Canrenone Restores Vasorelaxation Impaired by Marinobufagenin in Human Preeclampsia

Abstract

Previous studies implicated cardiotonic steroids, including Na/K-ATPase inhibitor marinobufagenin (MBG), in the pathogenesis of preeclampsia (PE). Recently, we demonstrated that (i) MBG induces fibrosis in rat tissues via a mechanism involving Fli1, a negative regulator of collagen-1 synthesis, and (ii) MBG sensitive Na/K-ATPase inhibition is reversed by mineralocorticoid antagonists. We hypothesized that in human PE elevated MBG level is associated with the development of fibrosis of the umbilical arteries and that this fibrosis can be attenuated by canrenone. Fifteen patients with PE (mean BP = 118 ± 4 mmHg; 34 ± 2 years; 38 ± 0.3 weeks gest. age) and twelve gestational age-matched normal pregnant subjects (mean BP = 92 ± 2 mmHg; 34 ± 1 years; 39 ± 0.2 weeks gest. age) were enrolled in the study. PE was associated with a higher plasma MBG level, with a four-fold decrease in Fli1 level and a three-fold increase in collagen-1 level in the PE umbilical arteries vs. those from the normal subjects (p < 0.01). Isolated rings of umbilical arteries from the subjects with PE exhibited impaired responses to the relaxant effect of sodium nitroprusside vs. control vessels (EC50 = 141 nmol/L vs. EC50 = 0.9 nmol/L; p < 0.001). The effects of PE on Fli1 and collagen-1 were blocked by the in vitro treatment of umbilical arteries by 10 μmol/L canrenone. Similar results were obtained for umbilical arteries pretreated with MBG. These data demonstrate that elevated MBG level is implicated in the development of the fibrosis of umbilical arteries in PE, and that this could be blocked by mineralocorticoid antagonists.

Keywords: Fli1; Na/K-ATPase; TGFβ; cardiotonic steroids; marinobufagenin; preeclampsia; sodium chloride; vascular stiffness.

Figure 1

Arteries in the chorionic villi of the placenta in healthy pregnant subjects (A) and patients with PE (B). Arrows point to the muscular layer of the arteries. Stained with hematoxylin and eosin, scale bars 200 µm. Immunohistochemical staining with anti-MBG antibodies in the placenta of healthy subjects (C) and patients with PE (D); results of quantitative analysis of the optical density of MBG in the cytotrophoblast of the chorionic villi (E). Solid arrows—the reaction in the cells of the chorionic villi, dashed arrows—the reaction in the trophoblast of the chorionic villi. In panel E–data are presented in arbitrary units (a.u.), *—the reliability of the difference from the control (p < 0.05). Scale 200 µm (A,B), 100 µm (C,D).

Figure 2

Plasma (A) and placental (B) levels of MBG, and mRNA levels of Fli1 in placenta (C) in subjects with uncomplicated pregnancy (Ctrl) and from patients with preeclampsia (PE). By two-way Student’s test: * p < 0.01 vs. Ctrl.

Figure 3

Levels of Fli1 (A), collagen-1 (B), TGFß (C) and collagen 4 (D) (Western blot) in the umbilical arteries from subjects with uncomplicated pregnancy (Ctrl) and from patients with preeclampsia (PE). Top panels. Western blotting representative bands; bottom panels, GAPDH for statistical analysis. Each bar represents individual samples pooled and averaged to give means and SEM of 5 measurements. By one-way ANOVA followed by Bonferroni test: * p < 0.01 vs. Ctrl.

Figure 4

(A)—Effect of normotensive pregnancy, control (EC₅₀ = 58 ± 14 nmol/L), PE alone (EC₅₀ = 58 ± 14 nmol/L; p < 0.01 vs. contol) and PE with canrenone (EC₅₀ = 40 ± 7 nmol/L, p < 0.01 in PE) on the responsiveness of umbilical artery rings to sodium nitroprusside-induced vasorelaxation following contractions induced by 100 nmol/l endothelin-1, concentration–response curves were determined 5 times and averaged to give means and SEM for IC₅₀. Effects of the normotensive pregnancy (Ctrl), PE (Nontr) and PE plus canrenone (Can) on (B) levels of Fli1 and (C) collagen-1 in umbilical arteries. Top analysis. Each bar represents individual samples pooled and averaged to give means and SEM of 5 measurements. By one-way ANOVA followed by Bonferroni test: * p < 0.01 versus nontr. # p < 0.01 vs. can.

Figure 5

Levels of Fli1 (A), collagen-1 (B), TGFβ (C) and collagen 4 (D) by. Western blotting in the umbilical arteries from subjects with uncomplicated pregnancy nontreated (Ctrl) and treated by 10& M MBG. Top panels, Western blotting representative bands; bottom panels, GAPDH for statistical analysis. Each bar represents individual samples pooled and averaged to give means and SEM of 5 measurements. By one-way ANOVA followed by Bonferroni test: * p < 0.01 vs. Ctrl.

Figure 6

(A)—Effect of the normotensive pregnancy, control (60 ± 15 nmol/L), normotensive pregnancy in the presence MBG (1.1 ± 30 μmol/L; p < 0.01 vs. contol), and normotensive pregnancy in the presence of MBG and canrenone (EC₅₀ = 7.9 ± 2.8 nmol/L, p < 0.01 vs. effect of MBG) on the responsiveness of umbilical artery rings to sodium nitroprusside-induced vasorelaxation following contractions induced by 100 nmol/L endothelin-1. Concentration–response curves were determined 5 times and averaged to give means and SEM for IC₅₀. Effects of the normotensive pregnancy (Ctrl), PE (Nontr) and PE plus canrenone (Can) on (B) levels of Fli1 and (C) collagen-1 in umbilical arteries. Top panels, Western blotting representative bands; bottom panels, GAPDH for statistical analysis. Each bar represents individual samples pooled and averaged to give means and SEM of 5 measurements. By one-way ANOVA followed by Bonferroni test: * p < 0.01 versus nontr. # p < 0.01 vs. can.