Oxytocin in the Central Amygdaloid Nucleus Modulates the Neuroendocrine Responses Induced by Hypertonic Volume Expansion in the Rat

The present study investigated the involvement of the oxytocinergic neurones that project into the central amygdala (CeA) in the control of electrolyte excretion and hormone secretion in unanaesthetised rats subjected to acute hypertonic blood volume expansion (BVE; 0.3 M NaCl, 2 ml/100 g of body weight over 1 min). Oxytocin and vasopressin mRNA expression in the paraventricular (Pa) and supraoptic nucleus (SON) of the hypothalamus were also determined using the real time‐polymerase chain reaction and in situ hybridisation. Male Wistar rats with unilaterally implanted stainless steel cannulas in the CeA were used. Oxytocin (1 μg/0.2 μl), vasotocin, an oxytocin antagonist (1 μg/0.2 μl) or vehicle was injected into the CeA 20 min before the BVE. In rats treated with vehicle in the CeA, hypertonic BVE increased urinary volume, sodium excretion, plasma oxytocin (OT), vasopressin (AVP) and atrial natriuretic peptide (ANP) levels and also increased the expression of OT and AVP mRNA in the Pa and SON. In rats pre‐treated with OT in the CeA, previously to the hypertonic BVE, there were further significant increases in plasma AVP, OT and ANP levels, urinary sodium and urine output, as well as in gene expression (AVP and OT mRNA) in the Pa and SON compared to BVE alone. Vasotocin reduced sodium, urine output and ANP levels, although no changes were observed in plasma AVP and OT levels or in the expression of the AVP and OT genes in both hypothalamic nuclei. The results of the present study suggest that oxytocin in the CeA exerts a facilitatory role in the maintenance of hydroelectrolyte balance in response to changes in extracellular volume and osmolality.

central amygdaloid nucleus; blood volume expansion; oxytocin; vasopressin; atrial natriuretic peptide

It is well established that blood volume expansion (BVE) induces, via reflex mechanisms, several regulatory responses, including the inhibition of sympathetic outflow to the kidney, heart and blood vessels; the reduction of renin and vasopressin secretion; and an increase in the release of atrial natriuretic peptide (ANP) and oxytocin (OT) that leads to diuresis and natriuresis [1] , [2] .

The central amygdaloid nucleus (CeA) was reported to be an important constituent of the central pathways that regulate cardiovascular function and the hormonal and behavioural responses that minimise and correct sodium deficits [3] . Lesions and chemical stimulation of the amygdaloid complex inhibit sodium chloride intake [4] . In addition, Galaverna et al. [5] showed that lesions of the CeA impair both need‐free and need‐induced NaCl intake. Rats with lesions of the CeA also fail to ingest 0.5 M NaCl in response to activation of angiotensin in the brain, or to systemic mineralocorticoids that produce a vigorous NaCl appetite in control rats. Andrade‐Franzé et al. [6] postulated that the facilitatory mechanisms present in the CeA are essential for the increase in water and hypertonic NaCl intake produced by blockade of the inhibitory mechanisms of the lateral parabrachial nucleus (LPBN) in furosemide + low dose of captopril treated rats. In addition, it was shown that the integrity of the CeA is crucial for the increase in fluid intake produced by serotonergic blockade or by α2 adrenergic activation of the LPBN.

Studies showing the involvement of the CeA in a neural circuit that controls the homeostasis of body fluids in response to volume expansion have been published. The first study to show the involvement of the amygdala in the mechanisms activated by BVE was demonstrated by Godino et al. [7] in which volume‐expanded rats (i.v. injection of 0.15 M NaCl, 2 ml/100 g body weight) showed a significant increase in Fos immunoreactive neurones in the lateral division of the CeA and in the bed nucleus of the stria terminalis (BNST). Both of these structures are principal components of the central extended amygdala and are related to blood pressure and fluid intake regulation [8] , [9] .

Using an injection of the retrograde tracer, Fluorogold, into the LPBN, Margatho et al. [10] demonstrated that blood volume‐expanded rats showed a significantly greater number of Fos‐Fluorogold double‐labelled cells in the CeA, BNST and the paraventricular nucleus (Pa). Therefore, it is plausible to consider that the retrogradely labelled neurones in these sites receive inputs from the LPBN and that they are activated by the BVE. Their activation may represent a feedback control related to the autonomic and behavioural aspects of cardiovascular functions [11] . Margatho et al. [12] also showed that the involvement of GABAergic mechanisms within the CeA is related to the control of urinary electrolyte excretion and hormone release in rats subjected to acute isotonic or hypertonic BVE. It was observed that in response to BVE, a previous injection of muscimol into the CeA inhibited the synaptic inputs mediated by the perinuclear GABAergic interneurones in the Pa and SON. Accordingly, it was hypothesised that the oxytocinergic magnocellular neurones in the Pa and the SON received these local GABAergic inhibitory inputs, leading to a decrease of plasma OT and ANP release and consequent sodium excretion in response to isotonic or hypertonic BVE.

Gray et al. [13] have shown that the CeA sends direct projections to the medial and lateral parvocellular subdivisions of the Pa that contain the oxytocinergic and vasopressinergic neurones that project to the median eminence, brainstem and autonomic nervous system. The hypothalamic oxytocinergic neurones project also to the central amygdala [14] , where a high density of oxytocin receptors is found [15] . Thus, the central amygdala has been described as a target structure of oxytocin action within the brain [16] . The electrophysiological effects of oxytocin in the amygdala were demonstrated in studies conducted by Condés‐Lara et al. [17] and Terenzi et al. [18] , who showed that oxytocin may act either as an excitatory neurotransmitter or as a neuromodulator.

The present study aimed to investigate the participation of the oxytocin in the CeA in response to isotonic and hypertonic BVE as measured by: (i) sodium excretion and urinary volumep (ii) plasma levels of oxytocin, vasopressin and atrial natriuretic peptide; and (iii) changes in OT and AVP gene expression, specifically in the Pa and SON of the hypothalamus.

Male Wistar rats (280–350 g) from the Central Animal Facility of the Campus of Ribeirao Preto – University of Sao Paulo were housed in individual stainless steel cages under a 12 : 12 h light/dark cycle (lights on 06.00 h) at 23 ± 2 °C and were given normal food pellets and tap water ad lib. The experiments were performed between 09.00 h and 13.00 h. All procedures for the care and use of the animals were approved by the Committee for Animal Use of the School of Medicine of Ribeirão Preto, University of São Paulo (009/2006).

Rats were anaesthetised with 2,2,2‐tribromoethanol (Sigma‐Aldrich Chemical Co., St Louis, MO, USA; 200 mg/kg of body weight) and placed in a Kopf stereotaxic apparatus (model 900; Kopf Instruments, Tujunga, CA, USA). The skull was levelled between the bregma and lambda. Stainless steel guide cannulas (inner diameter 0.4 mm, outer diameter 0.6 mm) were unilaterally implanted into the CeA using the coordinates: 2.2 mm caudal to bregma, 4.5 mm lateral to the midline and 7.2 mm below the dura mater. The tips of the guide cannulas were targeted to terminate 2 mm above the CeA. The cannulas were fixed to the cranium using dental acrylic resin and two jeweller's screws. A 30‐gauge metal wire filled the cannulas, except during injections. After surgery, the rats received a prophylactic injection of penicillin (20 000 units, i.m) and were allowed to recover for 7 days, during which they were handled daily and habituated to the removal of the 30‐guage wire of the guide cannula and to the gavage procedures.

After the experiments, the animals were deeply anaesthetised with sodium thiopental (80 mg/kg of body weight, i.p.) and received bilateral injections of 2% Evans blue solution (200 ηl/site) into the CeA. The rats were then perfused transcardially with a 0.15 M NaCl solution followed by 10% formalin. The brains were removed, fixed in 10% formalin, frozen, cut into 50‐μm coronal sections, stained with cresyl violet and analysed by light microscopy (Leica microscope equipped with a DC 200 Leica digital camera; Leica Microsystems, Wetzlar, Germany) to locate the injection site in the CeA.

Twenty‐four hours before the experiment, rats that were anaesthetised with 2,2,2‐tribromoethanol (200 mg/kg of body weight, i.p.) had a catheter inserted into their right external jugular vein that was then positioned in the right atrium, as described previously [19] . On the day of the experiment, extracellular volume expansion was performed in conscious, freely‐moving rats by an i.v. injection of hypertonic solution (0.30 M NaCl; 2 ml/100 g body weight) over 1 min.

Urinary sodium concentration was determined by flame photometry (model B 262; Micronal, São Paulo, Brazil). The rates of urinary sodium excretion were calculated by multiplying the urinary concentration by the urine volume and were expressed as μEq/100 g body weight. Urinary volume was expressed as ml/100 g body weight.

The levels of AVP, OT and ANP in plasma were measured by radioimmunoassay as described previously [20] , [21] , [22] . Oxytocin and vasopressin were extracted from 1 ml of plasma with acetone and petroleum ether, and ANP was extracted from 1 ml of plasma using Sep‐Pak C‐18 cartridges (Waters Corporation, Milford, MA, USA). The percentages of recovery after extraction were 83%, 85% and 90% for AVP, OT and ANP, respectively. The assay sensitivity and intra‐ and inter‐assay coefficients of variation were 0.9 pg/ml, 7.7% and 11.9% for AVP; 0.9 pg/ml, 7% and 12.6% for OT; and 7.0 pg/ml, 6.0% and 10.0% for ANP.

Thick sections (1500 μm) of the hypothalamus were obtained from −0.8 to −1.8 mm caudal to bregma (23) using a cryostat. A stainless steel punch needle with a diameter of 1.5 mm was used to collect the Pa and SON, which were immediately transferred to microtubes containing RNAlater reagent (Ambion, Austin, TX, USA) and were stored at 4 °C for a maximum of 24 h before RNA isolation. Total RNA was isolated from each micropunched hypothalamic tissue sample using TRIzol reagent (Invitrogen, AUCKLAND, New Zealand) in accordance with the manufacturer's instructions. The RNA concentration of each sample was determined using an ultraviolet spectrophotometer, and 500 ng of RNA were used for cDNA synthesis using a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). Quantitative real‐time PCR was performed using an Applied Biosystems 7500 real‐time PCR system. The quantitative expression of the AVP and OT genes, as well as of three housekeeping genes, was determined for each cDNA sample by the specific assays Rn00564446_g1 (OT), Rn00566449_m1 (AVP), Rn00667869_m1 (beta‐actin), eukaryotic 18S rRNA endogenous control‐4333760T and Rn99999916_s1 (GAPDH) (Taq‐Man; Applied Biosystems). Determination of the gene transcript levels in each sample was obtained by the ΔΔCT method. For each sample, the threshold cycle (Ct) was determined and normalised to the mean of the housekeeping genes (ΔCt = CtUnknown − CtHousekeeping genes). The fold‐change of mRNA expression in the unknown sample relative to the control group was calculated as 2−ΔΔCt, where ΔΔCt = ΔCtUnknown − ΔCtControl. The mRNA expression for the control group is set at 1 arbitrary unit or 100%, respecting the individual variations, mean and standard error. The experimental groups are then compared in relation to the control group. This estimative is also considered relative because the expression of target genes is expressed in relation to the expression of constitutive genes such as beta‐actin, 18S and GAPDH, whose expression do not differ among groups, independently of the experimental procedure.

One hour after BVE, the animals were deeply anaesthetised with 2.5% tribromoethanol (1 ml/100 g, i.p.) and perfused transcardially with 200 ml of 0.15 M NaCl diluted in diethylpyrocarbonate (DEPC) water followed by 400 ml of 4% paraformaldehyde. The brains were cryoprotected overnight at 4 °C in DEPC‐treated 0.1 M phosphate‐buffered saline (pH 7.4) containing 20% sucrose. The brains were cut (30‐μm sections) in the frontal plane using a freezing cryostat (model HM 5000 M; Microm, LEICA MICROSYSTEMS CMS GMBH Ernst‐Leitz‐Str., Wetzlar, Germany), preserved with cryoprotectant and stored at −70 °C.

We obtained the Pa and the SON from −1.3 mm to −1.80 mm from bregma according to the atlas of Paxinos and Watson [23] . To analyse the distribution of AVP and OT mRNA in the hypothalamus, one series of sections containing the Pa and the SON was collected for each group and was processed for in situ hybridisation. In situ hybridisation for AVP and OT mRNA was performed as described by Rondini et al. [24] .

The OT riboprobe includes positions 37–220 of GenBank accession number NM_012996.3 and the AVP riboprobe includes the exon C constituent of the glycoprotein part of the structure of the AVP gene, GenBank accession number NM_016992.2. The 35S‐labelled RNA riboprobes (AVP and OT) were generated from cDNA templates by in vitro transcription with the appropriate polymerases (T7), in accordance with the manufacturer's instructions (Promega, Madison, WI, USA). Hybridisation with sense probes was performed as a control.

Prior to hybridisation, sections were mounted onto Super‐Frost Plus slides (Fisher Scientific, Pittsburgh, PA, USA) and were pre‐treated with proteinase K (37 °C for 30 min) (Roche, Basel, Switzerland) and then with triethanolamine plus acetic anhydride. The 35S‐labelled probes were diluted (106 d.p.m./ml) in a hybridisation solution that was applied to each slide, and the sections were incubated overnight at 56 °C. The next day, the slides were incubated in 0.002% RNAse A followed by stringency washes. The sections were dehydrated in increasing concentrations of ethanol and delipidated for 15 min in xylenes. After washes in 100% and 95% ethanol, the tissue was air‐dried and the slides were placed in X‐ray film cassettes with BMR‐2 film (Kodak, Rochester, NY, USA) for 2–3 days. The slides were then dipped in NTB photographic emulsion (Kodak; VWR, Radnor, PA, USA) and were stored in foil‐wrapped slide boxes at 4 °C for 3–4 weeks. The slides were developed with Dektel developer (Kodak; VWR), dehydrated in increasing concentration of ethanol, cleared in xylenes and coverslipped with DPX mountant (Fluka, Buchs, Switzerland).

The photomicrographs were captured with a Zeiss Axioplan microscope (Carl Zeiss, Oberkochen, Germany) and qualitatively show the structures of the Pa and the SON that express OT and AVP mRNA. The qualitative analysis considered both the signal strength and the number of labelled cells. The hybridisation signal was estimated by comparing integrated optical density (pixels on a grey scale ranging from 0/black to 255/white) from a constant area. adobe photoshop, version 7.0 (Adobe Systems Inc., San Jose, CA, USA) was used to combine the photomicrographs into plates.

All drugs were purchased from Bachem (Torrance, CA, USA). OT (1 μg, CAT# H‐2510) and the oxytocin antagonist [d(CH2)51,Tyr(Me)2,Thr4,Orn8,des‐Gly‐NH29]‐vasotocin (OVT; 1 μg, CAT# H‐2908) were dissolved in isotonic saline (0.15 m NaCl). The doses of OT and OVT used in the present study were selected based on a previous study [25] . The CeA injections were performed using a 10‐μl Hamilton syringe connected by polyethylene tubing (PE‐10) to an injector needle (outer diameter 0.3 mm). The injector needle was 2 mm longer than the guide cannula. Each rat received only one CeA treatment.

The results are reported as the mean ± SEM. Renal, hormonal and mRNA expression responses were analysed using a two‐way anova, followed by a post‐hoc Newman Keuls test. Variables considered as factors were BVE (no BVE and hypertonic BVE) and treatment (drug injection in the CeA). P < 0.05 was considered statistically significant.

Rats had access to water but not food for 12 h before the experiment. After this period, the animals were weighed and received one intragastric water load (37 °C; 5% body weight) to maintain constant urine flow. The animals then received a unilateral CeA injection of OT (1 μg/0.2 μl), OVT (1 μg/0.2 μl) or isotonic saline (0.2 μl). For these injections, the rats were removed from their home cages and the injection cannula was introduced into the guide cannula. The injection was performed over 60 s. After the injections, the rats were placed in individual metabolic cages without access to food or water. Twenty minutes after the CeA injections, rats were subjected to hypertonic extracellular volume expansion. Two urine samples were collected at 20‐min intervals after the extracellular volume expansion. Complete voiding of urine was manually induced by gently pressing the suprapubic region at the end of each interval. The same CeA treatments were performed in control animals that received one water load without BVE.

In this experiment, hypertonic BVE was performed in freely moving animals that had been injected unilaterally with vehicle, OT or OVT into the CeA as described in Experiment 1. Five minutes after BVE, the unanaesthetised rats were decapitated and their trunk blood was collected into chilled plastic tubes containing heparin (10 μl/ml of blood) for OT and AVP determination, or ethylenediaminetetracetic acid (10 μg/ml) and proteolytic enzyme inhibitors (20 μl 1 mm phenylmethylsulfonyl fluoride and 20 μl 500 μm pepstatin A) for ANP determination. Blood samples were collected 5 min after BVE because the peak hormone levels occur around this time point [26] . The same CeA treatments were performed in control animals without BVE.

Hypertonic BVE was performed in rats that received a unilateral injection of vehicle, OT or OVT into the CeA, as described in Experiments 1 and 2. We used two different methodologies to assess changes in gene expression: quantitative real‐time PCR and quantitative in situ hybridisation. For real‐time PCR, unanaesthetised rats were decapitated 1 hour after BVE and the brains were collected in RNAse free conditions, immediately frozen in dry ice and stored at −80 °C. For in situ hybridisation, anaesthetised rats were transcardially perfused 1 h after BVE with 4% formaldehyde and the brains were sectioned (30 μm) in the frontal plane using a freezing cryostat. The sections were preserved in a cryoprotectant solution and stored at −70 °C. The same CeA treatments were performed in control animals without BVE.

The correct cannula placement (Fig. [NaN] ) was usually centred in the central amygdaloid nucleus in the lateral, medial and capsular portions of the CeA corresponding to −2.80 mm from bregma, in accordance with the coordinates of Paxinos and Watson [23] .

A two‐way anova showed significant main effects of BVE on sodium excretion (F5,38 = 108; P < 0.001) and urinary volume (F5,38 = 49; P < 0.01) (Fig. [NaN] ).

In the rats that were pre‐treated with vehicle in the CeA, hypertonic BVE increased sodium excretion (8.7 ± 0.7 versus control: 0.3 ± 0.6 μEq/100/g/min) (Fig. [NaN] a) and urinary volume (255 ± 16.6 versus control: 83 ± 11 μl/100/g/min) (Fig. [NaN] b). Unilateral pre‐injections of OT (1 μg) into the CeA enhanced the increases in sodium excretion (15.8 ± 1.0 μEq/100 g/min) (Fig. [NaN] a) and urinary volume (340 ± 32 μl/100/g/min) (Fig. [NaN] b).

Rats that were pre‐treated with unilateral injections of OVT (1 μg) in the CeA reduced the increase in sodium excretion (4.9 ± 0.7 μEq/100/g/min) (Fig. [NaN] a) and urinary volume (173 ± 20 μl/100/g/min) in response to hypertonic BVE (Fig. [NaN] b).

Unilateral injections of OT, OVT or isotonic saline into the CeA did not change the sodium and urinary volumes in control rats that were not subjected to hypertonic BVE (Fig. [NaN] ).

In the rats that were pre‐treated with vehicle in the CeA, hypertonic BVE increased the plasma levels of OT (47 ± 2 pg/ml versus no BVE: 5.5 ± 4 pg/ml) (F5,36 = 93; P < 0.001), ANP (306 ± 23 pg/ml versus no BVE: 65 ± 9 pg/ml) (F5,36 = 64; P < 0.001) and AVP (3.9 ± 0.4 pg/ml versus no BVE: 0.9 ± 0.1 pg/ml) (F5,36 = 31; P < 0.001) (Fig. [NaN] ).

In the rats that were pre‐treated with oxytocin in the CeA, hypertonic BVE enhanced the increase in the plasma concentrations of OT (77 ± 5 pg/ml), ANP (424 ± 25 pg/ml) and AVP (6 ± 0.6 pg/ml) (Fig. [NaN] ). An injection of OVT into the CeA decreased plasma ANP (161 ± 20 pg/ml) in rats subjected to hypertonic BVE compared to the conditions of vehicle into the CeA + BVE and OT into the CeA + BVE (Fig. [NaN] b). Additionally, a previous injection of OT antagonist (OVT) into the CeA decreased plasma OT (37 ± 2.5 pg/ml) (Fig. [NaN] a) and AVP (3.1 ± 0.8 pg/ml) compared to OT into the CeA + hypertonic BVE (Fig. [NaN] b).

Unilateral injections of OT, OVT or isotonic saline into the CeA did not change the OT, AVP and ANP levels in control rats that were not subjected to hypertonic BVE (Fig. [NaN] ).

In the rats that were pre‐treated with vehicle in the CeA, hypertonic BVE induced an increase in OT mRNA expression in the Pa (2.5 ± 0.3 versus 1.0 ± 0.05) (F3,16 = 8; P < 0.001) (Fig. [NaN] a) and in the SON (3.0 ± 0.2 versus 1.2 ± 0.2) (F3,16 = 19; P < 0.001) (Fig. [NaN] a). Hypertonic BVE also induced an increase in AVP mRNA expression in the Pa (3.0 ± 0.2 versus 0.9 ± 0.07) (F3,16 = 21; P < 0.001) (Fig. [NaN] a) and in the SON (2.0 ± 0.2 versus 1.0 ± 0.05) (F3,16 = 21; P < 0.001) (Fig. [NaN] a).

The previous administration of OT into the CeA significantly enhanced the OT mRNA expression induced by hypertonic BVE in the Pa (4.1 ± 0.8) (Fig. [NaN] a) and in the SON (4.7 ± 0.5) (Fig. [NaN] a). AVP mRNA expression was also significantly increased by OT treatment into the CeA in the Pa (4.7 ± 0.3) (Fig. [NaN] a) and the SON (4.3 ± 0.4) (Fig. [NaN] a). The previous administration of OVT into the CeA reduced OT mRNA levels in the Pa (1.5 ± 0.3) (Fig. [NaN] a) and in the SON (2.1 ± 0.3) compared to BVE + OT into the CeA (Figs [NaN] a and [NaN] a). In addition, OVT in the CeA reduced AVP mRNA in the Pa (2.3 ± 0.3) and SON (1.9 ± 0.2) compared to BVE + OT into the CeA (Figs [NaN] a and [NaN] a).

Figures [NaN] and [NaN] consist of representative dark field photomicrographs showing the distribution of the hybridisation signals (silver grains) with the 35S‐labelled OT riboprobe in the medial portion of the magnocellular paraventricular nucleus (PaMM) and the medial part of the SON in control rats (Figs [NaN] b and 5b) and in rats subjected to BVE NaCl 0.3 M (Figs [NaN] c and 5c), BVE NaCl 0.3 M plus OT into the CeA (Figs [NaN] d and 5d) and BVE NaCl 0.3 M plus OVT into the CeA (Figs [NaN] e and 5e)

Figures [NaN] and [NaN] consist of representative dark field photomicrographs showing the distribution of the hybridisation signals (silver grains) with the 35S‐labelled AVP riboprobe in the lateral portion of the magnocellular paraventricular nucleus (PaLM) and the medial part of the supra‐optic nucleus (SON) in control rats (Figs [NaN] b and 7b) and in rats subjected to BVE NaCl 0.3 M (Figs [NaN] c and 7c), BVE NaCl 0.3 M plus OT into the CeA (Figs [NaN] d and 7d) and BVE NaCl 0.3 M plus OVT into the CeA (Figs [NaN] e and 7e)

The results of the present study show that sodium excretion and urine output induced by hypertonic BVE were potentiated by oxytocin pre‐treatement in the CeA. It was also observed that an injection of oxytocin into the CeA induced a significant augment in the plasma AVP, OT and ANP levels, as well as significant increases in OT and AVP mRNA expression in the Pa and in the SON after hypertonic BVE, as determined by real‐time PCR. The use of in situ hybridisation (35S‐labelled AVP and OT riboprobes) allowed us to obtain additional information demonstrating specifically which subnuclei of the Pa and the SON (medial) were involved in altering the expression of AVP and OT mRNA.

In the present study, the effects of injecting OT or OVT into the CeA on renal (sodium excretion and urinary volume) and hormonal (plasma ANP, OT and AVP levels) parameters were studied in rats subjected to hypertonic BVE. The observed diuresis and natriuresis are the result of the acute effects of hypertonic BVE, which induced a reduction in sympathetic activity, a direct renal effect of increased plasma sodium that resulted in increased sodium filtration, which, combined with reduced sodium reabsorption caused by OT and ANP action, strongly increased renal sodium excretion and urinary volume.

It is well established that both isotonic and hypertonic BVE induce atrial distension, which increases ANP release from the heart and, by a neuroendocrine reflex, increases oxytocin secretion from neurohypophysial terminals that leads to diuresis and natriuresis [1] , [2] . A previous study showed that the increase in OT secretion induced by isotonic BVE is accompanied by an increase in the number of Fos and Fos‐OT [7] positive cells in the Pa and in the SON. More recently, Ruginsk et al. [27] have shown that hyperosmolality also induces an increase in vasopressinergic and oxytocinergic neurone activity in the Pa and in the SON. It has also been shown by real‐time PCR that hypertonic BVE increases the expression of OT and AVP mRNA in the Pa and the SON [28] . In addition, an anatomical analysis of the increased OT mRNA levels induced by salt loading, determined by quantitative in situ hybridisation, revealed parallel neurone activation in the Pa and the SON, as shown in previous studies [29] , [30] .

Studies using different models of salt appetite in rats have revealed that saline ingestion is present only when the circulating levels of the neurohypophysial hormone OT are supressed [31] . These observations led to the hypothesis that OT pathways participate in the mechanisms of inhibition of salt appetite [32] . Moreover, it has been demonstrated that systemic administration of OT stimulates sodium excretion at physiological concentrations [2] , [33] . Condés‐Lara et al. [17] showed that, in vivo, almost 50% of recorded neurones in the CeA responded to iontophoretic application of OT by an increase in their discharge firing rate, suggesting the presence of functional OT receptors in this area. Indeed, Huber et al. [34] demonstrated that oxytocin receptors are expressed in the lateral and capsular divisions of the CeA. These previous studies indicate that oxytocin is involved in the control of the ingestive and sodium balance responses that restrain the expansion of body fluid volume. The present results provide additional information showing the crucial role for oxytocin binding to its receptor in the CeA, thereby regulating OT and ANP release in response to hypertonic BVE.

In rats that were pre‐treated with an injection of oxytocin in the CeA, there was an increase in the OT and AVP mRNA in the PaMM and PaLM, respectively, after hypertonic BVE. The PaMM subdivision contains mostly oxytocinergic neurones and the lateral magnocellular subdivision (PaLM) contains mostly vasopressinergic neurones, which project to the neural lobe of the pituitary gland. Additionally, in rats that were pre‐treated with OT in the CeA, there was an increase in the levels of OT and AVP mRNA in the medial subdivision of the SON after hypertonic BVE. In the medial part of the SON, OT and AVP are located in the anterodorsal and posteroventral regions, respectively. These results further support the role of oxytocinergic mechanisms in the CeA in modulating not only the systemic, but also the genomic mechanisms related to hypertonic blood volume load. The molecular mechanisms by which OT in the CeA regulates the expression of AVP and OT encoding genes in the Pa and SON are unknown and deserve further study.

Considering the existence of reciprocal projections between the Pa and the CeA [13] , [35] , it has been determined that the oxytocinergic fibres that reach the CeA originate in the paraventricular nucleus [36] , and that the release of oxytocin could occur at axonal terminals or somatodendritically from the Pa to the CeA. Accordingly, the participation of the oxytocin in the CeA as a facilitatory mediator of renal and hormonal responses is confirmed by our data showing an increase in OT and AVP mRNA expression in the Pa and the SON after OT injection in the CeA in response to hypertonic BVE.

With regard to the effects of the unilateral injection of vasotocin into the CeA, it can be speculated that the significantly reduction in plasma ANP levels accounts for the reduction of natriuresis as a consequence of the effect of blocking of OT receptors (by OVT) in the CeA in response to hyperosmolality. Thus, it is reasonable to consider that bilateral CeA treatment with OVT probably would be more specific for inducing more pronounced effects. Although we have not observed significant changes in plasma AVP and OT levels in rats pre‐treated with OVT in CeA compared to vehicle in the CeA, in response to hypertonic BVE, these findings do not argue against the effects of endogenous OT released by the neurohypophysis. Perhaps of greatest importance, our results show that the OT in the CeA plays a key role (among other areas) with respect to integrating a neural circuit that participates in the regulation of the hypervolaemic states induced by hypertonic BVE.

The mechanism by which isotonic or hypertonic volume expansion triggers a neural circuit involving the activation of baroreceptors and volume cardiopulmonary receptors has been reported previously [7] , [10] . Baro‐ and cardiopulmonary receptor activation stimulates brainstem structures, such as the nucleus of the solitary tract (NTS), area postrema (AP) and the locus coeruleus (LC), in response to changes in extracellular volume or changes in arterial pressure. The NTS and the AP activate the LC neurones, which, in turn, send projections to integrative nuclei, such as the NPBL (lateral parabrachial nucleus), NDR (dorsal raphe nucleus) and CeA, which project to the Pa and the SON [37] , [38] . The dorsal raphe nucleus, once activated, could act through direct projections or connections with the CeA to activate the release of oxytocin from the magnocellular neurones in the Pa and in the SON [39] , [40] , [41] . The stimulation of OT secretion could be induced the by activation of intrahypothalamic ANPergic terminals [2] , [42] . These, in turn, could induce the synthesis and central release of ANP that stimulates the neurohypophysis release of OT. Once released into the circulation, OT induces the cardiac release of ANP, leading to diuresis, natriuresis and vasodilatation to reestablish body fluid homeostasis. Natriuresis induced by extracellular volume expansion is mediated by the release of ANP and OT [2] , [43] ; both hormones circulate to the kidney and act on renal specific receptors to reduce sodium reabsorption by a nitric oxide‐cGMP mechanism [2] , [44] . Thus, it can be hypothesised that oxytocin in the CeA may facilitate the release of OT and ANP that act in the kidney to increase electrolyte and water excretion.

We have previously shown [12] that, in rats pre‐treated with the GABAergic agonist muscimol in the CeA, which is rich in GABAA receptors, there was a reduction in plasma OT and ANP hormone levels, as well as in sodium excretion and urine output in response to hypertonic BVE. These responses are similar to the effects of the OT receptor antagonist OVT in the CeA on the BVE and OT induced responses observed in the present study. In view of these results, it can be suggested that oxytocin binds to its receptors (OTR) in the CeA, where OT can restrain the inhibitory GABAergic pathway via GABAA receptors in the magnocellular neuroendocrine system, thereby evoking an increase in the expression of OT and AVP mRNA in the magnocellular neurones of the Pa and SON, as well as increases in their plasma levels in response to hypertonic BVE. The molecular mechanisms and interactions with the receptors that regulate the pre‐ and postsynaptic modulation of magnocellular neuronal activity are not entirely known. The ability of oxytocin to control its own secretion through a neural pathway from the CeA to the Pa expands the complexity of the interactions between GABAergic and magnocellular neurones. In this context, magnocellular neurones show similarities to other neuronal populations by exhibiting autoregulation of neuronal activity by dendritic neuromodulator release [45] . A study that concurs with the hypothesis of the present study was reported by Leng et al. [46] , demonstrating that blockade of GABAA receptors with the antagonist bicuculline, delivered to the dendritic zone of the supraoptic nucleus, enhances oxytocin neurone activity in response to osmotic stimulation. Collectively, these results suggest that GABAergic and oxytocinergic mechanisms act in the CeA to modulate hypothalamic OT and AVP secretion, integrating into a regulatory circuit that controls body fluid homeostasis.

In conclusion, the results of the present study are the first to suggest that the oxytocin in the CeA exerts a facilitatory role in the maintenance of hydroelectrolyte balance, influencing the genomic and secretory neurohypophysial responses to changes in extracellular volume and osmolality.

The authors thank Maria Valci Silva and Rubens de Mello for their excellent technical assistance. This work was supported by grants from FAPESP and UT Southwestern Medical Center, Dallas, TX. Lisandra Oliveira Margatho holds a fellowship from FAPESP (2010/509170).

Graph: ( a ) Photomicrograph of a representative N issl‐stained section showing the central amygdaloid nucleus ( C e A ) corresponding to −2.8 mm from bregma, according to the atlas of P axinos and W atson 23. ( b ) Photomicrograph showing the injection site in the C e A (arrow) at the same level as ( a ). BLA , basolateral amygdaloid nucleus; C pu, caudate putamen/striatum; ic, internal capsule; LH , lateral hypothalamic area; M e A , medial amygdaloid nucleus; opt, optic tract. Scale bar = 3 mm.

Graph: Sodium excretion ( a ) and urinary volume ( b ) in rats pre‐treated with vehicle, oxytocin (1 μg/0.2 μl), or vasotocin (1 μg/0.2 μl) in the central amygdaloid nucleus ( C e A ) and were subjected (or not) to hypertonic blood volume expansion ( BVE) (0.3 M N a C l). The results are expressed as the mean ±  SEM. *P < 0.05 versus no BVE + vehicle; # P < 0.05 versus BVE + vehicle; + P < 0.05 from BVE + OT.

Graph: Plasma levels of oxytocin ( OT ; a ), atrial natriuretic peptide ( ANP ; b ) and vasopressin ( AVP ; c ) in control rats and in rats submitted to hypertonic blood volume expansion ( BVE ) that received unilateral injections of vehicle, oxytocin (1 μg/0.2 μl) or vasotocin (1 μg/0.2 μl) in the central amygdaloid nucleus ( C e A ). The results are expressed as the mean ±  SEM. *P < 0.05 from control + vehicle; # P < 0.05 from BVE + vehicle; + P < 0.05 from BVE + OT.

Graph: Relative oxytocin ( OT) m RNA expression [quantified by real‐time polymerase chain reaction (PCR) ] in the paraventricular nucleus of the hypothalamus ( P a) in rats that were not subjected to hypertonic blood volume expansion ( BVE ) or pre‐treated with a unilateral injection of OT (1 μg/0.2 μl), vasotocin ( OVT ; 1 μg/0.2 μl) or vehicle into the central amygdala ( C e A ) and subjected to hypertonic BVE ( a ). The results are expressed as the means ±  SEM. *P < 0.05 versus no BVE + vehicle; # P < 0.05 versus BVE + vehicle; + P < 0.05 from BVE + OT. ( b – e ) Representative dark field photomicrographs showing the distribution of in situ hybridisation histochemistry (ISHH) signals (silver grains) with a 35S ‐labelled OT riboprobe in the medial magnocellular division of the P a ( P a MM ), corresponding to −1.8 mm from bregma according to the atlas of P axinos and W atson 23. ( b ) Rats not submitted to BVE (n = 4). ( c – e ) Rats subjected to hypertonic BVE pre‐treated with vehicle (n = 4), OT (n = 4), OVT (n = 4) in the C e A , respectively. Scale bar = 100 μm. Note that ISHH provides additional information (qualitative) specifically demonstrating which subnuclei of the P a ( P a MM ) are involved in altering the expression of OT m RNA observed by quantitative RT ‐ PCR of these structures. 3V, third ventricle.

Graph: ( a ) Relative OT m RNA expression (quantified by r eal‐ t ime PCR ) in the supra‐optic nucleus of the hypothalamus ( SON ) in rats not subjected to hypertonic blood volume expansion ( BVE ) or pre‐treated with unilateral injection of oxytocin ( OT , 1 μg/0.2 μl), vasotocin ( OVT , 1 μg/0.2 μl) and vehicle into the central amygdaloid nucleus ( C e A ) and subjected to BVE. The results are expressed as the means ±  SEM. *P < 0.05 versus no BVE + vehicle; # P < 0.05 versus BVE + vehicle; + P < 0.05 from BVE + OT. ( b – e ) Representative dark field photomicrographs showing the distribution of in situ hybridisation histochemistry ( ISHH ) signals (silver grains) with a 35S ‐labelled OT riboprobe in the medial subdivision of the SON , corresponding to −1.8 mm from bregma, according to the atlas of P axinos and W atson 23. ( b ) Rats not submitted to BVE (n = 4). ( c – e ) Rats subjected to hypertonic BVE pre‐treated with vehicle (n = 4), OT (n = 4), OVT (n = 4) in the C e A , respectively. Scale bar = 100 μm. Note that the use of ISHH provides additional information (qualitative) specifically demonstrating which subnuclei of the SON (medial) are involved in altering the expression of OT m RNA observed by quantitative RT ‐ PCR of these structures. OX, optic chiasm.

Graph: ( a ) Relative vasopressin ( AVP ) m RNA expression (quantified by r eal‐ t ime PCR ) in the paraventricular nucleus of the hypothalamus ( P a) in rats not subjected to hypertonic blood volume expansion ( BVE ) or pre‐treated with a unilateral injection of oxytocin ( OT ; 1 μg/0.2 μl), vasotocin ( OVT ; 1 μg/0.2 μl) and vehicle into the central amygdaloid nucleus ( C e A ) and subjected to BVE. The results are expressed as the means ±  SEM. *P < 0.05 versus no BVE + vehicle; # P < 0.05 versus BVE + vehicle; + P < 0.05 from BVE + OT. ( b – e ) Representative dark field photomicrographs showing the distribution of in situ hybridisation histochemistry ( ISHH ) signals (silver grains) with a 35S ‐labelled AVP riboprobe in the lateral magnocellular division of the P a ( P a LM ), corresponding to −1.8 mm from bregma according to the atlas of P axinos and W atson 23. ( b ) Rats not submitted to BVE (n = 4). ( c – e ) Rats subjected to hypertonic BVE pre‐treated with vehicle (n = 4), OT (n = 4) and OVT (n = 4) in the C e A , respectively. Scale bar = 100 μm. Note that the use of ISHH provides additional information (qualitative) specifically demonstrating which subnuclei of the P a ( P a LM ) are involved in altering the expression of AVP m RNA observed by quantitative real‐time PCR of these structures. 3V, third ventricle.

Graph: ( a ) Relative vasopressin ( AVP ) m RNA expression [quantified by r eal‐ t ime polymerase chain reaction ( PCR )] in the supraoptic nucleus of the hypothalamus ( SON ) in rats not subjected to hypertonic blood volume expansion ( BVE ) or pre‐treated with a unilateral injection of oxytocin ( OT ; 1 μg/0.2 μl), vasotocin ( OVT ; 1 μg/0.2 μl) and vehicle into the central amygdaloid nucleus ( C e A ) and subjected to BVE. The results are expressed as the mean ±  SEM. *P < 0.05 versus no BVE + vehicle; # P < 0.05 versus BVE + vehicle; + P < 0.05 from BVE + OT. ( b – e ) Representative dark field photomicrographs showing the distribution of in situ hybridisation histochemistry ( ISHH ) signals (silver grains) with a 35S ‐labelled AVP riboprobe in the medial subdivision of the SON , corresponding to −1.8 mm from bregma, according to the atlas of P axinos and W atson 23. ( b ) Rats not submitted to BVE (n = 4). ( c – e ) Rats subjected to hypertonic BVE pre‐treated with vehicle (n = 4), OT (n = 4) and OVT (n = 4) in the C e A , respectively. Scale bar = 100 μm. Note that the use of ISHH provides additional information (qualitative) specifically demonstrating which subnuclei of the SON (medial) are involved in altering the expression of AVP m RNA observed by quantitative RT ‐ PCR of these structures. OX, optic chiasm.