Pharmacological treatment for heart failure: a view from the brain.

Systolic heart failure is a feed-forward phenomenon with devastating consequences. Impaired cardiac function is the initiating event, but central nervous system mechanisms activated by persistent altered neural and humoral signals from the periphery play an important sustaining role. Animals with experimentally induced heart failure have neurochemical abnormalities in the brain that, when manipulated, profoundly affect sympathetic drive, volume regulation, and cardiac remodeling--critical determinants of outcome. This brief review explores recent studies that provide a strong rationale for the development of pharmaceutical agents that target central nervous system abnormalities in heart failure.

Hypothalamic corticotrophin-releasing factor and norepinephrine mediate sympathetic and cardiovascular responses to acute intracarotid injection of tumour necrosis factor-alpha in the rat.

Systemic administration of tumour necrosis factor (TNF)-alpha induces the release of norepinephrine in the paraventricular nucleus (PVN) of hypothalamus and an increase in expression of corticotrophin-releasing factor (CRF) and CRF type 1 receptors. We explored the hypothesis that CRF and norepinephrine in PVN mediate the cardiovascular and sympathetic responses to acute systemic administration of TNF-alpha. In anaesthetised rats, the increases in arterial pressure and heart rate induced by intracarotid artery injection of TNF-alpha were attenuated by intracerebroventricular (ICV) injection of either the alpha 1-adrenergic antagonist prazosin or the CRF antagonist alpha-helical CRF. Prazosin blocked the TNF-alpha-induced increase in renal sympathetic nerve activity (RSNA), whereas alpha-helical CRF substantially reduced the RSNA response. Conversely, CRF and the alpha 1-adrenergic agonist phenylephrine, administered ICV, both elicited increases in PVN neuronal activity, RSNA, arterial pressure and heart rate. Microinjection of CRF and phenylephrine directly into PVN evoked smaller responses. These results are consistent with the hypothesis that norepinephrine and CRF in the PVN mediate the cardiovascular and sympathetic responses to acute systemic administration of TNF-alpha.

Central mineralocorticoid receptor blockade improves volume regulation and reduces sympathetic drive in heart failure.

The mineralocorticoid (MC) receptor antagonist spironolactone (SL) improves morbidity and mortality in patients with congestive heart failure (CHF). We tested the hypothesis that the central nervous system actions of SL contribute to its beneficial effects. SL (100 ng/h for 28 days) or ethanol vehicle (VEH) was administered intracerebroventricularly or intraperitoneally to rats with CHF induced by coronary artery ligation (CL) and to SHAM-operated controls. The intracerebroventricular SL treatment prevented the increase in sodium appetite and the decreases in sodium and water excretion observed within a week of CL in VEH-treated CHF rats. Intraperitoneal SL also improved volume regulation in the CHF rats, but only after 3 wk of treatment. Four weeks of SL treatment, either intracerebroventricularly or intraperitoneally, ameliorated both the increase in sympathetic drive and the impaired baroreflex function observed in VEH-treated CHF rats. These findings suggest that activation of MC receptors in the central nervous system plays a critical role in the altered volume regulation and augmented sympathetic drive that characterize clinical heart failure.

Progression of heart failure after myocardial infarction in the rat.

This study examined the early neurohumoral events in the progression of congestive heart failure (CHF) after myocardial infarction (MI) in rats. Immediately after MI was induced by coronary artery ligation, rats had severely depressed left ventricular systolic function and increased left ventricular end-diastolic volume (LVEDV). Both left ventricular function and the neurohumoral indicators of CHF underwent dynamic changes over the next 6 wk. LVEDV increased continuously over the study interval, whereas left ventricular stroke volume increased but reached a plateau at 4 wk. Plasma renin activity (PRA), arginine vasopressin, and atrial natriuretic factor all increased, but with differing time courses. PRA declined to a lower steady-state level by 4 wk. Six to 8 wk after MI, CHF rats had enhanced renal sympathetic nerve activity and blunted baroreflex regulation. These findings demonstrate that the early course of heart failure is characterized not by a simple "switching on" of neurohumoral drive, but rather by dynamic fluctuations in neurohumoral regulation that are linked to the process of left ventricular remodeling.

Neurohumoral regulation in ischemia-induced heart failure. Role of the forebrain.

Congestive heart failure (CHF) is characterized by neurohumoral excitation. Increased sympathetic drive and activation of the reninangiotensin-aldosterone system (RAAS), with vasoconstriction and volume retention, are hallmarks of the CHF syndrome. Treatment strategies have targeted the peripheral influences of these two systems, but have not addressed the central mechanisms that drive them. We monitored the development of CHF following coronary ligation in adult Sprague-Dawley rats. Left ventricular dysfunction characteristic of CHF was confirmed by echocardiography, and the CHF syndrome was validated by measurements of circulating hormones, sodium appetite, thirst, renal sodium and water retention, and renal sympathetic nerve activity (RSNA). In CHF rats, neuronal activity in the hypothalamic paraventricular nucleus (PVN), which mediates downstream effects of forebrain circumventricular organs, was increased and was inhibited by blocking components of the RAAS at the forebrain level. Forebrain (AV3V) lesions and intracarotid (forebrain directed) injections of agents (captopril, losartan, spironolactone) that block RAAS substantially attenuated the behavioral and physiological manifestations of CHF. Intravenous losartan and captopril, in doses that lower arterial pressure, increased RSNA. These findings demonstrate an important role for RAAS-activated forebrain mechanisms in CHF and suggest that the central neural mechanisms driving sympathetic nerve activity and volume retention may persist and promote the progression of CHF despite treatments directed toward the peripheral influences of RAAS.

Arterial chemoreflex in conscious normotensive and hypertensive adult rats.

Evidence from human and animal studies suggests that the arterial chemoreflex may be exaggerated in essential hypertension. In the present study, cardiorespiratory responses to peripheral chemoreceptor stimulation were compared in conscious unrestrained spontaneously hypertensive (SH) and normotensive Wistar-Kyoto (WKY) and Sprague-Dawley (SD) rats (13-14 wk old). Chemoreceptors were stimulated by injections of potassium cyanide (30-125 microgram/kg iv). Chemoreceptor stimulation elicited a pressor response and bradycardia. The peak change in mean arterial pressure evoked during chemoreceptor stimulation was not significantly different between SH (n = 18) and WKY (n = 18) rats but was significantly smaller in SD rats (n = 18). An evaluation of respiratory responses to chemoreceptor stimulation in conscious and anesthetized rats also demonstrated no significant difference between SH and WKY rats, but the response of the SD rats tended to be smaller. These results demonstrate that differences in the arterial chemoreflex response of SH vs. normotensive rats are not linked to hypertension but, rather, to differences between rat strains.

Electrophysiological properties of rat lateral parabrachial neurons in vitro.

Anatomical studies have demonstrated that the lateral parabrachial nucleus (LPBN) is composed of at least seven separate subnuclei distinguished by cell morphology, spatial clustering, and afferent and efferent connectivity. We hypothesized that neurons within the subnuclear clusters of the LPBN might have distinct electrophysiological properties that correlate with cellular morphology. An in vitro slice preparation was used to intracellularly record the intrinsic properties of 64 neurons located within the external lateral (EL) and central lateral (CL) subnuclei of the LPBN in adult rats. Analysis of intrinsic properties revealed that neurons in the EL subnucleus had significantly wider action potentials and on the average demonstrated more spike frequency adaptation during 2 s of depolarization compared with CL neurons. The majority of both EL and CL area neurons expressed delayed excitation (DE) after membrane hyperpolarization. DE was eliminated with the A-current blocker 4-aminopyridine (1.5-5 mM). Postinhibitory rebound was also observed in a subpopulation of EL and CL neurons. Morphological analysis of 11 LPBN neurons, which were electrophysiologically characterized and filled with 2% biocytin, failed to demonstrate an association between morphology and the electrophysiological profiles of LPBN neurons. The lack of distinct "type" of neuron within a single subnucleus of the LPBN is in agreement with recent findings reported from the neonatal rat.

Lateral parabrachial nucleus modulates baroreflex regulation of sympathetic nerve activity.

Previous studies have demonstrated that the lateral parabrachial nucleus (LPBN) is an important site for descending modulation of baroreflex control of heart rate. In the present study it was hypothesized that the LPBN neurons may also modulate baroreflex control of arterial pressure and sympathetic nerve activity. In urethan-anesthetized rats, electrical or chemical activation of the LPBN produced a significant reduction in the magnitude of the baroreflex inhibition of mean arterial pressure (MAP) and renal sympathetic nerve activity (RSNA) elicited by aortic depressor nerve stimulation. Chemical inactivation of the LPBN resulted in a small increase in baroreflex control of MAP, but baroreflex control of RSNA was not affected. The results suggest that LPBN neurons have little tonic influence over baroreflex control of MAP and RSNA in the anesthetized rat. When the LPBN is activated, however, LPBN neurons may function to reduce the capacity of the baroreflex to regulate sympathetically mediated increases in arterial pressure.

Vasopressin and V1-receptor antagonists modulate the activity of NTS neurons receiving baroreceptor input.

Arginine vasopressin (AVP) may act as a neurotransmitter or neuromodulator in the solitary tract nucleus (NTS). To determine whether AVP influences the activity of NTS neurons receiving cardiovascular afferent input, we used single-unit extracellular recording combined with local microinjection to test the effects of AVP and V1-receptor antagonists (antAVP) on spontaneously active NTS neurons in anesthetized rats. Phenylephrine-induced increases in arterial pressure were used to identify neurons receiving baroreceptor input. Phenylbiguanide was used to stimulate chemosensitive cardiopulmonary receptors. AVP excited 31 of 81 NTS neurons tested and inhibited 15 of 81 neurons. AntAVP had independent effects on NTS neurons: in addition to blocking the effects of AVP, antAVP inhibited 26 of 72 neurons but excited only 13. Eighty-two percent of NTS neurons receiving excitatory or inhibitory baroreceptor inputs responded to AVP; 61% of these were excited by AVP. Fifty-eight percent of neurons receiving cardiopulmonary receptor input responded to AVP. These results suggest that AVP in rat NTS has a tonic, predominantly excitatory influence on a significant proportion of baroreceptor-related neurons.

Altered expression of delayed excitation in medial NTS neurons of spontaneously hypertensive rats.

The spontaneously hypertensive (SH) rat has an exaggerated sympathetic discharge which may result from an enhanced neuronal excitability in the central nervous system. To test this hypothesis, we examined the electrophysiological properties of neurons in the medial region of the nucleus of the solitary tract (mNTS), a central nucleus involved in the processing of baroreceptor afferent information, in SH rats and normotensive Sprague-Dawley (SD) rats. An in vitro brainstem slice preparation was used to record intracellularly from 18 neurons in 4-5-month-old SH rats and 16 neurons in 4-5-month-old SD rats. Between the two groups there was no significant differences in resting membrane potential, input resistance, and spontaneous firing frequency, or in action potential amplitude, duration, and after-hyperpolarization (AHP). There were no significant differences in spike frequency adaptation and post-tetanic hyperpolarization (PTH). Delayed excitation (DE), a manifestation of A-current, occurred in 88% in SH and SD mNTS neurons, but the duration of DE was significantly (P < 0.05) shorter in SH mNTS neurons. We propose that attenuated expression of A-current may contribute to increased sympathetic drive in SH rats.

Influence of circulating angiotensin II and vasopressin on neurons of the nucleus of the solitary tract.

The central cardiovascular effects of the circulating peptides angiotensin II (ANG II) and arginine vasopressin (AVP) may be mediated through the nucleus of the solitary tract (NTS). In this study, we used single-unit extracellular recording techniques to determine the effects of peripheral infusions of ANG II and AVP on NTS neurons. Barosensitive NTS neurons were characterized by their response to an increase in arterial pressure induced with phenylephrine (PE). Of 74 NTS neurons, 17 were excited by ANG II infusion. Of these, nine were also excited by activation of baroreceptors. ANG II inhibited 15 of 74 neurons, 13 of which were also baroinhibited. Of 77 NTS neurons, 13 were excited by AVP infusion, 7 of which were also baroexcited. One AVP-excited neuron was inhibited by a PE-induced increase in arterial pressure. Of 77 neurons, 21 were inhibited by AVP, 10 of which were also baroinhibited. Out of 21 AVP-inhibited neurons, 2 were excited by an increase in arterial pressure. These results indicate that some barosensitive NTS neurons may be independently influenced by circulating peptides and that separate populations of barosensitive and peptide-sensitive neurons are found in the NTS.

Antagonism of vasopressin V1 receptors in NTS attenuates baroreflex control of renal nerve activity.

Previous work has shown that arginine vasopressin (AVP) present in the nucleus tractus solitarius (NTS) contributes to the control of peripheral cardiovascular parameters such as arterial pressure, heart rate, and sympathetic activity. In this study, we attempted to elucidate the influence of AVP in the NTS on baroreflex control of mean arterial pressure (MAP) and renal sympathetic nerve activity (RSNA) in urethan-anesthetized rats. To test the baroreflex, the aortic depressor nerve (ADN) was stimulated over a range of frequencies (1-15 Hz) to produce frequency-response curves for MAP and RSNA. The vasopressin V1-receptor antagonist [d(CH2)5Tyr(Me)]AVP (aAVP) was microinjected bilaterally or unilaterally into the caudal NTS to eliminate the influence of endogenous AVP on decreases in MAP and RSNA elicited by ADN stimulation. Bilateral microinjection of 10 or 100 ng aAVP significantly attenuated the decreases in RSNA elicited by ADN stimulation. Decreases in MAP were only attenuated following bilateral microinjection of 100 ng aAVP. Unilateral microinjection of the same doses of aAVP did not influence baroreflex control of MAP or RSNA. These results indicate that endogenous AVP within the NTS contributes to cardiovascular regulation. They also suggest that AVP within the NTS may specifically influence baroreflex control of RSNA.

Cardiac rhythmicity among NTS neurons and its relationship to sympathetic outflow in rabbits.

The caudal nucleus tractus solitarius (NTS) is the primary termination site for baroreceptor afferents, but distinct pulse phasic activity is rarely recorded from NTS neurons. In the present study cross-correlation analysis was used to identify pulse phasic activity in NTS neurons in the anesthetized rabbit. Cross-correlation analysis demonstrated that 19 of 38 pressure-sensitive neurons had activity that correlated with the cardiac cycle, including 5 of 11 neurons activated by aortic depressor nerve stimulation, and 12 of 28 pressure-sensitive NTS neurons had activity that correlated with renal sympathetic nerve activity. Autocorrelation analysis identified 17 of 38 pressure-sensitive neurons that had rhythmic activity not related to the cardiac cycle (mean frequency 14.1 +/- 2 Hz). The results indicate that many NTS neurons do have pulse phasic activity consistent with baroreceptor input, but this activity is difficult to detect because it is distributed throughout the cardiac cycle.

Sympathetic responses to stimulation of area postrema in decerebrate and anesthetized rats.

The effects of electrical and chemical stimulation of the area postrema (AP) on mean arterial pressure (MAP) and renal sympathetic nerve activity (RSNA) were examined in urethan- and pentobarbital sodium-anesthetized rats and in unanesthetized decerebrate rats. The AP was electrically stimulated over a range of frequencies (10-100 Hz) and intensities (10-80 microA) with a pulse duration of 0.2 or 1.0 ms. The excitatory amino acid L-glutamate (100 or 200 mM) was microinjected into the AP to preferentially stimulate neuronal cell bodies. In urethan-anesthetized rats, electrical stimulation of the AP decreased MAP and RSNA. In pentobarbital sodium-anesthetized rats, MAP and RSNA were markedly increased by AP stimulation. In unanesthetized decerebrate rats, increases in MAP and RSNA were also observed during electrical AP stimulation. Microinjection of L-glutamate had no effect on MAP and RSNA in anesthetized or in unanesthetized rats. These results indicate that electrical AP stimulation increases sympathetic output in the unanesthetized decerebrate rat and that anesthesia modifies this sympathetic response. The findings also suggest that peripheral responses to L-glutamate and electrical stimulation of the AP are mediated over different central pathways.

Peripheral chemoreceptor inputs to the parabrachial nucleus of the rat.

In the urethan-anesthetized rat, extracellular recordings were made from 60 neurons within the region of the parabrachial nucleus (PBN). The activity of 37 of 44 neurons was altered by selective stimulation of peripheral chemoreceptors via intracarotid injection of NaH2PO4 or CO2-saturated NaHCO3. Most of these neurons (28 of 37) were excited during chemoreceptor stimulation. Twenty-four of 48 neurons responded to changes in baroreceptor input via changes in blood pressure. Most of these neurons (18 of 24) were inhibited during baroreceptor stimulation. Eleven of 32 neurons were affected by both chemoreceptor and baroreceptor inputs. Seven of these neurons had opposite responses during selective afferent stimulation; that is, they were excited during chemoreceptor activation and inhibited during baroreceptor activation. Our observations reveal that neurons within the medial and lateral PBN are responsive to peripheral chemoreceptor input. A subgroup of PBN neurons was shown to integrate information from chemoreceptors and baroreceptors. These results suggest that both the medial and lateral PBN may play a role in the central integration of cardiovascular inputs.

Effects of L-arginine-derived nitric oxide synthesis on neuronal activity in nucleus tractus solitarius.

The purpose of these studies was to determine the effects of L-arginine-derived nitric oxide (NO) synthesis on neuronal activity in solitary tract nucleus (NTS) neurons. Single unit activity was recorded extracellularly from medial NTS neurons in Fischer-344 rats in vivo and in vitro. In anesthetized rats with arterial pressure maintained constant, NG-nitro-L-arginine methyl ester (L-NAME, 10 mg/kg iv), an inhibitor of NO synthesis, decreased the discharge rate in 12 of 14 neurons and increased the discharge rate in two. After injection of L-NAME, the slowing of neuronal activity began within 2-5 min, and maximal responses were observed 12-15 min after injection. The decreases in activity were reversed within 12-15 min with L-arginine (30 mg/kg iv) or immediately with nitroglycerin (NTG, 10-30 micrograms/kg iv). In superfused rat brain slices, the discharge rate was reduced by 1 mM L-NAME in seven neurons, increased in two, and unchanged in one. The decreases in discharge rate were reversed by 2 mM L-arginine (4 of 6 neurons) and by 10-30 microM NTG (6 of 7 neurons). The results show that L-arginine-derived NO can affect the spontaneous discharge rate of NTS neurons. We conclude that NO may influence the excitability of NTS neurons involved in central autonomic control.

Effects of aging on the intrinsic membrane properties of medial NTS neurons of Fischer-344 rats.

1. Recent studies have demonstrated that the arterial baroreflex is imparied with aging and have implicated central components of the baroreflex arc in this autonomic dysfunction. Neurons in the medial portion of the nucleus tractus solitarius (mNTS) receive a major input from the arterial baroreceptors. The present study was undertaken to characterize the intrinsic membrane properties of mNTS neurons in young rats and to test the hypothesis that these properties are altered with aging. An in vitro brain stem slice preparation was used to record intracellularly from mNTS neurons; passive membrane properties, action potential characteristics, and repetitive firing properties were examined and compared. 2. Neurons in the mNTS of young (3-5 mo old) Fischer-344 rats (F-344; n = 35) had a resting membrane potential of -57 +/- 6.9 mV (mean +/- SD), a membrane time constant of 18 +/- 9.0 ms, and an input resistance of 110 +/- 60 m omega. Action potential amplitude was 81 +/- 7.5 mV with a duration at half-height of 0.83 +/- 0.15 ms. The spontaneous firing rate in 24 cells was 4.3 +/- 2.9 Hz. The amplitude and duration of the action potential afterhyperpolarization (AHP) were 6.6 +/- 3.0 mV and 64 +/- 34 ms, respectively. All neurons expressed spike frequency adaptation, action potential AHP, and posttetanic hyperpolarization. Delayed excitation and postinhibitory rebound were present in 34 and 14% of neurons tested, respectively. Neurons from adult (10-12 mo old) F-344 rats (n = 34) were similar to the young F-344 rats with respect to all of these variables. 3. Neurons from aged (21-24 mo old) F-344 (n = 32) were similar to those from young and adult rats, but there were two potentially important differences: the mean input resistance of the aged neurons was higher (170 +/- 150 M omega), with a larger proportion (46% of aged neurons vs. 20% of young neurons and 21% of adult neurons) having input resistances > 150 M omega; and there was a tendency for a smaller percentage of aged neurons (16% of aged neurons vs. 34% of young neurons and 29% of adult neurons) to express delayed excitation. 4. The potential significance of a high input resistance was tested by comparing the steady-state current-voltage (I-V) relationships and the frequency-current (f-I) relationships among low-resistance (1-100 M omega), medium-resistance (101-200 M omega).(ABSTRACT TRUNCATED AT 400 WORDS)

Acute resetting of the carotid sinus baroreflex by aortic depressor nerve stimulation.

The effect of prolonged aortic depressor nerve (ADN) stimulation on carotid sinus baroreflex regulation of arterial pressure (AP) and renal sympathetic nerve activity (RSNA) was examined in anesthetized rabbits. Ramp increases in carotid sinus pressure (CSP) were repeated before and after 5 min of bilateral ADN stimulation. One minute after ADN stimulation the curve relating AP to CSP had shifted up and to the right, characterized by significant increases (P < 0.05) in the maximum (91 +/- 2 to 101 +/- 3 mmHg; mean +/- SE), midpoint (118 +/- 7 to 125 +/- 8 mmHg CSP), and minimum (45 +/- 3 to 53 +/- 4 mmHg) of the AP reflex curve. There was a parallel shift downward of the curve relating RSNA to CSP, characterized by significant decreases in the maximum [100 +/- 0 to 66 +/- 8% of maximum control RSNA value (%max)], the range (90 +/- 2 to 59 +/- 8%max), and the gain (-1.0 +/- 0.2 to -0.5 +/- 0.1%max/mmHg) of the RSNA reflex curve. Values returned to control within 10 min of cessation of ADN stimulation. These results suggest that central neurons processing baroreflex information from one set of mechanoreceptors can be reset by convergent signals arising from another baroreceptor site.

Central command increases sympathetic nerve activity during spontaneous locomotion in cats.

A controversial issue in exercise physiology is the relative contribution of central command versus afferent input from contracting muscles and baroreceptors in the regulation of sympathetic nerve activity (SNA) during exercise. Recent studies of exercising humans have suggested that central command increases cutaneous sympathetic sudomotor nerve activity but have challenged the concept that central command contributes importantly to increases in sympathetic vasoconstrictor nerve activity to skin and skeletal muscle. The purpose of this study was to examine the influence of central command on renal SNA and lumbar SNA during spontaneous locomotion in decorticate cats. Unanesthetized decorticate cats that developed locomotion spontaneously or during electrical stimulation of the subthalamic locomotor region were studied in the presence and absence of input from skeletal muscle and baroreceptor afferents. Spontaneous rhythmic locomotion in the unparalyzed state was associated with significant increases in mean arterial pressure (MAP) from 106 +/- 10 to 133 +/- 11 mm Hg (p less than 0.05) and increases in renal SNA of 301 +/- 100% (p less than 0.05). During spontaneous fictive rhythmic locomotion in paralyzed cats, there were also significant (p less than 0.05) increases in MAP (43 +/- 6%), renal SNA (183 +/- 32%), and lumbar SNA (223 +/- 83%). Baroreceptor denervation did not attenuate increases in MAP, renal SNA, and lumbar SNA during locomotion. During electrical stimulation of the subthalamic locomotor region in paralyzed cats, MAP increased by 43 +/- 17% (p less than 0.05), and renal SNA increased by 175 +/- 47% (p less than 0.05). These findings indicate that central command is capable of increasing sympathetic neural drive in unanesthetized decorticate cats. This increase in sympathetic drive occurs even in the absence of feedback from contracting muscles or from arterial and cardiopulmonary baroreceptors.

Effects of gamma-aminobutyric acid and glycine on synaptic excitability of neurones in the solitary tract nucleus.

In the solitary tract nucleus, neuronal responsiveness to synaptic input from peripheral afferent fibres has been found to decrease as the frequency of that input is increased. The present study investigated the possibility that glycine and gamma-aminobutyric acid (GABA) are (1) involved in this phenomenon of "frequency-dependent inhibition" or (2) capable of otherwise modifying neuronal responsiveness to synaptic input. In 32 of 57 neurones, application of glycine reduced responsiveness to input from visceral afferents. gamma-Aminobutyric acid reduced the responsiveness in 22 of 56 neurones but induced an increase in another 6. Selectivity of agonist effects was verified using glycinergic antagonist strychnine and the GABAergic antagonists picrotoxin and bicuculline. Of 67 neurones examined, 32 exhibited decreased numbers of action potentials with increasing stimulus frequency. Strychnine disrupted the frequency-dependent inhibition in 3 of 11 neurones, while picrotoxin and bicuculline prevented it in 2 of 7 and 4 of 9 neurones, respectively. These results indicate that both glycinergic and GABAergic systems may modulate the responsiveness of neurones in the solitary tract nucleus but that neither fully accounts for the expression of frequency-dependent inhibition.