Immunolabeling of mu opioid receptors in the rat nucleus of the solitary tract: extrasynaptic...

THE JOURNAL OF COMPARATIVE NEUROLOGY 371:522-536 (1996)

Immunolabeling of Mu Opioid Receptors in the Rat Nucleus of the Solitary Tract:

Extrasynaptic Plasmalemmal Localization and Association With Leu5-Enkephalin

P.Y. CHENG, L.Y. LIU-CHEN, C. CHEN, AND V.M. PICKEL Department of Neurology and Neuroscience, Cornell University Medical College, New York,

New York 10021 (P.Y.C., V.M.P.); Department of Pharmacology, Temple University School of Medicine, Philadelphia, Pennsylvania 19140 (L.Y.L.-C., C.C.)

ABSTRACT Activation of the mu opioid receptor (MOR) by morphine within the caudal nucleus of the

solitary tract (NTS) is known to mediate both cardiorespiratory and gastrointestinal responses. Leu5-enkephalin (LE), a potential endogenous ligand for MOR, is also present within neurons in this region. To determine the cellular sites for the visceral effects of MOR ligands, including LE, we used immunogold-silver and immunoperoxidase methods for light and electron microscopic localization of antisera against MOR (carboxyl terminal domain) and LE in the caudal NTS of rat brain. Light microscopy of coronal sections through the NTS at the level of the area postrema showed MOR-like immunoreactivity (MOR-LI) and LE labeling in punctate processes located within the subpostremal, dorsomedial and medial subnuclei. Electron microscopy of sections through the medial NTS at this level showed gold-silver particles identifying MOR-LI prominently distributed to the cytoplasmic side of the plasma membranes of axons and terminals. MOR labeled terminals formed mostly symmetric (inhibitory-type) synapses but sometimes showed multiple asymmetric junctions, characteristic of excitatory visceral afferents. MOR-LI was also present along extrasynaptic plasma membranes of dendrites receiving afferent input from unlabeled and LE-labeled terminals. We conclude that MOR ligands, possibly including LE, can act at extrasynaptic MORs on the plasma membranes of s o n s and dendrites in the caudal NTS to modulate the presynaptic release and postsynaptic responses of neurons. These are likely to include local inhibitory neurons and both gastric and cardiorespiratory afferents known to terminate in the subnuclei with the most intense MOR-LI. D 1996 Wiley-Liss, Inc.

Indexing terms: ultrastructure, anatomy, cardiorespiratory, neurotransmitters, brainstem

The nucleus of the solitary tract (NTS) is known to receive input from vagal afferents (Higgins et al., 1984; Kalia and Mesulam, 1980; Maley et al., 1987). These afferents show a somatotopic distribution with the barore- ceptor and gastrointestinal afferents terminating in the intermediate and medial subdivisions of the caudal NTS, whereas respiratory-related afferents from the trachea and lung project to the ventrolateral subdivisions, mainly of the more rostra1 NTS (Diz et al., 1987). Thus, the localization of opioid peptides, including Met5- and Leu5-enkephalin and @-endorphin in the medial subdivisions of the caudal NTS, suggests the involvement of opiates in the modulation of cardiovascular and gastrointestinal functions (De Jong et al., 1983; Mastrianni et al., 1989; Palkovits and Eskay, 1987). The importance of the mu opioid receptor (MOR) in centrally mediated visceral responses to opiates is sup-

ported by both immunocytochemical (Arvidsson et al., 1995) and autoradiographic studies (Sales et al., 1985; Mansour et al., 1988; Mansour et al., 1994b) showing MOR binding sites within the NTS.

The cellular sites within the NTS mediating the effects of MOR ligands have not been established. Several lines of evidence suggest that MOR selective ligands in the NTS have both pre- and postsynaptic actions within the NTS. Injection of the MOR selective ligand, D-Ala2,MePhe4,Gly- 015 enkephalin (DAMGO) into the NTS has been shown to inhibit transmitter release from afferent fibers in the NTS

~

Accepted April 15, 1996. Address reprint requests to Peter Y. Cheng, Ph.D., Department of

Neurology and Neuroscience, 411 East 69 th St., Cornell University Medical College, New York, NY 10021. E-mail: [email protected]

O 1996 WILEY-LISS, INC.

MU OPIOID RECEPTOR IMMUNOLABELING IN THE NTS 523

(Rhim et al., 1993), thus suggesting presynaptic actions. Sectioning of the vagus nerve, accompanied by nodose ganglion excision, which effectively removes a major source of visceral afferents, produces marked reduction of MOR in autoradiographic binding within the NTS (Dashwood et al., 19881, which provides further support for presynaptic localization of MOR within the NTS. MOR selective opioid ligands have also been shown, by using patch-clamp record- ings of neurons located in the NTS of brainstem slices, to produce a K+ dependent hyperpolarization of most neurons (Rhim et al., 19931, which suggests a postsynaptic effect.

The endogenous ligands for MOR within the NTS also have not been clearly established, but the data from several studies suggest that these may include the enkephalins (Magnan et al., 1982). Injection of the MOR-selective opioid agonist DAMGO has been shown to produce cardiovascular pressor responses and tachycardia as well as respiratory depressant effects at higher doses (Hassen et al., 1982). Enkephalin-immunoreactivity has been localized by light microscopy in perikarya and processes distributed in the dorsomedial, intermediate, and parasolitary subdivisions of the caudal NTS at the level of the area postrema (Velley et al., 1991). We have also shown that in this region, enkephalin immunoreactivity is present in perikarya, dendrites, and axon terminals (Pickel et al., 1989). Within the axon terminals, enkephalin immunoreactivity is frequently found within dense-core vesicles, not associated with the synaptic junction (Maley, 1985; Pickel et al., 1989; Velley et al., 1991). These peptides may be released by exocytosis from the large dense-core vesicles and diffuse within the extracellular space to reach and to activate extrasynaptic MORs (Schmitt, 1984; Herkenham, 1987).

To determine the cellular sites for MOR-mediated effects of opiates in the NTS, we used immunogold-silver and/or immunoperoxidase methods to determine the ultrastructural localization of an antibody generated to 16 amino acids within the carboxyl terminal domain of the cloned MOR (Wanget al., 1993; Chen et al., 1993; Thompson et al., 1993). The chosen peptide sequence has little homology to the delta opioid receptor (DOR; Evans et al., 1992; Kieffer et al., 1994) or the kappa opioid receptor (KOR; Yasuda et al., 1993; Zhu et al., 1995). Additionally, we combined immunogold-silver labeling of the MOR and immunoperoxidase labeling of Leu5-enkephalin (LE) to determine the distribution of MOR relative to LE. MOR was localized to extrasynaptic sites along plasma membranes of unmyelinated axons, axon terminals and dendrites. We also observed MOR-labeled profiles in proximity to LE-containing profiles which were mainly axonal. These results 1) implicate MOR in the extrasynaptic modulation of neurons and 2) suggest that LE is at least one of the endogenous ligands for the MOR in the NTS.

MATERIALS AND METHODS Tissue preparation

Detailed protocols for fixation and immunocytochemical labeling have been published (Sesack et al., 1994; Leranth and Pickel, 1989). For light and electron microscopy, four adult male Sprague-Dawley rats (200-300 g, Taconic Farms) were deeply anesthetized with sodium pentobarbital (100 mglkg i.p.) and perfused through the ascending aorta with 10 ml of heparin saline (1,000 U/ml) followed first by 50 ml of 3.75% acrolein (Polysciences) in 2% paraformaldehyde, then followed by 200 ml of 2% paraformaldehyde. The

caudal brainstems of these animals were removed and postfixed for 30 minutes in 2% paraformaldehyde. Fixatives were prepared in 0.1 M phosphate buffer (PB) at a pH of 7.4. Coronal sections (30-50 pm in thickness) through the caudal brainstem were cut by using a Lancer Vibratome and collected in an ice-cold solution of 0.1 M PB. Tissue sections were then transferred to a 1% sodium borohydride in 0.1 M PB solution for 30 minutes to neutralize alde- hydes. Tissue sections were freeze-thawed to enhance the antibody penetration, and subsequently placed in 0.5% bovine serum albumin (BSA) for another 30 minutes to reduce nonspecific labeling. The sections were then processed for immunocytochemistry .

Generation and characterization of MOR antiserum

Rabbit antiserum was generated against a synthetic peptide of the cloned rat MOR, composed of the amino acids CTN HQL ENL EAE TAP LP, representing a partial sequence within the cytoplasmic carboxyl terminal domain. In brief, the peptide was conjugated to keyhole limpet hemocyanin (KLH) or BSA using m-maleimidobenzoic acid N-hydroxysuccinimide ester (MBS), a heterobifunctional reagent linking the thiol group of cysteine on the peptide and the free amino groups of the protein carrier. Primary injection was done with the KLH conjugate, whereas booster injections were done with the BSA conjugate. The antiserum was able to immunoprecipitate solubilized [3H] p-funaltrexamine (P-FNA) labeled MOR or [3H]di- prenorphine bound MOR. When incubated with 1 p1 of this antiserum, 50% of 1 pmole of MOR in 200 ~1 was precipi- tated.

Immunodot-blot analysis (Larsson, 1981) was used in part to show the specificity of the rabbit antiserum against the MOR peptide. The MOR antiserum was characterized for cross-reactivity against unconjugated peptides corresponding to amino acid segments of the cloned rat MOR. These peptides included the N-terminal domain (CTS DCS DPL AQA S), the second extracellular loop (CRQ GSI DCT LTF SHP), and another segment of the carboxyl terminal domain (CQN TRE HPS TAN TVD RT). As a specificity control, we also examined the MOR antiserum for cross- reactivity with amino acids expressed in the cloned DOR and KOR. For DOR, peptide fragments corresponding to amino acids from the third cytoplasmic loop (CSG SKE KDK SLR RIT), the second cytoplasmic loop (CHP VKA LDF RTP AKA) and the N-terminal domain (CAS GSP GAR SAS SLA) of the cloned mouse DOR were tested. For KOR, peptide fragments from the carboxyl terminal domain (RNV DGV NKP V) and the N-terminal domain (QGP AQP ASE LPA R) of the cloned guinea pig KOR were tested. Peptides were blotted onto membranes, and the membranes were further processed by combining an enhanced chemiluminescent method (Dupont) and a previously published dot-blot procedure (Larsson, 1981).

Single immunoperoxidase and immunogold labeling for the MOR

Free-floating sections through the brainstem were incubated for 48 hours at 4°C with normal primary or pread- sorbed control antiserum against the MOR peptide. A 1: 10,0000 (immunoperoxidase) or 15,000 (immunogold) dilution of the rabbit MOR antiserum was prepared in 0.1 M Tris-saline (TS) containing 0.1% BSA. These sections were processed for immunoperoxidase labeling using the

524 P.Y. CHENG ET AL.

avidin-biotin Elite kit and conventional methods (Hsu, 1981), as elaborated below for the localization of LE. The immunogold-silver method was adapted from a previously described protocol (Chan et al., 1990). In brief, after removal from the antiserum, the brainstem sections were rinsed in 0.01 phosphate buffered-saline (PBS), blocked in 0.8% BSA and 0.1% gelatin in PBS, and incubated for 2 hours in 1:50 goat anti-rabbit IgG conjugated to 1 nm gold particles (Amersham). After rinsing in PBS and 0.2 M sodium citrate buffer (pH 7.4), the bound gold was visualized by deposition of silver using an enhancement kit (IntenSE, Amersham). Since the size of gold-silver deposits was dependent on the duration of the incubation, the time to enlarge the gold was empirically determined (6-7 minutes). All incubations and washes between each step were carried out in 0.1 M TS, at room temperature with continuous agitation.

Dual immunocytochemical labeling for the MOR and Leu5-enkephalin

Immunogold labeling of the rabbit anti-MOR antiserum was combined with immunoperoxidase labeling of mouse monoclonal anti-LE (Sera Labs; Chan et al., 1990). After initial preparation of tissue as described above for single labeling, the Vibratome sections were incubated for 48 hours in a 1:5,000 dilution of rabbit anti-MOR and a 1:200 dilution of mouse monoclonal anti-LE antiserum. These sections were then washed in buffer to remove excess antiserum and processed for immunoperoxidase labeling of the mouse LE antibody and immunogold labeling of the rabbit MOR antibody. To identify the bound mouse immunoglobulins, the sections were placed for 30 minutes each in 1) biotinylated horse anti-mouse IgG (1:400, Vector) in 0.1 % BSA, and 2) peroxidase-avidin-biotin complex (ABC, Elite Kit, Vector). The peroxidase was visualized by a 6-minute reaction with 3,3’-diaminobenzidine (22 mg/ 100 ml) and 0.01 % hydrogen peroxide. All incubations and washes between each step were carried out in 0.1 M TS, at room temperature with continuous agitation. Subsequently, to identify the bound rabbit immunoglobulins, the sections were rinsed and processed as described above for gold-silver immunolabeling of MOR.

For light microscopy, the sections of tissue were mounted on 1% gelatin-coated glass slides and examined using a Nikon microscope equipped with bright field or differential interference contrast optics. For electron microscopy, sections were processed using conventional methods (Sesack et al., 1994). Briefly, the immunolabeled Vibratome sections were postfixed for 1 hour in 2.0% osmium tetroxide in 0.1 M PB, dehydrated, and embedded in Epon. Ultrathin sections through the NTS, at the level of the area postrema, were cut with a diamond knife. The ribbons of thin sections were collected on grids and counter stained with uranyl acetate and Reynold’s lead citrate (Reynolds, 1963). These were examined with a Philip’s 201 electron microscope.

Data analysis Vibratome sections through the NTS at the level of the

area postrema from four rats showing the most optimal preservation of ultrastructure and immunocytochemical labeling were examined for detailed qualitative and quantitative analysis of immunoreactive profiles. Immunogold- silver particles (MOR-LI) and immunoperoxidase (LE-LI) were easily distinguishable by electron microscopy. Thin sections that contained intense clustering of gold-silver particles or immunoperoxidase product were further ana-

Fig. 1. Immunodot-blot analysis showing specificity of the MOR antiserum. Serial dilutions of the unconjugated mu opioid receptor (MOR) carboxyl terminal domain peptide fragment were blotted onto nitrocellulose paper and incubated with a 1:1,500 dilution of rabbit antiserum raised to MOR 381-398 peptide fragment. Dense immunore- action product is only seen against the immunizing MOR peptide at concentrations of 500,1,000, and 2,000 ngidot.

lyzed, by using unbiased cell counting methods, to determine the relative distribution of MOR and LE in axons and dendrites. Neuronal profiles were considered to be positive for MOR when the number of gold-silver deposits were at least three times higher then that seen in a comparable region of surrounding neuropil or selectively localized to the plasma membrane or vesicular membrane. An excep- tion was made for unmyelinated axons. Because of their small size, when even a single gold particle was selectively localized to the plasma membrane, and little spurious gold particles were seen within the same neuropil, axons were considered to be selectively labeled. Whenever possible, the labeling was confirmed in adjacent sections.

In regions near the surface of the tissue showing optimal immunogold-silver and immunoperoxidase labeling, the relationship between MOR and LE was quantitatively assessed. Electron micrographs of sections of tissue showing MOR-LI were first classified with regard to cellular and subcellular localization of the label in neuronal (axons, dendrites or soma) or glial profiles using nomenclature by Peters et al. (1991). Subsequently, the MOR containing profiles were further examined with respect to their co- existence, appositional or synaptic associations with LE- labeled and unlabeled structures. MOR-labeled terminals making contact with, or in apposition to profiles containing LE, were classified according to the presence or absence of a synaptic specialization and if present, the type of junction, whether thin (symmetric) or thick (asymmetric). If a vesicle-containing profile was in contact with another profile but no densities were observed, the two profiles were considered to be nonsynaptically associated.

RESULTS Specificity of antiserum

By immunodot-blot analysis, the rabbit antiserum against MOR (at 1: 1,500) specifically recognized the MOR peptide with a threshold of 500 ng (Fig. 1). No cross-reaction was detected with other nonimmunizing peptides from adjacent positions on the MOR or from segments of the cloned DOR or KOR (not shown). Additionally, preadsorption of the MOR antiserum with 1 or 10 pg/ml of the cognate peptide selectively abolished the immunoreactivity seen in coronal sections through the NTS processed for immunocytochemistry (Fig. 2).

Light microscopic distribution of MOR-LI Immunoperoxidase product for MOR-like immunoreactiv-

ity (MOR-LI) was heterogeneously distributed in punctate


Fig. 2. Light micrographs of coronal sections through the nucleus of the solitary tract (NTS) at the level of the area postrema. A: The immunoperoxidase labeling (black) for mu opioid receptor-like immunoreactivity (MOR-LI) is found most intensely distributed in processes within the subpostremal area (sp), the dorsomedial nucleus (d) as well as within the solitary tract (TR). B: A comparable section through the

NTS processed for immunocytochemical labeling using a MOR- antiserum after preadsorption with the cognate peptide shows absence of immunoreactive processes. Abbreviations: ap, area postrema; ng, gracile nucleus; X, dorsal nucleus of the vagus. Corner arrows point dorsal (D) and medial (MI. Scale bars = 0.26 mm.

processes throughout the medial NTS. The highest density of labeled varicose processes was seen in the subpostrema region (Fig. 2A). The dorsomedial nucleus bridging the region between the solitary tract and lateral border of the area postrema also contained abundant, intensely labeled processes. The more ventral and medial nuclei showed immunolabeling in processes whose densities appeared comparable to those seen in the adjacent motor nucleus of the vagus. MOR labeling was also seen in round profiles among bundles of axons in coronal sections through the solitary tract (Fig. 2A).

Subcellular localization of MOR-LI in unmyelinated axons and axon terminals

Electron microscopy of the medial NTS at the level of the area postrema showed that immunogold-silver particles indicative of MOR-LI were extensively distributed to the cytoplasmic surfaces of plasma membranes in unmyelinated axons (0.1-1.5 km) (Fig. 3) and axon terminals (> 1.5

pm) (Figs. 3A,B, 5, 7B, 8B). The gold-silver particles were most often localized to portions of the plasma membranes apposed to other axons, and axon terminals (Fig. 3). The remaining gold silver deposits in axon terminals were largely distributed either along plasma membranes apposed to astrocytes and/or dendrites. Occasionally, MOR-LI was also observed within the axonal cytoplasm, away from the plasma membrane. These cytoplasmic particles most often were in contact with vesicles or smooth endoplasmic reticulum (Figs. 3C, 5B, 7B, 8B).

The small unmyelinated axons accounted for 14% (14/ 99) of the profiles containing MOR-LI. Small axons (0.1-0.2 pm) were identified based on their regular contour and their location in bundles, seen in transverse sections (Pe- ters et al., 1991). These axons also occasionally contained small clear vesicles. Within a bundle of small unmyelinated axons, only a select few showed MOR-LI (Fig. 3). This distribution in bundles of labeled and unlabeled axons is reflected by the high frequency with which the MOR-


Fig. 3. Electron micrographs showing immunogold-silver labeling for MOR (black granules) and immunoperoxidase for Leu5-enkephalin (LE) (black precipitate) in axons. A MOR labeling (small arrows) is extensively localized to the plasma membrane of axons (MORa). Other dendrites (Ud) and axon terminals (Ut) within the same field are unlabeled. Immunoperoxidase product (p) for LE can be seen in a dendrite (LEd) and in a small axon (LEt). Open arrow shows a synaptic junction formed by a MOR-labeled axon terminal (MORt). B: MOR labeling (small arrows) is localized to the plasma membrane of small unmyelinated axons (MORa) in a field which also contains MOR-LI along the plasma membrane of a dendrite (MORd). This dendrite is ensheathed by an unlabeled glial process (*). Other small unmyelinated

axons (Ua) and dendrite (Ud) within the same field are unlabeled. C: An axon terminal (MORt), which has the morphological features of a primary afferent, exhibits MOR-LI (small arrows) prominently localized to the plasma membrane, but also present near smaI1 vesicles (v) within the cytoplasm. Puncta adhaerentia (arrowheads) marks the synaptic interface between the MORt and adjacent processes. MORt is also apposed to an unlabeled axon terminal (Ut), as well as an unlabeled glia (Ug, *), whose processes surround the MORt. Within the same field, smaller axons (MORa) also show MOR-LI localized to the plasma membrane. One of these is apposed to unmyelinated axom showing LE-like immunoreactivity (LE-LI) (LEa). Scale bars = 0.5 pm.


labeled axons were apposed to unlabeled axons (average of one labeled axon per five unlabeled axons). Additionally, a few MOR-labeled unmyelinated axons were apposed to unlabeled axon terminals or to dendrites (1114).

Fifty percent (50199) of profiles containing MOR-LI were considered axon terminals based on their content of synaptic vesicles and mean diameters 0.2 pm (see above) (Figs. 3A,C, 5, 7B, 8B). The MOR-labeled axon terminals were apposed with almost equal frequencies to other unlabeled axon terminals (241501, dendrites (22150) or glia (19150). Of the sections examined, 12150 of the MOR-labeled terminals showed thin and equally dense pre- and postsynaptic specializations, characteristic of symmetric synapses. These were mainly seen with unlabeled dendrites. Twenty-five percent (41 16) of the terminals containing MOR-LI (Fig. 5A,B) contacted dendrites showing postsynaptic densities, classified typically as asymmetric, based on the greater thickness of the postsynaptic as opposed to presynaptic specializations. Several of the MOR-labeled terminals were particularly large (0.5-1 pm diameter) and were observed to contact multiple targets, including unlabeled dendrites and axon terminals. These MOR-labeled terminals contained a mixed population of small clear and large dense- core vesicles (Figs. 3C, 8B). However, the majority of MOR-labeled terminals were smaller in size and less frequently showed multiple contacts.

Subcellular localization of MOR in soma and dendrites

Approximately 36% (36199) of the profiles containing MOR-LI were dendrites and less than 5% were soma. As observed in axons, the silver-intensified gold particles indi- cating MOR-LI were most often in direct contact with the cytoplasmic surfaces of the somatodendritic membrane. These were never detected along the postsynaptic density, but instead, were located near parasynaptic sites or extrasynaptic sites along plasma membranes (Figs. 3B, 4,6B, 7A, 8A). Examination of the microenvironment of dendrites showing MOR-LI revealed that only 28% (10136) either apposed or received synaptic contacts from unlabeled axon terminals. The junctional specializations on dendrites containing MOR-LI included both symmetric and asymmetric synapses. The extrasynaptic segments of the somatodendritic membrane in contact with gold particles usually opposed astrocytic processes (17/36)(Figs. 3B, 7A, 8A) or unmyelinated axons (15136). In coronal sections, smaller, presumably more distal dendrites immunolabeled for MOR were often cut in a transverse plane.

More infrequently, gold particles localized MOR to dendrites having closely apposed plasma membranes (Fig. 4). In this case, the gold-silver deposits were almost exclusively located at non-apposed segments of the plasma membrane. In this fortunate plane of section, isolated gold-silver deposits are also located in close proximity to the synaptic junctions (i.e., parasynaptic sites) formed by a single unlabeled terminal in contact with both dendrites. Gold particles were occasionally observed within the cytoplasm away from the plasma membrane in soma and large dendrites (Figs. 4, 6B). These were usually associated with vesicular organelles and saccules of smooth endoplasmic reticulum.

Localization of MOR-LI relative to LE In dually labeled sections through the NTS at the level of

the area postrema, axon, terminals, dendrites and soma contained almost exclusively either immunoperoxidase labeling for LE or immunogold silver labeling for MOR.

These processes were usually not in direct contact with each other (Figs. 5B, 6,7A).

LE was principally detected within axons and axon terminals. The reaction product was often associated with large vesicles whose diameter (80-150 nm) and distribution were similar to previously described large dense-core vesicles. However, the lumen of these vesicles were often obscured by the dense accumulation of peroxidase reaction product. Immunolabeling for LE was also seen rimming smaller vesicles in axon terminals. The peroxidase labeled large vesicles were usually located at sites distal to synaptic junctions and near glial processes that covered non- synaptic portions of the plasma membrane (Fig. 8B). In contrast, synaptic vesicles within LE-labeled terminals were more often seen near the active zone of the synapse. Nineteen percent (7137) of the dendrites containing MOR-LI were apposed to, or received synaptic contact from axon terminals immunolabeled for LE. Axon terminals presynaptic to MOR-labeled dendrites usually lacked immunoreactivity. Some of the MOR-labeled dendrites also received convergent input from other unlabeled axon terminals. In favorable planes of section, LE-labeled axon terminals formed symmetric contacts with soma and dendrites, with (Fig. 8A) or without (Figs. 6A, 7A) detectable MOR. In these terminals, the MOR-immunogold particles were located along portions of the plasmalemma distal to the synapse or associated with membranes of organelles within the cytoplasm.

In soma and dendrites containing LE-LI, the peroxidase reaction product was also largely localized to large dense- core vesicles or other vesicular organelles (Figs. 6A, 8A). Rarely, LE-labeled dendrites received synaptic contact, of the asymmetric type, from axon terminals containing MOR-LI (Fig. 8B).

DISCUSSION The overall distribution of MOR in the NTS is in

agreement with earlier studies on MOR binding sites within the NTS (Sales et al., 1985; Mansour et al., 1988, 1994b), in situ hybridization studies showing the anatomical localization of mRNA coding for this receptor (Delfs et al., 1994; Minami et al., 1994; Mansour et al., 1994a), and immunocytochemical studies. It was previously not possible to un- equivocally identify the cellular and subcellular location of MOR. The present results provide ultrastructural evidence that MOR in the caudal NTS are distributed mainly at extrasynaptic sites along plasma membranes of axons and terminals, but are also present along extrasynaptic membranes of dendrites receiving afferent input from unlabeled and LE-labeled terminals. These observations provide anatomical evidence that exogenous opiates, as well as LE may modulate the presynaptic release and postsynaptic responses of neurons in this region. The observed dense light microscopic distribution of MOR-labeled processes in subpostremal and dorsomedial NTS suggest that activation of the extrasynaptic MOR directly modulates cardiorespiratory and gastric reflex neurons in this'portion of the NTS.

Methodology The immunocytochemical methods used in this study

provide a major advance in subcellular studies of MOR. In contrast to ligand binding studies (Sales et al., 1985; Mansour et al., 1988; Mansour et al., 1994131, immunocytochemistry of MOR has greater resolution and is more compatible with dual labeling. The methods are, however,


Fig. 4. Electron micrographs showing immunogold-silver labeling for MOR in dendrites. Electron micrograph showing a transverse section of a pair of dendrites (MORdl, MORd2) that have immunogold- silver labeling of MOR (small arrows) predominantly localized to the nonapposed surfaces of plasma membrane. MORdl is synaptically

contacted by an unlabeled axon terminal (Utl). MORdl and MORd2 receive synaptic contacts (open arrows) from a single, unlabeled axon terminal (Ut2). er, endoplasmic reticulum; sp, dendritic spine. Curved arrows show the apposition between the two dendrites. Scale bar = 0.5 pm.


Fig. 5. Dendritic contacts formed by axon terminalsshowingplasma- lemmal immunogold-silver (arrows) labeling for MOR. A: The MOR- labeled terminal (MORt) forms a synaptic contact (open arrow) with an unlabeled dendrite (Ud). The gold-silver particles (small arrows) are discretely located along portions of the plasma membranejust lateral to the synaptic site (v) and along other segments of the plasma membrane apposed to other unlabeled terminals and astrocytic processes (*). Puncta adhaerentia (arrowheads) marks the synaptic interface between

the MORt and adjacent processes. Abbreviation: sv, synaptic vesicle. B: An axon terminal (MORtl) is seen closely apposed to an unlabeled axon terminal (Ut), whereas MORt2 forms an asymmetric synapse (open arrow) with an unlabeled dendrite (Ud). These terminals are separated from an axon terminal (LEO, showing immunoperoxidase reaction product for LE, by astrocytic processes (*). Small arrows point to the immunogold-silver MOR labeling. Scale bars = 0.5 pm.

highly complimentary and confirm and extend studies using ligand binding that showed an extrasynaptic localization of MOR in rat striatum (Hamel and Beaudet, 1987). The similarities in distribution of MOR-LI to autoradiogra- phy of radiolabeled ligands also support the specificity of the antiserum (see below). The results also confirm and extend our preliminary observations on the immunocytochemical localization of this MOR antiserum in the NTS (Cheng et al., 1995a).

Antiserum specificity. In the present study, we have referred to the localization of MOR as MOR-LI to acknowl- edge the possibility that the rabbit antiserum raised against the carboxyl terminal domain of the MOR may recognize other structurally similar proteins. However, we observed that the antiserum selectively recognized the peptide sequence corresponding to the carboxyl terminal domain of the MOR to which the antiserum was raised, and not other peptide sequences from the MOR, DOR or KOR. Addition- ally, we observed that within both NTS and the dorsal horn of the spinal cord, the labeling by this MOR antiserum was

similar to that seen with two other MOR antibodies (Arvids- son et al., 1995; Cheng et al., 1995b). These results, thus indicate that the protein recognized by the antiserum in the present study is MOR.

Quantification. Quantitative analysis of immunoreactive profiles in tissue labeled prior to plastic embedding can lead to erroneous results due to incomplete penetration of antibodies and/or secondary immunoreagents. The relative abundance of peroxidase and immunogold labeling may also be underestimated due to differences in penetration of peroxidase versus gold probes and to differing sensitivities of the methods (Leranth , and Pickel, 1989). Thus, in our quantification of immunoreactive profiles in tissue dually labeled for MOR and LE, care was taken to examine only the most superficial layers of the section containing both immunoreactivities. Additionally, to optimize the detection of both labels, the tissue was freeze-thawed to enhance penetration of both antibodies. However, despite these measures, it is still likely that the relative number of MOR- or LE-containing profiles were underestimated. In addi-


Fig. 6. Somatic localization of LE and MOR. A: A somata (LEs) and axon terminal (LEt) show immunoperoxidase reaction product for LE localized mainly to a dense large vesicle presumed to have a dense central core (dcv), as well as to segments of the plasma membrane (arrow heads). The LEt makes a synaptic contact (open arrow) with an unlabeled somata (Us).Within the same field, immunogold-silver labeling for MOR (arrows) is localized to the plasma membrane of a neuronal

process (MORp), in close proximity to an axon terminal (LEt). B: A somata (MORs) shows immunogold-silver labeling for MOR localized to segments of the plasma membrane (arrows), as well as to tubulovesicular organelles (tv) within the cytoplasm. Nearby, a small unmyelinated axon (LEa) shows immunoperoxidase labeling for LE, whereas other unmyelinated axons (Ua) are unlabeled. Abbreviations: G, Golgi appara- tus; nuc, nucleus. Scale bars = 0.5 pm.

Fig. 7. Axonal localization of LE and MOR. A: An axon terminal (LEt) shows immunoperoxidase product for LE, intensely localized to large vesicles having the diameters of large dense-core vesicles (dcv). This terminal forms a synaptic contact (open arrows) with a dendrite (Ud), presumed to be unlabeled even though a gold-silver particle (small arrow) is associated with smooth endoplasmic reticulum. The LEt and Ud are enclosed in the glial process (*). Small unmyelinated axons (LEa) located nearby also contain immunoperoxidase product for LE. The LEa’s are apposed to a process with MOR-LI (MORp). Within the same field is a transverse section of a dendrite (MORd) showing

immunogold-silver labeling for MOR (small arrows), localized along the plasma membrane. B: An axon terminal (MORt) shows immunogold- silver labeling for MOR that is localized to both the plasma membrane (arrows) and tubulovesicular organelles (tv). The terminal is closely apposed to an axon terminal (LEt) showing immunoperoxidase product for LE. The LEt is closely apposed to an unlabeled dendrite (Ud) and an unlabeled glia (*I. Within the same field is an axon (MORa), showing immunogold-silver labeling (arrows) for MOR, which is not contacted by LE-LI profiles. dcv, dense-core vesicles; cv, clatharin-coated vesicle; v, vesicle. Scale bars = 0.5 pm.


Fig. 8. Synaptic axodendritic associations between neurons containing MOR and LE. A: An axon terminal (LEt) with immunoperoxidase immunoreactivity for LE makes a synaptic contact (open arrow) with a dendrite (MORd). The LEt is ensheathed by an unlabeled glia (*). The MORd also receives synaptic contact (open arrow) from an unlabeled terminal (Ut). The gold particles (small arrows) are localized to both

cytoplasmic and plasmalemmal membranes. B: An axon terminal (MORt) showing immunogold-silver labeling (arrows) for MOR can be seen to make synaptic contact (open arrow) with a dendrite (LEd) showing immunoperoxidase product for LE. Within the same field is a small unmyelinated axon (LEa) showing immunoperoxidase product for LE. Abbreviation: v, vesicle. Scale bar = 0.5 pm in A 0.25 pm in B.


tion, since colchicine was not used in this study, this may account for the relatively small number of LE containing soma and dendrites. Intraventricular injections of colchicine have been used to elicit microtubule depolymerization and augmentation of perikaryal and dendritic labeling for enkephalin (Borisy and Taylor, 1967). We did not use colchicine in the present study in order to maximize the amount of LE labeling in axon terminals.

Plasmalemmal localization In the present study, most of the gold-silver particles

identifying MOR-LI were localized to the cytoplasmic side of axonal, as well as somatodendritic plasma membranes. This distribution is consistent with the proposed topogra- phy of the receptors which predicts that the amino acid sequence used for antiserum production is located within the cytoplasmic domain (Wang et al., 1993; Thompson et al., 1993; Chen et al., 1993). Similar localization was seen with the use of another antiserum to the cytoplasmic domain of the MOR (Cheng et al., 1995b1, but not seen with the use of antibodies which recognized the extracellular epitopes of DOR (Zerari et al., 1994; Cheng et al., 199%).

The extrasynaptic plasmalemmal sites of accumulation of gold-silver particles representing MOR-LI in the present study most likely represent the functional subcellular distribution of MOR. This conclusion is based on observed similarities using both inert gold particles and the more diffusible but highly sensitive immunoperoxidase reaction product (data not shown). The detection of MOR labeling only at extrasynaptic sites using both methods makes it unlikely that this distribution reflects incomplete penetration to postsynaptic zones as suggested for other receptors (Nusser et al., 1995). An extrasynaptic localization of MOR is also supported by observations that the large dense-core vesicles containing endogenous opioid peptides, unlike the amino acid storage vesicles, are never clustered near the synaptic specialization. Thus, endogenous opioid peptides are most likely released into the extracellular space to reach functional MOR, thus acting in a paracrine or parasynaptic manner (Herkenham, 1987; Schmitt, 1984).

In some cases, immunogold particles identifying MOR-LI were associated with the membranes of smooth endoplasmic reticulum and isolated vesicles within soma and dendrites, and more rarely axons within the NTS. Numerous studies have shown that membranes destined for both apical (axonal) and basolateral (dendritic) membranes of polarized epithelial cells as well as neurons are transported from the trans-Golgi network in tubulovesicular organelles similar to those showing MOR-LI labeling in the present study (Rodriguez-Boulan and Powell, 1992). The vesicles immunolabeled in the cytoplasm may also reflect internal- ized plasma membranes (Roettger et al., 1995).

Localization of MOR to unmyelinated axons and axon terminals

In the present study, MOR-LI was localized to varicose processes extensively distributed within the subpostrema and dorsomedial nuclei of the solitary tract by light microscopy and confirmed by electron microscopy as axons and terminals. The labeled axons that were seen in the subpostrema region of the present study have a distribution that is similar to that of subdiaphragmatic vagal afferents from the stomach (Leslie et al., 1982). In contrast, the labeling seen in the dorsomedial NTS in this study has certain similarities with baro- and chemosensory afferents (Bar- raco et al., 1992; Cechetto, 1987). These areas receive

prominent visceral afferent information associated with cardiorespiratory and gastrointestinal systems respectively. Inhibition of the presynaptic release of transmitters (mainly glutamate) from these afferents could thus contribute to the centrally mediated depression of respiration or gastrointestinal function by MOR agonists. This is supported by the observation that one group of MOR-labeled terminals was characterized by their large size, small round clear synaptic vesicles, presence of large dense-core vesicles, irregular outlines, and formation of multiple synaptic contacts. These morphological characteristics are similar to those already described for vagal afferents terminating in the NTS (Sumal et al., 1983; Kachidian and Pickel, 1993; Whitehead, 1993). Additionally, the release of glutamate, the excitatory transmitter found in many vagal afferents (Lewis et al., 1988), has been shown to be altered by MOR agonists (Kangrga and Randic, 1991). Glutamate containing terminals frequently form asymmetric junctions (De Biasi and Rustioni, 1988). Such terminals were infrequently MOR-labeled in the present study. Thus, it is likely that at least some of the presynaptic MOR is not restricted to visceral afferents.

The presently observed morphological heterogeneity of MOR-labeled axon terminals also suggests MOR involvement in the release of more than one neurotransmitter. Many of the MOR-labeled terminals were characterized by a smaller size, less frequent association with multiple targets, and more densely packed vesicles. Although these may be partial sections of larger multi-synaptic terminals, they are similar in size and shape to axons of previously described interneurons within the NTS (Whitehead, 1993). These MOR-labeled terminals formed mainly symmetric synapses on large to medium-sized dendrites. Symmetric synapses are more often associated with inhibitory transmitters such as gamma aminobutyric acid (GABA) (Carlin et al., 1980). We have also observed that catecholaminergic efferents from the area postrema and non-GABA containing efferents from the central amygdaloid nucleus form symmetric synapses with dendrites in the caudal NTS (Kachidian and Pickel, 1993). Opioid modulation of either local GABA neurons or afferents containing other inhibitory transmitters is strongly supported by physiological studies of opioid disinhibition (Hedner, 1983; Wang and Li, 1988). Other MOR containing terminals formed asymmetric synapses. Morphine also has been previously shown to inhibit the release of neurotransmmiters, such as serotonin, norepinephrine and GABA (Yaksh and Tyce, 1979; Mulder et al., 1987). Studies combining anterograde trans- port with immunocytochemical localization of MOR are needed to further establish the source of MOR-labeled axons in the NTS.

Extrasynaptic localization on dendrites The present localization of MOR-LI to the extrasynaptic

plasma membrane of a few soma and numerous dendrites receiving inputs from other unlabeled terminals suggests that opiates may act at these MOR either 1) to directly modulate the target neuron in a ‘parasynaptic’ fashion (Schmitt, 1984) or 2) to modulate the postsynaptic responses to other transmitters. One mechanism could involve the activation of K+ channel current (Twitchell and Rane, 1993). This would support electrophysiological findings that MOR selective agonists can hyperpolarize neurons located in the NTS of the rat by increasing K+ conductance (Rhim et al., 1993). Alternatively, recent studies also


suggest that postsynaptic MOR may be linked to Ca2+ channels within the NTS (Rhim and Miller, 1994). Either mechanism might contribute to the postulated role of opiates in direct modulation or gating the responses to other neurotransmitters. We occasionally observed that gold particles for MOR were localized to parasynaptic sites flanking asymmetric, excitatory-type junctions. However, many gold particles were not near any observed synapse, thus supporting direct, independent effects. Thus, we ex- pect that MOR-agonists may modulate viscerosensory reflexes in the NTS in a similar manner to that reported in the spinal cord where enkephalins reversibly inhibit sponta- neous, synaptically and L-glutamate-induced neuronal activity (Zieglgansberger and Tulloch, 1979). However, we also noted that many of the afferents to MOR-labeled dendrites formed inhibitory-type junctions thought to be associated with GABA containing axon terminals (Peters et al., 1991). Thus, in this context, our present results suggest that the postsynaptic activity of excitatory or inhibitory neurotransmitters on dendrites containing MOR-LI can be modulated by endogenous opioids such as LE.

A particularly close association between dendrites with MOR-LI and astrocytes or other dendrites was noted in the present study. These features have been described for catecholaminergic dendrites, some of which are known to colocalize with LE in the same region within the NTS (Pickel et al., 1989). Astrocytic processes that encapsulate either apposed dendrites or axons and dendrites may facilitate the transit of the peptide within a restricted extracellular space (Aoki and Pickel, 1992). Thus, astrocytes may play an active role in determining the extent to which extracellular peptides have access to MOR-containing neuronal processes in the caudal NTS and possibly other brain regions.

Localization of MOR relative to LE The present observations provide the first ultrastruc-

tural evidence that some of the dendrites expressing MOR in the NTS receive synaptic input from LE-containing terminals. However, since many MOR-labeled dendrites were not postsynaptic to these terminals within the sections examined, the localization of MOR along the plasma membrane does not appear to be dependent on contacts from LE-containing terminals. In addition, even when MOR-labeling was seen in targets of LE terminals, the receptor was not located at the postsynaptic junctions. In the present study, LE was also principally detected within large vesicles located at sites distant to synaptic junctions and near glial processes (see above discussion). These results suggest that the MOR-mediated effects of LE are produced in a paracrine manner, by exocytotic release from the dense-core vesicles as has been proposed for opioid and nonopioid modulators (Schmitt, 1984; Herkenham, 1987). Within the extracellular space, the peptide may diffuse to sites along the plasma membrane where access is not prevented by the junctional synaptic complex. We have recently reported a similar ultrastructural relationship between LE and MOR within the superficial layers of the spinal cord (Cheng et al., 1995b).

Occasionally, we observed LE-LI in dendrites postsynaptic to axon terminals containing MOR-LI. The localization of LE-LI in dendrites is consistent with earlier observations of peroxidase labeling for LE in neurons in the rat medial NTS (Pickel et al., 1989) and suggests feedback regulation from local intrinsic neurons. These findings also suggest

that LE is one of the endogenous ligands for MOR, and acts at the MOR, on both axons and dendrites, to modulate pre- and postsynaptic responses of neurons, respectively, within the NTS.

CONCLUSIONS The results of this study have important implications for

understanding respiratory depressant and gastroinhibitory effects of morphine. First, they indicate that one of the sites of MOR activation are extrasynaptic and may principally reflect modulation of the release of transmitters found in axons having the light microscopic distribution of function- ally distinct visceral afferents. Second, they indicate that the presynaptic effects of MOR-agonists may be opposed or augmented by modulation of the target neurons. The results of this study also indicate that LE and other endogenous opioid peptides most likely play a similar role in modulation of visceral reflexes within the caudal NTS. The findings are consistent with diffusions of both exogenous and endogenously released opioids within the extracellular spaces whose boundaries are at least partially limited by neighboring glial processes.

ACKNOWLEDGMENTS This work was supported in part by Aaron Diamond

Postdoctoral Fellowship to P.Y.C., NIDA grant DA04600, NIH HL 18974, and Career Award from NIMH MH00078 to VMP and DA04745 to L.Y.L.-C.

LITERATURE CITED Aoki, C., and V.M. Pickel (1992) C-terminal tail of beta-adrenergic receptors:

Immunocytochemical localization within astrocytes and their relation to catecholaminergic neurons in n. tractus solitarii and area postrema. Brain Res. 571.3549.

Arvidsson, U., M. Riedl, S. Chakrabarti, J.H. Lee, A.H. Nakano, R.J. Dado, H.H. Loh, P.Y. Law, M.W. Wessendorf, and R. Elde (1995) Distribution and targeting of a mu-opioid receptor (MOR1) in brain and spinal cord. J. Neurosci. 15.3328-334 1.

Barraco, R., M. el-Ridi, E. Ergene, M. Parizon, and D. Bradley (1992) An atlas of the rat subpostremal nucleus tractus solitarius. Brain Res. Bull. 29.703-765.

Borisy, G.G., and E.W. Taylor (1967) The mechanism of action of colchicine. Binding of colchicine-3H to cellular protein. J. Cell Biol. 34525-533.

Carlin, R.K., D.J. Grab, R.S. Cohen, and P. Siekevitz (1980) Isolation and characterization of postsynaptic densities from various brain regions: Enrichment of different types of postsynaptic densities. J. Cell Biol. 86.831-845.

Cechetto, D.F. (1987) Central representation of visceral function. Fed. Proc. 46.17-23.

Chan, J., C. Aoki, and V.M. Pickel (1990) Optimization of differential immunogold-silver and peroxidase labeling with maintenance of ultrastructure in brain sections before plastic embedding. J. Neurosci. Methods 33: 113-127.

Chen, Y., A. Mestek, J. Liu, J.A. Hurley, and L. Yu (1993) Molecular cloning and functional expression of a mu opioid receptor from rat brain. Mol. Pharmacol. 44.8-12.

Cheng, P.Y., L.Y. Liu-Chen, C. Chen, and V.M. Pickel (1995a) Light and electron microscopic localization of mu-opioid receptor immunoreactivity within the nucleus of the solitary tract. Analgesia 1.359-362.

Cheng, P.Y., A. Moriwaki, J.B. Wang, G.R. Uhl, and V.M. Pickel (1995b) Ultrastructural localization of mu-opioid receptor immunoreactivity and relationship to enkephalin in cervical dorsal horn of the rat spinal cord. Analgesia 1r363-366.

Cheng, P.Y., A.L. Svingos, H.W. Wang, C.L. Clarke, S. Jenab, I.W. Becz- kowska, C.E. Inturrisi, and V.M. Pickel (1995~) Ultrastructural immunolabeling shows prominent presynaptic vesicular localization of delta- opioid receptor within both enkephalin and non-enkephalin containing


axon terminals in the superficial layers of the rat cervical spinal cord. J. Neurosci. I5:5976-5988.

Dashwood, M.R., J.R. Muddle, and K.M. Spyer (1988) Opiate receptor subtypes in the nucleus tractus solitarii of the cat: The effect of vagal section. Eur. J. Pharmacol. 155r85-92.

De Biasi, S., and A. Rustioni (1988) Glutamate and substance P coexist in primary afferent terminals in the superficial laminae of spinal cord. Proc. Natl Acad. Sci. U S A . 85.7820-7824.

De Jong, W., M.A. Petty, and J.M. Sitsen (1983) Role of opioid peptides in brain mechanisms regulating blood pressure. Chest 83.306-308.

Delfs, J~M., H. Kong, A. Mestek, Y. Chen, L. Yu, T. Reisine, and M.F. Chesselet f 1994) Expression of mu opioid receptor mRNA in rat brain: An in situ hybridization study at, the single cell level. J. Comp. Neurol. 345t46-68.

Diz, D.I., K.L. Barnes, and C.M. Ferrario (1987) Functional characteristics of neuropeptides in the dorsal medulla oblongata and vagus nerve. Fed. Proc. 46.30-35.

Evans, C.J., D.E. Keith, Jr., H. Morrison, K Magendzo, and R.H. Edwards (1993) Cloning of a delta opioid receptor by functional expression. Science 258t1952-1955.

Hamel, E., and A. Beaudet (1987) Opioid receptors in rat neostriatum: Radioautographic distribution at the electron microscopic level. Brain Res. 401,239-257.

Hassen, A.H., G. Feuerstein, and A.I. Faden f 1982) Mu receptors and opioid cardiovascular effects in the NTS of rat. Peptides 3t1031-1037.

Hedner, J. (1983) Neuropharmacologxal aspects of central respiratory regulation. An experimental study in the rat. Acta Physiol. Scand. Suppl. 524:l-109.

Herkenham, M. (1987) Mismatches between neurotransmitter and receptor localizations in brain: Observations and implications. Neuroscience 23.1-38.

Higgins, G.A., G.E. Hoffman, S. Wray, and J.S. Schwaber (1984) Distribu- tion of neurotensin-immunoreactivity within baroreceptive portions of the nucleus of the tractus solitarius and the dorsal vagal nucleus of the rat. J Comp. Neurol. 226t155-164.

Hsu, S.M. (1981) Use of avidin-biotin peroxidase complex fABC) in immunoperoxidase techniques: A comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem~ Cytochem. 29577-580.

Kachidian, P., and V.M. Pickel (1993) Localization of tyrosine hydroxylase in neuronal targets and efferents of the area postrema in the nucleus tractus solitarii of the rat. J. Comp. Neurol. 329:337-353.

Kalia, M., and M.M. Mesulam (1980) Brain stem projections of sensory and motor components of the vagus complex in the cat: I. The cervical vagus and nodose ganglion. J. Comp. Neurol. 193t435-465.

Kangrga, I., and M. Randic (1991) Outflow of endogenous aspartate and glutamate from the rat spinal dorsal horn in vitro by activation of low- and high-threshold primary afferent fibers. Modulation by mu-opioids. Brain Res. 553r347-352.

Kieffer, B.L., K. Befort, C. Gaveriaux-Ruff, and C.G. Hirth (1994) The delta-opioid receptor: Isolation of a cDNA by expression cloning and pharmacological characterization. Proc. Natl. Acad. Sci. U.S.A. 9111193.

Larsson, L I. (1981) A novel immunocytochemical model system for specificity and sensitivity screening of antisera against multiple antigens. J. Histochem. Cytochem. 29t408-4 10.

Leranth, C., and V.M. Pickel (1989) Electron microscopic pre-embedding double immunostaining methods. In L. Heimer and L. Zaborszky feds): Neuroanatomical Tract-Tracing Methods 11, Recent Progress. New York: Plenum Publishing, pp. 129.172.

Leslie, R.A., D.G. Gwyn, and D.A. Hopkins (1982) The central distribution of the cervical vagus nerve and gastric afferent and efferent projections in the rat. Brain Res. 8.43

Lewis, S J., A.J. Verberne, R.J. Summers, P.M. Beart, and M. Cincotta (1988) Reduced glutamate binding in rat dorsal vagal complex after nodose ganglionectomy. Brain Res. Bull. 21t913-916.

Magnan, J., S.J. Paterson, and H.W. Kosterlitz (1982) The interaction of [met5lenkephalin and [leu5lenkephalin sequences, extended at the c-terminus, with the mu-, delta- and kappa-binding sites in the guinea- pig brain. Life Sci. 31:1359-1361.

Maley, B E. f 1985) The ultrastructural localization of enkephalin and substance P immunoreactivities in the nucleus tractus solitarii of the cat. J. Comp. Neurol. 233.490-496.

Maley, B.E., B.W. Newton, K.A. Howes, L.M. Herman, C.M. Oloff, K.C. Smith, and R.P. Elde (1987) Immunohistochemical localization of

substance P and enkephalin in the nucleus tractus solitarii of the rhesus monkey, macaca mulatta. J. Comp. Neurol. 260:483-490.

Mansour, A,, H. Khachaturian, M.E. Lewis, H. Akil, and S.J. Watson (1988) Anatomy of ens opioid receptors. Trends Neurosci. 1 It308-314.

Mansour, A., C.A. Fox, S. Burke, F. Meng, R.C. Thompson, H. Akil, and S.J. Watson f 1994a) Mu, delta, and kappa opioid receptor mRNA expression in the rat CNS: An in situ hybridization study. J. Comp. Neurol. 350.4 12438.

Mansour, A., C.A. Fox, R.C. Thompson, H. Akil, and S.J. Watson (1994b) Mu-opioid receptor mRNA expression in the rat CNS: Comparison to mu-receptor binding. Brain Res. 643:245-265.

Mastrianni, J.A., M. Palkovits, and G. Kunos (1989) Activation of brainstem endorphinergic neurons causes cardiovascular depression and facilitates baroreflex bradycardia. Neuroscience 33.559-566.

Minami, M., T. Onogi, T. Toya, Y. Katao, Y. Hosoi, K. Maekawa, S. Katsumata, K. Yabuuchi, and M. Satoh (1994) Molecular cloning and in situ hybridization histochemistry for rat mu-opioid receptor. Neurosci. Res. I8:3 15-322.

Mulder, A.H., F. Hogenboom, G. Wardeh, and A.N.M. Schoffelmeer (1987) Morphine and enkephalins potently inhibit [SHInoradrenaline release from rat brain cortex synaptosomes: Further evidence for a presynaptic localization of mu opioid receptor. J. Neurochem. 48.1043-1047.

Nusser, Z., J.D.B. Roberts, A. Baude, J.G. Richards, and P. Somogyi (1995) Relative densities of synaptic and extrasynaptic GABAa receptors on cerebellar granule cells as determined by a quantitative immunogold method J. Neurosci. 25r2948-2960.

Palkovits, M., and R.L. Eskay (19871 Distribution and possible origin of beta-endorphin and ACTH in discrete brainstem nuclei of rats. Neuropep- tides 9.123-137.

Peters, A., S.L. Palay, and H.D. Webster (1991) The fine structure of the nervous system: neurons and their supporting cells. New York: Oxford University Press.

Pickel, V.M., J. Chan, and T.A. Milner (19891 Ultrastructural basis for interactions between central opioids and catecholamines: 11. Nuclei of the solitary tracts. J. Neurosci. 9:2519-2535.

Reynolds, E.S. (1963) The use of lead citrate a t high pH as an electron- opaque stain in electron microscopy. J. Cell Biol. 17:208.

Rhim, H., S.R. Glaum, and R.J. Miller (1993) Selective opioid agonists modulate afferent transmission in the rat nucleus tractus solitarius. J. Pharmacol. Exp. Ther. 264.795-800.

Rhim, H., and R.J. Miller (1994) Opioid receptors modulate diverse types of calcium channels in the nucleus tractus solitarius of the rat. J. Neurosci. 14t7608-7615.

Rodriguez-Boulan, E., and S.K. Powell (1992) Polarity of epithelial and neuronal cells. Annu. Rev. Cell. Biol. 8.395-427.

Roettger, B.F., R.U. Rentsch, D. Pinon, E. Holicky, E. Hadac, J.M. Larkin, and L.J. Miller (1995) Dual pathways of internalization of the cholecysto- kinin receptor. J. Cell Biol. 128:1029-1041.

Sales, N., D. Riche, B.P. Roques, and M. Denavit-Saubie (1985) Localization of mu-and delta-opioid receptors in cat respiratory areas: An autoradiographic study. Brain Res. 344:382-386.

Schmitt, F.O. (1984) Molecular regulators of brain function: A new view. Neuroscience 13t991-1001.

Sesack, S.R., C. Aoki, and V.M. Pickel (1994) Ultrastructural localization of D2 receptor-like immunoreactivity in midbrain dopamine neurons and their striatal targets. J. Neurosci. 14.88-106.

Sumal, K.K., W.W. Blessing, T.H. Joh, D.J. Reis, and V.M. Pickel ( 1983) Synaptic interaction of vagal afferents and catecholaminergic neurons in the rat nucleus tractus solitarius. Brain Res. 277.3140.

Thompson, R.C., A. Mansour, H. Mil , and S.J. Watson (19931 Cloning and pharmacological characterization of a rat mu opioid receptor. Neuron 11:903-913.

Twitchell, W.A., and S.G. Rane (1993) Opioid peptide modulation of Ca2+-dependent K+ and voltage-activated Ca2+ currents in bovine adrenal chromaffin cells. Neuron 10:701-709.

Velley, L., T.A. Milner, J. Chan, S.F. Morrison, and V.M. Pickel (1991) Relationship of Met5-enkephalin-like immunoreactivity to vagal afferents and motor dendrites in the nucleus of the solitary tract: A light and electron microscopic dual labeling study. Brain Res. 550.298-312.

Wang, J.B., Y. Imai, C.M. Eppler, P. Gregor, C.E. Spivak, and G.R. Uhl (1993) Mu opiate receptor: cDNA cloning and expression. Proc. Natl. Acad. Sci. U.S.A. 90r10230-10234.

536

Wang, Q., and P. Li (1988) Inhibition of baroreflex following microinjection of GABA or morphine into the nucleus tractus solitarii in rabbits. J. Auton. New. Syst. 25:165-172.

Whitehead, M.C. (1993) Distribution of synapses on identified cell types in a gustatory subdivision of the nucleus of the solitary tract. J. Comp. Neurol. 332:32&340.

Yaksh, T.L., and G.M. Tyce (1979) Microinjection of morphine into the periaqueductal grey evokes the release of serotonin from spinal cord. Brain Res. 171:176181.

Yasuda, K., K. Raynor, H. Kong, C.D. Breder, J. Takeda, T. Reisine, and G.I.

MU OPIOID RECEPTOR IMMUNOLABELING IN THE NTS

Bell (1993) Cloning and functional comparison of kappa and delta opioid receptors from mouse brain. Proc. Natl. Acad. Sci. U S A . 90:6736-6740.

Zerari, F., D. Zouaoui, M. Gastard, E. Apartis, J. Fischer, F. Herbrrecht, A. Cupo, K. Cucumel, and M. Conrath (1994) Ultrastructural study of delta-opioid receptors in the dorsal horn of the rat spinal cord using monoclonal anti-idiotypic antibodies. J. Chem. Neuroanat. 7: 159-170.

Zhu, J., C. Chen, J.C. Xue, S. Kunapuli, J.K. De Riel, and L.Y. Liu-Chen (1995) Cloning of a human kappa opioid receptor from the brain. Life Sci. 56:PL201-7.

Zieglgansberger, W. and I.F. Tulloch (1979) The effects of methionine- and leucine-enkephalin on spinal neurones of the cat. Brain Res. 167t53-64.

Immunolabeling of mu opioid receptors in the rat nucleus of the solitary tract: extrasynaptic...

Documents

Transcript of Immunolabeling of mu opioid receptors in the rat nucleus of the solitary tract: extrasynaptic...