Association of circulating endothelin and noradrenaline with increased calcium-channel binding sites in the placental bed in pre-eclampsia


  • José A. Ferreira-de-Almeida,

    Consultant (Obstetrics and Gynaecology) , Corresponding author
    1. Department of Obstetrics-Gynaecology, Fuculty of Medicine of Oporto, Portugal
    Search for more papers by this author
  • Francesco Amenta,

    Associate Professor (Pharmacology)
    1. Institute of Phurmacology, Camerino University, Italy
    Search for more papers by this author
  • Filomena Cardoso,

    Senior Staff (Obstetrics and Gynaecology)
    1. Department of Obstetrics-Gynaecology, Fuculty of Medicine of Oporto, Portugal
    Search for more papers by this author
  • Jorge J. Polónia

    Consultant and Aggregated Professor (Internal Medicine)
    1. Clinical Pharmacology Unit, Pharmacology and Therapeutics Institute, Fuculty of Medicine of Oporto, Portugal
    Search for more papers by this author

Correspondence: Professor J. A. Ferreira-de-Almeida, Unidade de Farmacologia Clinica da Faculdade de Medicina, Hospital de S. JoBo, 4200 Porto, Portugal.


Objective To evaluate factors contributing to both placental hypoperfusion and maternal vasoconstriction in pre-eclampsia.

Design Single centre, comparative study of calcium-channel density and affinity in the placental bed of pregnant women with normotension and pre-eclampsia.

Setting Teaching hospital.

Participants Twenty-two primigravidae in the third trimester of pregnancy: 10 with pre-eclampsia and 12 normotensive.

Methods Plasma levels of endothelin-1 (by RIA) and noradrenaline (by HPLC-ED) were measured. Both pharmacological characterisation and anatomical localisation of dihydropyridine-sensitive binding sites (using radioligand-binding studies and autorradiographic techniques) were determined with 3H-isradipine in placental bed tissues to determine both the density (Bmax) and the affinity (Kd) of receptor sites.

Results Higher plasma levels of endothelin-1 and noradrenalin were found in women with pre-eclampsia compared with normotensive women. Placental bed tissues bound 3H-isradipine in a saturable, reversible time and temperature-dependent manner with very low Kd values. Study of the 3H-isradipine specificity binding included the use of several dihydropyridine displacers. In the group with pre-eclampsia the Scatchard analysis of the results showed a significant increase (P < 0.001) both in the affinity [Kd = 0.23 nmol (0.04) vs 0.45 nmol (0.03), pre-eclampsia vs norrnotensive] and in the density of calcium-channel binding sites [Brnax = 77.70 frnol/rng (1.30) vs 64.30 fmourng (1.80) tissue, pre-eclampsia vs normotensive]. Autoradiography confirmed that in the placental bed tissue of those with pre-eclampsia there was a much higher silver grain density in the arteries walls, compared with norrnotensive women.

Conclusions In pre-eclampsia there is an increase in the maternal circulation of two strong vasoconstrictor factors (endothelin-1 and noradrenalin) and a sharp increase both in the density and the affinity of calcium-channel binding sites in placental bed central area. The latter may strongly contribute to the perpetuation of the uteroplacental hypoperfusion either by itself or by amplifying the local actions of circulating factors, such as endothelin-1 and noradrenalin.


Pre-eclampsia, a common disorder of pregnancy, is responsible for significant fetal, maternal and neonatal morbidity and mortality1,2. The pathophysiology of pre-eclampsia remains unclear. In contrast to the normotensive pregnancy, the major pathophysiologi-cal feature of the disease consists of a marked increase in peripheral vascular resistance together with an intense vasoconstriction and plasma volume contraction3,4. While vascular reactivity to vasoconstrictor agents is decreased in normotensive pregnancy, the vasospasm in pre-eclampsia is due, at least in part, to an exaggerated vascular responsiveness to noradrenaline5, angiotensin II6 and possibly to an imbalance in the production of endothelial contractile and vasodilating substances7,8. This generalised vasospasm and associated vascular damage with endothelial cell dysfunction may determine a reduced perfusion of fetal-placental unit and of several maternal organs that presumably leads to the end-organ disturbances of pre-eclampsia4,9,10. Compromised uteroplacental perfusion, with ischemia due to placental pathology and vasospasm, is almost certainly an early and leading cause of pre-eclampsia11. Since placental vasospasm and uteroplacental hypoperfusion tends to be progressively augmented in spite of the increase in blood pressure that occurs in pre-eclampsia12 it is possible that the contraction of uteroplacental vessels may result both from increased levels of vasoconstrictor agents and to an intrinsic over-reactivity to those vasopressors. It is well known that the vasoconstrictor effects of several circulating hormones is dependent on receptor mediated activation of calcium influx into vascular smooth muscle cells through calcium channels that are sensitive to blockage by calcium antagonist13–15. Also, it has been shown that in vitro calcium antagonists of the dihydropyridine type effectively relax human uterine, as well as placental vessels16. We investigated whether an increase in both density and sensitivity of dihydropyridine sensitive calcium-binding sites could be found in the uteroplacental vascular bed of women with pre-eclampsia in the presence of high levels of circulating vasoconstrictor hormones that could explain the increased vasoconstrictor response of uteroplacental vessels to the circulating vasoconstrictor hormones.


Twenty-two primigravidae, aged between 22 and 34 years were recruited from the outpatient sector of the hospital department of gynecology and obstetrics over a 12 month period. Ten women had moderate or severe ‘pure’ pre-eclampsia after 32 weeks defined as hypertension (i.e. blood pressure > 140/90 mmHg after the 28th week, having been previously normotensive, with significant proteinuria and generalised edema)17. All women were evaluated before starting any anti-hypertensive treatment for measuring plasma hormones. At the time of delivery mor-pho-functional studies were performed. We excluded all women with an earlier onset (before the 28th week) as well as those with chronic hypertension and superimposed pre-eclampsia or with previously known maternal nephropathy and women with uncomplicated chronic hypertension. Twelve women with normotensive pregnancies in the third trimester without fetal growth deficiency acted as the control group. The study was approved by the local ethical committee and all women gave their informed consent.

Plasma levels of hormones

All subjects were studied after 20 minutes rest in the supine position. Brachial venous samples were collected in chilled tubes containing heparin or 8 mmol/L ethylenediaminetetraacetic acid and aprotinin (500,000 IU/m), centrifuged immediately at 4°C, and stored at −80°C. Plasma noradrenaline and its deaminated metabolite dihydroxyphenylglycol (DOPEG) were measured by high-performance liquid chromatography and electron detection18. Endothelin-1 levels were determined by radioimmunoassay19 after acidification of the plasma with 1 mL of 0.1% thrifluoroacetic acid and 2 N HCl, passage through C18 Sep-Pac columns, and elution with 60% acetonitrile. Endothelin-1 levels in the serum were determined in duplicate by means of the iodine 125-labelled endothelin-1 assay system (Peninsula Laboratories, Belmont, California, USA), Cross reactivity with ET-2 and ET-3 was 7% and with human big endothelin was 17%.

Pharmacological characterisation and anatomical localisation by radioligand techniques of the dihydropyridine sensitive binding sites in the human placental bed

Uterine tissue was collected during caesarean section. The reasons for caesarean section in normotensive pregnant women were mainly poor previous obstetric history, relative infertility for male reasons, premature rupture of fetal membranes and fetal prematurity. All caesarean sections were performed between weeks 36 and 38. There was a good match between normotensive and pre-eclamptic pregnancies (Table 1). At the time of surgery, placental bed samples were obtained under direct vision of the placental site central area (digitally marked by the assistant) using curved scissors in order to obtain almost perfect disks. A disk of about 1.5 cm in diameter and 1 cm in depth-including basal decidua and the underlying myometrium was removed and used for binding studies and autoradiographic experiments. Uterine samples were embedded in a cryoprotectant medium (OCT, Ames, Iowa, USA). All the radioligand experiments were performed by using tissue sections obtained in a cryostatic chamber ultramycrotome. Tissue sections proved to be better than homogenates for determination of binding parameters (Bmax values indicating receptor density, and Kd values indicating receptor afinity) and also allowed anatomical localisation of binding sites using autoradiography20,21. The chosen thickness for tissue sections was 8 μm for pharmacological characterisation of binding parameters and 6 μm for autoradiographic demonstration, respectively.

Table 1.  Demographic and clinical data of pregnant women who are normotensive and who have pre-eclampsia. Values are given as mean (SEM). S:D = systolic:diastolic.
 Normotensive (n= 12)Pre-eclampsia (n= 10)
  1. *P < 0.0001 vs normotensive pregnancy.

Age (yeas)27.6 (1.2)26.9 (0.8)
Maternal weight (kg)72.5(1.6)79.3(2.1)
Gestation (weeks)38.4 (06)36.1 (1.3)
Proteinuria (> 300 mg/24 h)011210/10*
Casual blood pressure (S:D mmHg)106 (2): 61(1)168 (2): 104 (2)*
Heart rate (bpm)88(5)93(4)
24-hour ambulatory blood pressure (SD mmHg)
  24-hour106 (3): 71(1)145 (4): 96 (2)*
  Daytime109 (3): 74(1)147 (4): 98 (2)*
  Nighttime97 (3): 63(1)141 (4): 91 (2)*

Determination of the binding parameters

Slides containing tissue samples were immersed in coplins with a buffer solution (TRIS HCL 170 nM) in the presence of increasing concentrations of the radioligand, in the absence (total binding) and in the presence of a displacer to define the nonspecific binding. After the incubation period slides were rinsed in the above mentioned buffer solution at 4°C in order to remove free radioligand. Tissue sections were then removed from slides through fiberglass filters (GF-B model, Whatman Inc., Clifton, New Jersey, USA) and put in vials containing 5 mL of scintilation liquid (Histagel, Hewlett Packard, Idaho, USA). Tissue radioactivity was measured in a scintilation counter (Beckman, Irvine, California, USA) and expressed in scintilations per minute (spm). Specific binding values were obtained by subtracting nonspecific binding values from total binding, expressed as fmol/mg tissue. Tissue net weight was previously calculated weighing the slides before and immediately after tissue cutting in a cryostatic chamber. Radioligand concentration and the respective displacer, the temperature (experiments were performed at 4°C, 25°C and 37°C), the buffer solution including its concentration and pH have been previously described22. After adequate rinsing and incubation, times for the radioligand were defined. We started the first saturation experiments in order to determine the first values of binding parameters. The dissociation constant was calculated through the Scatchard analysis of the binding results23. By using the radioligand concentration that binds to 50% of receptors sites (Kd), inhibition experiments were performed in the presence of constant concentrations of radioligand and increasing concentrations of several displacers. Inhibition constant values (Ki) were then obtained for each displacer. Initially we calculated displacer concentration values that inhibited 50% of the specific binding of the radioligand (IC50). IC50 values were calculated as described24 from competition experiments in which at least 12 different concentrations were used. Then these IC50 values were converted into Ki values according to the formula:


in which S represents the concentration of the radioligand used in competition experiments and Kd the value of its dissociation constant at steady-state conditions24. Finally a second group of saturation experiments were performed by using increasing concentrations of radioligand in the absence (total binding) and in the presence of constant concentrations of the displacer with the lowest Ki values or the most stereospecific displacer (nonspecific binding). The Scatchard's analysis of these results allowed the determination of the most accurate values of Bmax and Kd, that were also used in the autorradiographic experiments.


Autoradiographic experiments were performed by using a single radioligand and a single displacer concentration. After fast drying under forced air, tissue sections were fixed while being exposed to formaldehyde at 45°C for 45 minutes in order to avoid radioligand removal. Autoradiographic apposition was performed as previously described25,26. After exposure, development in Kodak D19, and fixation in Agefix Agfa were done as previously described26. After rinsing in destilled water and staining with toluidine blue, tissue sections were then observed in a dark and bright-field photomicroscope (Carl Zeiss, Zeiss 11, Oberkochen, Germany).

Pharmacological characterisation of dihydropyridine sensitive binding sites

Placental bed sections were incubated in the presence of increasing concentrations (0.01–2 nM) of3 (±) PN 200,110, a selective radioligand for dihydropyridine sensitive binding sites, in the absence (total binding) or in the presence of an excess of non-radioactive displacers: (+)-PN 200,110, (±)-PN 200,110, (-)-PN 200,110, darodipine (PY 108,068), BAY K 8644, nimodipine, flunarizine and verapamil. The buffer solution was the TRIS HCl 170 nM, with a final pH of 7.6. After an incubation period of 50 minutes, the slides were rinsed (2 × 10 minutes) in the buffer solution, free of drugs at 4°C. These experiments allowed the definition of the most appropriate displacer, (±)-PN 200 110, and the most accurate radioligand concentrations (0.5 nM) for autorradiography.

Autorradiographic studies

After the definition of the most appropriated concentration of the (3H)-(±)-PN 200 110 (0.5 nM), and the best displacer, (±)-PN 200 100 (1 μM), incubation experiments similar to the binding experiments were performed.

Drugs used

3H-(±)-PN 200 110 (specific activity 75 Cdmrnol) (Amersham Radiochemical Centre, Buckinghamshire, UK); (±)-PN 200 110, (±)-PN 200 110, (-)-PN 200 110 and PY 108 068 (Sandoz Pharma AG, Basel, Switzerland); Nimodipine and BAY K 8644 (Bayer AG, Leverkussen, Germany); Flunarizine (Janssen, Beerse, Belgium); Verapamil (Knoll AG, Liestal, Switzerland); all other substances were supplied by Sigma Chemical Co, St Louis, Missouri, USA.

Statistical analysis

Results are expressed as mean (standard error of the mean - SEM). The density of binding sites (Bmax values) and the dissociation constant values (Kd) were calculated through Scatchard's analysis of the binding results. The inhibition constant values (Ki) were determined as previously described24. Statistical comparisons of differences between normotensive and pre-eclamptic women were made using the student t test for unpaired observations. Analysis of data from ambulatory blood pressure monitoring to compare groups was made after repeated analysis of variance (ANOVA) using software developed in our department27. All data are expressed as mean values +SEM. Changes were considered to be statistically significant at the P < 0.05 level.


Demographic characteristics and plasma hormones

The clinical characteristics of the studied populations are shown in Table 1. There was no significant difference between the populations with regard to maternal age, weight and pregnancy duration. Women with pre-eclampsia presented higher blood pressure values both ‘in ofice’ and using 24-h ambulatory monitoring. Figure 1 shows that, in compare son with normotensive subjects, pre-eclamptic women had significantly higher plasma levels of noradrenaline, DOPEG and of endothelin-1.

Figure 1.

Plasma levels of the homones (supine) in normotensive pregnant women (n= 12) and in women with pre-eclampsia (n= 10). Values are mean (SEM) *P < 0.05 vs normotensive pregnancy.

Saturation curves of 3H-(±)-PN 200110 in the placental bed

The human placental bed binds the radioligand in a saturable, reversible, time- and temperature-dependent manner (Figs 2 and 3) with very low Kd values. The specificity (Ki±SEM) of 3H-(±)-PN 200 110 binding to sections of human placental bed was accomplished by the selective inhibition of the radioligand binding by both nonradioactive dihydropyridine calcium antagonists PY 108 068 (Ki values of 0.61±0.03, normotensives; 0.58±0.02, pre-eclamptics) and nimodipine (Ki values of 0.95±0.02, normotensives; 0.88±0.03, pre-eclamptics) and agonists BAY K 8644 (Ki values of 25.1±0.9, normotensives; 24.3±0.8, pre-eclamptics) but not by other nondihy-dropyridine calcium antagonists like verapamil and flunarizine which showed inhibition constants above 1 μM. The affinity of the different displacers demonstrated the selectivity of the radioligand binding to the dihydropiridinic sensitive binding sites. The Scatchard analysis of the results showed that women with pre-eclampsia had lower Kd values (Kd = 0.23±0.04 nM, PE; Kd = 0.45±0–03 nM, nor-motension, P < 0.001) and higher Bmax values (Bmax = 77.7±1.3 fmol/mg protein, PE; Bmax = 64.34143 fmol/mg protein, normotension, P < 0.001) than controls, meaning respectively an increase both in affinity and in density of dihydropiridinic sensitive binding sites in pre-eclampsia.

Autoradiographic localisation of 3H-(±)-PN 200 110 in the placental bed

As shown in Figs 3 and 4, in both normotensives and pre-eclamptic women, a higher silver grain density was found in the uteroplacental artery walls than in the adjacent myometrium. The density observed in both tissues was higher in pre-eclamptic women as compared with normotensives.


Uteroplacental underperfusion with ischemia, a feature of pre-eclampsia11, may cause vasoactive, mythogenic and procoagulant substances to be released into maternal circulation, resulting in the multisystemic manifestations of the syndrome4,9,10. Thus, uteroplacental bed vasospasm may be, not only an extension of the maternal generalised vasoconstriction, but also a contributing factor for perpetuating the disease. Another factor that may contribute to pre-eclamptic vasospasm is the increased levels of circulating vasopressor substances. Previous studies28,29 including ours30 have consistently found that, as opposed to normotensive pregnancy, pre-eclampsia is characterised by increased peripheral vascular resistances and plasma volume contraction along with an increase of sympathetic system activity and of endothelin plasma levels. In the present study a two-fold increase of noradrenaline and of its deaminated metabolite DOPEG levels, and a three fold increase of endothelin levels, were found in the plasma of pre-eclamptic women as compared with normotensive controls. Thus, at least two complementary mechanisms may be involved in the vasospasm in pre-eclampsia: the augmented circulating levels of two potent vasoconstrictors like endothelin and noradrenaline, and the well-known increased vascular sensitivity to these and other vasopressor substances.

Since endothelin also stimulates the release of cat-echolamines to the bloodstream31, this peptide may contribute to the vasospasm in pre-eclampsia, not only through direct effects but also by indirect mechanisms such as sympathetic stimulation. It has been shown4 that in spite of the maternal hypertensive state in pre-eclampsia, uteroplacental hypoperfusion tends to be progressively augmented. Thus, one possible explanation for the persistent reduction of uteroplacental vascular reserve could be that uteroplacental vessels are also hyperreactive to vasopressor circulating agents. The present study was also conducted to investigate in the placental vascular bed the presence of other underlying mechanisms that may facilitate vasospasm and particularly the increased vascular sensitivity to vasopressor agents at that level. We found in pre-eclamptic women a marked increase in circulating levels of two of the most potent endogenous vasoconstrictor agents, noradrenaline and endothelin, and an increased density and affinity of calcium channel dihydropyridine sensitive binding sites at the placental bed as compared with normotensive controls.

It is well known that in several hypertensive situations, vasoconstriction is partially dependent on the increase of vascular endomyocitic calcium concentration33 and it has been reported that vasoconstrictor effects of several agents like endothelin-I14,15, noradrenaline13 and angiotensin II34 may be dependent on the increase of smooth muscle cell calcium influx, mediated at least in part by receptor-dependent calcium channels, which may be antagonised by dihy-dropyridine calcium entry blockers. Indeed, intracellular free calcium is an important second messenger which controls the initiation of several cellular processes and in vascular smooth muscle, intracellular calcium concentration is known to be responsible for the basal tone and tension32. Also, in human platelets, intracellular calcium influx is known to be dependent on dihydropyridine-sensitive calcium channels35. Since platelets in pre-eclamptic subjects have an increased free calcium concentration which has been shown to be well correlated with blood pressure values36, and membrane calcium transport mechanisms have been reported to be rather similar in human platelets as in smooth vascular cells37, it is possible that vasospastic effects of several pressor agents like noradrenaline and endothelin may result from a calcium-channel dependent increase of calcium influx into vascular cells. Thus, a change in sensitivity and / or density of calcium channel receptor sites could participate in the augmentation of the vasoconstrictor effects of circulating pressor agents.

In the intramyometrial segments of uteroplacental arteries we demonstrated and identified the presence of calcium channel dihydropyridine binding sites by using the selective blocker 3H-PN 200 110. Compared with normotensive pregnancies, pre-eclampsia was associated with a 77% reduction of Kd values and a 21% increase of Bmax values, reflecting a significant increase both in the affinity and in the density of calcium receptor sites at the placental bed. Complementary autoradiographic experiments clearly demonstrated that the density of calcium channel dihydropyridine binding sites in the intramyometrial vessel wall was higher in pre-eclampsia as compared with normotensive pregnant women. These results prove that in pre-eclampsia there is a significant increase of calcium channels which is particularly evident at the vascular wall of the fetal-maternal interface.

The finding that pre-eclampsia occurs with an increased density and affinity of calcium channel dihydropyridine sensitive binding sites at the placental bed, suggests that in this vascular territory the vasospasm may be ascribed, at least in part, to an increased calcium influx into the vascular smooth muscle cells, which in turn, may be particularly amplified in pre-eclampsia by the increased levels of both endothelin and noradrenaline. We speculate that this may contribute to the uteroplacental hypo-perfusion which is still the most plausible explanation for both the placental and systemic manifestations of pre-eclampsia. There is good evidence that dihydropyridine calcium blockers38 do not reduce uteroplacental flow or even improve it39 in spite of a significant reduction in maternal blood pressure. That could be atributed to a specific reduction by these drugs of a uteroplacental vasospasm that is mediated by the blockade of calcium entry through their sensitive calcium-channels. It is possible that such an increase in vascular density and sensitivity of these calcium channels may also be present in other maternal vascular beds of pre-eclamptic women besides the uteroplacental territory. Although an overall vascular increase of calcium channels density and sensitivity could fulfill a good explanation for the generalised vasospasm and for the increased systemic vascular sensitivity to vasopressor agents in pre-eclampsia, this may be difficult to demonstrate since other maternal vascular beds are not so easily available as it is placental bed.

It has been reported that not only the infusion of angiotensin II induces an increase in intracellular calcium concentration in platelets of pre-eclamptic women as compared with platelets from normotensive pregnant40, but also that such an increase of calcium influx is proportional to the degree of the blood pressure rise36. It is interesting that human and experimental hypertensive states other than pre-eclampsia have also been associated to an increased density of calcium channels41. Similarly to what we have found in pre-eclampsia, studies using labelled nitrendipine and PN 200 11042 also reported an increased density and affinity of calcium channels in the vascular system of spontaneously hypertensive rats.

In conclusion, we have shown that, compared with normotensive pregnants, pre-eclampsia is associated with a marked increase in the density and in the affinity of calcium channel dihydropyridine sensitive binding sites at the placental bed. This occurs in association with a substantial increase in plasma concentrations of noradrenaline and of endothelin. However, to be conclusively proven, such an association needs to be confirmed in larger number of subjects. Nevertheless, this may represent an underlying mechanism which by promoting calcium influx into vascular cells may contribute directly to the under-perfusion at the uteroplacental bed, therefore perpetuating vasospasm by potentiating at this level the vasoconstriction induced by the increased circulating levels of vasopressor hormones. Although speculative, our results may also explain, in part, the patency of the uteroplacental vessels when calcium antagonists of the group of dihydropyridines are used to treat high blood pressure in pre-eclampsia.


The authors acknowledge the support of PRAXIS XXI-SAU-1302/95