The AMPA glutamate receptor GluR3 is enriched in oxytocinergic magnocellular neurons and is localized at synapses

Pergamon 0306-4522(94)00513-3

Neuroscience Vol. 65, No. 2, pp. 563-575, 1995 Elsevier Science Ltd

Copyright © 1995 IBRO Printed in Great Britain, All rights reserved

0306-4522/95 $9.50 + 0.00

THE A M P A G L U T A M A T E RECEPTOR GluR3 IS E N R I C H E D IN O X Y T O C I N E R G I C M A G N O C E L L U L A R

N E U R O N S A N D IS LOCALIZED AT SYNAPSES

S. D. GINSBERG,*tll D, L. PRICE,t~§ll C. D. BLACKSTONE,§¶ R. L. HUGANIR§¶ and L. J. MARTINt§II

Departments of tPathology, :~Neurology, and §Neuroscience, IjNeuropathology Laboratory and ¶Howard Hughes Medical Institute, The Johns Hopkins University School of Medicine, Baltimore,

MD 21205-2196, U.S.A.

Abstract--The cellular localization of ~-amino-3-hydroxy-5-methyl-4-isoxazole propionate glutamate receptor, GluR3, was identified using antibodies that recognize the N-terminus of the predicted polypeptide sequence of GIuR3. Regional immunoblot analysis of monkey brain homogenates identified a protein of ~ 102,000tool. wt that was enriched in hypothalamus. Immunocytochemistry demonstrated that GluR3 was enriched within the hypothalamic magnocellular neurosecretory nuclei and axons of the hypothalamo-neurohypophysial tract in rat and monkey. GIuR3 immunoreactivity co-localized to oxytocin-containing, but not vasopressin-containing, neurons of the hypothalamic paraventricular nucleus, supraoptic nucleus and accessory magnocellular nuclei. Ultrastructurally, GluR3 immunoreactivity was enriched throughout cytoplasm of the somatodendritic compartment and was associated with postsynaptic and presynaptic structures. GIuR3 immunoreactivity was frequently observed to be clustered at the plasma membrane of the somatodendritic compartment, consistent with the predicted localization of a membrane-bound ion channel. Additionally, GluR3-immunoreactive axon terminals in synaptic contact with unlabeled dendrites within the retrochiasmatic area and bed nucleus of the stria terminalis were observed, providing morphological evidence for a presynaptic a-amino-3-hydroxy-5-methyl-4- isoxazole propionate receptor.

By immunoblot analysis and immunocytochemistry using antibodies directed against a specific ct-amino- 3-hydroxy-5-methyl-4-isoxazole propionate receptor in rat and monkey brain, our findings suggest a highly selective hypothalamic distribution of the GluR3 subunit that may have functional significance in the glutamatergic regulation of oxytocinergic neurons.

Glutamatergic neurotransmission is thought to be an essential mediator of a variety of brain mechanisms including synaptic plasticity, development, sensory processing, and learning and memory. 5'37 Con- verging lines of anatomical, physiological and phar- macological evidence suggest that glutamatergic neurotransmission mediates responses of magnocellular neurosecretory neurons located within the anterior hypothalamus. For example, glutamate immunoreactivity has been localized to axon terminals in synaptic contact with neurons in the hypothalamic paraventricular nucleus (PVH) and supraoptic nucleus (SON). 34'35'57'60 Iontophoretic application of excitatory amino acids to hypothalamic slice preparations rapidly elicits depolarizing currents that are blocked by the application of glutamate

*To whom correspondence should be addressed at the Neuropathology Laboratory.

Abbreviations: AMPA, ~-amino-3-hydroxy-5-methyl-4- isoxazole propionate; CNQX, 6-cyano-7-nitroquinoxaline-2,3-dione; DAB, 3,3'-diaminobenzidine; GIuR, glutamate receptor; NMDA, N-methyl-D-aspartate; PBS, phosphate-buffered saline; PVH, hypothalamic paraventricular nucleus; SON, supraoptic nucleus.

antagonists) '2 Additionally, glutamatergic neurotransmission mediates the majority of spontaneous and evoked excitatory postsynaptic potentials in hypothalamic slice preparations and hypothalamic cell cultures. 59,64 The application of kynurenic acid, a broad-spectrum ionotropic glutamate receptor (GluR) antagonist, as well as the c~-amino-3-hydroxy- 5-methyl-4-isoxazole propionate (AMPA)/kainate antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) blocks these excitatory postsynaptic potentials, suggesting that N-methyl-t)-aspartate (NMDA) and non-NMDA receptors are located on magnocellular neurons that mediate these events. ~5'6°'64

Molecular cloning and functional expression studies have identified subtypes of receptors that bind glutamate and elicit excitatory postsynaptic currents. These GluRs have been categorized by their pharmacology, molecular structure and mechanisms of signal transduction. Ionotropic GluRs, channels composed of oligomers (presumed to be pentamers) of homologous subunits encoded by separate genes, consist of: N M D A receptors (NMDAR1, NMDAR2A-D); AMPA receptors (GluR1-4); and kainate receptors (GIuR5-7, KA1, KA2). H,3s,49 We have recently identified GluR distribution in the

563

564 S. D. Ginsberg et al.

macaque hypothalamus by immunocytochemist ry using affinity-purified antipeptide antibodies directed against specific subunits of the A M P A receptor [GluR1, a common epitope to the GIuR2 and GIuR3 subunits (designated GIuR2/3), and GIuR4], kainate receptor [a common epitope to the GluR6 and GluR7 subunits (designated as GluR6/7)] and a metabotropic GluR (mGluRlc0 . ~3 However, the PVH and SON demonst ra ted low levels of immunoreactivi ty for these GluR proteins. The N M D A R 1 subunit has been localized by immunocytochemist ry to magnocellular neurons within the rat PVH, SON and virtually all other neurons in the brain, 4~ al though specific n o n - N M D A GluRs utilized by these neurons are unknown. Parts of these data have been reported in abstract form. ~z

EXPERIMENTAL PROCEDURES

Genera tion of affinity -purified polyclonal an tibodies to GluR 3

To differentiate GIuR2 and GluR3 subunits immunocytochemically, which share considerable sequence homology at the C-terminus, 25 antibodies were generated against diver- gent regions of the GluR2 and GluR3 subunits. A synthetic peptide KVRLDEREFPEAKNAPLK, corresponding to the N-terminal amino acid residues 260-276, and a second synthetic peptide LTKNTQNFKPAPATNT, corresponding to the C-terminal amino acid residues 831-846 of the predicted polypeptide sequence encoded by the GIuR3 cDNA in rat, 25 were synthesized and purified as described previously. 3,4'29 A third antibody raised against a synthetic peptide NFKPAPATNTQ, corresponding to the C-terminal amino acid residues 837-847 (a generous gift from Dr R. J. Wenthold), was also used. These peptides do not share appreciable amino acid sequence identity with any known neuropeptides or neuropeptide receptors (BLAST network service, NCBI, Bethesda, MD). Briefly, these peptides were synthesized (430A Peptide Synthesizer, Applied Biosystems, Inc., Foster City, CA) and an N-terminal lysine residue was added to facilitate coupling to the carrier protein. Peptides were purified using reverse-phase high-performance liquid chromatography and conjugated to thyroglobulin with glutaraldehyde. Antibodies were generated in New Zealand white rabbits (Hazelton Research Products, Denver, PA), and the GluR3 antiserum was affinity purified on a column containing bovine serum albumin-conjugated GluR3 peptide coupled to Affi-Gel 15 (Bio-Rad, Rockville Center, NY).

Monkey brain homogenate preparation and immunoblotting

Tissue homogenates were prepared from unfixed frozen brain of monkey (Macaca mulatta) (n = 1). Samples were dissected and homogenized as described previously? °'32 For the PVH tissue homogenate, a 2-mm diameter biopsy punch (Acuderm, Inc., Fort Lauderdale, FL) was used to microdissect regions of monkey PVH that contained primarily magnocellular neurons. ~4 Tissue homogenates were subjected to sodium dodecyl sulfate~olyacrylamide gel electrophoresis (8% polyacrylamide gels), transferred to nitrocellulose by electroblotting (30V, overnight at 4°C) and immunoblotted with affinity-purified GluR3 antibody at a dilution of 1 : I00, as described previously. ~°'32 Immuno- reactive proteins were visualized with enhanced chemi- luminescence (Amersham, Arlington Heights, IL).

Single-label immunocytochemistry

Six cynomolgus monkeys (Macaca fascicularis) and six Sprague-Dawley rats (Charles River, Wilmington, MA) were exsanguinated with ice-cold I%o paraformaldehyde

in phosphate buffer (0.12M, pH 7.4) and perfused with ice-cold 4% paraformaldehyde in phosphate buffer, as described previously) 4 Brains were blocked in the coronal or sagittal plane (one monkey and one rat), postfixed in 4% paraformaldehyde in phosphate buffer for 2 h, cryopro- tected in a 20% glycerol solution in phosphate-buffered saline (PBS) and cut at a section thickness of 40/~m. Representative sections were stained with Cresyl Violet to assign cytoarchitectonic criteria for the delineation of nu- clear boundaries. Adjacent sections were processed for immunocytochemistry using GIuR3 antibodies. Tissue sections were blocked/permeabilized for 1 h in a PBS solution containing 4% normal goat serum and 0.4% Triton X-100 at 4°C. Tissue sections were then incubated for 48 h in a PBS solution containing the primary antiserum at a dilution of 1:200, 1%o normal goat serum and 0.2% Triton X-100 at 4°C. Secondary and tertiary steps were done using a Vectastain Elite kit (Vector Laboratories, Burlingame, CA) with 3,3'-diaminobenzidine (DAB; Sigma, St Louis, MO) as a chromogen. If the primary, secondary or tertiary reagents from the protocol were removed, there was no immunocytochemical labeling of the tissue. Incubation with rabbit immunoglobulin (Gibco, Grand Island, NY) fol- lowed by the histochemical procedure failed to produce labeling in tissue sections. To assess the specificity of the GluR3 antibody on hypothalamic tissue sections, GIuR3 antiserum was preadsorbed for 12 h at 4°C with 50 #g/ml of synthetic GIuR3 peptide prior to the addition of monkey hypothalamic sections and subsequent histochemical processing.

For ultrastructural analysis, four Sprague-Dawley rats (Charles River) were perfused with ice-cold 4% paraformaldehyde/0.1%0 glutaraldehyde/15% saturated picric acid in phosphate buffer, postfixed in the same solution for 2 h at 4°C and stored in PBS overnight. Tissue sections were cut at a section thickness of 50#m on a Vibratome and processed for immunocytochemistry as described for light microscopic analysis, with permeabilization omitted. Microdissected samples of selected immunostained tissue sections were placed in 2% osmium tetroxide for l h, dehydrated and flat embedded. Semithin sections (1 ,um) and thin sections (gold interference color) were cut. Selected thin sections were contrasted with uranyl acetate and lead citrate, whereas other thin sections were not contrasted. Sections were viewed with a Phillips CM 12 electron microscope.

Double-label immunocvtochemistry

A double-label immunocytochemical procedure was employed to assess the distribution of GluR3 immunoreactivity relative to oxytocin and vasopressin immunoreactivity within magnocellular neurons of the monkey PVH and SON. Using DAB and benzidine dihydrochloride as chromogens, ~8,29'32 hypothalamic sections were immunostained for GIuR3 and oxytocin or GluR3 and vasopressin. The distribution and density of GluR3-, oxytocin-, and vasopressin-immunoreactive perikarya were plotted using a computerized, video-based mapping system. 14,32 In tissues from three monkeys, regional maps were generated at a magnification of x 25; neurons were quantified at a magnification of x200 on four tissue sections evenly spaced throughout the rostrocaudal extent of the anterior hypothalamus per subject. Cross-adsorption controls were per- formed to assess the specificity of the double-label preparations. Oxytocin and vasopressin antisera (monoclonal antibodies) 2°'2~ were preadsorbed with 50/zg/ml of the synthetic GluR3 peptide; the GluR3 antiserum was preadsorbed with 50#g/ml of synthetic oxytocin and synthetic vasopressin (Cambridge Research Biochemicals, Wilmington, DE) prior to the histochemical processing of tissue sections. Double-label immunofluorescence was also employed for confocal microscopic analysis. Tissue sections were immunostained for GluR3/oxytocin and

GluR3 in oxytocinergic neurons 565

GluR3/vasopressin. The histochemical procedure has been described in detail elsewhere. 14 Briefly, GluR3 was visualized by a secondary antibody conjugated to Texas Red (Jackson ImmunoResearch, West Grove, PA); oxytocin and vasopressin antibodies were visualized by a secondary antibody conjugated to ftuorescein isothiocyanate (Jackson). Immunofluorescent preparations were evaluated on a confocal laser scanning microscope (MRC 600, Bio-Rad, Rockville Center, NY) equipped with image analysis software.

RESULTS

All three GIuR3 antibodies displayed similar patterns of immunoreactivity as assessed by western blot analysis and by immunocytochemistry. The N-terminal GluR3 antibody yielded the best signal- to-noise ratio and is described below.

lmmunoblotting

In regional monkey brain homogenates, one band of proteins with an approximate molecular mass of 102,000 was identified by western blot analysis using an antipeptide antibody against GluR3 (Fig. 1). An intense GluR3-immunoreactive band was observed on blots of total hypothalamic tissue and microdissected tissue homogenates that contained the PVH. A moderate signal of GiuR3 immunoreactivity was detected on blots of striatal and hippocampal tissue homogenates, whereas low GluR3 immunoreactivity was observed on blots of basal forebrain and cerebellar tissue homogenates. Detection of the band of proteins by immunoblotting was abolished when

the antibody was preadsorbed with 50/~g/ml of synthetic GluR3 peptide.

Irnmunocytochemistry

GluR3 irnmunoreactivity was enriched in the hypothalamus of rat and monkey. GluR3 was localized to a subpopulation of magnocellular neurosecretory neurons within the PVH, SON and accessory magnocellular nuclei (Figs 2, 3A, B). GluR3 immunoreactivity was not observed in PVH neurons that displayed parvicellular morphology. A bundle of lightly beaded, GluR3-immunoreactive axons was observed exiting the PVH dorsolaterally, arching past the fornix and joining GluR3-immunoreactive axons that originated from the SON to course within the hypothalamo-neurohypophysial tract. The neural lobe of the posterior pituitary was highly enriched in GluR3-immunoreactive varicosities, whereas the anterior pituitary expressed little GIuR3 immunoreactivity (Fig. 3E, F). Punctate, GluR3-immunoreactive varicosities were also visualized in the lateral dorsal subdivision of the rat bed nucleus of the stria terminalis, presumably originating from axon collaterals of magnocellular neurons (Fig. 3D). In rat, but not monkey, moderately GluR3-immunoreactive neurons were localized to the suprachiasmatic nucleus (Fig. 3C). The preoptic area, mediobasal hypothalamus and mammillary complex expressed very little GluR3 immunoreactivity in both rat and monkey.

Hyp P V Hip Str FB Cb C

1 1 6 - 9 7 - 6 6 -

4 5 -

Fig. 1. A single protein of ~ 102,000 mol. wt was identified by immunoblot analysis using an antipeptide antibody against GluR3 in regional homogenates of monkey brain. A dense band of GluR3 immunoreactivity was detected on blots containing total hypothalamic and PVH tissues, a moderate band of GluR3 immunoreactivity was detected on blots of striatal and hippocampal tissue homogenates, and weak or no GIuR3 immunoreactivity was observed in tissue homogenates of the basal forebrain and cerebellum. No signal was observed when the GluR3 antibody was preadsorbed with 50 # g/ml of synthetic GluR3 peptide. C, preadsorption control; Cb, cerebellum; FB, basal forebrain; Hip, hippocampus; Hyp, hypothalamus;

PV, paraventricular nucleus; Str, striatum.

566 S.D. Ginsberg et al.

A

B

¢

,j

t IP

p ~

J 11

I i

P

-I |

D

!

"I

B i l

- r J

f

|,

If

1

4

r '

,¢.-.g ~', "7 i

w

j -e ;~ v ~ ¸

¢

Fig. 2. Bright-field photomicrographs of coronal sections demonstrating GIuR3 immunoreactivity within the monkey magnocellular neurosecretory nuclei. Scale bar = 100/~m (A-D). (A) GluR3-immunoreactive neurons within the SON. (B) Photomontage of GluR3-immunoreactive neurons within the magnocellular PVH. (C) PVH tissue section labeled wtih the GluR3 antibody. (D) Adjacent tissue section to C where the GluR3 antibody was preadsorbed with synthetic GIuR3 peptide prior to histoehemical

processing.

567

• G

I t

¢ " ~ . . . . : : - . •

"5 "n .~ .~

• - ~ .= "~,

. . ~ o ,z, # .

~ - - ' = r 3 o

0 ~ 0 ~

II ~ "o ~

.~ ~ ~:~

. - & . ~

= ; > o ~ . -

._ .~ ~ . o ~..

~ z s 2 o 0

~.~- ~ ~

• = ~ ~ . ~ o • ~ ~ :~ ~ . . -

0 =1 ~D... o

:=rO ~ o,..=

" ~ = o o

. - ~ ~ ~ ' ~

q


Faint GluR3 immunoreactivity was also observed in non-pyramidal neurons within stratum oriens of the hippocampus, medium spiny striatal neurons and layer V pyramidal neurons in neocortex. GIuR3 immunoreactivity was expressed at low levels throughout the rest of the telencephalon, dien- cephalon and mesencephalon. Preadsorption of the GIuR3 antibody with synthetic GIuR3 peptide completely abolished immunocytochemical labeling (Fig. 2C, D).

GluR3 immunoreactivity was enriched within the somata and proximal dendrites of hypothalamic neurons that displayed magnocellular morphology. In the rat, ultrastructural analysis identified GluR3-immunoreactive somata and dendrites within the SON (Fig. 4A-C). The majority of synapses onto GluR3-immunoreactive somal profiles were asymmetrical. On dendritic profiles, clustering of GluR3 immunoreactivity on the plasma membrane was observed consistently. GIuR3 immunoreactivity was also observed within presynaptic axonal compartments in uncontrasted and contrasted thin sections, including myelinated axons and axon terminals within the bed nucleus of the stria terminalis and axon terminals in the retrochiasmatic area (Fig. 4D-F). Presynaptic axon terminals that contained GluR3 immunoreactivity formed symmetric and asymmetric synapses onto unlabeled dendrites within the retrochiasmatic area and bed nucleus of the stria terminalis of rat (Fig. 4E, F).

The neurochemical identity of GluR3- immunoreactive neurons in monkey hypothalamus was assessed by double-label immunocytochemistry. Preadsorption of GIuR3 antiserum with oxytocin or vasopressin did not alter immunostaining, nor did preadsorption of oxytocin and vasopressin antisera with synthetic GluR3 peptide. Qualitative observations indicated that a subpopulation of oxytocin-containing neurons co-localized GluR3 immunoreactivity, whereas GluR3 immunoreactivity was expressed minimally within vasopressin- immunoreactive neurons (Fig. 5). Quantitative analysis in the PVH and SON of three monkeys revealed that 99 .1+0.5% (S.E.M.) of GluR3- immunoreactive neurons contained oxytocin immunoreactivity. Variability in the percentage of oxytocin-containing neurons that co-expressed GIuR3 immunoreactivity was identified. Specifically, 55.7 + 1.7% of oxytocin-immunoreactive neurons quantified contained GIuR3 immunoreactivity within the PVH. In the SON, 45.8___ 0.9% of oxytocin- immunoreactive neurons quantified expressed GluR3 immunoreactivity. In terms of rostrocaudal distribution, the majority of GluR3-immunoreactive neurons was observed at mid levels of the PVH and SON, with the rostral and caudal poles containing relatively fewer GluR3-immunoreactive neurons. In contrast, 1.8 ± 0.9% and 1.1 _+ 0.3% of vasopressin- immunoreactive neurons quantified co-localized GluR3 immunoreactivity within the PVH and SON,

respectively. The assessment of oxytocin and GIuR3 immunoreactivity within the same z-axis plane using confocal microscopy (approximate section thickness <0.5/~m) 23 combined with double-label immunocytochemistry demonstrated flocculent oxytocin and vasopressin immunoreactivity dispersed throughout the cytoplasm, whereas GluR3 immunoreactivity was consistently observed at the plasma membrane, similar to the present ultrastructural assessment (Fig. 6).

DISCUSSION

Antibody characterization and regional distribution

This report identifies the cellular localization and highly selective enrichment of the AMPA receptor subunit GIuR3 within the hypothalamus of rat and monkey. Site-directed antibodies were raised against a synthetic peptide corresponding to the N-terminus and C-terminus of GluR3. 6'25 Character- ization of the GIuR3 antiserum by western blot analysis of monkey brain homogenates demonstrated a single band of proteins of an approximate molecular mass of 102,000, consistent with the size predicted by analysis of cloned GIuR3 cDNA, as well as immunoblot analysis using human brain tissue homogenates. 3'6'25 Although the conservation of amino acid residues between rat and monkey GluR3 peptides is presently unknown, a high degree of sequence identity between rat and human AMPA receptors has been demonstrated. For example, the full-length cDNA sequence of the human GluRI subunit predicts a protein that shares 97% identity with the rat GIuRI homolog. 4s Although the human GIuR3 cDNA sequence has not been fully characterized, oligonucleotide probes (approxi- mately 100 base pairs) that localized the GIuR3 subunit to the "long" arm of the human X chromosome share 96% identity with the rat GIuR3 homolog. 33

Western blot analysis demonstrated intensely GluR3-immunoreactive bands in monkey hypothalamus and PVH, whereas lower levels of immunoreactivity were detected in hippocampus and striatum; cerebellum and basal forebrain had minimal expression of the GluR3 subunit. This differential distribution is consistent with immunoblot analysis using the GIuR3 antibody in regional homogenates of human brain. 3 Immuno- cytochemistry on rat and monkey tissue sections detected GluR3-immunoreactive neurons within the magnocellular neurosecretory system, with lower expression of GluR3 throughout the remainder of the neuraxis. Using immunocytochemistry, this antibody appears to detect brain regions with relatively high levels of GIuR3 expression, but may be less sensitive in detecting lower levels of protein expression (e.g. within the hippocampus and striatum) relative to the PVH and SON. Alternatively, this difference in immunocytochemical localization may

Fig. 4. Ultrastructural assessment of GIuR3 immunoreactivity in rat hypothalamus. Scale bars = 0.5 ~tm (A F). (A) Unlabeled axon terminal (t) forming an asymmetric axosomatic synapse (arrowheads) onto a GluR3-immunoreact ive profile in the SON. (B) GIuR3 immunoreactivity within a dendrite (white d) in close apposition to an unlabeled dendrite (black d) and an unlabeled axon terminal (t) in the SON. (C) Clustering of DAB reaction product (arrows) in a GluR3-immunoreact ive dendrite (d) apposed to an unlabeled axon terminal (t) in the SON. (D) Myelinated axon (a) in the bed nucleus of the stria terminalis containing GluR3 immunoreactivity. An adjacent myelinated axon (asterisk) does not contain GIuR3 immunoreactivity, (E) An axon terminal (t) containing DAB reaction product (white arrows) forms a symmetric synapse (arrowhead) with an unlabeled dendrite (d) in the retrochiasmatic area. An unlabeled terminal (asterisk) forms an asymmetric synapse (arrowhead) with the same dendrite. (F) A GluR3-immunoreact ive axon terminal (t) forming

an asymmetric synapse (arrowhead) with an unlabeled dendrite (d) in the retrochiasmatic area.

569


Fig. 5. Double-label immunocytochemical preparation illustrating the relationship between GIuR3, oxytocin and vasopressin immunoreactivities in the monkey hypothalamus. Scale bars = 100#m (A); 50/~m (B-D). (A) GluR3 immunoreactivity (greenish-black crystals) does not appear to co-localize with vasopressin immunoreactivity (brown perikarya) within the PVH. Arrows indicate singleqabeled GluR3- immunoreactive neurons. The boxed area represents the area shown at higher power in B. (B) GIuR3 and vasopressin immunoreactivities are observed within separate pools of magnocellular neurons. Asterisks denote single-labeled vasopressin-immunoreactive neurons. Note that some vasopressin-immunoreactive neurons are under GluR3-immunoreactive neurons. (C) GluR3 immunoreactivity (greenish-black crystals) is co-localized to oxytocin-immunoreactive neurons (brown perikarya) within the magnocellular PVH. Asterisks denote oxytocin-immunoreactive neurons that do not contain GluR3 immunoreactivity. (D) Oxytocin-immunoreactive neurons within the SON also co-localize GluR3 immunoreactivity.

Asterisks denote oxytocin-immunoreactive neurons that do not contain GluR3 immunoreactivity.

be caused by regional differences in the post- t ransla- t ional processing of GluR3, tissue fixation and /o r permeabi l i ty of the an t ibody in tissue sections. In situ

hybr id iza t ion histochemical studies have localized GluR3 subuni t m R N A in rat brain. 6'2s48'5~ GIuR3 m R N A has a more restricted dis t r ibut ion relative to

the A M P A receptor subuni ts G l u R I , G luR2 and GluR4, consistent with our immunocytochemica l observat ions. Fo r example, GluR3 m R N A is detected in neocortex, h ippocampus , s t r ia tum and hypo tha l amus (including the PVH and SON). 48s~ Wi th in some regions, notably the mesencephalon, a

i

572 S. D. Ginsberg et al.

broader distribution of GIuR3 mRNA has been reported than what was observed using immunocytochemistry. Discrepancies between the detection of GIuR3 message and GIuR3 protein may be related to a variety of factors, including the kinetics of mRNA translation, the turnover rate of the GIuR3 subunit and stability of the protein. Presently, the rate of ionotropic GluR synthesis and the mechanism of receptor assembly and plasmalemmal insertion are unknown.

Immunocytochemistry using an antibody directed against GIuR3 allows for differentiation of the GIuR2 subunit from the GIuR3 subunit as compared to previous immunocytochemical analyses using antibodies that recognize a common epitope on GluR2 and GluR3.13'3°'32'4° For example, GIuR2/3 immunoreactivity is not observed within the monkey PVH and SON, but is localized to neurons within the adjacent anterior hypothalamic area, medial preoptic area, suprachiasmatic nucleus and dorsomedial nucleus, j3 Additionally, GluR2/3-immunoreactive neurons are widely distributed throughout several brain regions in rat and monkey, including the hippocampal formation, striatum and neocortex. 3°'32'4° The present report demonstrates that GluR3 is enriched primarily within hypothalamic magnocellular neurons, indicating that localization studies that use GluR2/3 antibodies may recognize the GIuR2 subunit predominantly when used for immunocytochemistry on tissue sections. Preliminary data using an antipeptide antibody directed against the GIuR2 subunit display a similar pattern of immunocytochemical labeling with GluR2/3 antibodies, but do not overlap with GIuR3 immunoreactivity in monkey hypothalamus. ~2 Immunocytochemical differentiation of these two AMPA receptor subunits is essential, because the GIuR2 subunit has a wide- spread distribution throughout the brain and is functionally distinct from other AMPA receptor subunits because of its impermeability to calcium ions and linear current-voltage relationship when expressed in Xenopus oocytes and HEK 293 cells. 19'22"61

We have identified GIuR3 immunoreactivity within hypothalamic neurons of the PVH and SON that display magnocellular morphology, similar to previous light and electron microscopic analyses that used Golgi techniques and immunocytochemistry for vasopressin and oxytocin. 27'42'43'5~'s2 GluR3 immunoreactivity was abundant throughout the somatodendritic compartment and was observed consistently in clusters at the plasma membrane, verifying observations of GIuR3 immunoreactivity using light and confocal microscopy. Postsynaptic clustering of AMPA receptor subunits has also been reported in vitro using hippocampal cultures. 8 However, the abundance of GluR3 immunoperoxidase labeling throughout the cytosol may represent non-functional and/or unassembled receptor proteins. It is likely that GluR3 exists within plasmalemmal and cytoplasmic corn-

partments. The precise subcellular localization of GluR3 to magnocellular cytosol, postsynaptic membranes and presynaptic membranes requires further electron microscopic (e.g. immunogold preparations) evaluation.

Functional implications of GluR 3 immunoreactivity in magnocellular neurons

Quantitative analysis of double-label immunocytochemical preparations in monkey PVH and SON indicated that >98% of GluR3-containing neurons were oxytocinergic, whereas only 1% of GluR3- immunoreactive neurons were vasopressinergic. Vasopressin- and oxytocin-containing neurons com- prise separate pools of magnocellular cells that differentially co-express a variety of neuropeptides. 9'~°'~4'36 These two subpopulations of magnocellular neurons are distinguished electrophysiologically by their distinct firing patterns. Specifically, vasopressin-containing neurons display phasic bursting activity when stimulated, whereas oxytocin-containing neurons are non-phasic and tend to display a fast continuous discharge of spikes upon activation. TM Differential distribution of the GIuR3 subunit may contribute to the distinct electrophysiological activities observed within magnocellular subpopulations, because glutamatergic neurotransmission mediates spontaneous and evoked excitatory postsynaptic potentials of magnocellular cells recorded in both hypothalamic slice preparations and cell cultures. 59'64 Pharmacologi- cal application of AMPA, but not NMDA or metabotropic agonists, to the SON of lactating rats stimulates the release of oxytocin, which is blocked by the application of CNQX. 39 Furthermore, vasopressin secretion upon AMPA application to the SON is less than oxytocin release, 39 which may be a result of the differential localization of GluR3 in these magnocellular neurons. Functionally, a well-documented role of oxytocinergic neurons is to fire synchronously from both the PVH and SON in response to rises in intramammary pressure during suckling to facilitate the milk ejection re f l ex . 62'63 Ultrastructural evidence for the reorganization of synaptic contacts, glial appositions and soma-soma contacts of oxytocin-immunoreactive neurons has been reported in lactating female rats as compared to virgin females or males. 16'17'55'56 Because synaptic plasticity has been attributed to the activation of GluRs in several brain regions, 5'37 it is likely that plastic changes in neuroendocrine circuitry are also regulated through GluR-mediated mechanisms.

Magnocellular PVH neurons differ morphologi- cally and physiologically from parvicellular PVH neurons, which project either through the external lamina of the median eminence to the portal capillary plexus and affect anterior pituitary function or terminate in the brainstem and/or upper spinal cord. 53'54 Although parvicellular neurons are more susceptible to damage by excitatory amino acids

GIuR3 in oxytocinergic neurons 573

than magnocellular neurons, 18,65 n o n - N M D A GluRs have not been localized immunocytochemically to parvicellular neurons. 13 It is likely that other types of GluR, which have yet to be identified morphologi- cally, are utilized by parvicellular neurons.

Axons of magnocellular neurons course in the hypothalamo-neurohypophysial tract, terminate in the perivascular spaces within the posterior pituitary, and release vasopressin and oxytocin to the systemic circulat ion? 4 Varicose axon collaterals of magnocellular neurons, a subpopulat ion of which are myelinated, have been observed to terminate within the retrochiasmatic area and bed nucleus of the stria terminalis 46,52 and correspond to GluR3- immunoreact ive processes visualized in the present investigation. Our ultrastructural localization of GluR3 to a subpopulat ion of axon terminals in synaptic contact with unlabeled dendrites in the rat retrochiasmatic area is a morphological demonstration of a presynaptic A M P A receptor. GIuR1 immunoreactivi ty has also been observed in Schaffer collateral axons without direct synaptic specializ- ations, 31 which suggests that A M P A receptor subunits can be localized to axonal compartments in some brain regions. Clusters o f GIuR3 immunoreactivity were observed in close association to presynaptic plasma membranes, as would be predicted o f membrane-bound ion channels, and were not visualized in secretory vesicles. This observation con- trasts with electron microscopic assessments of vasopressin and oxytocin immunoreactivities, which are localized to secretory vesicles and Herring bodies within magnocellular axon terminals. 24,4z43'5~ Pre- synaptic GluR3 ion channels may act as auto- receptors for coordinat ing synchronous firing of oxytocinergic neurons, and may play a permissive role in stimulus-secretion coupling. AlternativelY, GluR3-immunoreact ive axon terminals may receive glutamatergic axoaxonic synapses, as glutamate-

immunoreact ive terminals are observed throughout the retrochiasmatic area and perinuclear S O N . 34

In addition to the present report, A M P A receptor- containing varicosities (presumably axon terminals from magnocellular neurons) have been described in the posterior and intermediate lobes of the rat pituitary; however, these elements have not been characterized ultrastructurally. 26 The possibility exists that G l u R proteins (especially those localized outside of the b lood-bra in barrier) subserve non- receptor functions, as suggested by studies of mutant GIuR3 receptors expressed in Xenopus oocytes. 5° Moreover , serum antibodies to GIuR3 have been detected in patients with Rasmussen's encephalitis, thereby implicating GluR3 autoantibodies in the pathological process of seizure disorders. 47

CONCLUSIONS

The A M P A receptor GluR3 is enriched primarily within a populat ion o f oxytocinergic neurons in the PVH and SON, and appears to have the most restricted regional distribution o f the known GluRs. The localization of GluR3 immunoreactivi ty throughout the somatodendrit ic domain and to both the presynaptic and postsynaptic compartments suggests a functional role for GIuR3 in the glutamatergic activation and regulation of the hypothalamic magnocellular neurosecretory system.

Acknowledgements--The authors thank Dr Robert J. Wenthold for donating GluR3 antiserum, Dr Ann-Judith Silverman for donating oxytocin and vasopressin antisera, and Dr Linda C. Cork for providing monkey tissue. This work was supported by grants from the US Public Health Service (NS 20471, NS 07179, AG 09214), as well as the Metropolitan Life Foundation. Drs Price and Martin are the recipients of a Leadership and Excellence in Alzheimer's Disease (LEAD) award (AG 07914). Dr Price is the recipient of a Javits Neuroscience Investigator Award (NS 10580).

REFERENCES

1. Arnauld E., Cirino M., Layton B. S. and Renaud L. P. (1983) Contrasting actions of amino acids, acetylcholine, noradrenaline and leucine enkephalin on the excitability of supraoptic vasopressin-secreting neurons. A microionto- phoretic study in the rat. Neuroendocrinology 36, 187-196.

2. Arnauld E., Layton B. S., Padjen A. L. and Renaud L. P. (1983) Actions of acidic amino acids on the excitability of medial hypothalamic neurons in the rat. Neuroendocrinology 37, 184-192.

3. Blackstone C. D., Levey A. I., Martin L. J., Price D. L. and Huganir R. L. (1992) Immunological detection of glutamate receptor subtypes in human central nervous system. Ann. Neurol. 31, 680683.

4. Blackstone C. D., Moss S. J., Martin L. J., Levey A. I., Price D. L. and Huganir R. L. (1992) Biochemical characterization and localization of a non-N-methyl-D-aspartate glutamate receptor in rat brain. J. Neurochem. 58, 1118-1126.

5. Bliss T. V. P. and Collingridge G. L. (1993) A synaptic model of memory: long-term potentiation in the hippocampus. Nature 361, 31-39.

6. Boulter J., Hollmann M., O'Shea-Greenfield A., Hartley M., Deneris E., Maron C. and Heinemann S. (1990) Molecular cloning and functional expression of glutamate receptor subunit genes. Science 249, 1033-1037.

7. Cobbett P., Smithson K. G. and Hatton G. I. (1986) Immunoreactivity to vasopressin- but not oxytocin-associated neurophysin antiserum in phasic neurons of rat hypothalamic paraventricular nucleus. Brain Res. 362, 7-16.

8. Craig A.'M., Blackstone C. D., Huganir R. L.'and Banker G. (1993) The distribution of glutamate receptors in cultured rat hippocampal neurons: postsynaptic clustering of AMPA-seleetive subunits. Neuron 10, 1055-1068.

9. Dierickx K. and Vandesande F. (1977) Immunocytochemical localization of the vasopressinergic and the oxytocinergic neurons in the human hypothalamus. Cell Tiss. Res. 184, 15-27.

NSC 65/2~J


10. Dierickx K. and Vandesande F. (1979) Immunocytochemical demonstration of separate vasopressin-neurophysin and oxytocin-neurophysin neurons in the human hypothalamus. Cell. Tiss. Res. 196, 203-212.

11. Gasic G. P. and Hollmann M. (1992) Molecular neurobiology of glutamate receptors. A. Rev. PhysioL 54, 507-536.

12. Ginsberg S. D., Martin L. J., Blackstone C. D., Huganir R. L. and Price D. L. (1993) Non-NMDA glutamate receptor (GluR) subtypes in monkey hypothalamus are selectively distributed. Proc. Soc. Neurosci. 19, 918.

13. Ginsberg S. D., Price D. L., Blackstone C. D., Huganir R. L. and Martin L. J. (1994) Non-NMDA glutamate receptors are present throughout the primate hypothalamus. J. comp. Neurol. (in press).

14. Ginsberg S. D., Hof P. R., Young W. G. and Morrison J. H. (1994) Noradrenergic innervation of vasopressin- and oxytocin-containing neurons in the hypothalamic paraventricular nucleus of the macaque monkey: quantitative analysis using double-label immunohistochemistry and confocal laser microscopy. J. comp. Neurol. 341, 476-491.

15. Gribkoff V. K. and Dudek F. E. (1988) The effects of the excitatory amino acid antagonist kynurenic acid on synaptic transmission to supraoptic neuroendocrine cells. Brain Res. 442, 152-156.

16. Hatton G. I., Perlmutter L. S., Salm A. K. and Tweedle C. D. (1984) Dynamic neuronal~lial interactions in hypothalamus and pituitary: implications for control of hormone synthesis and release. Peptides 5, Suppl. 1, 121 138.

17. Hatton G. I. and Tweedle C. (1982) Magnocellular neuropeptide neurons in the hypothalamus: increases in membrane apposition and number of specialized synapses from pregnancy to lactation. Brain Res. Bull. 8, 197-204.

18. Herman J. P. and Weigand S. J. (1986) Ibotenate-induced cell death in the hypothalamic paraventricular nucleus: differential susceptibility of magnocellular and parvicellular neurons. Brain Res. 383, 367-372.

19. Hollmann M., Hartley M. and Heinemann S. (1991) Ca 2÷ permeability of KA-AMPA-gated glutamate receptor channels depends on subunit composition. Science 252, 851 853.

20. Hou-Yu A., Ehrlich P. H , Valiquette G., Engelhardt D. L., Sawyer W. H., Nilaver G. and Zimmerman E. A. (1982) A monoclonal antibody to vasopressin: preparation, characterization, and application in immunocytochemistry. J. Histochem. Cytochem. 30, 1249-1260.

21. Hou-Yu A., Lamme A. T., Zimmerman E. A. and Silverman A.-J. (1986) Comparative distribution of vasopressin and oxytocin neurons in the rat brain using a double-label procedure. Neuroendocrinology 44, 235 246.

22. Hume R. I., Dingledine R. and Heinemann S. F. (1991) Identification of a site in glutamate receptor subunits that controls calcium permeability. Science 253, 1028-1031.

23. Inou6 S. (1990) Foundations of confocal scanned imaging in light microscopy. In Handbook o f Biological Confocal Microscopy (ed. Pawley J. B.), pp. 1 14. Plenum Press, New York.

24. Jirikowski G. F., Sanna P. P. and Bloom F. E. (1990) mRNA coding for oxytocin is present in axons of the hypothalamo-neurohypophysial tract. Proc. natn. Acad. Sci. U.S.A. 87, 7400-7404.

25. Kein~/nen K., Wisden W., Sommer B., Werner P., Herb A., Verdoorn T. A., Sakmann B. and Seeburg P. H. (1990) A family of AMPA-selective glutamate receptors. Science 249, 556-560.

26. Kiyama H., Sato K. and Tohyama M. (1993) Characteristic localization of non-NMDA type glutamate receptor subunits in the rat pituitary gland. Molec. Brain Res. 19, 262-268.

27. Leontovich T. A. (1969) The neurons of the magnocellular neurosecretory nuclei of the dog's hypothalamus. A Golgi study. J. Hirnforsch. 6, 499 517.

28. Levey A. I., Bolan J. P., Rye D. B., Hallanger A. E., Demuth R. M., Mesulam M.-M. and Wainer B. H. (1986) A light and electron microscopic procedure for sequential double antigen localization using diaminobenzidine and benzidine dihydrochloride. J. Histochem. Cylochem. 34, 1449-1457.

29. Martin L. J., Blackstone C. D., Huganir R. L. and Price D. L. (1992) Cellular localization of a metabotropic glutamate receptor in rat brain. Neuron 9, 259-270.

30. Martin L. J., Blackstone C. D., Huganir R. L. and Price D. L. (1993) The striatal mosaic in primates: striosomes and matrix are differentially enriched in ionotropic glutamate receptor subunits. J. Neurosci. 13, 782 792.

31. Martin L. J., Blackstone C. D., Levey A. I., Huganir R. L. and Price D. L. (1993) AMPA glutamate receptor subunits are differentially distributed in rat brain. Neuroscience 53, 327-358.

32. Martin L. J., Blackstone C. D., Levey A. I., Huganir R. L. and Price D, L. (1993) Cellular localizations of AMPA glutamate receptors within the basal forebrain magnocellular complex of rat and monkey. J. Neurosci. 13, 2249-2263.

33. McNamara J. O., Eubanks J. H., McPherson J. D., Wasmuth J. J., Evans G. A. and Heinemann S. F. (1992) Chromosomal localization of human glutamate receptor genes. J. Neurosci. 12, 2555-2562.

34. Meeker R. B., Swanson D. J., Greenwood R. S. and Hayward J. N. (1993) Quantitative mapping of glutamate presynaptic terminals in the supraoptic nucleus and surrounding hypothalamus. Brain Res. 600, 112-122.

35. Meeker R. B., Swanson D. J. and Hayward J. N. (1989) Light and electron microscopic localization of glutamate immunoreactivity in the supraoptic nucleus of the rat hypothalamus. Neuroscience 33, 157-167.

36. Meister B., Villar M. J., Ceccatelli S. and H6kfelt T. (1990) Localization of chemical messengers in magnocellular neurons of the hypothalamic supraoptic and paraventricular nuclei: an immunohistochemical study using experimental manipulations. Neuroscience 37, 603-633.

37. Monaghan D. T., Bridges R. J. and Cotman C. W. (1989) The excitatory amino acid receptors: their classes, pharmacology, and distinct properties in the function of the central nervous system. A. Rev. Pharmac. Toxic. 29, 365-402.

38. Nakanishi S. (1992) Molecular diversity of glutamate receptors and implications for brain function. Science 258, 597-603.

39. Parker S. L. and Crowley W. R. (1993) Stimulation of oxytocin release in the lactating rat by central excitatory amino acid mechanisms: evidence for specific involvement of R,S-ct-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid- sensitive glutamate receptors. Endocrinology 133, 2847 2854.

40. Petralia R. S. and Wenthold R. J. (1992) Light and electron immunocytochemical localization of AMPA-selective glutamate receptors in the rat brain, J. comp. Neurol. 318, 329-354.

41. Petralia R. S., Yokotani N. and 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.

42. Piekut D. T. (1983) Ultrastructural characteristics of vasopressin-containing neurons in the paraventricular nucleus of the hypothalamus. Cell Tiss. Res. 234, 125-134.

GluR3 in oxytocinergic neurons 575

43. Piekut D. T. (1985) Ultrastructural studies of the afferent synaptic input to oxytocin-containing neurons in the paraventricular nucleus of the hypothalamus. J. ultrastruct. Res. 90, 71-79.

44. Poulain D. A. and Wakerley J. B. (1982) Electrophysiology of hypothalamic magnocellular neurons secreting oxytocin and vasopressin. Neuroscience 7, 773-808.

45. Puckett C., Gomez C. M., Korenberg J. R., Tung H., Meier T. J., Chen X. N. and Hood L. (1991) Molecular cloning and chromosomal localization of one of the human glutamate receptor genes. Proc. nam. Acad. Sci. U.S.A. 88, 7557-7561.

46. Rafols J. A., Aronin N. and DiFiglia M. (1987) A Golgi study of the monkey paraventricular nucleus: neuronal types, afferent and efferent fibers. J. comp. Neurol. 257, 5954513.

47. Rogers S. W., Andrews P. I., Gahring L. C., Whisenand T., Cauley K., Crain B., Hughes T. E., Heinemann S. F. and McNamara J. O. (1994) Autoantibodies to glutamate receptor GluR3 in Rasmussen's encephalitis. Science 265, 6484551.

48. Sato K., Kiyama H. and Tohyama M. (1993) The differential expression patterns of messengers RNAs encoding non-N-methyl-D-aspartate glutamate receptor subunits (GIuR1-4) in the rat brain. Neuroscience 52, 515-539.

49. Seeburg P. H. (1993) The molecular biology of mammalian glutamate receptor channels. Trends Neurosci. 16, 359 365.

50. Sekiguchi M., Doi K., Zhu W,-S., Watase K., Yokotani N., Wada K. and Wenthold R. J. (1994) A deletion in the second cytoplasmic loop of GIuR3 produces a dominant negative mutant of ct-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor. J. biol. Chem. 269, 14,559-14,565.

51. Silverman A.-J., Hou-Yu A. and Zimmerman E. A. (1983) Ultrastructural studies of vasopressin neurons of the paraventricular nucleus of the hypothalamus using a monoclonal antibody to vasopressin: analysis of synaptic input. Neuroscience 9, 141 155.

52. Sofroniew M. V. and Glasmann W. (1981) Golgi-like immunoperoxidase staining of hypothalamic magnocellular neurons that contain vasopressin, oxytocin or neurophysin in the rat. Neuroscienee 6, 6194543.

53. Swanson L. W. and Kuypers H. G. J. M. (1980) The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J. comp. Neurol. 194, 555-570.

54. Swanson L, W. and Sawchenko P. E. (1983) Hypothalamic integration: organization of the paraventricular and supraoptic nuclei. A. Rev. Neurosci. 6, 269 324.

55. Theodosis D. T. and Poulain D. A. (1984) Evidence that oxytocin-secreting neurons are involved in the ultrastructural reorganization of the rat supraoptic nucleus apparent at lactation. Cell Tiss. Res. 235, 217-219.

56. Theodosis D. T., Poulain D. A. and Vincent J.-D. (1981) Possible morphological bases for synchronization of neuronal firing in the rat supraoptic nucleus during lactation. Neuroscience 6, 919 929.

57. van den Pol A. N. (1991) Glutamate and aspartate immunoreactivity in hypothalamic presynaptic axons. J. Neurosci. I1, 2087-2101.

58. van den Pol A. N., Hermans-Borgmeyer I., Hofer M., Ghosh P. and Heinemann S. (1994) Ionotropic glutamate- receptor gene expression in hypothalamus: localization of AMPA, kainate, and NMDA receptor RNA with in situ hybridization. J. comp. Neurol. 343, 428-444.

59. van den Pol A. N. and Trombley P. Q. (1993) Glutamate neurons in hypothalamus regulate excitatory transmission. J. Neurosci. 13, 2829-2836.

60. van den Pol A. N., Wuarin J.-P. and Dudek F. E. (1990) Glutamate, the dominant excitatory transmitter in neuroendocrine regulation. Science 250, 1276-1278.

61. Verdoorn T. A., Burnashev N., Monyer H., Seeburg P. H. and Sakmann B. (1991) Structural determinants of ion flow through recombinant glutamate receptor channels. Science 252, 1715 1718.

62. Wakerley J. B., Dyball R. E. J. and Lincoln D. W. (1973) Milk ejection in the rat: the result of a selective release of oxytocin. J. Endocr. 57, 557 558.

63. Wakerley J. B. and Lincoln D. W. (1973) The milk-ejection reflex of the rat: a 20- to 40-fold acceleration in the firing of paraventricular neurons during oxytocin release. J. Endocr. 57, 477-493.

64. Wuarin J.-P. and Dudek F. E. (1993) Patch-clamp analysis of spontaneous synaptic currents in supraoptic neuroendocrine cells of the rat hypothalamus. J. Neurosci. 13, 2323 2331.

65. Zhang T.-X. and Ciriello J. (1985) Kainic acid lesions of paraventricular nucleus neurons reverse the elevated arterial pressure after aortic baroreceptor denervation in the rat. Brain Res. 358, 334-338.

(Accepted 23 September 1994)

The AMPA glutamate receptor GluR3 is enriched in oxytocinergic magnocellular neurons and is localized at synapses

Documents