点击显示 收起
【摘要】 Maternal renal disease is associated with high maternal and fetal morbidity. To establish an animal model to study renal dysfunction in pregnancy and its potential role in programming for renal disease and hypertension in adult life, a kidney was removed from each of 16 nonpregnant ewes, and a branch of the renal artery of the remaining kidney was ligated (STNx ewes). The 16 STNx and 15 intact ewes were time mated 2.5-17 mo later and studied at 119-132 days of gestation. STNx ewes demonstrated renal hypertrophy and glomerular hyperfiltration. They had higher diastolic arterial pressures ( P < 0.05) and larger left ventricles ( P < 0.0005), drank more water ( P < 0.01), were hypochloremic ( P < 0.01) and hyperglycemic ( P < 0.0005), and had higher plasma creatinine levels ( P < 0.0005) than intact ewes. Effective renal plasma flows and glomerular filtration rates were lower ( P < 0.01) and protein excretion was greater ( P < 0.05) in STNx than in intact ewes. Glomerulotubular balance was impaired in STNx ewes. Proximal tubular Na + reabsorption was reduced ( P < 0.05), so Na + excretion was increased ( P < 0.05). In STNx ewes, filtered K + loads were reduced ( P < 0.005), but K + excretion was the same as in intact ewes. There was net K + secretion in STNx ewes; in intact ewes, there was net reabsorption. Plasma renin and angiotensinogen concentrations in STNx and intact ewes were similar, so the hypertension in STNx ewes was not renin dependent. STNx fetuses grew normally, and their blood gases, blood pressure, and heart rates were normal. These alterations in maternal fluid and electrolyte balance and the potential risk of maternal salt depletion or hyperkalemia may adversely affect the fetus.
【关键词】 left ventricular hypertrophy glomerular filtration rate effective renal plasma flow glomerulotubular balance fetus
HYPERTENSION COMPLICATES 7-10% of human pregnancies ( 21, 38 ), occurs in association with renal disease, and can be programmed during intrauterine life. The Australian Aborigine has a reduced life expectancy. For indigenous Australians between 1999 and 2001, life expectancy was 55.7-56.8 yr for men and 61-63.8 yr for women compared with 77 yr for men and 82.4 yr for women in the nonindigenous population (16). This early mortality is related to a high incidence of hypertension, cardiovascular disease, and end-stage renal disease. In indigenous Australians, the rate of notification of end-stage renal disease between 1997 and 2001 was 645 per million compared with the nonindigenous rate of 75 per million ( 40a ). Moderate-to-severe renal disease in pregnancy is associated with hypertension in pregnancy and prematurity ( 21 ). Also there is a strong association between maternal and offspring blood pressure ( 11 ). At an early age in adult life, 26% of indigenous Australian adults have microalbuminuria and 24% have overt albuminuria ( 18 ). Therefore, in indigenous Australians, there is evidence of renal damage in many women of child-bearing age.
In 1999, 13.5% of infants born to indigenous women were of low birth weight compared with a nonindigenous incidence of 6.5% ( 40a ). Epidemiological studies have shown that failure to thrive in utero is associated with an increased risk of cardiovascular disease, hypertension, and type 2 diabetes in middle-aged adults ( 19 ). It is likely therefore that maternal renal disease, especially if associated with hypertension, not only adversely affects maternal health and increases fetal morbidity and mortality but may also program the fetus for renal disease and hypertension in adult life.
We previously studied renal and endocrine models of hypertension in pregnancy in the chronically catheterized pregnant ewe to determine the effects of hypertension on the ewe and the fetus ( 27 - 29 ). However, in these models, hypertension was induced acutely in late gestation, and our studies were limited to short-term investigations of its effects on mother and fetus.
Here we describe the effects of a reduction in renal mass of the nonpregnant ewe on her health during pregnancy and on fetal growth and development. Subtotal nephrectomy is known to predispose to hypertension in a number of animal models ( 6, 23, 36 ). Therefore, we postulated that if renal mass was reduced in ewes before pregnancy, they might develop hypertension in pregnancy. We believe that this animal model will be useful in studying the pathophysiology of gestational hypertension and renal dysfunction in pregnancy and their effects on fetal development, as well as their role (if any) in fetal programming for renal disease and hypertension in adult life.
METHODS
The experiments were approved by the University of New South Wales Animal Care and Ethics Committee.
Over a 2-yr period, 31 cross-bred Merino pregnant ewes were studied at 119-132 days of gestation (full term = 150 days).
Surgical Preparation of Ewes
All surgeries were carried out under general anesthesia induced by intravenous injection of 1-2 g of thiopentone sodium (Pentothal, Abbott Australasia) and maintained with 2-3% halothane (Fluothane, Zeneca) in O 2. Ewes were given 600 mg of penicillin (300 mg/ml; Ilium Propen, Troy Laboratories, New South Wales, Australia) and 288 mg of oxytetracycline (96 mg/ml im; Alamycin, Norbrook Laboratories) at induction of anesthesia and at completion of surgery. Pregnant ewes also received these antibiotics for 2 days postoperatively.
Unilateral nephrectomy and ligation of a branch of a renal artery. Sixteen nonpregnant ewes underwent subtotal nephrectomy (STNx). In 15 ewes the right kidney was removed through a paravertebral incision; in 1 ewe the left kidney was removed. Through another paravertebral incision, at least one branch of a renal artery in the hilus of the other kidney was ligated, so that there was a color change in 30-50% of the kidney surface. The kidney that had been removed was weighed. The animals were held until recovery from surgery and returned to pasture 1 wk later. STNx and intact ewes from the same flock were then time mated in groups at various times over the next 2 yr. STNx ewes were mated 2.5-17 mo after surgery.
Pregnant ewes. At 108-114 days of gestation, 16 STNx and 15 intact ewes were anesthetized, and polyvinyl catheters (2.70 mm OD, 1.50 mm ID) were inserted into a maternal femoral artery and vein as described previously ( 32 ). These techniques were also used for cannulation of the fetuses. Catheters were flushed daily with 0.15 M heparinized saline (100 IU heparin/ml; Heparin Injection, Pharmacia & Upjohn). The ewes were housed in metabolic cages, and room temperature was maintained at 18-23°C. They were given free access to 8 liters of water, 1.2 kg of lucerne chaff, and 300 g of oats per day. No experiments were carried out for 6 days after surgery.
Experimental Protocol
Food and water intake and urine and feces output were measured daily. Samples of urine were collected and stored at -20°C for further analysis. In most animals, a 24-h study of food and fluid intake, as well as a detailed investigation of cardiovascular and renal function, was undertaken. Ewes were removed from their cages on the morning of the day on which cardiovascular and renal function were to be measured, and a 14-gauge Foley catheter was inserted into the maternal bladder via the urethra. The ewes were returned to their cages and given an intravenous loading dose of LiCl (150 µmol/kg) and p -aminohippurate (4.8 mg/kg, PAH; Merck, Sharpe & Dohme or Sigma-Aldrich). During the course of the experiment, infusions of LiCl (10 µmol·kg -1 ·h -1 iv) and PAH (810 mg/h iv) were delivered through a 0.2-µm filter (Minisart) at a rate of 6.3 ml/h. After 45 min, maternal urine was collected at 30-min intervals for 1.5-4.5 h. Arterial blood samples (5 ml) were collected at the midpoint of the second and third or fourth periods and, when appropriate, the fifth, seventh, and ninth periods into tubes containing heparin (20 IU/ml). Throughout the course of the experiment, maternal blood pressure was logged using a pressure transducer (Easy Vent Deadender Cap, Ohmeda BOC) connected to a polygraph (Grass, Quincy, MA) and recorded on an IBM-compatible personal computer.
All ewes were killed by intravenous injection of 4-5 g of pentobarbitone sodium (Lethabarb, Virbac). The maternal heart and kidney(s) were removed and weighed, and the kidneys were photographed on a 1-cm grid with a digital camera. Samples of the heart and left kidney cortex were taken for histology or frozen in liquid nitrogen and stored at -80°C.
Biochemical Analysis
Arterial blood samples. Maternal P O 2, P CO 2, and pH were measured using a blood gas analyzer (ABL 715, Radiometer Pacific) at 37°C and corrected to 39.5°C. The blood gas analyzer was also used to measure K +, Na +, Cl -, glucose, and lactate concentrations. Hematocrits were determined in duplicate using a microhematocrit centrifuge and reader (Hettich, Tuttlingen, Germany). The remaining blood was centrifuged for 10 min at 3,000 rpm and 4°C. Plasma and urine samples were stored at -20°C until biochemical analysis.
Plasma and urinary analysis. Effective renal plasma flow (ERPF) was measured as the renal clearance of PAH. The methods were adapted from those described by Bauer et al. ( 3 ) and were similar to those we used previously ( 17 ) utilizing 1:1,000 dilutions of urine and 1:15 dilutions of plasma. Volumes and reagents were adjusted so that samples could be read using a 96-well microtiter plate and a Microplate spectrophotometer (SpectraMax Plus, Molecular Devices). Concentrations of PAH in samples were determined from standard curves with use of known concentrations of PAH made up in the appropriate dilutions of urine or plasma.
Na + and K + concentrations in urine and drinking water were measured using a flame photometer (model FLM3, Radiometer), and plasma and urinary osmolalities were measured using an osmometer (One-Ten, Fiske Associates).
The clearance of Li + was used to determine the amounts of Na + reabsorbed by the proximal and distal nephron ( 30 ). Li + is freely reabsorbed by the proximal, but not the distal, tubule. Urinary and plasma Li + concentrations were determined using an atomic absorption spectrophotometer (model AA5, Varian-Techtron, Melbourne, Australia).
Glomerular filtration rate (GFR) was measured as the renal clearance of endogenous creatinine. Creatinine levels in urine and plasma were measured by Laverty Pathology (Sydney, NSW, Australia) or according to methods described by Haeckel ( 15 ).
Urinary and plasma protein concentrations were determined using the method of Lowry et al. ( 25 ). Urine samples were diluted 1:100, and plasma samples were diluted 1:1,000. Samples were read against a standard curve at 750 nm on a spectrophotometer (model UNICAM5625 UV/VIS).
Plasma renin levels were measured as the rate of formation of angiotensin I in nanograms per milliliter per hour at 37°C and pH 7.5 in the presence of added nephrectomized sheep substrate. Plasma angiotensinogen levels were measured as the amount of angiotensin I formed in plasma in the presence of an excess of human renin ( 33 ). Angiotensin I was measured by radioimmunoassay ( 31 ). Values are expressed as micrograms per milliliter.
Statistics
The Na + content of drinking water was zero. Na + intake in food was calculated with the assumption that the Na + content of chaff was 33 mmol/kg ( 14 ). Twenty-four-hour urine collections were used to determine daily urine output and Na + excretion.
For comparison of the effects of reduced renal mass on renal function, urine flow rates and concentrations of solutes were determined for each 30-min sample, and the average of all the 30-min samples was used. If the measurement of a variable depended on the urinary and plasma concentration of a substance, the urine flow rate and concentration of solute in the 30-min sample during which the blood sample was taken at the midpoint of the collection were used in the calculation. The values are averages of these calculated variables.
Values are means ± SE; n is the number of animals. The Statistical Package for the Social Sciences (SPSS, Chicago, IL) was used to determine means and SE. Data were compared using Student's t -tests or Wilcoxon's and Mann-Whitney's nonparametric tests. Linear regression analysis was performed using Graph Pad Prism version 1.03 or SPSS. Equations of best fit were derived using Curvefit (SPSS). The equation was chosen from the value of r 2, and the highest value was chosen.
RESULTS
Effects of Maternal Subtotal Nephrectomy on Maternal Morphology
Sixteen STNx and 15 intact ewes were studied at 119-132 days of gestation. STNx ewes were heavier than intact ewes ( P < 0.02; Table 1 ). The interval between subtotal nephrectomy and pregnancy and body weight were related: the longer the interval, the greater the body weight ( r = 0.82, n = 16, P = 0.0001). STNx ewes had larger hearts ( P < 0.005) and larger left ventricles ( P < 0.0005). The heart weight-to-body weight ratios of STNx and intact ewes were similar, but the left ventricular weight-to-body weight ratio was greater in STNx ewes ( P < 0.05), as was the left ventricular weight-to-heart weight ratio ( P = 0.052). Neither heart weight-to-body weight ratio nor left ventricular weight-to-body weight ratio was influenced by the length of time after nephrectomy.
Table 1. Effects of subtotal nephrectomy on maternal morphology
Figure 1 shows the remaining kidney from a pregnant STNx ewe. The total kidney mass was less in STNx than in intact ewes with both kidneys ( P < 0.0005; Table 1 ); thus the kidney weight-to-body weight ratio was less in STNx ewes ( P < 0.0005; Table 1 ). Despite ligation of a branch of the renal artery, the remaining kidney of STNx ewes was larger than the kidney that was removed at surgery [115.3 ± 4.6 (left kidney) vs. 66.1 ± 1.8 g, n = 15, P < 0.0005] and larger than the left kidneys of intact sheep ( P < 0.0005; Table 1 ). The left kidney was removed from one animal, and the remaining right kidney weighed 145.7 g postmortem. It was larger than the right kidneys of the control animals (76.5 ± 4.9 g, n = 13). The remaining left kidneys of STNx ewes were longer and wider than the left kidneys of intact ewes ( P < 0.005 and P < 0.05), but the length-to-width ratios were similar ( Table 1 ). The one remaining right kidney from the STNx ewe from which the left kidney was removed was also longer and wider than the right kidneys of intact ewes (11.0 x 6.3 vs. 8.1 ± 0.2 x 4.4 ± 0.1 cm, n = 11).
Fig. 1. Hypertrophied remaining kidney taken postmortem from a ewe subjected to subtotal nephrectomy (STNx). Kidney is on a 1-cm grid.
24-Hour Salt and Water Balance in STNx and Intact Ewes
Over a 24-h period, food intake was similar in STNx and intact ewes, and both groups produced similar amounts of feces ( Table 2 ). STNx ewes drank significantly more water than intact ewes ( P < 0.01), but their 24-h urine volumes were only slightly greater than those of intact ewes (not significant), so STNx ewes appeared to retain more water ( P < 0.05). Na + intake was similar in both groups. Salt excretion over 24 h by STNx ewes was slightly greater, but the difference in the amount and fraction of salt intake that was excreted over 24 h in STNx ewes was not significant compared with that in intact ewes ( P < 0.1; Table 2 ). The urinary Na + -to-K + ratio (U NaK ) was higher in STNx ewes ( P < 0.05). Osmolar and K + excretions were similar in the two groups of ewes.
Table 2. Maternal daily fluid and food intake, urinary excretion, and balance studies
Composition of Blood in STNx and Intact Ewes
Arterial P O 2 and pH were similar in STNx and intact ewes, and P CO 2 was slightly higher in STNx ewes but not significantly so ( P < 0.1; Table 3 ). Bicarbonate levels were higher in STNx ewes ( P < 0.01). Plasma osmolality, Na + and K + levels, and Na + -to-K + ratios were similar in the two groups, but plasma Cl - levels were lower in STNx ewes ( P < 0.01). Plasma Li + levels were greater in STNx ewes ( P = 0.054). Plasma creatinine levels were higher in STNx ewes ( P < 0.0005). Hemoglobin, hematocrit, and plasma protein levels were similar. Plasma glucose levels were higher in STNx ewes ( P < 0.005), but the higher levels of lactate in STNx ewes were not significantly greater ( P < 0.1) than those in intact ewes. Plasma renin and angiotensinogen levels were similar in both groups.
Table 3. Composition of maternal blood
Cardiovascular Function of STNx and Intact Ewes
Diastolic pressures were higher ( P < 0.05) and heart rates were lower ( P < 0.005) in STNx than in intact ewes. Although systolic pressure tended to be higher in STNx than in intact ewes, this difference was not significant ( P < 0.08; Fig. 2 ). The rate-pressure products in the two groups of ewes were also similar: 11,724 ± 1,042 ( n = 10) and 11,252 ± 516 mmHg·beats·min -1 ( n = 15) in intact and STNx ewes, respectively. There was no relationship between renal mass and arterial pressure or between the time after reduction in renal mass and arterial pressure. In STNx ewes, heart weight was directly related to mean arterial pressure ( r = 0.69, n = 15, P = 0.007), as was the weight of the left ventricle ( r = 0.59, P = 0.02; Fig. 3 ). These relationships were not observed in intact sheep. In intact sheep, heart rate and mean arterial pressure were related: the higher the heart rate, the higher the mean arterial pressure ( r = 0.66, n = 10, P < 0.05). There was no relationship between heart rate and arterial pressure in STNx sheep. Although resistance to renal plasma flow tended to be higher in STNx ewes [0.18 ± 0.04 and 0.35 ± 0.10 mmHg·ml -1 ·min in intact ( n = 10) and STNx ( n = 15) ewes, respectively], this difference was not significant.
Fig. 2. Arterial blood pressure (BP) of intact ( n = 10) and STNx pregnant ewes ( n = 15). Higher systolic pressure (SP) of STNx ewes was not significantly greater than that of intact ewes ( P = 0.08). Diastolic pressure (DP) of STNx ewes was significantly higher, and mean heart rate (HR) was lower. Values are means ± SE. * P < 0.05. *** P < 0.005.
Fig. 3. Relationships between heart weight and mean arterial pressure ( r = 0.69, n = 15, P = 0.007) and between left ventricle weight and mean arterial pressure ( r = 0.59, n = 15, P = 0.02) in STNx pregnant ewes. There was no relationship between arterial pressure and heart weight or left ventricle weight in intact ewes. Dashed lines, 95% confidence intervals.
Renal Function of STNx and Intact Ewes After Insertion of a Urinary Bladder Catheter
Glomerular function. ERPF measured as clearance of PAH was less in STNx than in intact ewes ( P = 0.01; Fig. 4 ). ERPF per kilogram body weight was also less in STNx than in intact ewes ( P < 0.005; Table 4 ), but there was no difference between the two groups in ERPF per gram kidney weight. ERPF, ERPF per kilogram body weight, ERPF per gram kidney weight, and time after reduction in renal mass were not related. GFR, measured as creatinine clearance, was less in STNx than in intact ewes ( P < 0.005; Fig. 4 ), as was GFR per kilogram body weight ( P < 0.005), but GFR per gram kidney weight was not different (Table 4 ). Filtration fractions of intact and STNx ewes were similar. The interval between reduction in renal mass and pregnancy and GFR were related: the longer the interval between reduction in renal mass and pregnancy, the higher the GFR ( r = 0.59, n = 15, P = 0.02); this relation did not occur when GFR was expressed relative to body weight but was evident when GFR was expressed per gram kidney weight ( r = 0.55, n = 15, P = 0.04). The excretion of protein was higher in STNx than in intact ewes ( P < 0.05), but the higher rate of clearance of protein was not significantly different ( P < 0.1; Table 4 ). The filtered loads of Na + and K + were lower in STNx than in intact ewes ( P < 0.005). Arterial pressure did not influence GFR or ERPF.
Fig. 4. Effective renal plasma flow (ERPF, measured as renal clearance of p -aminohippurate) and glomerular filtration rate (GFR, measured as renal clearance of creatinine) in intact ( n = 12) and STNx ( n = 15) pregnant ewes. Values are means ± SE. ** P = 0.01. *** P < 0.005.
Table 4. Maternal renal function measured using average of timed urine collections obtained via a bladder catheter
Tubular function. More Na + was excreted by STNx than by intact ewes ( P < 0.05). Because the filtered load of Na + was much less in STNx than in intact ewes ( Table 4 ), it follows that STNx ewes reabsorbed much less Na + ( P < 0.005). When Li + clearance was used to differentiate between proximal and distal tubular function, it was found that the reabsorption of Na + by the proximal nephron (R Na P) was lower in STNx than in intact ewes ( P < 0.005), but the amount reabsorbed by the distal nephron (R Na D) was similar in the two groups ( Table 4 ). The reduction in tubular Na + reabsorption meant that the percentage of the filtered Na + load reabsorbed was reduced in STNx ewes [99.41 ± 0.23% ( n = 15) vs. 99.97 ± 0.01 ( n = 12), P < 0.05; Fig. 5 ], because fractional R Na P was reduced [76.2 ± 1.6 ( n = 14) vs. 86.5 ± 2.0% ( n = 12), P < 0.0005; Fig. 5 ]. Fractional R Na D was increased: 23.7 ± 1.67 vs. 12.8 ± 2.0% ( P < 0.0005; Fig. 5 ).
Fig. 5. Fractional reabsorption of Na + by intact and STNx pregnant ewes. Total, percentage of filtered Na + load reabsorbed ( n = 12 intact, n = 15 STNx); Prox, percentage of filtered Na + load reabsorbed by proximal tubule ( n = 12 intact, n = 14 STNx); Dist, percentage of filtered Na + load reabsorbed by distal tubule ( n = 12 intact, n = 14 STNx). Clearance of Li + was used to distinguish proximal from distal Na + reabsorption ( 30 ). * P < 0.05. *** P < 0.0005.
Even though the distal nephron reabsorbed a greater fraction of the total amount of Na + reabsorbed by the nephron, it could not fully compensate for the greater load delivered to it, so the percentage of the distally delivered Na + load that was reabsorbed in the distal nephron was less in STNx than in intact ewes ( P = 0.054; Table 4 ).
Arterial pressure was not related to excreted Na +.
In intact ewes, the amount of Na + reabsorbed was directly related to GFR ( r = 0.99, P < 0.001, n = 12), as was R Na P ( r = 0.995, P < 0.0001, n = 12; Fig. 6 ), but there was no relation between GFR and R Na D ( Fig. 6 ). The amount of Na + delivered to the distal nephron strongly influenced R Na D ( r = 0.99, n = 12, P < 0.0001).
Fig. 6. Glomerulotubular balance in intact and STNx ewes: relationship between Na + reabsorption by proximal and distal nephron and GFR. Clearance of Li + was used to distinguish proximal from distal Na + reabsorption ( 30 ). A : intact ewes. Proximal Na + reabsorption was directly related to GFR ( r = 0.99, n = 12, P < 0.0001). There was no relationship between amount of Na + reabsorbed by distal tubule and GFR. B : STNx ewes. Proximal Na + reabsorption was directly related to GFR ( r = 0.98, n = 14, P < 0.0001). There was a direct relationship between amount of Na + reabsorbed by distal tubule and GFR ( r = 0.80, n = 14, P = 0.0006).
Equations relating the amount of Na + reabsorbed to GFR are as follows
( 1 )
( 2 )
where TR Na is the total amount of Na + reabsorbed. In intact ewes, the relationships between R Na P and plasma protein concentration ( P = 0.07), between R Na P and filtration fraction, and between R Na D and plasma protein levels were not significant.
In STNx ewes, TRNa and R Na P were related directly to GFR ( r = 0.98, P < 0.0001, n = 15; r = 0.97, P < 0.0001, n = 14), but as well R Na D was dependent on GFR ( r = 0.80, P = 0.006, n = 14; Fig. 6 ). The equations relating amounts of Na + reabsorbed to the GFRs of STNx ewes are as follows
( 3 )
( 4 )
( 5 )
The amount of Na + delivered to the distal nephron of STNx ewes strongly influenced the amount of Na + reabsorbed ( r = 0.99, n = 14, P < 0.0005). In STNx ewes, R Na P was higher if plasma protein levels were higher ( r = 0.74, n = 14, P = 0.002); there was no relationship between R Na P and filtration fraction. R Na D and plasma protein levels were not related.
The filtered load of K + was significantly reduced in STNx ewes ( P < 0.005), but the amount of K + excreted in the urine was the same as in intact ewes ( Table 4 ), because there was net secretion of K +, rather than net reabsorption as in intact ewes ( Fig. 7 ); this difference was significant ( P = 0.01). U NaK was higher in STNx than in intact ewes ( P < 0.02).
Fig. 7. Tubular handling of K + by intact ( n = 12) and STNx ( n = 15) ewes. Horizontal line, 0 net reabsorption of K +. Values are means ± SE. ** P = 0.01.
Plasma renin and angiotensinogen levels were similar in intact and STNX ewes. Arterial pressure was not related to plasma renin or angiotensinogen levels in intact or STNx sheep, nor was plasma renin related to filtered Na + load and TR Na.
In intact sheep, the amount of Na + delivered to the distal tubule and R Na D were related to plasma renin: the greater the amount of Na + delivered to the distal tubule and the greater the R Na D, the lower the plasma renin. The relation was best described by an exponential ( r 2 = 0.95, n = 7, P < 0.0005 for distal delivery and plasma renin concentration and for R Na D and plasma renin concentration). Surprisingly, there was a similar, but weaker, relationship between R Na P and plasma renin concentration that was best described by a linear equation ( r 2 = 0.68, n = 7, P = 0.02). In intact sheep, there was no relation between the amount of Na + excreted and plasma renin concentration or between U NaK and plasma renin concentration.
In STNx sheep, there were no significant relationships between distal Na + delivery and plasma renin concentration, between R Na D and plasma renin concentration, or between R Na P and plasma renin concentration. There was an inverse relationship between excreted Na + and plasma renin concentration that was best described by a power function ( r 2 = 0.5, n = 8, P = 0.049).
With the use of data from both groups, it was found that angiotensinogen levels were inversely related to plasma Na + levels ( r = 0.64, P = 0.009, n = 15), but the inverse relationship between plasma renin concentration and plasma Na + levels was not significant ( P = 0.06). There were no relationships between urinary Na + excretion or U NaK and plasma angiotensinogen levels.
Effects of Maternal Renal Dysfunction on the Fetus
Body weights were similar in fetuses carried by intact and STNx ewes ( Table 5 ). There was no difference in the number of fetuses per pregnancy: 6 sets of twins and 9 singletons in the intact group and 3 sets of twins and 13 singletons in the STNx group. The male-to-female ratio was 12:9 in the intact group and 7:12 in the STNx group. The cotyledonary placental masses were similar ( Table 5 ). The weights of all fetal organs measured were similar in the two groups, as was nose-to-rump length. Interestingly, the abdominal circumferences of STNx fetuses measured at surgery were greater ( P < 0.01) than those of intact fetuses, but the values obtained postmortem were similar in the two groups. The fetuses in both groups had similar arterial blood gases and pH, arterial pressures, and heart rates ( Table 5 ).
Table 5. Characteristics of fetuses carried by STNx and intact ewes
DISCUSSION
We decided to remove one kidney and ligate a branch of the renal artery supplying the remaining kidney, because in a previous study uninephrectomy in the pregnant ewe did not cause hypertension, but reduction of the renal blood flow to the remaining kidney did ( 29 ). In our previous study, acute reductions in maternal renal blood flow had profound effects on fetal health ( 28 ). In the present study, we have shown the effects of a reduction in renal mass before pregnancy on maternal renal function, fluid and electrolyte balance, cardiovascular health, and reproductive performance. STNx ewes were capable of becoming pregnant and maintaining normal fetal growth and development but were at risk of fluid and electrolyte imbalance and had high blood pressures.
The weights of the right and left kidneys of intact ewes were the same. The remnant kidneys of STNx ewes were, despite ligation of a branch of the renal artery and evidence of scarring, 75% of the total renal mass of the intact ewes ( Table 1, Fig. 1 ). In addition, they were 78% larger than the kidneys that were removed at the first surgery. These kidneys were wider and longer; thus their compensatory growth must have been approximately symmetrical, except where there was scarring ( Table 1, Fig. 1 ). In the adult nonpregnant animal, the remnant kidney has been shown to increase in mass by 30% after unilateral nephrectomy (although this value is highly variable) ( 23 ). In animals with intact kidneys, pregnancy is also thought to increase renal mass, which may be due to hypertrophy ( 13 ) or an increase in water content ( 7 ). In the present study, the increase in mass was greater than most values reported for nonpregnant animals after a reduction in renal mass, suggesting that pregnancy and subtotal nephrectomy are additive stimuli for the hypertrophic process. This was also shown by Modai et al. ( 37 ), who found that unilateral nephrectomy and repetitive pregnancies in mice led to an increase in renal mass greater than that seen in virgin unilaterally nephrectomized mice.
To determine whether there were significant differences in renal function between intact and STNx ewes, we needed to obtain accurately timed urine collections. For this reason, we could not use urine volumes collected over 24-h periods, inasmuch as ewes do not comfortably tolerate a urinary bladder catheter for long periods of time. Therefore, on the morning of the experiment in which we were to study renal function, a bladder catheter was inserted into the urethra while the ewe was restrained, and renal function was studied for up to 4.5 h after she was returned to her cage and when plasma levels of PAH and Li + had equilibrated. These experiments provided accurate estimates of renal function, so that the differences between intact and STNx ewes could be measured. For example, although the threefold-greater salt excretion measured in STNx than in intact ewes using 24-h fluid balance studies was not significant ( Table 2 ), the difference in salt excretion between STNx and intact ewes measured more accurately after bladder cannulation was significant ( Table 4 ). However, these more accurate measurements cannot be used to estimate fluid and electrolyte balance over 24 h, because they do not take into account diurnal fluctuations in renal function.
Pregnancy alone increases GFR and ERPF in intact animals ( 5, 26 ). In nonpregnant intact ewes, GFR and ERPF are 0.66 and 3.8 ml·min -1 ·g -1, respectively ( 17 ). These values are less than those measured in pregnant intact ewes: 0.85 ± 0.07 and 4.78 ± 0.65 ml·min -1 ·g -1 for GFR and ERPF, respectively ( Table 4 ).
Because the mass of the remaining kidney in the STNx ewes was 75% of the total renal mass in the intact ewes, it was not surprising that ERPF and GFR were significantly lower in STNx than in intact ewes ( Fig. 4 ). Both were 60-65% of intact values. Expressed relative to kidney mass, ERPF and GFR were 80-85% of the values in intact ewes, and overall there was no difference in ERPF and GFR per gram kidney weight between intact and STNx ewes ( Table 4 50% (perhaps 60%) because of unilateral nephrectomy and ligation of a branch of the renal artery in STNx ewes. Therefore, because ERPF and GFR per gram kidney weight were not different from values for intact pregnant ewes, there must have been considerable hyperfiltration by the remaining nephrons of STNx ewes. GFR was also related to time after surgery: the longer the time between renal surgery and pregnancy, the greater the GFR and GFR per gram kidney weight. Because the weight of STNx ewes was also related to time after renal surgery, it would seem that GFR increased to cope with increased metabolic demand. This increase in GFR must be the result of increasing filtration by the remaining nephrons.
Although GFR increases in pregnancy, this physiological hyperfiltration does not damage the glomerulus, because there is no increase in glomerular capillary pressure ( 4 ). Furthermore, in pregnancies in animals with severe reductions in renal mass (5/6), Deng and Baylis ( 8 ) found levels of proteinuria and focal sclerosis that were similar to those in virgin rats with 5/6 renal mass reduction, suggesting that pregnancy did not exacerbate the damage in this model of glomerular injury. However, repeated pregnancies in unilaterally nephrectomized mice were associated with increasing hypertension and deterioration of renal function in the absence of renal histological damage ( 1 ).
Urinary protein excretion was significantly greater in STNx than in intact pregnant ewes. Inasmuch as proteinuria suggests renal damage ( 24 ), this might mean that over the long term these animals could show deterioration in renal function. However, urinary protein excretion was not greater in animals in which renal mass had been reduced for the longest periods of time. The impact of pregnancy on glomerular morphology in sheep with renal dysfunction requires further study.
Even with hyperfiltration that maintained the ERPF per gram and GFR per gram of STNx ewes at levels close to those of intact ewes and the increase in GFR per gram kidney weight with increased body weight, GFR was limited in STNx ewes, because their ERPF and GFR relative to body weight were only 55% of values in intact ewes ( P < 0.005; Table 4 ). This could account for the higher plasma Li + levels ( Table 3 ) in STNx ewes, inasmuch as the dose of Li + was based on body weight.
In catheterized STNx ewes, although the filtered load of Na + was substantially reduced by 62% ( P < 0.005), Na + excretion was 14-fold greater ( P < 0.05). Thus renal tubular Na + reabsorption was reduced in STNx ewes and glomerulotubular balance (GTB) was altered. In a previous study ( 30 ), we showed that, in normal nonpregnant adult sheep, GTB depended only on R Na P, and the proximal tubule reabsorbed 81% of the filtered Na + load. R Na D was independent of GFR and accounted for reabsorption of 18% of the filtered load. In the present study, we showed that, in the intact pregnant ewe, 87% of the filtered load was reabsorbed proximally and confirmed that R Na D (13%) was independent of GFR ( Fig. 6 ). In the STNx pregnant ewes, R Na P was only 76% compared with 87% in intact ewes, and, as in the immature kidney ( 30 ), the distal nephron participated in GTB ( Fig. 6 ), presumably as a result of the increase in distal Na + delivery. Participation of the distal nephron in GTB probably involved increased Na + reabsorption by the loop of Henle ( 30 ) as well as increased Na + reabsorption by more distal sites. Even though the distal nephron reabsorbed a greater amount of Na + ( Table 4 ), this did not compensate for the reduction in R Na P; thus Na + excretion was higher ( Table 4 ), and STNx ewes were hypochloremic ( Table 3 ).
Several factors may be responsible for the failure of the proximal tubules of STNx ewes to maintain GTB: a higher level of arterial pressure, alterations in peritubular oncotic pressure, increased levels of nonreabsorbable solutes in the proximal nephron, and/or altered levels of natriuretic hormones that act in the proximal nephron ( 35 ). The higher levels of arterial pressure in STNx ewes may have resulted in a greater peritubular capillary hydrostatic pressure. This could inhibit the movement of solute and water from tubule to peritubular capillary. Peritubular capillary oncotic pressure depends on the concentration of protein in the peritubular capillaries, which in turn depends on plasma protein levels, filtration fraction, and glomerular plasma flow ( 35 ). Glomerular plasma flow could not be measured in these studies. In STNx ewes, there was a direct relationship between plasma protein levels and R Na P, but there was no relation between R Na P and filtration fraction. However, STNx ewes did not have lower plasma protein levels than intact ewes ( Table 3 ), so it is unlikely that a reduced oncotic effect of circulating plasma proteins was responsible for the lower rates of R Na P in STNx ewes. STNx ewes also excreted more protein ( Table 4 ), but it is equally unlikely that this small increase in tubular protein concentration would have affected R Na P to any significant extent.
Using micropuncture, Pollock et al. ( 40 ) showed lower proximal intratubular concentrations of Na + in unilaterally nephrectomized rats, which could impede the rate of reabsorption. A lower intratubular Na + could result from the greatly increased single-nephron GFR of STNx ewes, a greater load of nonreabsorbable solute in the proximal tubule, and an osmotic diuresis.
Although STNx ewes appeared to retain more water than intact ewes ( Table 1 ), their hemoglobin levels, hematocrits, and osmolalities were similar ( Table 3 ), which suggests that they were not volume expanded; therefore, it is unlikely that the secretion of atrial natriuretic hormone and other natriuretic factors was increased.
The failure of the proximal tubule to maintain GTB and the greater loss of Na + ( Table 4 ) meant that STNx ewes were at risk of salt depletion and were, in fact, hypochloremic ( Table 3 ). In other experimental models in nonpregnant animals with reduced renal mass, Na + balance was often maintained ( 20, 40 ). Thus the combination of pregnancy and a reduction in renal mass may predispose STNx ewes to a negative salt balance. In unilaterally nephrectomized pregnant ewes in which renal blood flow to the remaining kidney was reduced to 30%, there was a fall in plasma Na + levels in 24-72 h, suggesting that Na + balance could not be maintained when renal function was acutely impaired ( 29 ). In rats, a maternal low-Na + diet programs for hypertension in offspring in adult life ( 2 ), so the instability in salt balance of our STNx ewes may also program the fetal sheep in a similar way.
Using combined data from intact and STNx ewes, we could show that plasma angiotensinogen levels were inversely related to plasma Na + levels, an observation that has not, as far as we know, been reported in pregnant sheep. Although plasma renin levels were similar in the two groups of sheep, it was only in intact sheep that plasma renin levels were inversely related to R Na D. Because the macula densa, which regulates renin release, is located in the distal nephron, it is not surprising that R Na D had a negative influence on plasma renin levels. However, it is not clear why this relationship was not apparent in STNx sheep. Two possible explanations are the higher blood pressure of STNx ewes and the participation of the distal nephron in GTB in STNx ewes. The higher arterial pressures in STNx ewes may have suppressed renin release. The reduction in R Na P and the participation of the distal tubule in GTB ( Fig. 6 ) may have overridden any effects of R Na D on renin release. The inverse relationship between plasma renin levels and Na + excretion in STNx ewes may have been mediated through aldosterone ( 22 ).
In a previous animal model of pregnancy-associated hypertension, in the unilaterally nephrectomized pregnant ewe in which renal blood flow was reduced acutely to 30% of resting level by inflation of a cuff around the renal artery, there was severe hypertension probably due to high renin levels ( 29 ). In the present model, renin levels were not raised, because 1 ) there was infarction of part of the remaining kidney, rather than a reduction in renal perfusion pressure; and 2 ) this procedure was carried out 6 mo before the experiment. Yet STNx ewes were hypertensive ( Fig. 2 ). Furthermore, their higher levels of arterial pressure were associated with left ventricular hypertrophy, suggesting that they were relatively longstanding. Thus this model of pregnancy-associated hypertension is not renin dependent and, in this respect, more closely resembles human hypertension in pregnancy, particularly where pregnancy-associated hypertension is the result of preexisting maternal hypertension or renal disease.
Rates of K + filtration were lower in STNx ewes, yet the rate of K + excretion by STNx ewes was not different from that by intact ewes, because they secreted more K + in the distal nephron ( Fig. 7 ). Thus, despite the reduction in filtered K + load, they remained in balance, and any potentially toxic effects of hyperkalemia were prevented. Despite the net secretion of K +, U NaK was greater in STNx than in intact ewes because of the failure of the distal nephron to completely reabsorb the greater amount of Na + presented to it as a result of failure of proximal tubular function ( Table 4 ). Sheep are herbivorous and have a high-K + -low-salt intake. Therefore, STNx ewes are at risk of salt depletion (as a result of their poor ability to maintain GTB) and hyperkalemia.
The inability of STNx ewes to reliably control their salt excretion ( Tables 2 and 4 ) may have stimulated their appetite ( 9 ), so they weighed more than intact ewes reared on the same pasture. They also had higher blood glucose levels ( Table 2 ). This was an unexpected finding, and the degree to which these higher blood glucose levels were related to their relatively greater body mass and were associated with insulin resistance was not determined. In women who have had gestational diabetes, the incidence of microalbuminuria is higher after the pregnancy ( 12 ).
STNx ewes were time mated and became pregnant. Their fetuses were similar in body weight to fetuses carried by intact ewes that were managed on the same pasture; there was no difference in the number of fetuses carried per ewe, and the cotyledonary placental masses were similar ( Table 5 ). Therefore, subtotal nephrectomy before mating did not affect reproductive performance. This meant that we had a suitable animal model with which to study the role of maternal renal dysfunction in fetal renal development and function and its potential role in fetal programming.
Maternal renal impairment throughout the whole of pregnancy did not restrict fetal growth ( Table 5 ). In fact, abdominal girth at surgery was greater in STNx fetuses. The increased abdominal girth of STNx fetuses at surgery suggests that they were growing faster than intact fetuses, but their growth rates slowed more after surgery. The reason for this increased growth rate of STNx fetuses before fetal surgery is not obvious, although it could be related to the unexpected finding of higher plasma glucose levels in STNx ewes.
Although epidemiological studies have linked fetal programming to low birth weight, experimental models of fetal programming for hypertension have been produced in which there is no reduction in birth weight. For example, fetal programming in sheep caused by early administration of dexamethasone ( 10 ) or unilateral nephrectomy ( 39 ) is not associated with low birth weight. Thus the lack of any effect of maternal subtotal nephrectomy on fetal growth does not mean that that fetal programming has not occurred. The changes in maternal cardiovascular physiology and altered maternal electrolyte levels may well program the fetus.
In summary, we have shown that subtotal nephrectomy in the ewe before mating does not affect fertility but is associated with alterations in maternal fluid and electrolyte balance, proteinuria, abnormal carbohydrate metabolism, hypertension, and left ventricular hypertrophy. This animal model of renal dysfunction in pregnancy is ideally suited for the study of gestational hypertension. The fact that the STNx ewe is compromised in terms of salt and K + homeostasis may indicate why the human fetus is at risk when there is moderate-to-severe maternal renal dysfunction. Also, there is the degree to which maternal renal dysfunction causes unstable maternal Na + balance, which may affect fetal renal development and program for renal disease and high blood pressure in offspring.
GRANTS
This work was supported by a grant from the National Health and Medical Research Council of Australia (Australia) to E. R. Lumbers.
ACKNOWLEDGMENTS
We thank P. Bode, Dr. J. Burrell, and V. Kumarasamy for technical assistance and Laverty Pathology for measuring creatinine levels.
【参考文献】
Averbukh Z, Weissgarten J, Cohn M, Rosenmann E, and Modai D. Dual challenge for glomerular damage in mice: uninephrectomy and repetitive pregnancies. Isr J Med Sci 26: 575-576, 1996.
Battista MC, Oligny LL, St.-Louis J, and Brochu M. Intrauterine growth restriction in rats is associated with hypertension and renal dysfunction in adulthood. Am J Physiol Endocrinol Metab 283: E124-E131, 2002.
Bauer JD, Ackerman PG, and Toro G. Bray's Clinical Laboratory Methods (7th ed.) St. Louis, MO: Mosby, 1968, p. 61-62.
Baylis C. Impact of pregnancy on underlying renal disease. Adv Ren Replace Ther 10: 31-39, 2003.
Berman LB. The pregnant kidney. JAMA 230: 111-112, 1974.
Cowley AW, Skelton MM, Papanek PE, and Greene AS. Hypertension induced by high salt intake in absence of volume retention in reduced renal mass rats. Am J Physiol Heart Circ Physiol 267: H1707-H1712, 1994.
Davison JM and Lindheimer MD. Changes in renal haemodynamics and kidney weight during pregnancy in the unanaesthetized rat. J Physiol 301: 129-136, 1980.
Deng A and Baylis C. Glomerular hemodynamic responses to pregnancy in rats with severe reduction of renal mass. Kidney Int 48: 39-44, 1995.
Denton D. The Hunger for Salt: An Anthropological, Physiological and Medical Analysis. New York, Springer-Verlag, 1982, p. 417-450.
Dodic M, May DN, Wintour EM, and Coghlan JP. An early prenatal exposure to excess glucocorticoid leads to hypertensive offspring in sheep. Clin Sci (Lond) 9: 149-155, 1998.
Drake AJ and Walker BR. The intergenerational effects of fetal programming: non-genomic mechanisms for inheritance of low birth weight and cardiovascular risk. J Endocrinol 180: 1-16, 2004.
Friedman S, Rabinerson D, Bar J, Erman A, Hod M, Kaplan B, Boner G, and Ovadia J. Microalbuminuria following gestational diabetes. Acta Obstet Gynecol Scand 74: 356-360, 1995.
Garland HO, Green R, and Moriarty RJ. Changes in body weight, kidney weight and proximal tubule length during pregnancy in the rat. Renal Physiol 1: 42-47, 1978.
Gibson KJ and Lumbers ER. Mechanisms by which the pregnant ewe can sustain increased salt and water supply to the fetus. J Physiol 445: 569-579, 1992.
Haeckel R. Simplified determinations of the "true" creatinine concentration in serum, and urine. J Clin Chem Clin Biol 18: 385-394, 1980.
Hill KJ and Lumbers ER. Renal function in adult and fetal sheep. J Dev Physiol 10: 149-159, 1988.
Hoy WE, Mathews JD, McCredie DA, Pugsley DJ, Hayhurst BG, Rees M, Kile E, Walker KA, and Wang Z. The multidimensional nature of renal disease: rates and associations of albuminuria in an Australian Aboriginal community. Kidney Int 54: 1296-1304, 1998.
Huxley RR, Shiell AW, and Law CM. The role of size at birth and postnatal catch-up growth in determining systolic blood pressure: a systematic review of the literature. J Hypertens 18: 815-831, 2000.
Ise T, Kobayashi K, Biller W, and Haberle DA. Sodium balance and blood pressure response to salt ingestion in uninephrectomized rats. Kidney Int 67: S245-S249, 1998.
Jones D and Hayslett JP. Outcome of pregnancy in women with moderate or severe renal insufficiency. N Engl J Med 335: 226-232, 1996.
Katz SA and Malvin RL. Renin secretion: control, pathways and glycosylation. In: The Renin-Angiotensin System, edited by Robertson JIS and Nicholls MG. New York: Gower, 1993, chapt. 24, p. 24.1-24.13.
Kleinknecht C, Terzi F, Burtin M, Laouari D, and Maniar S. Experimental models of nephron reduction: some answers, many questions. Kidney Int 49: S51-S54, 1995.
Kriz W and LeHir M. Pathways to nephron loss starting from glomerular diseases. Insights from animal models. Kidney Int 67: 404-419, 2005.
Lowry OH, Rosebrough NJ, Farr AL, and Randall JR. Protein measurement with the Folin reagent. J Biol Chem 193: 265-275, 1951.
Lumbers ER. The renal circulation in pregnancy. Adv Organ Biol 9: 301-316, 2000.
Lumbers ER, Bernasconi C, and Burrell JH. Effects of infusion of ACTH in the chronically catheterized pregnant ewe and her fetus. Am J Physiol Regul Integr Comp Physiol 274: R445-R452, 1998.
Lumbers ER, Burrell JH, Stevens AD, and Bernasconi C. Responses of fetal sheep to reduced maternal renal blood flow and maternal hypertension. Am J Physiol Regul Integr Comp Physiol 271: R1691-R1700, 1996.
Lumbers ER, Burrell JH, Stevens AD, and Weir BA. The effects of one clip one kidney hypertension in chronically catheterized pregnant ewes. Clin Exp Pharmacol Physiol 24: 336-343, 1997.
Lumbers ER, Hill KJ, and Bennett VJ. Proximal and distal tubular activity in chronically catheterized fetal sheep compared with the adult. Can J Physiol Pharmacol 66: 697-702, 1988.
Lumbers ER and Lee Lewes J. The actions of vasoactive drugs on fetal and maternal plasma renin activity. Biol Neonate 5: 23-32, 1979.
Lumbers ER and Stevens AD. Changes in fetal renal function in response to infusions of a hyperosmotic solution of mannitol to the ewe. J Physiol 343: 439-446, 1983.
Marsh AC, Gibson KJ, Wu JJ, Owens PC, Owens JA, and Lumbers ER. Chronic effects of insulin-like growth factor I on renin synthesis, secretion and renal function in fetal sheep. Am J Physiol Regul Integr Comp Physiol 281: R318-R326, 2001.
Mendez RE and Brenner BM. Glomerulotubular balance and the regulation of sodium excretion by intrarenal hemodynamics. In: The Regulation of Sodium and Chloride Balance, edited by Seldin DW and Giebisch G. New York: Raven, 1990, p. 105-131.
Meyer TW, Baboolal K, and Brenner BM. Nephron adaptation to renal injury. In: Brenner & Rector's The Kidney (5th ed.), edited by Brenner BM. Philadelphia, PA: Saunders, 1996, chapt. 44, p. 2011-2048.
Modai D, Averbukh Z, Weissgarten J, Bogin E, Rosenmann E, Cohn M, Goren E, and Shaked U. Uninephrectomy and repetitive pregnancies in mice: a combined stimulus for renal hypertrophy. Nephrol Dial Transplant 4: 701-706, 1989.
Montan S. Drugs used in hypertensive diseases of pregnancy. Curr Opin Obstet Gynecol 16: 111-115, 2004.
Moritz KM, Wintour EM, and Dodic M. Fetal uninephrectomy leads to postnatal hypertension and compromised renal function. Hypertension 39: 1071-1076, 2002.
Pollock CA, Bostrom TE, Dyne M, Gyory AZ, and Field MJ. Tubular sodium handling and tubuloglomerular feedback in compensatory renal hypertrophy. Pflügers Arch 420: 159-66, 1992.
Thomson N (Editor). Australian Indigenous Health Infonet . School of Indigenous Australian Studies, Edith Cowan University. http://healthinfonet.ecu.edu.au [2005].
作者单位：Department of Physiology and Pharmacology, School of Medical Sciences, University of New South Wales, Sydney, Australia