Preeclampsia is the most common hypertensive disorder of pregnancy, affecting 2–8% of pregnant women [1]. Preeclampsia is a syndrome of new-onset hypertension and proteinuria typically presenting after 20 weeks of gestation. Preeclampsia can lead to serious maternal complications, such as pulmonary edema, HELLP syndrome, placental abruption, eclampsia, and intracerebral hemorrhage. Preeclampsia is also associated with adverse fetal conditions such as growth restriction, oligohydramnios, and absent or reversed end-diastolic velocity in the umbilical artery. Consequently, preeclampsia is a leading cause of maternal and neonatal morbidity and mortality [1]. While it is well known that the placenta is essential to the development of preeclampsia, the heterogeneous clinical manifestations of the disease suggest multifactorial pathogenesis of preeclampsia. Because the pathogenesis has not been fully elucidated, prevention is not possible, and, even today, the fundamental therapy is to terminate the pregnancy. However, recent progress on genomic analyses using the placenta has gradually revealed molecular mechanisms underlying the pathogenesis of preeclampsia.

The etiology of preeclampsia remains unknown, and a number of theories have been proposed to explain its cause. Zweifel described preeclampsia as a “disease of theories” [2]. Preeclampsia is clinically heterogeneous in its manifestation including the onset of the disease and the fetal growth. Abnormal trophoblast invasion with insufficient uterine spiral artery remodeling often occurs in early-onset preeclampsia and fetal growth restriction [3]. In contrast, normal spiral artery remodeling is usually observed in late-onset preeclampsia with normal fetal growth. So, Ness and Roberts hypothesized that there are distinct origins of preeclampsia, and both poor placentation (placental preeclampsia) and latent vascular dysfunction (maternal preeclampsia) are attributed to the development of preeclampsia [4]. Abnormal placentation is hypothesized to be associated with early-onset preeclampsia with fetal growth restriction. It has been postulated that poor placentation (stage 1) results in the release of factors that lead to excessive maternal inflammatory response and endothelial dysfunction (stage 2) [5, 6]. Thus, in early-onset preeclampsia, the sequence of abnormal trophoblast invasion to systemic endothelial dysfunction has been proposed to occur in two stages. On the other hand, preexisting maternal endothelial dysfunction (as with chronic hypertension, obesity, or diabetes mellitus) is thought to predispose to late-onset preeclampsia, which is more often linked to a normal placenta and normal fetal growth [5, 6]. As a systemic inflammation increases with advancing gestation even in normal pregnancy, pregnancy is a “stress test for endothelial function” for women. This burden to latent vascular dysfunction is enough to induce a severe systemic inflammatory response and subsequent endothelial dysfunction that is characteristic of preeclampsia [6].

Poor placentation is typified by insufficient remodeling of spiral arteries [7–9]. During normal placentation, extravillous trophoblasts (EVTs) invade into the decidua and the inner one third of the myometrium either interstitially or via spiral arteries. EVTs disrupt the endothelium and the smooth muscle layer and replace the vascular wall. These conversions allow spiral arteries to get widely dilated independently of vasomotor control, thereby providing enough blood supply in intervillous space (IVS) to meet the requirements of the fetus [7–9]. On the other hand, in preeclampsia, endovascular trophoblast invasion is restricted to the decidual segments of the spiral arteries, resulting in less dilated vessels and intermittent hypoperfusion due to retention of vasoactive smooth muscle [7–9]. Burton proposed that the impaired vascular remodeling of the spiral arteries may not cause chronic placental hypoxia in itself, but that the retention of smooth muscle will increase the risk of spontaneous vasoconstriction and ischemia-reperfusion injury, leading to generation of oxidative stress [10]. Moreover, high-velocity and turbulent flow into the IVS will damage villous syncytiotrophoblast [10], which may enhance placentally derived stimuli, including pro-inflammatory cytokines, anti-angiogenic factors, and trophoblast microparticles.

Early placental development occurs in a low-oxygen environment, which appears to prevent trophoblast differentiation. As maternal blood flow into the IVS is blocked by endovascular plugs of extravillous trophoblasts prior to 10 weeks’ gestation, the oxygen concentration within the IVS is approximately 20 mmHg (2–3%) [11–13]. Once the plugs are displaced after 10–12 weeks’ gestation, chorionic villi are bathed in maternal blood. As a result, the oxygen concentration rises to 40–80 mmHg (5–10%) in the second trimester [11–13]. The oxygen concentration in the IVS gradually decreases thereafter to the third trimester (40 mmHg) in response to increased fetal oxygen consumption [11–13]. In preeclamptic placenta, the oxygen concentrations in the IVS are generally thought to be low because of defective vascular remodeling of the spiral arteries. However, to date there has been no direct evidence to support this assumption. Burton proposed that there was a slight alteration of total blood flow in the IVS due to insufficient spiral artery remodeling [10]. Moreover, the impairment of villous syncytiotrophoblasts will result in reduced oxygen removal in the IVS, which leads to placental hyperoxia and fetal hypoxia. So, it is unclear whether preeclamptic placentas are actually hypoxic [12], and it is plausible that exaggerated oxidative stress in preeclampsia is secondary to intermittent hypoperfusion rather than chronic placental hypoxia.

The hypoxia-inducible factors (HIFs) are central mediators of cellular adaptation to low oxygen and regulate placental development and maturation [12, 14, 15]. For instance, HIF-1 regulates TGF-beta3, which inhibits trophoblast differentiation and invasion and is overexpressed in preeclamptic placentas [14]. Transcription of HIF-1alpha and HIF-2alpha is differentially regulated under hypoxia in neuroblastoma [16]. HIF-1alpha protein is transiently stabilized under hypoxia (1% O₂), while HIF-2alpha protein gradually accumulates and controls prolonged hypoxic gene activation. Moreover, HIF-2alpha is increased to a greater extent than HIF-1alpha under mild hypoxia (5% O₂). In the placenta, inconclusive results have been reported regarding HIF regulation in response to acute versus prolonged hypoxia [12]. Intriguingly, HIFs are activated through oxygen-independent factors including hormones (angiotensin II), cytokines (IL-1b, TNF-alpha, and NF-kB), and growth factors (TGF-b and IGF) [12, 15], and activation of HIFs during pregnancy is considered to play an important role in the pathogenesis of preeclampsia.

Oxidative stress arises when the production of reactive oxygen species (ROS) exceeds the intrinsic antioxidant defenses [17]. ROS, such as superoxide anion, hydrogen peroxide, and hydroxyl radical, are chemically reactive molecules containing oxygen. ROS arise from the various sources such as mitochondria and cause cellular damage. The placenta is rich in mitochondria, is highly vascular, and is exposed to high maternal oxygen partial pressure, therefore resulting in increased production of ROS [18]. Indeed, increased production of lipid peroxides and oxidative stress are observed in normal pregnant women compared with nonpregnant women [19]. Moreover, preeclampsia is characterized by markedly decreased antioxidants such as glutathione or glutathione peroxidase activity, as well as increased oxidative stress [19, 20]. The higher production of oxidative stress in preeclampsia is probably due to insufficient remodeling of spiral arteries in the section of myometrial segments. Fluctuations in maternal blood flow to the placenta are thought to cause exaggerated oxidative stress. Consistently, labor, in which the placenta is exposed to repeated episodes of ischemia-reperfusion, produces high levels of oxidative stress [17, 21, 22].

Oxidative stress is one of many forms of cellular stress. ROS can activate redox-sensitive transcription factors (i.e., NF-kB, p53, AP1) and protein kinases (i.e., ERK1/2, SAPK/JNK, p38 MAPK) as physiological adaptive changes to alterations in the environment aimed at restoring homeostasis [17]. At higher levels, ROS lead to more extensive and irreparable cell damage, resulting in apoptosis or necrosis. ROS induce release of Ca²⁺ from the endoplasmic reticulum (ER) and are closely associated with mitochondrial and ER function [17]. Thus, oxidative stress rarely occurs in isolation, but is usually accompanied with other forms of cell responses, such as ER stress and unfolded protein responses [17, 23]. Placental oxidative stress causes diverse stress responses of syncytiotrophoblast including increased apoptosis and secretion of anti-angiogenic and pro-inflammatory products [6] and is thought to be an intermediary step in the pathogenesis of preeclampsia.

Inflammatory response is already well developed in normal pregnancy [24], and preeclampsia is a state of an exaggerated inflammatory response including abnormal cytokine production and neutrophil and endothelial cell activation [25]. Both normal and preeclamptic pregnancies exhibit an increased inflammatory response with advancing gestation [26]. There is now ample evidence to support that increased inflammatory response is associated with clinical manifestations of preeclampsia in animal models. For instance, chronic administration of endotoxin [27], TNF-alpha [28], IL-6 [29], and LPS [30] into pregnant rats causes preeclampsia-like symptoms. Activated neutrophils could cause vascular damage with activation of coagulation and platelets [31]. In addition to neutrophils, other immune effectors that appear to play important roles in preeclampsia are cytokines, which are major initiators and mediators of inflammation and endothelial dysfunction [25, 32, 33]. Cytokines can be generally classified as pro- and anti-inflammatory. Cytokines such as IL-1, IL-2, IL-8, IL-18, TNF-alpha, and IFN-gamma are pro-inflammatory. The expression of IL-1, IL-2, IL-18, and TNF-alpha are elevated in the preeclamptic placentas [34–38]. The anti-inflammatory cytokines, IL-4 and IL-10, are also secreted by placental tissues. IL-10 is a potent suppressor of pro-inflammatory cytokines such as TNF-alpha and IFN-gamma, and its placental production is decreased in the preeclamptic placentas [39, 40]. Indeed, IL-10 knockout mice exhibit mild hypertension with fetal growth restriction [41], and hypoxia induces preeclampsia-like features in pregnant IL-10 knockout mice [42]. Although pro-inflammatory cytokines are produced by trophoblasts and also by macrophages and stromal cells of the placenta, the high pro-inflammatory cytokine levels seen in peripheral blood in preeclamptic pregnancies are believed to be caused in great part by monocytes [25, 43, 44]. It has been proposed that the syncytiotrophoblast microparticles (STBM) stimulate the production of the pro-inflammatory cytokines through the activation of monocytes [43, 44]. As markedly increased amounts of STBM are shed into the maternal circulation in preeclampsia [45], STBM impairs maternal vascular endothelial function [46]. Cytokines may contribute to an increased release of STBM by stimulating enhanced trophoblast apoptosis [47], bringing about a vicious cycle of an excess of pro-inflammatory cytokines and the increased shedding of STBM.

Maternal-fetal immune maladaptation is believed to be one of the underlying mechanisms that contribute to the development of preeclampsia.

Preeclampsia can be thought of as a two-stage disorder, and the second stage is typified by the placenta-derived soluble angiogenic and anti-angiogenic factors into the maternal circulation [75].