INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The paraventricular nucleus (PVN) of the hypothalamus is one of the critical nuclei involved in neuroendocrine and autonomic regulation. These actions are mediated in part by the neurohypophysial hormones vasopressin (AVP) and oxytocin (OXY), which are synthesized in PVN and secreted from the pituitary (Swanson and Sawchenko 1980). The magnocellular neurons of the PVN, which are responsible for AVP and OXY production and release, can be distinguished both morphologically and electrophysiologically from parvocellular neurons in the same nucleus (Hoffman et al. 1991; Tasker and Dudek 1991). PVN neurons have been shown to receive afferents from other hypothalamic nuclei, the brain stem, and the subfornical organ (Swanson and Sawchenko 1983; van den Pol 1982). These cells integrate incoming signals from afferent fibers and send projections to the posterior pituitary, median eminence, brain stem, and spinal cord and ultimately influence autonomic regulation through these connections (Swanson and Sawchenko 1980). While the afferent signals to PVN provide the initial stimulus for excitation, it is thought that local synaptic regulation within the nucleus is responsible for preservation and propagation of the correct output. Interestingly, recent data suggest that a breakdown in these synaptic mechanisms may underlie the onset and/or maintenance of several well-described pathophysiological conditions (Ciriello et al. 1984; Earle and Pittman 1995; Patel and Zhang 1996).

The anatomical distribution and cellular actions of the predominant excitatory and inhibitory neurotransmitters of the CNS, glutamate and GABA in the hypothalamus are well documented (Decavel and van den Pol 1990; van den Pol and Trombley 1993; van den Pol et al. 1990). Furthermore, the actions of glutamate and GABA on the magnocellular neurons of the supraoptic nucleus (SON) and PVN and on parvocellular neurons in the PVN have been described (Bains and Ferguson 1997b,c; Renaud et al. 1992). Bains and Ferguson recently showed that a glutamate activated negative feedback loop within PVN, which may constitute a major regulatory mechanism modulating the strength of output signals (Bains and Ferguson 1997b). They observed that after N-methyl-D-aspartate (NMDA) receptor activation, magnocellular neurons were depolarized concurrent with an increase in both action potential firing and membrane conductance. Paradoxically, a subset of these neurons were not only excited but demonstrated an increase in GABAergic inhibitory postsynaptic potentials (IPSPs), the occurrence of which were dependent on nitric oxide (NO) production (Bains and Ferguson 1997b). Interestingly, AVP has also been observed to increase IPSPs in magnocellular neurons (Hermes et al. 2000). This increase in IPSPs could be abolished by tetrodotoxin (TTX) and bicuculline, suggesting the involvement of a GABAergic interneuron (Bains and Ferguson 1997b).

Angiotensin II (ANG) has been shown to influence a variety of neuroendocrine and autonomic functions (Culman et al. 1995; Ferguson and Washburn 1998; Ferguson et al. 1999; Lenkei et al. 1994). ANG-containing cell bodies, nerve terminals, and receptors have been localized in the PVN (Leikei et al. 1998; Lenkei et al. 1994, 1998; Lind et al. 1985; Phillips et al. 1993; Song et al. 1991). Dawson and Krukoff recently demonstrated that systemic infusion of ANG into PVN resulted not only in c-fos activation in neurons immunoreactive for AVP and OXY, but also neurons that stained positive for AT₁ receptors and NADPH-d (Dawson et al. 1998). Magnocellular neurons of both the SON and PVN are known to contain NO synthase (NOS) (Miyagawa et al. 1994; Nakamura et al. 1991; Vincent and Hope 1992), and our own recent studies have described a role for NO in synaptic transmission within PVN (Bains and Ferguson 1997b). In addition, in vivo studies indicate that microdialysis of NO into the PVN results in an increase in GABA release and a decrease in blood pressure (Horn et al. 1994).

We have utilized the whole cell patch-clamp technique to characterize the actions of ANG on synaptic activity within the PVN. We report that application of ANG results in an increased frequency of IPSPs in magnocellular neurons as a consequence of the activation of a NO-mediated GABAergic feedback loop in the PVN.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Slice preparation

Experiments were performed using hypothalamic slices prepared as previously described (Li and Ferguson 1996). Male, Sprague-Dawley rats (150-250 g, Charles River) were killed by decapitation; the brain was quickly removed from the skull and immersed in cold (1-4°C) artificial cerebrospinal fluid (ACSF). The brain was blocked, and 400-µm coronal sections were cut through the hypothalamus using a vibratome. Slices were hemisected, trimmed into blocks containing PVN, and incubated in oxygenated ACSF (95%O₂-5%CO₂) for 90 min at room temperature. Prior to recording, the slice was transferred into an interface-type recording chamber and continuously perfused with ACSF at a rate of 1 ml/min.

Electrophysiology

Electrophysiological experiments were performed using the whole cell configuration of the patch-clamp technique to record from PVN neurons. Patch pipettes were pulled to a resistance of 5-7 M and filled with the pipette solution described in the following text. Signals were processed with an Axoclamp-2A amplifier. An Ag-AgCl electrode connected to the bath solution via a KCl-agar bridge served as reference. All signals were digitized using the CED 1401 plus interface and stored on computer for off-line analysis. Drugs were applied by switching the perfusion solution from ACSF to a solution containing the desired drug. IPSPs were quantified based on frequency and amplitude (>1 mV) and shape (fast falling phase and slow decay) using Spike2 software (CED, Cambridge, UK). Each detected event was inspected visually to exclude obvious false IPSPs. Mean group values were compared with a Student's paired t-test. Dunnett's multiple comparison test was utilized when multiple means were analyzed versus a control group following a one-way ANOVA while a repeated-measures ANOVA and Newman-Keuls multiple comparison test were employed to statistically compare multiple groups. Cumulative probability plots of IPSP amplitude and frequency were compared with the Kolmogorov-Smirnov test.

Solutions

The ACSF composition was (in mM) 124 NaCl, 2 KCl, 1.25 KH₂PO₄, 2 CaCl₂, 1.3 MgSO₄, 20 NaHCO₃, and 10 glucose. Osmolarity was maintained between 285 and 300 mosM and pH between 7.3 and 7.4. The pipette solution contained (in mM) 140 Kgluconate, 0.1 CaCl₂, 2 MgCl₂, 1.1 EGTA, 10 HEPES, and 2 NaATP and had a pH of 7.25 (adjusted with KOH if necessary), except in voltage-clamp experiments where140 mM KCl was substituted for Kgluconate.

A stock solution of ANG (0.1mM, Phoenix Pharmaceuticals) was prepared from which daily aliquots were made to the required concentration. Tetrodotoxin (TTX, Alamone Laboratories), was used to block voltage-activated Na⁺ channels and was prepared daily from a stock solution (0.1mM). N-nitro-L-arginine methylester (L-NAME, Sigma) an inhibitor of NOS, kynurenic acid, a broad-based glutamate receptor antagonist (Sigma), and bicuculline methiodide (BMI, Sigma) were prepared daily as required.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Whole cell patch-clamp recordings were obtained from a total of 35 electrophysiologically identified magnocellular PVN neurons. These cells all demonstrated a linear I-V relationship and a prominent outward rectification (I_A) when depolarized from hyperpolarized potentials (Fig. 1A). These neurons had a mean resting membrane potential (RMP) of 59.7 ± 1.3 mV, displayed action potentials with a minimum spike amplitude of 60 mV, and had a mean input resistance of 1,200 ± 53 M. Action potential and IPSP amplitude was measured throughout the duration of the recordings, and no run-down was observed.

View larger version (26K): [in this window] [in a new window]   Fig. 1. Angiotensin II (ANG) increases inhibitory postsynaptic potentials (IPSPs) in magnocellular neurons. A: the electrophysiological characteristics of a magnocellular neuron are illustrated here. Magnocellular neurons demonstrate a linear I-V relationship and a prominent outward rectification (IA) when depolarized from hyperpolarized potentials. B, top: whole cell current-clamp recording depicts the response (56 mV resting membrane potential) of a magnocellular neuron to bath application of 0.1 µM ANG (30 s, application indicated by bar). Thirty picoampere hyperpolarizing current pulses are applied at 5-s intervals to evaluate the effects of ANG on input resistance in magnocellular neurons. Bottom: ANG depolarizes these neurons concurrent with an increase in IPSPs (traces have been expanded and labeled).

ANG activates IPSPs in magnocellular neurons

A total of 35 magnocellular neurons were tested for the effects of bath application of ANG using current-clamp techniques. After a control recording period of 5 min, ANG was administered by bath perfusion in concentrations ranging from 0.01 to 1 µM for a period of 30 s. Neurons tested with ANG typically responded with a depolarization (6.2 ± 0.7 mV; n = 20, 66% of neurons tested), and peak effects were observed at a dose of 0.1 µM (Fig. 1B, 0.1 µM; 5.0 ± 0.9 mV, n = 16, P < 0.001, unpaired t-test), which was therefore selected for the remainder of experiments. While changes in input resistance of cells influenced by ANG were variable (n = 11; 5 showing increases, 3 decreases, and 3 no change) the group mean was not altered significantly by ANG application (51 ± 183 M P > 0.05, paired t-test).

In 75% (12/16) of these neurons the ANG-induced depolarization was accompanied by an increase in IPSP frequency as illustrated in Figs. 1B and 2A. This ANG-induced increase in IPSPs showed a latency to onset ranging from 30 to 90 s and a duration of between 150 and 180 s in most neurons (Figs. 1B and 2B). The average increase in IPSPs for these 12 neurons (calculated in 30-s bins, normalized to control period) after ANG application was 0-30 s; 27.5 ± 14.9%, 30-60 s; 36.5 ± 12.1%*, 60-90 s; 31.4 ± 14.1%*, 90-120 s; 72.3 ± 10.9%*, 120-150 s; 42.9 ± 12.8%* (*P < 0.05 compared with control, 1-way ANOVA and Dunnett's multiple comparison test) as illustrated in Fig. 2C. The maximum increase in IPSPs (in a 30-s bin, average of 12 neurons: 88.7 ± 9.8% increase) occurred either 60-90 s (n = 3) or 90-120 s (n = 9) after ANG application, demonstrating a consistent time course of IPSP activation in these neurons. No significant receptor desensitization was observed as similar increases in IPSP frequency were observed in response to a second identical application of ANG (n = 3).

View larger version (19K): [in this window] [in a new window]   Fig. 2. Frequency analysis of ANG-induced increase of IPSPs in magnocellular neurons. A: current-clamp recordings demonstrating the increase in frequency of IPSPs after application of 0.1 µM ANG in a magnocellular neuron. B: instantaneous rate frequency histogram demonstrating the increase of IPSP frequency of a single magnocellular neuron in response to application of 0.1 µM ANG (application indicated, ). C: histogram depicting the increase in the IPSP frequency (as a %) after ANG vs. control [each bin (30 s, control and ANG) calculated as a percentage of the mean IPSP frequency (30 s) of the control period] (ANG; 0-30s; 27.5 ± 14.9%, 30-60s; 36.5 ± 12.1%*, 60-90s; 31.4 ± 14.1%*, 90-120s; 72.3 ± 10.9%*, 120-150s; 42.9 ± 12.8%*. * P < 0.05, 1-way ANOVA and Dunnett's multiple comparison test) in the population of magnocellular neurons (n = 12) that respond to 0.1 µM ANG with an increase in IPSP frequency (application indicated, ).

While ANG increased the frequency of synaptic events, the peptide was without effect on the amplitude of these synaptic events. As illustrated in the IPSP amplitude distributions shown in Fig. 3, control magnitude of PSPs (light bars) was similar to that observed after 0.1 µM ANG (dark bars: analyzed after ANG but prior to depolarization to avoid effects of increased driving force for Cl). The amplitude distributions demonstrate that ANG is without effect on the amplitude of synaptic events observed. Similarly analysis of cumulative probability versus amplitude plots also show ANG to be without effect on the amplitude of the observed IPSPs (n = 12; P > 0.05, Kolmogorov-Smirnov test). Cumulative probability/amplitude plots do, however, confirm that IPSPs measured during the depolarization induced by ANG are greater in amplitude (P < 0.05, Kolmogorov-Smirnov test) than the IPSPs observed in the control periods presumably as a result of the increased driving force for Cl movement across the membrane (data not shown). Application of BMI (10 µM) abolished the increase in IPSPs in response to ANG (Fig. 4). Previous studies in our laboratory have also demonstrated the GABAergic nature of these IPSPs (inhibition in the presence of bicuculline) and the effect of membrane potential shifts on their amplitude (Bains and Ferguson 1997b) supporting these observations.

View larger version (14K): [in this window] [in a new window]   Fig. 3. ANG is without effect on the amplitude of the IPSPs. A: amplitude distributions of IPSPs before (light bars) and after (dark bars) application of 0.1 µM ANG. The number of events falling into each amplitude bin in both distributions has been depicted as a percentage of the total number of events analyzed in the distribution (i.e., number of events of this magnitude/the total number of events recorded). The amplitude distributions demonstrate that ANG is without effect on the amplitude of synaptic events observed. B: a cumulative probability vs. amplitude plot reveals ANG is without effect on the amplitude of the observed IPSPs (P < 0.05, Kolmogorov-Smirnov test).

View larger version (13K): [in this window] [in a new window]   Fig. 4. Bicuculline methiodide (BMI) abolishes the ANG-induced IPSPs. A: current-clamp trace demonstrating application of ANG (0.1 µM) results in an increase in IPSP frequency in a magnocellular neuron. BMI (10 µM) abolishes the IPSPs demonstrating their GABAergic nature. Subsequent application of ANG in the same neuron fails to elicit a similar increase in IPSP frequency in the presence of BMI.

L-NAME abolishes ANG-stimulated IPSPs in magnocellular neurons

Previous studies showing that N-methyl-D-aspartate (NMDA) induced IPSPs in magnocellular neurons were NO dependent (Bains and Ferguson 1997b) led us to investigate whether the ANG-induced IPSPs were contingent on the production of NO. The increase in IPSP frequency in magnocellular neurons after application of ANG (145 ± 8.8%, P < 0.05 paired t-test vs. control) was abolished by prior administration of the nonselective NOS inhibitor L-NAME (L-NAME normalized to 100%; L-NAME+ANG; 82.3 ± 9.7%, n = 4) as illustrated in Fig. 5, A and B. Interestingly, L-NAME also decreased the basal frequency of IPSPs, suggesting a tonic role for such inputs in controlling magnocellular excitability but was without effect on the amplitude of the IPSPs. These observations suggest that NO plays a mandatory intermediary role causing the increased frequency of IPSPs in these neurons.

View larger version (15K): [in this window] [in a new window]   Fig. 5. The ANG-induced increase in IPSP frequency is dependent on the production nitric oxide. A: current-clamp recordings demonstrate that the increase in IPSPs in response to 0.1 µM ANG can be abolished by preapplication of the nitric oxide synthase inhibitor, N-nitro-L-arginine methylester (L-NAME, 10µM). Note that the basal level of IPSPs is also decreased in L-NAME, suggesting nitric oxide plays a role in maintaining tonic inhibitory input in these neurons. B: these bar graphs describe the effects of ANG on a population of magnocellular neurons (n = 4) after nitric oxide synthase inhibition. Left: the increase and recovery of IPSP frequency after 0.1 µM ANG application (145 ± 8.8%, P < 0.05 paired t-test, vs. control). Right: after L-NAME application (10 µM), ANG is without effect on IPSP frequency [L-NAME normalized to 100% (vs. recovery); L-NAME+ANG; 82.3 ± 9.7%, n = 4] in the same population of neurons.

IPSC amplitude and frequency are unchanged in TTX

The synaptic mechanisms underlying the ANG-mediated increase in IPSP frequency have not yet been identified. To establish whether the ANG/NO-mediated increase in IPSP frequency was due to an action at the cell body or synaptic terminal of an inhibitory interneuron, voltage-clamp recordings were obtained from an additional six magnocellular neurons of which four demonstrated clear statistically significant increases in IPSC frequency [in 10 µM kynurenic acid: applied to block glutamate receptor-mediated excitatory postsynaptic currents (EPSCs)] after application of 0.1 µM ANG (n = 4) as illustrated in Fig. 6, A and B (KA:1.0 ± 0.1 vs. KA+Ang: 1.8 ± 0.3 Hz; n = 4, P < 0.01 repeated-measures ANOVA and Newman-Keuls multiple comparison test). In contrast, after application of TTX (0.1 µM) to isolate miniature IPSCs, ANG was without effect on frequency (Fig. 6B; KA,TTX 0.6 ± 0.2 Hz vs. KA,TTX+ANG 0.4 ± 0.2 Hz; n = 4, P > 0.05, repeated-measures ANOVA and Newman-Keuls multiple comparison test) or amplitude (Fig. 6C; TTX,KA 60.2 ± 11.1 pA vs. TTX,KA+ANG 44.8 ± 7.8 pA; n = 4, P > 0.05, paired t-test) of these remaining events, supporting the conclusion that NO evokes these effects at the cell body of an inhibitory interneuron.

View larger version (28K): [in this window] [in a new window]   Fig. 6. ANG is without effect on the frequency and amplitude of mini inhibitory synaptic currents. A: voltage-clamp recordings showing that ANG (0.1 µM) application results in an increase in the frequency of spontaneous IPSCs. Recordings are done at -60mV with a KCl (substituted for Kgluconate) electrode with 10 µM kynurenic acid continuously bath applied to abolish EPSCs. After application of TTX (0.1 µM), however, miniature inhibitory synaptic activity was unaffected as illustrated in the bar charts summarizing effects of ANG on mIPSC frequency (Fig. 6B) and amplitude (Fig. 6C). *P < 0.01; repeated-measures ANOVA and Newman-Keuls multiple comparison test. $P > 0.05; repeated-measures ANOVA and Newman-Keuls multiple comparison test. #P > 0.05; paired t-test)

IPSPs modulate ANG responsiveness of magnocellular neurons

It is important to note that not all of the magnocellular PVN neurons recorded in this study showed IPSPs. Those neurons that demonstrated spontaneous IPSPs did however have an attenuated response (Fig. 7D, 3.2 ± 1.1 mV, 0.1 µM ANG, n = 10) to ANG compared with those neurons that did not show tonic IPSPs (8.0 ± 0.5 mV, 0.1 µM ANG, n = 6, P < 0.05 unpaired t-test). In addition, as illustrated in Fig. 7, ANG induced depolarizations (0.1 µM) in neurons in which this peptide caused increases in IPSP frequency were significantly enhanced in the presence of L-NAME (ANG: 2.0 ± 0.7 mV vs. L-NAME: 6.7 ± 0.7 mV, n = 4, P < 0.05; paired t-test). Consequently, magnocellular neurons were tested with ANG in the presence of BMI (10 µM). Application of BMI not only abolished the IPSPs (Fig. 4) demonstrating the GABAergic nature of these events but also amplified the depolarization in response to ANG (1.9 ± 1.2 vs. 11.9 ± 2.3 mV, n = 5, P < 0.001; paired t-test, Fig. 7D). These data suggest a physiological correlate exists between these neurons, an intrinsic inhibitory feedback loop within the nucleus, and their level of excitation and subsequent output.

View larger version (26K): [in this window] [in a new window]   Fig. 7. The magnitude of the ANG-induced change in membrane potential is dependent on the presence of inhibitory synaptic input. Ai: current-clamp recording demonstrating a magnocellular neuron (55 mV resting membrane potential) that showed a small response to bath application of ANG (0.1 µM, application indicated by bar). After application of L-NAME (10 µM, application indicated by light gray extended bar), the response to ANG (0.1 µM) in the same neuron is potentiated (53 mV resting membrane potential). Action potentials have truncated. Aii: the expanded traces demonstrate the high basal frequency of IPSPs observed in this neuron and the decrease in IPSP frequency in this neuron after L-NAME application. Note the lack of an increase in IPSP frequency before and after ANG application in this neuron in the presence of L-NAME. B: bar graph summarizing data demonstrating that neurons that respond to ANG with an increase in IPSPs show a reduced response to ANG (0.1 µM; 3.2 ± 1.1 mV, 0.1 µM ANG, n = 10) compared with neurons with no IPSPs (8.0 ± 0.5 mV, 0.1 µM ANG, n = 6; P < 0.05 t-test). C: this graph illustrates that the change in membrane potential to ANG (2.0 ± 0.7 mV, 0.1 µM ANG, n = 4) is exacerbated in those neurons with a high degree of IPSPs after bath application of the NOS inhibitor L-NAME (6.7 ± 0.7 mV, 0.1 µM ANG, n = 4; P < 0.05; t-test) and the subsequent removal of inhibitory input to these neurons. D: ANG-induced depolarization (1.9 ± 1.2 mV, 0.1 µM ANG, n = 5) is enhanced by application of BMI (10 µM), which abolished GABAergic IPSPs (11.9 ± 2.3 mV, 0.1 µM ANG, n = 5; P < 0.001; t-test).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Despite the well-recognized importance of ANG in neuroendocrine regulation the mechanisms by which this peptide modulates the excitability of PVN neurons are not well understood. We show here that ANG, in addition to depolarizing magnocellular neurons, activates an inhibitory feedback system, which can reduce the magnitude of this excitation. The production of NO, presumably produced by depolarized NOS-positive magnocellular neurons is essential to the activation of this inhibitory feedback loop (Miyagawa et al. 1994; Nakamura et al. 1991; Vincent and Hope 1992). We suggest that NO acts on inhibitory interneurons to increase the frequency of inhibitory synaptic events in magnocellular neurons in a manner similar to that reported after NMDA or vasopressin receptor activation (Bains and Ferguson 1997b; Hermes et al. 2000). These observations demonstrate that PVN possesses intrinsic circuitry that is capable of modulating the excitability of output neurons following multi-modally induced depolarization of magnocellular neurons.

Application of ANG results in the depolarization of magnocellular neurons (Latchford and Ferguson 1999) without a significant change in input resistance. While the mechanisms that underlie the ANG-induced depolarization remain unclear, potential candidates include an inhibition of I_A (Li and Ferguson 1996; Wang et al. 1997), activation of a calcium current (Sumners et al. 1996) or nonselective cationic conductance (Yang et al. 1992), and the possibility of indirect contributions after activation of a glutamate interneuron in PVN (Daftary et al. 1998, 2000).

We have shown here that a subpopulation of magnocellular neurons that were stimulated by ANG also demonstrated an increase in IPSP frequency. Intriguingly, we have previously reported that NO donors including S-nitroso-N-acetylpenicillamine (SNAP) and L-arginine (L-ARG) increase the frequency of GABAergic PSPs in magnocellular neurons (Bains and Ferguson 1997b), suggesting a potential role for NO as a mediator of ANG actions. Our observation that the ANG-induced increase in IPSPs in magnocellular neurons can be abolished by administration of a NOS inhibitor supports this hypothesis. Evidence documenting NO's role in facilitating acetylcholine, catecholamine, and neuroactive amino acid release (Kuriyama and Ohkuma 1995) suggests a potential role in the presynaptic terminal, whereas data reporting direct effects of NO on putative GABA neurons within this nucleus supports the possibility that NO controls the excitability of these interneurons (Bains and Ferguson 1997a). Our voltage-clamp data suggest these effects are most likely the result of peptide-induced effects on the firing frequency of GABAergic interneurons as while ANG increased the frequency of IPSCs, it was without effect on the amplitude of spontaneous synaptic events. The location of these GABAergic interneurons is likely the perinuclear zone (Boudaba et al. 1996; Decavel and van den Pol 1992; Tasker and Dudek 1993; van den Pol 1982). Their close proximity to PVN also makes them plausible targets for NO produced in the lateral magnocellular areas of PVN.

It is important to note that not all PVN magnocellular neurons recorded show IPSPs. Interestingly, those neurons showing IPSPs have an attenuated response to ANG (compared with cells not showing IPSPs). Furthermore, inhibition of inhibitory synaptic input with BMI or reduction of these events with L-NAME enhanced the depolarization observed in magnocellular neurons in response to ANG. In fact, in a number of magnocellular neurons application of BMI revealed an increase in excitatory postsynaptic potential frequency in response to ANG. These observations are in accordance with previous reports that NOS inhibition enhances the response of magnocellular neurons to NMDA receptor activation (Bains and Ferguson 1997b). It remains unknown whether those neurons that demonstrate an increase in IPSP frequency will have an augmented response to ANG if the feedback loop is rendered nonfunctional (by GABA antagonists or NOS inhibitors) in vivo. This is of particular importance pathophysiologically where PVN has been implicated in the onset of both congestive heart failure and chronic hypertension (Ciriello et al. 1984; Earle and Pittman 1995; Eilam et al. 1991; Eilam et al. 1994; Patel and Zhang 1996).

Perspectives

The PVN is essential in the regulation of neuroendocrine and autonomic functions, including body fluid homeostasis and cardiovascular regulation. Neurons in the PVN must be able to monitor the body's internal environment, integrate these messages, and send the appropriate signals to both other central nuclei and peripherally so that a constant environment is maintained. The negative feedback pathway we describe here represents an important circuit whereby the regulation of neuronal excitability and thus the extent of neuropeptide release can be tightly controlled. Physiologically, this inhibitory system has been documented to moderate PVN output signals essential to the control of both cardiovascular and endocrine systems.