Glutamate receptor subunits in neuronal populations of the gerbil lateral superior olive1

Hearing Research 137 (1999) 77^90 www.elsevier.com/locate/heares

Glutamate receptor subunits in neuronal populations of the gerbil lateral superior olive1 Ilsa R. Schwartz *, Patricia R. Eager Department of Surgery/Otolaryngology, Yale University School of Medicine, P.O. Box 20841, New Haven, CT 06520-8041, USA Received 25 March 1999; received in revised form 20 July 1999; accepted 30 July 1999

Abstract The distribution of AMPA-preferring ionotropic glutamate receptors (GluR) within the gerbil lateral superior olive (LSO) was investigated immunocytochemically using antibodies to GluR1, 2, 2/3 and 4. Light microscopy showed GluR1 antibody preferentially labeling a population of small neurons located in the dorsal hilus and a population mainly at or near the margins of the LSO. GluR4 antibody strongly stained most large LSO neuronal somata and proximal dendrites including all principal cells. GluR2/ 3 antibody showed very modest staining and appeared in most cell types. GluR2 showed less intense neuronal staining than GluR2/3 and was observed as a punctate accumulation at the surface of some neuronal profiles. GluR1, 2, 2/3 and 4 immunoreactivity was found along dendrites of most large LSO neurons and in their somata. Postsynaptic specializations positive for GluR2 were rare on LSO somata compared to the high frequency of GluR4 and 1 specializations. Double labeling studies showed that different portions of the distal dendrites showed a preponderance of GluR1 or GluR4 subunits. Electron microscopic observations confirm similarities in the localization of immunoreactivity for the antibodies tested in the cytoplasm of somata and dendrites, but reveal differences at the plasmalemma, at synaptic appositions and appositions with glial processes. Receptor composition varied with cell type and location on cells. ß 1999 Elsevier Science B.V. All rights reserved. Key words: Glutamate receptor subunits; Auditory system; Superior olive; AMPA-preferring; Immunochemistry; Light microscopy; Electron microscopy

1. Introduction Glutamate (Glu) is a major transmitter of the central nervous system, and its role in the auditory system is increasingly recognized. Glu mediates fast excitatory neurotransmission via ligand-gated cationic channels in ionotropic receptors and slower G-protein-coupled metabotropic receptors in the neuronal membrane. Several types of studies implicate its involvement at a number of auditory brainstem synapses. High a¤nity amino acid uptake demonstrated Glu uptake in synaptic terminals in the superior olivary complex (SOC) (Schwartz, 1984). Recent biochemical and electrophys-

* Corresponding author. Tel.: +1 (203) 785-6329; Fax: +1 (203) 737-2245; E-mail: [email protected] 1 Preliminary reports of some of these ¢ndings have appeared in abstract form (Schwartz, 1994; Schwartz and Eager, 1995, 1996a,b).

iological studies have shown that Glu plays an important role in the SOC (Caspary and Faingold, 1989; Caspary and Finlayson, 1991 ; Caspary et al., 1985; Wu and Kelly, 1992 ; Kandler and Friauf, 1995; Suneja et al., 1995), as well as most brain structures. This report focuses on the LSO of the gerbil, a major nucleus in the auditory pathway which integrates information from both ears and includes the cell bodies of the olivocochlear e¡erents. Since the action of Glu is dependent on the receptors with which it interacts, studies of auditory system function need to be concerned with the distribution of the di¡erent Glu receptor channels in di¡erent classes of auditory neurons as well as their location within each cell type. Ligand-gated glutamate channels are composed of various combinations of subunits: the AMPA-preferring channels of subunits GluR1^4 ; the kainatepreferring of subunits GluR5^7 and KA1^2; and

0378-5955 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 5 9 5 5 ( 9 9 ) 0 0 1 4 0 - 9

HEARES 3314 25-10-99

78

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

the NMDA-preferring of subunits NMDAR1 and NMDAR2A^D. The structure of the channels was ¢rst proposed to be pentameric, by analogy with acetylcholine receptors of muscle (Hollmann and Heinemann, 1994). More recent studies suggest that the channels are tetrameric (Wu et al., 1996; Mano and Teichberg, 1998). The distribution of subunits in di¡erent cell types throughout the brain has been studied with: (1) in situ hybridization to detect the mRNA encoding these proteins, (2) Northern blots of tissue homogenates of various brain subregions and (3) immunocytochemical methods. To date, there have been relatively few studies concerned with auditory areas. In situ hybridization studies demonstrated the presence of the di¡erent combinations of mRNA for GluR1^4 in di¡erent rat cochlear nucleus (CN) and superior olivary complex (SOC) neurons (Hunter and Wenthold, 1993 ; Hunter et al., 1992). DCN cartwheel and stellate cells and neurons in the ventral and lateral periolivary nuclei contained mRNA for GluR1^4. There was a general absence of GluR1 in a number of CN cell classes. Cells with little GluR1 mRNA included granule cells, AVCN globular, round, and spherical cells, DCN fusiform cells, (Hunter et al., 1993), and neurons of the medial nucleus of the trapezoid body (MNTB), medial and lateral superior olives (MSO, LSO). A subpopulation of VCN round cells, DCN fusiform cells and MNTB neurons showed much less mRNA for GluR3. In the developing rat brainstem (postnatal days 1, 3 and 5) GluR1 mRNA was expressed at high levels compared to the adult (50^65 days) in VCN, MNTB and LSO (Hunter and Wenthold, 1993). Northern blot data on the CN showed only a single major band recognized by each of the GluR1, 2/3 and 4 antibodies (Wenthold et al., 1992). Immunocytochemical studies have been largely directed at the whole brain or CN (Petralia and Wenthold, 1992a,b; Petralia et al., 1994, 1996; Bilak et al., 1996). This study addressed questions unresolved by in situ studies or blot studies of tissue homogenates. Are the peptides of the receptor subunit actually present in particular cells ? Where in the cell are they located? Here we report on the distribution of immunoreactivity to four antibodies generated against the C-terminal sequence of the ionotropic AMPA-preferring glutamate receptor subunits 1^4. 2. Materials and methods 2.1. Antibodies The antibodies (Abs) for GluR1 and GluR4 are to sequences unique to those subunits. The immunogen was the synthetic peptide sequence conjugated to BSA

with glutaraldehyde. The GluR2/3 Ab recognizes a sequence common to both GluR2 and GluR3. The GluR2 Ab recognizes a 16 amino acid sequence (residues 827^842) near the C-terminus (Petralia et al., 1997). The Ab to GluR2 and 3 also recognizes a splice variant of GluR4c (Gallo et al., 1992). Initially the polyclonal GluR1, 2, 2/3 and 4 antibodies were generously supplied by Dr. Robert Wenthold, NIH/NIDCD. They later became commercially available from Chemicon (Temecula, CA). Results with the GluR1^4 antibodies from both sources have been comparable and we have not distinguished between them in reporting the results. 2.2. Light microscopy Young adult Mongolian gerbils (n = 35) were anesthetized with sodium pentobarbital and perfused through the heart with cold saline nitrite, followed by cold mixed aldehyde ¢xative (4% paraformaldehyde (P) and 0.1% glutaraldehyde (G), 4% P and 0.5% G or 1% P and 1.25% G). The perfused animals were stored on ice for 30^60 min before removal of the brains and subsequent ¢xation in the cold mixed aldehyde ¢xative for 2 h or overnight. Brains were washed in cold 0.12 M phosphate bu¡er (pH 7.4) and stored in bu¡er at 4³C until they were cryostat-sectioned at 30 Wm. Time of storage in bu¡er had little e¡ect upon the preservation of antigenicity of GluR1, 2/3 and 4 between 2.5 weeks and 3 and 6 months of storage in the cold. Sections were collected in 0.12 M phosphate bu¡er, sorted, treated with blocking serum (5^10% goat) and then incubated for 24^72 h in the cold after at least 4 h on a shaker at room temperature (GluR1, 2 Wg/ml; GluR2/3, 0.2 Wg/ml; GluR2, 0.5^1 Wg/ml; GluR4, 1 Wg/ml). Alternatively, incubations were for 12^20 h on a shaker at room temperature. Sections were then washed, incubated with the biotinylated secondary Ab (Vecta Elite, Vector Laboratories, Burlingame, CA), followed by ABC and were visualized with diaminobenzidine (DAB) as the capture agent (0.0125% DAB with 0.0005% H2 O2 ), with or without intensi¢cation with nickel chloride (NiCl) (Adams, 1981). In double label experiments, the ¢rst Ab was visualized with the Vecta Elite ABC procedure with NiCl intensi¢cation producing a blue reaction product. The sections were then washed and incubated with the second Ab and visualized with HRP producing a brown reaction product. Sections were mounted on slides coated with 1% Elmer's glue, covered with Permount and coverslipped. Mounted sections were examined and photographed with a Nikon Biophot microscope. Alternatively, 2 Wm sections of plastic embedded unosmicated brain slices were cut with glass knives, heatmounted on slides and deplasticized with NaOH/absolute alcohol. Slides were incubated for 72 h in the cold

HEARES 3314 25-10-99

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

with double the above concentrations of antibody. After washing, biotinylated Ab and ABC incubations, visualization was with DAB. 2.3. Electron microscopy Tissue from eight animals was ¢xed with 4% P, 0.5% G for 2 h, bu¡er washed and stored in bu¡er until 50 Wm sections were cut with the vibratome. Sections were incubated with antibodies as above, followed by DAB-H2 O2 without NiCl. Sections were then exposed to 1% osmium tetroxide for 60 min (Schwartz, 1982), dehydrated and embedded in Polybed 812/Araldite on slides (Yu and Schwartz, 1989). After examination and documentation, areas of interest were cut out and remounted on blanks for thin sectioning and examination and photography with a Philips 300 electron microscope. In some instances, 1^2 Wm sections of these blocks were mounted on slides and photographed. 2.4. Controls Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of gerbil CN, performed in Dr. Wenthold's laboratory (following the procedure outlined in Tachibana et al., 1994), showed only a single major band reacting to each of the antibodies (Schwartz, unpublished) as Wenthold et al. (1992) have shown for rat tissue. The speci¢city of these antibodies was established by analysis of membranes of cells transfected with cDNAs of the four subunits (Fig. 2, Wenthold et al., 1992). Because of the similarity of gerbil CN to rat tissue in SDS-PAGE immunoblots, and because there was no indication of non-speci¢c binding in rat (Petralia and Wenthold, 1992a), only a no substrate and a serum control were run in each experiment and the antibodies incubated with adjacent sections run at the same time served as controls for each other. 2.5. Animals All animal protocols were described in NIH Grant DC00132 `Dynamic Aspects of Auditory Synaptic Terminals', approved by the Yale Animal Care and Use Committee and carried out in compliance with the regulations for the humane use of animals in research. 3. Results Observations of cryostat-sectioned tissue incubated in single and double labeling combinations for light microscopy and vibratome sections incubated for electron microscopy have demonstrated the presence of prefer-

79

ential staining of di¡erent groups of cells in the LSO with GluR1 and GluR4 antibodies. Similarities and differences have also been demonstrated in the patterns of localization of immunoreactivity within neurons and glia with all of the GluR antibodies tested. The correlation of localization patterns and cell types is considered in Section 4. While this report focuses on the LSO, it should be noted that immunoreactivity (IR) was found in neurons in all of the auditory brainstem nuclei. In the SOC, di¡erences in staining patterns between nuclei were much less apparent than in the CN (Schwartz, 1994). Cells in all nuclei stained and di¡erences in the relative intensity of staining among the nuclei were not great. Cells in the ventral and lateral nuclei of the trapezoid body (VNTB and LNTB) were the most intensely stained. 3.1. Light microscopic observations 3.1.1. Single labeling experiments GluR1 staining was present in both large and small cells of the LSO. Since most large cells were stained they must include principal, and probably also multipolar, neurons. Small neurons also were stained. On the basis of location some stained small cells were identi¢ed as marginal, but small and type 5 cells (Helfert and Schwartz, 1987) could not be distinguished in the body of the LSO. While most neurons and dendrites showed a light to moderate level of GluR1 staining, GluR1 preferentially strongly stained small neurons and long stretches of their dendrites in the LSO body, dorsal hilus and LSO periphery (Fig. 1A). A strongly stained small neuron in the body of the LSO is illustrated in Fig. 2A, and several small neurons in the dorsal hilus in Fig. 4. A labeled distal dendrite is seen in Fig. 3A. In addition, the nuclei of some larger neurons were moderately to strongly stained, as can be seen unambiguously in a 2 Wm plastic section (Fig. 3A). Unstained nuclei are illustrated in Fig. 2A. GluR2. The least intense somatic staining was observed with the GluR2 antibody in SOC nuclei (Fig. 1B). The low level of GluR2 staining of LSO neurons is illustrated in Fig. 2B, while the ¢ne granular nature of the staining is best appreciated in a 2 Wm plastic section (Fig. 3B). Punctate accumulations of immunoreactivity were present at the surface of some neuronal pro¢les (Fig. 3B). Also, with GluR2, glial cell staining (Fig. 2B) was heavier than neuronal staining in the LSO. A positive control for this antibody was the diffuse somatic staining of cerebellar Purkinje cells observed in the same sections. GluR2/3 stained large somata and dendrites lightly to moderately, but more heavily than GluR2 (Figs. 1C and 2C). The large somata were not as intensely stained

HEARES 3314 25-10-99

80

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

Fig. 1. Adjacent cryostat sections from a single animal demonstrating glutamate receptor immunoreactivity in the LSO to antibodies to GluR1 (A), GluR2 (B), GluR2/3 (C) and GluR4 (D). A: With GluR1 many neurons are lightly stained, but the most darkly stained neurons are marginal cells (arrows) and small cells. The block around the dorsal hilus indicates the region illustrated in a di¡erent section in Fig. 4. M = LSO medial limb, L = lateral limb. B: With GluR2 principal cells (arrows) are lightly stained. C: With GluR2/3 principal cells (arrows) are lightly to moderately stained. D: With GluR4 principal cells (small arrows) are darkly stained. Small darkly stained glial somata (large arrows) are present throughout the LSO. Bar = 100 Wm.

by GluR2/3 as the small neurons were with GluR1. No small cell staining was apparent and, at the light microscopic level, glial staining was not obvious. With GluR2/3, LSO neurons were less stained compared to cells in the superior paraolivary nucleus (SPN), and the VNTB and LNTB. GluR4 stained the most and largest cell bodies and long stretches of their dendrites, and stained them the most intensely (Figs. 1D and 2D). These stained fusiform shaped cells were uniformly distributed through-

out the LSO and their dendrites tended to be oriented orthogonally to a line following the curvature of the nucleus. In some animals many more neurons were labeled in the lateral limb. Principal cells appear to be the major cell type stained by GluR4. Since virtually all large somata were stained, we infer that multiplanar and type 5 neurons are also stained. The neuropil was generally clear, although the large number and extent of stained dendrites gave the body of the LSO a darker appearance than the periolivary region.

HEARES 3314 25-10-99

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

81

Fig. 2. Higher magni¢cation views of LSO neurons from cryostat sections. A: With GluR1 a small neuron (arrow) is intensely stained. Large neurons are lightly to moderately stained (*) and do not show nuclear staining. B: With GluR2 a few glial cells (arrows) are moderately stained. Adjacent principal cells are largely unstained. C: With GluR2/3 large neurons are moderately stained compared to GluR2. D: With GluR4 large neurons show the most intense di¡use stain and a few even more intense dark patches (arrows). Bar = 10 Wm.

In addition to the strong di¡use staining of somata and dendrites of LSO principal neurons, GluR4 also produced small intense patches of stain on some cell bodies. That these patches were within the somata and not at its surface are illustrated by their presence in 2 Wm plastic sections (Fig. 3C) and in electron micrographs (see below). These patches were not seen in tissue reacted for GluR1, 2 or 2/3. However, with GluR1, small patches were observed on neurons in the MSO, MNTB and SPN.

3.1.2. Double labeling experiments When GluR1 was the ¢rst (blue) antibody, and GluR2/3 the second (brown), a population of small cells in the LSO were darkly stained in blue (GluR1), including both their somata and all along their dendrites. Principal cells were lightly stained with blue (GluR1) or brown (GluR2/3) as well as a mixture of blue and brown. When GluR4 was the ¢rst antibody and GluR1 the second, portions of LSO principal cells and their den-

HEARES 3314 25-10-99

82

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90 Fig. 3. Illustrates two Wm plastic sections of tissue stained respectively with antibodies against GluR1, GluR2 and GluR4. A: A concentration of GluR1-IR is seen along the distal dendrite (arrow). Nuclear staining is present in many principal cells (arrowhead), although some nuclei are unstained (see Fig. 2A). B: GluR2-IR shows a ¢ne granular pattern over principal cells. C: GluR4-IR shows a more concentrated granular distribution over both somata and dendrites and a few intense patches of immunoreactivity can be seen (arrows). Bar = 10 Wm. 6

drites were stained either blue (GluR4) or brown (GluR1), or a mixture of blue and brown. Proximal dendrites of large neurons were usually blue, while exclusively brown patches occurred more frequently along the distal dendrites, although blue patches were also present distally. Some small neurons had very dark patches of brown stain. The extensive colocalization of GluR2/3 and GluR4 was also demonstrated. When GluR2/3 was the ¢rst antibody and GluR4 the second, principal cells and their dendrites were lightly stained blue (GluR2/3). No stained population of small neurons was seen. GluR2/3 seems to be present in all large LSO neurons. These observations con¢rm the presence of multiple subunits in principal cells and their dendrites, and the preponderance and possibly exclusive presence of GluR1 in a population of small neurons and their dendrites.

Fig. 4. Illustrates at a higher magni¢cation a group of small neurons (arrows) in the LSO dorsal hilus intensely stained with antiGluR1. The area shown is equivalent to that marked in Fig. 1A. M = medial limb, L = lateral limb. Bar = 25 Wm.

HEARES 3314 25-10-99

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

83

Fig. 5. Electron micrographs demonstrate the postsynaptic localization of antibodies to (A) GluR1, (B) GluR2, (C) GluR2/3 and (D) GluR4. Localization of immunoreactivity on ER and mitochondria close to the stained postsynaptic specialization is seen in A, B and C (arrows). Localization in a presynaptic terminal (p) is seen in D. Arrowheads mark immunopositive postsynaptic specializations. Bar = 1 Wm.

3.2. Ultrastructural observations Ultrastructurally, localization was observed postsynaptically immediately below the synaptic junction of a subpopulation of terminals, although the number of terminals so labeled and the length and density of the labeling varied from antibody to antibody and from nucleus to nucleus with the same antibody. In LSO,

terminals apposed to labeled postsynaptic specializations di¡ered, but many contained round or pleomorphic vesicles associated with a presumed excitatory, probably glutamatergic neurotransmitter (Fig. 5A^D). Vesicle morphology is not well preserved in tissue which is not uranyl block stained making it hard to di¡erentiate round and pleiomorphic vesicles. In addition to localization at postsynaptic membrane

HEARES 3314 25-10-99

84

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

Fig. 6. Illustrates the intense localization of staining in small glial processes (arrows) with antibodies to (A) GluR1, (B) GluR2, (C) GluR2/3, and (D) GluR4. Bars = 1 Wm.

sites, each antibody had characteristic patterns of distribution within the somatic and dendritic cytoplasm varying from nucleus to nucleus. With GluR4, in LSO principal neurons and many large dendrites, immunoreactivity was distributed: (1) at the somatic and dendritic surface beneath synaptic specializations at some terminals (Fig. 5D); (2) throughout the somatic and

dendritic cytoplasm around most organelles (Fig. 7B,C); and (3) in a concentrated way over various clumps of endoplasmic reticulum (ER) (Fig. 8). A similar distribution beneath the somata and dendritic surfaces and throughout the somatic and dendritic cytoplasm was seen in small neurons and dendrites after incubation with the GluR1 Ab (Fig. 5A). With

HEARES 3314 25-10-99

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

85

Fig. 7. A: A large neuron with much of its surface apposed to synaptic terminals show several areas of immunoreactivity to GluR2/3 within its cytoplasm (*), but few stained postsynaptic specializations (arrows). B: Shows the association of immunoreactivity and mitochondria (arrows) in a dendrite from tissue stained with A/GluR2/3. C: Shows the association of GluR4-IR and mitochondria and ER (arrows). Bars = 1 Wm.

HEARES 3314 25-10-99

86

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

croscopic level and are illustrated in a cell from the medial limb of the LSO (Fig. 8). 4. Discussion

Fig. 8. This electron micrograph shows the intense concentration of GluR4 immunoreactivity almost obscuring a stack of ER or Golgi lamellae, which corresponds to a dark patch seen at the LM level. Several smaller areas of immunoreactivity are distributed throughout the somatic cytoplasm, often in association with mitochondria and ER. Bar = 1 Wm.

GluR2/3, staining was sparse but, as with the other antibodies, cytoplasmic immunoreactivity was found associated with clumps of ER or near Golgi bodies or mitochondria (Figs. 5C and 7A,B). GluR2 immunoreactivity was very sparse at postsynaptic specializations on LSO neuronal somata, where it was found associated with mitochondria (Figs. 5B and 6B) and somewhat more frequent at postsynaptic specializations on dendrites. Comparable presynaptic localization of immunoreactivity (e.g., Fig. 5D) with all the antibodies was previously reported (Schwartz and Keh, 1997). Staining of glial processes was very prominent with GluR2 (Fig. 6B), but most pronounced with GluR4 (Fig. 6D). Some glial processes also showed staining with GluR1 (Fig. 6A) and GluR2/3 (Fig. 6C). The ¢ne nature of the glial processes stained accounts for the absence of detectability at the light microscopic level. At the electron microscopic level, heavy patches of GluR4 label were found over stacks of ER within the somatic cytoplasm rather than at synaptic sites. These heavy patches corresponded to the small intense patches seen over many LSO neurons at the light mi-

This study found that there are di¡erent and selective patterns of distribution of the GluR subunits on the well-characterized cell classes of the gerbil LSO (principal, multiplanar, marginal, small and type 5 cells) (Helfert and Schwartz, 1987). What is known about the distributions of various inputs to discrete locations on di¡erent LSO cell types allows us to associate receptors of di¡erent composition with speci¢c inputs. Principal cells, which show immunoreactivity for GluR1, 2, 2/3 and 4, compose approximately 75% of the total gerbil LSO neuronal population (Helfert and Schwartz, 1987). Since input to principal cells from the spherical cells of the anterior ventral cochlear nucleus, in the form of round vesicle-containing (R), presumably excitatory, synaptic terminals, is preferentially distributed to the distal dendrites (Cant and Casseday, 1986; Zook and DiCaprio, 1988; Helfert et al., 1992), our ¢ndings suggest that GluR1 subunits may be associated with the synapses of the spherical cell input. By contrast, GluR2/3-IR was generally restricted to the soma and GluR4-IR, the most strongly labeling, was preferentially distributed on the soma and proximal dendrites of principal cells. Since both GluR2/3 and GluR2 stain principal cells we cannot unequivocally identify the presence of GluR3. MNTB principal cell input to LSO principal cells in the form of £at vesicle (F) terminals immunoreactive for both glutamate and glycine are preferentially distributed on the soma and proximal dendrites (Helfert et al., 1992). Our material was not uranyl block stained or stained in the thin sections to maximize detection of GluR immunoreactivity. Thus it did not allow clear di¡erentiation of vesicle morphology. However, our ¢ndings of positional di¡erences suggest that while GluR2/3 and GluR4 may both be present in the specializations post-synaptic to F terminals from MNTB, GluR2/3 may also be found at another input more selectively distributed on the soma. It is not clear where on the principal cells the input from the presumed multipolar/stellate neurons of the contralateral rat PVCN project (Friauf and Ostwald, 1988), nor is it clear where the excitatory input from the VNTB (Warr and Spangler, 1989 ; Spangler et al., 1985) or the interstitial nucleus of the auditory nerve may be distributed. Results from double labeling studies indicating preferential localization of GluR2/3 in the soma, GluR4 in the soma and proximal dendrites and GluR1 more distally on dendrites could indicate a ¢nding similar to that on hippocampal pyramidal neurons of a greater

HEARES 3314 25-10-99

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

density of calcium permeable AMPA receptors on dendrites compared to their cell bodies (Lerma et al., 1994). Multiplanar neurons make up roughly 8% of the gerbil LSO neuronal population (Helfert and Schwartz, 1987). Principal and multiplanar neurons share similar cytoplasmic features, and greater than 65% of their perikaryal surface is in contact with synaptic terminals. Since they cannot be distinguished from principal cells in single transverse sections the arguments about GluR distribution on principal cells must be applied equally to them. Alternatively, multiplanar neurons could be candidates for the small number of LSO neurons which are not GluR4-IR. However, on the basis of location, shape and percentage of cells stained it appears that multiplanar neurons are, like principal cells, heavily immunoreactive for GluR4, and may also contain GluR1 and GluR2 or GluR3. Class 5 neurons share the same light microscopic features as principal neurons and can be identi¢ed electron microscopically based only on the percentage of somal surface occupied by synaptic terminals (about 31%) (Helfert and Schwartz, 1987). Their neuronal somata receive a similar number of axosomatic synaptic contacts as marginal neurons but are found well within the matrix of the LSO, aligned parallel to principal neurons. What percentage of the cells identi¢ed at the light microscopic level as principal cells are actually class 5 neurons is not known. They may correspond to the small number of LSO neurons which are not GluR4IR. They might also correspond to the subgroup of principal cells identi¢ed in rat by Rietzel and Friauf (1998) as banana-like cells which have a more elongated dendritic arborization, rarely show beading of the dendrites, showed no immunostaining for calbindin, and were identi¢ed only in the lateral limb. The criteria for identi¢cation of banana-like cells in the rat LSO (Rietzel and Friauf, 1998) rely on characteristics of the dendritic trees which were not detectable in our electron microscopic study. We have observed a similarity of staining pattern between GluR1-IR cells and those that are strongly immunopositive for another calcium binding protein, calreticulin (Korada and Schwartz, 1997). Lateral olivocochlear (LOC) neurons, identi¢ed by retrograde transport of D-aspartic acid (D-Asp), are found in or near the gerbil LSO. They are generally small in size, primarily fusiform in shape, and show fewer synaptic contacts than other LSO cells (Ryan et al., 1987; Helfert et al., 1988). All LOC neurons within the LSO that projected to the injected cochlea were labeled by D-Asp. LOC neurons include marginal, small and class 5 neurons (Helfert et al., 1988). Small neurons compose approximately 11% of the gerbil LSO neurons, have the lowest percentage of their somal surface contacted by synaptic terminals (approximately 8%), and

87

are found mostly in the middle/medial portions of the LSO (Helfert and Schwartz, 1987). Marginal neurons appear similar to principal neurons at the light microscopic level except that they are found along the contours of the LSO, oriented orthogonally to principal neurons. In the gerbil, they compose approximately 6% of the LSO population. On average, 28% of the somal surface of marginal neurons is in contact with synaptic terminals (Helfert and Schwartz, 1987). The presence in marginal and small cells of a high percentage of homomeric GluR1 receptors, which would be permeable to calcium, is suggested by the preferential localization of GluR1-IR in these cells and a marked absence of GluR2-IR and GluR2/3-IR. The presence of large numbers of calcium-£uxing channels is consistent with special properties of these cells for handling calcium. LOC neurons are the only LSO population to exhibit immunoreactivity for calcitonin gene-related peptide (CGRP) (Adams, 1986; Lu et al., 1987 ; Schweitzer et al., 1985 ; Simmons and Raji-Kubba, 1993; Spangler et al., 1987). A few CNS cell types have been reported to contain only GluR1 (Petralia and Wenthold, 1992a: rat brain ; Furuyama et al., 1993 : rat spinal cord ; Martin et al., 1993b : rat and monkey basal forebrain; Tachibana et al., 1994 : rat spinal cord). Thus, on the basis of the light microscopic criteria of number, location and shape it appears that marginal and small neurons are strongly, though not exclusively, immunoreactive for GluR1, and the majority of their AMPA channels are probably homomeric for GluR1. 4.1. Glia The localization of GluR subunits has previously been reported in a few glial populations in other brain regions (Joelsen and Schwartz, 1998 ; Luque and Richards, 1995 ; Burnashev et al., 1992 ; Muller et al., 1992 ; Martin et al., 1993a ; Lopez et al., 1994 ; Petralia et al., 1996). The location of GluR4-IR in glial processes surrounding synaptic terminals and at the neuronal surface apposed to glial processes suggests the possibility of their involvement in a tonic glial regulation of the activity of the synapse, possibly through regulation of the feedback of glutamine, or of the responsiveness of the postsynaptic cell. Immunoreactivity for GluR1, GluR2, and GluR2/3 was also found in LSO glial processes, although to a lesser extent than GluR4. There is no di¡erence in the immunoreactivity for GluR1, 2/3 and 4 antibodies present presynaptically within synaptic terminals of the LSO (Schwartz and Keh, 1997). 4.2. Implications of staining patterns Immunoprecipitation studies by Wenthold et al. (1996) have shown that the antibodies used in this study

HEARES 3314 25-10-99

88

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

can recognize their respective subunits both in a heteromeric receptor and in an unassembled form. While staining in the cytoplasm may re£ect the presence of subunits and/or assembled receptors, it is likely that staining at the membrane surface re£ects mainly assembled receptors. Electron microscopic observations of the same areas suggest that moderate to intense staining of somata and dendrites at the light microscopic level re£ects deposition of reaction product throughout the cytoplasm, or widely distributed around organelles dispersed through the cytoplasm of a neuronal somata or dendrite, rather than antibody binding restricted to a thin layer at the cell membrane. When staining is extensive but largely restricted to the membranes it may be perceived as a di¡use neuropil stain. Strong staining associated with ER, such as the intensely stained somatic patches seen with GluR4 antibody, suggests that we are seeing a production site of the subunit. Several proteins of the ER have been shown to function as molecular chaperones and in£uence the folding and assembly of complex proteins (Hurtley and Helenius, 1989; Hammond and Helenius, 1995 ; Frydman et al., 1994). Two such proteins, immunoglobulin-binding protein and calnexin, have been shown to associate with the nicotinic acetylcholine receptor (Forsayeth et al., 1992; Gelman et al., 1995). Thus, the ER may be a production site of the receptor complex as well as of the subunit protein. Staining throughout the cytoplasm suggests the presence of a cytoplasmic pool of the subunit prior to its assembly into a receptor, although it could also re£ect the presence of the heteromeric receptor. It is likely that immunoreactivity found at the cell surface where the receptors are functioning re£ects the presence of the heteromeric receptor. It is unclear why some nuclei in large LSO neurons showed strong immunoreactivity for the GluR1 antibody. It appears to be speci¢c to the subunit and not a general artifact of the staining protocols. We have observed nuclear labeling with another glutamate receptor subunit, NMDAR2B (Joelsen and Schwartz, 1998). According to Lerea (1997), ``Immediate early genes (IEGs) have been implicated in the conversion of short-term stimuli to long-term changes in cellular phenotype by regulation of gene expression. Many of the long-term consequences of glutamate receptor activation correlate with increases in speci¢c IEGs; the intracellular signalling pathways coupling activation of receptors at the cell surface with induction of IEGs in the nucleus are incompletely understood.'' It is possible that the GluR subunits which produce nuclear staining indicate their involvement in the intracellular signalling pathway.

Because of di¡erences between the GluR4 and GluR2/3 patterns (see also Schwartz and Keh, 1997), it seems likely that most of the GluR2/3 pattern does not re£ect staining of the GluR4c subunit. The di¡erential distribution of GluR1^4-IR on di¡erent LSO neuronal classes, and in ¢ne glial processes around synaptic terminals on LSO neurons, are re£ections of true di¡erences in the localization and amounts of the AMPA subunits present. 4.3. Correlation of stain intensity and subunit protein concentration The ¢nding of di¡erences in the relative intensity of staining for each antibody in di¡erent cell classes is valid. The concentrations of antibodies used in this study were chosen for their ability to provide a comparable range of staining intensity. However, intensity of staining has not been calibrated with absolute protein concentrations in individual cell types. Thus the relative concentrations indicated by staining intensity of di¡erent antibodies in particular cell types remains to be proved. Colloidal gold methods might provide a more quantitative estimate of protein amounts, but di¡erences in distribution at the synapse are clearly detectable with DAB immunocytochemistry in quantitative and serial section analyses (Schwartz and Eager, 1995, 1996b; Keh and Schwartz, 1999). The presence of subunit proteins identi¢ed by the DAB immunocytochemical methods in particular cell populations was not equal to the amounts of mRNA identi¢ed in these cell types in rat (Hunter et al., 1992, 1993; Hunter and Wenthold, 1993). GluR1 immunoreactivity was notably present in most cell types which had little GluR1 mRNA. The mere presence in the neuron of GluR subunits does not prove their relative concentration in assembled receptors at postsynaptic sites. That requires a separate analysis of localization at postsynaptic densities, such as we have done for two CN cell types (Keh and Schwartz, 1999). However, the results of the present study suggest that di¡erences in the amounts of the various subunits present in particular LSO cell types do exist which may be re£ected in the composition of receptors. Acknowledgements This work was supported by NIH Grant DC00132. We are grateful to William Hormuzdiar, Agnes Keh and Dr. Sailaja Korada for excellent technical assistance.

HEARES 3314 25-10-99

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

References Adams, J.C., 1981. Heavy metal intensi¢cation of DAB-based HRP reaction product. J. Histochem. Cytochem. 29, 775. Adams, J.C., 1986. Cells of origin of cochlear e¡erents in human. Abstr. ARO Midwinter Mtg. 9, 5. Bilak, M.M., Bilak, S.R., Morest, D.K., 1996. Di¡erential expression of N-methyl-D-aspartate receptor in the cochlear nucleus of the mouse. Neuroscience 75, 1075^1097. Burnashev, N., Khodorova, A., Jonas, P., Helm, P.J., Wisden, W., Monyer, H., Seeburg, P.H., Sakmann, B., 1992. Calcium-permeable AMPA-kainate receptors in fusiform cerebellar glial cells. Science 256, 1566^1570. Cant, N.B., Casseday, J.H., 1986. Projections from the anteroventral cochlear nucleus to the lateral and medial superior olivary nuclei. J. Comp. Neurol. 247, 457^476. Caspary, D.M., Faingold, C.L., 1989. Non-N-methyl-D-aspartate receptors may mediate ipsilateral excitation at lateral superior olivary synapses. Brain Res. 503, 83^90. Caspary, D., Finlayson, P.G., 1991. Superior olivary complex: functional neuropharmacology of the principal cell types. In: Altschuler, R.A., Bobbin, R.P., Clopton, B.C., Ho¡man, D.W. (Eds.), Neurobiology of Hearing: The Central Auditory System. Raven Press, New York, pp. 141^161. Caspary, D.M., Rybak, L., Faingold, C.L., 1985. The e¡ects of inhibitory and excitatory amino acid neurotransmitters on the response properties of brainstem auditory neurons. In: Drescher, D. (Ed.), Auditory Biochemistry. Charles C. Thomas, Spring¢eld, IL, pp. 198^226. Forsayeth, J.R., Gu, Y., Hall, Z.W., 1992. BiP forms stable complexes with unassembled subunits of the acetylcholine receptor in transfected COS cells and in C2 muscle cells. J. Cell Biol. 117, 841^ 847. Friauf, E., Ostwald, J., 1988. Divergent projections of physiologically characterized rat ventral cochlear nucleus neurons as shown by intra-axonal injection of horseradish peroxidase. Exp. Brain Res. 73, 263^284. Frydman, J., Nimmesgern, E., Ohtsuka, K., Hartl, F.U., 1994. Folding of nascent polypeptide chains in a high molecular mass assembly with molecular chaperones. Nature 370, 111^117. Furuyama, R., Kiyama, H., Sato, K., Park, H.T., Maeno, H., Takagi, H., Tohyama, M., 1993. Region-speci¢c expression of subunits of ionotropic glutamate receptors (AMPA-type, KA-type and NMDA receptors) in the rat spinal cord with special reference to nociception. Mol. Brain Res. 18, 141^151. Gallo, V., Upson, L.M., Hayes, W.P., Vyklicky, L., Jr., Winters, C.A., Buonanno, A., 1992. Molecular cloning and development analysis of a new glutamate receptor subunit isoform in cerebellum. J. Neurosci. 12, 1010^1023. Gelman, M.S., Chang, W., Thomas, D.Y., Bergeron, J.J.M., Prives, J.M., 1995. Role of the endoplasmic reticulum chaperone calnexin in subunit folding and assembly of nicotinic acetylcholine receptors. J. Biol. Chem. 270, 15085^15092. Hammond, C., Helenius, A., 1995. Quality control in the secretory pathway. [Review]. Curr. Opin. Cell Biol. 7, 523^529. Helfert, R.H., Schwartz, I.R., 1987. Morphologic evidence for the presence of ¢ve cell types in the gerbil lateral superior olivary nucleus. Am. J. Anat. 179, 55^69. Helfert, R.H., Schwartz, I.R., Ryan, A.F., 1988. Ultrastructural characterization of gerbil olivocochlear neurons based on di¡erential uptake of 3 H-D-aspartic acid and a wheatgerm agglutinin-horseradish peroxidase conjugate from the cochlea. J. Neurosci. 8, 3111^3123. Helfert, R.H., Juiz, J.M., Bledsoe, S.C., Jr., Bonneau, J.M., Wenthold, R.J., Altschuler, R.A., 1992. Patterns of glutamate, glycine,

89

and GABA immunolabeling in the lateral superior olive of the guinea pig. J. Comp. Neurol. 323, 305^325. Hollmann, M., Heinemann, S., 1994. Cloned glutamate receptors. Annu. Rev. Neurosci. 17, 31^108. Hunter, C., Wenthold, R.J., 1993. Expression of the AMPA-selective glutamate receptor, GluR1, is limited in principal cell populations of the rat brain stem auditory nuclei. Abstr. ARO Midwinter Mtg. 16, 124. Hunter, C., Vu, T., Wenthold, R.J., 1992. Localization of glutamate receptor subunit mRNAs in morphologically de¢ned cells of the rat cochlear nucleus and superior olivary complex. Abstr. Mol. Biol. Hear. Deafness 83. Hunter, C., Petralia, R.A., Vu, T., Wenthold, R.J., 1993. Expression of AMPA-selective glutamate receptor subunits in morphologically de¢ned neurons of the mammalian cochlear nucleus. J. Neurosci. 13, 1932^1946. Hurtley, S.M., Helenius, A., 1989. Protein oligomerization in the endoplasmic reticulum. Annu. Rev. Cell Biol. 5, 277^307. Joelsen, D., Schwartz, I.R., 1998. Development of N-methyl-D-aspartate receptor subunit immunoreactivity in the neonatal gerbil cochlear nucleus. Microsc. Res. Tech. 41, 246^262. Kandler, K., Friauf, E., 1995. Development of glycinergic and glutamatergic synaptic transmission in the auditory brainstem of perinatal rats. J. Neurosci. 15, 6890^6904. Keh, A., Schwartz, I.R., 1999. Di¡erences in the distribution of synaptic specializations containing GluR1^4 subunits on cartwheel and octopus cells in the gerbil cochlear nucleus. Abstr. ARO Midwinter Mtg. 22, 68. Korada, S., Schwartz, I.R., 1997. Colocalization of calbindin and the ionotropic glutamate receptor subunits GluR1^4 in gerbil auditory brainstem. Soc. Neurosci. Abstr. 23, 182. Lerea, L.S., 1997. Glutamate receptors and gene induction: signalling from receptor to nucleus. [Review]. Cell. Signal. 9, 219^226. Lerma, J., Morales, M., Ibarz, J.M., Somohano, F., 1994. Recti¢cation properties and Ca2‡ permeability of glutamate receptor channels in hippocampal cells. Eur. J. Neurosci. 6, 1080^1088. Lopez, T., Lopez-Colome, A.M., Ortega, A., 1994. AMPA/KA receptor expression in radial glia. NeuroReport 5, 504^506. Lu, S.M., Schweitzer, L., Siegel, L., Cant, N.B., Dawbarn, D., 1987. Immunoreactivity to calcitonin gene-related peptide in the superior olivary complex and cochlea of cat and rat. Hear. Res. 31, 137^ 146. Luque, J., Richards, J.G., 1995. Expression of NMDA 2B receptor subunit mRNA in Bergmann glia. Glia 13, 228^232. Mano, I., Teichberg, V.I., 1998. A tetrameric subunit stoichiometry for a glutamate receptor-channel complex. NeuroReport 9, 327^ 331. Martin, L.J., Blackstone, C.D., Levey, A.I., Huganir, R.L., Price, D.L., 1993a. AMPA glutamate receptor subunits are di¡erentially distributed in rat brain. Neuroscience 53, 327^358. Martin, L.J., Blackstone, C.D., Levey, A.I., Huganir, R.L., Price, D.L., 1993b. Cellular localizations of AMPA glutamate receptors within the basal forebrain magnocellular complex of rat and monkey. J. Neurosci. 13, 2249^2263. Muller, T., Mo«ller, T., Berger, T., Schnitzer, J., Kettenmann, H., 1992. Calcium entry through kainate receptors and resulting potassium-channel blockage in Bergmann glial cells. Science 256, 1563^1566. Petralia, R.S., Wenthold, R.J., 1992a. Light and electron immunocytochemical localization of AMPA-selective glutamate receptors in the rat brain. J. Comp. Neurol. 318, 329^354. Petralia, R.S., Wenthold, R.J., 1992b. Light and electron microscopic localization of AMPA-selective glutamate receptors in the cochlear nuclei of the rat using subunit-speci¢c antipeptide antibodies. Abstr. Mol. Biol. Hear. Deafness 106.

HEARES 3314 25-10-99

90

I.R. Schwartz, P.R. Eager / Hearing Research 137 (1999) 77^90

Petralia, R.S., Yokotani, N., Wenthold, R.J., 1994. Light and electron microscope distribution of the NMDA receptor subunit NMDAR1 in the rat nervous system using a selective anti-peptide antibody. J. Neurosci. 14, 667^696. Petralia, R.S., Wang, Y.X., Zhao, H.M., Wenthold, R.J., 1996. Ionotropic and metabotropic glutamate receptors show unique postsynaptic, presynaptic, and glial localizations in the dorsal cochlear nucleus. J. Comp. Neurol. 372, 356^383. Petralia, R.S., Wang, Y.X., Mayat, E., Wenthold, R.J., 1997. Glutamate receptor subunit 2-selective antibody shows a di¡erential distribution of calcium-impermeable AMPA receptors among populations of neurons. J. Comp. Neurol. 385, 456^476. Rietzel, H.J., Friauf, E., 1998. Neuron types in the rat lateral superior olive and developmental changes in the complexity of their dendritic arbors. J. Comp. Neurol. 390, 20^40. Ryan, A.F., Schwartz, I.R., Helfert, R.H., Keithley, E., Wang, Z.X., 1987. Selective retrograde labeling of lateral olivocochlear neurons in brainstem based on preferential uptake of 3 H-D-aspartic acid in the cochlea. J. Comp. Neurol. 255, 606^616. Schwartz, I.R., 1982. A simple method for £at embedding of large thin tissues sections. Stain Technol. 57, 52^54. Schwartz, I.R., 1994. Axonal organization in the cat medial superior olivary nucleus. In: Ne¡, W.D. (Ed.), Contributions to Sensory Physiology, Vol. 8, pp. 99^129. Schwartz, I.R., 1994. Immunocytochemistry of glutamate receptor subunits in the gerbil cochlear nucleus and superior olivary complex. Abstr. ARO Midwinter Mtg. 17, 14. Schwartz, I.R., Eager, P., 1995. Ultrastructural localization of glutamate receptor subunit immunoreactivity in the gerbil superior olivary complex. Abstr. ARO Midwinter Mtg. 18, 39. Schwartz, I.R., Eager, P., 1996a. Developmental changes in glutamate receptor subunit immunocytochemistry in the neonatal gerbil lateral superior olive. Abstr. ARO Midwinter Mtg. 19, 120. Schwartz, I.R., Eager, P., 1996b. Ultrastructural localization of glutamate receptor subunit immunoreactivity in the gerbil lateral superior olive (LSO). Soc. Neurosci. Abstr. 22, 125. Schwartz, I.R., Keh, A., 1997. Distribution of GluR receptor immunoreactivity in relation to synaptic terminals on gerbil lateral superior olivary (LSO) neurons: a three dimensional study. Abstr. ARO Midwinter Mtg. 20, 83. Schweitzer, L., Lu, S.M., Dawburn, D., Cant, N.B., 1985. Calcitonin

gene-related peptide in the superior olivary complex of cat and rat: a speci¢c label for the lateral olivocochlear system. Neurosci. Abstr. 11, 1051. Simmons, D.D., Raji-Kubba, J., 1993. Postnatal calcitonin gene-related peptide in the superior olivary complex. J. Chem. Neuroanat. 6, 407^418. Spangler, K.M., Warr, W.B., Henkel, C.K., 1985. The projections of principal cells of the medial nucleus of the trapezoid body in the cat. J. Comp. Neurol. 238, 249^261. Spangler, K.M., Cant, N.B., Henkel, C.K., Farley, G.R., Warr, W.B., 1987. Descending projections from the superior olivary complex to the cochlear nucleus of the cat. J. Comp. Neurol. 259, 452^465. Suneja, S.K., Benson, C.G., Gross, J., Potashner, S.J., 1995. Evidence for glutamatergic projections from the cochlear nucleus to the superior olive and the ventral nucleus of the lateral lemniscus. J. Neurochem. 64, 161^171. Tachibana, M., Wenthold, R.J., Morioka, H., Petralia, R.S., 1994. Light and electron immunocytochemical localization of AMPAselective glutamate receptors in the rat spinal cord. J. Comp. Neurol. 344, 431^454. Yu, S.-M., Schwartz, I.R., 1989. Flat embedding for immunocytochemistry. Stain Technol. 64, 143^146. Warr, W.B., Spangler, K.M., 1989. A novel projection of the ventral nucleus of the trapezoid body in the rat. Neurosci. Abstr. 15, 745. Wenthold, R.J., Yokotani, N., Doi, K., Wada, K., 1992. Immunochemical characterization of the non-NMDA glutamate receptor using subunit-speci¢c antibodies: evidence for a hetero-oligomeric structure in rat brain. J. Biol. Chem. 267, 501^507. Wenthold, R.J., Petralia, R.S., Blahos, J., Niedzielski, A.S., 1996. Evidence for multiple AMPA receptor complexes in hippocampal CA1/CA2 neurons. J. Neurosci. 16, 1982^1989. Wu, S.H., Kelly, J.B., 1992. Synaptic pharmacology of the superior olivary complex studied in mouse brain slices. J. Neurosci. 12, 3084^3097. Wu, T.-Y., Lie, C.-I., Chang, Y.-C., 1996. A study of the oligomeric state of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid-preferring glutamate receptors in the synaptic junctions of porcine brain. Biochem. J. 319, 731^739. Zook, J.M., DiCaprio, R.A., 1988. Intracellular labeling of a¡erents to the lateral superior olive in the bat, Eptesicus fuscus. Hear. Res. 34, 141^147.

HEARES 3314 25-10-99