Pregnancy-Induced High Plasma Levels of Soluble Endoglin in Mice Lead to Preeclampsia Symptoms and Placental Abnormalities

Abstract

Preeclampsia is a pregnancy-specific disease of high prevalence characterized by the onset of hypertension, among other maternal or fetal signs. Its etiopathogenesis remains elusive, but it is widely accepted that abnormal placentation results in the release of soluble factors that cause the clinical manifestations of the disease. An increased level of soluble endoglin (sEng) in plasma has been proposed to be an early diagnostic and prognostic biomarker of this disease. A pathogenic function of sEng involving hypertension has also been reported in several animal models with high levels of plasma sEng not directly dependent on pregnancy. The aim of this work was to study the functional effect of high plasma levels of sEng in the pathophysiology of preeclampsia in a model of pregnant mice, in which the levels of sEng in the maternal blood during pregnancy replicate the conditions of human preeclampsia. Our results show that wild type pregnant mice carrying human sEng-expressing transgenic fetuses (fWT(hsEng⁺)) present high plasma levels of sEng with a timing profile similar to that of human preeclampsia. High plasma levels of human sEng (hsEng) are associated with hypertension, proteinuria, fetal growth restriction, and the release of soluble factors to maternal plasma. In addition, fWT(hsEng⁺) mice also present placental alterations comparable to those caused by the poor remodeling of the spiral arteries characteristic of preeclampsia. In vitro and ex vivo experiments, performed in a human trophoblast cell line and human placental explants, show that sEng interferes with trophoblast invasion and the associated pseudovasculogenesis, a process by which cytotrophoblasts switch from an epithelial to an endothelial phenotype, both events being related to remodeling of the spiral arteries. Our findings provide a novel and useful animal model for future research in preeclampsia and reveal a much more relevant role of sEng in preeclampsia than initially proposed.

Keywords: murine model; placenta; preeclampsia; soluble endoglin.

Figure 1

Description of the animal model. Wild type (WT) female mice crossed with transgenic male mice overexpressing human soluble endoglin (hsEng⁺), named as fWT(hsEng⁺), carry both WT and hsEng⁺ embryos. WT female mice crossed with WT male mice, named as fWT(WT), carry only WT embryos.

Figure 2

Pregnant fWT(hsEng⁺) mice show high plasma levels of soluble endoglin (sEng) and signs of preeclampsia. (a) Plasma levels of hsEng in fWT(hsEng⁺) mice (n = 6) at different time points during gestation; (b) qRT-PCR analysis of hEng mRNA expression in placentas from 5 different fWT(hsEng⁺) mice litters. Each column represents one litter, and each value corresponds to one placenta from that litter. Placentas with higher levels of hEng expressions are inferred to be from hsEng⁺ embryos, whereas placentas with almost no expression are inferred to be from WT embryos; (c) Systolic arterial pressure of fWT(WT) (n = 9) and fWT(hsEng⁺) (n = 12) mice at different time points during gestation; (d) Urinary protein excretion, normalized to urinary creatinine concentration, of fWT(WT) and fWT(hsEng⁺) at Day 13 (fWT(WT) n = 8 and fWT(hsEng⁺) n = 9) and Day 18 (fWT(WT) n = 19 and fWT(hsEng⁺) n = 21) of gestation; (e) Weight of the litters from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 5 litters and fWT(hsEng⁺) n = 6 litters), and Day 18 (fWT(WT) n = 18 litters and fWT(hsEng⁺) n = 22 litters) of gestation; (f) Percentage of aborted fetuses over total fetuses observed on Day 13 or Day 18 of gestation; (g) Plasma levels of murine sEng in fWT(WT) and fWT(hsEng⁺) at Day 13 (fWT(WT) n = 12 and fWT(hsEng⁺) n = 15) and Day 18 (fWT(WT) n = 10 and fWT(hsEng⁺) n = 13) of gestation; (h) sFlt1:PlGF ratio in plasma from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 6 and fWT(hsEng⁺) n = 7) and Day 18 (fWT(WT) n=12 and fWT(hsEng⁺) n = 11) of gestation. Mean ± SEM are displayed. * p < 0.05, ** p < 0.01, and *** p < 0.001 of fWT(hsEng⁺) vs. fWT(WT).

Figure 3

Pregnant fWT(hsEng⁺) mice present placental alterations. (a) Representative image of the analysis of placental and fetal perfusion by laser Doppler imaging. For each scan, the algorithm builds up a color-coded image representing tissue perfusion (left) and records a color photograph at the same time (right). Perfusion was determined in two different areas defined by color photograph, i.e., uterine area and fetal area. As shown in the scale, the pseudocolor ranges from left to right, therefore, blue represents the lowest perfusion and red the highest perfusion; (b) Quantification of the uterine-area perfusion from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 7 and fWT(hsEng⁺) n = 7) and Day 18 ((fWT(WT) n = 11 and fWT(hsEng⁺) n = 15) of gestation; (c) Quantification of the fetal-area perfusion from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 7 and fWT(hsEng⁺) n = 7) and Day 18 (fWT(WT) n = 11 and fWT(hsEng⁺) n = 15) of gestation; (d) Weight of the placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 7 and fWT(hsEng⁺) n = 7) and Day 18 (fWT(WT) n = 9 and fWT(hsEng⁺) n = 8] of gestation; (e) Lipid peroxidation, measured as presence of thiobarbituric acid-reactive substances (TBARS), of placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 20 and fWT(hsEng⁺) n = 21) and Day 18 (fWT(WT) n = 13 and fWT(hsEng⁺) n = 12) of gestation; (f) qRT-PCR analysis of IL-1β expression in the placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 12 and fWT(hsEng⁺) n = 13) and Day 18 (fWT(WT) n = 12 and fWT(hsEng⁺) n = 15) of gestation; (g) qRT-PCR analysis of IL-6 expression in the placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 12 and fWT(hsEng⁺) n = 13) and Day 18 (fWT(WT) n = 12 and fWT(hsEng⁺) n = 15) of gestation. Mean ± SEM are displayed. ** p < 0.01 and *** p < 0.001 of fWT(hsEng⁺) vs. fWT(WT).

Figure 4

Histological alterations of placentas from fWT(hsEng⁺) mice. (a–d) Hematoxylin-eosin (H&E) staining of placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (a,b) and Day 18 (c,d) showing the presence of clusters of glycogen-containing trophoblast (GLYT)-like cells in the basal area of some placentas (arrows); (e) Quantification of the area occupied by GLYT-like clusters in placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 13 (fWT(WT) n = 20 and fWT(hsEng⁺) n = 17) and Day 18 (fWT(WT) n = 13 and fWT(hsEng⁺) n = 16) of gestation; (f) Glycogen content in placentas from fWT(WT) (n = 13) and fWT(hsEng⁺) (n = 15) mice at Day 18 of gestation; (g,h) Periodic acid–Schiff (PAS) staining of placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 18 colocalizing high levels of polysaccharides with GLYT-like clusters; (i,j) Immunohistochemical staining for HIF-1α of placentas from fWT(WT) and fWT(hsEng⁺) mice at Day 18 showing nuclear HIF-1α expression in GLYTs from fWT(hsEng⁺). Mean ± SEM are displayed. ** p < 0.01 of fWT(hsEng⁺) vs. fWT(WT).

Figure 5

Soluble endoglin modifies cytotrophoblast cell biology. (a,b) Effect of rhsEng on JAr cell proliferation measured by MTT assay after 5 days in culture (n = 5) (a) or BrdU incorporation during 2.5 h (b) (n = 5); (c) Effect of rhsEng on trophoblast cell necrosis evaluated measuring extracellular levels of lactate dehydrogenase (LDH) of JAr cells treated or not with rhsEng (n = 4); (d) Caspase activity in cell lysates of JAr cells treated or not with rhsEng (n = 7); (e) qRT-PCR analysis of PECAM1, VE-cadherin, and E-cadherin expression in human placental extracts treated or not with rhsEng (n = 6); (f) Quantification of JAr invasiveness through the Matrigel^®-coated transwell in fetal bovine serum (FBS) gradient, with or without rhsEng treatment (n = 6); (g) Quantification of JAr migration through the uncoated transwell in FBS gradient, with or without rhsEng treatment (n = 5). Mean ± SEM are displayed. * p < 0.05 of 100 ng/mL of rhsEng vs. control.

Figure 6

Hypothetical model of sEng role in preeclampsia. The high levels of soluble endoglin (sEng) are directly involved in the appearance of maternal preeclampsia symptoms, such as hypertension or proteinuria (dashed arrowhead on the right), as shown in non-pregnant animal models. But during pregnancy, sEng can also induce placental alterations such as inflammation, oxidative stress, and hypoxia. These alterations seem to be due to a defective pseudovasculogenesis and a diminished proliferative and invasive capacity of trophoblasts, and lead to the appearance of maternal symptoms, as well as the abnormal levels of soluble factors. As sEng is one of these factors whose expression is increased, it could trigger a positive feedback loop that may contribute to aggravate the disease (continuous arrowhead on the left). Created with BioRender.com.