A review of the mechanism for poor placentation in early-onset preeclampsia: the role of autophagy in trophoblast invasion and vascular remodeling

Abstract

Shallow trophoblast invasion and impaired vascular remodeling of spiral arteries have been recognized in early-onset preeclampsia. Placentation and vascular remodeling are multistep processes, and hypoxia, placental oxidative stress, excessive or atypical maternal immune response to trophoblasts, exaggerated inflammation, and increased production of anti-angiogenic factors such as the soluble form of the vascular endothelial growth factor (VEGF) receptor (sFlt-1) and soluble endoglin (sENG) may play a role in poor placentation in preeclampsia. Recent findings suggest that autophagy plays an important role in extravillous trophoblast (EVT) invasion and vascular remodeling under hypoxia, and sENG inhibits EVT invasion and vascular remodeling by the inhibition of autophagy under hypoxic conditions. In this review, we discuss the relationship between inadequate autophagy and poor placentation in preeclampsia.

Introduction

The pathogenesis of preeclampsia remains largely unknown. However, many researchers support the two-step theory (Steegers et al., 2010) or the three-step theory (Redman and Sargent, 2010). In normal pregnancy, extravillous trophoblasts (EVTs) deeply invade the uterine spiral arteries, disrupt the muscular coat and elastica, and replace the vascular endothelial cells (Pijnenborg et al., 1980). This remodeling dilates the spiral arteries and triggers increased uteroplacental blood flow. The placental oxygen curve estimated by Jauniaux et al. shows that this vascular remodeling starts from a gestational age of 10–12 weeks (Jauniaux et al., 2001). Therefore, failed remodeling at this stage leads to reduced uteroplacental blood flow and hypoxic stress to the fetus and placenta. In stage 1 of early-onset preeclampsia, impaired EVT invasion into maternal spiral arteries causes poor vascular remodeling and induces placental and endothelial damage (Khong et al., 1986).In stage 2, these damaged tissues release anti-angiogenic factors such as the soluble form of the vascular endothelial growth factor (VEGF) receptor (sFlt-1) and soluble endoglin (sENG), a co-receptor for transforming growth factor (TGF)-β₁ and -β₃, which induces maternal intravascular systemic inflammatory responses and endothelial dysfunction, resulting in hypertension and proteinuria after 20 weeks’ gestation, especially in early-onset preeclampsia (Venkatesha et al., 2006, Levine et al., 2006).

During early pregnancy, the placental oxygen concentration is only 2%, while the decidual oxygen concentration is around 8% (Jauniaux et al., 2001). Furthermore, glucose concentration in the intervillous space at 5–12 weeks’ gestation is only one quarter to one fifth of that in maternal serum, coelomic fluid, and amniotic fluid (Jauniaux et al., 2005). Nevertheless, EVTs invade the maternal decidua and myometrium, and induce vascular remodeling under harsh conditions. Indeed, the hypoxia-inducible factor (HIF) 1 system plays a critical role in EVT functions. Therefore, we should understand the mechanism by which EVTs can invade decidua or myometrium and induce vascular remodeling of spiral arteries under physiological conditions of low oxygen in normal pregnancy. In addition, we should understand why shallow EVT invasion and inadequate vascular remodeling occur in early-onset preeclampsia. Burton et al. have shown that total blood flow and hence oxygenation of the placenta was only slightly changed in patients with failure to transform the spiral arteries (Burton et al., 2009); thus, not only hypoxia, but also other unknown factors, may play important roles in the pathogenesis of early-onset preeclampsia.

Recent studies have demonstrated that autophagy is a process of self-degradation of cellular components in which double-membrane autophagosomes sequester organelles and fuse with lysosomes so that the contents can be digested by lysosomal enzymes (Mizushima et al., 2010, Mizushima et al., 2011, Ichimura and Komatsu, 2011). By using this system, cells can survive under starvation or stress conditions such as hypoxia or oxidative stress.

The expression of autophagy-related proteins in the placenta (Oh et al., 2008, Signorelli et al., 2011) and activated autophagy in the placenta of intrauterine growth restriction (IUGR) pregnancies (Hung et al., 2012) have been reported. Recently, we reported that autophagy is recognized in deeply invaded EVTs in the uterine myometrium and perivascular region. We have found that autophagy is essential for EVT functions, invasion, and vascular remodeling under physiological conditions of low oxygen (Nakashima et al., 2013, Saito and Nakashima, 2013). We also reported that anti-angiogenic factor sENG inhibits autophagy in EVTs under hypoxia, resulting in poor EVT invasion and vascular remodeling (Nakashima et al., 2013).

In this review, we discuss the mechanisms of poor placentation in preeclampsia from the perspective of autophagy.

In normal pregnancy, invasive extravillous trophoblasts (EVTs) express integrin α₁β₁, a receptor for collagen 1, collagen IV, and laminin. However, in preeclampsia, the expression of integrin α₁β₁ is downregulated, and this failure to acquire the vascular repertoire of adhesion molecules may explain the impaired invasion of EVTs (Zhou et al., 1993) (Fig. 1). Early in the first trimester (<10 weeks), placental oxygen tension is very low (∼2%; 25.6 mmHG O₂) (Jauniaux et al., 2001), and this low-oxygen environment maintains trophoblasts in an immature, proliferative state mediated by TGF-β₃ through HIF-1α (Caniggia et al., 1999, Caniggia et al., 2000). After the gestational age of 10 weeks, increased placental oxygen tension increases, and may reduce the pool of proliferating trophoblasts and increase the number of invasive trophoblasts. Caniggia et al. (1999) speculated that increased placental oxygen tension reduces the expression of HIF-1α and TGF-β₃ and the failure of TGF-β₃ production at around 9 weeks’ gestation results in shallow trophoblast invasion. However, Lyall et al. (2001) reported that TGF-β₁ and -β₂, and to a much lesser extent, TGF-β₃, were present within the placental bed, and no change in the expression of either isoform of TGF-β was found in the placenta and placental bed in preeclampsia and fetal growth restriction (FGR) compared with those in normal pregnancy.

Smith et al. (2009) reported that NK cells and macrophages were present in the vascular wall at the stage of remodeling (gestational age of 9 ∼ 10 weeks). These NK cells produce matrix metalloproteinase −7 and −9, and urokinase plasminogen activator (uPAR) (Fig. 1) (Smith et al., 2009, Naruse et al., 2009a, Naruse et al., 2009b). These enzymes can break down the extracellular matrix and induce the separation of vascular smooth vessel cells. Cell culture supernatant of uterine NK cells at a gestational age of 12–14 weeks stimulated EVT invasion (Lash et al., 2010, Lash and Bulmer, 2011), and this effect was partially abrogated in the presence of neutralizing antibodies to IL-8 and IP-10 (Fig. 1) (Hanna et al., 2006). Adequate NK cell stimulation might be necessary for EVT invasion and vascular remodeling of uterine spiral arteries. In this regard, Fraser et al. (2012) reported some very interesting findings. They studied the resistance indices using uterine artery Doppler ultrasound. Uterine NK cells isolated from pregnant women with higher resistance indices, i.e., impaired vascular remodeling, were less able to promote invasive behavior of trophoblasts. Furthermore, uterine NK cells isolated from high-resistance-index pregnancies failed to induce vascular apoptosis (Fraser et al., 2012). These findings suggest that dysregulation of uterine NK cells may contribute to the impaired vascular remodeling (Fig. 1). Indeed, Hiby et al. (2004) reported that the combination of maternal killer-cell immunoglobulin-like receptor (KIR) AA and fetal HLA-C2 was a risk factor for preeclampsia (Fig. 1). KIR AA lacks the activation receptor for HLA-C2; therefore, inadequate NK cell activation might induce poor EVT invasion and vascular remodeling, and adequate NK cell activation might be necessary for the placentation. Kaufmann et al. (2003) reported that activated macrophages induce trophoblast apoptosis by the secretion of TNFα and by the expression of indoleamine 2,3-dioxygenase (IDO), which depletes the local level of tryptophan. They speculated that activated macrophages reduce EVT invasion by the induction of apoptosis of EVTs in preeclampsia (Fig. 1). Indeed, the serum level of TNFα is elevated (Meekins et al., 1994) and peripheral blood mononuclear cells produce a lot of TNFα in such cases (Saito et al., 1999).