Evidence of a Functional α7-Neuronal Nicotinic Receptor Subtype Located on Motoneurons of the Dorsal Motor Nucleus of the Vagus

Materials and Methods

Compounds.

α-Chloralose, urethane, fast green dye, and (−)-nicotine hydrogen tartrate salt were purchased from Sigma (St. Louis, MO). MLA, α-bungarotoxin (α-BGTx), and strychnine were purchased from Research Biochemicals International (Natick, MA). Dexamethasone was purchased from Elkins-Sinn (Cherry Hill, NJ). Xylazine was purchased from Haver Pharmaceuticals (Shawnee, KS). The following compounds were purchased from other vendors: ketamine (Phoenix Scientific, St. Joseph, MO), buprenorphine (Reckitt and Colman Pharmaceuticals, Richmond, VA), and Nembutal (Abbott Laboratories, North Chicago, IL). All compounds were dissolved in physiological saline (final pH 7.2–7.4). Because of solubility limitations, moderate heating was required for the more concentrated dose of strychnine (1500 pmol).

Chronic Vagotomy Surgery.

Sprague-Dawley male rats weighing 260 to 300 g (Taconic Farms, Germantown, NY) were anesthetized with a ketamine-xylazine solution (90–100 mg/kg ketamine in combination with 5–10 mg/kg xylazine administered by the intraperitoneal route). Body temperature was maintained at 37 ± 1°C with a heating lamp. Using aseptic technique, a small incision was made on the ventral part of the neck. The fascia was teased away and muscle layers were retracted, thus exposing the carotid sheath. The cervical vagus nerve was cut. A 3-mm section of the nerve was removed followed by cauterization of both cut ends (to prevent regeneration). In the case of the sham control animals, the vagus nerve was separated from the carotid sheath, but the nerve was never cut. The surgical site was then rinsed with sterile saline and sutured with monofilament nylon 4.0. All rats were given buprenorphine (0.02–0.08 mg/kg i.p.), for postoperative attenuation of pain. After recovery from surgery, rats were followed for a minimum of 2 weeks (no more than 3 weeks) to allow for adequate neuronal degeneration in the DMV. This time period has been documented as an adequate time for the degeneration of vagal motoneurons and loss of choline acetyltransferase, following cervical vagotomy (Ruggiero et al., 1993). Brainstem tissue was then taken for performing either in vitro ¹²⁵I-α-BGTx autoradiography or anti-α7 immunohistochemical examination of the DMV (see below).

¹²⁵I-α-BGTx Autoradiography.

For these studies four animals underwent ipsilateral chronic vagotomy as described above, and two animals underwent sham ipsilateral vagotomy also described above. Two additional animals were used as control animals and their brains studied (i.e., no vagotomy or sham surgery was performed). Two to 3 weeks later, animals were sacrificed by rapid decapitation and brains were rapidly removed and frozen on dry ice. Serial coronal sections (20 μm) of the medulla oblongata encompassing the full length of the DMV were cut with a cryostat, thaw mounted on gelatin-coated slides, and stored at −80°C until used. For autoradiography, tissue sections were allowed to equilibrate to room temperature and then preincubated for 30 min (at room temperature) in Tris-HCl buffer (50 mM, pH 7.3) containing 0.1% bovine serum albumin (Sorenson and Chiappinelli, 1992). The sections were then incubated with ¹²⁵I-α-BGTx (116 Ci/mmol, 0.8 nM; NEN Life Sciences, Boston, MA) for 60 min at room temperature, followed by three 10 min rinses in the same buffer (but ice-cooled), and then quickly rinsed in ice-cold distilled water. Nonspecific binding of¹²⁵I-α-BGTx was determined by adding 100 μM nicotine bitartrate to the preincubation and incubation buffers, and to a control group of adjacent coronal sections. Nonspecific binding (i.e., binding of ¹²⁵I-α-BGTx that remained in the presence of 100 μM nicotine bitartrate) was typically less than 10% of total binding. The slides were then dried under a stream of air and desiccated overnight. To generate autoradiographs, slides containing labeled brain sections were exposed to Hyperfilm βmax (Amersham Pharmacia Biotech, Piscataway, NJ) for 3 to 7 days. The exposed film was developed for 2 min in Kodak D-19 developer and fixed for 4 min in Kodak rapid fixer. Film images were then digitized using the Loats Inquiry System, and figures printed from these digitized images. Brain regions were identified by light microscopic examination of adjacent coronal sections stained with neutral red to visualize the nuclear groups of the medulla. Confirmation of these nuclear groups was made with reference to the rat brain atlas ofPaxinos and Watson (1998). In describing the results, only a rough approximation of quantification of ¹²⁵I-α-BGTx has been made using descriptors such as very strong signal, moderate signal, weak signal, and background signal.

α7-Fluorescence Immunohistochemistry.

Immunofluorescent staining procedures were performed essentially as described previously (Ebert et al., 1994), with the following modifications. Adult male Sprague-Dawley rats (vagotomized and sham) were anesthetized with Nembutal (50 mg/kg) and their tissues fixed by intracardiac perfusion with freshly prepared 2% paraformaldehyde in phosphate-buffered saline (PBS), pH 7.3. The brainstem was then removed, transferred to a 30% sucrose solution in PBS, pH 7.3, and allowed to equilibrate at 4°C for at least 48 h. Brainstem sections 30-μm thick were prepared and mounted onto Superfrost Plus (Fisher Scientific; Pittsburgh, PA) microscope slides by FD Neurotechnologies, Inc. (Catonsville, MD).

For single-labeled immunofluorescent staining, an α7-nAChR-specific mouse monoclonal antibody (mAb 306) was obtained from Research Biochemicals International and used at a dilution of 1:200. Adjacent sections were incubated with a choline acetyltransferase (ChAT)-specific mouse monoclonal antibody (mAb 305) (Chemicon, Inc., Temecula, CA) at a dilution of 1:2000. Primary antibody signals were amplified using the cyanine 3 TSA-Direct kit from NEN Life Sciences according to the manufacturer's instructions.

Dual-immunofluorescent labeling of brainstem neurons was performed using a modified version of the above-described protocol. For the primary antibody incubations, the α7-nAChR-specific antibody mAb 306 was used at dilutions of 1:20 to 1:50, whereas a rabbit anti-ChAT antibody (Chemicon, Inc.) was used at dilutions of 1:200 to 1:500. No amplification was used for the dual-labeling experiments. Instead, secondary antibodies (FITC-conjugated donkey anti-mouse IgG and Texas Red-conjugated donkey anti-rabbit IgG) were obtained from Jackson/Immunoresearch (West Grove, PA) and used at a dilution of 1:200 (in PBS). All images were visualized and captured using an Olympus Fluoview confocal laser-scanning microscope.

Positively stained neurons were quantitated for the purposes of describing the effect of unilateral vagotomy and of comparing data obtained from DMV nuclei with data obtained from hypoglossal nuclei. For these purposes, only the fluorescently stained cells with a clearly identifiable dark (unstained) nucleus were included. To ensure that the same neurons were not counted more than once, neurons were counted from every sixth section in a series through the DMV (at the rostral to caudal extent of the area postrema) from three separate rat brains. In addition, neuron counting was independently performed by two different investigators, and the results were averaged.

Acute Anesthetized Rat Surgical Preparation.

Experiments were performed on male Sprague-Dawley rats weighing 270 to 360 g (Taconic Farms), and procedures used have been described in detail in another article (Ferreira et al., 2000). Briefly, before each experiment, food was withheld overnight but water was provided ad libitum. Animals were anesthetized with an α-chloralose (60 mg/ml) and urethane (800 mg/ml) solution (dissolved in 0.9% saline) in a dose of 3 ml/kg given intraperitoneally. Toe-pinch and corneal reflexes were performed to assess depth of anesthesia. Body temperature was monitored by a rectal thermometer and maintained at 37 ± 1°C with an infrared heating lamp. The trachea and carotid artery were exposed, and animals were intubated to facilitate breathing. The carotid artery was cannulated to monitor arterial blood pressure. A laparotomy was performed, and the stomach was gently pulled from the abdominal cavity for insertion of an intragastric balloon (made from the finger of a latex glove and tied to polyethylene 160 tubing). The balloon was inflated with warm water to give a baseline pressure of approximately 10 mm Hg. Dexamethasone (0.8 mg) was given subcutaneously to prevent swelling of the brain. Both the arterial and intragastric pressure lines were connected to pressure transducers. These transducers were, in turn, connected to bridge amplifiers (Analog Digital Instruments, Milford, MA), and each of these amplifiers was fed into a MacLab motherboard. Data were saved on an Apple Macintosh G3 for analysis at a later time. Rats were then placed in a stereotaxic apparatus (David Kopf, Tujunga, CA), an incision was made on the dorsal aspect of the cranium, the occipital plate was exposed, and the atlanto-occipital membrane was cut and gently pulled away. The occipital bone was removed, the dura was cut, and the cerebellum retracted to expose the area postrema and calamus scriptorius.

Following each experiment, animals were sacrificed with an overdose of pentobarbital. Brains were removed and placed in a 4% paraformaldehyde and 20% sucrose solution for at least 24 h. Brains were sectioned into 50-μm sections. These sections were stained with neutral red to visualize the nuclear groups of the medulla. The location of nuclear groups was studied in relation to microinjection sites using the atlas of Paxinos and Watson (1998). Camera lucida drawings of pipette tracts were made and analyzed using a projection microscope. Animal care and experimental procedures were performed in accordance with the National Institutes of Health guidelines and with the approval of the Animal Care and Utilization Committee of Georgetown University, Washington, DC.

Microinjection Technique.

Double-barreled pipettes were used as described previously (Ferreira et al., 2000). All compounds were dissolved in saline and the pH was brought to 7.2 to 7.3. In some experiments, fast green (1 mg/ml) was dissolved in the drug or vehicle solution to aid in viewing microinjection sites. Coordinates for the DMV ranged from 0.3 to 0.5 mm rostral to calamus scriptorius, medial-lateral 0.3–0.5 mm lateral from the midline, and dorsal-ventral 0.5–0.7 mm from the dorsal surface of the medulla. Coordinates for the medial subnucleus of the tractus solitarius (mNTS) ranged from 0.3 to 0.5 mm rostral to calamus scriptorius, medial-lateral 0.5–0.7 mm lateral from the midline, and from 0.4 to 0.6 mm from the dorsal surface of the medulla. Microinjections were all made in 60-nl volumes by observing the movement of the fluid meniscus of the pipette against a reticule. In some experiments, vehicle (saline or saline with fast green dye, 1 mg/ml) was microinjected into the DMV or mNTS. For the DMV experiments, nicotine (100 pmol/60 nl) was microinjected to elicit a response. When robust increases in intragastric pressure were elicited, antagonists (α-BGTx, MLA, and strychnine) were microinjected 15 min later. Nicotine was then microinjected 5 min following microinjections of the antagonists, to explore blockade of the response. For the mNTS, nicotine (100 pmol/60 nl) was microinjected and when robust decreases in blood pressure were elicited, antagonists were microinjected 5 min later. Nicotine was then microinjected after 5 more min. When an antagonist was able to block the response from the DMV, we tested the same doses of the compounds against nicotine in the mNTS. This was done to determine antagonist specificity for α7-containing nicotinic receptors.

Data Analysis.

Data were analyzed using the Chart Software for data analysis made for MacLab (ADI Instruments, Milford, MA). Before microinjections were performed, the lowest points of the intragastric pressure trace were obtained over a 3-min baseline period, and a single value was calculated as the mean of all of these points. This value was used as an index of gastric tone. Phasic stomach contractions (frequency of 2–5/min) were often noted in the intragastric pressure trace but this index of gastric function was not analyzed for drug effects in this study. Phasic contractions will be measured more directly in future studies by extraluminal strain gauge force transducers attached to specific areas of the stomach. After microinjections into the DMV, the maximum value in the trace was taken as the largest increase in gastric tone. The percentage of change from baseline in intragastric pressure was then calculated. For studies using α-BGTx, the percentage of the response that was blocked was calculated by comparing the post-antagonist response to the pre-antagonist response. For the recovery experiments, the response to nicotine before α-BGTx microinjection was treated as a 100% normalized response. Each subsequent response to nicotine was compared with this initial response to calculate the percentage of the response that was blocked. For blood pressure calculations, the change in mean blood pressure was taken (mm Hg). The mean blood pressure over a 3-min period was taken before microinjections. These baseline values were compared with the mean of the blood pressure trace 30 s following microinjections (which corresponded to the peak response, in this case a decrease in blood pressure) (Ferreira et al., 2000). Data appear as means (% change from baseline for intragastric pressure and change in mm Hg for blood pressure changes) ± standard error of the mean.

For the calculations of the IC₅₀ value for α-BGTx, the inhibition curve was constructed using different concentrations of the antagonist. The effects of varying doses of α-BGTx (0.01–300 pmol) to inhibit the increase in IGP elicited from the DMV with 100 pmol of nicotine were included in the calculations, and compared with the effect after microinjection of vehicle. These values were used to define an approximate IC₅₀value, for the doses of αBGTx tested in this study. The IC₅₀ value for α-BGTx was determined by nonlinear least-squares regression analysis (sigmoidal dose-response; GraphPad, San Diego, CA).

In all cases, statistical analysis was performed on both percentage of change and on raw data. The data are presented here as percentage of change from baseline (intragastric pressure) or raw changes (mm Hg, blood pressure). Paired samples t test was performed when animals served as their own controls. Independent-sample ttest was performed on data from separate control and experimental groups. Comparisons among more than two means from different groups of rats were made by analysis of variance followed by Duncan's multiple range test. Differences were considered significant atP < 0.05. All values are expressed as mean ± S.E.M.

Results

Effect of Vagotomy on ¹²⁵I-α-BGTx Binding in the DMV.

We initially performed light microscopic autoradiographic studies on the medulla oblongata of two rats using coronal sections taken at the level of the DMV. Similar findings were made in both rats, i.e., specific labeling for ¹²⁵I-α-BGTx was very strong in the DMV. In contrast, the labeling signal was weak and/or not detectable in nuclei in proximity to the DMV such as the mNTS, the area postrema, and the hypoglossal nucleus. There was a moderate signal in several ventrolateral structures, namely, the inferior olive and the lateral reticular nucleus.

Binding of ¹²⁵I-α-BGTx in coronal sections of the medulla oblongata obtained from either sham-operated animals (N = 2) or animals subjected to chronic vagotomy (one cervical vagus nerve was sectioned and the other vagus nerve was left intact) (N = 4) is compared in Fig.1. The distribution of¹²⁵I-α-BGTx binding in the sham-operated animal is similar to that found in the two control (non-operated) animals (data not shown). A very strong signal was detected over the DMV bilaterally (Fig. 1A). A moderate signal was noted for the lateral reticular nucleus and the inferior olive. Very weak or rare signals were noted in the mNTS, area postrema, and the hypoglossal nucleus. Most importantly, in all rats subjected to chronic unilateral vagotomy (N = 4), no signal for¹²⁵I-α-BGTx labeling was detectable in the ipsilateral DMV (Fig. 1B). On the contralateral side, where the vagus nerve was left intact, the DMV exhibited the usual very strong signal¹²⁵I-α-BGTx binding (Fig. 1B). Furthermore,¹²⁵I-α-BGTx binding in the inferior olive and the lateral reticular nucleus, which exhibit a moderate signal, was almost identical on both sides of the brain (Fig. 1B).

Anti-α7-Immunohistochemical Examination of the DMV of Sham and Chronically Vagotomized Rats.

To obtain more definitive proof for a postsynaptic (on the DMV cell bodies) location of the α7-subunit, we performed immunofluorescence histochemical staining of brainstem sections from rats that had been subjected to unilateral (right) cervical vagotomy 2 to 3 weeks before fixation. We observed a significant reduction (P < 0.01, N = 3) in the number of positively stained neurons in the right (vagotomized side) DMV compared with the left (nonvagotomized control side) DMV (Figs. 2A and 3A). Despite the vagotomy, approximately one-third of the total number of α7-positive neurons remained in the DMV (Fig. 3A). As a control, we also counted α7-positive cells in the hypoglossal regions from the same brainstem sections used to quantitate α7-expression in the DMV. The hypoglossal neurons serve as a useful control in this regard because they are peripherally projecting via the hypoglossal nerve and are not damaged by cervical vagotomy (Ruggiero et al., 1993). As shown in Fig. 3A, no significant differences in the number of α7-stained hypoglossal neurons were observed (P > 0.5, N = 3).

As an additional control, we stained alternate brainstem sections for expression of the cholinergic neuronal marker ChAT. As shown in Figs.2B and 3B, we observed nearly complete elimination of ChAT-labeled neurons in the DMV ipsilateral to the vagotomy, whereas the contralateral DMV contained clearly labeled neurons. No significant differences were observed in the number of ChAT-expressing neurons in either side of the hypoglossal region following unilateral cervical vagotomy (Fig. 3A). In sham-operated animals (N = 3), no left-right differences were observed in the number of DMV or hypoglossal neurons labeled for either ChAT or α7-nAChRs (data not shown).

To determine whether α7-nAChRs are expressed on cholinergic neurons in the DMV, coimmunofluorescent staining for α7-nAChR and ChAT was undertaken. Before proceeding with the double-staining experiment, however, we performed control experiments to confirm that there was no cross-reactivity of secondary antibodies. For these experiments, α7-nAChR expression was visualized using an FITC-conjugated anti-mouse secondary antibody and ChAT expression was identified via staining with a Texas Red-conjugated anti-rabbit secondary antibody. As shown in Fig. 4, α7-nAChR expression is only evident in the presence of the anti-α7-nAChR primary antibody and the FITC-conjugated secondary antibody (Fig. 4A). No α7-nAChR staining was observed when the Texas Red-conjugated anti-rabbit antibody was used (Fig. 4B), despite the fact that the anti-α7-nAChR primary antibody was present. Conversely, ChAT expression was only observed when the Texas Red-conjugated anti-rabbit secondary antibody was used following incubation with the rabbit anti-ChAT primary antibody (Fig. 4, C and D). When neither primary antibody was present, no staining was observed following incubation with both secondary antibodies (Fig. 4, E and F). Thus, these results demonstrate that the primary and secondary antibodies used for this study are specific and do not show any appreciable cross-reactivity.

Next, we performed simultaneous staining of rat brainstem sections with anti-α7 and anti-ChAT antibodies. Every DMV neuron that was clearly stained for ChAT also showed α7-staining (Fig.5). The patterns of expression for ChAT and α7-nAChRs were different within the cells, but it is clear that the same DMV neurons express both of these proteins (Fig. 5C).

Pharmacological Studies Designed to Assess Whether the α7-nAChR Subtype Resembles the Heterologously Expressed Homomeric α7-nAChR.

To explore whether the α7-nAChR is pharmacologically similar to the homomeric α7-nAChR expressed in heterologous systems (e.g., oocytes) we first characterized the full dose-response curve for the antagonist effect of α-BGTx on DMV neurons. From our previous study we determined that 100 pmol/60 nl was very close to the ED₅₀ value (89 pmol) for nicotine in evoking an increase in intragastric pressure (Ferreira et al., 2000). We also demonstrated that 100 pmol/60 nl αBGTx could produce a significant blockade of the nicotine-induced increase in IGP (Ferreira et al., 2000). We also learned in this previous study that a 15-min interval between microinjections of a 100-pmol dose of nicotine was sufficient to allow reproducibility of the nicotine response. The new data obtained in this study are the results from studying doses of 0.01 to 300 pmol/60 nl α-BGTx (Fig. 6). As can be noted from this figure, 0.01 pmol/60 nl α-BGTx exerted no effect on nicotine-induced increase in intragastric pressure, and 100 pmol/60 nl α-BGTx evoked a maximal antagonistic effect. Furthermore, analysis of data obtained with all six doses of α-BGTx tested indicated an IC₅₀ value of 0.23 pmol/60 nl for α-BGTx (Fig.6).

The blocking effect observed with 100 pmol/60 nl α-BGTx was reversible (N = 5) (Fig.7). In these studies 100 pmol/60 nl nicotine was microinjected unilaterally into the DMV. This was followed 15 min later by a microinjection of 100 pmol/60 nl α-BGTx. Nicotine, 100 pmol/60 nl, was again microinjected into the same site in the DMV 5 min after the α-BGTx. As can be seen from the summarized data of Fig.7A, a 60% block of nicotine's response occurred at this 5-min time point. Over the next 1.5 h, the α-BGTx-induced block gradually wore off (Fig. 7). A representative experiment depicting the time of peak antagonism with α-BGTx and near restoration of the nicotine-induced response 105 min later appears as Fig. 7B. As controls for this type of experiment, nicotine, 100 pmol/60 nl, was microinjected into the DMV of three additional animals, at approximately four successive 15-min intervals. Intragastric pressure responses were determined and α-BGTx was replaced with saline (Fig.7A). The initial intragastric pressure response was designated as 100% response. The intragastric pressure responses evoked by the second, third, fourth, and fifth administration of 100 pmol/60 nl nicotine into the DMV were 114 ± 14, 101 ± 9, 112 ± 13, and 104 ± 10%, respectively. These results indicate a very reproducible effect of successive doses of nicotine into the DMV, and contrast the significant changes in intragastric pressure when successive doses of nicotine are tested with α-BGTx present (Fig. 7).

α-BGTx-induced antagonism of nicotine-induced increase in intragastric pressure appeared to be specific for an α7-containing nAChR subtype. Evidence for this appears in Fig.8. Nicotine microinjected into the mNTS produced a dose-related decrease in mean arterial pressure (Ferreira et al., 2000). Although the increase in IGP elicited by nicotine microinjected into the DMV is blocked by α-BGTx, the blood pressure-lowering effect of nicotine is not (Fig. 8; Ferreira et al., 2000). As can be noted, α-BGTx in a dose (100 pmol/60 nl) that produced maximal antagonism of nicotine-evoked increases in intragastric pressure (Fig. 8A) had no effect on the nicotine-evoked decrease in mean blood pressure (Fig. 8B, N = 3).

We also tested MLA for antagonist activity against nicotine-induced increases in intragastric pressure in doses that would most assuredly block the homomeric α7-nAChR subtype, namely, 1 pmol/60 nl (N = 5) and 10 pmol/60 nl (N = 5) (Fig.8A) (Briggs and McKenna, 1996; Palma et al., 1996). As can be noted from our summarized data, neither 1 nor 10 pmol of MLA microinjected unilaterally into the DMV significantly antagonized the effect of 100 pmol of nicotine microinjected into the same site. An excessively high dose of MLA was also tested, e.g., 30 pmol/60 nl (N = 5). This dose did exert a significant antagonistic effect against nicotine's effect at the DMV (Fig. 8A). This dose was clearly not selective for the α7-subtype because it also counteracted nicotine's effect at the mNTS to lower mean blood pressure (Fig. 8B).

MLA antagonism was tested in the above-mentioned studies by retesting nicotine 5 min following microinjection of the presumed antagonist (under Materials and Methods). It is possible that MLA could “wash out” quickly and not be present in adequate concentrations at the 5-min time point. To avoid such a possibility five experiments were performed wherein 10 pmol of MLA was coadministered with nicotine (100 pmol/60 nl) as a “cocktail” (Fig. 8A). The percentage of increase in intragastric pressure with nicotine alone was 21.8 ± 2.8 and the corresponding increase with the “cocktail” was 18.5 ± 3.0 (P > 0.05). Thus, 10 pmol/60 nl MLA was still ineffective when coapplied with nicotine. In addition to these studies, we conducted two experiments whereby following an initial nicotine dose (100 pmol/60 nl, to verify DMV placement), MLA (10 pmol) was microinjected 20 and 5 min before nicotine (100 pmol). In one experiment, the percentage of increase in intragastric pressure elicited by nicotine was 19.2 versus 20.6 following MLA. In the second experiment, nicotine elicited an 18.1% increase in intragastric pressure and a 17.9% increase following the two doses of MLA. Therefore, the 10-pmol/60 nl dose of MLA was still not able to block nicotine's effects on the DMV, even after 20 min.

We attempted to block the nicotine-induced responses with two different doses of strychnine, namely, 70 (N = 4) and 1500 pmol (N = 6). Data obtained are presented in Fig. 8A and indicate that strychnine, 1500 pmol/60 nl, does counteract the effect of nicotine on the α7-nAChR subtype in the DMV. This dose was clearly acting specifically since it did not significantly counteract nicotine's effect at the mNTS to lower mean blood pressure (Fig. 8B). The 70-pmol/60 nl dose of strychnine had no effect. This finding is noteworthy because 67 pmol/60 nl has been reported to act in the NTS to abolish cardiovascular responses induced by maximal doses of glycine (Talman and Robertson, 1989). This dose of strychnine that has been shown by others to block the glycine receptor had no effect on the nicotine-evoked response; instead, a higher dose of strychnine (1500 pmol/60 nl) was required to counteract some of the nicotine-evoked effect. The brainstem sites where these drug tests were performed are depicted in Fig. 9.

Discussion

Our new findings are 1) the α7-nAChR subtype in the DMV is located postsynaptically on DMV neurons that project into the vagus nerve, and 2) the DMV α7-nAChR subtype's pharmacological characterization and recovery profile from α-BGTx blockade separates it from the homomeric nAChR α7-subtype described in expression systems (Anand et al., 1993).

Autoradiographic studies of ¹²⁵I-α-BGTx binding, combined with unilateral cervical vagotomy, clearly demonstrated that binding sites for α-BGTx disappear once DMV neurons degenerate. The demonstration that α-BGTx blocks most of the nicotine-induced increases in intragastric pressure argues strongly for a predominant influence of an α7-nAChR subtype receptor at the DMV mediating a functional response.

Ashworth-Preece et al. (1998) also performed in vitro autoradiography using ¹²⁵I-α-BGTx on coronal sections taken from the rat medulla oblongata of sham controls and of rats subjected to chronic (2-week duration) unilateral vagotomy. Their technique for “vagotomy” involved unilateral nodose ganglionectomy (in contrast, we used unilateral cervical vagotomy, where the nerve was cut distal to the nodose ganglion, with a 2-week recovery). These investigators concluded that ¹²⁵I-α-BGTx binding sites were located presynaptically on vagal afferent terminals in the mNTS. Nodose ganglionectomy removes cell bodies of afferent vagal nerves, resulting in loss of vagal terminals in the NTS (Helke et al., 1983). Nodose ganglionectomy also produces loss of cell bodies of efferent vagal neurons in the medulla oblongata because efferent vagal fibers travel through the nodose ganglia (Helke et al., 1983). Thus, using the procedure of nodose ganglionectomy, it is impossible to tell whether¹²⁵I-α-BGTx binding is presynaptic on afferent vagal nerve terminals in the mNTS, or postsynaptic on cell bodies of efferent vagal neurons, i.e., postsynaptic on DMV neurons. We used the unilateral cervical vagotomy to test the hypothesis that¹²⁵I-α-BGTx binding was postsynaptic on DMV neurons. Indeed, unilateral cervical vagotomy in our study resulted in nearly 100% loss of ¹²⁵I-α-BGTx binding in the DMV.

Additional evidence supporting the postsynaptic location of the α7-nAChR subtype was obtained with anti-α7-immunohistochemical examination of the DMV of sham and chronically vagotomized rats. By using antibodies against ChAT, we identified DMV neurons that project into the vagus nerve and presumably innervate the abdominal viscera, including the stomach. It is well accepted that the vagus nerve is cholinergic in nature. Immunocytochemistry for ChAT demonstrates that the DMV contains many cholinergic neurons (Ruggiero et al., 1993). Consistent with this notion are our own data showing complete disappearance of ChAT-expressing neurons in the DMV following unilateral cervical vagotomy. Simultaneous staining of rat brainstem sections with anti-α7 and anti-ChAT antibodies revealed that every DMV neuron stained for ChAT also showed α7-staining. Furthermore, following unilateral cervical vagotomy, no ChAT-positive neurons expressing α7-nAChRs remained in the DMV. Although roughly one-third of the α7-expressing cells were still present in the DMV following unilateral cervical vagotomy, none of these exhibited positive staining for ChAT. These remaining DMV neurons probably do not project out of the brainstem (McLean and Hopkins, 1982).

In the autoradiographic studies, the hypoglossal nucleus did not show binding of ¹²⁵I-α-BGTx, suggesting that the α7-nAChR subtype was not present. On the other hand, immunofluorescent staining of hypoglossal neurons using an anti-α7-nAChR antibody indicated the presence of α7-subunits (our data; Dominguez del Toro et al., 1994). Thus, we are left with the dilemma that ¹²⁵I-α-BGTx does not bind to hypoglossal neurons but that α7-subunit antibody fluorescence is present in the hypoglossal nucleus. Breese et al. (1997) reported α7-mRNA in the hypoglossal nucleus, showing that these neurons are probably synthesizing the α7-protein, but they also reported near background levels of α-BGTx radiolabeling. One explanation may be that α7-nAChRs produced in the soma of hypoglossal neurons do not form functional α-BGTx binding sites. Alternatively, α7-nAChRs produced in hypoglossal neurons may not be accessible for¹²⁵I-α-BGTx binding until they are transported out of the nucleus and reach nerve terminals.

The α7-nAChR subtype that is present in the DMV of the rat appears to be functionally and pharmacologically distinct from heterologously expressed homomeric α7-nAChRs and from α7-nAChRs characterized in most neural tissue. Homomeric α7-nAChRs are nearly irreversibly blocked by α-BGTx (Couturier et al., 1990). In addition, these α7-nAChRs are blocked by MLA (Ward et al., 1990) and strychnine (Seguela et al., 1993). Current responses evoked by nAChR agonists at α7-nAChRs in hippocampal neurons (Alkondon and Albuquerque, 1993;Frazier et al., 1998), in olfactory bulb neurons (Alkondon et al., 1996), and in parasympathetic ganglion neurons (Vijayaraghavan et al., 1992) behave similarly.

The two common properties that the α7-nAChR in the DMV has in common with other α7-nAChRs and α7-nAChRs is that α-BGTx and strychnine are effective antagonists of the nicotine-induced response evoked from the DMV. However, in contrast to the α7-nAChRs in the hippocampus (Frazier at al., 1998), the α7-nAChR subtype in the DMV is not irreversibly blocked by α-BGTx. The α-BGTx blockade of nicotine-induced increases in intragastric pressure had faded within 1.5 h. It should be noted that the reversibility of α-BGTx blockade in the Frazier et al. (1998) study was followed for only 35 min. Recent data reported by Cuevas and Berg (1998) and Cuevas et al. (2000) indicate that there is a type of functional α7-nAChR that is blocked by α-BGTx but in a reversible manner. Our data raise the important question of whether the α7-nAChR can be synaptically activated. Although not mentioned under Results, we did not observe any change in intragastric pressure from microinjecting α-BGTx into the DMV in a dose that significantly antagonized nicotine-induced increase in intragastric pressure. This would argue against the existence of a tonic cholinergic input to neurons of the DMV. In agreement with this finding are data of Schafer et al. (1998)using antibodies directed against the C terminus of rat vesicular acetylcholine transporter as a marker of intrinsic cholinergic innervation, which indicate little cholinergic innervation of the DMV relative to nearby brainstem structures such as facial and hypoglossal motor nuclei, and the nucleus tractus solitarius. Strychnine, an antagonist of the α7-nAChR when used in doses known to be acting selectively (Peng et al., 1994) is effective in blocking nicotine at the DMV, whereas MLA (Alkondon and Albuquerque, 1993) is not. Taken together, these findings raise the possibility that the native α7-nAChR at the DMV may exist as a heteromeric complex, comprised of the α7 with an as yet unknown nicotinic receptor subunit. However, we can not rule out post-translational changes in an α7-homomeric nAChR.

Evidence does exist that some neurons may express heteromeric α7-nAChRs (Anand et al., 1993; Pugh et al., 1995; Cuevas and Berg, 1998; Guo et al., 1998; Yu and Role, 1998a,b). Guo et al. (1998)studied the presynaptic nicotinic receptors in the lateral geniculate nucleus of the chick. Exposure to nicotinic receptor agonists resulted in increases in the frequency of glutamatergic spontaneous postsynaptic currents that were sensitive to α-BGTx. This response, like ours in the DMV, recovered from α-BGTx blockade. MLA failed to alter the response unless high non-α7-selective doses were used. In contrast to the present finding, Guo et al. (1998) found that strychnine failed to block the α-BGTx-sensitive response. However, the strychnine concentration studied may have been too low to achieve α7-selective blockade. The data of Peng et al. (1994) using an α7-nAChR homomeric expression system created with chick gene product indicate that a much higher concentration of strychnine than used byGuo et al. (1998) is required for attenuating this receptor's response. The strychnine concentration used by Guo et al. (1998) was 1 to 3 μM and was based on concentrations found to be effective for attenuating α-BGTx-sensitive currents in the hippocampus of rats (Matsubayashi et al., 1998).

In our study, we microinjected micromolar concentrations of compounds into the DMV. That is, nicotine was used as a dose of 100 pmol contained in 60 nl (1.66 mM solution). Other investigators who perform microinjection studies of nicotine into brain tissue of rats use similar high concentrations (Nagata and Osumi, 1991; Tseng et al., 1993). Nicotine will be diluted once it diffuses and equilibrates in the brain tissue. According to Fu et al. (1999), a 1 mM concentration of nicotine in the microdialysate would yield a tissue concentration of 17.6 μM. These data indicate a dilution factor of 57-fold, and extrapolating back to our data using 1.66 mM nicotine, our dose of 100 pmol/60 nl would be anticipated to result in a tissue concentration of approximately 27 μM. Compared with the data of Fu et al. (1999), 27 μM is in the range of concentrations that induces a significant response in their experimental system.

We are not the first to suggest that the α7-nAChR subtype in the DMV has different properties from α7-homomeric nAChRs. This MLA-sensitive current evoked in the brain slice of 3- to 9-day-old rats had a relatively long time-to-peak (Zaninetti et al., 1999). These investigators were able to demonstrate that MLA was an antagonist for currents evoked by acetylcholine. No clear explanation exists for why we failed to observe an antagonist effect of MLA in our test system. It is possible that the α7-nAChR subtype in 3- to 9-day-old rats may differ from that found in adult rats. There are data indicating changes in α7-subunit immunoreactivity between postnatal day 3 and postnatal day 15 in rats (Dominguez del Toro et al., 1997). Thus, α7-nAChR subunit expression is developmentally regulated. It is possible that sensitivity of the α7-nAChR subtype to MLA varies with the stages of development. Zaninetti et al. (1999) also reported that some DMV neurons possess a non-α7-containing nAChR. This finding fits the data we obtained, indicating that α-BGTx does not fully block nicotine-induced increases in intragastric pressure (Fig. 7A).