Hemodynamic Responses to Angiotensin-(1–7) in Women in Their Third Trimester of Pregnancy

Abstract

Background:

To understand the role of Angiotensin-(1–7) (Ang-(1–7)) in vasculature of pregnant women, we compared cardiac output (CO), total peripheral resistance (TPR) and forearm blood flow (FBF) responses to Ang-(1–7) infusion between normotensive pregnant women in their third trimester and healthy age matched non-pregnant women. The responses of skin microcirculation to Ang-(1–7) were tested in preeclamptic, normotensive pregnant and non-pregnant women. Responses to Angiotensin II (Ang II) were also determined.

Methods:

Non-invasive methods for systemic (bioimpedance and VascuMAP), FBF (venous occlusion strain gauge plethysmography), and skin (laser Doppler) hemodynamics assessments were used.

Results:

Compared to non-pregnant women, systemic infusion of Ang-(1–7) (2000 pmol/min) resulted in a greater increase in CO (9.4±6.4 vs. −3.3±2.1%, n=9–10) in normotensive pregnant women. Brachial local infusion of Ang-(1–7) had no effect on FBF in either group. In non-pregnant and normotensive pregnant women, local Ang II induced a dose-dependent decrease in FBF and increase in forearm resistance at 50 and 100 pmol/min (p<0.05 vs. corresponding baseline, n=7–10). Following iontophoretic application of 5 mmol/l dose of Ang-(1–7), the change in skin flow was higher in normotensive pregnant versus preeclamptic women (182.5±93 vs. 15.76±19.46%, n=14–15). Skin flow was lower in normotensive pregnant versus preeclamptic women (−46.5±48.7 vs. 108.7±49.1%, n=14–15) following Ang II infusion at 1.0 pmol/min.

Conclusion:

In the third trimester of pregnancy Ang-(1–7) induces alterations in CO and differentially regulates micro- and macrocirculations, depending on the dose. Dysregulation in skin vasculature may contribute to the development of vascular dysfunction and hypertension in preeclampsia.

Introduction

The function of local vascular beds during pregnancy depends on systemic and local vasoactive factors. Components of the renin-angiotensin system (RAS) – such as renin, angiotensinogen, angiotensin II (Ang II), and angiotensin-(1–7) (Ang-(1–7)) – are elevated in the circulation of pregnant women¹. However, plasma Ang-(1–7) levels are reduced in women with preeclampsia in their third trimester¹. Since the RAS regulates cardiovascular function and body fluid and electrolyte balance, Ang-(1–7) and Ang II – two RAS peptides with opposing actions – may be involved in vascular adaptations in pregnancy. Outside of pregnancy, Ang-(1–7) induces vasodilation and anti-proliferation, and Ang II elicits vasoconstriction, cell proliferation, and hypertrophy. The peripheral vasculature of pregnant women is resistant to pressor effects of Ang II, in part due to reduced sensitivity of arterial smooth muscle and down-regulation of AT1 receptors^2–4. However, the contribution of Ang-(1–7) to systemic or peripheral circulation in pregnant women is not known.

Ang-(1–7) is expressed in vascular endothelial and smooth muscle cells (VSMC), and can influence the release of endothelium-derived vasodilatory factors or factors attenuating VSMC hypertrophy, suggesting that Ang-(1–7) can contribute to the vasodilatory state in pregnancy^5–7. Ang-(1–7) induces vascular effects via activation of the Mas/AT_1–7 receptor⁸. In the third trimester of pregnancy Ang-(1–7) may act to reduce total peripheral resistance (TPR) and increase blood flow via systemic vasodilatory effects^{1, 9}. In spontaneously hypertensive rats (SHR) and Wistar Kyoto (WKY) rats, Ang-(1–7) regulates CO and TPR without much change in blood pressure¹⁰. In normal pregnant rats, Ang-(1–7) significantly dilates mesenteric arteries, an effect blocked by the Ang-(1–7) antagonist [D-Ala7]-Ang-(1–7)⁹. To elucidate the contributions of Ang-(1–7) in vascular responses during pregnancy, we compared the influence of Ang-(1–7) on systemic hemodynamics and peripheral regional circulation (brachial artery) in normotensive pregnant and non-pregnant women. We also studied hypertensive women with preeclampsia, but only for assessments of skin (finger) microcirculation.

Methods

The protocol for these experiments was reviewed and approved by the Wake Forest School of Medicine and Zablocki VA Medical Center Institutional Review Boards. Consent forms were signed by all participants at the time of their enrollment in the study. All pregnant subjects were in their third trimester of pregnancy, and free of any known pre-existing cardiovascular, endocrine, or connective tissue disease.

Study 1: Ang-(1–7) contribution to systemic hemodynamics (blood pressure, cardiac output, heart rate, and total peripheral resistance) and forearm blood flow (FBF) in normotensive pregnant and non-pregnant women:

Group 1 (n=7) was composed of non-pregnant, normotensive women who were not taking oral contraceptives; the stage of the menstrual cycle was not determined. Group 2 (n=13) was composed of nulliparous, normotensive pregnant women. The non-pregnant and normotensive pregnant subjects are described in Table 1. All subjects were given systemic (intravenous) infusions of Ang-(1–7) in supine position with a 15° elevation from the back to the head. Three doses of Ang-(1–7) were administered (1000, 2000, and 3000 pmol/min) with each infusion dose lasting 30 minutes. Arterial pressures were measured using a radial artery catheter and VascuMAP (Carolina Medical, King, NC). The VascuMAP was placed on the arm contralateral to the infusion. CO and TPR were measured by bioimpedance using the BioZ device (CardioDynamics International Corporation, San Diego, CA). The advantage of the BioZ System is its ability to assess accurately the changes in perfusion/flux in response to a drug or maneuver. The procedure has been validated against the thermal dilution and Fick methods in animal studies and in adult critically ill surgical patients^{11, 12}. CO data were expressed as a percent change relative to control (saline infusion). Blood samples were obtained for measurement of Ang-(1–7) before Ang-(1–7) infusion and at the end of each dosage (a total of four samples). FBF was measured by bilateral venous occlusion strain gauge plethysmography. Pediatric cuffs were placed around both wrists and inflated to 200 mmHg during each blood flow measurement. Strain gauges were placed around the forearms 5 cm distal to the olecranon.

Table 1.

Systemic Infusion Study: Patient Population

	Non-Pregnant	Normotensive Pregnant
Number of Subjects	7	13
Race (Percent of Caucasian/African American, American Indian and Hispanic)	85.7/14.3	54/46
Age (years)	24.6±3.1	21.8±0.8
Body Mass Index (kg/m2)	21.9±1.1	29.3±1.4*
Gestation age at the time of study (wks)		37.4±0.3

^*p<0.05 vs. non-pregnant.

Study 2: Ang-(1–7) and forearm (brachial) blood flow in normotensive pregnant and non-pregnant women:

Group 1 (n=7–13) was composed of non-pregnant, normotensive women. Group 2 (n=5–10) was composed of nulliparous, normotensive pregnant women (Table 2) as characterized in the Study 1. All subjects were studied in the supine position.

Table 2.

Forearm Blood Flow (FBF) Study: Patient Population

	Non-Pregnant Low dose	Non-Pregnant High Dose	NormotensivePregnant Low Dose	NormotensivePregnant High Dose
Number of Subjects	13	7	5	10
Race (Percent of Caucasian/African American, American Indian and Hispanic)	77/23	100/0	40/60	50/50
Age (years)	25.8±7.7	20.7±1.9	21.5±5.9	22.6±3.9
Body Mass Index (kg/m2)	22.9±3.7	23.8±3.7	30.1±2.9	33.5±6.4
Gestation age at the time of study (wks)			35.3±1.9	37.2±0.9

^*p<0.05 vs. non-pregnant

FBF was measured by bilateral venous occlusion strain gauge plethysmography as described in the Study 1. Blood pressure cuffs placed around the upper arm were inflated to 40 mmHg and deflated every 15 sec. All drugs were infused via a catheter (27 gauge) inserted into the brachial artery of the non-dominant arm under local anesthesia with 1% lidocaine. Baseline FBF was measured at least 60 minutes after the cannulation. Each subject received Ang II or Ang-(1–7) intra-arterially with measurement of forearm blood flow in the order indicated and with increasing doses. Angiotensin II was given at low (1, 5, 10, 20 pmol/min) and high (5, 25, 50, 100 pmol/min) doses. Ang-(1–7) was given at low (1, 5, 10, 50 pmol/min) and high (10, 50, 100, 500 pmol/min) doses. We used three different doses of Ang-(1–7), previously reported to have vasodilatory effects in non-pregnant women or men^{13, 14}. Each dose was given for 10 minutes with a 30 minute washout period between each agent. The doses of drugs were selected to produce local changes in the infused arm only. Systemic arterial pressure was measured by VascuMAP in the contralateral arm. Since venous occlusion plethysmography provides only a relative assessment of blood flow, the data were expressed as a ratio of measurements recorded in infused - to a non-infused (control) arm to reduce the possibility of small changes in blood pressure or sympathetic nervous system affecting the responses to vasoactive substances¹⁵.

Study 3: The contribution of Ang-(1–7) to skin microcirculation.

Group 1 (n=14) was composed of non-pregnant women and Group 2 (n=15) was composed of normotensive pregnant women according to the same criteria as described in the Study 1. Group 3 (n=15) was composed of nulliparous preeclamptic women, matched to Group 2 by gestational age. Preeclampsia was defined by the onset of pregnancy-induced hypertension (blood pressure ≥ 140/90 mmHg) and proteinuria (≥ 300 mg/24 hours or 2+ or more protein levels on random sample collection) (Table 3).

Table 3.

Skin Microcirculation Study: Patient Population

	Non-Pregnant	NormotensivePregnant	Preeclamptic
Number of Subjects	14	15	15
Race (Percent of Caucasian/African American, American Indian and Hispanic)	63.3/35.7	66. 7/33.3	71.4/28.6
Age (years)	21.5±5.9	20.7±1.9	24.3±1.7
Body Mass Index (kg/m2)	26.3±1.8	28.5±1.6	36.6±1.9*#
SBP (mmHg; baseline)	108.9±3.4	105.4±2.4	142.6±1.9*#
DBP (mmHg; baseline)	65.8±0.4	65.7±1.7	86.2±3.1*#
Proteinuria (mg)		Negligible	507.3±214.4#
Gestation age at the time of study (wks)		30.9±1.2	31.5±0.8
Gestational age at delivery (wks)		36.8±1.2	32.9±0.9#
Birthweight, g		2750±249.4	2031±241.0#
Birth centile		47.7±0.1	39.3±0.1
Baby gender (female-to-male ratio)		10/5	5/10

^*p<0.05 vs. non-pregnant;

^#p<0.05 vs. pregnant

Skin blood flow was determined by laser Doppler and Ang II and Ang-(1–7) were delivered using iontophoresis, a drug delivery method based on the stimulation of charged ions migration across the skin¹⁶. The iontophoresis laser Doppler unit (PeriIont Micropharmacology System, Perimed AB, Stockholm, Sweden) allows for direct measurements of skin blood flow over the iontophoresized area^{17, 18}. Since this method does not provide absolute perfusion values but arbitrary perfusion units that represent blood “flux” (F) changes, the data are shown as percent change determined by the following equation:

$$\mathit{\text{Percentage change in flux}}=\left[\frac{\mathit{\text{F post}}-\mathit{\text{F pre}}}{\mathit{\text{F pre}}-\mathit{\text{biologic zero}}}\right]\times 100$$

All participants were examined in the supine position. Subjects were warmed with a blanket until the finger temperature was 30 ± 2°C. Skin (finger) blood flow responses to each of the following iontophoretically applied agents were studied: a) Ang II (0.5, 1.0 and 2.0 mmol/l), b) Ang-(1–7) (1.0, 5.0 and 20.0 mmol/l), and c) physiological saline (control). Drugs were administered in the order and dose indicated. Iontophoresis (200 μA) was performed for 2 minutes with each agent at each concentration. Skin blood flow was measured at the same site as the iontophoresis, with a 15-minute interval between each drug and dose.

Angiotensin-(1–7) radioimmunoassay:

Blood (14 ml) was taken in a cocktail of inhibitors¹⁹ and plasma was extracted using Sep-Pak columns. Ang-(1–7) was measured by an in-house radioimmunoassay, as previously described¹⁹.

Statistical analysis

Statistical analysis of data was performed using GraphPad Prism IV (San Diego, CA) plotting and statistical software. Comparisons of demographic indices between normotensive pregnant and preeclamptic women in the systemic infusion study were performed using unpaired t-tests. The comparisons of demographic parameters among non-pregnant, normotensive pregnant, and preeclamptic women in the skin flow study were performed using one-way analysis of variance (ANOVA) with Tukey posttests. Hemodynamic responses were evaluated using two-way ANOVA with Bonferroni post hoc tests. Dose-dependent responses in each group were evaluated using one-way ANOVA with repeated measures test for matched observations. All measurements were expressed as the mean ± standard error of the mean (SEM). The criterion for statistical significance was p < 0.05.

Results

Study 1:

In the systemic hemodynamics study, most non-pregnant and pregnant women were Caucasians, but there were more African American or Hispanic women in the normotensive pregnant group compared to non-pregnant women group (Table 1). Systemic infusion of Ang-(1–7) at 2000 pmol/min resulted in a greater change in CO (pregnant: 9.4±6.4 vs. non-pregnant: −3.3±2.1% p<0.05) and less change in TPR (pregnant: −12.9±5.5 vs. non-pregnant: 1.7±3.1%) in pregnant women compared to non-pregnant women (Figure 1). There were no differences between groups in response to the systemic infusion of Ang-(1–7) at 1000 or 3000 pmol/min. The change in FBF was greater in pregnant women after systemic administration of Ang-(1–7) at 2000 pmol/min (Pregnant: 12.7±7.5 vs. Non-pregnant: −13.8±5.1 percent, p<0.01, n=9–10) and at 3000 pmol/min (Pregnant: 21.2±11.6 vs. Non-pregnant: −20.3±6.2 percent, p<0.01, n=9–10) but not at 1000 pmol/min. There were no significant changes in blood pressure, heart rate or stroke volume.

Figure 1.

Responses to systemic Ang-(1–7) infusion in non-pregnant and normotensive pregnant women. Data are mean ± SEM; *p<0.05 vs. non-pregnant women, n=7–13. FBF = forearm blood flow.

Plasma levels of Ang-(1–7) were significantly elevated following the infusion of 1000 pmol/min (69.9±5.6 pmol/ml), 2000 pmol/min (122.5±9.9 pmol/ml), and 3000 pmol/min (127.3±11.8 pmol/ml) of the peptide in comparison to pretreatment state (21±3.6 pmol/ml). Plasma Ang-(1–7) levels were measured in normotensive pregnant women; p<0.05; n=15 in each group. In addition, the plasma levels of Ang-(1–7) were increased in a dose-dependent manner such as the peptide levels detected following 2000 and 3000 pmol/min infusion were significantly higher compared to the levels of the peptide detected following 1000 pmol/min infusion (p<0.05; n=15 in each group).

Study 2:

In the forearm (brachial) circulation study, most non-pregnant women were Caucasians. However, there were more participants of African American, American Indian or Hispanic heritage among the normotensive pregnant women compared to non-pregnant women (Table 2). Local infusion of Ang-(1–7) (Figures 2) had no effect on forearm flow, brachial resistance, or mean blood pressure at any dose. Ang II induced a dose-dependent decrease in FBF at 25, 50 and 100 pmol/min (2.1 fold at 25 pmol/min, 2.5 fold at 50 pmol/min and 2.7 fold at 100 pmol/min; all vs. baseline in non-pregnant women, p<0.05) and an increase in resistance in non-pregnant women at 50 and 100 pmol/min (2.9 fold at 50 pmol/min and 3.1 fold at 100 pmol/min p<0.05 vs. baseline in non-pregnant women). In normotensive pregnant women, FBF was lower at 50 and at 100 pmol/min (1.65 fold at 50 pmol/min and 2.1 fold at 100 pmol/min, all vs. baseline in pregnant women, p<0.05) and resistance was higher at 50 and 100 pmol/min following Ang II administration (2.1 fold at 50 pmol/min and 2 fold at 100 pmol/min; all vs. baseline in pregnant women, p<0.05) (Figure 3). These differences were not observed in response to Ang II infusion at low doses. Mean blood pressure did not change in any group in response to brachial infusion of Ang II, regardless of dose.

Figure 2.

Responses of forearm (brachial) circulation to local infusion of Ang-(1–7) in non-pregnant and normotensive pregnant women. Data are mean ± SEM, n=5–13 (low dose), n=7–10 (high dose).

Figure 3.

Responses of forearm (brachial) circulation to local infusion of Ang II in non-pregnant and normotensive pregnant women. Data are mean ± SEM, n=5–13 (low dose), n=7–10 (high dose); *p<0.05 vs. baseline in non-pregnant women; #p<0.05 vs. baseline in pregnant women; ^p<0.05 vs. 5 pmol/min.

Study 3:

Participants in the skin microcirculation experiment were predominantly Caucasian. As expected, resting systolic and diastolic blood pressures and proteinuria were greater in preeclamptic women versus normotensive pregnant and non-pregnant women (Table 3). Body mass index was greater in preeclamptic subjects compared to normotensive pregnant or to non-pregnant women. On the average, preeclamptic women had shorter pregnancies and smaller babies. There were more boys born to preeclamptic mothers in this cohort of women.

Changes in cutaneous flow were significantly higher in normotensive pregnant women versus preeclamptic women in response to 5 mmol/l of Ang-(1–7) (Pregnant: 182.5±93 vs. Preeclamptic: 15.76±19.46 percent, p<0.05) (Figure 4A), but not in response to Ang-(1–7) at either 1 or 20 mmol/l dose. Preeclamptic women lost the vasodilatory response to 5 mmol/l. Skin flow was reduced in normotensive pregnant women compared to preeclamptic women in response to Ang II (1.0 mmol/l) (Figure 4B). There were no differences in skin flow among studied groups in response to Ang II at either 0.5 or 2.0 mmol/l dose (Figure 4B).

Figure 4.

Responses of skin microcirculation to local infusion of Ang-(1–7) (part A) and Ang II (part B) in non-pregnant, pregnant normotensive, and preeclamptic women. Data are mean ± SEM, n=14–15; *p<0.05 vs. preeclamptic women.

Discussion

In this study, systemic infusion of Ang-(1–7) resulted in greater increase in the change of CO and lower decrease in the change of TPR in normotensive pregnant women compared to non-pregnant women. In contrast, in our forearm circulation experiments, Ang-(1–7) had no effect on the change of FBF or resistance, whereas high doses of Ang II induced a dose-dependent decrease in FBF and increased resistance in normotensive pregnant and non-pregnant women. We also found that the change in skin blood flow in response to Ang-(1–7) (5mmol/l) was greater in normotensive pregnant women and had little or no effect in preeclamptic women. These findings suggest a possible role for Ang-(1–7) in the regulation of systemic and local skin hemodynamics in the third trimester of normotensive pregnancy.

Here, we demonstrate for the first time that systemic infusion of a single dose of Ang-(1–7) in the third trimester of pregnancy increases CO and decreases TPR without altering blood pressure. Changes in stroke volume probably led to the increase in CO in pregnant women. However, our study was limited by the low number of subjects to detect differences in stroke volume between studied groups. Since plasma levels of Ang-(1–7) are increased in the third trimester of normotensive pregnancy¹, our data suggest that Ang-(1–7) could be important in the regulation of systemic hemodynamic changes during pregnancy.

In animal studies, cardiac index is increased in response to continuous systemic infusion of Ang-(1–7)²⁰. Continuous infusion of Ang-(1–7) in rats also reduces TPR and blood pressure in some studies^{20, 21}. We did not have preeclamptic participants in this arm of the study; however, low circulating Ang-(1–7) at late gestation¹ may contribute to augmented peripheral resistance and hypertension in preeclampsia. Moreover, increased FBF in pregnant women in response to systemic Ang-(1–7) may reflect reduced TPR. This FBF response signifies the contribution of systemic Ang-(1–7) to local (forearm) cardiovascular adaptations during normal pregnancy.

Local infusion of Ang-(1–7) was used to eliminate the influence of systemic hemodynamic effects and the possibility of the down-regulation of sympathetic nervous system²². We used three different doses of Ang-(1–7), previously reported to have vasodilatory effects in non-pregnant women or men^{13, 14}. We report, for the first time, the responses of forearm circulation to local infusion of Ang-(1–7) in pregnant women. We found no FBF alterations in response to local infusion of Ang-(1–7). However, FBF was greater in pregnant women compared to non-pregnant women in response to systemic Ang-(1–7) infusion. Differential effects of Ang-(1–7) on FBF could be explained by a higher dose of Ang-(1–7) (2000 and 3000 pg/ml) used for systemic infusion versus the dose used for local infusion (1–500 pg/ml). In addition, higher CO and lower TPR in normotensive pregnant women induced by systemic Ang-(1–7) infusion could facilitate a greater FBF response. Receptor internalization may occur following the administration of increasing doses of Ang-(1–7)^{23, 24}.

Previous reports showed that Ang-(1–7) induces vasodilation in forearm circulation of normotensive and hypertensive subjects¹³. However another study could not confirm the vasodilatory effect of Ang-(1–7) in forearms of normotensive subjects²⁵. Ueda et al also reported that in men, Ang-(1–7) brachial artery infusion attenuated vasoconstriction induced by Ang II and not by noradrenaline, suggesting the specificity of this effect and interference with Ang II-induced contractility responses²⁶. Our data suggest that brachial bed of pregnant women is unresponsive to the influence of Ang-(1–7) at the studied doses. Although we must admit that due to a small sample size the detection of differences in FBF was limited. Nevertheless, some previous reports describe no brachial artery response to flow-mediated dilation in pregnant compared to non-pregnant women^{27, 28}, suggesting that large conduit vessels are not sensitive to vasodilation induced by Ang-(1–7) or other agents.

Although previous studies in normotensive pregnant women found that brachial artery circulation was resistant to pressor effects of Ang II^2–4, our study did not confirm those observations. We had a high number of African American and Hispanic women; therefore, the differences we observed may be related to racial differences. Interestingly, in a previous report, African American men and women had reduced responsiveness of conduit arteries to nitric oxide, suggesting an increased predisposition for vasoconstriction²⁹. Therefore, the vasculature of pregnant African American women may be more responsive to Ang II than that of Caucasians, perhaps leading to a higher incidence of preeclampsia in African American women^{30, 31}.

Skin vasculature of pregnant women had a greater sensitivity to both Ang-(1–7) (at 5 mmol/l) and Ang II compared to preeclamptic women. These results agreed with known physiological roles of these peptides in vasculature. The lack of a response to Ang-(1–7) at the higher dose may reflect that Ang-(1–7) may act on the angiotensin II type 1 receptor at a high pharmacological dose, and the lower response would be explained by its competing effects at both the mas/angiotensin-(1–7) and angiotensin II type 1 receptors. Skin microcirculatory responses in preeclamptic women were similar to those of non-pregnant women, suggesting that preeclamptic women show less sensitivity to vascular vasodilatory stimuli. Preeclampsia is associated with vascular endothelial dysfunction and increased stiffness³². Our data suggest that preeclampsia alters skin microcirculation, possibly leading to endothelial dysfunction and less compliant vasculature. Interestingly, moderate obesity is associated with endothelial dysfunction in non-pregnant women and is linked to systemic inflammation³³. We can only hypothesize that an increased body mass index may be partially responsible for the lack of changes in skin circulation in preeclamptic group. Only a few earlier reports have assessed the cutaneous circulation in pregnancy^34–36. A study by Tur et al³⁴ described alterations of skin vasculature in women with gestational hypertension, and found differences between normotensive pregnant women and women with gestational hypertension. In that study, responses of skin vasculature in non-pregnant women were similar to those of women with gestational hypertension. Although the skin microcirculatory responses may be different in gestational hypertension versus preeclampsia-induced hypertension, study by Tur et al. and our findings suggest that hypertension alters reactivity of conduit arteries and skin vessels, perhaps contributing to the overall increase in peripheral vascular resistance in hypertension. Moreover, other reports have shown that the flow-mediated dilatation of large vessels (e.g. brachial and radial arteries) is reduced in preeclamptic women in their third trimester and in women in their second trimester who later became preeclamptic^{28, 37, 38}, suggesting that preeclampsia is associated with vascular abnormalities of not only skin but also conduit vessels.

In summary, our data imply that Ang-(1–7) is involved in the control of hemodynamic changes in the third trimester of pregnancy. The loss of this regulation in microvessels, as assessed by skin vasculature, may contribute to the development of vascular dysfunction in preeclampsia.