INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MICROINJECTIONS OF L-GLUTAMATE (L-Glu) into the nucleus tractus solitarii (NTS) of conscious rats elicits decreased heart rate (HR) and increased arterial pressure, which could be mediated by sites to which chemoreceptor afferents project in NTS (6-8, 13). The pressor response is abolished by systemic administration of the -adrenoreceptor antagonist prazosin, and the bradycardia is abolished by systemic administration of the muscarinic receptor antagonist methylatropine (6). Hemodynamic studies have shown that initial pressor responses and bradycardia produced by L-Glu microinjections are associated with vasoconstriction in aortic (hindquarter), renal, and mesenteric vascular beds. Vasoconstriction is followed by vasodilation in the hindquarter bed. Systemic administration of prazosin and the nitric oxide (NO) synthase inhibitor N^G-nitro-L-arginine methyl ester (L-NAME) reduces vasodilation. Therefore, it is probable that the hemodynamic effects of L-Glu microinjections into the NTS of conscious rats involve both activation of the sympathetic nervous system and release of nitrosyl factors in hindquarter vessels (7).

Cardiovascular responses induced by glutamate microinjected into the NTS of conscious animals is blocked by microinjection into the same site of kynurenic acid, an ionotropic glutamatergic blocker (6). In this study, we sought to test the hypothesis that the cardiovascular responses that distinguish effects of N-methyl-D-aspartic acid (NMDA) from those of non-NMDA receptor activation result from divergent hemodynamic responses produced by the two types of stimuli. To test that hypothesis and to better understand how glutamatergic neurons in NTS modulate arterial pressure, we sought to define the hemodynamic effects produced by microinjection of ionotropic glutamate agonists (NMDA and non-NMDA) into the NTS of conscious rats.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. All experiments were performed in conscious, freely moving, male Wistar rats weighing between 250 and 350 g. Animals were maintained on a 12:12-h light-dark cycle in temperature-controlled rooms. Food and water were available ad libitum except during the experiments. The Medical Ethics Committee of the Universidade Federal de São Paulo approved the study.

Surgical procedures. Five days before the experiments were performed, rats were anesthetized by an intraperitoneal injection of ketamine (50 mg/kg, Holliday-Scott, Buenos Aires, Argentina). The rats were placed in a stereotaxic frame (model 1940, David Kopf Instruments) with the incisor bar fixed at 3.5 mm. The skull surface was exposed, and a screw was inserted into the parietal bones bilaterally. Guide cannulae were implanted into the brain using the stereotaxic coordinates of Paxinos and Watson (15) as described previously (6). Briefly, a 15-mm-long stainless steel cannula (22 gauge) was introduced through a cranial window made caudal to lambda. The cannula was positioned 14.5 mm caudal to bregma, 0.5 mm lateral to the midline, and 5.5 mm below the skull surface at the level of bregma. The tip of the guide cannula was placed in the cerebellum 1.0 mm above the dorsal surface of the brain stem. The guide cannula was fixed with methacrylate both to the skull and to watch screws inserted in the skull. An occluder closed the cannula until needed for the experiments. The animals were allowed to recover fully from anesthesia and were kept in their home cages. At least 24 h before the experiments were performed, the rats were anesthetized with a mixture of halothane (2%) in oxygen (100%). Femoral arterial and venous polyethylene cannulas (PE-10 connected to PE-50, Clay Adams; Parsippany, NJ) were implanted for measurement of pulsatile arterial blood pressure, mean arterial pressure (MAP), and HR and for the administration of drugs, respectively. After catheterization, a midline laparotomy was performed, and miniature pulsed Doppler flow probes (Iowa Doppler Products; Iowa City, IA) were placed around the lower abdominal aorta (1.3 mm of lumen), superior mesenteric artery (1.0 mm of lumen), and left renal artery (0.8 mm of lumen) for measurement of hindquarter, mesenteric, and renal blood flow, respectively. The probes were sutured in place, the leads and catheters were tunneled subcutaneously and exteriorized between the scapulae, and the wounds were closed. To protect the probe wires and polyethylene tubing while allowing the animals to have unrestricted movement during recovery and experimental testing, free ends of the catheters and Doppler leads were passed through a stainless steel skin button connected to a spring-swivel assembly. The assembly was mounted to a ring stand clamp and suspended above the cage. The skin button was attached to the skin incision in the scapular region using stainless steel sutures. Details of the Doppler technique, including the reliability of the method for estimation of flow velocity, have been described previously by Haywood and his colleagues (10). Relative mesenteric, renal, and hindquarter vascular conductance (%) were calculated as the ratio of Doppler shift and MAP. Data were expressed as percent change from the baseline.

Protocols. All studies were performed in conscious animals after recovery from instrumentation (at least 24 h). Arterial cannulae were connected to a Statham P23Db transducer, and flow probe leads were connected to a Doppler flowmeter (Department of Bioengineering; University of Iowa, Iowa City, IA) for recording MAP, HR, and blood flows, respectively. The data were recorded on a polygraph (Grass model 7; 8 channels). A needle (33 gauge) 1.5 mm longer than the guide cannula previously placed in the brain was connected by PE-10 tubing to a 1-µl syringe (Hamilton; Reno, NV). The syringe was used to deliver microinjections into the NTS. Injections of NMDA (Sigma, St. Louis, MO) and S--amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA; Sigma) (0.1, 1.0, 5.0, and 10.0 pmol/100 nl) into the NTS defined dose-related responses for each agonist. Unilateral microinjections were made after baseline hemodynamic values were established. The interval between injections of different doses was 20 min. After dose-response analysis, hemodynamic effects produced by NMDA (5.0 pmol/100 nl) and AMPA (10.0 pmol/100 nl) were determined after systemic administration of methylatropine (1 mg/kg iv, Sigma).

Histology. After the experiments, methylene blue (100 nl of a 2% solution) was microinjected at the same site in the NTS for histological analysis. The animals were killed by an overdose of urethane (1.8 g/kg iv) and saline was perfused through the heart followed by 10% buffered formalin. The brains were stored in buffered formalin for 2 days after which serial coronal sections (40 µm) were cut and stained by the Nissl method (9). Only rats whose microinjection sites were located in the NTS at the level of the calamus scriptorius were used for data analysis.

Statistics. All data were expressed as means ± SE and were analyzed by repeated-measures ANOVA followed by Student's modified t-test with Bonferroni correction for multiple comparisons among means by paired t-test. Significance was accepted at a 0.05 confidence level.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Hemodynamic effects elicited by microinjection of NMDA into the NTS. A typical example of effects elicited by microinjection of NMDA (5 pmol/100 nl) into the NTS of a conscious rat is shown in Fig. 1A. Increasing doses of NMDA microinjected into the NTS of conscious rats (n = 6) produced a depressor response. Lower doses (0.1 and 1.0 pmol/100 nl) decreased MAP from a baseline of 122 ± 5 mmHg (MAP 4 ± 2 and 28 ± 12 mmHg, respectively; P < 0.05). Higher doses (5.0 and 10.0 pmol/100 nl) produced a biphasic response, which was a decrease in MAP followed by an increase (from 32 ± 12 to 22 ± 8 and from 48 ± 8 to 23 ± 9 mmHg, respectively; P < 0.05). All doses induced significant dose-related bradycardia (from 39 ± 13 to 175 ± 25 beats/min; P < 0.05) and decreases in hindquarter, renal, and mesenteric blood flow. The dose-related changes in regional blood flows elicited by NMDA microinjections are shown in Fig. 2A. Vascular conductance (expressed as percent change from baseline) decreased in renal (from 22 ± 12 to 44 ± 7%), mesenteric (from 16 ± 12 to 31 ± 13%), and hindquarter arteries (from 4 ± 4 to 41 ± 10%; P < 0.05). Responses were divided into two phases defined by MAP changes. Maximum hypotension defined phase I, and maximum hypertension, when it occurred, defined phase II. Recovery was observed about 3 min after phase I. A summary of hemodynamic effects produced by NMDA represented by phases I and II and recovery are shown in Fig. 3.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Typical tracing showing changes in pulsatile arterial pressure (PAP), mean arterial pressure (MAP), heart rate (HR), hindquarter blood flow (HBF), renal blood flow (RBF) and mesenteric blood flow (MBF) elicited by N-methyl-D-aspartic acid (NMDA) (5.0 pmol/100 nl) (A) and S--amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) (10.0 pmol/100 nl) (B) microinjection into the nucleus tractus solitarii (NTS) of conscious rats. Cardiovascular responses were phase I (pI), nadir of the initial fall of arterial pressure; phase II (pII), subsequent peak rise of arterial pressure, and phase III (pIII), peak elevations of HBF (seen only in animals treated with AMPA). Recovery (R) bars were shown about 3 min after phase I. Arrows, time of microinjection.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Dose-related changes in regional blood flows elicited by microinjections (doses of 0.1, 1.0, 5.0, and 10.0 pmol/100 nl) of NMDA (A; n = 6) and AMPA (B; n = 8) into NTS of conscious rats. Data were expressed as means ± SE; *P < 0.05, significant response; P < 0.05, significant change from basal.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Dose-related hemodynamic responses to microinjections of NMDA (0.1, 1.0, 5.0, and 10.0 pmol/100 nl) into the NTS of conscious rats (n = 6). Percentile changes are expressed as RVC%, renal vascular conductance; MVC%, superior mesenteric vascular conductance, and HVC%, hindquarter vascular conductances. Data were expressed as means ± SE; *P < 0.05, significant response; P < 0.05, significant change from basal.

Hemodynamic effects elicited by microinjection of AMPA into NTS. A typical example of effects elicited by microinjection of AMPA (10 pmol/100 nl) into the NTS of a conscious rat are shown in Fig. 1B. Increasing doses of AMPA microinjected into the NTS of conscious rats (n = 8) produced MAP responses (baseline 112 ± 4 mmHg) similar to those produced by NMDA microinjections. Low doses (0.1 and 1.0 pmol/100 nl) decreased MAP (30 ± 8 and 46 ± 9 mmHg, respectively). Higher doses (5.0 and 10.0 pmol/100 nl) first decreased MAP but then increased it (from 31 ± 11 to 36 ± 10 and 23 ± 12 to 41 ± 7 mmHg, respectively; P < 0.05). All doses produced bradycardia (HR from 77 ± 33 to 163 ± 24 beats/min) and decreases in hindquarter, renal, and mesenteric blood flow. The initial decrease in hindquarter blood flow was followed by a pronounced increase in flow. The dose-related changes in regional blood flows elicited by AMPA microinjections are shown in Fig. 2B. Vascular conductance (expressed as percent change from baseline) decreased in renal (from 1 ± 8 to 34 ± 6%; P < 0.05), mesenteric (from 4 ± 3 to 41 ± 9%; P < 0.05), and hindquarter vessels (from 2 ± 17 to 33 ± 8%), but the higher doses of AMPA also elicited delayed increase in vascular conductance in the hindquarter bed (from 10 ± 10 to 35 ± 13%, P < 0.05). Responses were divided into phases to facilitate data analysis. As in the NMDA studies, maximal hypotension and hypertension defined phases I and II, respectively, but maximal hindquarter vasodilation during the hypertensive phase defined a phase III not present in NMDA studies. Recovery was observed ~3.5 min after phase I. A summary of hemodynamic effects produced by AMPA represented by phases I, II, and III and recovery are shown in Fig. 4.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Dose-related hemodynamic responses to microinjections of AMPA (0.1, 1.0, 5.0, and 10.0 pmol/100 nl) into the NTS of conscious rats (n = 8). Data were expressed as means ± SE; *P < 0.05, significant response; P < 0.05, significant change from basal.

Effects of methylatropine treatment on hemodynamic responses to NMDA and AMPA. To determine whether hypotension might be a response to decreased cardiac output associated with bradycardia, we assessed cardiovascular responses to NMDA or AMPA before and after systemic administration of the muscarinic receptor antagonist methylatropine (1 mg/kg iv).

Microinjection sites within the NTS. All microinjections that led to hemodynamic effects were located in the NTS either on the left or the right side. The side of injection did not influence hemodynamic responses. A diagrammatic representation of microinjection sites within the NTS of each rat considered in this study are shown in Fig. 5. No hemodynamic responses occurred when microinjection sites fell outside the NTS (hypoglossal nucleus, intravenous ventricle, dorsal motor nucleus of vagus, and area postrema; data not shown). Photomicrography of one rat is shown on Fig. 6.

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Diagrammatic representation of injection sites (black shading) in each rat. Coronal sections of brain stem are depicted at the same rostral-caudal level. A: animals that received NMDA. B: animals that received AMPA. C and D: animals that received methylatropine with NMDA (C) and AMPA (D). AP, area postrema; cc, central canal; X, dorsal motor nucleus of vagus; XII, hypoglossal nucleus. Arrow, section used for the photomicrograph in Fig. 6; bars, 1.3 mm of neural tissue.

View larger version (136K): [in this window] [in a new window]   Fig. 6.   Photomicrography of a coronal section of the brain stem showing a typical unilateral microinjection site in the NTS. Arrow, center of microinjections. Scale, 3 cm correspond to 1 mm of neural tissue.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study shows that hemodynamic responses resulting from activation of one type of ionotropic glutamate receptor in the NTS of conscious rats differ from those produced by activation of other ionotropic receptors. Vasodilatory responses produced by activation of AMPA, but not NMDA, receptors resemble those produced by stimulation of the NTS with glutamate or by stimulation of afferent baroreceptor nerves. Finally this study shows that hypotension produced by stimulation of AMPA and NMDA receptors in conscious rats, like that produced by stimulation of NTS with glutamate, is blocked if the HR responses are themselves blocked by a muscarinic antagonist.

This study provides additional insights into the role played by NMDA and AMPA receptors in mediating cardiovascular responses through the NTS. NMDA and AMPA both promoted vasoconstriction in all vascular beds but only AMPA produced a delayed vasodilation in the hindquarter bed. Previous neuroanatomical, electrophysiological, and neuropharmacological studies suggested that NMDA and non-NMDA receptors within the NTS participate in cardiovascular control by that nucleus (1, 3, 5, 14, 21, 22).

NMDA receptors seem to be involved in cardiovascular regulation in NTS. However, intracellular studies (2) (in vitro) suggested that non-NMDA receptors transmitted input from visceral afferents to primary baroreceptors neurons and NMDA receptors played only a modulatory role within NTS. In vivo studies (21, 22) suggested that NMDA and non-NMDA receptors in the NTS participated in reflex transmission at different levels of the reflex pathway. The first synapse in NTS (activated monosynaptically by stimulation of baroreceptor afferents) depended on non-NMDA, but not NMDA, receptor activation. On the other hand, activation of second and higher order neurons was more related to NMDA receptor activation (22). Double-labeling studies employing Fos immunoreactivity and glutamate receptor immunoreactivity revealed that Fos expression, as a result of baroreceptor activation, localized with GluR1 (non-NMDA receptor subunit) immunoreactivity but not with NMDAR1 immunoreactivity (5). This interesting finding further supported a role for glutamate in baroreceptor reflex transmission through NTS and suggested that the glutamate receptor subunits may contribute differentially to transmission in that pathway. Different glutamate receptors are differentially located with respect to the synaptic order of neurons, but we (11, 12) and others (1) have also shown that their distribution in NTS is remarkably similar.

Consideration of methods. On-line measurement of arterial blood pressure, HR, and regional blood flow by pulsed Doppler flow probes is standard and has been used in many laboratories for years (7, 10). Microinjection of biologically active substances into the brain has also become a standard neurobiological tool, but its utilization in conscious animals has been perfected only over the past decade (13). Damage to the NTS is minimized with this method by placing guide cannulas well above the nucleus and leaving the injection pipette in place only during microinjection. Therefore, as shown in Figs. 5 and 6, injections can be placed accurately into regions of NTS and the injectate limited in its spread to other regions. However, the method clearly is unable to selectively stimulate specific neurons whose membranes express receptors with affinity for any given agonist delivered into the region. Nonetheless, the method does have an advantageous feature (not seen with iontophoresis) of delivering a predictable dose of an agent at a known time and over a short interval. Therefore, when comparing responses to AMPA and NMDA, we can be assured that cells near the point of injection will have the same time course of stimulation regardless of the agonist injected. The limitation of this method is that the limits of diffusion of any two agents may differ. Therefore, cells and their receptors at the extremes of the diffusion could be reached more by one agent than by another. If neurons responsible for vasodilation lay predominantly at the extremes of diffusion and were reached by AMPA, but not by NMDA, the dilation would be errantly attributed to AMPA receptor activation. Had NMDA reached the same cells, it too, would have potentially elicited dilatation. However, this seems an unlikely explanation for our results for four reasons. First, we have shown that mGLUR1 (AMPA receptor subunit) and NMDAR1 (NMDA receptor subunit) immunoreactivity is similarly distributed in each NTS region (11, 12). Second, there is no evidence to suggest that vasodilation is controlled only from one region of the NTS. Third, whereas the microinjection technique allows reproducible activation of regions of NTS, there is some variability in the actual site of injection so that some NMDA injections, or some NMDA doses, would be expected to reach vasodilatory neurons and elicit increase in hindquarter conductance even though others did not. None did. Finally, we chose high volumes (100 nl) of injectate to ensure maximal exposure of NTS to each agent. Talman et al. (18) showed that unilateral injections of 10- to 25-nl volumes diffuse throughout the ipsilateral NTS. It would then be reasonable to expect a 100-nl volume to carry critical concentrations of each agent throughout the nucleus.

Consideration of results. This is the first study that combines specific stimulation of AMPA and NMDA receptors in NTS with measurement of regional blood flow responses in conscious, freely moving rats. Therefore, some conclusions can be inferred from our findings. The studies suggest that hindquarter vasodilation produced when glutamate is injected into NTS (7) is the product of AMPA, not NMDA, receptor activation. Stimulation of baroreceptor afferent fibers in the superior laryngeal nerve leads to similar hindquarter vasodilation in anesthetized rats (16) as does stimulation of NTS with glutamate (20). However, in both cases, the stimuli cause bradycardia and hypotension, not hypertension, as seen with both glutamate (13) and now AMPA injections in conscious animals. We conjecture that hypotension is not seen in awake animals because the agonist stimulates both baroreceptor (more rostral and lateral) and chemoreceptor (more commissural) regions of NTS. With simultaneous stimulation of both, the chemoreceptor response (hypertension) predominates. After lesions of the commissural region, microinjections of glutamate into the more rostral intermediate NTS leads to hypotension and bradycardia (6) even in awake animals.

Perspectives. In summary, this study demonstrates that hemodynamic responses elicited by NMDA and AMPA in the NTS differ. The contrasting hemodynamic effects could be explained if the two agonists acted on the same NTS neuron to produce differential activation of specific axonal projections from that neuron. This seems a very unlikely explanation. It is possible, however, that NMDA and AMPA might act on the same NTS neuron to affect release of a different transmitter or modulator from terminals of that neuron. The actions of those different affectors could then explain differing vascular responses. We think it most likely that the two agonists differentially activate subsets of NTS neurons. Preferential activation of one subset could produce the NMDA-related responses, whereas activation of another could elicit AMPA-related responses. It seems unlikely that differences in diffusion of NMDA and AMPA explain any differences in subsets of neurons reached by the agonist in that injection sites varied somewhat between animals, and, yet, responses were consistent throughout the study. Thus we conjecture that within the NTS there are subsets of neurons that preferentially respond to agonists for different ionotropic glutamate receptors and, on activation, preferentially influence different regional vascular beds.