Discussion
In the present study, we demonstrated for the first time that placental 11β-HSD2 dysfunction leads to hallmark of PE features, deficient trophoblast invasion, abnormal placentation and reduced placental blood flow in pregnant rats. Further, using cultured human placental cells, we revealed that 11β-HSD2 downregulation suppresses migration and invasion of EVTs and promotes sFlt1 release. These data strongly indicate that dysregulation of placental 11β-HSD2 play critical roles in PE development.
Prior studies have demonstrated that placental 11β-HSD2 expression and activity are suppressed by systemic administration of CBX (Lindsay et al., 1996; Sanna et al., 2016). Similarity, we showed that systemic administration and placenta-targeted delivery of CBX resulted in a decrease in 11β-HSD2 expression and activity in placentas. In accordance with that 11β-HSD2 converts corticosterone into 11-hydroxycorticosterone, an increase in corticosterone concentration in placenta was exhibited in the pregnant rats with administration of CBX (1.2mg/kg & 2.4mg/kg ) or CBX-CSA, indicating that inhibition of 11β-HSD2 lead to excess GC in placenta. As expected, systemic administration of CBX could lead to increased corticosterone concentration in maternal circulation because 11β-HSD2 in maternal peripheral tissues besides placenta was inhibited. Of note, our data also demonstrated that specific inhibition of placental 11β-HSD2 by CBX-CSA could increase corticosterone level in maternal circulation in late gestation. It indicates that a large amount of corticosterone is metabolized by the placenta in late gestation, and confirms that placental 11β-HSD2 constitutes a barrier for glucocorticoids across placenta.
As mentioned, local GC concentration in tissues is controlled by 11β-HSD1 and 11β-HSD2(Hunter & Bailey, 2015). Some studies have demonstrated that 11β-HSD1 in fetal membranes regulates prostaglandin synthesis and metabolism through control of local GC concentration (Li et al., 2014; Mirazi et al., 2004). Since no specific GC receptor antagonist is commercially available, we investigated the effect of DEX administration on pregnant outcome in order to mimic the effects of excess GCs. PE-like features including hypertension, nephropathy and increased level of sFlt1 in maternal circulation were induced in the pregnant rats with DEX administration. In the cultured human placental cells, reduced migration and invasion and increased sFlt-1 secretion were induced by cortisol in the cells with 11β-HSD2 knockdown. Together, it strongly suggests that excess GCs mediate the development of PE caused by placental 11β-HSD2 downregulation.
Maternal endothelial dysfunction in PE mainly results from the factor(s) released by the ischemic placenta. Among them, imbalance release of sFlt1 and PlGF plays critical roles in PE development (Seki, 2014; Szpera-Gozdziewicz & Breborowicz, 2014). Many studies have shown that sFlt1/PlGF ratio is markedly increased in PE patients and in women with pre-existing conditions predisposing or mimicking PE (Hodel et al., 2019; Saleh et al.,;2017; Zeisler et al., 2016). In consistence with these studies, we demonstrated that an increased circulatory sFlt-1/PlGF ratio was found in the rats with PE-like features induced by placental 11β-HSD2 downregulation. Moreover, we found that 11β-HSD2 downregulation promoted sFlt1 release but did not affect PlGF release, indicating that increased sFlt1 secretion from placenta is one of the mechanisms of the PE development induced by placental 11β-HSD2 dysfunction.
In the present study, we demonstrated that 11β-HSD2 modulates sFlt1 release through proteolytic process of Flt1 in placenta. Membrane anchored metalloproteases of the ADAM family assume central functions in the living cell by the controlled cleavage and release of biologically active proteins and peptides from the membrane surface (Jones et al, 2016; Zunke &Rose-John,2017). ADAM10 and ADAM17, the well characterized members of the ADAM family, have been shown to mediate the release of a number of cytokines and various receptors from cell membrane in human placentas (Hung et al., 2008; Ma et al.,2011; Yang et al., 2012; Zhao et al., 2010). Our previous study has shown that ADAM10 can mediate sFlt1 release in human placenta (Hu et al., 2015). In the present study, we confirmed that ADAM17 is also involved in sFlt1 secretion in placenta. The regulatory mechanisms responsible for ADAM10 and ADAM17 expression in placentas remain to be elusive. In the present study, we firstly proved that excess GCs caused by 11β-HSD2 dysfunction could upregulate ADAM17 but not ADAM10 mRNA expression. As bioinformatic analysis shows seven putative GREs in ADAM17 gene promoter, we then demonstrated that these elements conferred GCs upregulation ofADAM17 gene transcription using the molecular biological approaches. Of note, Yang et al (2012) have proposed that increased expression of ADAM10 and ADAM17 is associated with oxidative stress, a common phenomenon in PE placenta. Interestingly, our previous study has shown that decreased H2S production contributes to increased ADAM10 expression in PE placentas (Hu et al., 2015). The present study demonstrated that excessive GCs could promote ADAM17 expression in placentas. All together, it may indicate that ADAM10 and ADAM17 are the final paths of various factors linked to PE development.
As mentioned, the first stage of PE is attributed to impaired functions of EVTs. In the cultured human cell lines of EVTs, we demonstrated that 11β-HSD2 regulates migration and invasion function of EVTs. In the animal model, we showed that placental 11β-HSD2 maintains normal interstitial trophoblast invasion, endovascular SA remodeling, placental blood flow and placental morphology, indicating that placental 11β-HSD2 is crucial for normal placentation and placental development. Of note, a number of prior studies have proposed that IUGR in rodents caused by administration of CBX is attributed to excess active GCs because GCs have been known to inhibit fetal growth (Murphy et al., 2012; Welberg et al., 2000). In human studies, it has been assumed that downregulation of placental 11β-HSD2 is associated with IUGR because GCs have detrimental effects on fetal growth (Causevic & Mohaupt, 2007; Hofmann et al, 2001). Our findings indicate that abnormal placentation and reduced placental blood flow caused by excess GCs due to 1β-HSD2 dysfunction also contribute to IUGR.
The limitation of this study is that we did not elucidate the mechanisms responsible for excessive GC suppression of migration and invasion function in EVTs. It is known that invasion function is associated with expression of MMPs in EVTs, in particular, MMP2 and MMP9 are crucial for invasion function of EVTs (Peng et al., 2016; Zhang et al., 2013). Interestingly, some studies have shown that synthetic GCs can regulate MMP2 and MMP9 expression in trophoblast cell lines (Mandl et al., 2006; Pavek et al., 2007). More recently, Kisanga et al (2018) have shown that synthetic GCs can suppress migration and invasion of EVTs and lead to more than 3,000 genes changed, suggesting that GCs modulate EVT function via various signaling pathways. Nevertheless, the mechanisms underlying GCs regulation of trophoblast invasion remains to be further investigated.
In conclusion, our study has demonstrated a previously unrecognized role of 11β-HSD2 in placental development and function. Excessive active GCs caused by placental 11β-HSD2 downregulation can lead to impaired invasion function of EVTs, abnormal placentation and increased sFlt1 release from placenta via elevated ADAM17 expression. Thus, our data provide in vivo evidence that dysregulation of placental 11β-HSD2 plays a critical role in pathogenesis of PE. It is therefore valuable to develop strategies to treat PE by using specific activators of 11β-HSD2 or specific inhibitor to block GCs in placenta.