0021-972X/91/7303-0489$03.00/0 Journal of Clinical Endocrinology and Metabolism Copyright © 1991 by The Endocrine Society
Vol. 73, No. 3 Printed in U.S.A.
Measurement of Individual Plasma Angiotensins in Normal Pregnancy and Pregnancy-Induced Hypertension* MYRIAM HANSSENS, MARC J. N. C. KEIRSE, BERNARD SPITZ, AND F. ANDRE VAN ASSCHE Department of Obstetrics and Gynecology, University of Leuven (M.H., M.J.N.C.K., B.S., F.A.V.A.), U.Z. Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium; and the Department of Obstetrics, Gynecology, and Reproduction, Leiden University (M.J.N.C.K.), Rijnsburgerweg 10, 2333 AA Leiden, The Netherlands
ABSTRACT. Individual angiotensin peptides were measured by a high pressure liquid chromatography-RIA (HPLC-RIA) technique in the plasma of 20 nonpregnant women, 17 women with normal pregnancy, and 49 with pregnancy-induced hypertension. Immunoreactive angiotensin-II (ANG-II) consisted mainly of ANG-(l-8) octapeptide (>65%), variable proportions (15-25%) of ANG-(4-8) pentapeptide, and small to negligible proportions of ANG-(2-8) heptapeptide and ANG-(3-8) hexapeptide. Levels of ANG-(l-8) were significantly higher in women with normal pregnancy than in both nonpregnant women (P < 0.0006) and women with pregnancy-induced hypertension (P < 0.008); in the latter, levels were lower with increasing severity
of disease. Levels of ANG-(4-8) were higher in women with normal pregnancy than in women with pregnancy-induced hypertension or nonpregnant women. When expressed as a proportion of ANG-(l-8) levels, however, ANG-(4-8) levels were not higher in normal pregnancy than in women with pregnancyinduced hypertension or nonpregnant women. It is concluded that the well known increase in ANG-II levels in normal pregnancy relates predominantly to the active ANG-(l-8) octapeptide and to a far lesser extent to the smaller ANG peptides. Similarly, lower ANG-II levels in pregnancy-induced hypertension relate predominantly to lower ANG-(1-8) levels. (J Clin Endocrinol Metab 73: 489-494,1991)
D
aldosterone release in the adrenal. A disproportional increase in circulating ANG-(2-8) compared with ANG(1-8) levels in pregnancy could, thus, form part of the explanation for some of the physiological changes of normal pregnancy, such as sodium and water retention, decreased blood pressure, and a blunted pressor response to exogenous ANG. Similarly, a higher proportion of ANG-(l-8) vs. ANG-(2-8) and other ANG metabolites could, regardless of changes in total immunoreactive ANG-II concentrations, explain some of the clinical manifestations of pregnancy-induced hypertension or preeclampsia. Variations in the relative proportions of different ANG peptides could also be at the basis of conflicting opinions in the literature on whether ANGII levels are equal (14, 15), higher (16, 17), or lower (18) in pregnancy-induced hypertension than in normal pregnancy. We, therefore, investigated circulating levels of individual immunoreactive ANG peptides during normal pregnancy and in pregnancy-induced hypertension and compared these with values in nonpregnant women. The purpose was to examine whether differences, if any, in ANG-II levels could be due to differences in the proportions of the parent ANG-(l-8) and its metabolites.
URING pregnancy, the renin-angiotensin system is activated, resulting in high circulating levels of angiotensin-II (ANG-II) (1). Immunoreactive measurements of ANG-II represent not only the active ANG octapeptide [referred to as ANG-(1-8)], but also various COOH-terminal ANG peptide fragments (Fig. 1) that result from the action of aminopeptidases on the parent octapeptide (2, 3). One of these, the heptapeptide ANG(2-8) (Des^Asp-ANG-II) may be important, especially during pregnancy, because the uterus is the main source of this peptide (4), and aminopeptidase levels are raised during pregnancy (5, 6). The ANG-(2-8) heptapeptide occupies the same receptors as ANG-(l-8) in the blood vessel walls and adrenals (7-9). Its pressor activity is only 30% that of ANG-(18) octapeptide (10-12), while it is stated to be at least as active (7, 12, 13) as ANG-(l-8) in the regulation of Received October 5,1990. Address all correspondence and requests for reprints to: Dr. M. Hanssens, Department of Obstetrics and Gynecology, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium. * This work was supported by FWGO (Grant 3.0069.85), KULResearch Foundation (OT 85/54), and the Queen Elisabeth Foundation, Brussels.
489
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 18:23 For personal use only. No other uses without permission. . All rights reserved.
HANSSENS ET AL.
490
Subjects and Methods Patients and sample collection Eighty-six Caucasian women were included in the study. Forty-nine of them were admitted to hospital with pregnancyinduced hypertension, 17 were normal pregnant volunteers recruited at random from the antenatal clinic, and 20 were nonpregnant volunteers with regular menstrual cycles recruited among hospital staff members and medical students. All women had an unrestricted salt intake; sodium excretion in a random sample (n = 20) of the hypertensive women ranged between 25-150 mmol/24 h. The clinical characteristics of the subjects in the various groups are summarized in Table 1. Hypertension was defined as a diastolic blood pressure of 90 mm Hg or more on two occasions, and proteinuria was defined as the presence of more than 0.3 g protein excretion in 24 h. Proteinuric pregnancy-induced hypertension was classified as severe if the condition was associated with any of the following: one diastolic blood pressure reading above 120 mm Hg or two readings above 110 mm Hg, proteinuria of more than 3.0 g/24 h, hemolysis, elevated liver enzymes, platelet count below 100,000/mL, or evidence of neurological irritability. Each woman rested for half an hour in the left lateral position before 20 mL blood were withdrawn from an antecubital vein in the right arm and collected in tubes containing 8-hydroxyquinoline-5-sulfonic acid (17.5 mM final dilution) on ice. All samples were taken by the same individual and were further handled as described previously (19). Measurement ofANG peptides The method used has previously been described and evaluated in detail (19). In short, the peptide fraction was extracted from 2 mL plasma by reversible adsorption to phenylsilyl silica cartridges (Bond Elut-pH, Analytichem International, Harbor City, CA). The methanol eluate was filtered through a Millex 0.45-jum filter and vacuum dried in a Buhler Vortex Evaporator at 40 C. The residue, redissolved in 0.1 N acetic acid was then applied to a 250 x 4.6-mm id column packed with 6-^ni particles of trimethylsilane (duPont de Nemours, Hertogenbosch, The Netherlands). Separation of the individual ANG peptides was achieved in an isocratic ion-pairing system using, initially, 0.024% and, later, 0.034% trifluoroacetic acid as the ion-pairing substance and 23% acetonitrile as the organic modifier. The 0.024% trifluoroacetic acid concentration was used before an Asp
Arg
Val
Tyr
I le
His
Pro
Phe
ANG 4-8 ANG 3-8 ANG 2-8 ANG 1-8
FIG. 1. Amino acid sequence of ANG peptides, which are known to cross-react with antiserum raised against the ANG-(1-8) octapeptide.
JCE & M • 1991 Vol 73 • No 3
ANG-(3-8) peptide standard became available to us. Under these conditions ANG-(1-8) and ANG-(3-8) were eluted together. Additional isolation of ANG-(3-8) from ANG-(l-8) was achieved later by the use of 0.034% trifluoroacetic acid. The immunoreactive ANG content was measured in 0.5-mL HPLC eluate fractions. The Medical Research Council of London ANG-(1-8) octapeptide research standard A 70/302 was used as the reference standard in the RIA. The results of fractions corresponding to an ANG peptide peak were summed after subtraction of the background. Corrections were made for the mol wt of the individual peptides and their cross-reactivity in the RIA. Values below the limit of detection were considered at the detection limit for calculation of the mean and SD. When the sum of the values above the detection limit was less than the sum of the levels at the detection limit, the latter was given as an upper limit, and no mean or SD was calculated. Cross-reactivities of the antiserum raised against 5-Ile-ANG(1-8) amide with the various angiotensin peptides were: ANG(1-8), 100%; ANG-(2-8), 165%; ANG-(3-8), 500%; ANG-(48), 130%; and ANG-(1-10), 2%. Cross-reactions with the tetrapeptides ANG-(5-8) and ANG-(l-4) were less than 0.001%. Recoveries of [3H] ANG-(1-8) ranged from 70-75%, and interassay variations were 7.9% at the 55 fmol level and 18% at the 10 fmol level, but up to 35% at the limits of detection of the assay. The detection limits of the assay were 1.9, 0.8, 0.15, and 1.45 fmol for fractions containing, respectively, ANG-(18), ANG-(2-8), ANG-(3-8), and ANG-(4-8) peptide. Combining fractions resulted in a lower limit of detection of 3.8 for ANG-U-8), 1.6 for ANG-(2-8), 0.3 for ANG-(3-8), and 2.9 fmol for ANG-(4-8). The Wilcoxon rank sum and confidence interval (CI) analysis for differences between proportions (20) were used for statistical analysis.
Results In all groups of women, nonpregnant as well as pregnant with or without hypertension, ANG-(1-8) was the main immunoreactive ANG peptide. In 36 women the measurement of ANG-(1-8) may have included ANG(3-8), since complete separation of ANG-(1-8) and ANG-(3-8) was only achieved for the last 50 women included in this study. There were no differences in ANG-(l-8) levels regardless of whether ANG-(3-8) had been excluded from the sample. ANG-(1-8) levels were significantly higher in normal pregnant women than in nonpregnant women (Table 2; P < 0.0006, by Wilcoxon test). Women with pregnancy-induced hypertension, whether superimposed on preexisting hypertension or not, had significantly lower levels than women with normal pregnancy (Table 2; P < 0.008, by Wilcoxon test). Levels in hypertensive women tended to approximate nonpregnant values, a trend that was more pronounced in the severest forms of disease, such as superimposed pregnancy-induced hypertension (Table 2) or severe proteinuric hypertension (Table 3). Of the smaller ANG peptides, ANG-(4-8) was quan-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 18:23 For personal use only. No other uses without permission. . All rights reserved.
INDIVIDUAL ANGIOTENSINS IN PREGNANCY
491
TABLE 1. Clinical characteristics of the women and maximum arterial blood pressures and uric acid levels obtained before ANG measurement in the various study groups No. of women
No. of nulliparae
Age (yr)
Weeks gestation
Maximum systolic pressure
Maximum diastolic pressure
Uric acid (mmol/L)
Pregnancy-induced hypertension without preexistent hypertension Nonproteinuric Mild proteinuric Severe proteinuric
37
24
26.9 (4.0)
33.5 (3.8)
159.8 (16.6)
102.3 (8.3)
0.42 (0.12)
17 6 14
8 6 10
27.3 (3.9) 28.5 (5.4) 25.8 (3.4)
35.4 (2.9) 35.2 (3.0) 30.5 (3.4)
152.9 (11.5) 148.7 (8.8) 172.8 (16.3)
99.9 (6.4) 96.7 (6.0) 107.6 (8.5)
0.36 (0.10) 0.40 (0.10) 0.48 (0.11)
Superimposed pregnancy-induced hypertension
12°
12
32.7 (5.6)
31.7 (3.5)
170.0 (17.6)
110.0 (8.8)
0.43 (0.15)
Normal pregnancy Nonpregnant normotensive women
17 20
7
27.9 (3.6) 28.7 (5.5)
34.0 (2.2)
130.3 (9.9) ND
75.6 (6.8) ND
ND ND
Values are given as a mean with the SD in parentheses. ND, not determined. Seven had proteinuria.
0
TABLE 2. Individual ANG peptides in women with pregnancy-induced hypertension compared with women with normal pregnancy and nonpregnant women No. of women
No. with ANG-(3-8) separation
ANG-U-8)
ANG-(2-8)
ANG-(3-8)
ANG-(4-8)
Pregnancy-induced hypertension Not superimposed Superimposed
49 37 12
22 15 7
10, 12.6 (11.2)° 9, 13.8 (12.4) 1, 8.6 (5.3)
41,
Vol. 73, No. 3 Printed in U.S.A.
Measurement of Individual Plasma Angiotensins in Normal Pregnancy and Pregnancy-Induced Hypertension* MYRIAM HANSSENS, MARC J. N. C. KEIRSE, BERNARD SPITZ, AND F. ANDRE VAN ASSCHE Department of Obstetrics and Gynecology, University of Leuven (M.H., M.J.N.C.K., B.S., F.A.V.A.), U.Z. Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium; and the Department of Obstetrics, Gynecology, and Reproduction, Leiden University (M.J.N.C.K.), Rijnsburgerweg 10, 2333 AA Leiden, The Netherlands
ABSTRACT. Individual angiotensin peptides were measured by a high pressure liquid chromatography-RIA (HPLC-RIA) technique in the plasma of 20 nonpregnant women, 17 women with normal pregnancy, and 49 with pregnancy-induced hypertension. Immunoreactive angiotensin-II (ANG-II) consisted mainly of ANG-(l-8) octapeptide (>65%), variable proportions (15-25%) of ANG-(4-8) pentapeptide, and small to negligible proportions of ANG-(2-8) heptapeptide and ANG-(3-8) hexapeptide. Levels of ANG-(l-8) were significantly higher in women with normal pregnancy than in both nonpregnant women (P < 0.0006) and women with pregnancy-induced hypertension (P < 0.008); in the latter, levels were lower with increasing severity
of disease. Levels of ANG-(4-8) were higher in women with normal pregnancy than in women with pregnancy-induced hypertension or nonpregnant women. When expressed as a proportion of ANG-(l-8) levels, however, ANG-(4-8) levels were not higher in normal pregnancy than in women with pregnancyinduced hypertension or nonpregnant women. It is concluded that the well known increase in ANG-II levels in normal pregnancy relates predominantly to the active ANG-(l-8) octapeptide and to a far lesser extent to the smaller ANG peptides. Similarly, lower ANG-II levels in pregnancy-induced hypertension relate predominantly to lower ANG-(1-8) levels. (J Clin Endocrinol Metab 73: 489-494,1991)
D
aldosterone release in the adrenal. A disproportional increase in circulating ANG-(2-8) compared with ANG(1-8) levels in pregnancy could, thus, form part of the explanation for some of the physiological changes of normal pregnancy, such as sodium and water retention, decreased blood pressure, and a blunted pressor response to exogenous ANG. Similarly, a higher proportion of ANG-(l-8) vs. ANG-(2-8) and other ANG metabolites could, regardless of changes in total immunoreactive ANG-II concentrations, explain some of the clinical manifestations of pregnancy-induced hypertension or preeclampsia. Variations in the relative proportions of different ANG peptides could also be at the basis of conflicting opinions in the literature on whether ANGII levels are equal (14, 15), higher (16, 17), or lower (18) in pregnancy-induced hypertension than in normal pregnancy. We, therefore, investigated circulating levels of individual immunoreactive ANG peptides during normal pregnancy and in pregnancy-induced hypertension and compared these with values in nonpregnant women. The purpose was to examine whether differences, if any, in ANG-II levels could be due to differences in the proportions of the parent ANG-(l-8) and its metabolites.
URING pregnancy, the renin-angiotensin system is activated, resulting in high circulating levels of angiotensin-II (ANG-II) (1). Immunoreactive measurements of ANG-II represent not only the active ANG octapeptide [referred to as ANG-(1-8)], but also various COOH-terminal ANG peptide fragments (Fig. 1) that result from the action of aminopeptidases on the parent octapeptide (2, 3). One of these, the heptapeptide ANG(2-8) (Des^Asp-ANG-II) may be important, especially during pregnancy, because the uterus is the main source of this peptide (4), and aminopeptidase levels are raised during pregnancy (5, 6). The ANG-(2-8) heptapeptide occupies the same receptors as ANG-(l-8) in the blood vessel walls and adrenals (7-9). Its pressor activity is only 30% that of ANG-(18) octapeptide (10-12), while it is stated to be at least as active (7, 12, 13) as ANG-(l-8) in the regulation of Received October 5,1990. Address all correspondence and requests for reprints to: Dr. M. Hanssens, Department of Obstetrics and Gynecology, U.Z. Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium. * This work was supported by FWGO (Grant 3.0069.85), KULResearch Foundation (OT 85/54), and the Queen Elisabeth Foundation, Brussels.
489
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 18:23 For personal use only. No other uses without permission. . All rights reserved.
HANSSENS ET AL.
490
Subjects and Methods Patients and sample collection Eighty-six Caucasian women were included in the study. Forty-nine of them were admitted to hospital with pregnancyinduced hypertension, 17 were normal pregnant volunteers recruited at random from the antenatal clinic, and 20 were nonpregnant volunteers with regular menstrual cycles recruited among hospital staff members and medical students. All women had an unrestricted salt intake; sodium excretion in a random sample (n = 20) of the hypertensive women ranged between 25-150 mmol/24 h. The clinical characteristics of the subjects in the various groups are summarized in Table 1. Hypertension was defined as a diastolic blood pressure of 90 mm Hg or more on two occasions, and proteinuria was defined as the presence of more than 0.3 g protein excretion in 24 h. Proteinuric pregnancy-induced hypertension was classified as severe if the condition was associated with any of the following: one diastolic blood pressure reading above 120 mm Hg or two readings above 110 mm Hg, proteinuria of more than 3.0 g/24 h, hemolysis, elevated liver enzymes, platelet count below 100,000/mL, or evidence of neurological irritability. Each woman rested for half an hour in the left lateral position before 20 mL blood were withdrawn from an antecubital vein in the right arm and collected in tubes containing 8-hydroxyquinoline-5-sulfonic acid (17.5 mM final dilution) on ice. All samples were taken by the same individual and were further handled as described previously (19). Measurement ofANG peptides The method used has previously been described and evaluated in detail (19). In short, the peptide fraction was extracted from 2 mL plasma by reversible adsorption to phenylsilyl silica cartridges (Bond Elut-pH, Analytichem International, Harbor City, CA). The methanol eluate was filtered through a Millex 0.45-jum filter and vacuum dried in a Buhler Vortex Evaporator at 40 C. The residue, redissolved in 0.1 N acetic acid was then applied to a 250 x 4.6-mm id column packed with 6-^ni particles of trimethylsilane (duPont de Nemours, Hertogenbosch, The Netherlands). Separation of the individual ANG peptides was achieved in an isocratic ion-pairing system using, initially, 0.024% and, later, 0.034% trifluoroacetic acid as the ion-pairing substance and 23% acetonitrile as the organic modifier. The 0.024% trifluoroacetic acid concentration was used before an Asp
Arg
Val
Tyr
I le
His
Pro
Phe
ANG 4-8 ANG 3-8 ANG 2-8 ANG 1-8
FIG. 1. Amino acid sequence of ANG peptides, which are known to cross-react with antiserum raised against the ANG-(1-8) octapeptide.
JCE & M • 1991 Vol 73 • No 3
ANG-(3-8) peptide standard became available to us. Under these conditions ANG-(1-8) and ANG-(3-8) were eluted together. Additional isolation of ANG-(3-8) from ANG-(l-8) was achieved later by the use of 0.034% trifluoroacetic acid. The immunoreactive ANG content was measured in 0.5-mL HPLC eluate fractions. The Medical Research Council of London ANG-(1-8) octapeptide research standard A 70/302 was used as the reference standard in the RIA. The results of fractions corresponding to an ANG peptide peak were summed after subtraction of the background. Corrections were made for the mol wt of the individual peptides and their cross-reactivity in the RIA. Values below the limit of detection were considered at the detection limit for calculation of the mean and SD. When the sum of the values above the detection limit was less than the sum of the levels at the detection limit, the latter was given as an upper limit, and no mean or SD was calculated. Cross-reactivities of the antiserum raised against 5-Ile-ANG(1-8) amide with the various angiotensin peptides were: ANG(1-8), 100%; ANG-(2-8), 165%; ANG-(3-8), 500%; ANG-(48), 130%; and ANG-(1-10), 2%. Cross-reactions with the tetrapeptides ANG-(5-8) and ANG-(l-4) were less than 0.001%. Recoveries of [3H] ANG-(1-8) ranged from 70-75%, and interassay variations were 7.9% at the 55 fmol level and 18% at the 10 fmol level, but up to 35% at the limits of detection of the assay. The detection limits of the assay were 1.9, 0.8, 0.15, and 1.45 fmol for fractions containing, respectively, ANG-(18), ANG-(2-8), ANG-(3-8), and ANG-(4-8) peptide. Combining fractions resulted in a lower limit of detection of 3.8 for ANG-U-8), 1.6 for ANG-(2-8), 0.3 for ANG-(3-8), and 2.9 fmol for ANG-(4-8). The Wilcoxon rank sum and confidence interval (CI) analysis for differences between proportions (20) were used for statistical analysis.
Results In all groups of women, nonpregnant as well as pregnant with or without hypertension, ANG-(1-8) was the main immunoreactive ANG peptide. In 36 women the measurement of ANG-(1-8) may have included ANG(3-8), since complete separation of ANG-(1-8) and ANG-(3-8) was only achieved for the last 50 women included in this study. There were no differences in ANG-(l-8) levels regardless of whether ANG-(3-8) had been excluded from the sample. ANG-(1-8) levels were significantly higher in normal pregnant women than in nonpregnant women (Table 2; P < 0.0006, by Wilcoxon test). Women with pregnancy-induced hypertension, whether superimposed on preexisting hypertension or not, had significantly lower levels than women with normal pregnancy (Table 2; P < 0.008, by Wilcoxon test). Levels in hypertensive women tended to approximate nonpregnant values, a trend that was more pronounced in the severest forms of disease, such as superimposed pregnancy-induced hypertension (Table 2) or severe proteinuric hypertension (Table 3). Of the smaller ANG peptides, ANG-(4-8) was quan-
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 16 November 2015. at 18:23 For personal use only. No other uses without permission. . All rights reserved.
INDIVIDUAL ANGIOTENSINS IN PREGNANCY
491
TABLE 1. Clinical characteristics of the women and maximum arterial blood pressures and uric acid levels obtained before ANG measurement in the various study groups No. of women
No. of nulliparae
Age (yr)
Weeks gestation
Maximum systolic pressure
Maximum diastolic pressure
Uric acid (mmol/L)
Pregnancy-induced hypertension without preexistent hypertension Nonproteinuric Mild proteinuric Severe proteinuric
37
24
26.9 (4.0)
33.5 (3.8)
159.8 (16.6)
102.3 (8.3)
0.42 (0.12)
17 6 14
8 6 10
27.3 (3.9) 28.5 (5.4) 25.8 (3.4)
35.4 (2.9) 35.2 (3.0) 30.5 (3.4)
152.9 (11.5) 148.7 (8.8) 172.8 (16.3)
99.9 (6.4) 96.7 (6.0) 107.6 (8.5)
0.36 (0.10) 0.40 (0.10) 0.48 (0.11)
Superimposed pregnancy-induced hypertension
12°
12
32.7 (5.6)
31.7 (3.5)
170.0 (17.6)
110.0 (8.8)
0.43 (0.15)
Normal pregnancy Nonpregnant normotensive women
17 20
7
27.9 (3.6) 28.7 (5.5)
34.0 (2.2)
130.3 (9.9) ND
75.6 (6.8) ND
ND ND
Values are given as a mean with the SD in parentheses. ND, not determined. Seven had proteinuria.
0
TABLE 2. Individual ANG peptides in women with pregnancy-induced hypertension compared with women with normal pregnancy and nonpregnant women No. of women
No. with ANG-(3-8) separation
ANG-U-8)
ANG-(2-8)
ANG-(3-8)
ANG-(4-8)
Pregnancy-induced hypertension Not superimposed Superimposed
49 37 12
22 15 7
10, 12.6 (11.2)° 9, 13.8 (12.4) 1, 8.6 (5.3)
41,
