Estrogen receptors are located in important brain areas that integrate cardiovascular and hydroelectrolytic responses, including the subfornical organ (SFO) and supraoptic (SON) and paraventricular (PVN) nuclei. The aim of this study was to evaluate the influence of estradiol on cardiovascular and neuroendocrine changes induced by hemorrhagic shock in ovariectomized rats. Female Wistar rats (220–280 g) were ovariectomized and treated for 7 days with vehicle or estradiol cypionate (EC, 10 or 40 μg/kg, sc). On the 8th day, animals were subjected to hemorrhage (1.5 ml/100 g for 1 min). Hemorrhage induced acute hypotension and bradycardia in the ovariectomized-oil group, but EC treatment inhibited these responses. We observed increases in plasma angiotensin II concentrations and decreases in plasma atrial natriuretic peptide levels after hemorrhage; EC treatment produced no effects on these responses. There were also increases in plasma vasopressin (AVP), oxytocin (OT), and prolactin levels after the induction of hemorrhage in all groups, and these responses were potentiated by EC administration. SFO neurons and parvocellular and magnocellular AVP and OT neurons in the PVN and SON were activated by hemorrhagic shock. EC treatment enhanced the activation of SFO neurons and AVP and OT magnocellular neurons in the PVN and SON and AVP neurons in the medial parvocellular region of the PVN. These results suggest that estradiol modulates the cardiovascular responses induced by hemorrhage, and this effect is likely mediated by an enhancement of AVP and OT neuron activity in the SON and PVN.
- female rats
- supraoptic nuclei
- paraventricular nuclei
- arterial pressure
- heart rate
the precise regulation of body fluids is essential for the metabolic function of virtually all cells in the body. A variety of mechanisms are activated to maintain plasma osmolality and blood volume within a very narrow range of values (3). For example, hypovolemia and/or hypotension induce vasopressin (AVP) and oxytocin (OT) release from the magnocellular neurons of the supraoptic (SON) and paraventricular (PVN) nuclei in the hypothalamus (3, 56). It is estimated that a decrease of 10–20% in total blood volume induces the release of AVP in several species. This neurosecretory response is modulated by peripheral baroreceptors in the aortic arch and carotid sinus, cardiopulmonary volume receptors, and angiotensin II (ANG II) (50, 65).
The precise role of estrogen in maintaining body fluid homeostasis is not yet fully understood (16). Our group has previously reported that ANG II type 1 (AT1) receptors are involved in the regulation of water and hypertonic saline intake in ovariectomized (OVX) rats during the nocturnal period (37, 38). In addition, several reports have suggested that estrogen has modulatory effects on cardiovascular function, as evidenced by the cardiovascular changes observed in postmenopausal women, OVX rats, and, possibly, females of other species during senescence (12, 46). These effects occur by estrogen's action on the renin-angiotensin, atrial natriuretic peptide (ANP), and other peptidergic systems (33). Thus, it is plausible to assume that investigating the association between electrolyte and cardiovascular homeostasis and estrogen's mechanisms of action may help clarify the controversial cardioprotective effect of estrogen reported in experimental and clinical trials (11).
In a recent review, Somponpun (61) described possible mechanisms by which estrogen regulates electrolyte homeostasis, focusing on AVP and OT. The fact that estrogen receptors (ER) are expressed in key brain nuclei involved in body fluid maintenance strongly suggests a role for estrogen in hydroelectrolytic homeostasis; in rats, ERβ are expressed in magnocellular AVP and OT secretory neurons, and ERα are widely distributed in neurons in the basal forebrain nuclei (53, 55, 57, 58, 59, 60, 61).
Despite the apparently controversial results in the literature, estrogen's effects on cardiovascular and hydroelectrolytic homeostasis seem to be at least partially due to its influence on AVP secretion (55). In women, thirst and AVP secretion induced by osmotic stimuli are diminished during the luteal compared with the follicular phase of the menstrual cycle (67). In postmenopausal women, Forsling et al. (24) observed an increase in basal AVP plasma levels after estrogen therapy. In animal models, Crofton et al. (14) observed no changes in plasma AVP levels in OVX rats treated with estrogen, whereas Skowsky et al. (54) observed an increase in AVP secretion during proestrus compared with diestrus in intact female rats and OVX rats treated with estrogen. However, Crofton et al. (14) observed that estrogen plus progesterone reversed the OVX-induced increase in AVP release, whereas Peysner and Forsling (45) observed a reduction in plasma AVP concentrations after OVX and either an increase (low dose) or reduction (high dose) in AVP levels after estrogen treatment.
Although the influence of OT on reproductive function is well established, knowledge of its effects on electrolyte balance and cardiovascular function is scarce. A positive correlation between estrogen and plasma OT concentrations in young women has been reported previously (2). Furthermore, gonadal steroid hormones have been shown to be required for increases in OT mRNA expression in magnocellular neurons of the SON and PVN in rats subjected to osmotic stimulus (15).
Hemorrhagic hypovolemia is a common clinical condition that occurs as a consequence of trauma, surgeries, gastrointestinal diseases, and anticoagulant therapy (34). Understanding the mechanisms involved in the control of homeostatic changes after hemorrhage is extremely important for the development of new therapeutic approaches. Because estrogen is closely related to neuroendocrine systems, integrating cardiovascular and hydroelectrolytic functions, the aim of this study was to evaluate the influence of estrogen on cardiovascular and neuroendocrine changes induced by hemorrhagic hypovolemia in OVX rats. Thus, we evaluated the effects of estradiol treatment on ANG II, ANP, AVP, OT, and prolactin (PRL) secretion and activation of vasopressinergic and oxytocinergic neurons in the SON and PVN in OVX rats subjected to hemorrhage.
MATERIALS AND METHODS
Female Wistar rats (220–280 g) obtained from the Animal Facility of the campus of Ribeirao Preto, University of Sao Paulo, Brazil, were group-housed (4 animals/cage) under controlled temperature and light conditions (23 ± 2°C, lights on between 0600 and 1800 h) with free access to standard food pellets and tap water. All experiments were performed in the morning (between 0700 and 1100 h).
This study was conducted according to the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85–23, revised 1996), and the experimental protocols were approved by the Ethics Committee on Animal Experiments of the Faculty of Medicine of Ribeirao Preto at the University of Sao Paulo under protocol 041/2009.
All surgical procedures were performed under anesthesia induced by 2.5% 2,2,2-tribromoethanol (250 mg/kg, ip; Sigma) and were followed by prophylactic doses of veterinary pentabiotic (Fort Dodge).
Ovariectomy and estradiol treatment.
Female rats were subjected to bilateral OVX and randomly separated into groups treated with corn oil (OVX-oil) or estradiol cypionate (EC; Pfizer) at 10 (OVX-EC 10) or 40 (OVX-EC 40) μg/kg. Vehicle or EC administration (0.1 ml/rat, sc) started 24 h after OVX and was conducted one time per day for eight consecutive days between 0700 and 1000 h. The efficiency of the OVX procedure and estradiol treatment was confirmed by estradiol plasma concentration and a uterine index (uterus weight/body weight expressed as mg/100 g of body wt) after 8 days of treatment.
We observed a significant increase in plasma estradiol concentrations in a dose-dependent manner in the OVX-EC 10 and OVX-EC 40 groups compared with the OVX-oil group (72.2 ± 6.1 and 201.4 ± 30.4 vs. 21.6 ± 2.0 pg/ml, N = 8, 10, and 7, respectively, P < 0.05). We also verified a significant increase in the uterine index in the OVX-EC 10 and OVX-EC 40 groups when compared with the OVX-oil group (216 ± 9.0 and 367.0 ± 23.5 vs. 53.5 ± 6.5 mg/100 g body wt, N = 15, 14, and 15, respectively, P < 0.001) (37, 38). These results confirm the efficiency of estrogen treatment in reestablishing estradiol plasma levels after ovariectomy and the trophic uterine index. Additionally, published data in the literature support the evidence that 10 μg/kg of EC is sufficient for estrogen replacement as a physiological dose (similar to estradiol concentrations found in proestrus rats) and 40 μg/kg as a supraphysiological dose (62). However, because of different doses of estradiol used, several reports have found diverse responses mediated by this steroid (24, 27, 45). For this reason, we chose two different doses of estradiol in this study.
Femoral artery cannulation.
For the hemorrhage procedure and cardiovascular recordings, animals were subjected to right femoral artery cannulation; a polyethylene cannula (PE-10 connected to PE-50, Intramedic; Becton-Dickinson, Sparks, MD) was inserted and then externalized in the dorsal cervical region. Soon after, the cannula was flushed with isotonic saline containing heparin (100 IU/ml Liquemine; Roche) to prevent obstruction. Animals were allowed to recover for 24 h before the experimental procedure.
Fifteen minutes before the hemorrhage procedure, isotonic saline containing heparin (150 IU in 150 μl) was administered through the arterial catheter to prevent blood clotting during blood removal. The hemorrhage procedure was achieved by blood removal through the arterial catheter (15 ml/kg body wt, corresponding to ∼25% of total blood volume) for 1 min. False hemorrhage was performed using the same protocols without blood withdrawal.
For mean arterial pressure (MAP, mmHg) and heart rate (HR, beats/min) assessment, the arterial catheter was connected to a pressure transducer (P23Gb; Statham Instruments, Hato Hey, Puerto Rico) and data acquisition system (Windaq/200; Dataq Instruments). After 20 min of adaptation, baseline data were collected for 10 min, and the blood withdrawal was then initiated, followed by monitoring of hemodynamic changes for 30 min. The data represent several data points from a continuous recording.
Plasma Hormone Determination
For plasma hormone measurements, animals were decapitated, and the blood was collected from the trunk at 0 (baseline), 5, 15, and 30 min after hemorrhage induction in chilled tubes containing heparin (for estradiol, AVP, OT, and PRL) or peptidase inhibitors (for ANG II and ANP). Plasma was obtained after centrifugation (20 min, 3,000 rpm, 4°C) and stored at −20°C until specific extraction and radioimmunoassay procedures.
The radioimmunoassay for estradiol was performed with a commercial kit from Diagnostic System Laboratories (DSL-4400; Webster, TX). The specific antibodies for ANG II, AVP, and OT radioimmunoassay were obtained from Peninsula (ANG II, T4007; AVP, T4561; and OT, T4084; San Carlos, CA). PRL was obtained from the National Institute of Diabetes and Digestive and Kidney Diseases (PRL AFP131581570; Baltimore, MD), and ANP was donated by Jolanta Gutkowska (Hotel Dieu, Univ. of Montreal, Montreal, Quebec, Canada). The radioimmunoassay sensitivity and intra- and interassay coefficients of variation were 20 pg/ml, 1.4–27.1% for estradiol; 0.5 pg/ml, 10.9–17.1% for ANG II; 0.7 pg/ml, 4.8–10.0% for ANP; 0.8 pg/ml, 7.7–11.9% for AVP; 0.9 pg/ml, 7.0–12.6% to OT; and 0.9 ng/ml, 5.2–11.8% for PRL.
Perfusion, Tissue Preparation, and Immunohistochemistry
Ninety minutes after hemorrhage induction, the anesthetized animals were transcardially perfused with 200 ml of isotonic saline containing heparin (50 IU/l) followed by 400 ml of 4% paraformaldehyde in 0.1 M phosphate buffer (PB), pH 7.2. The brains were removed, fixed for 4 h in the perfusion solution, and stored at 4°C in PB containing 30% sucrose.
Coronal sections of 30-μm thicknesses were obtained using a cryostat (Micron) and collected in 0.01 M PB. Briefly, sections were incubated with H2O2 solution (0.03%) for 30 min and washed in 0.01 PB. Sections were then incubated with 5% bovine albumin in 0.1 M PB for 1 h. Sections were subsequently incubated at room temperature for 12–14 h with the primary anti-Fos antibody (1:10,000, produced in rabbits, Ab-5; Oncogene), washed, and then incubated with a biotin-labeled antibody (1:200, produced in goats; Vector Laboratories, Burlingame, CA) for 1 h. The avidin-biotin-peroxidase complex (ABC) was used for staining (Vector Laboratories), followed by 0.5% diaminobenzidine hydrochloride (DAB; Sigma Chemical, St. Louis, MO) intensified with 5% cobalt chloride and 1% nickel ammonium sulfate, which labels the cell nuclei black.
For double labeling, after concluding the Fos protocol described above, sections were incubated with anti-OT or anti-AVP (1:10,000, both produced in rabbits; Peninsula) for 48 h at 4°C. Thereafter, sections were washed and subjected to the protocol described for Fos labeling, using appropriate secondary biotinylated antibodies followed by the ABC. The brown cytoplasmatic color was detected using a nonintensified DAB solution. Finally, the sections were mounted on gelatinized slides, air-dried overnight, dehydrated, cleared in xylene, and placed under a cover slip with Ethelan.
Immunostained cells were quantified using a Leica microscope equipped with a DC 200 Leica digital camera and coupled to a computer using Leica Application Suite software. Visual counting was performed unilaterally in one section per animal and was repeated at least two times for each analyzed section.
The PVN and SON were identified and delimited according to the Paxinos and Watson atlas (43). The subfornical organ (SFO) and SON were examined on their medial portions (−0.84 and −1.08 mm from the bregma, respectively). The PVN was counted in the following four different areas: medial magnocellular (PaMM, −1.44 mm from the bregma), lateral magnocellular (PaLM, −1.72 mm from the bregma), medial parvocellular (PaMP, −1.72 mm from the bregma), and posterior parvocellular (PaPO, −2.04 mm from the bregma). The schematic brain areas are shown in Fig. 1.
Results are expressed as means ± SE. For statistical purposes, estradiol results below the detection limit of the assay were assigned the value of the detection limit. Statistical analyses were conducted using one-way or two-way ANOVA (factors: time or hemorrhage and treatment) followed by the Bonferroni posttest. The level of significance was set at P < 0.05.
Protocol 1: Effects of estradiol on MAP and HR in OVX rats submitted to hemorrhage.
Animals were connected to the acquisition system, and, after 20 min of adaptation, baseline parameters were measured for 10 min. Next, blood withdrawal was performed, animals were reconnected to the system, and recording of the cardiovascular parameters resumed. Our results show the values of MAP and HR at 0 (baseline), 5, 15, and 30 min after hemorrhage induction.
Protocol 2: Effects of estradiol on hemorrhage-induced hormone secretion in OVX rats.
Blood collected during hemorrhage induction (basal) and by decapitation (5, 15, or 30 min after hemorrhage) was used for radioimmunoassay measurements of ANG II, ANP, AVP, OT, and PRL plasma levels.
Protocol 3: Effects of estradiol on SFO, SON, and PVN neuronal activation in OVX rats submitted to hemorrhage.
Ninety minutes after hemorrhage or false-hemorrhage induction, animals were anesthetized and subjected to brain perfusion for subsequent immunohistochemical procedures. Our results show the absolute number of Fos-positive neurons in the SFO, SON, and PVN subdivisions and the percentage of double-labeled neurons for Fos/AVP and Fos/OT in the SON and PVN subdivisions.
Protocol 1: Effects of Estradiol on MAP and HR in OVX Rats Subjected to Hemorrhage
Basal MAP was similar among the OVX-oil (110.2 ± 2.4 mmHg), OVX-EC 10 (107.0 ± 3.1 mmHg), and OVX-EC 40 (104.3 ± 3.1 mmHg) groups. However, we observed an estradiol effect on HR in basal conditions [F(2,18) = 4.3; P < 0.05]. Estradiol treatment induced a significant reduction in basal HR in the OVX-EC 10 (385.6 ± 20.5 beats/min, P < 0.01) and OVX-EC 40 (395.1 ± 12.5 beats/min, P < 0.01) groups compared with the OVX-oil group (452.3 ± 18.2 beats/min).
Figure 2 shows the difference (Δ) in basal MAP (Fig. 2A) and HR (Fig. 2B) at 5, 15, and 30 min after hemorrhage in all groups. There was a significant effect of time on ΔMAP [F(8,54) = 5.2; P < 0.01] and ΔHR [F(8,54) = 6.8; P < 0.01]. We also observed an effect of treatment (oil, EC 10, and EC 40) on ΔMAP [F(8,54) = 7.2; P < 0.01] and ΔHR [F(8,54) = 3.7; P < 0.05]. Furthermore, there was an interaction between treatment and time on ΔMAP [F(8,54) = 5.2; P < 0.01] but not ΔHR.
In OVX-oil rats, hemorrhage induced a decrease in MAP (89.7 ± 3.9 vs. 110.2 ± 2.4 mmHg, P < 0.01) and HR (400.4 ± 11.3 vs. 452.3 ± 18.2 beats/min, P < 0.05) at 5 min compared with basal values. However, estradiol treatment abolished the hypotension 5 min after hemorrhage in OVX-EC 10 (P < 0.001) and OVX-EC 40 (P < 0.001) rats compared with the OVX-oil group. Likewise, EC 10 and EC 40 treatment also abolished (P < 0.05) the HR decrease response induced by blood withdrawal compared with the oil-treated rats simultaneously with hemorrhage.
Protocol 2: Effects of Estradiol on Hemorrhage-Induced Hormone Secretion in OVX Rats
Plasma concentrations of ANG II (Fig. 3A) and ANP (Fig. 3B) are shown in Fig. 3. We observed an effect of time on plasma ANG II [F(11,99) = 14.6; P < 0.001] and ANP [F(11,98) = 12.5; P < 0.001] concentrations. There was no interaction between treatment and time after hemorrhage on these parameters.
In the OVX-oil group, hemorrhage induced a significant increase in plasma ANG II at 5 (P < 0.01), 15 (P < 0.01), and 30 (P < 0.01) min after the procedure compared with basal conditions. However, hemorrhage decreased plasma ANP concentrations at 5 (P < 0.01), 15 (P < 0.01), and 30 (P < 0.01) min in the OVX-oil rats compared with basal conditions. Estradiol treatment (both doses) did not alter ANG II or ANP secretion compared with oil treatment.
Plasma AVP (Fig. 4A), OT (Fig. 4B), and PRL (Fig. 4C) concentrations are shown in Fig. 4. There was an effect of time on plasma AVP [F(11,102) = 29.3; P < 0.001], OT [F(11,102) = 13.4; P < 0.001], and PRL [F(11,106) = 39.2; P < 0.001] concentrations. Plasma PRL basal levels were also increased by estradiol treatment [F(11,106) = 30.7; P < 0.001]. Finally, we observed a significant interaction between estradiol treatment and time after hemorrhage on AVP [F(11,102) = 3.2; P < 0.01], OT [F(11,111) = 3.2; P < 0.01], and PRL [F(11,106) = 7.8; P < 0.01] secretion.
In OVX-oil rats, hemorrhage induced an increase in plasma AVP concentrations at 5 (P < 0.001), 15 (P < 0.001), and 30 (P < 0.05) min postprocedure compared with basal conditions. Five minutes after hemorrhage, EC 10 and 40 treatments potentiated the AVP secretion induced by hemorrhage compared with the OVX-oil group (P < 0.001). Hemorrhage also induced an increase in plasma OT levels at 5 (P < 0.05), 15 (P < 0.05), and 30 (P < 0.05) min in the OVX-oil rats compared with baseline. However, only the EC 40 treatment enhanced OT secretion 5 min after hemorrhage (P < 0.05) compared with the OVX-oil group.
We also observed an increase in plasma PRL concentrations 5 min after blood withdrawal (P < 0.05) in OVX-oil rats compared with baseline. At both 5 and 15 min after blood withdrawal, PRL secretion was strongly potentiated by EC 10 (5 min, P < 0.001; 15 min, P < 0.05) and EC 40 (5 min, P < 0.001; 15 min, P < 0.001) compared with the OVX-oil group.
Protocol 3: Effects of Estradiol on SFO, SON, and PVN Neuron Activation in OVX Rats Submitted to Hemorrhage
Estradiol effects on hemorrhage-induced Fos in SFO and Fos/AVP or OT neuronal activation in the PVN and SON are summarized in Table 1. Hemorrhage induced a significant increase in the number of Fos-labeled neurons in the SFO, SON, and PVN subdivisions and the percentage of Fos/AVP and Fos/OT double-labeled neurons in the PVN subdivisions and SON of the OVX-oil rats. The EC 10 and 40 treatments induced a significant increase in Fos expression in the SFO nucleus induced by hemorrhage in OVX rats. Additionally, after hemorrhage, EC 10-treated rats showed a significant increase in the number of Fos-positive neurons in the PaLM and PaPo and in Fos/AVP double-labeled cells in the SON and PaMM compared with the OVX-oil group. The results also showed that hemorrhage increased the number of Fos-immunoreactive neurons in the SON and all PVN subdivisions in OVX-EC 40 rats compared with the OVX-oil group. Furthermore, the higher EC dose potentiated the percentage of Fos/AVP double-labeled neurons in the SON and PaMM, PaLM, and PaMP subdivisions of the PVN after blood withdrawal compared with oil treatment. EC 40 treatment also induced a significant increase in the percentage of Fos/OT double-labeled neurons in the SON and magnocellular PVN of hemorrhagic rats compared with the OVX-oil group.
The representative photomicrographs from SFO showing Fos-labeled neurons, SON showing Fos/AVP double-labeled neurons, and the lateral magnocellular PVN showing Fos/OT double-labeled neurons in sham or hemorrhage rats pretreated with oil or EC are shown in Figs. 5, 6, and 7, respectively.
In the present report, we observed that blood withdrawal reduced MAP and HR only at 5 min in the OVX-oil group. Hypotension induced by mild to moderate hemorrhage is usually accompanied by a tachycardic reflex response (29); however, we used moderate to severe hemorrhage in the present study, which is known to produce a bradycardic response as a consequence of the vasovagal reflex, i.e., an acute central hypovolemia with a consequent parasympathetic activation (1, 20, 40, 48). The bradycardic reflex induced by parasympathetic activation is likely to involve activation of the periaqueductal gray substance and serotoninergic system that sends inhibitory projections to the rostral ventrolateral medulla (RVLM), a crucial area implicated in the generation and modulation of sympathetic activity to the cardiovascular system (7, 9, 49, 51).
We observed an inhibition of the hypotension and bradycardic responses induced by hemorrhage in OVX-EC rats compared with the OVX-oil group. The sex steroid hormones are required for changes in baroreflex sensitivity, as shown by the changes that occur in the cardiovagal baroreflex sensitivity and renal sympathetic nerve activity during the estrous cycle of female rats that are abolished by OVX (25). Additionally, several studies have demonstrated the modulation of hypotensive responses by estrogen, although the data are controversial. Estrogen attenuated the hypotensive response after the administration of a β-adrenergic agonist (isoproterenol) but did not influence the tachycardic reflex response (32). However, the recovery of MAP and HR after hemorrhage in Long Evans female rats was impaired by estrogen. This did not occur in AVP-deficient animals (40). Our results contrast with those obtained by these authors, possibly because of methodological differences in time and total blood volume removal. Women in the postmenopausal period and during OVX, rats present an increase in sympathetic activity with higher α1-adrenergic activation and peripheral vasoconstriction associated with a reduction in nitric oxide synthesis, possibly because of estrogen deficiency (68). As a result, we suggest that both the lower basal HR in OVX-EC rats and attenuation of cardiovascular responses induced by hemorrhage in the OVX-EC group found in the present study could be a consequence of estrogen effects on autonomic activity.
Plasma ANP concentrations were reduced in all groups after blood withdrawal in accordance with well-established data in the literature demonstrating a decrease in its concentrations in response to hypovolemia (69). However, despite previous demonstrations of the colocalization of estrogen and ANP receptors in cardiomyocytes and ANP secretion induced by estrogen perfusion in atrial cardiac cells, we did not observe an influence of estrogen on ANP secretion under basal or hemorrhagic conditions (4, 19). Although estrogen was shown to modulate ANP secretion in response to osmolality changes and extracellular volume expansion (66), it seems that the reduction of ANP release during hemorrhage is not influenced by estrogen treatment.
We also showed an important increase in plasma ANG II concentrations in all groups after the hemorrhage procedure. Systemic ANG II interacts with AT1 receptors located in the vascular smooth muscle, leading to its contraction and consequently elevating arterial pressure, an important mechanism in reestablishing tissue perfusion after blood withdrawal (17). Furthermore, circulating ANG II also acts on the central nervous system (CNS) structures devoid of blood-brain barrier, such as the SFO, contributing to the arterial pressure maintenance (17, 36). Estradiol had no effect on the ANG II increase induced by hemorrhage.
It is possible that blood loss may not be the best experimental paradigm to evaluate the effects of estrogen on ANP secretion and RAS activity because hemorrhage is a potent stimulus for RAS and inhibits ANP secretion. Therefore, the potent neuroendocrine stimulation or inhibition may overcome the modulatory effects of estradiol on these systems.
Our results show an increase of Fos-immunoreactive neurons in the SFO after hemorrhage in OVX-EC-treated animals. Because there was no difference between plasma ANG II of OVX-oil and OVX-EC groups after hemorrhage and because it is well established that estrogen treatment reduces AT1 mRNA expression (18), we suggest that the increase in SFO neuronal firing by estrogen is not related to the enhancement of ANG II signaling. The SFO integrates the information from other areas that participate in the control of cardiovascular function with those from plasma ANG II concentrations and retransmits this information to the PVN and SON, a possible pathway through which plasma ANG II at least partially modulates AVP and OT secretion (64). The SFO receives (and sends) important projections from (and to) areas involved with hydroelectrolytic and cardiovascular balance, including the PVN, SON, RVLM, and nucleus tractus solitarius (23). The influence of estrogen on SFO neuronal activation could be exerted by direct action on the ERα in the SFO, decreasing the threshold activation of these neurons or acting through the above-mentioned areas to increase stimulatory inputs to the SFO.
As expected, plasma AVP concentration was increased 5 min after blood withdrawal in all groups evaluated, and this response is mediated by baroreceptors, volume receptors, and increased plasma ANG II concentrations (3). Interestingly, there was an increase in AVP secretion induced by hemorrhage in the OVX-EC rats, which is in accordance with previous studies observing a similar estrogen effect in rats and women both in basal conditions and after an osmotic or hypovolemic stimulus (24, 27, 45, 54).
Hypovolemia induced by hemorrhage increases AVP and OT plasma levels and the number of Fos-positive magnocellular neurons in the PVN and SON (8, 52). This response is likely modulated by estrogen because it changes the number of Fos-labeled cells in these nuclei observed in response to a variety of stimuli (27, 31). Estrogen also attenuates hypotension and the number of Fos-positive cells in the area postrema and lateral parabrachial nucleus, areas closely related to these neuroendocrine responses (32).
During hemorrhage, vasomotor brain nuclei relay peripheral information to the PVN and SON, culminating with the marked magnocellular vasopressinergic neuron activation seen in the present study and subsequent enhancement of plasma AVP levels (6). It has been demonstrated that exogenous administration of AVP induces significant bradycardia in AVP-deficient rats, suggesting that the endogenous hormone may contribute to the vagally mediated bradycardia induced by hemorrhage (28). Peuler et al. (44) demonstrated that sympathetic nerve activity and HR during hemorrhage were consistently higher in diabetes insipidus rats compared with normal Long-Evans or AVP-treated rats both before and after vagotomy. In addition, these authors show that vagotomy attenuated the inhibitory effect of AVP on the HR response but not renal sympathetic nerve activity following hemorrhage in the same experimental model of AVP-deficient animals. Taken together, these results suggest that AVP may play a major role in the cardiovascular adjustments following hemorrhage.
In the present report, we observed an increase in OT and AVP magnocellular neuronal activation, assessed by the number of neurons double labeled for Fos/OT and Fos/AVP and hormone plasma concentrations in response to hemorrhage. Previous studies have demonstrated that AVP and OT could present negative chronotropic effects mediated by cardiac V1 receptors (13, 22). However, the estrogen effect attenuating the bradycardic response seems not to be related to OT and AVP plasma concentrations because estrogen potentiated hormone secretion after hemorrhage.
We also report a potentiation of AVP magnocellular neuronal activation mediated by estrogen following hemorrhage. Previous reports have already shown estrogen effects on AVP secretion and SON neuronal activation in female rats subjected to hemorrhage. However, they did not demonstrate the neuronal phenotype involved in this response, which could be AVP- or OT-producing cells (27). Because PVN and SON neurons express ERβ, the modulation exerted by estrogen on AVP magnocellular activity and AVP secretion could be produced by a direct effect on the hypothalamic neuroendocrine system or mediated by the activation of other central areas controlling plasma volume and arterial pressure, such as the SFO (53, 55, 57, 58, 59, 60, 61). Thus, the increase of AVP secretion mediated by estrogen after blood withdrawal is consistent with the attenuation of hypotension mediated by estrogen observed 5 min after the hemorrhage.
Additionally, we demonstrated that parvocellular PVN neurons, mostly vasopressinergic neurons, were activated in response to hemorrhage, as previously described (5). Hypotensive hemorrhage culminates in the simultaneous activation of AVP- and corticotropin-releasing hormone parvocellular neurons in the PVN, and it increases adrenocorticotropin hormone secretion (8). Parvocellular PVN neurons maintain reciprocal communication with other CNS areas related to the integrated control of cardiovascular and hydromineral balance. The increased activation of vasopressinergic parvocellular PVN neurons could indicate that AVP may act as a neurotransmitter and/or neuromodulator in areas related to arterial pressure control in hemorrhagic conditions (3, 8, 43, 56, 52). Furthermore, we also demonstrated that estrogen treatment potentiated AVP neuronal activation in the medial parvocellular PVN.
In our study, we observed a significant increase in magnocellular oxytocinergic activation in the PVN and SON and a concomitant increase in plasma OT concentrations following blood withdrawal, as previously described in response to hypotension and/or hypovolemia (56). In addition, we demonstrate an intense parvocellular oxytocinergic neuronal activation induced by hemorrhage, which could indicate some physiological role for OT release as a neurotransmitter in hypovolemia and/or hypotension conditions. Treatment with estrogen potentiated the effects of hemorrhage on the activation of OT magnocellular neurons in the PVN, which corresponded with an enhancement of OT secretion in OVX-EC rats. However, estrogen did not change the response induced by hemorrhage on the number of Fos/OT double-labeled cells in the parvocellular groups, suggesting that estrogen acts selectively on AVP parvocellular neurons from the PVN in hypovolemic hemorrhagic conditions.
The potent increase in AVP and OT secretion observed in response to an acute hemorrhage induced a simultaneous reduction of hormone content in the SON and PVN under the same experimental conditions in male rats (47). Previous reports have also demonstrated the increased secretion of PRL after hemorrhage in male rats (35).
Some studies have shown a positive correlation between stress intensity, OT hypothalamic synthesis, and its release to systemic circulation, which was not found for AVP (21). Furthermore, PRL, a stress-related hormone in rodents, was increased by hemorrhage in the present study. This evidence also supports the hypothesis that the observed increase in OT secretion following hemorrhage could be a nonspecific stress response rather than an effect induced by hypotension or decreased circulating volume (30). Our study also showed that estrogen induces a further increase in OT magnocellular neuronal activation and OT and PRL secretion in response to hemorrhage, which corroborates previous findings showing that estrogen can potentiate some effects demonstrated by stress models (39).
Somponpun and Sladek (60) showed a clear influence of hypovolemia on reducing ERβ expression in the PVN and SON. However, these authors did not perform acute challenges because ERβ expression was evaluated 8 h after polyethylene glycol administration plus water deprivation or 20 and 26 h after water deprivation. In our study, we found a potentiation of AVP and OT secretion mediated by estrogen (5 min after hemorrhage) and the double labeling of Fos/AVP and Fos/OT 90 min after hemorrhage. Thus, it is important for future studies to verify the ability of an acute hypovolemic stimulus and its time-dependent profile in altering the expression of ERβ. However, since hormonal evaluations were performed 5 min after hypovolemia in the present study, it is unlikely that a change in ERβ expression induced by hemorrhage is responsible for these observed responses. Instead, this response could be mediated by estrogen signaling through ERβ in the PVN and SON and/or β- and α-receptors located in other important brain areas for cardiovascular maintenance.
Nomura et al. (41) reported a decrease in AVP transcript levels induced by estrogen, which was abolished in ERβ knockout mice. Thus, we suggest that the potentiating effect on AVP secretion and PVN and SON neuronal activation induced by estrogen may not be related to ERβ signaling. In a recent report, Grassi et al. (26) demonstrated that the increase in AVP immunoreactivity in the PVN and SON induced by estrogen was not due to ERβ signaling, but it could be mediated by ERα. Because magnocellular neurons in the PVN and SON do not express ERα, this response (as in our experimental design) could be mediated by afferent ERα-expressing neurons from different areas, such as the SFO.
Estrogen has important genomic actions regulating the expression of several genes related to cellular excitability, including the synthesis of neurotransmitters, expression of ion channels, and activation of intracellular cascades (10). Therefore, we hypothesize that estrogen, through its genomic and nongenomic actions, can change the excitability threshold of circuits related to AVP and OT secretion. However, treatment with estrogen has been shown to decrease the expression of ERβ in the PVN (42). Moreover, ERβ is mainly expressed in AVP neurons, whereas <7–10% of ERβ reside in magnocellular OT cells in both male and female rats (58, 63). However, we observed an estrogen-mediated potentiation of OT release and neuronal activity after hemorrhage compared with the oil-treated OVX group. In this context, Nomura et al. (41) demonstrated an increase in OT mRNA levels in the PVN after estrogen therapy that was abolished in ERβ knockout mice, suggesting a positive effect of ERβ signaling in OT neurons. Additionally, inhibitory afferents (i.e., activation of baroreceptors) also participate in the tonic control of OT and AVP secretion (via the brain stem, which express ERα and ERβ) (53). Together with the potentiation of SFO neuronal activation induced by estrogen, these data led us to again hypothesize that estrogen could also act in other brain areas potentiating stimulatory pathways and/or decreasing inhibitory pathways to the PVN and SON, thus acting on the AVP and OT.
Significance and Perspectives
Taken together, our results suggest that estrogen modulates some cardiovascular and neuroendocrine responses induced by acute hemorrhage, and the data provide valuable evidence on the influence of estrogen on body fluid homeostasis. Furthermore, hemorrhage is a very common condition in medical clinics, and our results discriminate between several cardiovascular and neuroendocrine responses in OVX rats receiving or not receiving estrogen therapy after blood loss, which could give new insights into hemorrhage management during the postmenopausal and/or postpartum period. Further studies are needed to investigate the role of ERβ on AVP and OT secretion and the importance of estrogen signaling in other brain areas in the acute hypovolemia model performed in this study.
This work was supported by the Fundação de Amparo a Pesquisa do Estado de São Paulo, Brazil, and the Conselho Nacional de Desenvolvimento Científico e Tecnológico.
No conflicts of interest, financial or otherwise, are declared by the authors.
We thank Rubens Fazan, Jr., for kindly providing equipment for the cardiovascular assessments and Maria Valci dos Santos Silva, Milene Mantovani, Rubens Fernando de Mello, and Carlos Alberto A. Silva for excellent technical assistance.
- Copyright © 2011 the American Physiological Society