ang II is the main biologically active product of the renin angiotensin system and is well established as having major roles in the regulation of blood pressure and fluid homeostasis. It is produced by the cleaving of the precursor, ANG I, by angiotensin-converting enzyme (ACE). ANG I, however, in the presence of tissue-specific endopeptidases, can be cleaved into a different product, ANG-(1–7) (4). This peptide can also be produced from ANG II but exhibits some biological actions that are quite distinct from ANG II (10). Both ANG-(1–7) and ANG II can stimulate arginine vasopressin release (27), but recent studies indicate that ANG-(1–7) may oppose the pressor actions of ANG II (for review, see Ref. 12). It has been demonstrated that ANG-(1–7) is a vasodilator and has antihypertensive activity, particularly in animals with high basal pressures (1). ANG-(1–7) stimulates prostaglandin release from vascular endothelial cells in culture (17) and can inhibit vascular smooth muscle cell proliferation (13). ANG-(1–7) also augments bradykinin-induced relaxation of isolated blood vessels by inhibiting ACE and releasing nitric oxide (20).

ANG-(1–7) has been suggested to have agonist activity at the ANG II type 2 receptor (AT₂ receptor) and antagonistic actions on the type 1 receptor (AT₁; Ref.21). However, some effects of ANG-(1–7), such as the inhibition of vascular smooth muscle growth, cannot be blocked in the presence of both AT₁ and AT₂receptor antagonists (13). A novel binding site for ANG-(1–7) has been identified in bovine aortic endothelial cells, and this site can be blocked by a specific ANG-(1–7) antagonist, [d-Ala⁷]-ANG-(1–7) (31).

An important target organ for ANG-(1–7) is the kidney. The kidney itself contains the necessary enzymes required for ANG-(1–7) production, and thus the ANG-(1–7) may act in a paracrine manner in this organ. It is produced at the brush border of the proximal tubule (29) and is present in large quantities in rat urine (6). In the isolated rat kidney, exogenous ANG-(1–7) stimulated a marked diuresis and natriuresis as well as increased glomerular filtration rate (GFR) (8). Production of renal prostaglandins may play a role in the natriuretic actions of ANG-(1–7) (16). In the spontaneously hypertensive rat treated for 2 wk with ANG-(1–7), urine and sodium output were increased over the first 3 days but then returned to preinfusion levels. This effect was not observed in Sprague-Dawley or Wistar-Kyoto rats (15). Other investigators have reported that ANG-(1–7) causes an antinatriuresis in water-loaded rats, an effect that could be blocked by a specific ANG-(1–7) antagonist (26). These conflicting results indicate that sodium and water status, as well as basal blood pressure, may be important in determining the overall effect of infused ANG-(1–7).

The developing ovine fetus produces relatively large amounts of urine, which is hypotonic when the fetus is not stressed (36). This urine is excreted into the amniotic fluid that surrounds the fetus or into a second fluid-filled compartment, the allantois. In many ways, the fetus is similar to the water-loaded adult. The renin angiotensin system is present in the fetus from early in gestation and plays important roles in kidney development (34, 35). Plasma levels of renin are higher in the fetus over the last third of gestation than in the adult (22). The high levels of renin in the fetus may indicate increased production of ANG I and thus provide large amounts of substrate for ANG-(1–7) production, and we therefore hypothesized that ANG-(1–7) concentrations would be high in the fetus. We also hypothesized that ANG-(1–7) infusion may have an antidiuretic effect on the fetus similar to that observed in water-loaded animals (26). As some effects of ANG-(1–7) may be mediated by the AT₁ or AT₂ receptor, we examined gene expression levels of these receptors as well as renin, to determine any role of ANG-(1–7) in the long-term regulation of these genes.

METHODS

All experiments were approved by the Animal Ethics Committee of the Howard Florey Institute. Fetuses were cannulated between 98 and 102 days under general anesthesia as described previously (21). This involved the placement of a cannula (internal diameter 0.76 mm, outer diameter 1.65 mm) into one carotid artery, one jugular vein, and the fetal bladder. A cannula was also placed in the amniotic fluid and a maternal jugular vein. Animals were allowed to recover from surgery for at least 7 days and had free access to food and water at all times. No infusions were performed unless the fetal urine osmolality was <180 osmol/kgH₂O.

Basal hormone concentrations.

On day 111 of gestation, basal blood samples were taken from the fetal arterial cannula (n = 10) for plasma ANG I, ANG-(1–7), ANG II, and renin. A total of 12 ml blood was taken, and this was replaced with an equal volume of isotonic saline. A similar sample was taken from the maternal jugular vein. Samples for ANG I, ANG-(1–7), and ANG II were taken rapidly (<5 s) and immediately placed into chilled tubes containing an inhibitor cocktail (0.01 mM SR-42128, 0.015 mM pepstatin, 5 mM phenanthroline, 12.5 mM EDTA, 0.2 g/l neomycin sulfate, 0.2% DMSO, 0.2% ethanol; all shown as final concentration in blood sample). The SR-42128 has been shown to inhibit sheep renin with 50% inhibition at 30 nM and 90% inhibition at 200 nM. The SR-42128 was a generous gift from Sanofi, Direction des Recherches, Centre de Montpellier, France. Samples for renin were taken in tubes coated with EDTA. Samples were spun at 4°C, and plasma was stored at −80°C until assayed.

Measurement of blood pressure and GFR.

Fetal urine was drained for 1 h, and a background sample (2 ml) was taken. A blood sample (2 ml) for background counts was taken from the arterial line. This cannula was then connected to a Cobe transducer for measurement of blood pressure. The amniotic cannula was also attached to a pressure transducer, and the blood pressure was recorded as arterial pressure minus amniotic fluid pressure. Blood pressure was measured continuously (a 10-s sample every 10 min) and stored on a personal computer using a data-acquisition system. The fetus then received an infusion of [⁵¹Cr]EDTA (∼10 μCi/h) intravenously. After a 1-h equilibration period, urine was collected continuously into a measuring cylinder for 1-h periods for a total of 3 h. At the midpoint of each urine collection, a 2-ml blood sample was taken from the arterial cannula. Duplicate samples of plasma and urine (0.5 ml) were counted on a Cobra 5010 gamma counter (Packard Instruments, Downers Grove, IL).

Chronic infusions of ANG-(1–7).

After taking basal blood samples and measuring GFR, fetuses received an intravenous infusion of ANG-(1–7) (∼9 μg/h, Auspep, Parkville, Australia, n = 10) or saline (n = 5) at a rate of 0.19 ml/h via a Braun perfusor pump for 3 days. At 24, 48, and 72 h after commencement of the infusion, urine flow rate and composition were measured in all fetuses. On these days, urine was drained for 2 h. All measurements of urine flow were performed at the same time of the day. Blood samples (12 ml) were taken for all hormones as described above. During the last 3 h of the infusion, fetal GFR was measured. Four fetuses that had received ANG-(1–7) infusion and five saline-infused fetuses were killed with an overdose of Lethobarb (Arnolds, Reading, Australia, 100 mg/kg), and fluid volumes were measured. Samples of amniotic and allantoic fluid were taken for biochemical analysis. Approximately 3 g of kidney from each animal was frozen in liquid nitrogen and stored at −80°C. Blood pressure and urinary data from the animals receiving saline have been published elsewhere (24).

Infusions of [d-Ala⁷]-ANG (1-7).

Four fetuses that had received an infusion of ANG-(1–7) for 3 days were maintained for a further week and were infused with the specific antagonist of ANG-(1–7), [d-Ala⁷]-ANG-(1–7) (20 μg/h, Auspep) for 3 days. The protocol was identical to that for the ANG-(1–7) infusions with respect to measurement of blood pressure and renal function. However, blood samples were taken only for analysis of plasma renin.

Sample analysis.

Sodium, potassium, chloride, urea, creatinine, magnesium, phosphate, calcium, glucose, lactate, and fructose were measured in fetal plasma, fetal urine, amniotic fluid, and allantoic fluid using a Synchron CX-5 Clinical System (Beckman, Fullerton, CA). Osmolality was measured in these fluids by freezing-point depression using an osmometer (Advanced Instruments). Hematocrits were measured in duplicate. A blood gas machine (Ciba Corning 278, Australian diagnostics, Melbourne, Australia) was used to measure pH, Po₂, and Pco₂ in fetal blood.

Hormones.

Plasma renin concentrations were measured by the generation of ANG I and measured in picomoles ANG I per milliliter per hour (9). Angiotensin levels (ANG I, ANG II, and ANG-(1–7)) were measured using HPLC-based radioimmunoassays as described previously (3).

Gene expression studies.

Total RNA was extracted using the method of Chirgwin et al. (7). Samples were treated with DNase to remove any genomic DNA. One microgram of total RNA was reverse transcribed in a 10 μl reaction containing 1× TaqMan RT buffer, 5.5 mM MgCl₂, 500 μM each 2′-deoxynucleoside 5′-triphosphate, 2.5 μM random hexamers, 0.4 U/μl RNase inhibitor, and 1.25 U/μl MultiScribe reverse transcriptase (PE Biosystems). Controls were set up where no reverse transcriptase was included in RT reactions with all total RNA samples. The RT reactions were carried out in a GeneAmp PCR System 9600 (PE Applied Biosystems) at 25°C for 10 min, 48°C for 30 min, and 95°C for 5 min. Each reaction was then diluted 1/10 in 0.01 M EDTA pH 8.0 and stored at −80°C. A real time PCR assay (ABI PRISM 7700 Sequence Detector, PE Biosystems) was used to assess relative gene expression levels of renin and the AT₁ and AT₂ receptors. This method has been used previously to examine expression levels of these genes in fetal kidney and primer and probe sequences, along with optimal conditions for its use in real time PCR (25). The intra-assay coefficients of variation for this method are 14, 7, and 5% for renin, the AT₁, and the AT₂ receptor, respectively. In brief, a comparative cycle of threshold fluorescence (C_T) method was used to assess relative gene expression levels, where a C_T value reflects the cycle number at which DNA amplification is first detected. Ribosomal RNA, 18S, was used as an endogenous reference. One sample of kidney from an animal that had received saline was assayed five times, and the mean value was calculated. This mean value was used as a “calibrator” to which all other samples were compared. Thus comparative C_Tcalculations for the expression of renin, AT₁ receptor, and AT₂ receptor were all relative to an internal control. First, 18S C_T values were subtracted from renin, AT₁, and AT₂ receptor values for each sample to give a ΔC_T value. ΔΔC_T values were achieved by subtracting the calibrator ΔC_T value from each ΔC_T value. The expression of renin, AT₁receptor, and AT₂ receptor relative to the calibrator was evaluated using the expression 2^−ΔΔC_T.

Statistics.

All data are reported as means ± SE. A two-way repeated-measures ANOVA was used to examine changes over time. A t-test was used to compare the fluids at postmortem between the ANG-(1–7) and the saline-infused fetuses and between the basal fetal and adult hormone concentrations. Where appropriate, values were log transformed before analysis.

RESULTS

Blood gases.

Fetal blood gases were in the normal range for fetuses in this laboratory. On the day of basal blood samples, pH, Pco₂, and Po₂ were 7.44 ± 0.01, 44.4 ± 0.8, and 22.0 ± 1.6 mmHg, respectively. Hematocrit was 32 ± 1%. These parameters did not change over the course of the 3-day infusion of ANG-(1–7) or the infusion of [d-Ala⁷]-ANG-(1–7).

Basal hormone concentrations.

Basal concentrations of ANG 1 and ANG-(1–7) in the fetus were 86 ± 21 and 13 ± 2 fmol/ml plasma, respectively (see Fig. 1). These were significantly higher than in the ewe, where concentrations were 20 ± 4 and 2.9 ± 0.6 fmol/ml, respectively (P < 0.05). ANG II concentrations, however, were similar (14 ± 2 in the fetus and 10 ± 1 fmol/ml in the ewe). Renin concentrations in the fetus were 4.8 ± 1.1 pmol · ml⁻¹ · h⁻¹, a value significantly higher than in the ewe (0.9 ± 0.2 pmol · ml⁻¹ · h⁻¹).

Infusions of ANG-(1–7).

The peptide content of the ANG-(1–7) was 68% of the actual weight, making the estimated infusion rate 13.6 μg/h. The concentration of the ANG-(1–7) measured in the infusate was 53 ± 5 nmol/ml (∼9 μg/h, n = 4). The plasma concentrations of ANG I, ANG-(1–7), and ANG II over the course of infusion can be seen in Fig. 1. Plasma concentrations of ANG-(1–7) increased significantly over the infusion protocol to a maximum on day 3 of 448 ± 146 fmol/ml (P < 0.05). ANG I and ANG II concentrations were unchanged over this period (Fig. 1, Band C). Plasma renin concentrations were also unchanged by the infusion protocol (Fig. 2). In the ewe, the concentrations of all hormones measured did not differ over the course of the infusion protocol (see Fig. 1).

Clearance rate.

As the infusate was measured to be 53 nmol/ml and the infusion rate was 0.19 ml/h, the fetus received 10.07 nmol/h. The clearance rate of ANG-(1–7) from the fetal circulation was calculated to be 57 ± 11, 81 ± 11, and 65 ± 40 l/h over each 24-h period of infusion. The large standard error in the measurement over the final 24 h reflects the large range of ANG-(1–7) concentrations in fetuses on day 3 of infusion.

Blood pressure.

Fetal blood pressure can be seen in Fig.3 and did not change over the 3-day infusion period. Basal pressure was 41 ± 1 mmHg, and values at 24, 48, and 72 h of infusion were 40 ± 1, 39 ± 2, and 39 ± 1 mmHg, respectively. Values at the corresponding time points in saline-infused fetuses were 39 ± 2, 40 ± 1, 41 ± 2, and 43 ± 3.

Kidney function.

Fetal urine flow rates can be seen in Fig. 3. Urine flow was not altered by infusion of ANG-(1–7) and was not different from fetuses receiving saline. GFR was 83 ± 7 ml/h before the infusion and 83 ± 6 ml/h after 3 days of ANG-(1–7). These values were not significantly different, nor were they different from values in saline-infused fetuses (94 ± 4 ml/h basal, 104 ± 6 ml/h post 3 day saline). There was no change in fetal urine osmolality or in the excretion rates of sodium, potassium, or chloride (see Table1).

Fetal fluids.

There was no difference in the volumes of either amniotic or allantoic fluid compared with saline-infused fetuses of similar age. Because of large variation between animals, ion concentrations in these fluids were log transformed before analysis. Osmolality in both the amniotic and allantoic fluid of the ANG-(1–7)-infused fetuses was significantly higher than in saline-infused fetuses (see Table2, P < 0.05). The allantoic fluid also had lower chloride concentrations (P < 0.05) and higher total protein. The sodium-to-potassium ratio in the allantoic fluid was not different between the groups, although the ANG-(1–7)-infused group tended to have lower sodium and higher potassium concentrations.

Infusion of [d-Ala⁷]-ANG-(1–7).

The peptide content of this preparation was 71%. Infusion of the specific antagonist of ANG-(1–7) had no effect on GFR, urine flow rate, or urinary composition (Table 1). There was no change in fetal blood pressure (data not shown). Renin concentrations tended to decline over the course of infusion, but this did not reach the level of significance (P = 0.07, treatment × time interaction, see Fig. 2).

Gene expression studies.

Results from the real time PCR analysis of gene expression levels of the AT₁ and AT₂ receptors along with renin are shown in Table 3. Both receptors showed increased gene expression levels in the kidneys of fetuses that had received ANG-(1–7) (P < 0.05). Renin gene expression was also increased significantly by the ANG-(1–7) infusion (P < 0.01).

DISCUSSION

This study shows for the first time that ANG-(1–7) is present in fetal plasma and circulates at much higher levels than observed in the adult sheep. A previous study had showed jugular vein concentrations of ANG-(1–7) in nonpregnant adult sheep to be 2.2 ± 0.4 fmol/ml (19), which was similar to samples obtained from the pregnant ewes in this study. Studies in the human have shown similar circulating concentrations of ANG-(1–7) as seen in the adult sheep, with concentrations being ∼10-fold higher in people on ACE inhibitor therapy for hypertension (19). Basal ANG II concentrations in fetuses in this study were between 4 and 20 fmol/ml, which is similar to adult values in this study. Thus, in the adult, ANG-(1–7) concentrations are only 10% of ANG II levels, whereas in the fetus, concentrations of each peptide are approximately equal.

ACE activity has been detected in the ovine fetal kidneys (meso- and metanephros), lung, brain, and liver from as early as 41 days of gestation (35), in the lungs of human fetuses in the second trimester (28), and in the lungs and kidneys of late-gestation rats (33). There is sufficient ACE in the fetus to produce comparable plasma concentrations of ANG II to those seen in the ewe, but the higher levels of circulating ANG I in the fetus indicate a large percentage of ANG I is not converted to ANG II. Cleavage of ANG I into ANG-(1–7) can occur by a number of pathways (for review, see Ref. 11), and because the different enzymes are located in various tissues, it is thought that regulation of ANG-(1–7) production occurs at the tissue level (12). From the results in this study, we could speculate that concentrations of the enzymes required for ANG-(1–7) production may be high in some tissues of the fetus. No study has yet been conducted to examine this. In the adult sheep, high levels of these enzymes were found in the median eminence, which may account for the high concentrations of ANG-(1–7) in hypophysial portal blood (19).

The infusion of ANG-(1–7) to the fetus, although raising plasma concentrations by 15- to 35-fold over the course of the 3 days, did not have any effect on blood pressure or renal function. Plasma changes of this magnitude have been seen for ANG-(1–7) in rats on ACE treatment (4). The dose used in this study was 53 nmol/ml or ∼10 nmol/h, and, given that fetuses of this age weigh ∼1.5 kg, this equates to over 600 pmol · 100 g body wt⁻¹ · h⁻¹. This is considerably higher than the doses used of 20–80 pmol/100 g body wt to produce a potent antidiuresis in a study in water-loaded rats (26) but very similar to doses of 100 pmol · kg⁻¹ · min⁻¹ that caused a diuresis in the anesthetized rat (32). Infusion of the ANG-(1–7) antagonist alone had no effect on urine volume or composition in the ovine fetus at the dose used in this study. We cannot rule out that this dose may not have been sufficient to completely block the endogenous ANG-(1–7), but it was a similar dose on a body weight basis to that which blocked the antidiuretic effects caused by infusion of ANG-(1–7) in another study (26).

Interestingly, the infusion of ANG-(1–7) caused a significant increase in osmolality in both amniotic and allantoic fluids and some small changes in solute concentrations. A rise in osmolality of the allantoic fluid has been observed in ovine fetuses during acute urine drainage (14), and, although the urine of fetuses in this study was drained for 2–4 h per day, an identical protocol occurred in the saline-infused controls. Increases in allantoic osmolality can occur only by water leaving the cavity through the intramembranous pathway (absorption of water by the vascularized chorion). Although not examined in this study, ANG-(1–7) may have had effects on the permeability of these membranes.

An interesting result was the significant increase in mRNA levels for the AT receptors and renin in the kidneys of fetuses that had received infusions of ANG-(1–7). In the ovine fetus at this age, there is abundant expression of both the AT₁ and AT₂ receptor in the kidney (2, 34). Infusions of ANG II for 3 days to fetuses at 80 days of gestation age lead to a downregulation of mRNA in the kidney for the AT₁ receptor as well as for renin (25). We also found that infusion of ANG I to the fetus around 120 days of gestation causes a decrease in AT₁ receptor gene expression (unpublished). The increase in gene expression in this study suggests an opposing action of ANG-(1–7) to that of ANG II in the regulation of AT receptors in the fetal kidney. High concentrations of ANG-(1–7) during the infusion may compete with ANG II for the AT₁ and AT₂ receptors. It is possible this may lead to upregulation of gene expression for renin and the AT receptors. Although not measured in this study, we previously showed that protein levels closely mirror mRNA expression levels at least for the AT₁ receptor in the fetal kidney (2). It is not known if any other binding site(s) for ANG-(1–7) are present in the ovine fetus. A binding site with high affinity for ANG-(1–7) has been identified in bovine aortic endothelial cells (31), but no examination has been made of fetal tissues. Lack of any physiological effect of the ANG-(1–7) antagonist suggests they may be absent. This would not be surprising, inasmuch as the distribution of both the AT₁ and AT₂ receptors is distinctly different in the ovine fetus compared with the adult in both blood vessels (17) and kidney (2). The increase in mRNA for renin was unexpected because plasma concentrations were unaltered by the ANG-(1–7) infusion. However, the tissue levels of renin in the kidney may be higher than circulating concentrations. The ANG-(1–7) may also have caused changes in the processing or clearance rate of renin.

Perspectives

In conclusion, basal fetal plasma concentrations of ANG-(1–7) were significantly higher than in the adult, although ANG II concentrations are similar. Raising plasma levels of ANG-(1–7) did not appear to affect renal function or blood pressure, but changes in fetal fluids and gene expression in the kidney suggest that ANG-(1–7) may play a role in the developing ovine kidney.

FOOTNOTES

• This work was supported by a block grant (983001) to the Howard Florey Institute from the National Health and Medical Research Council of Australia. The real time PCR machine (ABI PRISM detector system) was purchased with grants from the Sylvia and Charles Viertel Foundation, the Clive and Vera Ramaciotti Foundation, the Phillip Bushell Foundation, and the Harold and Cora Brennen Benevolent Trust.

• Address for reprint requests and other correspondence: M. Wintour, Howard Florey Institute of Experimental Physiology and Medicine, Univ. of Melbourne, Parkville 3010, Australia (E-mail:[email protected]unimelb.edu.au).

• The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.