Main

Several factors may contribute to the development of arterial hypertension. They include neuronal control of peripheral vascular resistance, alterations in circulating humoral and local paracrine factors, a genetic predisposition, dietary sodium excess, and "life-style" factors such as smoking, obesity, and lack of exercise. Recently it was also suggested that babies with low birth weights may have higher risk of developing arterial hypertension later in adult life. A correlation between arterial pressure and birth weight has been found in several studies(1–4). Barker et al.(5–7) have shown in a number of epidemiologic studies that subjects born with low birth weight have a higher risk of developing hypertension, non-insulin-dependent diabetes mellitus, and hyperlipidemia, leading to premature death in adulthood.

A number of mechanisms have been proposed to explain the association between birth weight and adult arterial pressure. One such mechanism, posed by Brenner and Chertow(8), hypothesizes that an inborn reduction in nephron number results in reduced glomerular filtration surface area followed by a compensatory hypertrophy; hyperfiltration then occurs in the remaining nephrons, contributing to the development of essential hypertension. Clinical and experimental data support this hypothesis. In an autopsy study, low birth weight newborns had fewer nephrons than did normal birth weight newborns(9). In addition, experimental rat models have shown that either low protein intake during pregnancy or reduced placental perfusion causes fetal growth retardation associated with permanent oligonephronia(10). It has also been shown that the offspring of dams fed a protein-deficient diet during pregnancy have a low GFR(11). Thus, low birth weight may be associated with reduced nephrogenesis and oligonephronia promoting the development of hypertension in adulthood.

Recently it has been suggested that an excess of glucocorticoids crossing the placenta may result in fetal growth retardation and the development of adult hypertension(12). The aim of this study was to evaluate the effects of DEX, a synthetic glucocorticoid able to cross the placenta, on fetal renal development and postnatal maturation of renal function. In addition, a separate group of pregnant rats was treated with HYD, which is metabolized and inactivated in the placenta.

METHODS

Animals. Experiments were performed on pregnant female Sprague-Dawley rats. The rats were treated with DEX (Decadron, Merrk Sharp& Dohme, Sweden; 0.1 mg·kg birth weight-1 d-1 i.p.), HYD(Solu-Cortef, Upjohn Pharmacia, Sweden: 2.5 mg·kg-1 d-1) or vehicle (distilled water) from d 1 of pregnancy until parturition. After delivery, the pups were left with their mothers until weaning; after weaning pregnant rats and offspring received standard rat chow and water ad libitum. Newborns were studied at 1, 20, or 60 d after delivery.

Glomerular count. A modification of the method described by Larsson et al.(13) was used to count the number of glomeruli. The animals were anesthetized (Inaktin-Byk; 80 mg/kg) and their kidneys were rapidly removed, freed of perirenal tissues, and weighed. Each kidney was digested in 1 mL of 8 N hydrochloric acid with gentle agitation to obtain a homogenous suspension (1.5-2.0 h at room temperature). The incubation time was predetermined to digest tubuli but not glomeruli. After digestion, each sample was diluted with distilled water to give a final volume of 10 mL. Ten samples of 25 µL each were then transferred to an object slide, and the number of glomeruli was counted under a light microscope. The examiner was unaware of the different experimental groups.

Renal functional studies. After anesthesia with Inaktin-Byk the rats were intubated, and one jugular vein and one carotid artery were cannulated with polyethylene catheters. The ureters were exposed through a midline incision and cannulated with polyethylene catheters that were advanced almost to the entrance of the renal pelvis. Clearance of polyfructosan-S (Inutest; Laevosan-Gesellchaft, Linz, Austria) was used to determine GFR, as previously described(14). The rats were given an i.v. prime dose (1 mL/100 g of birth weight) of a solution containing 5% Inutest in isotonic saline, followed by a continuous infusion of the same solution at the rate of 1 mL·100 g birth weight-1 h-1. Clearance periods were started 60 min after the prime dose. In each rat, timed urine samples were collected, and GFR and sodium excretion rate were calculated from the mean of two to three clearance periods(approximately 30 min each).

Plasma and urine were analyzed for inulin using previously described methods(14) and were analyzed for sodium using a flame photometer. The mean arterial pressure was monitored using a Statham transducer and recorded on a polygraph (Grass model 7B). Urinary albumin concentration was determined on a MIRA autoanalyzer (CBAS, Japan) using immunoturbidimetry.

Tissue Na+ content. Tissue samples (femoral skeletal muscle and liver) were dried at 105°C for 24 h, and their dry weight was determined. Samples were dissolved in concentrated nitric acid at 38°C, and the sodium concentration was determined using flame photometry.

Histochemistry. For histochemical analysis, four to five kidneys were studied in each experimental group. Pregnant rats (20-d of pregnancy) were anesthetized, and fetal kidneys were rapidly removed and placed in B5 fixative (4% formalin + 6% mercury chloride) or 4% formalin. The fixed kidneys were dehydrated in ethanol and xylene in a vacuum infiltration processor (Histolab, Gothenburg, Sweden) and embedded in paraffin(15). Longitudinal/coronal sections (6 µm) were cut and placed on glass slides pretreated with 3-aminopropyltriethoxysilane(Sigma Chemical Co.) to prevent detachment of the sections during the incubation procedure. The paraffin-embedded sections were deparaffinized in xylene, dehydrated in ethanol, and washed in water. The B5-fixed sections were incubated for 5 min in an iodine solution and for 5 min in a 5% sodium thiosulfate solution to remove mercury deposits. For morphologic analysis, sections were stained with hematoxylin and eosin. For immunohistochemical detection of the PCNA, the sections were covered with 1% Nonidet P40 (Kebo) and rinsed in water. To inactivate endogenous peroxidase activity, the sections were incubated in a 1% hydrogen peroxide solution for 30 min and rinsed in PBS. The sections were incubated in horse serum (1:50) for 20 min, and then in a MAb against PCNA (1:100) (Dakopatts) overnight at room temperature. After rinsing in PBS, slides were incubated in a biotinylated anti-mouse antibody (1:200) (Dakopatts) for 30 min at room temperature. In the following incubation, avidin and biotinylated peroxidase were introduced(Vectastain; Vector, Burlingame, CA). The staining was performed using 3,3′-diaminobenzidine tetrahydrochloride (Sigma Chemical Co.) as a chromogen giving a brownish color to the positive cells. Some sections were lightly counterstained with hematoxylin. Some sections were incubated as described above without primary antibody. The sections were examined in an Olympus BM 60 microscope, and Tri-X black and white film (Kodak) was used for photography.

Statistical analysis. Values are presented as means ± SEM. The statistical analyses were performed with a t test and Wilcoxon test. A p value <0.05 was considered significant.

RESULTS

Prenatal treatment with DEX or HYD did not have a significant effect on the duration of pregnancy (parturition occurred on d 21 of pregnancy) or the litter size. The number of pups per litter was 10.2 ± 0.17(n = 5) for the sham-injected dams, 11.0 ± 0.35 (n= 4) for the DEX-treated dams, and 11.3 ± 0.5 (n = 3) for the HYD-treated dams. On postnatal d 1, pups whose dams were treated with DEX(DEX-rats) were 30% smaller compared with pups whose dams were sham injected(control rats) (Table 1). Kidney weight was also 30% lower in DEX-rats (Table 1), and, therefore, the kidney/body weight ratio was not statistically different between the two groups of rats(0.51 ± 0.04% versus 0.52 ± 0.03%, respectively). No significant differences were observed between control rats and pups whose dams were treated with HYD (HYD-rats) (Table 1). Postnatally, DEX-rats grew faster than control rats. By the time of weaning (i.e. 20 d of age) body and kidney weight had normalized in DEX-rats, and, thereafter, they remained comparable to those of control rats(Table 1). Postnatal body and kidney weights in HYD-rats were always similar to control values (Table 1).

Table 1 Body and kidney weights in rats during postnatal development
Full size table

Arterial blood pressure was measured on postnatal d 60 and was significantly higher in the DEX-rats than in control rats (Fig. 1). No significant difference was observed between HYD-rats and control rats.

Figure 1
figure 1

Effect of prenatal DEX or HYD treatment on mean arterial pressure in 60-d-old rats. Values are means ± SE(n = 5 in each group). The asterisk (*) indicates significant difference compared with control rats.

Full size image

Because HYD treatment does not seem to alter fetal development or adult blood pressure, the following experiments were performed in control and DEX-rats. Renal function was determined in 60-d-old rats (Table 2). Diuresis and fractional water reabsorption were similar in control and DEX-rats. However, GFR was significantly lower in DEX-rats compared with control rats. In addition, urinary Na+ excretion rate and fractional Na+ excretion were both significantly lower in DEX-rats than in control rats.

Table 2 Renal function in 60-d-old rats
Full size table

The plasma Na+ concentration was evaluated in 60-d-old rats, and no significant difference was found between DEX-rats and control rats (151± 1 versus 150 ± 0.8 mEq/L, n = 8). However, the Na+ content in both liver and skeletal muscle from 60-d-old DEX-rats was significantly elevated compared with control values(Table 3). The Na+ content was already higher in skeletal muscle from 1-d-old DEX-rats compared with control rats(Table 3). Urinary albumin excretion rate was significantly higher in 60-d-old DEX-rats (Table 4).

Table 3 Sodium tissue content in 1- and 60-d-old rats
Full size table
Table 4 Urinary albumin excretion rate in 60-d-old rats
Full size table

In addition to renal function in adult rats, we also studied the morphology of fetal kidneys (20 d of pregnancy). The rat kidney is still immature in late gestation and continues to differentiate until 1 wk postpartum. All stages of nephron differentiation are thus detectable. Nephrogenesis takes place at the periphery of the kidney. In DEX-rats, we observed a disproportionate reduction in the thickness of the cortex. Kidneys from DEX-rats appeared to have a greater proportion of mesenchymal-like connective tissue and fewer proximal tubule convolutions. The nephrogenic zone at the periphery, containing undifferentiated mesenchymal cells or cells in their early differentiation stages, seemed to be thinner in DEX-rats(Fig. 2). When kidney sections were probed with an antibody against a PCNA, we observed an intense staining in the nephrogenic zone in the control rats (Fig. 2). On the contrary, the staining for PCNA immuno-like reactivity was remarkably reduced in DEX-rats.

Figure 2
figure 2

Bright-field micrographs of kidney sections from fetal (20-d) control (CON) (A and C) and DEX-rats. Hematoxylin and eosin-stained sections(A and B) show a thinner nephrogenic zone (nz)(indicated by arrowheads) in DEX-rats compared with control rats.(C and D) Fetal kidney sections stained with an antibody against PCNA. Note that fewer positive cells are present in DEX-rats(D) compared with control rats (C). Figures representative of five experiments in each group. Bar: 100 µm in A and B; 200 µm in C and D.

Full size image

The number of nephrons was determined in 20-d-old rats, i.e. when nephrogenesis was completed and the permanent number of nephrons had been achieved. No immature glomeruli were observed in control rats nor in DEX-rats. Prenatal DEX treatment permanently altered kidney morphology; the number of glomeruli and consequently of functional nephrons was significantly decreased by approximately 60% in mature DEX-rats compared with control rats (Fig. 3).

Figure 3
figure 3

Effect of prenatal DEX treatment on number of nephrons in 20-d-old rats. Values are means ± SE(n = 5 in each group). The asterisk (*) indicates significant difference compared with control rats.

Full size image

DISCUSSION

This study shows that fetal growth retardation induced by prenatal DEX leads to oligonephronia that persists throughout life. Moreover, this study confirms previous reports that moderate doses of prenatal DEX increase arterial blood pressure in the offspring(12) and, for the first time, shows that DEX-treated rats have lower GFR, higher albuminuria, and seem to retain sodium 2 mo after the treatment has been discontinued.

New nephrons are formed in the periphery when mesenchymal cells are induced to differentiate by the branching of the ureteral bud. In rats, this process of nephron formation occurs from fetal d 13 and continues to d 7 postnatal(13). Because we did not observe immature glomeruli in control or in treated rats, the number of glomeruli we defined at 20 d of age represents the permanent number of functional nephrons. In the fetal kidney, undifferentiated mesenchymal cells continue to proliferate centrifugally until the branching of a new ureteral bud induces additional mesenchymal cells to differentiate into epithelial cells. Mesenchymal stem cells are found in the upper cortex above the branching tips of the ureteric tree, whereas stroma cells found in the more medullary regions cannot be induced to differentiate into epithelial cells(16). Mesenchymal cells in the upper cortex are rapidly proliferating, whereas fewer mitoses are observed in the medulla(17). In this study, we found a high mitotic rate in the nephrogenic zone in the fetus at d 20 of gestation. Prenatal DEX treatment markedly reduces total DNA content in several fetal organs including the heart, kidney, and lungs(18,19). PCNA, a cofactor for DNA-polymerase σ, plays an important role in DNA replication during cellular proliferation, and its expression is highest during the transition from the G1 to S phase of the cell cycle(20). After prenatal DEX, the PCNA staining and thus the number of cells in mitosis was reduced in the nephrogenic zone of fetal kidneys, resulting in a thinner superficial layer of mesenchymal cells.

The activity of several paracrine growth factors is also altered in fetal tissues exposed to DEX(21). Therefore, impairment of tissue growth by prenatal DEX treatment may reflect a deficit in cell proliferation that extends to a variety of cell types. Oligonephronia, therefore, may be a consequence of the reduction in the number of undifferentiated mesenchymal cells. Alternatively, glucocorticoids may directly alter the synthesis of specific proteins which may be essential for nephron formation. Indeed, several intra- and extracellular proteins are required for normal nephron induction and differentiation(16), and alteration in their gene expression elicits oligonephronia(22).

Even though DEX-treated rats were born with low weight, their postnatal body and kidney growth rates were higher, and by the time of weaning both body and kidney weight had normalized. The number of nephrons was lower in the offspring of DEX-treated dams, implying that the nephrons in DEX-treated rats had a compensatory growth. This compensatory adaptation to nephron deficit, which at first appears beneficial, may have a harmful long-term effect. Compensatory renal growth may accelerate the development of focal glomerulosclerosis, particularly when starting early in life(14,23), and it is associated with glomerular hyperfiltration(8). Long-term follow-up studies, after unilateral nephrectomy in humans(23) and in experimental animals, have shown morphologic and functional differences between compensatory renal growth starting in infancy and compensatory growth starting in adult life. When compensatory renal growth starts in infancy it is associated with cellular hyperplasia and a 2-fold increase in GFR(24). At later time periods, albuminuria and focal glomerulosclerosis occur and GFR rapidly begins to decline(14).

In this study, we found evidence that renal function was deteriorated in fetal growth-retarded rats when they approached adult life. GFR was reduced by 40% and nephron number by 60% in adulthood. This finding suggests that the nephrons in the offspring of DEX-treated rats are hyperfiltrating. In addition, the urinary albumin excretion rate was significantly higher in the offspring of DEX-treated dams. Microalbuminuria is thought to be a predictor of hypertensive kidney damage leading to proteinuria and a decline in GFR(25). These results support the hypothesis proposed by Brenner and Chertow(8) that congenital oligonephronia leads to a decreased surface filtration area which in turn induces glomerular hyperfiltration and hypertension. A link between congenital oligonephronia and hypertension is also supported by the finding that rat strains with genetic hypertension such as the spontaneously hypertensive rat and the Milan hypertensive rat have been shown to have fewer nephrons compared with normotensive control rats [for review, see Brenner et al.(26)].

In addition to alteration in GFR, we found that the urinary Na+ excretion rate was lower in offspring of DEX-treated dams. This may be simply a consequence of a lower Na+ load due to the lower GFR. However, fractional Na+ excretion was also decreased, suggesting that the nephrons of DEX-rats are reabsorbing Na+ at a higher rate than those of control rats. Diuresis and fractional water reabsorbtion were not altered in DEX-rats, suggesting a specific alteration in tubular mechanisms regulating Na+ balance. Although plasma Na+ was maintained in the normal range, tissue Na+ content was increased in adult rats that were treated prenatally with DEX, indicating that they were in a positive sodium balance. Tissue Na+ content was already increased in 1-d-old pups, a finding that can be explained by a direct effect of DEX on fetal metabolism. The hypothesis that sodium retention precedes the onset of arterial hypertension has been extensively investigated(27). A positive Na+ balance in infancy may have permanent effects on the mechanisms regulating sodium excretion and contribute to the development of hypertension. For instance, it has recently been shown that high Na+ diet in the early postnatal period sensitizes the adult CNS to angiotensin II, resulting in sustained elevation of arterial blood pressure that is dependent upon intact sympathetic outflow to the kidneys(28).

It may be interesting to note that demographic groups in whom hypertension is more prevalent, such as Afro-Americans, have in general smaller kidneys, attain higher blood pressure in response to sodium load, and display a reduced capacity to excrete a sodium load(29). Fetal cortisol levels are elevated in human intrauterine growth retardation(30). Therefore, in these populations, congenital oligonephronia and sodium retention may be associated with fetal growth retardation and a higher incidence of hypertension.

Glucocorticoids reduce birth weight and alter organ maturation in both experimental animals and humans(31). Glucocorticoid levels are lower in fetal circulation than in maternal blood, and fetal cortisol levels have been shown to be elevated in newborns with intrauterine growth retardation(30). The fetus is primarily protected from excess maternal glucocorticoids by the placental enzyme 11β-OHSD, which converts cortisol (corticosterone in rats) to its inactive metabolites. In the present study, HYD, which is metabolized by the placental enzyme 11β-OHSD, does not affect birth weight, kidney weight, or adult blood pressure in the offspring. On the contrary, DEX, which is not metabolized by the placenta, exhibits profound effects on fetal body and kidney weight, and on adult blood pressure. The dosages of HYD and DEX we used (2.5 versus 0.1 mg) are of similar biologic potency. These results further support the hypothesis that an alteration in this placental enzyme may be harmful for fetal development(32). When the activity of 11 β-OHSD is lowered, the fetus may be exposed to higher levels of glucocorticoids, leading to reduced body weight in the newborn. It has been found that low activity of placental 11 β-OHSD correlates with low birth weight in rats(12) and in humans(33), and inhibition of the enzyme activity by carbenoxolone results in fetal growth retardation and hypertension in the offspring(34). Intriguingly, protein deprivation, which causes fetal growth retardation, is also associated with lower placental 11β-OHSD activity and hypertension in the offspring(35).

In summary, we have shown that rats exposed to moderate fetal levels of DEX but not HYD are growth-retarded in utero and develop hypertension in adulthood. Fetal growth retardation is associated with oligonephronia and a positive Na+ balance which may cause adult onset of hypertension. Future studies are required to determine precisely the mechanism(s) responsible for the development of arterial hypertension in offspring exposed to excess intrauterine glucocorticoids.