Vascular Contraction and Preeclampsia
Downregulation of the Angiotensin Receptor 1 by Hemopexin In Vitro
During normal pregnancy, in contrast to preeclampsia, plasma hemopexin activity is increased together with a decreased vascular angiotensin II receptor (AT1) expression. We now tested the hypothesis that hemopexin can downregulate the AT1 receptor in vitro. Analysis of human monocytes or endothelial cells by flow cytometry showed decreased membrane density of AT1 exclusively after incubation with active hemopexin, whereas in supernatants of these cell cultures, AT1 molecules could be detected (dot blotting). Also, diminished AT1 was observed in endothelial cell lysates after contact with hemopexin (Western blotting). Hemopexin also induced extracellular signal–regulated kinase 1/2 pathway inhibition in cells after stimulation with angiotensin II in vitro, indicating downregulation of AT1 by hemopexin. In addition, functional loss of AT1 occurred after incubation of rat aortic rings with active hemopexin, as reflected by decreased contraction of the aortic rings on stimulation with angiotensin II. It was further demonstrated that plasma from normal pregnant women decreased the AT1 receptor expression on monocytes as compared with plasma from nonpregnant women or preeclamptic women. Finally, it was shown that plasma hemopexin activity increases during normal gestation from week 10 onward. We concluded that active hemopexin is able to downregulate the AT1 receptor in human monocytes, endothelial cells, and rat aortic rings. We propose that the physiological role of enhanced hemopexin activity during healthy pregnancy is to downregulate the vascular AT1 receptor, promoting an expanded vascular bed. Inhibition of hemopexin activity during preeclampsia may result in an enhanced AT1 receptor expression and a contracted vascular bed.
Preeclampsia (PE) is a major complication of pregnancy affecting ≈6% of pregnant women in the Western world.1 The syndrome is clinically characterized by hypertension, proteinuria, and often edema in the second half of pregnancy. Both severe (early onset) and mild forms of this disorder are recognized, as well as severe variants with liver involvement (hemolysis, elevated liver enzymes, and low platelets syndrome).2 The etiology of PE is unknown, whereas the pathophysiology is poorly understood. The general consensus at this moment is that a systemic inflammatory response associated with endothelial cell dysfunction plays a central role in the pathophysiology of PE. Current thinking focuses on activation of both inflammatory cells and circulating factors, causing endothelial cell dysfunction. For instance, a soluble fms-like kinase 1, an antagonist of vascular endothelial growth factor, may lead to systemic endothelial alterations, including glomerular endotheliosis.3
An important physiological adaptation of normal pregnancy is the increase in cardiac output and blood volume, together with a decrease in peripheral vascular resistance.4,5 These physiological changes are already apparent very early in pregnancy and appear to be maximal between weeks 15 and 25 and remain relatively stable until the end of pregnancy.4,5 The hemodynamic changes during pregnancy are probably caused (at least in part) by changes in the renin-angiotensin-aldosterone system. One of those changes includes decreased vascular responsiveness to angiotensin II (Ang II), starting approximately at week 10.6,7 This drop in responsiveness to Ang II may be attributed to decreased vascular expression of the Ang II receptor (AT1) receptor during normal pregnancy.8–11 In PE, however, maternal hemodynamics are characterized by relative hypovolemia, enhanced total peripheral resistance, and impaired uteroplacental perfusion.6,7 The contracted vascular bed in these patients is associated with the persisting vascular responsiveness on Ang II during the entire gestational period.6,7 Accordingly, an upregulation of vascular AT1 receptors, as compared with normal pregnancy, was shown in preeclamptic women.8–11
Plasma hemopexin (Hx), a heme binding glycoprotein, has been shown recently to also exert protease activity.12–14 Thus, Hx is able to affect extracellular matrix (ECM) molecules of renal tissue and endothelial cells in vitro, which can be inhibited by protease inhibitors or nucleotides like ATP.15 Interestingly, enhanced plasma Hx activity occurs in normal pregnant women as compared with nonpregnant or preeclamptic women.16 Because this enhanced plasma activity occurs together with downregulation of the vascular AT1 in healthy pregnancy,8–11 we proposed that enhanced Hx activity in healthy pregnant women may be responsible for this downregulation. Because of inhibition of Hx activity by enhanced plasma ATP levels observed in PE,16 it seems likely that AT1 downregulation does not occur, resulting in a contracted vascular bed in these patients.
In the present article, it is shown that the AT1 expression can be downregulated by the active isoform of Hx in vitro in various cell types expressing AT1, including human endothelial cells. Moreover, functional nonresponsiveness on Ang II in blood vessel fragments can be induced by incubation of thoracic artery rings of the rat with active Hx; these data support the notion that the active isoform of Hx may (during pregnancy) downregulate AT1, resulting in a decrease of peripheral vascular resistance. The lack of vascular expansion in PE in vivo may be related to the absence of active Hx in these patients.
Materials and Methods
Incubation of Peripheral Blood Mononuclear Cells or Human Endothelial Cells With Hx
Peripheral blood mononuclear cells (PBMCs), isolated from EDTA blood samples obtained from healthy volunteers using standard methods, were washed in Hank’s balanced salt solution and resuspended in tissue culture medium (RPMI 1640; 1×106 cells per mL) and incubated with either active Hx (200 μg/mL) or the same amount of heat-inactivated Hx (Hxi) or washed in saline under standard conditions for 2 hours (37°C; 5% CO2; Hxi was prepared by heating at 80°C for 60 minutes). After the incubation, cells were centrifuged (2000 rpm) for 10 minutes; supernatants were collected and frozen (−20°C) until measurement of the AT1 level using Western blot analysis. The cells were resuspended in PBS and analyzed for AT1 receptor expression using flow cytometry (please see the online data supplement at http://hyper.ahajournals.org). In another set of experiments, cells were incubated with plasma diluted in PBS (1:10) obtained from patients with PE (n=4), healthy pregnant women matched for gestational age (n=5), or age-matched nonpregnant control women (n=5; please see the online data supplement).
Human umbilical venous endothelial cells (HUVECs), passage 3, were obtained from the University Medical Centre Groningen Division of Medical Biology and cultured according to standard methods using 6-well tissue culture plates. Confluent cultures were incubated with either 200 μg/mL of active Hx or the same amount of Hxi or saline as described for PBMCs, harvested using a rubber cell scraper (Corning Inc) after 2 hours, and subsequently washed and prepared for flow cytometry for measurement of AT1 receptor expression (please see the online data supplement).
The present study was performed after approval by the medical ethics committee of the University Medical Center Groningen. Written informed consent was obtained from all of the patients. PE patients and healthy pregnant controls were recruited from the antenatal ward of the University Medical Center Groningen. Blood samples were taken from PE patients and healthy pregnant control women, as well as from nonpregnant women. The nonpregnant women were recruited from hospital staff and students (please see the online data supplement).
Dot Blot Assays for the AT1 Receptor in Supernatants From PBMC Cultures
For dot blot analysis of the AT1 receptor in culture supernatants in PBMCs, BioBlot nitrocellulose membranes (Costar Cambridge Canada) and the BioDot apparatus (BioRad Laboratories) were used. The AT1 receptor was measured in supernatants of cultures from PBMCs stimulated with active Hx, Hxi, or saline according to standard methods (please see the online data supplement).
Western Blotting for the AT1 Receptor in Endothelial Cell Extracts
Confluent human endothelial cell cultures (cell line ECV340) were cultured in medium 199 containing 10% fetal calf serum under standard conditions. After washing with Hanks’ balanced salt solution, cultures were incubated for 120 minutes with Hx (200 μg/mL) under serum-free conditions and subsequently washed. Control incubations were done with Hxi. After washing with PBS, cell lysates were prepared and subjected to 7.5% SDS-PAGE and transferred to a nitrocellulose membranes (BioRad Laboratories) according to standard methods (please see the online data supplement).
AT1 Receptor Downregulation by Hx in Smooth Muscle Cell Lysates as Shown by Inhibition of the Phosphorylation of Mitogen-Activated Protein Kinase Extracellular Signal–Regulated Kinase 1/2
This test system is based on phosphorylation of the mitogen-activated protein kinase extracellular signal–regulated kinase (ERK) 1/2 in cells expressing the AT1 receptor. This phosphorylation occurs on binding of Ang II to the AT1 receptor. The protocol used was carried out with vascular smooth muscle cells (VSMCs) of Wistar rats, which were kindly provided by L.E. Deelman (Department of Clinical Pharmacology, University Medical Centre Groningen), according to standard methods (please see the online data supplement).
Evaluation of Hx Activity in Plasma From Pregnant Women
For measurement of Hx activity during pregnancy, after signing informed consent, blood samples (EDTA, Vacutainer) were taken of 6 pregnant women throughout the entire period of pregnancy. Sampling started at week 5, and samples were taken every 5 weeks until the end of pregnancy. These women were under standard control with midwives, and their pregnancy remained normal throughout the entire gestational period. Exclusion criteria were the presence of known diseases. Samples were measured for Hx activity using the standard ECM stripping assay.12–15 This assay is based on the protease activity of Hx evaluated after incubation of kidney tissue with plasma samples with or without anti-Hx IgG. Decrease of expression of glomerular ectoapyrase, reflected by a loss of reaction product, indicates Hx activity of the sample tested and is calculated and expressed as arbitrary units of Hx activity.14
Aortic-Ring Contraction Studies
The effect of Hx on stimulation of aorta tissue with Ang II was studied using standard isotonic contraction experiments with thoracic aorta rings of the rat pretreated with NG-monomethyl-L-arginine (l-NNMA) (10−4 mol/L for 15 minutes) to inhibit NO synthase according to standard methods.17 Because the vascular response to Ang II is mediated through a variety of receptors, including AT1, Ang II type 2 (AT2), and possibly additional receptors for cleavage products, without l-NNMA the vascular response for Ang II is highly variable because of differences in receptor types in individual animals. To obtain reproducible and specific responses, l-NNMA is used in this standard assay.17 Aortic rings from normal Wistar rats were kept in Krebs solution (at 37°C, 5% CO2, 95% O2), and the cumulative concentration-response curves for Ang II (10−10 to 10−6 mol/L) were evaluated after incubation of the tissue with either active Hx (10−5 mol/L; n=5) or control medium (n=5). Vasoconstriction responses were expressed as a percentage of maximal contraction after stimulation with phenylephrine (10−5 mol/L).
For evaluation of the experiments in which PBMCs or HUVECs were incubated with Hx or plasma, statistical analysis was performed using the Wilcoxon’s signed-rank test, and differences were considered significant if P<0.05. For the experiments in which aortic rings were incubated with Hx, differences between concentration-response curves were evaluated by 1-way ANOVA for repeated measurements followed by the Bonferroni posthoc test (SPSS for Windows Standard version 8.0, SPSS Inc). Differences were considered significant at a level of P<0.05.
Data obtained after flow cytometry for the detection of the expression of AT1 receptor on monocytes or HUVECs are shown in Figure 1. After incubation of PBMCs with active Hx, the mean percentage of AT1 receptor–positive monocytes shows a significant decrease as compared with monocytes incubated with Hxi (Figure 1A and 1B; Because incubation with saline showed the same results as compared with Hxi throughout all of the experiments, data with saline incubation are not shown). In HUVECs, no significant drop in the percentage of positive cells could be observed (Figure 1C). However, when the mean channel brightness of the AT1 receptor–positive cells was measured, reflecting mean AT1 receptor density, a significant reduction of AT1 receptor expression was seen after incubation with active Hx versus Hxi or saline (Figure 1D). In Figure 2, it is shown that the expression of the AT2 receptor has increased after stimulation of cells with active Hx versus Hxi or control medium. With respect to the effect of active Hx on Ang II–mediated contraction of aorta rings in vitro, Hx significantly reduced aortic contraction to cumulative doses of Ang II compared with incubation medium (Figure 3).
Dot blot analysis of supernatants of PBMC cultures after incubation with active Hx shows the presence of increased amounts of AT1 receptor in contrast to control supernatants (ie, after incubation with Hxi or saline; Figure 4A and 4B). In addition, Western blotting of cell lysates prepared from endothelial cell cultures after incubation with Hx showed a decreased amount of the AT1 receptor as compared with incubation with heat-inactivated Hx (Figure 5).
Immunostaining for phosphorylated ERK1/2 using Western blots represented activation of this mitogen-activated protein kinase through stimulation of the AT1 receptor by Ang II. Also this test system showed downregulation of the AT1 receptor by Hx, because it was shown that previous incubation of VSMCs with Hx, in contrast to Hxi, leads to decreased staining of the phosphorylated ERK1/2, as reflected in Figure 6.
The studies with plasma samples from either subjects with PE or healthy pregnant women show decreased AT1 receptor expression in monocytes after incubation with plasma from pregnant individuals as compared with cells incubated with plasma from nonpregnant women (Figure 7). The expression of the AT1 receptor in monocytes after contact with PE plasma also showed a loss of AT1 receptor expression, although to a lesser extent as compared with plasma from normal pregnant women. Figure 8 shows mean activities of plasma Hx in samples from healthy pregnant women during the entire period of pregnancy. It appeared that Hx activity gradually increased during pregnancy, starting around week 10.
The aim of the present study was to test our hypothesis that Hx may be able to downregulate the AT1 receptor in vitro from cells expressing this receptor. Because both PBMCs and endothelial cells showed a loss of expression of the AT1 receptor after contact with active Hx (Figure 1), it is likely that active Hx was indeed able to downregulate this receptor. The data in Figure 4A and 4B clearly show that shedding of the AT1 receptor seems the most likely mechanism. It is clear from Figure 1C and 1D that the outcome of the Hx effect depended on the method of expression of the data. Thus, whereas the mean decrease of the percentage of AT1–positive monocytes reflected the loss of all of the AT1 receptors per cell leading to a decrease of a number of positive cells, the mean decrease of intensity (mean channel brightness) reflected the loss of only a part of the receptors per cell after contact with active Hx. It can be deduced from these data that the downregulation of the AT1 receptors in HUVECs by active Hx was less prominent as compared with this effect in monocytes. The reason for this remains to be investigated. It is conceivable that the nature of AT1 receptor expression in both cell types is not identical, which may explain this discrepancy.
Be this as it may, it is clear that also in endothelial cells a significant reduction of AT1 expression occurred after contact with active Hx. As concomitant with downregulation of AT1, upregulation of the AT2 receptor occurred after stimulation with Hx (Figure 2), and it is clear that exclusively shedding of the AT1 receptor was induced by active Hx. Both Ang II receptor subtypes AT1 and AT2 were able to bind Ang II. However, downregulation of the AT1 receptor often leads to upregulation of the AT2 receptor.18 The mechanism of AT1/ AT2 interaction after binding of Ang II is poorly understood.19 Because Ang II receptor trafficking between cell membrane and cytosol occurs,20 it is conceivable that nongenomic mechanisms are involved in upregulation of AT2 receptors concomitantly with enzymatic removal of AT1 receptors after contact with active Hx. Be this as it may, the present data show an antagonistic effect regarding AT1 versus AT2 expression, which is in line with data from other authors.21
Downregulation of the AT1 receptor seems also reflected by inhibition of the contraction of aorta rings by active Hx after stimulation with Ang II (Figure 3). However, although a significant decrease of aorta ring contraction during contact with active Hx is seen in these experiments, it cannot be concluded from these data whether this is because of either enzymatic shedding or blocking of this receptor by active Hx. In view of the present data with isolated cells, however (Figures 1 and 4⇑), it is suggested that enzymatic cleavage of the AT1 receptors by active Hx may be considered as a mechanism by which this nonresponsiveness on Ang II stimulation is induced. Because the contraction induced by the -adrenergic receptor agonist phenylephrine was not affected by Hx, it seems unlikely that the effect of Hx was caused by a general influence on signal transduction pathways or calcium handling. Therefore, we feel that downregulation of this vascular response on Ang II by active Hx may be specific.
If active Hx is indeed able to promote shedding of the AT1 receptor from cell surfaces, it should be expected that this molecule is detectable in supernatants of cell cultures treated with active Hx. This indeed appears to be the case. The observed increase of the AT1 receptor in supernatants of PBMC cultures after incubation with active Hx, as shown in Figure 4, therefore supports the assumption that active Hx removed the AT1 receptor from the cells studied. This notion is enhanced by the Western blot data of Figure 5, showing reduced AT1 staining in endothelial cell lysates after incubation of endothelial cells with active Hx. Taking advantage of the immediate phosphorylation of the mitogen-activated protein kinase ERK1/2 after stimulation of the AT1 receptor by Ang II, also using this test system, it is clear that downregulation of the AT1 receptor occurred after contact of the cells with active Hx. Because active Hx contains protease activity, able to enzymatically affect various cell membrane constituents and ECM,12–14 it is highly likely that this was mediated by the protease activity of active Hx. However, the precise mechanism of this putative shedding of AT1 receptors from cell surfaces remains to be established. Because the blotting technique is based on immunologic detection of AT1 receptor antigens using a polyclonal antibody, it is possible that both the AT1 receptor and fragments of this molecule are recognized. If active Hx also induces shedding of AT1 in vivo, the question as to the function of soluble AT1 emerges. Up to now, a potential function for shedded AT1 receptors or receptor fragments in vivo is unknown. We considered the possibility that active Hx does not enzymatically hydrolyze AT1 directly but acts through stimulation of other membrane-bound enzymes, eg, adamalysine 17. These so-called sheddases are able to promote shedding of various membrane-associated molecules, eg tumor necrosis factor α.22 Preliminary data, however, showed that inhibition of adamalysine 17, with its inhibitor tumor necrosis factor α protease inhibitor 1, did not affect the downregulation of the AT1 receptor after incubation of the cells with active Hx, suggesting that Hx is able to enzymatically remove the AT1 receptor directly from cell membranes. To exclude the involvement of other sheddases, studies are in progress to test more sheddase inhibitors, including metalloproteinase inhibitors. With respect to the AT2 receptor expression, it is clear that AT2 is significantly upregulated by active Hx, suggesting that the balance between AT1 and AT2 expression may be under the control of active Hx (Figure 2). This is in line with the observation that the AT2 receptor may also be involved in vascular responses in normal and complicated pregnancies.18 However, as stated above, the molecular mechanism underlying AT1/ AT2 interactions is unclear.
Similar downregulation of the AT1 receptor occurs when PBMCs are incubated with plasma from healthy pregnant individuals (Figure 7). Because these plasma samples contain active Hx,16 we ascribe this effect to the presence of Hx in these samples. Incubation of cells with plasma samples from subjects with PE unexpectedly also showed significant downregulation of AT1 receptor as compared with plasma from nonpregnant and pregnant control individuals. This may reflect the presence of Hx activity in these plasma samples. However, using the ECM stripping assay, we demonstrated previously that PE plasma does not show Hx activity. This inhibition is because of the relatively high titers of ATP, which inhibit Hx activity.16 The apparent discrepancy obtained in the current experiments, however, can be explained by assuming that contact of PE plasma with living cells expressing ectophosphatases on their surface promotes hydrolysis of ATP, leading to reactivation of Hx in the PE plasma samples tested. Indeed, experimental evidence is accumulating that contact of PE plasma with living PBMCs may partially activate Hxi in this plasma (results not shown). In other words, it is likely that this reactivation of Hxi to active Hx has occurred during the incubation steps necessary for the flow cytometry assays. In spite of this disturbing reactivation of Hxi in PE plasma samples, it is clear that plasma from pregnant women contains more Hx activity as compared with that of subjects with PE.
Previously we described enhanced Hx activity in plasma from healthy pregnant women, in contrast to women with PE, occurring in the third trimester of pregnancy.16 From Figure 8, it can be seen that the rise of Hx activity starts approximately on week 10. Also around the gestational age of week 10 in healthy pregnant individuals, the nonresponsiveness on Ang II is observed.5,6 This persists throughout the entire period of pregnancy, leading to a relatively expanded vascular bed, beneficial for proper placental perfusion, as compared with age-matched nonpregnant women. Because the nonresponsiveness of vessels on Ang II may be attributable to vascular loss of AT1,7 it is tempting to speculate that the function of enhanced Hx activity in normal pregnancy is to promote shedding of the vascular AT1 receptor, leading to vascular expansion. Interestingly, in healthy pregnant rats, enhanced plasma Hx activity is also associated with decreased AT1 receptor expression (MM Faas et al, unpublished data, 2009). The hypothesis is in line with the situation in PE lacking Hx activity because of enhanced titers of plasma ATP.16 In PE, the AT1 receptor expression is relatively increased8–11 and vascular responsiveness on Ang II is persisting, leading to a contracted vascular bed and resulting in detrimental consequences for placental perfusion. We feel that potential treatment strategies for PE should focus on reactivation of Hx activity, ie, by treatment of PE patients with alkaline phosphatase, or other ATP hydrolyzing principles. This suggestion is corroborated in vivo, in which smoking, per se, is associated with increased plasma alkaline phosphatase activity and decreased incidence of PE.23
Source of Funding
This study was supported by the Dutch Kidney Foundation (grant C05-2123).
- Received December 12, 2008.
- Revision received January 5, 2009.
- Accepted April 3, 2009.
Levine RJ, Ewell MG, Hauth JC, Curet LB, Catalano PM, Morris CD, Choudhary G, Sibai BM. Should the definition of preeclampsia include a rise in diastolic blood pressure of >15 mm Hg to a level of >90 mm Hg in association with proteinuria? Am J Obstet Gynecol. 2000; 187: 787–792.
Cho FN, Chen SN, Kan YY, Lee TC, Wang JS. Successful management of a pregnant woman with HELLP syndrome, pulmonary edema, post-partum hemorrhage and acute renal failure, using early hemodialysis, intravenous immunoglobulin and non-invasive monitoring: a case report. J Reprod Med. 2007; 52: 661–663.
De Swiet M. The cardiovascular system. In: Hytten F, Chamberlain G, eds. Clinical Physiology and Obstetrics. Oxford, United Kingdom: Blackwell Science Ltd; 1998: 33–70.
Hoe SH. The kidney in pregnancy. In: Greenberg A, eds. Primer on Kidney Diseases. 2nd ed. London, United Kingdom: Acad Press San Diego; 1998: 388–394.
Shah DM. Role of renin-angiotensin system in the pathogenesis of pre-eclampsia. Am J Physiol. 2005; 288: F614–F625.
Groenendijk R, Trimbos JB, Wallenburg HC. Hemodynamic measurements in pre-eclampsia: preliminary observations. Am J Obstet Gynecol. 1984; 50: 232–236.
Arenas IA, Xu Y, Davidge ST. Expression of angiotensin II type II receptors in the systemic microvasculature of women with normal and pre-eclamptic pregnancy [abstract]. J Soc Gynecol Investig. 2005; 12 (suppl): 150.
Herse F, Dechend R, Harsem NK, Wallukat G, Janke J, Qadri F, Hering L, Muller N, Luft FC, Staff AC. Dysregulation of the circulation and tissue-based renin-angiotensin system in preecampsia. Hypertension. 2007; 49: 604–611.
Cheung PK, Stulp B, Immenshuh S, Borghuis T, Baller JF, Bakker WW. Is 100KF an isoform of hemopexin? immunochemical characterization of the vascular plasma factor 100KF. J Am Soc Nephrol. 1999; 10: 1700–1708.
Buikema H, Monnink SHJ, Tio RA, Crijns HJGM, De Zeeuw D, Van Gilst WH. Comparison of zofenopril and liginopril to study the role of the sulfhydryl-group in improvement of endothelial dysfunction with ACE inhibitors in experimental heart failure. Br J Pharmacol. 2000; 30: 199–207.
Takeda-Matsubara Y, Iwai M, Cui TX, Shiuchi T, Liu HW, Okumura M, Ito M, Horiuchi M. Role of angiotensin type 1 and 2 receptors in pregnancy associated blood pressure change. Am J Hypertens. 2004; 7: 684–689.
Zheng Y, Saftig P, Hartmann D, Blobel C. Evaluation of the contribution of different ADAMs to tumor necrosis factor alpha (TNF alpha) shedding and of the function of the TNFalpha ectodomain in ensuring selective stimulated shedding by the TNF alpha convertase (TACE/ADAM17). J Biol Chem. 2004; 41: 42898–42906.