Immunolocalization of metabotropic glutamate receptor 7 in the rat olfactory bulb

Immunolocalization of MetabotropicGlutamate Receptor 7 in the Rat

Olfactory Bulb

J. MARK KINZIE,1,2 MICHI M. SHINOHARA,1,2 ANTHONY N. VAN DEN POL,4

GARY L. WESTBROOK,1,3 AND THOMAS P. SEGERSON1,2*1Vollum Institute, Oregon Health Sciences University, Portland, Oregon 97201

2Department of Medicine, Oregon Health Sciences University, Portland, Oregon 972013Department of Neurology, Oregon Health Sciences University, Portland, Oregon 97201

4Section of Neurosurgery, Yale University School of Medicine,New Haven, Connecticut 06510

ABSTRACTMetabotropic glutamate receptors (mGluRs) constitute a large family of G-protein-

coupled receptors that are subdivided into three groups based on sequence similarity,pharmacological profiles, and coupling to second messengers. Although mRNAs for seven ofthe eight mGluRs are expressed in the olfactory system, the localization and function ofspecific subtypes have not been fully characterized. Mitral cells of the olfactory bulb expressmRNA for several mGluRs, including mGluR7, which has been suggested as a presynapticglutamate autoreceptor. To investigate the immunolocalization of mGluR7 in the olfactorysystem, we used a polyclonal antiserum specific for the carboxy terminus of the receptor.Mitral cell somata and proximal dendrites were strongly labeled by the mGluR7 antibody.Electron microscopic analysis revealed that most of the mitral cell somatic staining wascytoplasmic. In olfactory bulb glomeruli, immunoreactivity was present in axons anddendrites. In the piriform cortex, diffuse staining was present in layer Ia that was markedlyreduced following bulbectomy, consistent with expression of mGluR7 in mitral cell axonterminals. Electron microscopic analysis of this region confirmed the presence of mGluR7 inmultiple axon terminals. Distinct labeled fibers in all levels of layer I appeared to originatefrom labeled piriform cortex pyramidal cells in layers II and III. Our results indicate thatmGluR7 is primarily presynaptic at olfactory bulb synapses. However, the postsynapticlocalization of mGluR7 at selected synapses indicates that mGluR7 is not targeted exclusivelyto axonal compartments. J. Comp. Neurol. 385:372–384, 1997. r 1997 Wiley-Liss, Inc.

Indexing terms: presynaptic; confocal; 2-amino-4-phosphonobutyrate; metabotropic glutamate

receptor 7; mitral

The olfactory system is designed to discriminate diluteconcentrations of thousands of airborne odorant mol-ecules. Afferent inputs from the olfactory epithelium reachthe olfactory bulb and synapse on the apical dendrites ofmitral and tufted cells within glomeruli. In the bulb, bothcentrifugal and intrinsic pathways are responsible for thespatial encoding and synaptic modulation of incomingsensory information. Many studies suggest that glutamateis the primary excitatory neurotransmitter at olfactorynerve terminals, at mitral cell output of dendrodendriticsynapses, and at mitral cell axon terminals in olfactorycortex (Trombley and Shepherd, 1993). Recent evidencehas emphasized that metabotropic glutamate receptors(mGluRs) may have a significant role in olfactory informa-tion processing at dendrodendritic reciprocal synapses in

the bulb (Hayashi et al., 1993; Kaba and Nakanishi, 1995).Likewise, mGluR agonists can inhibit the output of thebulb by reducing transmitter release from mitral cellaxons that project through the lateral olfactory tract (LOT)to olfactory cortical structures, including the piriformcortex. Immunolocalization of mGluRs is particularly im-portant to an understanding of olfactory bulb function,

Grant sponsor: NIH; Grant numbers: MH10314, DC01783, NS10174.*Correspondence to: Thomas P. Segerson, Department of Medicine,

Division of Endocrinology, Oregon Health Sciences University, L607, 3181S.W. Sam Jackson Park Road, Portland, OR 97201-3098.E-mail: [email protected]

Received 19 June 1996; Revised 19 March 1997; Accepted 14 April 1997

THE JOURNAL OF COMPARATIVE NEUROLOGY 385:372–384 (1997)

r 1997 WILEY-LISS, INC.

given the predominance of dendrodendritic synaptic plas-ticity and the juxtaposition of pre- and postsynaptic special-izations.

Activation of mGluRs by glutamate modulates synaptictransmission through the inhibition or stimulation of ionchannels and intracellular signal transduction systems(Abe et al., 1992; Westbrook, 1994; Pin and Bockaert,1995). Cloning of mGluR cDNAs has identified eightsubtypes that fall into three groups based on amino acidhomology, signal transduction mechanism, and pharmaco-logical profile. Group I receptors (mGluR1,5) are coupled toinositol phospholipid hydrolysis, whereas group II(mGluR2,3) and III (mGluR4,6,7,8) are negatively coupledto adenylate cyclase (Knopfel et al., 1995). In addition,alternative mRNA splicing of some mGluR subtypes givesrise to several distinct molecules that may vary in theirdistribution and signaling characteristics (Pin et al., 1992;Hampson et al., 1994; Minakami et al., 1994; Joly et al.,1995). Messenger RNAs coding for seven of eight mGluRsubtypes are expressed in the olfactory bulb (Abe et al.,1992; Shigemoto et al., 1992; Ohishi et al., 1993a,b, 1995a;Duvoisin et al., 1995; Kinzie et al., 1995). The onlyexception is mGluR6, which is restricted to retinal depolar-izing bipolar cells, wherein it activates a postsynapticphosphodiesterase (Nomura et al., 1994). However, thecellular pattern of expression of most mGluR proteins isonly beginning to be explored. For example, whetherparticular subtypes are exclusively pre- or postsynaptic isunclear. In many brain regions, one splice variant ofmGluR1, mGluR1a, has been localized to postsynapticspecializations of dendrites (Martin et al., 1992; Baude etal., 1993; Craig et al., 1993; Fotuhi et al., 1993). Bycontrast, mGluR1a immunoreactivity has been found atthe presynaptic specialized membrane at dendrodendriticsynapses in the external plexiform layer (EPL; van denPol, 1995).

In this study, we focused on mGluR7, a group III mGluRthat is expressed in mitral cells, tufted cells, and groups ofjuxtaglomerular neurons in the bulb (Kinzie et al., 1995;Ohishi et al., 1995a). Group III receptors are activatedselectively by 2-amino-4-phosphonobutyrate (L-AP4),which, in physiological experiments, causes inhibition oftransmitter release from mitral cell axons (Anson andCollins, 1987; Trombley and Westbrook, 1992), suggestingthat group III receptors, including mGluR7, are presynap-tically localized. We used a polyclonal rabbit antiserumagainst rat mGluR7 to determine the cellular and subcellu-lar localization of mGluR7 in olfactory bulb neurons. Ourdata indicate that olfactory bulb mGluR7 is principallypresynaptic, although dendritic mGluR7 immunoreactiv-ity suggests that postsynaptic actions of mGluR7 should bereassessed.

MATERIALS AND METHODS

Antisera

An antiserum to the carboxy terminus of the rat mGluR7(peptide sequence KNSPAAKKKYVSNNLVI) was gener-ously provided by Stefania Risso Bradley and P. JeffreyConn of Emory University (Bradley et al., 1996). Anaffinity-purified antibody to rat mGluR1a was provided byEileen Mulvihill and Betty Haldeman of Zymogenetics(Baude et al., 1993; Hampson et al., 1994). Monoclonalantibodies against glial fibrillary acidic protein (GFAP),microtubule-associated protein (MAP) 2A/B/C, MAP5, and

a polyclonal goat antibody against calretinin were ob-tained from Chemicon International, Inc. (Temecula, CA).

Unilateral bulbectomy

Two adult female Sprague-Dawley rats were deeplyanesthetized with ketamine (40 mg/kg) and acepromazine(1 mg/kg) and were positioned in a stereotaxic apparatus.The olfactory bulb on one side was exposed and removed byaspiration. Ablation was considered complete by exposureof the cribriform plate. The site of the lesion was filled with1% agar, and the wound was closed. After 7 days, rats wereanesthetized, killed by decapitation, and their brains wereprocessed for immunohistochemistry.

Immunoblot analysis of mGluR7

Baby hamster kidney-21 (BHK) cells stably expressingcDNAs encoding mGluRs 1a, 4, or 7 were used to deter-mine antiserum specificity (Houamed et al., 1991; Thom-sen et al., 1992; Saugstad et al., 1994). Wild type andmGluR-expressing BHK cells were grown to 75% conflu-ency, rinsed with phosphate-buffered saline (PBS; 137 mMNaCl, 2.7 mM KCl, 4.3 mM Na2HPO4 ?7H20, 1.4 mMKH2PO4, pH , 7.3), and harvested with membrane-harvesting buffer [10 mM sodium bicarbonate, pH 7.2,containing 1 mM phenyl methyl sulfonyl fluoride (PMSF)and 1 µg/ml each leupeptin, pepstatin A, aprotinin; Sigma,St. Louis, MO]. Membranes were centrifuged at 316,000gfor 5 minutes, resuspended in the membrane-harvestingbuffer, and stored at 280°C until use. Oocyte membraneswere harvested by trituration in disruption buffer (7.5 mMsodium phosphate, dibasic, pH 7.4, 10 mM EDTA, 1 mMPMSF, and 1 µg/ml each leupeptin, pepstatin A, aprotinin)followed by centrifugation at 3300g for 5 minutes. Thesupernatant was centrifuged at 316,000g for 30 minutes,and the membrane pellets were resuspended in disruptionbuffer containing 1% sodium dodecyl sulfate (SDS), dena-tured at 37°C for 30 minutes, and stored at 220°C untiluse.

For immunoblot of rat brain regions, brains were rapidlyremoved from four anesthetized rats, and the olfactorybulb, hippocampus, and cerebellum were dissected andhomogenized in 20 ml of 0.32 M sucrose, 10 mM Tris, and 1mM MgCl2, pH 7.0, containing protease inhibitors (1 µg/mlaprotinin, 1 µg/ml pepstatin, and 1 mM PMSF). Aftercentrifugation (31,000g for 10 minutes at 4°C) the super-natant was centrifuged at 330,000g for 20 minutes, andthe pellet was resuspended in 2 mM HEPES and 2 mMEDTA, pH 7.5, with protease inhibitors. The crude synap-tosomes were homogenized for 10 seconds and incubatedon ice for 30 minutes. Lysed membranes were isolated bycentrifugation (320,000g for 20 minutes at 4°C) andquantified. Ten micrograms of protein were heated to 50°Cfor 10 minutes prior to gel fractionation.

Membrane protein aliquots were subjected to SDS-polyacrylamide gel electrophoresis (SDS-PAGE; 10% acryl-amide) and transferred to nitrocellulose membranes (Bio-Rad, Hercules, CA) by electrophoresis. The blots wereblocked with 5% nonfat dry milk in Tris-buffered saline,pH 8.0, with 0.1% Tween-20 (TBST) at 4°C overnight.Blots were then incubated for 4 hours at room temperaturewith anti-mGluR7 antiserum (1:8,000) containing sodiumazide (0.025%) in TBST.

Membranes were rinsed and incubated for 45 minutes inTBST with horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin (IgG; 1:10,000; Amersham, Arling-

mGluR7 IN THE RAT OLFACTORY BULB 373

ton Heights, IL) at room temperature. Following severalwashes in TBST, immunoreactive proteins were visualizedwith enhanced chemiluminescence (Amersham), as recom-mended by the manufacturer.

Immunohistochemistry

Adult female Sprague-Dawley rats (n 5 18) were deeplyanesthetized with pentobarbital and perfused transcardi-ally with ice-cold saline, followed by ice-cold 4% paraformal-dehyde in 100 mM sodium borate, pH 9.5. The brains wereremoved quickly, blocked, and postfixed for 4 hours at 4°Cin 4% paraformaldehyde in 100 mM borate buffer, pH 9.5.The brains were rinsed six times in PBS, followed bysectioning (75 µm) on a Vibratome 1000 Sectioning System(Lancer, St. Louis, MO).

For diaminobenzidine (DAB) immunohistochemistry, sec-tions were rinsed in 14 mM NaCl, 2 mM K2HPO4 (KPBS)and blocked in 2% normal donkey serum and 0.02% non-fat dry milk in KPBS (1 hour at room temperature). Thesections were then incubated in anti-mGluR7 serum(1:10,000) in 2% normal donkey serum and 0.02% nonfatdry milk in KPBS (3 days at 4°C). After rinsing in KPBS,sections were treated for endogenous peroxidase activity(1% H202 in KPBS; 30 minutes at room temperature).Sections were then incubated in biotin-SP donkey anti-rabbit IgG at 1:200 (Jackson ImmunoResearch, WestGrove, PA) for 4 hours, followed by incubation in avidin-biotin-peroxidase complex (Vector Laboratories, Burlin-game, CA) for 2 hours. Sections were developed with an Ni-and Co-enhanced DAB/peroxide reaction (Pierce ChemicalCo., Rockford IL), mounted on gelatin-coated slides, dehy-drated in graded alcohols, cleared in xylene, and thencoverslipped with DPX.

For a control for specificity, some sections were incu-bated without primary antisera. In addition, some sectionswere incubated in antiserum that had been preincubatedin peptide (10 µg/ml) corresponding to the C-terminalsequence of mGluR7 (KNSPAAKKKYVSNNLVI) fromwhich the antiserum had been generated.

For fluorescent immunohistochemistry, two methodswere used with equal efficacy. Some sections were cut on afreezing microtome (50 µm), rinsed, then permeabilized in0.3% Triton X-100/KPBS, and blocked as described abovewith the addition of 0.3% Triton X-100. The sections werethen incubated in anti-mGluR7 (1:1,000) or anti-mGluR1a(1:500) sera overnight at 4°C. After rinsing sections in0.3% Triton X-100/KPBS, sections were incubated in bio-tin-SP donkey anti-rabbit IgG at 1:200 (Jackson Immu-noResearch) for 4 hours, followed by Cy3/streptavidin at1:200 (Jackson ImmunoResearch). Sections were mountedand coverslipped described as above. Other sections werecut on a Vibratome (75 µm), rinsed in KPBS, and blockedwithout permeabilization (three animals). The sectionswere then incubated in anti-mGluR7 (1:10,000) or anti-mGluR1a (1:500) sera for 3 days at 4°C. After rinsing inKPBS, sections were then incubated in Cy3 donkey anti-rabbit secondary antibody (1:200; Jackson ImmunoRe-search), rinsed, and mounted. Slides were then imaged ona BioRad MRC1024 confocal imaging system with a NikonDiaphot 200 microscope. The scanned images were ad-justed for brightness, contrast, and size in Photoshop 2.0.1(Adobe Systems, Inc., Mountain View, CA) and labeled inFreehand 3.1 (Macromedia, San Francisco, CA) on a

Macintosh 8100/80. Anatomical sites were identified byusing the Brain Maps atlas (Swanson, 1992).

Double-labeled fluorescent immunohistochemistry wasperformed as described above, using a mouse monoclonalprimary antibody (anti-MAP2, anti-MAP5, anti-GFAP)coincubated with the anti-mGluR7 primary antibody. La-beling by monoclonal antibodies was detected using aCy5-labeled anti-mouse secondary antibody (1:200) coincu-bated with the Cy3-labeled anti-rabbit secondary antibody(1:200). Controls included the omission of one primaryantiserum and incubation with two different fluorescentsecondary antibodies. The slides were visualized with red(excitation l: 568 nm/10 nm dichroic filter, emission l: 605nm/32 nm dichroic filter) and far red (excitation l: 647nm/10 nm dichroic filter, emission l: 680 nm/32 nmdichroic filter) filter sets to insure that there was nocross-over detection of the respective emission spectra ofthe fluorescent labels.

Immunoelectron microscopy

Adult Sprague-Dawley rats were heavily anesthetizedwith Nembutal (80 mg/kg) and perfused transcardiallywith physiological saline, followed by a fixative that con-tained 4% paraformaldehyde and 0.05% glutaraldehyde ina 0.1 M phosphate buffer. Fifty-micrometer-thick Vibra-tome sections were immersed in a graded series of sucroseconcentrations up to 30% sucrose. Thick sections thenwere frozen twice in liquid nitrogen to lyse membranesand to increase penetration of immunoreagents. Sectionsof the olfactory bulb and forebrain caudal to the olfactorypeduncle were incubated for 30 minutes in normal goatserum containing 0.1% lysine, 0.1% glycine, and 1% bovineserum albumin and were then placed in the anti-mGluR7antiserum. After incubation in biotinylated goat anti-rabbit IgG and then in avidin-biotin-peroxidase complex(Vector Laboratories), sections were treated with DAB andhydrogen peroxide, osmicated for 30 minutes in 1% os-mium tetroxide, and dehydrated through an ascendingseries of ethanol concentrations to 100%. To increasemembrane contrast, the 70% ethanol contained 1% uranylacetate. Some sections were also stained with 1% leadacetate, except where indicated. For controls, some sec-tions were incubated without primary antibody, secondaryantibody, or both. Ultrathin Epon sections were cut on aReichert ultramicrotome (Reichert-Jung, Nussloch, Ger-many) and picked up on Formvar-coated, carbon-stabi-lized, single-slot grids. Sections were examined in a JEOL1200 EXII transmission electron microscope (JEOL USA,Inc., Peabody, MA) at 60,000 accelerating volts. All animalprocedures were approved by the Oregon Health SciencesUniversity Animal Care Committee and were in conform-ance with NIH guidelines on the Care and Use of Animalsin Research.

RESULTS

Antibody specificity

To determine the specificity of the carboxy-terminalantipeptide antiserum for mGluR7, we performed immuno-blots with this antibody of cell membranes obtained fromdissected rat brain regions and from mGluR1a-, mGluR4-,and mGluR7-expressing and wild type BHK cells. Immuno-blots of olfactory bulb and hippocampal membrane proteinresolved a single band of approximately 130 kDa, whereas

374 J.M. KINZIE ET AL.

mGluR7-expressing BHK cell immunoblots demonstrateda doublet of similar mobility (Fig. 1A). This doublet is mostlikely due to alternate glycosylation of the mGluR7 protein(Bradley et al., 1996). In addition, the antisera recognizeda single band in membranes prepared from Xenopusoocytes expressing mGluR7, but not mGluR8, demonstrat-ing further that the antiserum was specific for mGluR7(not shown). Expression of mGluR7 and mGluR8 in oo-cytes was confirmed by coexpression with G-protein-coupled, inward-rectifying potassium channel (Saugstadet al., 1997).

Preincubation of sections in peptide corresponding tothe carboxy-terminal sequence of mGluR7 from which theantisera had been generated reduced DAB staining prod-uct to background levels (compare Fig. 1B with Fig. 1C).An antiserum purified by affinity for the immunizingpeptide that recognizes a single protein species in immuno-blots of olfactory bulb protein (Bradley et al., 1996) gave

the same results as the antiserum used in these studies(not shown). Therefore, because the affinity-purified antise-rum was not available in sufficient quantity, crude antise-rum for mGluR7 was used for the remainder of theexperiments. No staining was observed when the antise-rum was omitted from the primary incubation (data notshown). Each experimental result was confirmed by exami-nation in multiple (two to six) animals.

mGluR7 immunoreactivityin the olfactory bulb

Sections of adult rat olfactory bulb and olfactory cortexwere initially stained by using the avidin-biotin complexmethod, followed by development with an Ni- and Co-enhanced DAB/peroxide reaction. At low-magnificationlight microscopy, somata of mitral cells and internal andintermediate tufted cells of the main olfactory bulb (MOB)were the most visibly stained, but granule cells werelightly and inconsistently stained (Fig. 1B). The EPL wasnot strongly stained, especially when compared with thestaining pattern of mGluR1a (van den Pol, 1995). Themost prominent staining was in dendritic processes ofmitral cells that traversed the EPL in perpendicularorientation to the mitral cell layer. No staining wasapparent in the olfactory nerve layer or in the LOT.

To trace labeled fibers coursing through the tissueplanes, the remaining studies were analyzed with fluores-cence by using a BioRad MRC 1000 confocal microscope,which provides better contrast and resolution. The methodhas the added benefit of reducing artifact from diffusion ofthe reaction product. At lower magnification, the stainingpatterns were identical, with the most predominant stain-ing seen in the cell bodies of mitral and internal tuftedcells (Fig. 2A). The labeled dendrites of the mitral cellscould be traced nearly to the glomerular layer. In addition,secondary dendrites projecting parallel to the mitral celllayer were labeled. Dendrites of internal and intermediatetufted cells, often more than one per cell, were also labeled.There was light, diffuse, punctate staining of the glo-meruli, but small, stained, juxtaglomerular cell bodies,consistent with periglomerular cells, as well as larger,juxtaglomerular, stained cell bodies, suggestive of superfi-cial tufted cells, could be discerned (Fig. 2B).

At higher magnification, mGluR7 immunoreactivity waspresent in the MOB mitral cell layer with a cytoplasmicstaining pattern (Fig. 2C). The nucleus was not labeled.The staining intensity was greater in proximal dendritesand was reduced more distally. Double-labeling experi-ments with MAP5, a cytoskeletal protein expressed inmitral cell bodies, axons, and dendrites, demonstratedcolocalization with mGluR7 in mitral cell bodies anddendrites, but no mGluR7 immunoreactivity was seen inMAP5-labeled axons (data not shown). Within the innerplexiform layer, occasional punctate staining of fibers justdeep and parallel to the mitral cell layer was observed(Fig. 2D). These fibers could be traced to nongranule cellsin the granule cell layer and may represent short axoncells.

Staining of mitral and granule cell bodies was observedin the accessory olfactory bulb (AOB; Fig. 3), whereas thebright staining with immunofluorescence (Fig. 3B) was notseen if the tissue was preincubated with the immunizingpeptide. Because the cell bodies of the mitral cells in thisregion are densely packed, process staining could not be

Fig. 1. Specificity of metabotropic glutamate receptor 7 (mGluR7)antiserum. A: Immunoblot analysis was used with polyclonal antise-rum directed against mGluR7 to determine reactivity in protein frommembranes of dissected rat olfactory bulb (OB), hippocampus (Hip),and cerebellum (C) as well as from membranes obtained from wildtype baby hamster kidney (BHK) cells (wt) and BHK cells expressingmGluR1a (1a), mGluR4 (4), and mGluR7 (7). Relative mobility, asdetermined by molecular mobility standards, is shown at the left inkDa. An immunoreactive band of about 130 kDa was observed inprotein extracts from the olfactory bulb and hippocampus, with nodetection of mGluR7 in the cerebellum. In cell lines, mGluR7immunoblot detected a doublet of similar mobility only in extracts ofcells expressing the mGluR7 cDNA. B: Diaminobenzidine (DAB)staining of main olfactory bulb for mGluR7. Mitral cell somata (Mi)and dendrites (arrow) are labeled in sections incubated with theantiserum used in these studies. C: Sections stained with the antise-rum preincubated with 10 µg/ml immunizing peptide had muchreduced staining, with a mitral cell body (Mi) barely visible. Scalebar 5 50 µm.


Fig. 2. Confocal immunofluorescence analysis of mGluR7 in the ratolfactory bulb. A: Staining with the anti-mGluR7 antiserum revealslabeling of cell bodies and their proximal dendrites (arrows) in themitral cell layer (MCL) and in the external plexiform layer (EPL).Smaller cells were labeled around the glomeruli (Glo). Cell bodies inthe granule cell layer (GCL) were only slightly stained. B: Immunofluo-rescent labeling of a glomerulus showing diffuse, punctate staining of

neuropil. Groups of small juxtaglomerular cells (small arrow) and lessnumerous larger juxtaglomerular cells (large arrow) are also labeled.C: Higher magnification of the MCL of the main olfactory bulb revealsnuclear exclusion of staining and gradual decrease of labeling inten-sity in dendrites (arrow). D: Coursing parallel to the MCL, distinctlabeled fibers could be visualized (arrow). Scale bars 5 150 µm in A,B,300 µm in C,D.


evaluated. The granule cell layer of the AOB was alsoimmunostained, but the lateral olfactory tract in thisregion was unlabeled. Glomeruli in this region were notdistinctly discernible.

Immunoelectron microscopyof the olfactory bulb

The light microscopic finding of mGluR7 in mitral celldendrites suggested that mGluR7 might be postsynaptic.This was not expected, because the group III mGluRagonist L-AP4 has no described postsynaptic effects exceptin the retina. To investigate the ultrastructural localiza-tion of mGluR7, immunoelectron microscopic analysis ofthe olfactory bulb was performed by using the anti-mGluR7 antiserum. Examination of mitral cells showedmGluR7 in the rough endoplasmic reticulum and the Golgiapparatus (Fig. 4). No staining was observed in nuclei,mitochondria, or lysosomes. Small amounts of stainingwere found on the cell membrane, with no appreciableincrease at synaptic junctions or at reciprocal dendroden-dritic synapses in the EPL. In the granule cell layer, somecytoplasmic staining of narrow processes was seen (notshown), suggestive of astrocyte processes. Granule cellsdid not stain.

Within glomeruli, mGluR7 immunoreactivity was vis-ible in several synaptic compartments. Immunoreactivitywas present in presynaptic specializations that synapsedon other unlabeled presynaptic specializations (Fig. 5A).At axodendritic synapses, mGluR7 immunoreactivity couldbe found both pre- and postsynaptically (Fig. 5B and Fig.5C, respectively). The neuropil of the glomerular layer is

composed of processes from a number of cell types (Shep-herd and Greer, 1990). Immunoreactive profiles, therefore,could include dendrites of mitral, tufted, or periglomerularcells and axons of periglomerular cells or centrifugalfibers.

mGluR7 in the piriform cortexand anterior olfactory nucleus

Within the olfactory peduncle, there was minimal label-ing of the LOT, although a minority of the fibers within thetract appeared moderately labeled (Fig. 6A). In the ante-rior olfactory nucleus (lateral part), there was diffusestaining of the molecular layer and many stained peri-karya within the pyramidal layer (Fig. 6A). In the rostralpiriform cortex, there was diffuse staining in layer Ia butmuch less in layer Ib (Fig. 6A). Perikarya in layer IIpyramidal neurons were immunoreactive. We observed asimilar pattern in the piriform cortex caudal to the pe-duncle, with strong immunoreactivity in layer Ia, theregion of synaptic contacts between mitral cell axons anddendrites of the piriform cortex pyramidal neurons (Fig.6B). However, layer Ib, which contains synapses betweenrecurrent collateral axons and dendrites of the piriformcortex pyramidal neurons, was not stained. Layer IIcontained many immunoreactive cell bodies, as did layerIII. In addition, layer III contained many distinct fiberscoursing through it in a parallel orientation to layer I.Longitudinal sections of the LOT failed to demonstratestrong fiber staining.

In layer Ia, in addition to diffuse staining, distinctarborizing fibers could be seen (Fig. 6C). These fibers

Fig. 3. A: DAB-reacted immunostaining of accessory olfactory bulb for mGluR7, showing staining inthe mitral cell layer (aMCL) and granule cell layer (aGCL) as well as more distinct staining in theadjacent mitral cell layer (MCL) of the main olfactory bulb. B: Immunofluorescence of the accessoryolfactory bulb revealed bright staining of both the aMCL and the aGCL. Scale bars 5 150 µm.


appeared to originate in layer II neurons, process throughto layer Ia, and branch out. Labeled fibers that appeared tooriginate in deeper layers and penetrate into superficiallayers colabeled with MAP2 and mGluR1a, suggestingthat these fibers were dendrites. In contrast to mGluR7,antisera to mGluR1a did not demonstrate bright, diffuselabeling of layer Ia.

Immunoreactivity following ablationof olfactory bulb

To investigate whether the diffuse staining in layer Iarepresented mGluR7 localized to mitral cell axon termi-nals, we examined the contralateral and ipsilateral piri-form cortex 1 week after unilateral bulbectomy (Fig. 7).

Fig. 4. Mitral cells-perikarya. A: Low-magnification micrographshowing peroxidase labeling (arrows) in the cytoplasm of a mitral cellbody. Labeling is associated with rough endoplasmic reticulum (RER).Little label is found on the plasma membrane. B: Higher magnifica-tion of another mitral cell showing peroxidase (long arrows) on the

RER and associated with the Golgi apparatus (GA). Mitochondria (M)and lysosomes show no label. Along the plasma membrane (smallarrows), even at regions of synaptic contact, little labeling is found.Scale bars 5 1 µm in A, 0.5 µm in B.


Staining for GFAP demonstrated the expected reactive chan-ges only in the ipsilateral piriform cortex, but the diffuselayer Ia staining for mGluR7 was not present. However,mGluR7 staining in layer Ia on the contralateral side wassimilar to that seen in nonbulbectomized rats. In contrastto mGluR7 immunoreactivity, the labeling of fibers andcell bodies in layers II and in some fibers of layer I of thepiriform cortex by antisera to mGluR1a was unaltered bybulbectomy, as was the staining for MAP2, a cytoskeletalprotein localized to dendrites. Staining for calretinin, acalcium-binding protein present in mitral cells (Wouter-lood and Hrtig, 1995), was strong in the LOT and nearly asstrong in layer Ia of the piriform cortex. After bulbectomy,there was a significant reduction in the ipsilateral stainingof both regions, demonstrating that mitral cell axons andtheir terminals were reduced following bulbectomy.

Immunoelectron microscopyof piriform cortex

We also examined by ultrastructural analysis of peroxi-dase-stained tissue whether mGluR7-containing axon ter-minals were present in the layer of the piriform cortex inwhich mGluR7 immunostaining was altered by bulbec-

tomy. Some sections were counterstained with uranyl andlead to enhance contrast in nonimmunoreactive cells (Fig.8A,B). Large numbers of immunoreactive axon terminalswere found in the rostral piriform cortex (Fig. 8A–D). Thisstaining was particularly evident in tissue sections thatwere not further stained with heavy metals (Fig. 8C,D).The immunoreactive axons made asymmetrical Gray typeI synaptic contacts, typical of excitatory axons. Presynap-tic axons had small, clear vesicles (Fig. 8A,D), and densecore vesicles sometimes were also evident (Fig. 8D). In manycases, multiple labeled axon terminals made contact withone dendrite (Fig. 8C). In immunoreactive axons, the cyto-plasmic side of the presynaptic membrane was consistentlystained. The apparent labeling of synaptic vesicles wasprobably due to diffusion and precipitation of the peroxi-dase reaction product near the antigen site. Consistent withthis view, vesicles in immunoreactive axons that were far-ther from the presynaptic membrane often were not labeled.

DISCUSSION

Electrophysiological evidence has suggested that groupIII mGluRs generally have presynaptic actions. Because

Fig. 5. Neuropil labeling. A: Labeled presynaptic process in theglomerular layer contains a large number of synaptic vesicles and is insynaptic contact (arrows) with other unlabeled processes that alsocontain synaptic vesicles. B: Immunoreactive, vesicle-rich processwith the appearance of an axon makes synaptic contact (arrows) withunlabeled dendrites in the glomerular layer. Immunoreactivity is

found in the presynaptic axon and is particularly dense near thepresynaptic membrane specialization. C: Dendrite in the glomerularlayer shows peroxidase labeling, particularly associated with thepostsynaptic specialization at contacts with multiple unlabeled, vesicle-containing processes (arrows). Scale bars 5 0.3 µm in A,C, 0.25 µmin B.


mGluR7 mRNA is expressed in mitral cells and L-AP4inhibits release from mitral cell axons (Anson and Collins,1987; Collins and Howlett, 1988; Hasselmo and Bower,1991; Trombley and Westbrook, 1992), we expected thatmGluR7 might be targeted to axon terminals of mitralcells. Our results are consistent with an axonal distribu-tion of mGluR7 in mitral cell processes. However, thecoincident dendritic localization of mGluR7 at postsynap-tic sites in the glomerular layer suggests that mGluR7may also have postsynaptic actions.

Distribution of mGluR7 mRNA and protein

The cellular distribution of mGluR7 immunoreactivityclosely matched that expected from mRNA distributionpatterns; however, the unexpected dendritic staining madeus examine closely the specificity of the antiserum. Al-though most of our experiments were performed with acrude antiserum due to limited availability of affinity-purified antiserum, preadsorption of the antibody with theimmunizing peptide in the absence of carrier proteinabolished staining in olfactory bulb sections. The aminoacid sequence of mGluR7 is most similar to other group IIImGluRs (mGluR4 and mGluR8); thus, one might expectnonspecific cross reactivity to be most apparent with theseproteins. However, the C-terminal sequence used for theimmunizing peptide is only 39% and 44% identical to thecorresponding regions of mGluR4 and mGluR8, the othergroup III receptors expressed in the olfactory bulb. Consis-tent with these sequence differences, the antiserum recog-nized a single band in extracts from dissected rat brainregions and did not recognize mGluR4 expressed in BHK

cells (see Results) or in mGluR8 in Western blots fromoocytes expressing functional mGluR8 receptors (unpub-lished observations). Because identified splice variants ofmGluRs, including those of mGluR7 (van den Pol, unpub-lished), are divergent in their carboxy-terminal sequences,an antibody raised to this part of the mGluR7 molecule isunlikely to recognize other potential mGluR7 splice vari-ants. Furthermore, antibody affinity purified from thisantiserum and well characterized in hippocampal neurons(Bradley et al., 1996) gave an identical pattern of stainingin our experiments. Thus, we conclude that, under theconditions of our experiments, the antiserum specificallyrecognized mGluR7.

The pattern of mGluR7 immunoreactivity in the olfac-tory bulb and piriform cortex closely parallels the distribu-tion of mGluR7 mRNA as demonstrated by in situ hybrid-ization (Kinzie et al., 1995; Ohishi et al., 1995a). Strongimmunoreactivity matched the high level of mRNA label-ing in mitral cells, whereas mGluR7 immunoreactivitywas difficult to detect in juxtaglomerular and granule cellsthat have low levels of mGluR7 mRNA. It is possible,however, that some of the synaptic labeling in glomerulirepresents processes of periglomerular cells, other juxtaglo-merular interneurons, or centrifugal input (Luskin andPrice, 1983). One unexpected finding in our experimentswas the more intense labeling of the granule cell layer inthe AOB relative to the MOB. This difference was far lessapparent by in situ hybridization (Kinzie et al., 1995;Ohishi et al., 1995a).

We used bulbectomy to determine whether immunostain-ing of the piriform cortex represented mitral cell axon

Fig. 6. Confocal immunofluorescence analysis of mGluR7 in theolfactory peduncle and piriform cortex. A: Immunohistochemistry ofthe anterior olfactory nucleus, lateral part (AONl), and rostral piri-form cortex at the level of the olfactory peduncle. The molecular layerof the anterior olfactory nucleus (1) is diffusely stained; whereas, inthe pyramidal layer (2), numerous immunoreactive cell bodies arepresent. There is staining of a few fibers traversing the lateralolfactory tract (LOT). In the rostral piriform cortex, there is intense,

diffuse staining of layer Ia. Layer Ib is not strongly stained. Somata inlayer II (II) are clearly immunoreactive. B: Immunofluorescence in thepiriform cortex more caudally also shows diffuse staining in layer Iawithout labeling of the LOT. Cell bodies in layers II and III areimmunoreactive, as are fibers coursing through layer III. C: Highermagnification of layers I and II show distinct fibers, which appear tooriginate in layer II, branching out in layer Ia. Scale bars 5 300 µm inA,B, 75 µm in C.


terminals that contained mGluR7. The marked reductionof staining in layer Ia of the piriform cortex after bulbec-tomy supports this idea. One caveat is that pyramidal cellsin the piriform cortex can undergo transsynaptic degenera-tion as soon as 24 hours after bulbectomy (Heimer andKalil, 1978; Friedman and Price, 1986; Westrum, 1988).However, MAP2 and mGluR1a, which are expressed inpiriform cortex pyramidal dendrites, were not affected atthe time point examined. In addition to layer Ia, distinctlabeled fibers in all levels of layer I could be visualized.These fibers appeared to originate from layers II and III ofthe piriform cortex and may be recurrent axons. Indeed,some electrophysiological evidence suggests that recur-rent axons in layer Ib originating from neurons in layers IIand III contain L-AP4-sensitive receptors (Collins andHowlett, 1988). However, anatomical evidence suggeststhat these recurrent axons do not enter layer Ia (Price,1973; Schwob and Price, 1984).

To confirm that our bulbectomy results reflect the pres-ence of mGluR7 in mitral cell axon terminals, we exam-ined mGluR7 immunoreactivity by electron microscopy inlayer Ia of the piriform cortex. This analysis showedabundant staining of large axon terminals with small,clear vesicles, suggestive of olfactory bulb terminals. Fromthese results and those from our bulbectomy experiments,we conclude that mGluR7 is present at axon terminals of

the LOT, where it functions as a presynaptic autoreceptor.However, we cannot discount the suggestion from ourstudies that mGluR7 may be in part dendritically targetedin piriform cortex pyramidal neurons.

Are mGluRs targeted to axons vs.somatodendritic compartments?

Early electrophysiological and immunohistochemicalstudies of mGluRs suggested that individual mGluR sub-types are localized either to somatodendritic or to axonalcompartments. For example, mGluR1a immunoreactivityis found in postsynaptic membranes and is targeted tosomatodendritic compartments in cultured hippocampalneurons (Martin et al., 1992; Craig et al., 1993; Shigemotoet al., 1994). Likewise, physiological activation of group Ireceptors increases excitability of dendrites. In contrast,the most obvious action of group II and III receptors is toinhibit transmitter release by a presynaptic mechanism(Schoepp and Conn, 1993; Saugstad et al., 1995). Thus, itwas reasonable to suppose that group I receptors weretargeted to dendrites, whereas group II/III receptors weretargeted to axons. However, there is increasing evidencethat this dichotomy is incorrect. For example, presynapticimmunoreactivity has been seen for both the group Ireceptors mGluR1a and mGluR5 (Fotuhi et al., 1993;Romano et al., 1995). In mitral cells of the olfactory bulb,

Fig. 7. Top: Immunohistochemistry with mGluR7, mGluR1a, mi-crotubule-associated protein 2 (MAP2), and calretinin antisera ofcoronal sections of the piriform cortex ipsilateral to the bulbectomy(indicated on the left by the depiction of a bulbectomized parasagittalsection of rat brain). Bottom: Immunohistochemistry for the antiserain the piriform cortex of the contralateral side from the same rat brain

section (indicated on the left by the depiction of an intact rat brainsection). Unlike mGluR1a and MAP2 staining, mGluR7 and calretininstaining is reduced in layer Ia in sections ipsilateral, but not contra-lateral, to bulbectomy. Graphics from Swanson (1992). Scale bar 5300 µm.


mGluR1a is found in dendrites, but this immunoreactivityis actually localized presynaptic to the mitral-granule celldendrodendritic synapses (van den Pol, 1995).

Of the group III receptors, mGluR6 is the only subtypethat appears to be exclusively postsynaptic (Nomura et al.,1994). However, this receptor is unusual, in that it may beexpressed in only a single cell type (retinal ON-bipolarcells) and has a specialized transduction pathway (Slaugh-

ter and Miller, 1985; Nawy and Jahr, 1990). The immuno-localization of the extraretinal group III receptors (mGluR4,mGluR7, and mGluR8) is just beginning to emerge. FormGluR4, prominent postsynaptic as well as presynapticlabeling has been demonstrated in hippocampal neurons(Bradley et al., 1996). However, current evidence suggeststhat mGluR7 and mGluR8 (Kinoshita et al., 1996) areprimarily axonal. For example, mGluR7 is localized pre-

Fig. 8. Electron microscopic immunohistochemistry for mGluR7 inlayer Ia of the rostral piriform cortex (OLF CRT). A and B arecounterstained with lead citrate and uranyl acetate. C and D receivedno lead staining, allowing for easier visualization of immunoreactiveaxons, but leaving nonimmunoreactive profiles with low contrast. A,B:Immunoreactive axons (Ax; open arrow) make asymmetrical synaptic

contact (thin solid arrow) with a dendrite (DEN, thick solid arrow). C:Three immunoreactive axons (Ax; open arrows) contact a centraldendrite (solid arrow). D: Single immunoreactive axon (Ax) makessynaptic contact with a dendrite (thick arrow). Postsynaptic density isindicated by the thin arrow. Scale bars 5 650 µm in A, 300 µm in B, 800µm in C, 700 µm in D.


dominantly to presynaptic axon terminals of asymmetricalsynapses in the hippocampus (Bradley et al., 1996). Simi-larly, mGluR7 is localized to axon terminals in laminae Iand II in the dorsal horn (Ohishi et al., 1995b). ThemGluR7 staining in mitral cell axon terminals in ourexperiments further supports a presynaptic localization.However, our results provide evidence that mGluR7 in theolfactory system is not exclusively axonal. Thus, it wouldappear that most, if not all, nonretinal group III mGluRsmay be expressed in somatodendritic as well as axonalcompartments.

Function of mGluRs in the olfactory system

The olfactory system may prove to be a unique system toexamine the roles of individual mGluRs in synaptic trans-mission. The synaptic connections of the bulb are special-ized to provide lateral inhibition that presumably medi-ates an essential component of olfactory discrimination.Therefore, the pathways that affect lateral inhibition atthe dendrodendritic synapses could powerfully modulateincoming sensory information. This has been beautifullydemonstrated by recent studies in the AOB. In thesestudies, activation of mGluR2 in granule cell dendrites atreciprocal synapses with mitral cell dendrites resulted ininhibition of g-aminobutyric acid (GABA) release. Thus,mGluR2 activation at these synapses relieves the excitedmitral cell from GABA-mediated feedback inhibition yetretains the feed-forward inhibition of surrounding mitralcells (Hayashi et al., 1993). This mechanism appears to beimportant for formation of olfactory memory during mat-ing in mice (Kaba and Nakanishi, 1995). The function ofthe other six mGluRs in the olfactory system is less clear.

The somatodendritic staining of mGluR7 suggests thatthis receptor may be postsynaptically localized, although,in the case of mitral cells, this may represent targeting topresynaptic release sites of dendrodendritic synapses (vanden Pol, 1995). Nonetheless, our electron microscopicresults suggest that, in some neurons, dendritic mGluR7 ispostsynaptically localized. It is not clear why a postsynap-tic mGluR7-mediated response has not been detectedpreviously. This discrepancy may result from postsynaptictargeting of mGluR7 in only a few neurons. Alternatively, apostsynaptic mGluR7 response might have been over-looked because of the low apparent agonist affinity of thereceptor (Okamoto et al., 1994; Saugstad et al., 1994) orbecause the transducing signal for mGluR7 action atpostsynaptic sites, such as a decrease in cyclic adenosinemonophosphate (cAMP), is not easily detected by physi-ological techniques.

Another question raised by our results is why receptorsthat are as similar as mGluR7 and mGluR8 are coex-pressed on mitral cell axons. Again, the approximate500-fold less apparent affinity of mGluR7 for agonistcompared with mGluR8 (Saugstad et al., 1997) may pro-vide an explanation. This difference suggests that mGluR7might be activated only near released quanta of gluta-mate. By contrast, mGluR8 could be activated at lowerglutamate concentrations, for example, by spill over fromadjacent synapses. Thus, mGluR7 and mGluR8 couldoperate in tandem to mediate homosynaptic and heterosy-naptic inhibition, respectively.

ACKNOWLEDGMENTS

We thank Stefania Risso-Bradley and P. Jeffrey Conn forproviding the mGluR7 antibody, Betty Haldeman andEileen Mulvihill for providing the mGluR1a antibody, EvaShannon for assistance with confocal microscopy, andStephen Gancher for assistance with olfactory bulbectomy.This work was supported by NIH grants MH10314 (J.M.K.),DC01783 (T.P.S.), and NS10174 (A.N.v.d.P.)

LITERATURE CITED

Abe, T., H. Sugihara, H. Nawa, R. Shigemoto, N. Mizuno, and S. Nakanishi(1992) Molecular characterization of a novel metabotropic glutamatereceptor mGluR5 coupled to inositol phosphate/Ca21 signal transduc-tion. J. Biol. Chem. 267:13361–13368.

Anson, J., and G.G. Collins (1987) Possible presynaptic actions of 2-amino-4-phosphonobutyrate in rat olfactory cortex. Br. J. Pharmacol. 91:753–761.

Baude, A., Z. Nusser, J.D. Roberts, E. Mulvihill, R.A. McIlhinney, and P.Somogyi (1993) The metabotropic glutamate receptor (mGluR1 alpha)is concentrated at perisynaptic membrane of neuronal subpopulationsas detected by immunogold reaction. Neuron 11:771–87.

Bradley, S.R., A.I. Levey, S.M. Hersch, and P.J. Conn (1996) Immunocyto-chemical localization of group III metabotropic glutamate receptors inthe hippocampus with subtype specific antibodies. J. Neurosci. 16:2044–2056.

Collins, G.G., and S.J. Howlett (1988) The pharmacology of excitatorytransmission in the rat olfactory cortex slice. Neuropharmacology27:697–705.

Craig, A.M., C.D. Blackstone, R.L. Huganir, and G. Banker (1993) Thedistribution of glutamate receptors in cultured rat hippocampal neu-rons: Postsynaptic clustering of AMPA-selective subunits. Neuron10:1055–1068.

Duvoisin, R.M., C. Zhang, and K. Ramonell (1995) A novel metabotropicglutamate receptor expressed in the retina and olfactory bulb. J.Neurosci. 15:3075–3083.

Fotuhi, M., A.H. Sharp, C.E. Glatt, P.M. Hwang, M. von-Krosigk, S.H.Snyder, and T.M. Dawson (1993) Differential localization of phosphoino-sitide-linked metabotropic glutamate receptor (mGluR1) and the inosi-tol 1,4,5-trisphosphate receptor in rat brain. J. Neurosci. 13:2001–2012.

Friedman, B., and J.L. Price (1986) Plasticity in the olfactory cortex:Age-dependent effects on deafferentation. J. Comp. Neurol. 246:1–19.

Hampson, D.R., E. Theriault, X.P. Huang, P. Kristensen, D.S. Pickering,J.E. Franck, and E.R. Mulvihill (1994) Characterization of two alterna-tively spliced forms of a metabotropic glutamate receptor in the centralnervous system of the rat. Neuroscience 60:325–336.

Hasselmo, M.E., and J.M. Bower (1991) Selective suppression of afferentbut not intrinsic fiber synaptic transmission by 2-amino-4-phosphonobu-tyric acid (AP4) in piriform cortex. Brain Res. 548:248–255.

Hayashi, Y., A. Momiyama, T. Takahashi, H. Ohishi, R. Ogawa-Meguro, R.Shigemoto, N. Mizuno, and S. Nakanishi (1993) Role of a metabotropicglutamate receptor in synaptic modulation in the accessory olfactorybulb. Nature 366:687–690.

Heimer, L., and R. Kalil (1978) Rapid transneuronal degeneration anddeath of cortical neurons following removal of the olfactory bulb in adultrats. J. Comp. Neurol. 178:559–609.

Houamed, K.M., J.L. Kuijper, T.L. Gilbert, B.A. Haldeman, P.J. O’Hara,E.R. Mulvihill, W. Almers, and F.S. Hagen (1991) Cloning, expression,and gene structure of a G protein-coupled glutamate receptor from ratbrain. Science 252:1318–1321.

Joly, C., J. Gomeza, I. Brabet, K. Curry, J. Bockaert, and J.-P. Pin (1995)Molecular, functional, and pharmacological characterization of themetabotropic glutamate receptor type 5 splice variants: Comparisonwith mGluR1. J. Neurosci. 15:3970–3981.

Kaba, H., and S. Nakanishi (1995) Synaptic mechanisms of olfactoryrecognition memory. Rev. Neurosci. 6:125–141.

Kinoshita, A., H. Ohishi, A. Neki, S. Nomura, R. Shigemoto, M. Takada, S.Nakanishi, and N. Mizuno (1996) Presynaptic localization of a metabo-tropic glutamate receptor, mGluR8, in the rhinencephalic areas—Alight and electron microscope study in the rat. Neurosci. Lett. 207:61–64.

Kinzie, J.M., J.A. Saugstad, G.L. Westbrook, and T.P. Segerson (1995)Distribution of metabotropic glutamate receptor 7 messenger RNA inthe developing and adult rat brain. Neuroscience 69:167–176.


Knopfel, T., R. Kuhn, and H. Allgeier (1995) Metabotropic glutamatereceptors: Novel targets for drug development. J. Med. Chem. 38:1417–1426.

Luskin, M.B., and J.L. Price (1983) The topographic organization ofassociational fibers of the olfactory system, including centrifugal fibersto the olfactory bulb. J. Comp. Neurol. 216:264–291.

Martin, L.J., C.D. Blackstone, R.L. Huganir, and D.L. Price (1992) Cellularlocalization of a metabotropic glutamate receptor in rat brain. Neuron9:259–270.

Minakami, R., F. Katsuki, T. Yamamoto, K. Nakamura, and H. Sugiyama(1994) Molecular cloning and the functional expression of two isoformsof human metabotropic glutamate receptor subtype 5. Biochem. Bio-phys. Res. Commun. 199:1136–1143.

Nawy, S., and C. Jahr (1990) Suppression by glutamate of cGMP-activatedconductance in retinal bipolar cells. Nature 346:269–271.

Nomura, A., R. Shigemoto, Y. Nakamura, N. Okamoto, N. Mizuno, and S.Nakanishi (1994) Developmentally regulated postsynaptic localizationof a metabotropic glutamate receptor in rat rod bipolar cells. Cell77:361–369.

Ohishi, H., R. Shigemoto, S. Nakanishi, and N. Mizuno (1993a) Distribu-tion of the messenger RNA for a metabotropic glutamate receptor,mGluR2, in the central nervous system of the rat. Neuroscience53:1009–1018.

Ohishi, H., R. Shigemoto, S. Nakanishi, and N. Mizuno (1993b) Distribu-tion of the mRNA for a metabotropic glutamate receptor (mGluR3) inthe rat brain: An in situ hybridization study. J. Comp. Neurol. 335:252–266.

Ohishi, H., C. Akazawa, R. Shigemoto, S. Nakanishi, and N. Mizuno(1995a) Distributions of the mRNAs for l-2-amino-4-phosphonobutyrate-sensitive metabotropic glutamate receptors, mGluR4 and mGluR7, inthe rat brain. J. Comp. Neurol. 360:555–570.

Ohishi, H., S. Nomura, Y. Ding, R. Shigemoto, E. Wada, A. Kinoshita, J. Li,A. Neki, S. Nakanishi, and N. Mizuno (1995b) Presynaptic localizationof a metabotropic glutamate receptor, mGluR7, in the primary afferentneurons: An immunohistochemical study in the rat. Neurosci. Lett.202:85–88.

Okamoto, N., S. Hori, C. Akazawa, Y. Hayashi, R. Shigemoto, N. Mizuno,and S. Nakanishi (1994) Molecular characterization of a new metabo-tropic glutamate receptor mGluR7 coupled to inhibitory cyclic AMPsignal transduction. J. Biol. Chem. 269:1231–1236.

Pin, J.-P., and J. Bockaert (1995) Get receptive to metabotropic glutamatereceptors. Curr. Opin. Neurobiol. 5:342–349.

Pin, J.-P., C. Waeber, L. Prezeau, J. Bockaert, and S.F. Heinemann (1992)Alternative splicing generates metabotropic glutamate receptors induc-ing different patterns of calcium release in Xenopus oocytes. Proc. Natl.Acad. Sci. USA 89:10331–10335.

Price, J.L. (1973) An autoradiographic study of complementary laminarpatterns of termination of afferent fibers to the olfactory cortex. J.Comp. Neurol. 150:87–108.

Romano, C., M.A. Sesma, C.T. McDonald, K. O’Malley, A.N. van den Pol,and J.W. Olney (1995) Distribution of metabotropic glutamate receptormGluR5 immunoreactivity in rat brain. J. Comp. Neurol. 355:455–469.

Saugstad, J.A., J.M. Kinzie, E.R. Mulvihill, T.P. Segerson, and G.L.Westbrook (1994) Cloning and expression of a new member of theL-2-amino-4-phosphonobutyric acid-sensitive class of metabotropic glu-tamate receptors. Mol. Pharmacol. 45:367–372, pp. 77–78.

Saugstad, J.A., T.P. Segerson, and G.L. Westbrook (1995) Modulation of ionchannels and synaptic transmission by metabotropic glutamate recep-tors. In H. Wheal and A. Thomson (eds): Excitatory Amino Acids andSynaptic Transmission. San Diego, CA: Academic Press.

Saugstad, J.A., J.M. Kinzie, M.M. Shinohara, T.P. Segerson, G.L. West-brook (1997) Cloning and expression of rat metabotropic glutamatereceptor 8 reveals a unique pharmacological profile. Mol. Pharmacol.51:119–125.

Schoepp, D.D., and P.J. Conn (1993) Metabotropic glutamate receptors inbrain function and pathology. Trends Pharmacol. Sci. 14:13–20.

Schwob, J.E., and J.L. Price (1984) The development of lamination ofafferent fibers to the olfactory cortex in rats, with additional observa-tions in the adult. J. Comp. Neurol. 223:203–222.

Shepherd, G.M., and C.A. Greer (1990) Olfactory bulb. In G.M. Shepherd(ed): The Synaptic Organization of the Brain. New York: OxfordUniversity Press, pp. 133–169.

Shigemoto, R., S. Nakanishi, and N. Mizuno (1992) Distribution of themRNA for a metabotropic glutamate receptor (mGluR1) in the centralnervous system: An in situ hybridization study in adult and developingrat. J. Comp. Neurol. 322:121–135.

Shigemoto, R., T. Abe, S. Nomura, S. Nakanishi, and T. Hirano (1994)Antibodies inactivating mGluR1 metabotropic glutamate receptor blocklong-term depression in cultured Purkinje cells. Neuron 12:1245–1255.

Slaughter, M.M., and R.F. Miller (1985) Characterization of an extendedglutamate receptor of the on bipolar neuron in the vertebrate retina. J.Neurosci. 5:224–233.

Swanson, L.W. (1992) Brain Maps: Structure Of The Rat Brain. Amster-dam: Elsevier Science Publishers.

Thomsen, C., P. Kristensen, E. Mulvihill, B. Haldeman, and P.D. Suzdak(1992) L-2-amino-4-phosphonobutyrate (L-AP4) is an agonist at thetype IV metabotropic glutamate receptor which is negatively coupled toadenylate cyclase. [Published erratum appears in Eur. J. Pharmacol.15;244(2):187.] Eur. J. Pharmacol. 227:361–362.

Trombley, P.Q., and G.M. Shepherd (1993) Synaptic transmission andmodulation in the olfactory bulb. Curr. Opin. Neurobiol. 3:540–547.

Trombley, P.Q., and G.L. Westbrook (1992) L-AP4 inhibits calcium currentsand synaptic transmission via a G-protein-coupled glutamate receptor.J. Neurosci. 12:2043–2050.

van den Pol, A.N. (1995) Presynaptic metabotropic glutamate receptors inadult and developing neurons: Autoexcitation in the olfactory bulb. J.Comp. Neurol. 359:253–271.

Westbrook, G.L. (1994) Glutamate receptor update. Curr. Opin. Neurobiol.4:337–346.

Westrum, L.E. (1988) Electron microscopy of plasticity in rat olfactorycortex. Brain Res. 470:29–43.

Wouterlood, F.G., and W. Hrtig (1995) Calretinin-immunoreactivity inmitral cells of the olfactory bulb. Brain Res. 682:93–100.


Documents

Immunolocalization of metabotropic glutamate receptor 7 in the rat olfactory bulb