Preeclampsia: animal models for a human cure.

P reeclampsia (PE) affects ∼5% of human pregnancies and is a leading cause of perinatal mortality, preterm birth, and maternal morbidity (1). Through positive effects on vascular tone and glomerular capillary health, vascular endothelial growth factor (VEGF) and placental growth factor (PGF) are necessary for normal pregnancy (2–4). In PE, these proteins are antagonized by excessive placental production of soluble fms-like tyrosine kinase-1 (sFLT1), a splice variant of the VEGF receptor (2, 3), and in later pregnancy, by soluble endoglin (sENG), a soluble form of a transforming growth factor beta (TGFB1) receptor that prevents binding of TGFB1 to membrane-bound endoglin (5). The report by Kumasawa et al. (6) in PNAS describes the use of a mouse model in which placenta-specific expression of human sFLT1 is present from the time of implantation. The model faithfully reproduces many of the human findings of late-onset PE. The investigators also show that the lipid-lowering drug, pravastatin, ameliorates sFLT1-induced PE in these sFLT mice (6). The etiology of PE is likely multifactorial and may have several forms. The disease can be characterized by early involvement of some combination of abnormal placental inflammation and hypoxia/ reoxygenation injury that is mediated by reactive oxygen and nitrogen species (7, 8). The placental beds of preeclamptic women typically exhibit abnormally shallow trophoblast invasion. It remains unclear whether the invasion abnormalities are the cause or effect of inflammation, hypoxic injury, and changes in angiogenic mediators that are explored in the investigation by Kumasawa et al. (6–8). Problematic in the study of PE is its delayed clinical manifestation until after 26– 28 wk gestation (in its severest form) or more commonly, after 34–36 wk (8). Early (10–12 wk gestation) alterations in soluble factors (9) detected in maternal blood can be used to predict those patients who will subsequently develop PE. However, corresponding placental tissues from this stage of gestation are not generally accessible. Study of placental aspects of the disease relies almost exclusively on specimens acquired after delivery, and information gleaned from these placentas is limited to relatively late disease manifestations. Inherent difficulties in the in vivo and in vitro investigations of placental contributions to this human disease require the development of model systems. It is here that Kumasawa et al. (6) have made their most important contribution. Although the architecture of the murine placenta has remarkable similarities to that of the human, there are several important differences that are outlined in Fig. 1 (10, 11). By transducing murine trophectoderm at the blastocyst stage, Kumasawa et al. (6) drive trophoblast-specific expression of sFLT1 and show that a PE-like disease can be specifically promoted in mice by placentally derived sFLT1. The disease caused by such expression seems to best mimic Fig. 1. Mouse (Upper) and human (Lower) placentas share a discoid shape, hemochorial exchange, and analogous cell types (syncytiotrophoblast, lavender purple; cytotrophoblast/giant cell, turquoise) and cell layers (labyrinth/villi and junctional zone/basal plate). Fetal vessels are bordered by a discontinuous cytotrophoblast layer inboth species, but this is surroundedby twocontinuous syncytiotrophoblast layers inmouse labyrinthand only one in human villi. The exchange surface in the mouse consists of a complex labyrinth of interconnecting villous structures surrounding narrowmaternal blood sinuses, whereas that of the human consists of tree-like, branching villous structures, many of which lie free-floating in a continuous pool of maternal blood; however, some traverse the intervillous space to adhere to thematernal decidua (10, 11). Inmice, invasion is fairly shallow andmostly limited to those areas of thematernal decidua that lie closest to the fetus (light green and light blue cells), although vascular remodeling is evident far upstream of the most deeply invaded trophoblast (10). In humans, endovascular trophoblast cells (light green) remodel maternal uterine veins and spiral arteries by replacingvasoactiveendothelial structures, and interstitial trophoblast cells (lightgreen) invadethroughat least the inner one-third of the uterine myometrium. [Reproduced with permission from The Curators of the University of Missouri (Copyright 2010, The Curators of the University of Missouri).]