isolation, localization, and cloning of a kainic acid binding protein

0022-1~14/90/$3.30 The Journal of HktocbcmLuy and Cytochcmisuy Copyright Q 1990 by The Histochemical Society. Inc.

Vol. 38. NO. 12, pp. 1717-1723, 1990 Printed in USA.

Symposium I

Isolation, Localization, and Cloning of a Kainic Acid Binding Protein from Frog Brain'

ROBERT J. WENTHOLD,' DAVID R. HAMPSON, KEIJI WADA, CHYREN HUNTER, MICHAEL D. OBERDORFER, and CLAUDE J. DECHESNE Section of Neurochemistry, Laboratory of Molecular Otologyv Notional Institute on Deafness and Other Communication Disorders, National Institutes of Health, Bether&, Maryland.

Received for publication August 21, 1990: accepted August 22. 1990 (OS2076).

Excitatory amino acids ( E M ) are major neurotransmitters in the vertebrate central nervous system. E M rmptors have been divided into three major subtypes on the basis of electrophysiological and ligand binding studies: N-methyl-D- aspartate, kainate, and quisqdate receptors. To understand their molecular properties, we undertook a project aimed at isolation and cloning of these rmptor subtypes. We pu- d i e d a kainate binding protein (KBP) from frog brain, in which kainate binding sites are about fortyfold more abundant than in rat brain, using domoic acid affinity chromatography, and made monoclonal and polydonal antibodies to the purified protein. These antibodies immunoprecipitate the frog KBP but not KBPs from other species. Immu- nocytochemical analyses show that KBP has a synaptic and extrasynaptic localization in frog optic tectum, with most labeling being extrasynaptic. The D N A encoding frog brain KBP was isolated by screening a frog brain D N A library with oligonucleotide probes that were based on the amino

acid sequence of the purified protein. The deduced amino acid sequence of the KBP has a hydrophobic profile similar to those of other ligand-gated ion channel subunits, such as the nicotinic acetylcholine receptor, the GABAA receptor, and the glycine rmptor. Frog brain KBP is very similar (36% amino acid identity to the carboxyl half) to rat brain kainate receptor, suggesting that these two proteins evolved from a common ancestor. The function of KBP in frog brain remains a major question. Preliminary results showed that Xenopuslaevis oocytes injected with KBP RNA did not produce a detectable electrophysiological response when per- fused with kainate. These results suggest that additional subunits may be required to form a functional receptor or that KBP is not functionally related to a neurotransmitter r a p t o r . (J Hstochem Cytochem 38:1717-1723, 1990) KEY WORDS: Excitatory amino acids: Neurotransmitters: Kainate binding protein: Frog brain; Immunocytochemistry.

Introduction Excitatory amino acids (EAA), including glutamate, aspartate, and related compounds, are the major excitatory neurotransmitters in the mammalian central nervous system. Electrophysiological studies have led to the identification of three major classes of EAA receptors named after the agonist that preferentially excites the receptor: kainate (U), quisqualate (QA) or a-amino-3-hydroxy-5-methyl-4- isoxazolepropionic acid (AMPA), and N-methyl-D-qmte (NMDA) receptors (6,13,18). Although these classes are known to be made up of channel-linked receptors, recent data suggest the presence of at least one type of metabotropic glutamate receptor that is phar-

I Correspondence to: Dr. Robert J. Wenthold, National Institutes of Health, Building 36, Room 5D08, Bethesda, MD 20892.

Presented in part at the symposium on EXCITATORY AMINO ACID NEUROTRANSMITERS AND THEIR RECEPTORS, August 9,1990, as part of the program at the third joint meeting of the Japan Society of Histochemistry and Cytochemistry and The Histochemical Society, held in Seattle, Washington, August 8-11.

macologically distinct from the ionotropic receptors (14~5) . Radi- oligand binding studies to tissue homogenates and in vitro autoradiography on tissue sections have demonstrated binding sites which have pharmacological properties similar to those of the three classes of receptors (6,13).

With their widespread distribution and apparent involvement in many neurobiological functions, including neurotransmission, excitotoxicity, long-term potentiation, and neurodegenerative dis- eases, an intense interest in EAAs and their receptors has developed. The capability to localize EAA receptors in the CNS would greatly facilitate their characterization and also would provide a usdul approach to identdying synapses at which an EAA is a neurotransmitter. Neurons using EAAs as neurotransmitters have been notoriously difficult to localize owing to the lack of selective histochemical markers. No selective enzymes in the biosynthesis of EAAs have been conclusively identified, and the direct immunocytochemical localization of EAAs suffers from the inability to differentiate between neurotransmitter and metabolic pools of the amino acids. Therefore, the receptor may be the most specific marker available for identdying synapses that use EAAs as neurotransmit-

1717

by guest on September 29, 2016jhc.sagepub.comDownloaded from

http://jhc.sagepub.com/

1718 WENTHOLD, HAMPSON, WADA, HUNTER, OBERDORFER, DECHESNE

ten. Receptor autoradiography has been useful in determining the general distributions of EAA receptors. but it does not have the molution needed for cellular and ululsuuctural localization. Owing largely to the lack of specific high-affinity ligands, the molecular characterization of EAA receptors has been slow to develop rela- tive to that of other neurotransmitter receptors. However, the recent purification and cloning of two KA binding proteins (KBP) (7,17) and the cloning of a KA receptor (11) have opened the molecular characterization of EAA receptors and have provided the necessary tools for localization of EAA receptors through immunocytochemistry and in situ hybridization. In this manuscript we briefly review our work on the purification, biochemical characterization, and cloning of a frog brain KBP and present in more de- tail the localization of this protein in frog brain using receptor autoradiography, immunocytochemistry, and in situ hybridization.

Biochemical Properties of Frog Brain KBP Frog brain was chosen as a source of KBP because its concentration of KBP is nearly fiftyfold that of mammalian brain, and the pharmacological characteristics of KA binding to frog brain are similar to those of mammalian brain (8,9). Scatchard analysis for both sources shows the presence of two binding sites: the values of the dissociation constants are 4.8 and 39 nM for the solubilized frog brain KBP, while in a similar preparation from mammalian brain, the values are 6.4 and 53 nM. Optimal solubilization of KBP was obtained with a combination of Triton X-100, digitonin, and glycerol in phosphate buffer. The crucial step in the purification of KBP was domoic acid-affinity chromatography, which produced an apparently homogeneous preparation of KBP (9). The purified protein displayed pharmacological properties indistinguishable from those of the membrane-associated or the unpurified solubilized KBP. The purified KBP contained a single broad band migrating with an Mr of 48.000 on SDS-polyacrylamide gel electrophoresis (PAGE). Two-dimensional electrophoresis/electrofocusing showed at l ev t eight forms of the protein, having the same mol& wight but differing isoelectric points. The nature of these multiple forms has not been resolved, but peptide mapping studies indicate that they are closely related and could arise from post-translational events such as glycosylation. Gel filtration chromatography in non-dena- turing buffer showcd that the KA binding site migrated with Mr of 570,000, suggesting that the active unit in detergent is composed of several of the MI 48,000 subunits.

The frog brain q P was extensively characterized using polyclonal and monoclonal antibodies made to the purified protein (10). Three monoclonal antibodies (MAb) which immunoprecipitate O%, 12%, and 85% of the [3H]-KA binding activity, and a polyclonal antibody which immunoprecipitates 8096, were used in these studies. All antibodies produced similar patterns on immunoblots of gels showing intense labeling at Mr 48,000 and a minor band at Mr 99,000. Under these conditions, all antibodies appear to recognize the same protein, indicating that the antibodies’ capability for immunoprecipitation is not due to different molecules being recog nized. Rather, it appears to be due to the area of the molecule that is recognized and the conformation of the protein; in cases where there is little or no immunoprecipitation, the epitope is probably not exposed on the native protein but is exposed after it is dena-

tured by SDS. There is also no indication that these antibodies can select for pharmacological subtype, since both high- and Iow-afFinity binding sites were equally immunoprecipitated with all antibodies. Because a photoaffinity ligand that binds to the KBP has not been developed, an alternative method to demonstrate that the KA binding site is associated with the Mr 48,000 protein is to iden- tify an MAb that both immunoprecipitates KA binding activity and labels the MI 48,000 band after PAGE. One MAb, KAR-AI, immunoprecipitated up to 85% of solubilized KA binding activity and labeled a protein at Mr 48,000. Further analysis with two- dimensional electrofocusinglelectrophoresis showed that the major forms of KBP, with different isoelectric points, are also hetero- geneous in the molecular size dimension, being made up of two or more components. In labeling of these subforms, there appears to be some difference among the antibodies, but this labeling does not appear to correlate with the antibody’s immunoprecipitation of binding activity. The significance of this remains to be determined, but a likely explanation is that these forms represent different stata of glycosylation and that the degree of glycosylation influences the binding of the antibody to the protein portion of the KBP. We have assumed that the immunoreactive labeling seen at Mr 99,000 is a dimer of the lower molecular weight form, since the amount of labeling of this band is increased in the absence of reduc- ing agents. However, in view of the findings of a KA receptor in mammalian brain with a similar size (discussed below), the possibility that the antibodies recognize a related but distinct protein must be considered.

The specific antibodies were used to determine if the KBP is found outside the CNS of the frog and ifa similar molecule is present in other species. Thew analyses show that KBP is essentially limited to the CNS in frog. The only notable exception is the inner ear where there is heavy labeling of a non-neuronal fibroblast-like cell (2). With respect to other species, one MAb, KAR-B1, showed an intense reaction with a band with MI 99,000 in rat brain but did not immunoprecipitate [ 3H]-KA binding activity. The protein recognized by this antibody is enriched in cerebellum, cortex, and hippocampus but sparse in brainstem, thalamus, and globus pal- lidus. Although such a distribution is consistent with that expected of an EAA receptor, a molecular weight twice that of the frog KBP, which is also much larger than any previously isolated channel-linked neurotransmitter receptor, raised the possibility that this labeling resulted from a nonspecific interaction. Howevcr, more recent data support the contention that the protein recognized by KAR-B1 is a component of an EAA receptor. We have obtained a highly purified AMPA binding protein from bovine brain (12). The major protein of this fraction has an Mr 114,000 and is recognized by KAR-B1. Therefore, in bovine brain KAR-B1 recognizes an AMPA binding protein and apparently not a KA binding protein. Fur- thermore, a KA receptor cDNA was cloned from rat brain (11) and encodes a protein with a calculated molecular weight of 99.8 KD, which is wry similar to the size of the protein recognized by KAR-B1.

Isolation of cDNA Encoding the KBP A partial amino acid sequence was obtained from V8 protease- generated peptides of the affinity-purified KBP and was used to design synthetic oligonucleotide probes to Screen a frog brain cDNA



KAINIC ACID BINDING PROTEIN IN FROG BRAIN 1719

Figure 1. Comparison of deduced amino acid sequences of frog brain KBP (17), chick brain KBP (7). and the carboxyl terminal portion of rat brain GIuR-K1 (11). Amino acids that are identical among the three proteins are shaded and conservatively substituted amino acids are boxed. Putative transmembrane regions are indicated by bold lines over the amino acid sequence. Computer analysis of these sequences shows that 35% of the amino acids are identical for the three proteins and 62% are conservatively substituted.

library (17). A cDNA encoding the entire KBP was constructed from four overlapping clones. This cDNA codes for 487 amino acids, with a 17 amino acid hydrophobic segment at the N terminus which corresponds to a signal peptide (Figure 1). The proposed mature protein contains 470 amino acids and has a calculated molecular weight of 52,500. Analysis of regional hydrophobicity of the deduced amino acid sequence of the KBP shows four putative transmembrane regions. Such a structure has been found for all neurotransmitter-gated ion channels thus far cloned (1). Two glycosylation sites are found in the proposed N-terminal extracellular segment. To characterize the [3HJ-KA binding properties of the KBP cDNA. an expressible cDNA containing the enure open read- ing frame was transfected into COS-7 cells. Scatchard analysis of binding to the membrane fraction of harvested cells showed the presence of a single class of [ 3H]-KA binding site with a Q of 5.6 nM (17). This fmding contrasts with that ofthe KBP affiity-purified from frog brain, which contains both high (Q 5.5 nM)- and low (Q 34 nM)-affinity sites (9). The absence of the low-affinity site in COS-7 cells may indicate that the presence of a second subunit is required for expression of this site. Alternatively, different post- translational modifications between frog brain and transfected COS-7 cells could explain the absence of the low-affinity site. In- j&n of RNA transcribed from the KBP cDNA into Xenopus krevli oocytes faded to produce a physiologically active KA-gated ion channel (17).

With the recent publication of the cloning of a KBP from chick brain (7) and a KA receptor from rat brain (GluR-K1) ( l l ) , the structures and properties of the three proteins can be compared

(Figure 1). The size of the two binding proteins is similar, and they are structurally similar as well, with 5 5 % amino acid identity. The rat brain receptor is nearly twice the size of the binding proteins, but comparing the carboxy terminal half of the rat brain receptor with the entire sequence of the frog and chick brain KBPs shows a very similar spacing of the four hydrophobic segments and sig- nificant sequence homology among the three proteins. With this alignment, 35% of the amino acid residues are identical among the three proteins. This high degree of homology and the conserved spacing of the hydrophobic domains provides convincing evidence that the two KBPs and the carboxy terminal region of GluR-K1 are evolutionarily related. The structural similarity of the frog KBP to the rat brain receptor suggests that these two proteins may be functionally similar as well. Howcver, GluR-K1 forms a functional receptor when its mRNA is injected into oocyte ( l l ) , whereas KBP does not.

Localization of KBP in Frog Brain The localization of KBP was used to address several questions con- cerning its function, raised by the biochemical and cloning experi- ments. Three approaches were used to localize KBP in frog brain: receptor autoradiography, immunocytochemistry, and in situ hybridization. To determine if the antibodies were selective for the major KA binding site in frog brain, the distribution of KA binding sites determined with receptor autoradiography and the distribution of antibody binding sites determined with immunocytochemistry were compared (3) . [ )H J-Ka binding was done at both



1720 WENTHOLD, HAMPSON, WADA, HUNTER, OBERDORFER. DECHESNE

A L

8 and 80 nM to favor labeling of the high- and low-affinity sites. respectively. These studies showed that throughout the frog CNS (a) the binding done at 8 and 80 nM [3H]-KA gave similar labeling patterns, and (b) the immunostaining patterns obtained with both polyclonal and monoclonal antibodies were similar to the binding pattern obtained through receptor autoradiography (Figure 2). These results support the biochemical data (9) suggesting that the high- and low-affinity [ 3H]-KA binding sites are associated with the same molecule in frog brain. Further analysis of KBP im- munoreactivity at the light microscopic level showed the labeling to be primarily punctate, with some fiber labeling. No immunoreactive cell bodies were seen. At the electron microscopic level, immunostaining was found associated with synapses as well as at extrasynaptic sites in the neuropile (Figure 3) (3). At synapses, staining was found on the cytoplasmic side of the postsynaptic membrane, with cytoplasmic staining often associated with this postsynaptic

Figure 2. Comparison of KBP localization in the frog optic tectum obtained with (A) immunocytochemistry, (B) high and (C) low- affinity I3HJ-KA binding, and (0) in situ hybridization histochemistry. KA binding sites and immunostaining are apparent in layers 3,5,7 and lamina F of the optic tectum, while KBP mRNA is found in layers 2,4 and 6. OT, optic tectum; PCTN, posterocentral thalamic nucleus; VIN, infundibular nucleus. Bar = 05 mm. (Reproduced with permission from reference 3)

staining. However, synaptic localization of KBP antibodies was very rare. The major staining was extrasynaptic and appeared as dense patches on the cytoplasmic membrane of unmyelinated processes.

The mRNA encoding the KBP was localized through in situ hybridization histochemistry using synthetic oligonucleotides based on the cDNA sequence of KBP. The distribution of KBP mRNA was compared to the distributions of high- and low-affinity [3H]- KA binding and to the distributions of anti-KBP antibody binding in the optic tectum (Figure 2). These results showed that immunostaining and [3H]-KA binding sites are found in the plexiform layers 3, 5 , 7 and lamina F. while transcripts are heaviest in the cellular layers 2.4 and 6 (Figure 2). Dendritic processes of the neurons of layers 2 .4 and 6 largely contribute to the structure of layers 3 and 5 ; similarly, dendritic collaterals of neurons in layer 6 extend into lamina F. Therefore, the results of the immunocytochemistry and of the in situ hybridization are concordant. Consistent with



KAIMC ACID BINDING PROTEIN IN FROG BRAIN 1721

Figure 3. Electron micrographs of immune staining in the optic tectum show (A,B) synaptic and (C) extrasynaptic labeling. In A and B. the postsynaptic membrane is intensely stained (arrowhead) and intracellular staining is present in the postsynaptic neurite. In C, intensely stained patches are found on neurite cytoplasmic membranes (arrowhead). A and B am not counterstained. Bars: A,B = 0.25 pm; C = 0.5 pm. (Reproduced with permission from reference 3)

the results of the receptor autoradiography and immunocytochemistry studies, the heaviest labeling obtained with in situ histochemistry is seen in the telencephalon (Figures 4-6). However, the KBP mRNA is concentrated in the area surrounding the lateral ventricles, whereas immunocytochemical labeling was distributed throughout the telencephalon. This pattern is expected, since most of the neuronal cell bodies in the telencephalon, especially in the pallium. are near the lateral ventricles, while their dendrites extend into the more superficial areas of the telencephalon. In the less densely labeled areas, clusters of silver grains appear over some neurons, confirming the presence of KBP in neurons (Figures I and

6). The heavy labeling seen near the ventricle may be due. in part, to ependymal cell labeling.

Discussion From these studies it can be seen that KBP is a protein that has two important properties of a neurotransmitter receptor: (a) it binds a ligand in a pharmacologically appropriate manner, and (b) it is structurally related to the family ofchannel-linked neurotransmitter receptors and is very similar to a functional EAA receptor. How- ever, the lack of physiological function after injection of KBP mRNA

Figure 4. Parasagittal section through the frog brain showing distribution of KBP mRNA. Labeling is seen throughout the brain but is most intense in telencephalon in areas bordering the lateral ventrlcles. Hy- bridization was done with %-labeled 48- base oligonucleotide probe complementary to KBP nucleotide sequence at residues 421-468 (17). After hybridization. slides were exposed to X-ray films to detect labeling. An oligonucleotide probe encoding the sense strand of KBP was used as a control and showed no specific labeling. a. optic tectum; TS. torus semicircularis; AA, amygda- lar area; Pa, pallium; Ven, lateral ventricle. Bar = 0.83 pm.

M

I 'If A A I Pa

Ven



1722 WENTHOLD, HAMPSON. WADA. HUNTER, OBERDORFER. DECHESNE

Figure 5. Darkfield autoradiograph of in situ hybridization localization of KBP mRNA in frog brain telencephalon showing intense labeling (asterisks) near border of lateral ventricle. Arrows show clusters of silver grains over neuronal cell bodies. In situ hybridization was carried out as described in Figure 4. but slides were dipped in Kodak NTB-3 nuclear emulsion. Bar = 67 pm.

into oocytes raises questions about its function in frog CNS. A num- ber of explanations can be proposed for this lack of activity, and these fit into two major categories. The first is that KBP is a receptor or part of a receptor but does not exhibit functional activity in the oocyte expression system. The oocyte may not be an appropriate system for expressing KBP, or KBP may be a component of a functional receptor but by itself it is not functional. It is known from work on other neurotransmitter receptors that multiple subunits are often needed for physiological function (4.5.16). The second explanation is that KBP is not part of a functional receptor

complex. Its structud shdarity to channel-linked neurotransmitter receptors and its ability to bind KA may be due to the fact that KBP evolved from a functional receptor but now has a distinct function. The distribution studies also do not support a typical neurotransmitter receptor role for KBP in frog brain. KBP is largely confined to the CNS and appears to be primarily neuronal, and it is present in only certain populations of neurons, which is consistent with the interpretation that it is related to the neurotransmitter used by the neuron. However, only a small percentage of KBP is located at the synapse, where a neurotransmitter receptor

c .f'

Figure 6. Brightfield autoradiograph of in situ hybridization localization of KBP mRNA in frog brain telencephalon. There is intense labeling of neuronal cell bodies bordering the lateral ventricle (A) and scattered labeling of cell bodies (arrows) in more superficial areas of the telencephalon (9). Methods as in Figures 4 and 5. Sections were counterstained with methyl green. Bars: A = 16.7 p; B - 33 pm.



KAINlC ACID BINDING PROTEIN IN FROG BRAIN 1723

would be expected to be concentrated. Such a localization suggests a dual function for KBP in frog brain, where only the synaptic KBP would be functioning as a receptor.

I t remains to be determined i f a homologue of KBP is present in mammalian brain. The binding analyses showing pharmacological and biochemical similarities between the binding sites from frog and mammalian brains support the presence of a mammalian form of KBP, although much less abundant than in many non- mammalian sources. A n obvious candidate for the mammalian homologue of KBP is GluR-K1. Although the sequence homology between KBP and GluR-K1 clearly indicates that they are related proteins, i t is probably not a sufficient basis o n which to conclude that GluR-K1 is the rat homologue of KBP. More likely, GluR-K1 and KBP represent subpopulations of a large family of EAA receptors that are structurally and functionally related but are distinct gene products. As presented above, one of the monoclonal antibodies (KAR-Bl) to the frog brain KBP labeled a protein with an Mr of about 100,000 in rat brain. We now have evidence that this antibody recognizes GluR-K1 and probably similar members of this family a t the same molecular weight (unpublished observation). This is not unexpected, given the sequence similarity between the two proteins. However, our preliminary data also suggest that GluR- K1 is not responsible for KA binding in mammalian brain. There- fore, the possibility remains that a KBFWre protein, which is responsible for KA binding and has an Mr of about 48,000, is present in mammalian brain. The fact that such a protein is not seen on immunoblots of mammalian brain using KAR-B1 would be ex- plained by the low abundance of this protein, which would not be detected with this method.

Literature Cited 1. Barnard EA, Darlison MG, Seeburg P: Molecular biology of the

GABAA receptor: the receptorkhannel superfamily. Trends Neuroxi 10502, 1987

2. Dechesne CJ, Hampson DR, Goping G, Wheaton KD. Wenthold RJ: Evidence for kainic acid receptor on non-neuronal cells in the frog hby- rinth. Neurosci Abst 1S:209, 1989

3. Dechesne CJ, Oberdorfcr ME, Hampson DR, Wheaton KD, N a z d i AJ, Wenthold RJ: Distribution of a putative kainic acid receptor in the frog central nervous system determined with monoclonal and polyclonal antibodies: evidence for synaptic and extrasynaptic localization. J Neuroni 10:479, 1990

4. Deneris ES, Connolly J, Boulter J, Wada E, Wad? K, S w n W, Patrick J, Heinemann S Primary structure and expression of beta 2: a novel subunit of neuronal nicotinic acetylcholine receptors. Neuron 1:4S, 1988

5 . Duvoisin RM, Deneris ES, Ruick J, Heinemann S The functional diw- sity of the neuronal nicotinic acetylcholine receptors is increased by a now1 subunit: W. Neuron 3:487, 1989

6. Foster AC, Fagg GE: Acidic amino acid binding sites in mammalian neuronal membranes: their characteristics and relationship to synaptic

7. Gregor P, Mano I, Maoz I, McKeown M. Teichbcrg VI: Molecular structure of the chick cerebellar hinate-binding subunit ofa putative glutamate receptor, Nature 342:689, 1989

8. Hampson DR. Huie D, Wenthold RJ: Solubilization ofkainic acid binding sites from rat brain. J Neurochem 49:1209, 1987

9. Hampson DR. Wenthold RJ: A kainic acid receptor from frog brain purified using domoic acid Sinity chromatography. J Biol Chem 263:2500, 1988

10. Hampson DR, Wheaton KD, Dechesne CJ, Wenthold RJ: Idcntifca- tion and characterization of the ligand binding subunit o f a kainic acid receptor using monoclonal antibodies and peptide mapping. J Biol Chem 26413329, 1989

11. Hollmann M, O’Shea-Greenfield A, Rogers SW, Heinemann S: Clon- ing by functional expression of a member of the glutamate receptor family. Nature 342:643, 1989

12. Hunter C, Wcnrhold RJ: Characterization of solubilized AMPA binding sites with polydonal and monoclonal antibodies. Neuroxi Abst 16:543, 1990

13. Monaghan M: Bridge RJ, Cotman CW: The excitatory amino acid mcp- tors. Annu Rev Phannacol Toxicol 29:365. 1989

14. Sladeczek F, Pin JP, Recvens M, Bockaert J, Wciss S: Glutamate stimu- lates inositol phosphate formation in striatal neurones. Nature 317:717, 1985

IS. Sugiyama H, Ito I, Watanabe M: Glutamate receptor subtypes may be classified into two major categories: a study on Xenopus oocytes injected with rat brain mRNA. Neuron 3:129, 1989

16. Wadi K, Ballivet M, Boulter J , Connolly J, Wad? E, Deneris E, Swm- son JSV, Heinemann S, Patrick J: Functional expression of a new phu- macological subtype of brain nicotinic acetylcholine receptor. Science 240:330, 1988

17. Wada K, Dechesne CJ, Shimvaki S, King RG, Kusano K, Buonanno A, Hampson DR, Banner C, Wenthold RJ, Nakatani Y Sequence and expression of a frog brain complementary DNA encoding a kainate- binding protein. Nature 342:684, 1989

18. Watkins JC, Evans RH: Excitatory amino acid transmitters. Annu Rev Pharmacol Toxicol 21:165, 1981

fCCCptOK. Bf& RCS RCV 7:103, 1984



isolation, localization, and cloning of a kainic acid binding protein

Documents