J.F., Lambolez, B., Rossier, J., Audinat, E., Monaghan, D.T., Cotman, C.W., 1982. The distribution of [3H]kainic acid 1998. Properties of bipolar VIPergic interneurons and their excitation binding sites in the rat CNS as determined by autoradiography. Brain by pyramidal neurons in the rat neocortex. Eur. J. Neurosci. 10, 3617– Res. 252, 91–100. 3628. Monaghan, D.T., Bridges, R.J., Cotman, C.W., 1989. The excitatory amino Procter, M.J., Houghton, A.K., Faber, E.S.L., Chizh, B.A., Ornstein, acid receptors: their classes, pharmacology, and distinct properties in P.L., Lodge, D., 1998. Actions of kainate and AMPA selective the function of the central nervous system. Ann. Rev. Pharmacol. glutamate receptor ligands on nociceptive processing in the spinal cord. Toxicol. 29, 365–402. Neuropharmacology 37, 1287–1297. Monyer, H., Sprengel, R., Schoepfer, R., Herb, A., Higuchi, M., Lomeli, Raymond, L.A., Blackstone, C.D., Huganir, R.L., 1993. Phosphorylation H., Burnashev, N., Sakmann, B., Seeburg, P.H., 1992. Heteromeric and modulation of recombinant GluR6 glutamate receptors by NMDA receptors: molecular and functional distinction of subtypes. cAMP-dependent protein kinase. Nature 361, 637–641. Science 256, 1217–1221. Renard, A., Crepel, F., Audinat, E., 1995. Evidence for two types of Moriyoshi, K., Masu, M., Ishii, T., Shigemoto, R., Mizuno, N., Nakanishi, non-NMDA receptors in rat cerebellar Purkinje cells maintained in S., 1991. Molecular cloning and characterization of the rat NMDA slice cultures. Neuropharmacology 34, 335–346. receptor. Nature 354, 31–37. Robinson, J.H., Deadwyler, S.A., 1981. Kainic acid produces Mulle, C., Sailer, A., Perez-Otano, I., Dickinson-Anson, H., Castillo, P.E., depolarization of CA3 pyramidal cells in the in vitro hippocampal Bureau, I., Maron, C., Gage, F.H., Mann, J.R., Bettler, B., Heinemann, slice. Brain Res. 221, 117–127. S.F., 1998. Altered synaptic physiology and reduced susceptibility to Roche, K.W., Raymond, L.A., Blackstone, C., Huganir, R.L., 1994. kainate-induced seizures in GluR6-deficient mice. Nature 392, 601– Transmembrane topology of the glutamate receptor subunit GluR6. J. 605. Biol. Chem. 269, 11679–11682. Mulle, C., Sailer, A., Swanson, G.T., Brana, C., O’Gorman, S., Bettler, Rodriguez-Moreno, A., Herreras, O., Lerma, J., 1997. Kainate receptors B., Heinemann, S.F., 2000. Subunit composition of kainate receptors presynaptically downregulate GABAergic inhibition in the rat in hippocampal interneurons. Neuron 28, 475–484. hippocampus. Neuron 19, 893–901. Nadler, J.V., Perry, B.W., Cotman, C.W., 1978. Intraventricular kainic Rodriguez-Moreno, A., Lerma, J., 1998. Kainate receptor modulation of acid preferentially destroys hippocampal pyramidal cells. Nature 271, GABA release involves a metabotropic function. Neuron 20, 1211– 676–677. 1218. Nicoll, R.A., Mellor, J., Frerking, M., Schmitz, D., 2000. Kainate receptors Rodriguez-Moreno, A., Lopez-Garcia, J.C., Lerma, J., 2000. Two and synaptic plasticity. Nature 406, 957. populations of kainate receptors with separate signaling mechanisms O’Neill, M.J., Bond, A., Ornstein, P.L., Ward, M.A., Hicks, C.A., Hoo, in hippocampal interneurons. Proc. Natl. Acad. Sci. U.S.A. 97, 1293– K., Bleakman, D., Lodge, D., 1998. Decahydroisoquinolines: novel 1298. competitive AMPA/kainate antagonists with neuroprotective effects in Ruano, D., Lambolez, B., Rossier, J., Paternain, A.V., Lerma, J., 1995. global cerebral ischaemia. Neuropharmacology 37, 1211–1222. Kainate receptor subunits expressed in single cultured hippocampal Paarmann, I., Frermann, D., Keller, B.U., Hollmann, M., 2000. Expression neurons: molecular and functional variants by RNA editing. Neuron of 15 glutamate receptor subunits and various splice variants in tissue 14, 1009–1017. slices and single neurons of brainstem nuclei and potential functional Ruscheweyh, R., Sankuhler, J., 2002. Role of kainate receptors in implications. J. Neurochem. 74, 1335–1345. nociception. Brain Res. Rev. 40, 215–222. Paschen, W., Schnitt, J., Gissel, C., Dux, E., 1997. Developmental changes Sahara, Y., Noro, N., Iida, Y., Soma, K., Nakamura, Y., 1997. Glutamate of RNA editing of glutamate receptor subunits GluR5 and GluR6: in receptor subunits GluR5 and KA2 are coexpressed in rat trigeminal vivo versus in vitro. Dev. Brain Res. 98, 271–280. ganglion neurons. J. Neurosci. 17, 6611–6620. Paternain, A.V., Morales, M., Lerma, J., 1995. Selective antagonism Sailer, A., Swanson, G.T., Perez-Otano, I., O’Leary, L., Malkmus, S.A., of AMPA receptors unmasks kainate receptor-mediated responses in Dyck, R.H., Dickinson-Anson, H., Schiffer, H.H., Maron, C., Yaksh, hippocampal neurons. Neuron 14, 185–189. T.L., Gage, F.H., O’Gorman, S., Heinemann, S.F., 1999. Generation and Paternain, A.V., Vicente, A., Nielsen, E.O., Lerma, J., 1996. Comparative analysis of GluR5(Q636R) kainate receptor mutant mice. J. Neurosci. antagonism of kainate-activated kainate and AMPA receptors in 19, 8757–8764. hippocampal neurons. Eur. J. Neurosci. 8, 2129–2136. Sakimura, K., Morita, T., Kushiya, E., Mishina, M., 1992. Primary Paternain, A.V., Rodriguez-Moreno, A., Villarroel, A., Lerma, J., 1998. structure and expression of the g2 subunit of the glutamate receptor Activation and desensitization properties of native and recombinant channel selective for kainate. Neuron 8, 267–274. kainate receptors. Neuropharmacology 37, 1249–1259. Savinainen, A., Garcia, E.P., Dorow, D., Marshall, J., Liu, Y.F., 2001. Paternain, A.V., Herrera, M.T., Nieto, M.A., Lerma, J., 2000. GluR5 and Kainate receptor activation induces mixed lineage kinase-mediated GluR6 kainate receptor subunits coexist in hippocampal neurons and cellular signaling cascades via postsynaptic density protein 95. J. Biol. co-assemble to form functional receptors. J. Neurosci. 20, 196–205. Chem. 276, 11382–11386. Patneau, D.K., Mayer, M.L., 1991. Kinetic analysis of interactions between Schiffer, H.H., Swanson, G.T., Heinemann, S.F., 1997. Rat GluR7 and a kainate and AMPA: evidence for activation of a single receptor in carboxy-terminal splice variant, GluR7b, are functional kainate receptor mouse hippocampal neurons. Neuron 6, 785–798. subunits with a low sensitivity to glutamate. Neuron 19, 1141–1146. Pelletier, J.C., Hesson, D.P., Jones, K.A., Costa, A.M., 1996. Substituted Schmitz, D., Frerking, M., Nicoll, R.A., 2000. Synaptic activation of 1,2-dihydrophthalazines: potent, selective and non-competitive presynaptic kainate receptors on hippocampal mossy fiber synapses. inhibitors of the AMPA receptor. J. Med. Chem. 39, 343–346. Neuron 27, 327–338. J.E. Huettner / Progress in Neurobiology 70 (2003) 387–407 407

Schmitz, D., Mellor, J., Nicoll, R.A., 2001a. Presynaptic kainate receptor Unnerstall, J.R., Wamsley, J.K., 1983. Autoradiographic localization of mediation of frequency facilitation at hippocampal mossy fiber high affinity [3H]kainic acid binding sites in the rat forebrain. Eur. J. synapses. Science 291, 1972–1976. Pharmacol. 86, 361–371. Schmitz, D., Mellor, J., Frerking, M., Nicoll, R.A., 2001b. Presynaptic Verdoorn, T.A., Johansen, T.H., Drejer, J., Nielsen, E.O., 1994. Selective kainate receptors at hippocampal mossy fiber synapses. Proc. Natl. block of recombinant glur6 receptors by NS102, a novel non-NMDA Acad. Sci. U.S.A. 98, 11003–11008. receptor antagonist. Eur. J. Pharmacol. 269, 43–49. Semyanov, A., Kullmann, D.M., 2001. Kainate receptor-dependent axonal Vignes, M., Collingridge, G.L., 1997. The synaptic activation of kainate depolarization and action potential initiation in interneurons. Nat. receptors. Nature 388, 179–182. Neurosci. 4, 718–723. Vignes, M., Bleakman, D., Lodge, D., Collingridge, G.L., 1997. The Sheng, M., Pak, D.T., 2000. Ligand-gated ion channel interactions with synaptic activation of the GluR5 subtype of kainate receptor in area cytoskeletal and signaling proteins. Annu. Rev. Physiol. 62, 755–778. CA3 of the rat hippocampus. Neuropharmacology 36, 1477–1481. Simmons, R.M., Li, D.L., Hoo, K.H., Deverill, M., Ornstein, P.L., Iyengar, Vignes, M., Clarke, V.R., Parry, M.J., Bleakman, D., Lodge, D., Ornstein, S., 1998. Kainate GluR5 receptor subtype mediates the nociceptive P.L., Collingridge, G.L., 1998. The GluR5 subtype of kainate receptor response to formalin in the rat. Neuropharmacology 37, 25–36. regulates excitatory synaptic transmission in areas CA1 and CA3 of Sloviter, R.S., Damiano, B.P., 1981. On the relationship between kainic the rat hippocampus. Neuropharmacology 37, 1269–1277. acid-induced epileptiform activity and hippocampal neuronal damage. Vissel, B., Royle, G.A., Christie, B.R., Schiffer, H.H., Ghetti, A., Tritto, T., Neuropharmacology 20, 1003–1011. Perez-Otano, I., Radcliffe, R.A., Seamans, J., Sejnowski, T., Wehner, Smith, T.C., Wang, L.Y., Howe, J.R., 1999. Distinct kainate receptor J.M., Collins, A.C., O’Gorman, S., Heinemann, S.F., 2001. The role phenotypes in immature and mature mouse cerebellar granule cells. J. of RNA editing of kainate receptors in synaptic plasticity and seizures. Physiol. 517, 51–58. Neuron 29, 217–227. Smolders, I., Bortolotto, Z.A., Clarke, V.R., Warre, R., Khan, G.M., Wang, L.Y., Taverna, F.A., Huang, X.P., MacDonald, J.F., Hampson, D.R., O’Neill, M.J., Ornstein, P.L., Bleakman, D., Ogden, A., Weiss, B., 1993. Phosphorylation and modulation of a kainate receptor (GluR6) Stables, J.P., Ho, K.H., Ebinger, G., Collingridge, G.L., Lodge, by cAMP-dependent protein kinase. Science 259, 1173–1175. D., Michotte, Y., 2002. Antagonists of GLU(K5)-containing kainate Watkins, J.C., Evans, R.H., 1981. Excitatory amino acid transmitters. Ann. receptors prevent pilocarpine-induced limbic seizures. Nat. Neurosci. Rev. Pharmacol. Toxicol. 21, 165–204. 5, 796–804. Watkins, J.C., Krogsgaard-Larsen, P., Honoré, T., 1990. Structure-activity Sommer, B., Kohler, M., Sprengel, R., Seeburg, P.H., 1991. RNA editing relationships in the development of excitatory amino acid receptor in brain controls a determinant of ion flow in glutamate-gated channels. agonists and competitive antagonists. Trends Pharmacol. Sci. 11, 25– Cell 67, 11–19. 33. Sommer, B., Burnashev, N., Verdoorn, T.A., Keinänen, K., Sakmann, B., Werner, P., Voigt, M., Keinänen, K., Wisden, W., Seeburg, P.H., Seeburg, P.H., 1992. A glutamate receptor channel with high affinity 1991. Cloning of a putative high-affinity kainate receptor expressed for domoate and kainate. EMBO J. 11, 1651–1656. predominantly in hippocampal CA3 cells. Nature 351, 742–744. Stanfa, L.C., Dickenson, A.H., 1999. The role of non-N-methyl-d-aspartate Westbrook, G.L., Lothman, E.W., 1983. Cellular and synaptic basis of ionotropic glutamate receptors in the spinal transmission of nociception kainic acid-induced hippocampal epileptiform activity. Brain Res. 273, in normal animals and animals with carrageenan inflammation. 97–109. Neuroscience 93, 1391–1398. Wilding, T.J., Huettner, J.E., 1995. Differential antagonism of ␣-amino-3- Stern-Bach, Y., Bettler, B., Hartley, M., Sheppard, P.O., O’Hara, P.J., hydroxy-5-methyl-4-isoxazolepropionic acid-preferring and kainate- Heinemann, S.F., 1994. Agonist selectivity of glutamate receptors preferring receptors by 2,3-benzodiazepines. Mol. Pharmacol. 47, 582– is specified by two domains structurally related to bacterial amino 587. acid-binding proteins. Neuron 13, 1345–1357. Wilding, T.J., Huettner, J.E., 1997. Activation and desensitization of Sutton, J.L., Maccecchini, M.L., Kajander, K.C., 1999. The kainate hippocampal kainate receptors. J. Neurosci. 17, 2713–2721. receptor antagonist 2S,4R-4-methylglutamate attenuates mechanical Wilding, T.J., Huettner, J.E., 2001. Functional diversity and developmental allodynia and thermal hyperalgesia in a rat model of nerve injury. changes in neuronal kainate receptors. J. Physiol. 532, 411–421. Neuroscience 91, 283–292. Wisden, W., Seeburg, P.H., 1993. A complex mosaic of high-affinity Swanson, G.T., Feldmeyer, D., Kaneda, M., Cull-Candy, S.G., 1996. Effect kainate receptors in rat brain. J. Neurosci. 13, 3582–3598. of RNA editing and subunit co-assembly single-channel properties of Wong, L.A., Mayer, M.L., 1993. Differential modulation by cyclothiazide recombinant kainate receptors. J. Physiol. (Lond.) 492, 129–142. and concanavalin A of desensitization at AMPA- and kainate-preferring Tölle, T.R., Berthele, A., Zieglgansberger, W., Seeburg, P.H., Wisden, glutamate receptors. Mol. Pharmacol. 44, 504–510. W., 1993. The differential expression of 16 NMDA and non-NMDA Zalutsky, R.A., Nicoll, R.A., 1990. Comparison of two forms of long-term receptor subunits in the rat spinal cord and in periaqueductal gray. J. potentiation in single hippocampal neurons. Science 248, 1619– Neurosci. 13, 5009–5028. 1624. Traynelis, S.F., Wahl, P., 1997. Control of rat GluR6 glutamate receptor Ziegra, C.J., Willard, J.M., Oswald, R.E., 1992. Coupling of a purified open probability by protein kinase A and calcineurin. J. Physiol. 503, goldfish brain kainate receptor with a pertussis toxin-sensitive G 513–531. protein. Proc. Natl. Acad. Sci. U.S.A. 89, 4134–4138.