Abstract
ABSTRACT. Age-dependent glomerulosclerosis with reduced GFR develops earlier among men than among women. Therefore, whether female sex hormones could prevent the age-dependent decrease in GFR was investigated. The kidney function in oophorectomized rats treated with placebo (OOX group), estrogen (OOX+E2 group), or estrogen plus progesterone (OOX+E2+P group) for 5 mo and in sham-operated rats (sham group) was examined. The rats were 13 mo of age at the time of the investigation. They were conscious and chronically instrumented. The results demonstrated that estrogen, alone or in combination with progesterone, was without effect on the baseline GFR and effective renal plasma flow but prevented severe decreases in the renal functional reserve and fractional proximal tubular fluid output (lithium clearance technique, fractional lithium excretion), which were observed in the OOX group. The renal functional reserve (estimated by stimulation with glycine) was −223 ± 151, 483 ± 129, 675 ± 76, and 208 ± 140 μl/min in the OOX, OOX+E2, OOX+E2+P, and sham groups, respectively. Fractional lithium excretion was 12.4 ± 3.1, 26.8 ± 2.0, 31.8 ± 2.5, and 23.6 ± 3.3% in the OOX, OOX+E2, OOX+E2+P, and sham groups, respectively. In conclusion, oophorectomy at the age of 8 mo did not produce a decrease in baseline GFR in female rats within a period of 5 mo. However, oophorectomy led to severe decreases in the renal functional reserve and fractional proximal tubular fluid output. Both effects were prevented with administration of estrogen. Sham-operated rats demonstrated values for renal functional reserve and fractional lithium excretion that were between those observed for the OOX group and the groups treated with sex hormones.
Gender differences in the progression of kidney failure represent a well known phenomenon. Male patients exhibit faster development of age-dependent glomerulosclerosis than do female patients (1–3⇓⇓). Furthermore, male patients exhibit a more rapid decline in renal function attributable to nephrotic syndrome (4), polycystic kidney disease, or chronic glomerulonephritis (5). Finally, the incidence of subnormal GFR in IgA nephropathy is higher among male patients (6).
The rapid progression of kidney failure among male subjects is also evident in animal models. Unilateral nephrectomy and contralateral ligation of two-thirds of the renal circulation produce more rapid increases in proteinuria, segmental glomerulosclerosis, and collagen α1 IV mRNA in male rats than in female rats (7). In line with this, estrogen protects male rats against the development of glomerulosclerosis (8). In mesangial cell cultures, estrogen inhibits proliferation, suppresses synthesis of collagen type I and IV, and inhibits the oxidation and uptake of LDL by macrophages (9,10⇓). Other in vitro studies demonstrated that estrogen inhibits procollagen type III and increases the production of collagenase (11–13⇓⇓). Furthermore, castration of male rats seems to have a beneficial effect on the development of glomerulosclerosis (14), whereas treatment with testosterone accelerates the disease (15). Recent studies demonstrated that glomerulosclerosis susceptibility is associated with diminished estrogen receptor expression in mesangial cells (16). Taken together, these findings indicate the involvement of both hormones in age-dependent changes in renal function, through direct effects on the kidney, and they suggest that estrogen may protect against these changes. To date, no studies have been performed to explore the effects of menopause on normal kidney function.
The aim of this study was to examine whether estrogen protects premenopausal rats from age-dependent changes in renal function and whether the protection disappears when estrogen is withdrawn, through either menopause or oophorectomy. Wistar rats develop age-dependent glomerulosclerosis at 13 mo of age and demonstrate gender differences comparable to those observed for human subjects (17). To study whether estrogen could prevent age-dependent changes in renal function, we performed a detailed study involving determinations of effective renal plasma flow (ERPF), GFR, and lithium clearance in 13-mo-old conscious rats that had been oophorectomized at the age of 8 mo. One subgroup was supplemented with estrogen and another was given a combination of estrogen and progesterone. The results indicated that proximal tubular fluid output (lithium clearance) in rats supplemented with both sex hormones was more than twice the level in oophorectomized rats. Furthermore, the renal functional reserve was abolished in oophorectomized rats but was preserved in rats supplemented with ovarian sex hormones.
Materials and Methods
Animals and Physical Environment
Specific pathogen-free, adult, female, Wistar rats (M & B, Ry, Denmark) weighing between 240 and 300 g were used. The studies were performed in accordance with the European Community guidelines and were approved by the Danish Animal Experiments Inspectorate. Animals were housed two per cage in a room with a 12/12-h artificial light cycle (lights on from 8 a.m. to 8 p.m.), a temperature of 21 ± 2°C, and a humidity level of 55 ± 2%. The rats were fed a special synthetic estrogen-free diet (C-1000; Altromin GmbH, Lage, Germany) from the age of 8 mo and throughout the experimental period (until the rats were 13 mo of age). The diet contained 17.3% protein, 110 mmol/kg sodium, and 180 mmol/kg potassium. Three days before clearance measurements, this diet was changed to one containing lithium chloride (12 mmol/kg dry weight), to yield measurable plasma lithium concentrations without affecting renal function (18). Rats had free access to food and tap water throughout the studies.
Surgical Preparation
Eight-month-old rats were randomized into four groups, as follows: (1) oophorectomized and substituted with placebo (placebo pellet, 90-d release) (OOX group), (2) oophorectomized and substituted with 17β-estradiol (0.85 mg/pellet, 90-d release) (OOX+E2 group), (3) oophorectomized and substituted with 17β-estradiol and progesterone (0.85 mg/pellet and 200 mg/pellet, respectively, 90-d release) (OOX+E2+P group), and (4) sham-operated (sham group). The pellets (Innovative Research of America, Sarasota, FL) were implanted subcutaneously between the scapulae, during anesthesia with fentanyl-fluanisone (0.0945 mg/kg body wt and 0.3 mg/kg body wt, respectively; Janssen, Copenhagen, Denmark) and midazolam (0.25 mg/kg body wt, Dumex; Alpharma, Copenhagen, Denmark). The implantation of pellets was repeated 90 d after the first surgical procedure. Five months after the oophorectomy/sham operation, the rats underwent a kidney function study. One week before the functional study, the rats were anesthetized with halothane/N2O, because that anesthesia was better tolerated at that stage of the investigation. With semiaseptic surgical techniques, sterile Tygon catheters (Norton Performance Plastics, Akron, OH) were advanced into the abdominal aorta and the inferior vena cava via the femoral vessels. A sterile chronic suprapubic catheter was implanted in the bladder. All catheters were prepared and fixed, with minor modifications, as described previously (19). After instrumentation, the rats were given saline solution (5 ml, administered subcutaneously) and the long-acting analgesic buprenorphine (10 μg/animal, administered subcutaneously, Temgesic; Reckitt & Colman, Hull, UK). The animals were housed individually and, after a recovery period of 3 to 5 d, were accustomed to restriction with daily training sessions in restraining cages. The duration of the daily sessions was gradually increased from 1 to 3 h.
Experimental Design
On the basis of vaginal smears, the sham-operated animals were determined to be in proestrus the day before the kidney function experiments. From all groups, two rats were randomly chosen for time-control experiments the day after the kidney function experiments. Renal function was measured during a control period, followed by a period during which glycine was infused. Time-control animals were treated similarly except that glycine was replaced by vehicle alone. Each experiment included a 15-min bolus period for infusion of renal markers, a 105-min equilibration period, three 20-min baseline urine collection periods, a 60-min glycine equilibration period, and three 20-min test urine collection periods, during which glycine was infused.
Renal Clearance Protocol
The experiments were performed between 9 a.m. and 2 p.m. Each rat was transferred to a restraining cage and was connected to infusion pumps via the venous catheter and to a BP transducer (Baxter Uniflow; Bentley Laboratories Europe BV, Uden, Holland) via the arterial catheter.
Through the pressure transducer, a continuous intra-arterial infusion of 25 to 50 mM glucose solution containing heparin (100 U/ml) was administered at a rate of 5 μl/min, to keep the arterial catheter open. In addition, the animals received, throughout the experiments, an intravenous infusion of 25 to 50 mM glucose solution (bolus, 0.6 ml; sustained infusion, 10 μl/min) containing [14C]tetraethylammonium bromide (TEA) (bolus, 31 kBq; sustained infusion, 0.52 kBq/min; New England Nuclear, Boston, MA), [3H]inulin (bolus, 67 kBq; sustained infusion, 1.1 kBq/min; Amersham International, Rainham, UK), and lithium (bolus, 7.2 μmol; sustained infusion, 0.12 μmol/min), as markers of ERPF, GFR, and proximal tubular fluid output, respectively. Furthermore, the animals received 25 to 50 mM glucose solution at a rate of 30 μl/min to maintain an adequate urine flow, which was necessary for elimination of bladder-emptying errors. Glycine (4 mg/min) was administered with the glucose solution. During glycine infusion, net losses of water and sodium were replaced with a personal computer-controlled servo-system, as previously described (20). The fluid infusion rate was calculated by taking into account the 45 μl/min of fluid continuously administered throughout the study and the variable amount of fluid administered by the NaCl solution pump.
Blood samples (200 μl) were withdrawn from the arterial catheter at 0, 120, 180, 240, and 300 min. Replacement donor blood was administered after each blood sample was withdrawn. The time-control rats underwent the experiment once more but were treated with vehicle instead of glycine after 180 min.
Analyses
Urine volume was determined by gravimetry. Concentrations of sodium and potassium in plasma and urine and of lithium in plasma were determined by flame emission photometry. Lithium concentrations in urine were determined by atomic absorption spectrophotometry. [14C]TEA and [3H]inulin concentrations in plasma and urine were determined by dual-label liquid scintillation counting with a Packard Tri-Carb liquid scintillation analyzer (model 1900CA; Packard Instruments, Greve, Denmark). The sample (15 μl) and 285 μl of water were mixed with 2.5 ml of Ultima Gold scintillation liquid (Packard Instruments). Concentration calculations were performed with automatic efficiency control.
Calculations
Renal clearances (C) and fractional excretions (FE) were calculated with the standard formulas C = U × V/P and FE = C/GFR, where U is the urinary concentration, V is the urine flow rate, and P is the plasma concentration. In previous studies, the renal extraction fraction of TEA was demonstrated to be approximately 90% and the validity of using TEA to estimate ERPF was documented (21,22⇓). At the concentration used in this study, TEA is without effects on efferent renal sympathetic nerve activity in rats (23). Using the clearances of TEA, inulin, and lithium, the following parameters were calculated: ERPF (TEA clearance), GFR (inulin clearance), filtration fraction (100 × inulin clearance/TEA clearance), proximal tubular fluid output (lithium clearance), and renal vascular resistance [(mean arterial pressure − 5) × (1 − hematocrit)/TEA clearance]. In all calculations, it was assumed that the renal venous pressure was 5 mmHg throughout the experiment.
Bioelectric Impedance Measurements
Before the rats were euthanized, body weights were measured and the animals were anesthetized with phenobarbital (50 mg/kg body wt, administered intraperitoneally). Bioelectric impedance was then measured in the frequency range of 4 to 1024 kHz with a multifrequency bioimpedance spectroscopic analyzer (SFB3; UniQuest Ltd., Brisbane, Australia), as described previously (24). In brief, stainless-steel electrodes were inserted 3 to 5 mm subcutaneously. The source electrodes were inserted in the midline between the auricles and at the midpoint between the iliac crests, and the detector electrodes were inserted at the base of the tail and in the midline between the anterior edges of the orbits. Assuming an hydration value of 0.732, the following equation was used to calculate the fat-free mass: fat-free mass = [(131.2981 × L2/Zc) + 37.3184]/0.732, where L is the conductor length (in centimeters) and Zc is the magnitude of impedance (Ward et al., personal communication). The fat mass was calculated as fat mass = body weight − fat-free mass.
Organ Weights
After euthanasia, the wet weights of the uterus, kidneys, heart, gastrocnemius muscle, and liver were recorded for all animals.
Data Presentation and Statistical Analyses
All values are presented as mean ± SEM. Overall statistical analyses were performed with ANOVA between groups or one-way ANOVA within groups. The test periods were compared with the baseline period with the t test for paired data or a nonparametric analysis (Wilcoxon test), when the variance differed significantly between the periods. Differences between groups were tested with the t test or a nonparametric analysis (Mann-Whitney test), when the variance differed significantly between the groups. Differences were considered statistically significant at the 0.05 level.
Results
Plasma Values
The plasma values are presented in Table 1. In the baseline period and the glycine period, there were no differences in the plasma concentrations of sodium, potassium, and lithium between the groups. The hematocrit values were significantly lower in the OOX+E2 and OOX+E2+P groups, compared with the OOX group, whereas there was no change in the sham group. There were no significant differences in mean BP between the groups.
Table 1. Plasma values, hematocrit, and mean arterial BP in rats that were oophorectomized at the age of 8 mo and monitored for 5 moa
The plasma concentrations of sodium and potassium were not affected by glycine administration in any of the groups. In all of the groups given glycine, there was a significant decrease in the mean plasma lithium concentration. In the time-control group, there was no change in the plasma lithium concentration. Hematocrit and mean arterial pressure values were not affected by glycine.
Renal Hemodynamics and Segmental Tubular Handling of Sodium and Water
In the baseline period, there was no difference between the groups with respect to ERPF, renal vascular resistance, or GFR (Figure 1). In contrast, the OOX+E2, OOX+E2+P, and sham groups demonstrated significantly higher values for proximal tubular fluid output and fractional lithium excretion, compared with the OOX group (Table 2). Also, the clearance and fractional excretion of potassium were higher in the groups treated with sex hormones than in the placebo-treated group.
Figure 1. Effective renal plasma flow (ERPF), renal vascular resistance (RVR), and GFR in four groups of rats that were oophorectomized at the age of 8 mo and were monitored for 5 mo, as follows: oophorectomized and given placebo (OOX group), oophorectomized and substituted with 17β-estradiol (OOX+E2 group), oophorectomized and substituted with 17β-estradiol and progesterone (OOX+E2+P group), and sham-operated (sham group). *P < 0.05, OOX group versus any other group. †P < 0.05, ††P < 0.01, †††P < 0.001, vehicle versus glycine. White bars, baseline values; black bars, values during infusion of glycine.
Table 2. Renal variables in rats that were oophorectomized at the age of 8 mo and monitored for 5 moa
Infusion of glycine led to ERPF increases that were significantly greater in the OOX+E2 and OOX+E2+P groups, compared with the OOX group, in which there was no change (Figure 1). Similarly, glycine led to GFR increases that were significantly greater in the OOX+E2, OOX+E2+P, and sham groups, compared with the OOX group, in which there was no change. Absolute proximal reabsorption (APR) decreased significantly in the OOX and sham groups, suggesting that glycine inhibits APR. In the two hormone-substituted groups, large increases in GFR were accompanied by no changes in APR (−84 and 15 μl/min). This is also an indication of a direct inhibitory effect on proximal reabsorption, because an increase in GFR should cause proximal tubular reabsorption to increase. Assuming that the glycine-induced changes in GFR would lead to proportional increases in APR with unaltered fractional proximal reabsorption, we calculated hypothetical APR values. Those values were compared with the actual measured values to yield an estimate of the effect of glycine on APR in the various groups. As demonstrated in Figure 2, this comparison revealed that glycine led to inhibition of APR in all groups. The inhibition seemed to be similar in all groups, i.e., independent of the preceding hormone treatment. The increases in proximal tubular fluid output, sodium clearance, and potassium clearance were similar to the changes in the OOX group (Table 2). The time-control group demonstrated no alterations in any of the variables except the urine flow rate, which exhibited a small increase (data not shown).
Figure 2. Changes in GFR versus measured (○ and hypothetical (•) changes in absolute proximal reabsorption (APR) in response to glycine in the four groups under investigation. The hypothetical changes in APR were calculated as ΔGFR × fractional proximal reabsorption before glycine.
The weights of various organs in the different groups are presented in Table 3. The weights of the uterus and liver were increased and the body weight and fat mass were decreased in the OOX+E2, OOX+E2+P, and sham groups, compared with the OOX group.
Table 3. Body weight and weight of various organs in rats that were oophorectomized at the age of 8 mo and monitored for 5 moa
Discussion
This study examined for the first time the effect of estrogen on the preservation of kidney function in aging female rats. The results indicated that rats that were oophorectomized at the age of 8 mo and were monitored for 5 mo demonstrated normal baseline GFR, compared with sex hormone-supplemented rats or sham-operated rats with intact estrous cycles. However, when the renal functional reserve was examined with maximal stimulation with glycine, interesting significant differences between the groups were observed. Rats deprived of estrogen were unable to increase the GFR, whereas rats treated with estrogen exhibited a preserved renal functional reserve. The sham-operated group, being middle-aged rats and therefore having waning estrous cycles, demonstrated some renal functional reserve, although it was significantly lower than that observed in the substituted rats, which had been given a full substitution dose of sex hormones during the preceding 5 mo. These observations suggest that, although a protective effect of estrogen on GFR could not be demonstrated, protection of the renal functional reserve in female Wistar rats was evident.
ERPF exhibited the same pattern as GFR. There was no difference between the groups in the baseline period, whereas groups receiving or producing estrogen were able to increase ERPF after the administration of glycine, because of a reduction of the renal vascular resistance. The placebo-treated group was the only group that did not respond to glycine with an increase in ERPF. These observations suggest that estrogen preserved the vascular vasodilating reactivity to glycine in oophorectomized rats.
It has long been suggested that estrogen protects women against hypertension and coronary artery disease and that the effect of estrogen is to potentiate endothelium-dependent vasodilation, rather than being vasodilating itself (25). For example, it was demonstrated that long-term estrogen replacement therapy increased vasodilation in response to ischemia in the forearm resistance arteries in postmenopausal women (26), and it was demonstrated that ovariectomy of rats caused a loss of acetylcholine-mediated dilation within weeks (27). However, to fully appreciate the interaction between estrogen and glycine observed in this study, it is necessary to further explore the mechanisms underlying the vascular effects of estrogen and glycine.
Effects of Estrogen
Estrogen increases bioactive nitric oxide (NO) contents in vascular endothelial cells. This could be the result of stimulated NO production with increased NO synthase expression and activity (28), or it could be attributable to reduced degradation of NO because of inhibition of superoxide anion production (29).
In addition, estrogen has been demonstrated to interfere with the renin-angiotensin system. Estrogen inhibits circulating renin and angiotensin-converting enzyme, decreases circulating angiotensin II (AngII) levels, and downregulates AngII type 1 receptor expression in vascular smooth muscle cells, adrenal cortex, and hypothalamus (30). Estrogen deficiency as a result of menopause or ovariectomy has the opposite effects on the renin-angiotensin system (30). Furthermore, the vasoconstrictor response to AngII is attenuated by estrogen, whereas estrogen deficiency enhances the vasoconstrictor response to AngII by upregulating AT1 receptors (30). These data suggest that NO and AngII may interact physiologically to maintain vascular endothelial function, and an imbalance of these interactions may play an important role in endothelial dysfunction in hypertension and cardiovascular disease (30). Estrogen changes the balance between NO and AngII, leading to a higher NO/AngII ratio that favors vasodilation.
Furthermore, estrogen may produce its vasodilating effect by interfering with prostaglandins. For example, estrogen stimulates prostacyclin production in human vascular endothelial cells (31), and urinary excretion of prostacyclin increases in postmenopausal women receiving hormone therapy (32). In addition, the effect of estrogen may be related to the release of or sensitivity to endothelium-derived contracting factors such as endothelin-1 (33,34⇓).
Effects of Glycine on GFR
Recent experimental studies demonstrated that the vasodilatory response to glycine depends on the presence of a proper balance between NO and AngII (35), and the renal functional reserve has been proposed as an index of the intrarenal balance between NO and AngII (36,37⇓). Under normal conditions, glycine induces a decrease in renal vascular resistance and a consequent increase in GFR. It has been demonstrated that the response to glycine is inhibited during blockade of the NO system and pretreatment with an AngII inhibitor normalizes the response to glycine. It has also been demonstrated that increases in AngII levels diminish the vasodilatory response. In fact, the absence of a renal functional reserve was observed in experimental models of renal damage, such as renovascular hypertension (38), diabetes mellitus (39), and glomerulonephritis (40). In all of those conditions, angiotensin-converting enzyme inhibitors restored a normal renal functional reserve. Additional studies with human subjects confirmed a role of AngII in the reduction of the renal functional reserve. Among hypertensive patients on a low-salt diet, the consequent activation of the renin-angiotensin system blunted the hemodynamic response to amino acids and the response was restored to normal with angiotensin-converting enzyme inhibitors (41). In other pathophysiologic conditions among human subjects, angiotensin-converting enzyme inhibitors restored the hemodynamic response to amino acids (42,43⇓). Therefore, in different conditions with a prevalence of AngII over NO as a common feature, the renal functional reserve is blunted or absent. This is likely to be caused by overwhelming actions of vasoconstrictor stimuli that offset the normal vasodilatory response to amino acids.
Therefore, the decreased vasodilatory response to glycine in the OOX group in this study could be attributable to an abnormal intrarenal balance between NO and AngII. However, an effect of estrogen on the production of prostaglandins or endothelin-1 cannot be excluded. Estrogen does interfere with both hormone systems, as mentioned above.
Effects of Glycine on Proximal Reabsorption
This study demonstrated fractional lithium excretion from the proximal tubules that was similar to values normally observed under control conditions in our laboratory (44). Only one group demonstrated an abnormal fractional lithium excretion value, namely, the group that neither produced nor received estrogen. Fractional lithium excretion was almost halved in that group, compared with the other groups. It is known that fractional lithium excretion is decreased in conditions involving extracellular volume depletion, such as during treatment with thiazides and in some clinical conditions associated with cardiovascular disease (45). It has also been demonstrated that fractional lithium excretion increases during volume expansion and that NO plays an important role in modulating this response (45). Taken together, these studies suggest that fractional lithium excretion is decreased in conditions with a low renal NO/AngII ratio and is increased in conditions with a high NO/AngII ratio. It may therefore be suggested that fractional lithium excretion is decreased in the OOX group because a lack of estrogen in this group leads to a decrease in the renal NO/AngII ratio, as discussed above.
The administration of glycine led to inhibition of proximal sodium reabsorption in all groups. It has been suggested that inhibition of APR leads to loss of the renal functional reserve because of increased sodium delivery to the macula densa, resulting in activation of the tubuloglomerular feedback mechanism (35). The results of our study did not support this hypothesis. We observed the same increase in proximal tubular fluid output in all groups and should, according to the hypothesis, expect the same reduction in the renal functional reserve in all groups. In fact, only one group, the OOX group, failed to increase GFR; all of the other groups demonstrated normal renal functional reserve. Therefore, we suggest that in our study the glomerular response to glycine was dissociated from the tubular response, as also observed in some other studies (38,39,46,47⇓⇓⇓). The lack of renal functional reserve observed in the OOX group may be attributable to a decreased NO/AngII ratio at the local level, as suggested above.
In conclusion, oophorectomy at the age of 8 mo did not, within a period of 5 mo, produce a decrease in baseline GFR in female rats. However, oophorectomy led to severe decreases in the renal functional reserve and fractional proximal tubular fluid output. Both effects were prevented with the administration of estrogen. The effects of oophorectomy might be attributable to a decrease in the intrarenal NO/AngII ratio.
Acknowledgments
This study was supported by grants from the Danish Medical Research Council (Grant 9700592) and the Novo-Nordisk Foundation. We thank Jette Birk and Kirsten Nyborg for excellent technical assistance.
- © 2003 American Society of Nephrology
References
