This study examined the effects of ovariectomy (OVX) and 17β-estradiol (E2) replacement (OVX + E2) on renal function in Sprague-Dawley rats. OVX caused a 40% decrease in the fractional excretion of potassium (FEK+) that was prevented by E2 replacement [Sham, 24.2 ± 2.9%; OVX, 14.5 ± 2.1% (P < 0.05 vs. OVX + E2); and OVX + E2, 26.2 ± 2.7%; n = 7–11] and that corresponded to significant increases in plasma potassium [(in mmol/l): Sham, 3.15 ± 0.087; OVX, 3.42 ± 0.048 (P < 0.05 vs. OVX + E2); and OVX + E2, 3.19 ± 0.11; n = 7–11]. No effects of OVX were detected on plasma levels of sodium and aldosterone. Angiotensin II type 1 receptor (AT1R) densities in ovariectomized rats were 1.4-fold and 1.3-fold higher in glomerular [maximum binding capacity (Bmax; in fmol/mg protein): Sham, 482 ± 21; OVX, 666 ± 20 (P < 0.05 vs. OVX + E2); and OVX + E2, 504 ± 26; n = 7–11] and proximal tubular [Bmax (in fmol/mg protein): Sham, 721 ± 16; OVX, 741 ± 24 (P < 0.05 vs. OVX + E2); and OVX + E2, 569 ± 23; n = 7–11] membranes compared with E2 replete animals, respectively. Both the angiotensin-converting enzyme inhibitor captopril and the AT1R antagonist losartan prevented the OVX-induced decrease in the FEK+ and the increase in renal AT1R densities, suggesting that E2 deficiency reduces potassium excretion in an ANG II/AT1R-dependent manner. These findings may have implications for renal function in postmenopausal women as well as contribute to the reasons underlying the age-induced increase in susceptibility to hypertension-associated disease in women.
- renal cortex
- mean arterial pressure
the renin angiotensin aldosterone system plays a critical role in controlling blood pressure and electrolyte homeostasis through its effects in key target tissues including the adrenal, kidney, vasculature, and brain (29, 31). Angiotensin II (ANG II) acting on the type 1 angiotensin receptor (AT1R) induces aldosterone release from the adrenal gland. Subsequently, aldosterone acts on apical sodium channels and the sodium-potassium-ATPase in basal membranes of renal epithelial cells to cause sodium reabsorption (23). Aldosterone also causes an increase in the potassium gradient across the apical membrane causing potassium excretion into the urine.
We have shown that 17β-estradiol (E2) modulates adrenal responsivity to ANG II. AT1R densities in the adrenal cortex and aldosterone release in response to acute and chronic infusions of ANG II are significantly higher in ovariectomized rats compared with ovariectomized rats replaced with E2 (24, 35). Other investigators have shown that ovariectomized rats exhibit increased ANG II-induced vascular contractility in aortic rings, which is prevented by E2 replacement (17, 25).
There is evidence from human studies that E2 also modulates ANG II actions in the kidney. A report by Miller et al. (16) showed that the effective renal plasma flow (ERPF) response to ANG II infusion inversely correlated with plasma levels of E2: the higher the level of plasma E2, the smaller the magnitude of the ANG II-induced decrease in ERPF, and Chidambaram et al. (3) reported that renal and peripheral hemodynamic responses to ANG II blockade were blunted during the luteal phase of the menstrual cycle. In this study, we investigated the effects of E2 on renal function in the normal female Sprague-Dawley rat and how these E2-mediated effects are modulated by an inhibitor of ANG II synthesis and the AT1R antagonist.1
Female Sprague-Dawley rats weighing 250–300 g were purchased from Harlan (Indianapolis, ID) and individually housed in a temperature-controlled animal facility. All rats were maintained on a phytoestrogen-free diet (Harlan) and tap water ad libitum under controlled conditions (12-h:12-h light-dark schedule at 24°C). All procedures were approved by the Georgetown University Animal Care and Use Committee.
Under methoxyflurane anesthesia, bilateral incisions were made. The vascular supply was ligated, and the ovaries were then removed. The muscle layer was sutured, and the incisions were closed with wound clips. In the sham-operated rats, the animals were subjected to surgery and the ovaries were manipulated but left intact.
E2 benzoate (2.5 μg E2/kg body wt) was dissolved in 200 μl peanut oil and injected subcutaneously every day for 15 days (unless otherwise indicated); sham-operated and ovarectomized animals were injected subcutaneously with vehicle (0.2 ml peanut oil) as previously described (13). Losartan (0.1 g/l) and captopril (0.5 g/l) were delivered in the drinking water for the duration of the experiments as previously described (35).
Plasma electrolytes and hormone assays.
Blood samples (0.5–1 ml) were collected in 3-ml Vacutainer blood collection tubes (Becton Dickinson) containing heparin sodium, and the plasma was separated by centrifugation before the renal function experiments were initiated. Plasma sodium and potassium were determined by flame photometry (IL-943). The detection limits of sodium and potassium were 100–200 mmol/l and 1.0–15 mmol/l, respectively; the intra-assay coefficients of variability were 7% and 8%, respectively. Plasma levels of E2 and aldosterone were measured according to the radioimmunoassay protocols of Diagnostic Products. The radioimmunoassay detection limits of E2 and aldosterone were 20–3,600 and 25–1,200 pg/ml, respectively; the intra-assay coefficients were 5.8% and 3.8%, respectively.
Blood pressure and heart rate.
Rats were anesthetized with Nembutal (50 mg/kg ip). A ventral midline skin incision was made from the lower mandible posterior to the sternum (∼3 cm long). Three silk ligatures (3-0, Braintree) were passed under the left carotid artery vessel and used for both ligation and retraction. The anterior silk suture, placed just caudal to the bifurcation of the interior and exterior carotid arteries, was used to ligate the artery. After a loose knot on the middle suture was made, the posterior suture was placed loosely about 0.5 cm from the anterior tie for temporary occlusion of the artery during catheter insertion. Gentle tension was applied to both end ligatures to slightly retract and lift the vessel before a tiny incision was made in the carotid artery just posterior to the suture using Vannas spring microscissors and inserted into the carotid artery with the aid of fine forceps that were used to hold the vessel incision open. The occlusive posterior suture was released, and the catheter was advanced to the point where the small notch on the tubing (the transition between the thin-walled tubing and remaining outer tubing, 10 mm from the tip) resided at the vessel opening. Inserting the catheter up to this landmark notch ensured the critical placement of the pressure-sensing tip just inside of the thoracic aorta. This notch also served as a useful tool for anchoring the catheter into the vessel; the middle suture was then tied, and the loose ends of the suture were tied securely around the catheter at the notch. The loose ends of the occlusive posterior suture were next tied securely around the vessel/catheter. The middle suture was used to hold the catheter in place after the vessel was cannulated.
Through the same ventral neck incision, a subcutaneous pouch was formed for the placement of the transmitter body along the right flank of the animal. With the use of a pair of blunt dissecting scissors, the skin was gently dissected free from the underlying tissue starting at the right neck region and proceeding posteriorly to form a “pocket” along the right flank. It was important that the pocket be made sufficiently large enough to house the transmitter without unduly stretching the skin, because pressure necrosis could result. The transmitter was slipped under the skin and down into the pocket along the flank as close to the right hindlimb as possible. The neck incision was closed using silk sutures (3-0). Animals were kept on the warming pad until they were fully conscious. Recording of blood pressure and heart rate began on the 7th day after transmitter implantation at 10-s intervals every 10 min for 1 wk.
Animals were anesthetized with Inactin (100 mg/kg ip), and catheters were placed in the jugular vein for infusion of inulin for the determination of glomerular filtration rate (GFR), in the carotid artery for blood sampling, and in the bladder for urine collection. Infusions of 2% inulin in 0.9% NaCl were given intravenously at a rate of 2% body weight (BW)/h. This protocol is known to be effective in maintaining arterial hematocrit and plasma volume during preparatory surgery and throughout the period of the experiment at preanesthesia values (6). After a 60-min equilibration period, two 20-min urine clearances were taken to measure renal function, including urinary sodium and potassium excretion, fractional excretion of sodium and potassium (FEK+), and urine flow rate. Blood was collected at the midpoint of each urine clearance. The inulin concentration in plasma and urine was determined by the anthrone method (5). GFR was equated with the clearance of inulin. Urine concentrations of sodium and potassium were determined by flame photometry (IL-943).
AT1R radioligand binding.
Membranes were prepared from the adrenal cortex and from isolated glomeruli and a tubular enriched fraction from the renal cortex as described previously (11, 36). Membranes (5–10 μg protein/tube) were incubated for 1–2 h at room temperature with increasing concentrations of 125I-labeled [Sar1,Ile8]ANG II (Peptide Radioiodination Service Center, University of Mississippi) in the presence of 1 μM PD-123319, an ANG II type 2 receptor antagonist (so only AT1R expression was measured) (36). Binding reactions were terminated by rapid filtration through a Brandel cell harvester. Regarding quantitation, specific AT1R binding was defined as the total amount of radioligand bound minus the nonspecific binding, defined as the amount bound in the presence of 200 nM ANG II (100 × Kd for ANG II). Data points were obtained in triplicate. Kd and Bmax values from Scatchard plots were determined using the nonlinear regression analysis program PRISM.
Sample sizes were selected using a power of 0.8 and an alpha of 0.05 and are based on a minimal desired detectable difference in means of ∼20% and the expected standard deviation of the data set. Statistical significance for the differences between groups was determined by one-way ANOVA, and for time relationships, by two-way ANOVA, followed by Student-Newman-Keuls post hoc tests. Differences were considered significant at P < 0.05.
In these studies, female Sprague-Dawley rats were sham-operated or received ovarectomy (OVX) and injected subcutaneously with vehicle or E2 every day for 2 wk unless otherwise indicated. There were no significant differences in BW at the outset of the study [BW: Sham, 179 ± 3.1 g; OVX, 180 ± 3.4 g; and OVX + E2, 177 ± 2.2 g]. There were also no significant differences in BW between the sham-operated and OVX + E2 animal groups at the end of the study although both groups significantly increased their BW by 1.1-fold (P < 0.01) compared with their initial BWs. The OVX group increased their BW by 1.4-fold (P < 0.001) compared with their initial BWs [Sham, 197 ± 3.5 g; OVX, 256 ± 4.2 g; and OVX + E2, 194 ± 3.9 g].
Radioimmunoassays confirmed that plasma E2 levels were below that typically detected at estrus (5–10 pg/ml) in the vehicle-treated OVX rats (2.4 ± 0.5 pg/ml, n = 6), whereas the OVX rats replaced with E2 were maintained at levels similar to peak levels observed in the rat at diestrous (91 ± 10 pg/ml, n = 6) (8).
Blood pressure and heart rate.
No differences in systolic, diastolic, or mean arterial blood pressure were observed between the OVX and OVX + E2 animal groups during the day or at night (Table 1). Heart rate was also indistinguishable between these two animal groups (Table 1).
Plasma sodium, potassium, and aldosterone.
OVX had no detectable effect on plasma sodium and aldosterone compared with the sham-operated or OVX + E2 animal groups. In contrast, OVX significantly increased plasma potassium by 0.3 mmol/l compared with sham-operated controls; this effect of OVX on plasma potassium was prevented by E2 replacement (Table 2).
OVX had no significant effect on GFR based on BW (Table 3) or when calculated by kidney weight [in ml·min−1·g−1: sham, 1.9 ± 0.1 (n = 7) vs. OVX, 2.0 ± 0.2 (n = 11)] compared with the sham-operated group. No significant differences were observed in GFR in the ovariectomized rats treated with and without captopril [BW (in ml·min−1·100 g−1): OVX, 0.63 ± 0.05 (n = 8) vs. OVX + captopril, 0.69 ± 0.05 (n = 7)] or in the E2-treated ovariectomized rats treated with and without losartan [BW (in ml·min−1·100 g−1): OVX + E2, 0.60 ± 0.06 (n = 9) vs. OVX + E2 + losartan, 0.51 ± 0.05 (n = 6)].
Urine flow rate, urinary sodium excretion, the fractional excretion of sodium, and urinary potassium excretion were also unaffected by OVX compared with sham-operated animals (Table 3). In contrast, OVX significantly reduced the FEK+ by 40% compared with the sham-operated animals; this inhibitory effect of OVX on the FEK+ was prevented by E2 replacement (Table 3). Treatment with the angiotensin-converting enzyme inhibitor captopril (Fig. 1A) or the AT1R antagonist losartan (Fig. 2A) throughout the duration of the experiment both prevented the OVX-induced decrease in the FEK+.
Renal AT1R binding.
OVX increased the density of AT1Rs in membranes prepared from isolated glomeruli as early as 2 days after surgery compared with the sham-operated and OVX + E2 animal groups (Fig. 3). Maximal receptor upregulation occurred by 8 days. Two weeks after OVX, glomerular AT1R densities were increased by 1.4-fold compared with the E2-replaced group.
Treatment with the angiotensin-converting enzyme inhibitor captopril (Fig. 1B) or the AT1R antagonist losartan (Fig. 2B) throughout the 2-wk duration of the experiment prevented the OVX-induced increase in glomerular AT1R densities.
Scatchard analysis also revealed that AT1R densities were significantly higher in the OVX group compared with the OVX + E2 animals in the proximal tubular-enriched membrane fraction [Bmax (in fmol/mg protein): OVX, 742 ± 24 vs. OVX + E2, 569 ± 24] as well as in the adrenal cortex membrane fraction [Bmax (in fmol/mg protein): OVX, 183 ± 7.4 vs. OVX + E2, 134 ± 7.4], and both losartan [Bmax (in fmol/mg protein): OVX + losartan, 122 ± 2.33] and captopril [Bmax (in fmol/mg protein): OVX + captopril, 114 ± 11] prevented the OVX-induced increase in adrenal AT1R densities.
The major finding of this study was that OVX caused a decrease in the FEK+ compared with sham-operated animals. The fact that E2 replacement prevented this effect suggests that the observed decrease in the FEK+ induced by OVX was a result of E2 deficiency. We also found that OVX caused a significant increase in plasma potassium. This latter finding is consistent with our previous study in Sprague-Dawley rats maintained on a low-sodium diet, which showed that plasma levels of potassium were higher in E2-deficient rats compared with E2-replete animals and that E2 deficiency attenuated the ability of aldosterone infusion to reduce plasma potassium levels (37). In addition, our observation that E2 replacement was able to prevent the effects of OVX on plasma potassium is consistent with our previous study, which showed that E2 replacement prevented the effect of OVX on plasma potassium, whereas progesterone replacement had no effect on plasma potassium levels (37). Taken together, our current findings suggest that the OVX-induced decrease in the FEK+ contributes to the increase in plasma potassium that we have previously described in the E2-deficient rat.
Aldosterone stimulates potassium excretion in the kidney, resulting in decreased plasma potassium (34). Under the experimental conditions of this study, however, OVX did not result in detectable increases in plasma aldosterone. Previously, we showed that E2 deficiency induced by OVX increases ANG II-induced aldosterone secretion (24); however, this previous study was conducted in the presence of dexamethasone to separate the effects of E2 on ANG II-induced aldosterone from its effects on adrenocorticotropic hormone-induced aldosterone. Thus the effects of E2 deficiency on ANG II-induced aldosterone in this study may have been masked by adrenocorticotropic hormone-induced aldosterone secretion.
The finding that renal AT1R densities were higher in the glomeruli and proximal tubules of E2-deficient animals compared with the E2-replete females supports previous studies. We and others reported that the density of renal AT1Rs was higher in the ovariectomized Dahl salt-sensitive rat (7, 9), spontaneously hypertensive rat (28), Wistar-Kyoto rat (1), and beagle dog (19) compared with the E2-replete females. This observation raises the possibility that the decreased FEK+ is in part a result of increased renal AT1R activity. AT1Rs act in the kidney to regulate the expression of ion channels and transporters, and thus increased AT1R activity might modulate the ability of aldosterone to regulate potassium excretion, resulting in a decrease in the FEK+. The finding that both an inhibitor of ANG II synthesis and an AT1R antagonist prevented the OVX-induced decrease in the FEK+ supports this hypothesis. Furthermore, previous studies have shown that blocking AT1Rs or inhibiting ANG II action decreases plasma potassium by increasing potassium excretion (2, 15, 20).
At this time, we have no information on which AT1R-signaling pathways in the kidney are modulated by changes in E2 levels. One possibility is that E2 deficiency increases the ability of AT1Rs to modulate potassium channel activity. It is interesting that the AT1R has been shown to regulate the intracellular distribution and gating properties of the transient-outward potassium channel Kv4.3 (4) and has dual effects on the activity of the apical 70 pS potassium channel in the thick ascending limb (14). However, the possibility remains that the OVX-induced increase in renal AT1R density is coincidental rather than causative and that other effects of E2 deficiency are responsible for the inhibition of potassium excretion. In this regard, our recent study demonstrated that E2 has mineralocorticoid-like effects on the abundance of α-sodium-potassium ATPase and the thiazide-sensitive sodium-chloride cotransporter and protects females from aldosterone-induced hypertension (26). Thus a loss of E2 regulation of these transporters might contribute to the observed effects of OVX in this current study.
The effect of OVX on renal function in the Sprague-Dawley rat is consistent with clinical observations. Renal sodium reabsorption after a hypertonic saline infusion is greater in postmenopausal women on estrogen replacement therapy compared with those on placebo (30), suggesting that aldosterone activity is greater in the estrogen-replete woman compared with the estrogen-deficient state. It is important to keep in mind that this current study was conducted in healthy young animals. It is possible that the effects of E2 on potassium excretion might be altered in the setting of chronic cardio-renal disease. This point is highlighted by clinical studies that report differential effects of antihypertensive therapies on potassium homeostasis in chronic kidney disease and heart failure (27, 32). Furthermore, we did not observe any significant effects of OVX on blood pressure and heart rate. It is possible that OVX would have increased blood pressure if these animals were hypertensive. In this regard, OVX has been shown to increase mean arterial pressure in both the young and old Dahl salt-sensitive rat (7, 9, 10) and in Dahl-Iwai salt-sensitive rats (18). The incidence of hypertension is higher in postmenopausal women compared with estrogen-replete premenopausal women (22). Thus the blood pressure-raising effects of E2 depletion may require an underlying pathological state such as salt-sensitive hypertension to be observed. Alternatively, the fact that we did not observe any differences in blood pressure between the E2-replete and -depleted state in our animal study may indicate that a 2-wk exposure to an E2-deficient state is not sufficient to observe changes in blood pressure.
Although our results suggest that E2 deficiency reduces the FEK+ by upregulating ANG II-mediated AT1R activity through increased AT1R density in the kidney, it is also possible that AT1Rs in other key target tissues contribute to this effect. In this regard, the fact that in this study E2 deficiency increased AT1R densities in the adrenal cortex raises the possibility of adrenal AT1R-mediated changes in aldosterone actions in the kidney through aldosterone-regulatory pathways. Alternatively, E2 deficiency could amplify aldosterone action through mechanisms that are independent of the AT1R. For example, E2 deficiency may increase the renal responsivity to aldosterone by increasing the expression or activity of the mineralocorticoid receptor. In this regard, Verlander et al. (33) have shown that E2 (12) enhances the density of the aldosterone-regulated (12) thiazide-sensitive sodium-chloride cotransporter in the distal tubule of ovariectomized rats (26).
In conclusion, these studies indicate that E2 deficiency modulates renal function by reducing the FEK+ in an ANG II/AT1R-mediated manner, resulting in a reduction in the ability of aldosterone to excrete potassium. These findings may contribute to the clinical observations that the ERPF response to ANG II infusion inversely correlated with plasma levels of E2 [i.e., the higher the level of plasma E2, the smaller the magnitude of the ANG II-induced decrease in ERPF (16)] and that the renal hemodynamic responses to ANG II blockade were blunted during the luteal phase (peak E2 levels) of the menstrual cycle (3). Furthermore, these effects of E2 deficiency on renal function and renal AT1R densities may have implications for renal function in postmenopausal women, especially if there is any underlying renal or vascular pathology (21).
This research was supported by a National Kidney Foundation of the National Capital Area grant-in-aid (to H. Ji) and National Institute on Aging Grant AG-19291 (to K. Sandberg).
We thank Dr. William Welch for critical review of the manuscript.
↵1 This paper was presented at the 9th Cardiovascular-Kidney Interactions in Health and Disease Meeting at Amelia Island Plantation, Florida, on May 26–29, 2006.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by the American Physiological Society