Lesions of the dorsal vagal complex abolish increases in meal size induced by NMDA receptor blockade

Brain Research 872 (2000) 37–43 www.elsevier.com / locate / bres

Research report

Lesions of the dorsal vagal complex abolish increases in meal size induced by NMDA receptor blockade B.R. Treece, R.C. Ritter, G.A. Burns* College of Veterinary Medicine, Department of VCAPP, Room 205 Wegner Hall, Washington State University, Pullman, WA 99164 -6520, USA Accepted 18 April 2000

Abstract Rats increase meal size and duration after intraperitoneal injection of MK-801, a non-competitive N-methyl-D-aspartate (NMDA) receptor antagonist. This effect depends upon intact vagal fibers, since the antagonist does not increase intake when visceral afferent and efferent pathways have been interrupted by bilateral subdiaphragmatic vagotomy. NMDA receptors have been demonstrated on vagal afferent fibers and on second-order neurons in the medial subnucleus of the solitary tract (NTS), the area postrema (AP), and the dorsal motor nucleus of the vagus. To determine whether neurons in these structures are crucial for NMDA receptor effects on feeding, we examined the effect of MK-801 on intake of 15% sucrose in rats with aspiration lesions of the AP and adjacent NTS. MK-801 (100 mg / kg, i.p.) significantly increased sucrose intake in these lesioned rats compared to sham-lesioned rats (32.360.1 ml versus 23.360.1 ml, P,0.001). However, when the AP/ NTS aspiration lesions were combined with bilateral electrolytic destruction of the medial NTS and the DMV, lesioned rats consumed nearly the same amount of sucrose after either saline or MK-801 (25.962.4 ml versus 24.363.0 ml; P50.687). By contrast, sham-lesioned controls ingested significantly more sucrose following MK-801 compared to saline (19.861.0 ml versus 13.160.8 ml, P,0.001). These results suggest that an intact caudomedial NTS and / or DMV are necessary for increases in intake induced by NMDA receptor blockade. While the AP might participate in MK-801-induced enhancement of intake, it is not essential for this effect.  2000 Elsevier Science B.V. All rights reserved. Theme: Neurotransmitters, modulators, transporters, and receptors Topic: Excitatory amino acid receptors: physiology, pharmacology and modulation Keywords: Satiety; Glutamate; Nucleus of the Solitary Tract; Brainstem; MK-801

1. Introduction We previously reported that rats increase their intake of preferred foods but not water, following an intraperitoneal (i.p.) injection of Dizocilpine (MK-801), a non-competitive antagonist of N-methyl-D-aspartate (NMDA)-activated ion channels [5]. When administered i.p., the antagonist does not elicit feeding, but rather increases the size and duration of the meals that occur in response to deprivation or introduction of a palatable food. Treece et al. [18] demonstrated robust increases in food intake following

*Corresponding author. Tel.: 11-509-335-7645; fax: 11-509-3354650. E-mail address: gil [email protected] (G.A. Burns) ]

nanoliter injections of MK-801 directly into the caudomedial nucleus of the solitary tract (NTS). Here, second order neurons, some of which express NMDA receptor message [1], receive vagal afferent inputs from the gut [2]. In addition, interconnections exist between NTS neurons and those in the overlying area postrema (AP). The AP is another region of the brainstem that has been implicated in the intake of food [3] and NMDA receptors also have been located in this structure. Furthermore, neurons in the AP and NTS express Fos in response to injections of the antagonist into the fourth ventricle [19]. Since small volumes of MK-801 specifically administered into the medial NTS increase meal size [18], ablation of this portion of the NTS and the overlying AP might abolish MK-801’s effects on feeding. To evaluate this possibility, we also recorded intakes of 15% sucrose in sham- and

0006-8993 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0006-8993( 00 )02432-X

38

B.R. Treece et al. / Brain Research 872 (2000) 37 – 43

AP/ NTS-lesioned rats, following systemic injection of MK-801.

2. Materials and methods

2.1. Animals Adult (350–400 g) male Sprague–Dawley rats were individually housed in a climate-controlled vivarium with ad libitum access to a standard pelleted or powdered rat chow and water, except during experiments. In addition, rat chow was not available during overnight fasts. The rats were maintained on a 12:12 light:dark schedule (lights on at 0700) and were habituated to the laboratory conditions for at least 5–7 days prior to surgery.

2.2. Drugs MK-801 (Research Biochemical International, Natic, MA) and CCK (CCK-8) (Squibb Laboratories, CA) were dissolved in isotonic saline (pH 5.3) at a concentration of 100 mg / ml and 1 mg / ml, respectively, in preparation for i.p. injection.

2.3. Surgical procedures 2.3.1. Experiment 1. Aspiration of the AP and adherent NTS In ten rats, the AP and underlying NTS were aspirated according to a previously described method [16]. Briefly, the rats were anesthetized with an intramuscular injection of ketamine (Fort Dodge) and xylazine (Amherst) at a dose of 0.1 mg / kg and 1 mg / kg, respectively. An inhaled anesthetic (Metafane) was used to maintain a surgical plane of anesthesia. The rats were positioned in a stereotaxic frame with their heads in moderate ventroflexion. The atlanto-occipital ligament was exposed in a routine fashion and incised to reveal the dorsal surface of the medulla. The dorsal aspect of the foramen magnum was enlarged slightly with rongeurs. The AP was removed by suction through a blunt 23-gauge aspiration needle. In the case of sham animals (n54), the exposed AP and NTS were touched with the aspiration needle with the vacuum turned off (Fig. 3A). Finally, the surgical incisions were closed in a routine manner. 2.3.2. Experiment 2. Aspiration of the AP and electrolytic lesioning of the NTS Aspiration of the AP was followed by making bilateral electrolytic lesions of the underlying NTS in a second group of rats (n59), using slight modifications of previously described procedures [10,7]. Briefly, the caudal brainstem was exposed for aspiration. Near the rostral border of the AP, an insulated nickel / chrome probe with a 0.25 mm uninsulated tip was inserted into the NTS, 1.0

mm lateral of the midline and 0.5 mm below the surface of the brainstem. An electrolytic lesion was created at this site with a lesion making device (Grass Medical Instruments, Quincy, MA) set at 0.2 mA for 30 s. Near the obex, a second ipsilateral lesion (same amperage and duration) was made in the NTS, 0.5 mm lateral to the midline and 0.7 mm below the surface of the brainstem. The AP was then removed by suction through a blunted 23-gauge aspiration needle (Fig. 3C). In the case of sham-lesioned rats (n59), the probe was inserted into the NTS at the same locations, but no electric current was passed through the probe (Fig. 3A). The surgical incisions were closed in a routine manner. On post-operative day 15, the rats were re-anesthetized and returned to the stereotaxic frame. Contralateral rostral and caudal NTS lesions were made with the electrolytic lesion maker set at the same amperage and duration. Sham lesions consisted of inserting the probe into the NTS without the passage of any current. The surgical wounds were closed in a routine fashion. The rats were allowed to recover from surgery for at least 2 weeks prior to the start of experimental testing.

2.4. Experimental protocol Rats were trained to ingest a solution of 15% sucrose after an overnight (16 h) fast, according to a previously described protocol [5]. All experiments were conducted at intervals of 48 h. Rat chow was removed at 1700 h the day prior to each experiment and was not returned until the completion of an experiment the following day (1000 h). At 0900 the day of the experiment, the rats were injected i.p. with either MK-801 (100 mg / kg) or saline vehicle. Fifteen min later, calibrated drinking tubes filled with 15% sucrose were presented. Intake was measured to the nearest 0.1 ml every 5 min over a 30 min feeding period. Each series of vehicle and MK-801 tests was performed four times. Immediately following each experimental trial, the test solutions were removed and replaced with a normal ration of rat chow.

2.5. Verification of lesion placement At the completion of the experiments, assessing the ability of exogenous cholecystokinin (CCK-8) to reduce food intake behaviorally tested the extent of lesioning. Attenuation of CCK-induced reduction of food intake following AP/ NTS lesions is well documented and the procedure for testing has been previously described [10,13]. Following a 16 h fast, both lesioned and sham lesioned rats received 1 mg / kg i.p. injections of CCK. Five min later, intake of sucrose was recorded at 5 min intervals over a 30 min feeding period. Sham lesioned rats reduced their intake by at least 40% following CCK. Therefore, those rats, in which CCK produced reduction of intake in excess of 40% of their intake after i.p. saline, were

B.R. Treece et al. / Brain Research 872 (2000) 37 – 43

considered incompletely lesioned (Fig. 5). Data from such animals were not further analyzed. Following behavioral testing, the rats were anesthetized and transcardially perfused with 0.1 M phosphate buffer followed by 4% formalin solution. Brains were removed and placed in a 20% sucrose cryoprotectant solution until they sank. The brains were cut into 30 mm sections on a cryostat and thaw mounted onto glass slides. The sections were stained with cresyl violet and coverslipped. The sections then were examined microscopically to evaluate the size and position of the aspiration and / or electrolytic lesions. The lesions were reconstructed based on both actual reference to landmarks in intact brains and on maps from a rat brain atlas [15].

2.6. Data analysis The results, which represent the average of four vehicle or four MK-801 trials, are all expressed as means6S.E. Differences between groups were analyzed using a combination of Students t-test and two-way repeated measures analysis of variance (ANOVA), followed by Bonferroni’s pair-wise multiple comparison tests. Data from the CCK tests are expressed as % reduction of intake calculated according to the formula: % reduction5100hIntake after NaCl 2 Intake after CCK / Intake after NaClj

3. Results Examination of the histologic preparations from the rats in Experiment 1 revealed uniform removal of the entire AP with varying amounts of damage to the underlying NTS. Damage was limited to dorsal portions of the NTS with sparing of most of the medial subnucleus of the NTS (Figs. 3, 4). In Experiment 2, the AP was missing in each of the

39

sections from lesioned rats (n59). In addition, the extent of the electrolytic lesions varied from complete bilateral destruction of the caudal NTS to extensive unilateral damage with moderate contralateral involvement (Figs. 3, 4). In these lesioned rats, moderate to severe gliosis and neuronal loss were also evident in the dorsal motor (DMV) and hypoglossal nuclei. No histologic changes were seen in any of the sections from sham-lesioned rats (Fig. 3). As shown in Fig. 1, rats from Experiment 1, in which only the AP and closely adjacent NTS were destroyed (n510), consumed significantly more 15% sucrose than sham-treated controls (n54), following saline injection (24.660.1 ml versus 14.160.1 ml, P50.001) over a 60 min feeding period. Intraperitoneal injection of MK-801 (100 mg / kg) also significantly increased 60 min sucrose intake in aspiration-lesioned rats compared to the saline condition (32.360.01 ml versus 24.660.01 ml; P50.006) (Fig. 1B). Likewise, sham-lesioned rats significantly increased their 60 min intake of sucrose after an i.p. injection of MK-801 compared to an injection of saline (23.060.03 ml versus 14.160.01 ml; P50.029) (Fig. 1A). However, the drug3surgery interaction was not significant hF(1,27)50.0768; P50.786j. Rats from Experiment 2, with AP aspiration and electrolytic lesions of the NTS (n59) ingested significantly more sucrose after i.p. saline than sham-treated controls (25.462.4 ml versus 13.160.8 ml; P,0.001) (Fig. 2A). Moreover, rats with AP/ NTS lesions did not ingest significantly more sucrose under the influence of MK-801 than saline over the 30 min period (25.962.4 ml versus 24.363.0 ml; P50.687) (Fig. 2B). However, for this experiment, the drug3surgery interaction was significant hF(1,35)552.457; P50.001j. In contrast to the lesioned rats, the sham-lesioned rats (n59) consumed significantly (P,0.001) more sucrose over the 30 min feeding period, following an i.p. injection of MK 801 (19.861.0 ml),

Fig. 1. Graph A: Cumulative intake of 15% sucrose following i.p. injection of MK-801 (100 mg / ml) or saline vehicle in sham-lesioned rats. Sham-lesioned rats (n54) increased the size and duration of their sucrose meal after MK-801 injection (s) compared to saline injection (d). Graph B: Rats with aspiration lesions of the AP and NTS (n510) consumed more sucrose than their sham-treated counterparts. In addition, AP lesioned rats increased their sucrose intake after i.p. MK-801 (s) compared to saline injection (d).

40

B.R. Treece et al. / Brain Research 872 (2000) 37 – 43

Fig. 2. Graph A: Cumulative intake of 15% sucrose over a 30 min feeding period in sham-lesioned rats (n510) given either MK-801 or saline vehicle i.p. Sham-lesioned rats consume significantly more sucrose after MK-801 injection (s) compared to saline injection (d). Graph B: AP/ NTS lesioned rats (n59) ingested more sucrose after an injection of i.p. saline (d) than sham-treated rats (see graph A). However, these lesioned rats did not increase sucrose intake after MK-801 injection (s) compared to their intakes after saline injection.

compared to an injection of saline (13.160.8 ml) (Fig. 2A). Rats with aspiration lesions of the AP and dorsal NTS and their sham-treated counterparts (Experiment 1) did not (P,0.001) exhibit attenuation of CCK-induced reduction of sucrose intake (Fig. 5). For Experiment 2, rats with AP/ NTS aspiration and moderate electrolytic lesions (n5 6) significantly attenuated their intake of 15% sucrose following i.p. CCK injection (P50.019). However, rats (n53) with AP aspiration and extensive NTS electrolytic lesions exhibited no significant difference in sucrose intake following i.p. CCK or saline (P50.855).

3.1. Discussion Lesions that destroy the AP and substantial amounts of the dorsal vagal complex attenuate increases in food intake induced by MK-801. Our results indicate that surgical aspiration of the AP and dorsal NTS caused saline-treated rats to increase their intake of 15% sucrose compared to sham-lesioned controls. However, systemic injection of MK-801 induced AP/ NTS-lesioned rats to ingest even larger quantities of sucrose (Fig. 1). Moreover, in addition to increased meal size, meal termination was delayed for nearly 60 min following treatment with MK-801. When AP/ NTS aspiration was combined with electrolytic lesions that encompassed the medial subnucleus of the NTS and the underlying DMV, MK-801 no longer increased sucrose intake compared to the vehicle condition (Fig. 2). The AP/ NTS aspiration experiment provides compelling evidence to suggest that the AP is not essential for NMDAmediated control of meal size. Work by Edwards and Ritter [11] indicates that lesioning of the AP with superficial damage to the NTS is followed by hyperphagia of palatable novel foods. Studies combining AP lesions and vagotomy indicate that AP functions are not dependent upon an intact vagus nerve [11]. Thus, hyperphagia

resulting from AP/ NTS lesions may reflect an altered response to orosensory or humoral cues, rather than vagal cues. Our results are consistent with this conclusion. Surgical aspiration of the AP and dorsolateral NTS, both of which receive vagal afferent inputs, resulted in an enhancement, rather than an attenuation of MK-801-mediated increases in meal size. The characteristic hyperphagia that accompanies lesions of the AP suggests that a qualitative modification of control of food intake by the caudal hindbrain has occurred. The exact nature of this modification is not known, although elevation of NPY-like immunoreactivity in the paraventricular nuclei has been reported in conjunction with these lesions [12]. Nevertheless, treating APlesioned rats with MK-801 resulted in increased consumption and marked delays in termination of the meal similar to the pattern observed in sham lesioned rats. Thus, the effect of MK-801 seems to be additive with the effect of AP lesions. The mechanisms, which produce MK-801induced increases in meal size, remain operative following AP lesioning (Fig. 1). This suggests that, in the absence of the AP, non-AP dependent satiety cues take longer to reach threshold, but once the new set point for satiety has been reached, termination of the meal ensues. A consistent feature of our aspiration lesions was that they spared most of the medial aspect of the NTS, including those portions rostral to the AP (Figs. 3, 4). We previously have reported that 30 nl volumes of MK-801, delivered into the medial subnucleus of the NTS, produce notable increases in 15% sucrose intake [18]. Our present data suggest that, when bilateral electrolytic lesions included these portions of the NTS in conjunction with AP/ NTS aspiration, MK-801-induced increases in sucrose intake were eliminated. This result would be consistent with a role for visceral afferent neurons in NMDA-mediated satiety. However, other recent reports are less supportive of this conclusion. Zheng et al. [19] demon-

B.R. Treece et al. / Brain Research 872 (2000) 37 – 43

Fig. 3. Photomicrographs of representative histology sections taken at the level of the AP between obex and the fourth ventricle from rats with: (A) a sham lesion; (B) an AP aspiration lesion; or (C) an AP/ NTS electrolytic lesion. Abbreviations: hypoglossal nucleus (XII), area postrema (AP), dorsal motor nucleus of vagus (DMV), nucleus of the solitary tract (NTS), central canal (cc).

strated that MK-801 does not reduce gastric distentioninduced Fos expression in the medial subnucleus of the NTS, AP, and dorsal motor nucleus (DMV). These findings indicate that MK-801 does not enhance food intake by directly antagonizing gastric mechanoreceptive signals. Covasa and co-workers [9], who reported that MK-801 does not increase intake when ingesta are retained within the stomach by an occluding pyloric cuff, have since corroborated this observation. Recent work also indicates

41

that systemic and 4th-ventricular (unpublished data) doses of MK-801 do not reverse macronutrient-induced suppression of sucrose intake by real or sham feeding rats [8]. Furthermore, early in the course of a 60 min feeding period, MK-801-induced increases in intake by capsaicintreated rats are attenuated, but later in the feeding period the effects of capsaicin and MK-801 treatment become additive [14]. This finding suggests that binding to capsaicin-insensitive fibers sufficient to provide an additive response does not occur until late in the feeding period. The authors interpret this additivity to be an indication that the relevant neural substrates are not of vagal origin. Taken together, these results do not support the assertion that MK-801 increases meal size by interfering with direct vagal afferent input to the CNS. One difficulty in interpreting the results of vagal-vagal manipulations is the close proximity of the afferent and efferent limbs of this system. For example, bilateral subdiaphragmatic vagotomy has been shown to abolish increased intake by MK-801 [6]. However, this procedure can not be used to distinguish the importance of afferent versus efferent fibers to a given effect, since vagotomy involves the interruption of both pathways. Shinozaki and co-workers [17] found that an intravenous dose of MK-801 increases spontaneous gastric motility in anesthetized rats. Our more recent study provides compelling evidence that systemic administration of the antagonist accelerates emptying of loaded and real-fed ingesta from the stomach [9]. The net result of this accelerated emptying may be an indirect reduction of gastric mechanoreceptive satiety cuesa motor effect with sensory implications. Thus, it could be argued that, in the present study, the elimination of MK801-induced feeding effects by electrolytic lesioning reflects the destruction of DMV neurons, rather than those in the NTS. If this interpretation is correct, then nearby DMV motor neurons might have been responsible for the previously reported [18] robust response to nanoliter injections of MK-801 into the medial subnucleus of the NTS. Neurons in the DMV have been shown to contain NMDA receptors [4]. Furthermore, examination of our histologic preparations does not exclude this possibility. They reveal that, in all instances, the electrolytic lesions were extensive enough to destroy neurons in the medial NTS, as well as those in the subjacent DMV (Fig. 4). However, an equally plausible scenario is that glutamatergic neurons in the NTS interact with those in the DMV to yield the observed responses to MK-801. Regardless of which scenario is correct, it is clear that lesions encompassing both the DMV and medial NTS abolish the feeding effects of MK-801 (Fig. 2B). Unfortunately, the extensive nature of our lesions precludes making further distinctions. Our histologic preparations also revealed that the larger electrolytic lesions also produced moderate damage to hypoglossal nuclei (Fig. 3C). However, the resulting performance deficits were surprisingly minimal. Total sucrose intakes at the end of the 30-min challenge ex-

42

B.R. Treece et al. / Brain Research 872 (2000) 37 – 43

Fig. 4. Representative schematic of: (A) a typical AP aspiration lesion; and (B) a typical AP/ NTS aspiration with underlying electrolytic lesion of the NTS and surrounding structures. In the larger lesion, note the damage to the MSN, DMV and HYP, which are intact in the AP aspiration only preparation.

ceeded sham intakes and persisted in reaching intakes at comparable volumes to those rats with no hypoglossal damage at all (Fig. 2). This was evident in both the saline and MK-801 injection tests. In summary, the results of this study suggest that neurons in the medial NTS and / or the DMV are crucial for increases in meal size and duration mediated by MK-801. While the AP might participate in NMDA-induced satiety, this structure is not essential for the effect. Our results add to the existing evidence indicating that NMDA receptors in

the dorsal vagal complex contribute to the control of food intake.

Acknowledgements This work was supported by a grant (DK-52849) from the National Institute of Diabetes, Digestive, Kidney Disorders.

B.R. Treece et al. / Brain Research 872 (2000) 37 – 43

43

Fig. 5. Cumulative intakes of 15% sucrose in rats with: (A) AP aspiration lesions (n510); (B) AP/ NTS aspiration and moderate electrolytic lesions (n56); or (C) AP/ NTS aspiration and extensive electrolytic lesions (n53). Response to CCK following AP/ NTS lesions of various sizes has been well described (Ritter et al., 1992). Sucrose intakes following systemic CCK-8 (1 mg / kg) were measured as a functional indication of lesion size and location.

References [1] S.A. Aicher, S. Sharma, V.M. Pickel, NMDA receptors in the nucleus tractus solitarius (NTS): ultrastructural localization and relation to vagal afferents, Soc. Neurosci. Abstr. 23 (1997) 1255. [2] R.E. Allchin, T.F. Batten, P.N. McWilliam, P.F. Vaughan, Electrical stimulation of the vagus increases extracellular glutamate recovered from the nucleus tractus solitarii of the cat by in vivo microdialysis, Exp. Physiol. 79 (2) (1994) 265–268. [3] M.L. Alwin, J.M. Horowitz, A.C. Bonham, Non-NMDA and NMDA receptors in the synaptic pathway between area postrema and nucleus tractus solitarius, Am. J. Physiol. 275 (1992) H1236– H1246. [4] D.L. Broussard, H. Li, S.M. Altschuler, Colocalization of GABA(A) and NMDA receptors within the dorsal motor nucleus of the vagus nerve (DMV) of the rat, Brain Res. 763 (1) (1997) 123–126. [5] G.A. Burns, R.C. Ritter, The non-competitive NMDA antagonist MK-801 increases food intake in rats, Pharmacol. Biochem. Behav. 56 (1997) 145–149. [6] G.A. Burns, R.C. Ritter, Visceral afferent participation in delayed satiation following NMDA receptor blockade, Physiol. Behav. 65 (2) (1998) 361–366. [7] E.R. Cosman, B.S. Nashold, P. Bedenbaugh, Stereotactic radiofrequency lesion making, Appl. Neurophysiol. 46 (1983) 160–166. [8] M. Covasa, R.C. Ritter, G.A. Burns, NMDA receptor blockade does not with macronutrient-induced feedback signals from the small intestine, Appetite Abstr. 1999, in press. [9] M. Covasa, R.C. Ritter and G.A. Burns, Gastric participation in increased food intake following NMDA receptor blockade, Appetite Abstr. 1999 in press.

[10] G.L. Edwards, E.E. Ladenheim, R.C. Ritter, Dorsomedial hindbrain participation in cholecystokinin-induced satiety, Am. J. Physiol. 251 (1986) R971–R977. [11] G.L. Edwards, R.C. Ritter, Area postrema lesions: cause of overingestion is not altered visceral nerve function, Am. J. Physiol. 251 (1986) R575–R581. [12] G.L. Edwards, B.D. White, B. He, R.G. Dean, R.J. Martin, Elevated hypothalamic neuropeptide Y levels in rats with dorsomedial hindbrain lesions, Brain Res. 755 (1) (1997) 84–90. [13] J.E. Morley, A.S. Levine, J. Kneip, M. Grace, The effect of vagotomy on the satiety effects of neuropeptides and naloxone, Life Sci. 30 (1982) 1943–1947. [14] L.M. Patterson, H. Zheng, H.R. Berthoud, Additive satiety-delaying effects of capsaicin-induced visceral deafferentation and systemic NMDA receptor blockade, Appetite Abstr. (1999) in press. [15] G. Paxinos, C. Watson, The Rat Brain in Stereotaxic Coordinates, in: 3rd Edition, Academic Press, San Diego, 1997, pp. 64–74. [16] S. Ritter, S.L. Stone, Area postrema lesions block feeding induced by systemic injections of monosodium glutamate, Physiol. Behav. 41 (1) (1987) 21–24. [17] H. Shinozaki, Y. Gotoh, M. Ishida, Selective N-methyl-D-aspartate (NMDA) antagonists increase gastric motility in the rat, Neurosci. Lett. 113 (1) (1990) 56–61. [18] B.R. Treece, M. Covasa, R.C. Ritter, G.A. Burns, Delay in meal termination follows blockade of NMDA receptors in the dorsal hindbrain, Brain Res. 810 (1998) 34–40. [19] H. Zheng, L. Kelly, L.M. Patterson, H.R. Berthoud, Effect of brainstem NMDA-receptor blockade with MK-801 on rat behavioral and Fos responses to vagal satiety signals, Am. J. Physiol. (1999) in press.