- Email: [email protected]
Quantitation of five forms of high molecular weight angiotensinogen from human placenta
AJH
1996; 9:1029-1034
Quantitation of Five Forms of High Molecular Weight Angiotensinogen From Human Placenta Duane A. Tewksbuvy
In addition to the well characterized predominant form of plasma angiotensinogen, which will be termed low molecular weight angiotensinogen (LM,A), significant quantities of a high molecular weight angiotensinogen (HM,A) occur in the human pregnant state. HM,A is the predominant form of angiotensinogen in the placenta and amniotic fluid. The ratio of HM,A/LM,A is elevated in approximately half of the women who develop pregnancy induced hypertension (PIH). This report documents a simple two step methodology for the separation and quantitation of different forms of HM,A. The two steps are gel filtration and ion exchange chromatography. The application of this methodology to amniotic fluid
and plasma from normotensive pregnant women and extracts of amnion, chorion, and placental tissue has shown five distinct forms in the tissue extracts and amniotic fluid, but only three significant forms in plasma. The elution position of all forms of HM,A are highly reproducible and are the same for each tissue or fluid studied, thus lending support to the concept that the three forms seen in plasma of normotensive pregnant women are the same as three of the forms of HM,A seen in extracts of placental tissue. Am J Hypertens 1996;9:1029-1034
n the renin-angiotensin cascade, the first reaction is the cleavage of angiotensinogen by renin to release angiotensin 1 (Ang I). This is the rate limiting step in the renin-angiotensin system, and in plasma both renin and angiotensinogen are rate limiting factors. Angiotensinogen has been isolated and characterized,l,’ its amino acid sequence determined from cDNA studies,3 and its entire gene isolated.4 This well characterized angiotensinogen occurs in two major forms in plasma with apparent molecular weights of 61.4 and 65.4 kDa. The difference can largely be attributed to a difference in glycosylation. These forms of angiotensinogen have been collec-
I
tively designated low molecular weight angiotensinogen (LM,A). In the human pregnant state there are significant quantities of angiotensinogens that exhibit much higher molecular weights. These forms have been collectively called high molecular weight angiotensinogen (HM,A) . It should be noted that, by definition, angiotensinogen is any protein that releases Ang I upon digestion with renin. HM,A has been operationally defined as those forms of angiotensinogen that elute before LM,A on a gel filtration column. Gordon and Sachin’ were the first to note that significant quantities of HM,A were present in gel filtration profiles of plasma. Tewksbury and Dart6 quanti-
Received November 16,1995. Accepted March 19,1996. From the Marshfield Medical Research and Education Foundation, Marshfield, Wisconsin. Presented in part at the Tenth Scientific Meeting of the American Society of Hypertension in New York, May 1995, and was supported
in part by grant 93011310 from the American Heart Association. Address correspondence and reprint requests to Duane A. Tewksbury, PhD, Marshfield Medical Research and Education Foundation, 1000 North Oak Avenue, Marshfield, WI 54449.
0 1996 by the American Journal of Hypertension, Published by Elseaier Science, Inc.
Ltd.
Angiotensinogen, high molecular weight angiotensinogen, low molecular weight angiotensinogen, renin substrate.
KEY WORDS:
0895.7061/96/$15.00 Pll SO895-7062(96)00175-6
1030
TEWKSBURY
AJH-OCTOBER
fied plasma HMA in men, menstruating women, normotensive pregnant women, and women with pregnancy-induced hypertension (PIH) . Low, but detectable, levels of HM,A were found in the nonpregnant state (3% to 5% of the total angiotensinogen). Significant HM,A levels were found in normotensive pregnant women (15% of total angiotensinogen) and significantly elevated levels of HM,A (24% of total angiotensinogen) were found in women with PIH. The original group of women was expanded to include 45 women with PIH and 22 women whose preexisting hypertension was exacerbated during pregnancy.7 In both groups of hypertensive pregnant women, the ratio of HM,A/LM,A was elevated in 47% of the women. Thus, there is a subgroup of hypertensive pregnant women who show significant alterations in their renin-angiotensin system, at least with respect to the angiotensinogen component. Prospective studies have shown that there is no relationship between the development of hypertension during pregnancy and an elevation in plasma HM,A. In some women a significant elevation in plasma HM,A precedes the development of hypertension, and in other women it occurs after the development of hypertension (unpublished data). The concentration of LM,A and HM,A in amniotic fluid from 12 normotensive and 11 hypertensive pregnant women has also been reported.8 HM,A is the predominant form of angiotensinogen in amniotic fluid from normotensive women (62% of total angiotensinogen), whereas the HM,A concentration in amniotic fluid from hypertensive women was slightly, but not significantly, higher (69% of total angiotensinogen) .’ Also, in extracts of amnion, chorion, and placental tissue, HM,A is the predominant form of angiotensinogen (accounting for 60% to 70% of the total angiotensinogen).’ Enzyme kinetic studies with the complete HM,A fraction from plasma of normotensive pregnant women have documented that this preparation of HM,A is a very good substrate for human renin.” The kinetic parameters derived from this study showed that at pH 6.0, given an equimolar solution of HM,A and LM,A, renin would release Ang I from both substrates at equal rates, whereas at pH 7.4, renin would consume LM,A 1.9 times faster than HM,A. That study also confirmed that angiotensinogen is a rate limiting factor in the production of Ang I. Since HM,A appeared to be a significant component of the renin-angiotensin system in the human pregnant state, it was deemed important to isolate and characterize HM,A. A method for the separation and quantitation of multiple forms of HM,A was established, the results of which are presented in this article. METHODS Tissue Preparation centas were perfused
The major vessels of intact plawith saline at 4°C to remove exog-
1996VOL.
9, NO.
20, PART
1
enous blood. The amnion, chorion, and placental tissue were separated, minced, and mixed with three parts (w:v) 0.05 mol! L tris-HCl - 0.15 mol/ L NaCl, 1 mmol/ L leupetin, 1 mmol/ L EDTA, 0.29 mmol/ L PMSF, pH 7.4. This mixture was homogenized in a Brinkman PTlO-35 homogenizer at a setting of 3 four times for 30 set, with an intervening 2-min pause at 4°C. The homogenate was centrifuged at 100,000 g for 5 min at 20°C. The supernatant was used in the following experiments. Fluid Preparation Plasma or amniotic fluid was diluted 1:l (v:v) with 0.05 mol/L tris-HCl, 0.15 mol/L NaCl, pH 7.4 and centrifuged at 100,000 g for 5 min at 20°C. Gel Filtration on Superose 12 The sample, 0.5 mL, was applied to a Pharmacia Superose 12 HR lo/30 column and eluted with 0.05 mol / L tris-HCl, 0.15 mol / L NaCl, pH 7.4 at a flow rate of 0.4 mL/min with 1 min fractions being collected. All the fractions were assayed for angiotensinogen. Fractions containing HM,A, as defined by its elution position, were pooled. Ion Exchange Chromatography The HM,A fraction from the previous step was concentrated and equilibrated with 0.05 mol / L tris-HCl, pH 7.0. Two milliliters containing 7 to 8 mg protein was applied to a MonoQ HR 5 / 5 (Pharmacia ) column equilibrated with 0.05 mol/ L tris-HCl, pH 7.0 and eluted with a combined linear and step gradient at a flow rate of 1 mL/min with 1 min fractions being collected. The elution gradient was as follows: linear from 0 to 0.15 mol/ L NaCl (0 mL to 15 mL), hold at 0.15 mol/L NaCl(15 mL to 20 mL); linear from 0.15 mol/ L NaCl to 0.25 mol/L NaCl (20 mL to 25 mL) , hold at 0.25 mol / L NaCl (25 mL to 30 mL); linear from 0.25 mol! L NaCl to 0.40 mol / L NaCl (30 mL to 35 mL), hold at 0.40 mol / L NaCl (35 mL to 40 mL); linear from 0.40 mol / L NaCl to 0.50 mol/ L NaCl (40 mL to 45 mL), hold at 0.50 mol / L NaCl (45 mL to 50 mL); linear from 0.5 mol/ L NaCl to 0.70 mol/L NaCl (50 mL to 60 mL), and linear from 0.70 mol / L NaCl to 1 mol / L NaCl(60 mL to 65 mL). Angiotensinogen Assay Angiotensinogen was assayed by exhaustive digestion with renin followed by a radioimmunoassay of the release Ang I as previously described.6 Renin was prepared from autopsy kidney tissue by the method of Haas et al,” procedure A. The intraassay coefficient of variation was 3.3% (n = 10) and the interassay coefficient of variation was 6.6% (n = 19). RESULTS In the initial studies on the isolation of HM,A from placental tissue, which was chosen as a starting material because it contained high concentrations of HM,A and was very plentiful, it was found that five forms
AJH-OCTOBER
1996-VOL. 9, NO. IO, PART 1
QUANTITATION
of HM,A were identified. Since HM,A occurs in significant quantities in a number of different tissues and fluids, it was deemed important to quantitate these different forms of HM,A in order to eventually determine their significance. A two step procedure for the separation and quantification of five major forms of HM,A has been devised. The first step was the separation of HM,A from LM,A by means of gel filtration using Superose 12. Sephadex S-200 produced equivalent results. Figure 1 shows a typical separation on Superose 12. The combined HM,A fraction was further fractionated by ion exchange chromatography on an anion exchange column (Mono-Q). Figures 2 and 3 show the resolution of the various fractions of HM,A contained in the three tissue extracts and three fluids analyzed. Peak elution times for all five forms in amniotic fluid and in extracts of amnion, chorion, and placental tissue were the same. Likewise, the peak elution times for the first three forms in plasma from pregnant women and the first two forms in male plasma were identical to the first three and the first two forms in placental tissue, respectively. This lends credence to the postulate that the forms from any of the fluids or tissues studied that elute at the same elution volume are similar in structure. Definitive proof will require isolation of these different forms. It should also be noted that there are five forms in amniotic fluid and extracts of amnion, chorion, and placental tissue, but only significant quantities of three forms in plasma from normotensive pregnant women and only two forms in male plasma. The five forms were numbered in the order of elution from the anion exchange column. Since the various forms of HM,A were quite well resolved, it was possible to quantitate each form from each source. The quantitative data are presented in Table 1. As expected, the greatest variability occurred in the forms present in the lower concentrations. The overall variability is low enough to allow the establishment of some general
04
15
OF
HM,A
1031
6
16 12
C
6 4 0
15
Fraction FIGURE
2.
Chvomafogmm
30
45
60
Number of the ion exchange
chuomafogra-
high molecular zueigkt fraction of angiotensinogen on Mono-Q. (A), amnion; (B), ckorion; and (C), amniotic fluid. The nomenclature used in this article numbers the five forms of HM,A according to their elution position with the first form eltlted beingdesignated I and the lasf form being designated V. phy
of the
patterns. Table 2 presents the percent of each form of HM,A in each type of tissues or fluids. The relative proportions of the different forms were very similar for amniotic fluid and extracts of amnion, chorion, and placental tissue. There were appreciable amounts of all five forms with the predominant form being form III or IV. In contrast, in plasma of normotensive pregnant women, over 50% of the HM,A was form II with appreciable quantities of forms I and III. There were insignificant quantities of forms IV and V. There are two possible explanations for the presence of only three forms in plasma of normotensive pregnant women. Forms IV and V may be cleared faster from the circulation than forms I, II, and III. Alternatively, forms IV and V may not be released into the circulation. The concentration of HM,A in male plasma was very low. Form I tended to be the major form, but a considerable amount of variation was noted. DISCUSSION
I
20
25 Fraction
30
35
Number
FIGURE 1. Ckromatogram of the gel filtration tissue extract on Superose 12.
of a placental
One must question whether these five forms of HM,A observed this study are artifacts and whether indeed HM,A per se is an artifact, as has been suggested.” In response to these concerns it should be noted that
1032
TEWKSBURY
AJHCKTOBER
C
1.0 0.5 0.0
15
30
Fraction
45
60
Number
FIGURE 3. Chvomatogram of the ion exchange ckuomafoguapky of the high molecular weight fraction of angiotensinogen on Mono-Q. (A ),maleplasma; (B ), normotensive pregnant women plasma; and (C), placental tissue.
total HM,A appears to be under physiological control since its plasma concentration rises rapidly from very low to moderate levels throughout the first half of pregnancy, remains constant until parturition, and then returns to nonpregnant levels within 6 to 8 weeks after parturition. Also total HM,A is elevated in a significant proportion of pregnant women with PIH. Only two mild chromatographic procedures were used to quantitate the various forms of HM,A in amniotic fluid and plasma making it unlikely that the five forms were created during the purification procedure.
TABLE
Sample
N
Amnion Chorion Placental tissue Amniotic fluid NPP Male plasma
5 4 6 4 4 4
1. QUANTITATION
Total Amount Applies to Column* 121 110 30 424 380 14.2
OF THE
Form I (Mean 2 SD) 18.5 19.7 4.4 74.0 57.7 10.6
2 ? i i? -c
DIFFERENT
Form II (Mean k SD)
10.4 9.3 1.1 21.7 24.2 2.6
9, NO.
10, PART
1
Extracts of placental tissue and membranes exhibited the same five forms that were found in amniotic fluid. When the inhibitors used in the extraction of placental tissue were omitted, the same yield and relative distribution of the five forms of HM,A were obtained (data not shown). When each of the five major forms of HM,A were individually rechromatographed on Mono Q, all five eluted as a single peak at the original elution volume (data not shown). No evidence was obtained for either the interconversion of one form of HM,A to another form or for the spontaneous dissociation of any form of HM,A observed during extraction or chromatography. Thus, all the available evidence would indicate that the five forms of HM,A represent unique molecular species and are not artifacts. The possibility of other forms being present in either plasma or placental tissue has not been ruled out and can only be addressed by the isolation and characterization of all forms of HM,A. Although this has not been accomplished at this time, much information about the composition of HM,A has been obtained. Tewksbury and DartI determined the molecular weight of a preparation of HM,A from plasma of pregnant women and of purified LM,A by means of gel filtration in 6 M guanidine hydrochloride after reduction and alkylation. The molecular weight of the polypeptide chain containing Ang I from HM,A was the same as that of purified LM,A. Tewksbury and Tryon” showed that immobilized IgG from antisera specific for LM,A effectively bound HM,A from amniotic fluid, plasma of pregnant women, and plasma of men. Western blots of sodium dodecyl sulfate polyacrylamide gel electrophoresis using anti-LM,A sera showed that reduced HM,A from each of the three sources contained a subunit that was identical to LM,A with respect to molecular weight and heterogeneity. These data indicate that HM,A contains at least one subunit that is very similar, if not identical to LM,A. Oxvig et alJ5 have recently provided some addi-
2.5 2.0 1.5
1996-VOL.
24.9 25.2 7.0 70.6 202 3.7
-c + + + + +
10.0 4.7 0.4 19.0 17.7 2.6
FORMS
Form III (Mean ? SD) 36.8 23.8 9.3 86.5 104 0
+ 2 2 ? 2
11.6 7.6 0.6 6.6 16.2 0
OF HM,A Form IV (Mean + SD) 23.6 27.0 5.3 125 12.5 0
t 16.4 -+ 8.8 k 0.7 -t 19.2 2 10.7 0
Form V (Mean ? SD) 17.2 14.2 4.0 67.6 3.7 0
? 5 ? ? ?
15.0 7.2 0.7 24.9 42 0
All values are nanograms/Ang I/milliliter. NPP = normotensive
pregnancy
plasma.
* The actual amount of HM,A applied to the column varied slightly within defermined and the assay values obtained zuere standardized accordingly.
any given group of tissue extracts
or fluids.
Thus a mean amount
applied was
Al&OCTOBER
TABLE
1996VOL.
9, NO.
10, PART
1
QUANTITATION
2. RELATIVE QUANTITIES OF THE DIFFERENT FORMS OF HM,A Percent of Total
HM,A
Sample
Form I
Form II
Form III
Form IV
Form V
Amnion Chorion Placental tissue Amniotic fluid NIT Male plasma
15.3 18.0 14.6 17.5 15.2 74.4
20.6 22.9 23.2 16.7 53.1 25.6
30.4 21.6 31.1 20.4 27.5 0
19.5 24.6 17.7 29.5 3.3 0
14.3 12.9 13.4 16.0 1.0 0
NPP = normotensiue
pregnancy
plasma.
tional information about HM,A. This group has been studying proeosinophil major basic protein (proMBP) in plasma from pregnant women. This 38 kDa protein is known to be of placental origin that is present in plasma complexed by disulfide bonds to other proteins. The major protein involved in this complex was pregnancy-associated plasma protein A (PAPP-A) . However, it was found that the PAPP-A-proMBP complex could not account for all the proMBP in plasma. Oxvig et al l5 partially purified the other complexes, and by means of Western blots, two bands of approximately 200 and 300 kDa were found to contain proMBP. When sequenced, the 200 kDa band contained proMBP and angiotensinogen, while the 300 kDa band contained proMBP, angiotensinogen, and complement C3dg. They suggested that HM,A is a 2:2 complex between angiotensinogen and proMBP and that a fraction of this complex further binds two molecules of complement C3dg in a 2:2:2 complex. Although this report is interesting, these investigators were unable to show that immobilized antisera to proMBP bound these previously unidentified complexes. Our results suggest that neither of these complexes are any of the five forms of HM,A described in this article, because the two complexes with proMBP eluted at relatively high salt ( >0.5 mol/ L NaCl) concentration from a Mono-Q anion exchange column, whereas the three forms of HM,A from plasma of pregnant women all eluted at a relatively low salt concentration (0.15 mol! L, 0.22 mol/ L, and 0.33 mol/ L) . Oxvig et al” did not attempt to measure the renin releasable Ang I in their preparations and thus could not determine what percent of the total HM,A their two complexes constituted. Also in the current study, if the angiotensinogen-proMBP complexes did not release Ang I upon digestion with renin, they would not have been detected. In conclusion, the methodology described in this article separates and quantitates five different forms of HM,A in extracts of chorion, amnion and placental tissue, and in amniotic fluid. Three forms in plasma
OF
HM,A
1033
from pregnant women and two forms in plasma from men have been identified. Significant differences in the relative concentrations of these forms were observed in placental tissue and plasma of normotensive pregnant women. It has not been substantiated that the forms of HM,A in plasma are identical to their respective forms in placental tissue. The fact that the three forms in plasma elute at the same position on ion exchange chromatography as three of the forms in the placental extract would lend credence to the premise that they are identical. Further studies are required to determine which form(s) of HM,A are elevated in PIH and, if there is any relationship between elevation in HM,A and the genotype of angiotensinogen which has been reported to be associated with PIH.i6 ACKNOWLEDGMENTS The author wishes to acknowledge the editorial of Alice Stargardt and the technical assistance Frome, Steve Kaiser, and Kai Qi Zhang.
assistance of Wayne
REFERENCES 1. Tewksbury DA, Dart RA, Travis J: The amino terminal amino acid sequence of human angiotensinogen. Biothem Biophys Res Commun 1981;99:1311-1315. 2.
Tewksbury DA: 4212724-2728.
Angiotensinogen.
Fed
Proc
1983;
3.
Kageyama R, Ohkubo H, Nakanishi S: Primary structure of human preangiotensinogen deduced from the cloned cDNA sequence. Biochemistry 1984;23:36033609.
4.
Gaillard I, Clauser E, Corvol P: Structure angiotensinogen gene. DNA 1989;8:87-99.
5.
Gordon DB, Sachin IN: Chromatographic separation of multiple renin substrates in women: effect of pregnancy and oral contraceptives. Proc Sot Exp Biol Med 1977;156:461-464.
6.
Tewksbury DA, Dart RA: High molecular weight angiotensinogen levels in hypertensive pregnant women. Hypertension 1982;4:729-734.
of human
7. Tewksbury DA, Burrill RE, Tryon ES, et al: Study on high molecular weight angiotensinogen, in Placental and Endometrial Proteins: Basic and Clinical Aspects. VSP, Utrecht, 1988, pp 651-654. 8.
Tewksbury DA, Tryon ES, Burrill RE, et al: High molecular weight angiotensinogen: a pregnancy associated protein. Clin Chim Acta 1986;158:7-12.
9.
Lenz T, Sealey JE, Tewksbury DA: Regional distribution of the angiotensinogens in human placentas. Placenta 1993;14:695-699.
10.
Tryon ES, Tewksbury DA: Kinetic analysis of the reaction of human renin with human high and low molecular weight angiotensinogen. Hypertens Pregn 1995; 14:327-338.
11.
Haas E, Goldblatt H, Gipson EC, Lewis L: Extraction, purification, and assay of human renin free angiotensinase. Circ Res 1966;19:739-749.
12.
Campbell
DJ, Bouhnik
J, Coezy E, et al: Characteriza-
1034
AJH-OCTOBER
TEWKSBURY
tion of precursor and secreted forms of human tensinogen. J Clin Invest 1985;75:1880-1893.
angio-
13.
Tewksbury DA, Dart RA: Molecular weight studies on high molecular weight human angiotensinogen, in Sambhi Ml’ fed): Heterogeneity of Renin and Renin Substrate. New York, Elsevier North-Holland, 1981, pp 261-270.
14.
Tewksbury DA, Tryon son of high molecular
ES: Immunochemical compariweight angiotensinogen from
15.
16.
1996VOL.
9, NO.
10, PART
amniotic fluid, plasma of men and plasma of pregnant women. Am J Hypertens 1989;2:411-413. Oxvig C, Haaning J, Kristensen L, et al: Identification of angiotensinogen and complement C3dg as novel proteins binding the proform of eosinophil major basic protein in human pregnancy serum and plasma. J Biol Chem 1995;270:13645-13651. Ward K, Hata A, Jeunemaitre X, et al: A molecular variant of angiotensinogen associated with preeclampsia. Nature Genet 1993;4:59-61.
1
1996; 9:1029-1034
Quantitation of Five Forms of High Molecular Weight Angiotensinogen From Human Placenta Duane A. Tewksbuvy
In addition to the well characterized predominant form of plasma angiotensinogen, which will be termed low molecular weight angiotensinogen (LM,A), significant quantities of a high molecular weight angiotensinogen (HM,A) occur in the human pregnant state. HM,A is the predominant form of angiotensinogen in the placenta and amniotic fluid. The ratio of HM,A/LM,A is elevated in approximately half of the women who develop pregnancy induced hypertension (PIH). This report documents a simple two step methodology for the separation and quantitation of different forms of HM,A. The two steps are gel filtration and ion exchange chromatography. The application of this methodology to amniotic fluid
and plasma from normotensive pregnant women and extracts of amnion, chorion, and placental tissue has shown five distinct forms in the tissue extracts and amniotic fluid, but only three significant forms in plasma. The elution position of all forms of HM,A are highly reproducible and are the same for each tissue or fluid studied, thus lending support to the concept that the three forms seen in plasma of normotensive pregnant women are the same as three of the forms of HM,A seen in extracts of placental tissue. Am J Hypertens 1996;9:1029-1034
n the renin-angiotensin cascade, the first reaction is the cleavage of angiotensinogen by renin to release angiotensin 1 (Ang I). This is the rate limiting step in the renin-angiotensin system, and in plasma both renin and angiotensinogen are rate limiting factors. Angiotensinogen has been isolated and characterized,l,’ its amino acid sequence determined from cDNA studies,3 and its entire gene isolated.4 This well characterized angiotensinogen occurs in two major forms in plasma with apparent molecular weights of 61.4 and 65.4 kDa. The difference can largely be attributed to a difference in glycosylation. These forms of angiotensinogen have been collec-
I
tively designated low molecular weight angiotensinogen (LM,A). In the human pregnant state there are significant quantities of angiotensinogens that exhibit much higher molecular weights. These forms have been collectively called high molecular weight angiotensinogen (HM,A) . It should be noted that, by definition, angiotensinogen is any protein that releases Ang I upon digestion with renin. HM,A has been operationally defined as those forms of angiotensinogen that elute before LM,A on a gel filtration column. Gordon and Sachin’ were the first to note that significant quantities of HM,A were present in gel filtration profiles of plasma. Tewksbury and Dart6 quanti-
Received November 16,1995. Accepted March 19,1996. From the Marshfield Medical Research and Education Foundation, Marshfield, Wisconsin. Presented in part at the Tenth Scientific Meeting of the American Society of Hypertension in New York, May 1995, and was supported
in part by grant 93011310 from the American Heart Association. Address correspondence and reprint requests to Duane A. Tewksbury, PhD, Marshfield Medical Research and Education Foundation, 1000 North Oak Avenue, Marshfield, WI 54449.
0 1996 by the American Journal of Hypertension, Published by Elseaier Science, Inc.
Ltd.
Angiotensinogen, high molecular weight angiotensinogen, low molecular weight angiotensinogen, renin substrate.
KEY WORDS:
0895.7061/96/$15.00 Pll SO895-7062(96)00175-6
1030
TEWKSBURY
AJH-OCTOBER
fied plasma HMA in men, menstruating women, normotensive pregnant women, and women with pregnancy-induced hypertension (PIH) . Low, but detectable, levels of HM,A were found in the nonpregnant state (3% to 5% of the total angiotensinogen). Significant HM,A levels were found in normotensive pregnant women (15% of total angiotensinogen) and significantly elevated levels of HM,A (24% of total angiotensinogen) were found in women with PIH. The original group of women was expanded to include 45 women with PIH and 22 women whose preexisting hypertension was exacerbated during pregnancy.7 In both groups of hypertensive pregnant women, the ratio of HM,A/LM,A was elevated in 47% of the women. Thus, there is a subgroup of hypertensive pregnant women who show significant alterations in their renin-angiotensin system, at least with respect to the angiotensinogen component. Prospective studies have shown that there is no relationship between the development of hypertension during pregnancy and an elevation in plasma HM,A. In some women a significant elevation in plasma HM,A precedes the development of hypertension, and in other women it occurs after the development of hypertension (unpublished data). The concentration of LM,A and HM,A in amniotic fluid from 12 normotensive and 11 hypertensive pregnant women has also been reported.8 HM,A is the predominant form of angiotensinogen in amniotic fluid from normotensive women (62% of total angiotensinogen), whereas the HM,A concentration in amniotic fluid from hypertensive women was slightly, but not significantly, higher (69% of total angiotensinogen) .’ Also, in extracts of amnion, chorion, and placental tissue, HM,A is the predominant form of angiotensinogen (accounting for 60% to 70% of the total angiotensinogen).’ Enzyme kinetic studies with the complete HM,A fraction from plasma of normotensive pregnant women have documented that this preparation of HM,A is a very good substrate for human renin.” The kinetic parameters derived from this study showed that at pH 6.0, given an equimolar solution of HM,A and LM,A, renin would release Ang I from both substrates at equal rates, whereas at pH 7.4, renin would consume LM,A 1.9 times faster than HM,A. That study also confirmed that angiotensinogen is a rate limiting factor in the production of Ang I. Since HM,A appeared to be a significant component of the renin-angiotensin system in the human pregnant state, it was deemed important to isolate and characterize HM,A. A method for the separation and quantitation of multiple forms of HM,A was established, the results of which are presented in this article. METHODS Tissue Preparation centas were perfused
The major vessels of intact plawith saline at 4°C to remove exog-
1996VOL.
9, NO.
20, PART
1
enous blood. The amnion, chorion, and placental tissue were separated, minced, and mixed with three parts (w:v) 0.05 mol! L tris-HCl - 0.15 mol/ L NaCl, 1 mmol/ L leupetin, 1 mmol/ L EDTA, 0.29 mmol/ L PMSF, pH 7.4. This mixture was homogenized in a Brinkman PTlO-35 homogenizer at a setting of 3 four times for 30 set, with an intervening 2-min pause at 4°C. The homogenate was centrifuged at 100,000 g for 5 min at 20°C. The supernatant was used in the following experiments. Fluid Preparation Plasma or amniotic fluid was diluted 1:l (v:v) with 0.05 mol/L tris-HCl, 0.15 mol/L NaCl, pH 7.4 and centrifuged at 100,000 g for 5 min at 20°C. Gel Filtration on Superose 12 The sample, 0.5 mL, was applied to a Pharmacia Superose 12 HR lo/30 column and eluted with 0.05 mol / L tris-HCl, 0.15 mol / L NaCl, pH 7.4 at a flow rate of 0.4 mL/min with 1 min fractions being collected. All the fractions were assayed for angiotensinogen. Fractions containing HM,A, as defined by its elution position, were pooled. Ion Exchange Chromatography The HM,A fraction from the previous step was concentrated and equilibrated with 0.05 mol / L tris-HCl, pH 7.0. Two milliliters containing 7 to 8 mg protein was applied to a MonoQ HR 5 / 5 (Pharmacia ) column equilibrated with 0.05 mol/ L tris-HCl, pH 7.0 and eluted with a combined linear and step gradient at a flow rate of 1 mL/min with 1 min fractions being collected. The elution gradient was as follows: linear from 0 to 0.15 mol/ L NaCl (0 mL to 15 mL), hold at 0.15 mol/L NaCl(15 mL to 20 mL); linear from 0.15 mol/ L NaCl to 0.25 mol/L NaCl (20 mL to 25 mL) , hold at 0.25 mol / L NaCl (25 mL to 30 mL); linear from 0.25 mol! L NaCl to 0.40 mol / L NaCl (30 mL to 35 mL), hold at 0.40 mol / L NaCl (35 mL to 40 mL); linear from 0.40 mol / L NaCl to 0.50 mol/ L NaCl (40 mL to 45 mL), hold at 0.50 mol / L NaCl (45 mL to 50 mL); linear from 0.5 mol/ L NaCl to 0.70 mol/L NaCl (50 mL to 60 mL), and linear from 0.70 mol / L NaCl to 1 mol / L NaCl(60 mL to 65 mL). Angiotensinogen Assay Angiotensinogen was assayed by exhaustive digestion with renin followed by a radioimmunoassay of the release Ang I as previously described.6 Renin was prepared from autopsy kidney tissue by the method of Haas et al,” procedure A. The intraassay coefficient of variation was 3.3% (n = 10) and the interassay coefficient of variation was 6.6% (n = 19). RESULTS In the initial studies on the isolation of HM,A from placental tissue, which was chosen as a starting material because it contained high concentrations of HM,A and was very plentiful, it was found that five forms
AJH-OCTOBER
1996-VOL. 9, NO. IO, PART 1
QUANTITATION
of HM,A were identified. Since HM,A occurs in significant quantities in a number of different tissues and fluids, it was deemed important to quantitate these different forms of HM,A in order to eventually determine their significance. A two step procedure for the separation and quantification of five major forms of HM,A has been devised. The first step was the separation of HM,A from LM,A by means of gel filtration using Superose 12. Sephadex S-200 produced equivalent results. Figure 1 shows a typical separation on Superose 12. The combined HM,A fraction was further fractionated by ion exchange chromatography on an anion exchange column (Mono-Q). Figures 2 and 3 show the resolution of the various fractions of HM,A contained in the three tissue extracts and three fluids analyzed. Peak elution times for all five forms in amniotic fluid and in extracts of amnion, chorion, and placental tissue were the same. Likewise, the peak elution times for the first three forms in plasma from pregnant women and the first two forms in male plasma were identical to the first three and the first two forms in placental tissue, respectively. This lends credence to the postulate that the forms from any of the fluids or tissues studied that elute at the same elution volume are similar in structure. Definitive proof will require isolation of these different forms. It should also be noted that there are five forms in amniotic fluid and extracts of amnion, chorion, and placental tissue, but only significant quantities of three forms in plasma from normotensive pregnant women and only two forms in male plasma. The five forms were numbered in the order of elution from the anion exchange column. Since the various forms of HM,A were quite well resolved, it was possible to quantitate each form from each source. The quantitative data are presented in Table 1. As expected, the greatest variability occurred in the forms present in the lower concentrations. The overall variability is low enough to allow the establishment of some general
04
15
OF
HM,A
1031
6
16 12
C
6 4 0
15
Fraction FIGURE
2.
Chvomafogmm
30
45
60
Number of the ion exchange
chuomafogra-
high molecular zueigkt fraction of angiotensinogen on Mono-Q. (A), amnion; (B), ckorion; and (C), amniotic fluid. The nomenclature used in this article numbers the five forms of HM,A according to their elution position with the first form eltlted beingdesignated I and the lasf form being designated V. phy
of the
patterns. Table 2 presents the percent of each form of HM,A in each type of tissues or fluids. The relative proportions of the different forms were very similar for amniotic fluid and extracts of amnion, chorion, and placental tissue. There were appreciable amounts of all five forms with the predominant form being form III or IV. In contrast, in plasma of normotensive pregnant women, over 50% of the HM,A was form II with appreciable quantities of forms I and III. There were insignificant quantities of forms IV and V. There are two possible explanations for the presence of only three forms in plasma of normotensive pregnant women. Forms IV and V may be cleared faster from the circulation than forms I, II, and III. Alternatively, forms IV and V may not be released into the circulation. The concentration of HM,A in male plasma was very low. Form I tended to be the major form, but a considerable amount of variation was noted. DISCUSSION
I
20
25 Fraction
30
35
Number
FIGURE 1. Ckromatogram of the gel filtration tissue extract on Superose 12.
of a placental
One must question whether these five forms of HM,A observed this study are artifacts and whether indeed HM,A per se is an artifact, as has been suggested.” In response to these concerns it should be noted that
1032
TEWKSBURY
AJHCKTOBER
C
1.0 0.5 0.0
15
30
Fraction
45
60
Number
FIGURE 3. Chvomatogram of the ion exchange ckuomafoguapky of the high molecular weight fraction of angiotensinogen on Mono-Q. (A ),maleplasma; (B ), normotensive pregnant women plasma; and (C), placental tissue.
total HM,A appears to be under physiological control since its plasma concentration rises rapidly from very low to moderate levels throughout the first half of pregnancy, remains constant until parturition, and then returns to nonpregnant levels within 6 to 8 weeks after parturition. Also total HM,A is elevated in a significant proportion of pregnant women with PIH. Only two mild chromatographic procedures were used to quantitate the various forms of HM,A in amniotic fluid and plasma making it unlikely that the five forms were created during the purification procedure.
TABLE
Sample
N
Amnion Chorion Placental tissue Amniotic fluid NPP Male plasma
5 4 6 4 4 4
1. QUANTITATION
Total Amount Applies to Column* 121 110 30 424 380 14.2
OF THE
Form I (Mean 2 SD) 18.5 19.7 4.4 74.0 57.7 10.6
2 ? i i? -c
DIFFERENT
Form II (Mean k SD)
10.4 9.3 1.1 21.7 24.2 2.6
9, NO.
10, PART
1
Extracts of placental tissue and membranes exhibited the same five forms that were found in amniotic fluid. When the inhibitors used in the extraction of placental tissue were omitted, the same yield and relative distribution of the five forms of HM,A were obtained (data not shown). When each of the five major forms of HM,A were individually rechromatographed on Mono Q, all five eluted as a single peak at the original elution volume (data not shown). No evidence was obtained for either the interconversion of one form of HM,A to another form or for the spontaneous dissociation of any form of HM,A observed during extraction or chromatography. Thus, all the available evidence would indicate that the five forms of HM,A represent unique molecular species and are not artifacts. The possibility of other forms being present in either plasma or placental tissue has not been ruled out and can only be addressed by the isolation and characterization of all forms of HM,A. Although this has not been accomplished at this time, much information about the composition of HM,A has been obtained. Tewksbury and DartI determined the molecular weight of a preparation of HM,A from plasma of pregnant women and of purified LM,A by means of gel filtration in 6 M guanidine hydrochloride after reduction and alkylation. The molecular weight of the polypeptide chain containing Ang I from HM,A was the same as that of purified LM,A. Tewksbury and Tryon” showed that immobilized IgG from antisera specific for LM,A effectively bound HM,A from amniotic fluid, plasma of pregnant women, and plasma of men. Western blots of sodium dodecyl sulfate polyacrylamide gel electrophoresis using anti-LM,A sera showed that reduced HM,A from each of the three sources contained a subunit that was identical to LM,A with respect to molecular weight and heterogeneity. These data indicate that HM,A contains at least one subunit that is very similar, if not identical to LM,A. Oxvig et alJ5 have recently provided some addi-
2.5 2.0 1.5
1996-VOL.
24.9 25.2 7.0 70.6 202 3.7
-c + + + + +
10.0 4.7 0.4 19.0 17.7 2.6
FORMS
Form III (Mean ? SD) 36.8 23.8 9.3 86.5 104 0
+ 2 2 ? 2
11.6 7.6 0.6 6.6 16.2 0
OF HM,A Form IV (Mean + SD) 23.6 27.0 5.3 125 12.5 0
t 16.4 -+ 8.8 k 0.7 -t 19.2 2 10.7 0
Form V (Mean ? SD) 17.2 14.2 4.0 67.6 3.7 0
? 5 ? ? ?
15.0 7.2 0.7 24.9 42 0
All values are nanograms/Ang I/milliliter. NPP = normotensive
pregnancy
plasma.
* The actual amount of HM,A applied to the column varied slightly within defermined and the assay values obtained zuere standardized accordingly.
any given group of tissue extracts
or fluids.
Thus a mean amount
applied was
Al&OCTOBER
TABLE
1996VOL.
9, NO.
10, PART
1
QUANTITATION
2. RELATIVE QUANTITIES OF THE DIFFERENT FORMS OF HM,A Percent of Total
HM,A
Sample
Form I
Form II
Form III
Form IV
Form V
Amnion Chorion Placental tissue Amniotic fluid NIT Male plasma
15.3 18.0 14.6 17.5 15.2 74.4
20.6 22.9 23.2 16.7 53.1 25.6
30.4 21.6 31.1 20.4 27.5 0
19.5 24.6 17.7 29.5 3.3 0
14.3 12.9 13.4 16.0 1.0 0
NPP = normotensiue
pregnancy
plasma.
tional information about HM,A. This group has been studying proeosinophil major basic protein (proMBP) in plasma from pregnant women. This 38 kDa protein is known to be of placental origin that is present in plasma complexed by disulfide bonds to other proteins. The major protein involved in this complex was pregnancy-associated plasma protein A (PAPP-A) . However, it was found that the PAPP-A-proMBP complex could not account for all the proMBP in plasma. Oxvig et al l5 partially purified the other complexes, and by means of Western blots, two bands of approximately 200 and 300 kDa were found to contain proMBP. When sequenced, the 200 kDa band contained proMBP and angiotensinogen, while the 300 kDa band contained proMBP, angiotensinogen, and complement C3dg. They suggested that HM,A is a 2:2 complex between angiotensinogen and proMBP and that a fraction of this complex further binds two molecules of complement C3dg in a 2:2:2 complex. Although this report is interesting, these investigators were unable to show that immobilized antisera to proMBP bound these previously unidentified complexes. Our results suggest that neither of these complexes are any of the five forms of HM,A described in this article, because the two complexes with proMBP eluted at relatively high salt ( >0.5 mol/ L NaCl) concentration from a Mono-Q anion exchange column, whereas the three forms of HM,A from plasma of pregnant women all eluted at a relatively low salt concentration (0.15 mol! L, 0.22 mol/ L, and 0.33 mol/ L) . Oxvig et al” did not attempt to measure the renin releasable Ang I in their preparations and thus could not determine what percent of the total HM,A their two complexes constituted. Also in the current study, if the angiotensinogen-proMBP complexes did not release Ang I upon digestion with renin, they would not have been detected. In conclusion, the methodology described in this article separates and quantitates five different forms of HM,A in extracts of chorion, amnion and placental tissue, and in amniotic fluid. Three forms in plasma
OF
HM,A
1033
from pregnant women and two forms in plasma from men have been identified. Significant differences in the relative concentrations of these forms were observed in placental tissue and plasma of normotensive pregnant women. It has not been substantiated that the forms of HM,A in plasma are identical to their respective forms in placental tissue. The fact that the three forms in plasma elute at the same position on ion exchange chromatography as three of the forms in the placental extract would lend credence to the premise that they are identical. Further studies are required to determine which form(s) of HM,A are elevated in PIH and, if there is any relationship between elevation in HM,A and the genotype of angiotensinogen which has been reported to be associated with PIH.i6 ACKNOWLEDGMENTS The author wishes to acknowledge the editorial of Alice Stargardt and the technical assistance Frome, Steve Kaiser, and Kai Qi Zhang.
assistance of Wayne
REFERENCES 1. Tewksbury DA, Dart RA, Travis J: The amino terminal amino acid sequence of human angiotensinogen. Biothem Biophys Res Commun 1981;99:1311-1315. 2.
Tewksbury DA: 4212724-2728.
Angiotensinogen.
Fed
Proc
1983;
3.
Kageyama R, Ohkubo H, Nakanishi S: Primary structure of human preangiotensinogen deduced from the cloned cDNA sequence. Biochemistry 1984;23:36033609.
4.
Gaillard I, Clauser E, Corvol P: Structure angiotensinogen gene. DNA 1989;8:87-99.
5.
Gordon DB, Sachin IN: Chromatographic separation of multiple renin substrates in women: effect of pregnancy and oral contraceptives. Proc Sot Exp Biol Med 1977;156:461-464.
6.
Tewksbury DA, Dart RA: High molecular weight angiotensinogen levels in hypertensive pregnant women. Hypertension 1982;4:729-734.
of human
7. Tewksbury DA, Burrill RE, Tryon ES, et al: Study on high molecular weight angiotensinogen, in Placental and Endometrial Proteins: Basic and Clinical Aspects. VSP, Utrecht, 1988, pp 651-654. 8.
Tewksbury DA, Tryon ES, Burrill RE, et al: High molecular weight angiotensinogen: a pregnancy associated protein. Clin Chim Acta 1986;158:7-12.
9.
Lenz T, Sealey JE, Tewksbury DA: Regional distribution of the angiotensinogens in human placentas. Placenta 1993;14:695-699.
10.
Tryon ES, Tewksbury DA: Kinetic analysis of the reaction of human renin with human high and low molecular weight angiotensinogen. Hypertens Pregn 1995; 14:327-338.
11.
Haas E, Goldblatt H, Gipson EC, Lewis L: Extraction, purification, and assay of human renin free angiotensinase. Circ Res 1966;19:739-749.
12.
Campbell
DJ, Bouhnik
J, Coezy E, et al: Characteriza-
1034
AJH-OCTOBER
TEWKSBURY
tion of precursor and secreted forms of human tensinogen. J Clin Invest 1985;75:1880-1893.
angio-
13.
Tewksbury DA, Dart RA: Molecular weight studies on high molecular weight human angiotensinogen, in Sambhi Ml’ fed): Heterogeneity of Renin and Renin Substrate. New York, Elsevier North-Holland, 1981, pp 261-270.
14.
Tewksbury DA, Tryon son of high molecular
ES: Immunochemical compariweight angiotensinogen from
15.
16.
1996VOL.
9, NO.
10, PART
amniotic fluid, plasma of men and plasma of pregnant women. Am J Hypertens 1989;2:411-413. Oxvig C, Haaning J, Kristensen L, et al: Identification of angiotensinogen and complement C3dg as novel proteins binding the proform of eosinophil major basic protein in human pregnancy serum and plasma. J Biol Chem 1995;270:13645-13651. Ward K, Hata A, Jeunemaitre X, et al: A molecular variant of angiotensinogen associated with preeclampsia. Nature Genet 1993;4:59-61.
1