ELABELA deficiency promotes preeclampsia and cardiovascular malformations in mice

See allHide authors and affiliations

Science  18 Aug 2017:
Vol. 357, Issue 6352, pp. 707-713
DOI: 10.1126/science.aam6607

Modeling a pregnancy disorder

Preeclampsia, a dangerous pregnancy disorder marked by high blood pressure, can lead to premature birth and be life-threatening to the mother and baby. Research leading to effective treatments has been hampered by a lack of informative animal models. Ho et al. identified ELABELA as a hormone produced by the placenta whose levels are lower in preeclampsia (see the Perspective by Wirka and Quertermous). ELABELA-deficient pregnant mice showed clinical signs of preeclampsia, including high blood pressure and elevated urine protein. A proportion of embryos lacking ELABELA displayed defective heart development, and full-term pups had low birth weights.

Science, this issue p. 707; see also p. 643


Preeclampsia (PE) is a gestational hypertensive syndrome affecting between 5 and 8% of all pregnancies. Although PE is the leading cause of fetal and maternal morbidity and mortality, its molecular etiology is still unclear. Here, we show that ELABELA (ELA), an endogenous ligand of the apelin receptor (APLNR, or APJ), is a circulating hormone secreted by the placenta. Elabela but not Apelin knockout pregnant mice exhibit PE-like symptoms, including proteinuria and elevated blood pressure due to defective placental angiogenesis. In mice, infusion of exogenous ELA normalizes hypertension, proteinuria, and birth weight. ELA, which is abundant in human placentas, increases the invasiveness of trophoblast-like cells, suggesting that it enhances placental development to prevent PE. The ELA-APLNR signaling axis may offer a new paradigm for the treatment of common pregnancy-related complications, including PE.

The placenta is a mammalian-specific organ and a critical source of factors responsible for remodeling the maternal cardiovascular system to accommodate the needs of the growing fetus. Defects in placentation often result in intrauterine growth restriction (IUGR) for the fetus and gestational complications such as preeclampsia (PE) for the mother. PE affects 5 to 8% of all pregnancies and remains the leading cause of fetal and maternal morbidity and mortality. Current challenges in PE include early detection and the availability of effective drugs that do not adversely affect fetal development. ELABELA encodes an endogenous ligand for the apelin receptor (APLNR, or APJ). It is first detected in preimplantation human blastocysts and controls the self-renewal of embryonic stem cells (1). In the adult, its expression is restricted to a few tissues, including two endocrine organs, the kidneys and the placenta (1). In rodents, the onset of Ela expression coincides with zygotic transcription (fig. S1A), peaks at the blastocyst stage, and is similarly restricted in the adult. In lower vertebrates, Ela is required for proper endoderm development, and Ela-deficient zebrafish have profound cardiac malformations resulting from impaired migration of cardiac progenitors (2, 3). Zebrafish lacking both Ela and Apelin (Apln), the alternate ligand for Aplnr, have defects in vasculogenesis owing to impaired migration of angioblasts to the midline (4). At present, the molecular effects of ELA signaling downstream of APLNR are unknown, and its involvement in mammalian development and physiology has not been addressed.

To delineate the contribution of ELA to mammalian development, we generated Ela knockout (ElaΔ/Δ) mice using homologous recombination to delete exon 3 encoding the mature ELA peptide (Fig. 1, A and B, and fig. S1, B and C). This strategy did not result in nonsense-mediated decay of the ElaΔ mRNA (Fig. 1C) and presumably preserves the potential noncoding functions of the Ela transcript (5). Only half of the expected ElaΔ/Δ mice from heterozygous intercrosses were obtained at weaning (Fig. 1D) (P < 0.001, chi-square test with df = 1). Notably, this reduced recessive Mendelian inheritance was even more pronounced for ElaΔ/Δ embryos carried by ElaΔ/Δ mothers (67%) than by ElaΔ/+ mothers (51%) (Fig. 1D). This apparent maternal contribution is not due to Ela mRNA being deposited in the oocyte, because the onset of Ela transcription is strictly zygotic (fig. S1A). Rather, we surmise that ELA might be provided by the maternal circulation or uterine environment. At embryonic day 10.5 (E10.5), ElaΔ/Δ embryos can be grouped into three classes: 48.9% were phenotypically normal (class 1), 8.5% were delayed with a hypovascular yolk sac (class 2), and 42.6% had avascular yolk sacs and severe embryonic vascular malformations (class 3) (fig. S1D) that are similar to those previously reported for Apj knockouts (Fig. 1, E to J). In affected ElaΔ/Δ embryos, vasculogenesis appears to be initiated, as evidenced by the presence of a CD31/Pecam+ endothelial plexus, which subsequently fails to undergo remodeling and angiogenic sprouting to form organized vitelline vessels, dorsal aorta, outflow tract, and intersomitic vessels (Fig. 1, K to S, and fig. S1, E to J). The heart tube is poorly looped, with reduced smooth actin muscle (SMA) staining (Fig. 1, Q to S), and the most severely affected embryos (class 3) have pericardial edema (fig. S1, K and L). These cardiac defects are consistent with the first postgastrulation expression of Ela in the primitive foregut overlying the developing heart tube (Fig. 1, T and U) (6). Surprisingly, Ela is not detected in endothelial precursors of the yolk sac (Fig. 1W), whereas Apj expression is ubiquitous in embryonic, allantoic, and yolk sac mesoderm, which gives rise to endothelial cells (Fig. 1, V and X). The expression patterns of Ela and Apj suggest that the observed cardiac defects are partly due to insufficient blood flow to stimulate angiogenesis. Outside of the developing heart tube, Ela is first detected in the chorionic trophoblast of the developing placenta (Fig. 1U and fig. S1, M and N) and is robustly up-regulated after allantoic fusion (Fig. 2A), becoming restricted to syncytiotrophoblasts (STs) at E10.5 (Fig. 2, C and C′). Accordingly, ELA protein is detected by immunohistochemistry in wild-type (WT) STs but not in ElaΔ/Δ placentas (Fig. 2, E and F). ELA-positive STs are juxtaposed to Apj-expressing fetal endothelial cells (Fig. 2, B, D, and D′). Hence, ELA may signal to APJ-expressing cells in a paracrine manner but may also be circulating systemically because the chorioallantoic placenta is perfused by maternal and fetal blood. Indeed, endogenous ELA is detected by enzyme-linked immunosorbent assay (ELISA) in the serum of pregnant females, peaking at midgestation, but not in nonpregnant mice (Fig. 2G). Systemic ELA in a pregnant mother is contributed both maternally and embryonically (Fig. 2H), the former reflecting secretion from the maternal endometrial stroma and kidneys (fig. S2, A to C) and the latter from embryonically derived STs (Fig. 2C). We therefore conclude that ELA is a pregnancy-associated hormone secreted by the developing conceptus and placenta.

Fig. 1 Zygotic deletion of Ela causes midgestation lethality due to cardiovascular defects and phenocopies loss of Apj.

(A) Exon 3 of murine Ela was flanked with loxp sites and excised with cre recombinase to generate the ElaΔ allele lacking the ELA mature peptide (MP) coding region. (B) Schematic of cDNA from WT and ElaΔ alleles. SP, signal peptide. (C) Semi-qPCR of Ela locus from genomic DNA (gDNA) and cDNA. Primer locations are indicated in (B). (D) Distribution of genotypes at E10.5 and at weaning from intercrosses and ElaΔ/Δ (mother) x Ela+/Δ (father) crosses. %P, penetrance; L, number of litters. Data were tested using a chi-square test with 1 degree of freedom for significant deviation from the expected distribution. (E to G) At E10.5, Ela+/Δ embryos are indistinguishable from WT, whereas 43% (n = 17 of 39) of ElaΔ/Δ embryos and 14% (n = 3 of 22) of ApjΔ/Δ embryos display cardiovascular defects along with IUGR. Scale bars, 1 mm. (H to J) At E10.5, Ela+/Δ yolk sacs have normal vitelline vessels, whereas affected ElaΔ/Δ and ApjΔ/Δ embryos have avascular yolk sacs with a ruffled appearance. Scale bars, 1 mm. (K to M) CD31 staining of Ela+/Δ, ElaΔ/Δ, and ApjΔ/Δ yolk sacs reveals poorly matured vasculature in mutant embryos. Scale bars, 50 μm. (N to P) CD31 staining of Ela+/Δ, ElaΔ/Δ, and ApjΔ/Δ head vasculature at E10.5. Scale bars, 300 μm. (Q to S) CD31 (green) and SMA (red) staining of Ela+/Δ, ElaΔ/Δ, and ApjΔ/Δ hearts. Scale bars, 300 μm. (T) In situ hybridization of Ela at E8, showing mRNA localization in the region overlying the developing heart tube (ht) and chordal neural hinge (cnh). Scale bar, 200 μm. (U and V) RNAScope of Ela and Apj in E8 embryo within its decidua showing expression in the primitive foregut (fg) and hindgut (hg) endoderm. Arrowheads indicate the start of Ela expression in the chorionic trophoblast. Scale bars, 100 μm. (W and X) RNAScope of Ela and Apj in E8 yolk sac layers adhering to underlying decidua. en, endoderm; me, mesoderm. Scale bars, 40 μm.

Fig. 2 ELA is a pregnancy hormone required for placental angiogenesis.

(A and B) At E9, Ela is expressed in the chorionic plate (cp) of the chorioallantoic placenta, and its receptor Apj is expressed in fetal allantoic endothelial cells. d, decidua. Scale bars, 1 mm. (C and D) At E10.5, Ela expression in the placenta labyrinth (lb) is restricted to STs, whereas Apj expression is restricted to endothelial cells adjacent to STs. Scale bars, 100 μm. (C′ and D′) Higher magnification showing Ela expression in STs surrounding maternal blood spaces (mbs) and Apj expression in endothelial cells (EC) lining fetal blood spaces (fbs). Scale bars, 200 μm. (E and F) ELA can be detected by immunohistochemistry using an ELA-specific antibody (α C) in WT E10.5 labyrinth in cells lining blood spaces (arrowheads) but not in ElaΔ/Δ placentas. (G) ELISA detects circulating ELA in maternal serum harvested at indicated gestational days (GD). n = number of mice assayed at each gestational time point. NP, nonpregnant. Error bars indicate SEM of three independent experiments. Data were tested with one-way analysis of variance (ANOVA) (red asterisk) and with two-sample Student’s t test (black asterisks). (H) ELISA of GD 10.5 maternal serum harvested from WT or ElaΔ/Δ mothers mated with the WT or ElaΔ/Δ fathers, indicating a maternal and zygotic origin of circulating ELA during pregnancy. n = number of mice assayed at each gestational time point. Error bars, SEM of three independent experiments. Data were tested using one-way ANOVA. In (G) and (H), ELA detected in maternal zygotic knockout is attributed to assay background. (I and J) Hematoxylin and eosin staining of Ela+/Δ and ElaΔ/Δ E10.5 placentas showing poor invasion and angiogenesis of ElaΔ/Δ placentas. Scale bars, 250 μm. al, allantois. (K and L) CD31/Pecam-1 staining of E10.5 Ela+/Δ and ElaΔ/Δ showing a paucity of fetal endothelial cells in the labyrinth. Scale bars, 100 μm. (M and N) Alpp (placenta alkaline phosphatase) staining of E10.5 Ela+/Δ and ElaΔ/Δ placentas showing lack of trophoblasts in the labyrinth. Scale bars, 50 μm. *P < 0.05, **P < 0.01 from indicated tests of significance.

ElaΔ/Δ placentas from affected embryos have thin labyrinths (Fig. 2, I and J, and fig. S2, D and E) with poor vascularization (Fig. 2, K and L), increased apoptosis (fig. S2, F and G), and reduced proliferation (fig. S2, H and I). ElaΔ/Δ placentas from unaffected (class 1) or mildly affected (class 2) embryos, which are intermediately vascularized, nonetheless exhibit delayed ST differentiation, as indicated by reduced alkaline phosphatase and syncytin-1 staining at E10.5 (Fig. 2, M and N, and fig. S2, J and K). Although such placentas eventually develop, allowing embryo survival, the labyrinth of mutant versus WT placentas remains thinner until the end of gestation (fig. S2E).

To understand the pathogenesis of Ela deficiency causing placental dysfunction, we isolated placentas denuded of maternal decidua from WT and ElaΔ/Δ conceptuses (Fig. 3A). We chose to carry out the analysis by E9.5 to avoid the confounding transcriptional changes brought about by major cardiovascular anomalies seen at E10.5. ElaΔ/Δ placentas were categorized into class 1 or class 3 based on the gross morphology of the corresponding embryos (fig. S3A). RNA sequencing (RNA-seq) and principal components analysis revealed that both class 1 and 3 ElaΔ/Δ placentas clustered closer to each other and away from WT placentas (fig. S3B). Because class 1 placentas are grossly indistinguishable from WT counterparts, these results indicate that the observed transcriptional changes are due to ELA deficiency rather morphological defects already present at the time of specimen collection. Gene set enrichment analysis (GSEA) (7) revealed that class 1 and 3 ElaΔ/Δ placentas have a gene signature indicative of an elevated hypoxic response (Fig. 3B; fig. S3, C and D; and table S1). Consistent with this observation, ElaΔ/Δ placentas have high levels of stabilized Hif1α (fig. S3, E and F) and decreased levels of prolyl-hydroxylated Hif1α (fig. S3, G and H), which is targeted for degradation under normoxia (8). Concurrently, and possibly as part of the elevated hypoxic response, Ela deficiency results in an up-regulation of pro-angiogenic genes, even in class 1 placentas that are bereft of discernible vascular defects (Fig. 3C).

Fig. 3 Loss of Ela causes hypoxic response and up-regulation of a pro-angiogenic program.

(A) Schematic of RNA-seq experiment of E9.5 WT versus ElaΔ/Δ labyrinths. (B and C) GSEA analysis of ElaΔ/Δ (classes 1 and 3) showing an up-regulation of hypoxic response and pro-angiogenic genes in ElaΔ/Δ labyrinths, even in morphologically normal class 1 placentas. (D) Gene ontology analysis of genes up-regulated in ElaΔ/Δ labyrinths. In red are pathways enriched in tip cells. P values are derived from a binomial distribution with Bonferroni correction. (E) GSEA detects an up-regulation of endothelial tip cell genes in class 1 ElaΔ/Δ labyrinths. (F) qPCR validation of tip cell–enriched and angiogenic genes in ElaΔ/Δ labyrinths (n = 6 WT; n = 6 ElaΔ/Δ). Error bars indicate SEM of two independent experiments. (G) Esm1 immunofluorescence on E9.5 placenta vibratome sections taken from medial planes containing the maternal central canal. Dotted line marks the position of the transitional zone. al, allantois; lb, labyrinth; d, decidua. Scale bars, 40 μm. (H) Number of Esm1+ cells per section (n = 6 WT; n = 7 ElaΔ/Δ; each section represents a distinct placenta). (I) Seventy-fifth percentile integrated density of Esm1+ cells in each placental sample quantified in (H). Data presented as arbitrary units (A.U.). (J) β-galactosidase (LacZ transgene in AplnΔ allele) staining of Ela+/+;Apln+/Δ and ElaΔ/Δ;Apln+/Δ placentas indicate increased Apln expression in ElaΔ/Δ labyrinths. Scale bars, 300 μm. Data are depicted as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 of two-sample Student’s t test.

A close examination of differentially regulated genes revealed a dramatic enrichment in genes and pathways defining endothelial tip cells (Fig. 3, D and E) (9). Tip cells form the leading edge of sprouting endothelial cells and migrate in response to pro-angiogenic signals (10). Functioning in the same way as axonal growth cones, tip cells extend filopodia to determine the direction of the angiogenic sprout, whereas trailing stalk cells proliferate to enable lumenogenesis and extension of the vascular sprout (11, 12). Gene ontology analysis confirmed functional hallmarks of tip cell identity such as vascular endothelial growth factor (VEGF) and semaphorin signaling, hormone secretion, axonogenesis, and filopodia extension (Fig. 3D). Quantitative polymerase chain reaction (qPCR) analysis validated the up-regulation of tip cell markers and angiogenic genes, including Vegfa, Apln, Plgf, Esm1, Igfbp3, Flt4, and Adm (Fig. 3F). This was confirmed by direct immunostaining against endogenous Esm1, a specific tip cell marker (9, 13, 14), which was significantly up-regulated in both number and intensity in ElaΔ/Δ placental sections, indicating ectopic tip cell differentiation (Fig. 3, G to I). Using a Apln-LacZ knockin reporter (15) as an alternate marker of tip cell identity, we found that there are indeed more Apln-positive tip cells in ElaΔ/Δ labyrinths, with an overall stunted architecture characterized by little or no extension and branching of angiogenic sprouts (Fig. 3J and fig. S3, I and J). This finding suggests that the absence of ELA causes an expansion of tip versus stalk cells, which is expected to impair lumenogenesis and sprout extension, such as in mice haplo-insufficient for the Notch ligand Dll4 (16, 17). Such an imbalance in tip cell identity is consistent with proliferative perturbations seen in ElaΔ/Δ labyrinths, which are hypoproliferative (fig. S2, H and I), and ElaΔ/Δ yolk sacs, which are, conversely, hyperproliferative (fig. S3, K to O), with an overall negative effect on angiogenesis. Our results suggest that the ELA actively suppresses the expression of tip cell genes such as Esm1 and Apln.

Defects in genes required for placental development and angiogenesis frequently lead to preeclampsia in mice (18, 19). In light of the placental defects seen in ElaΔ/Δ conceptuses, including a prominent gene signature of increased inflammatory response (fig. S4A), and increased expression of Esm1 (20, 21) and Adm (22), which have been linked to PE in humans, we hypothesized that ElaΔ/Δ mothers might exhibit symptoms of PE. We thus assessed ElaΔ/Δ pregnant mice for signs of proteinuria and hypertension, two diagnostic hallmarks of PE. Indeed, by decreasing the number of WT Ela alleles in fetuses and their pregnant mothers, the urine protein/creatinine ratio at gestational day (GD) 15 increased dramatically, indicating an inverse correlation between endogenous ELA levels and the severity of proteinuria (Fig. 4A). At the end of pregnancy, histology and transmission electron microscopy of kidney glomerular sections revealed signs of endotheliosis in ElaΔ/Δ pregnant mothers (fig. S4, B to G), a unique renal pathology of women suffering from PE (23, 24). Glomeruli from ElaΔ/Δ pregnant mothers were swollen and had occluded capillaries, with evidence of protein and vesicular deposition on endothelial cells, absence of proper endothelial fenestration, and coagulation of red blood cells in capillary lumens (fig. S4, B to G). Podocytes, on the other hand, appeared normal. Proteinuria was not observed in nonpregnant ElaΔ/Δ mice (fig. S4H), indicating that the renal pathology is unique to pregnancy. Next, we employed a tail-cuff method to measure systolic blood pressure (BP) throughout pregnancy, after training the mice for a minimum of 3 days before mating. Although there were no significant differences in the nonpregnant baseline BP between WT and ElaΔ/Δ mice, pregnant ElaΔ/Δ mothers (mated to ElaΔ/Δ fathers) had significantly higher systolic BP, which returned to normal levels postparturition (Fig. 4B), and delta BP (pregnant BP minus baseline BP) on GD 16 and 18 (Fig. 4C). In addition, ElaΔ/Δ pups from ElaΔ/Δ mothers collected by caesarean section at E18.5 were significantly lower in weight compared with WT pups from WT mothers (Fig. 4D). This is reminiscent of IUGR, which frequently accompanies PE and placental insufficiency. Our findings indicate that ElaΔ/Δ mice suffer from PE and suggest that ELA is necessary for regulating maternal cardiovascular homeostasis to prevent gestational hypertension. To determine whether the loss of ELA is upstream of well-established biomarkers of human PE, we measured both maternal plasma and placental mRNA levels of sFlt1 (24), Vegf (25), and Plgf (19, 26). At late gestation, ElaΔ/Δ placentas have increased levels of sFlt1, Vegfa, and Plgf mRNA (fig. S4I), although these transcriptional changes did not translate into significantly elevated plasma levels of the respective proteins (fig. S4, J to L). Altogether, these data indicate that ElaΔ/Δ mice are not developing PE symptoms simply due to a decrease in the Plgf/sFlt1 ratio but suggest that ELA acts independently of, and possibly earlier than, these angiogenic factors in the pathogenesis of PE.

Fig. 4 Endogenous ELA prevents preeclampsia (PE), and exogenous ELA administration rescues PE symptoms in Ela-deficient mice.

(A) Urine protein/creatinine ratios from GD 15 pregnant mothers (♀) of indicated genotype mated with fathers (♂) of indicated genotypes. Each dot represents an individual mouse. Error bars indicate SEM. (B) Repeated tail-cuff systolic blood pressure measurements of WT mothers (n = 7, mated to WT fathers) and ElaΔ/Δ mothers (n = 5, mated to ElaΔ/Δ fathers) at the indicated gestational age. Dotted line indicates day of parturition. Error bars indicate SEM. Two-way ANOVA analysis detected a significant interaction between time and genotype, F(7,49) = 2.074; P = 0.0413; i.e., ElaΔ/Δ females developed significantly higher systolic BP compared to the controls as pregnancy progressed. Asterisks indicate significance of two-sample unpaired t test between WT and ElaΔ/Δ on the indicated GD. (C) BP readings from (B) calculated in the form of delta BP (BP of indicated GD minus baseline nonpregnant BP of the same mother). Each dot represents BP of one mouse averaged over 20 readings; error bars indicate SEM. Two-way ANOVA test detected a statistically significant difference in mean delta BP between WT and ElaΔ/Δ mice; F(1,16) = 11.28, P = 0.0100. Asterisks indicate significance of two-sample unpaired t test. (D) Weight of pups at E18.5 collected by caesarean section. Each dot represents one pup. Error bars indicate SEM. (E) Urine protein/creatinine ratios of WT mothers (mated to WT fathers) (black squares) and ElaΔ/Δ mothers (mated to ElaΔ/Δ fathers) (red circles) implanted at GD 7 with infusion pumps containing either phosphate-buffered saline (PBS) (closed symbols) or synthetic ELA peptide (open symbols) measured at GD 15. Each dot represents one mouse; error bars indicate SEM. (F) Systolic delta BP measurements of subjects in (E) measured at GD 14, 16, and 18. Each dot represents one mouse; error bars indicate SEM. Two-way ANOVA test detected a statistically significant difference in mean delta BP between ElaΔ/Δ + PBS and ElaΔ/Δ + ELA mice; F(1,16) = 6.938, P = 0.0300. Asterisks indicate significance of two-sample unpaired t test. (G) Immunohistochemistry of ELA with α C biotinylated ELA-specific antibody on human first trimester (8 + 3 weeks) placental formalin-fixed paraffin-embedded sections. Scale bars, 200 μm. (H) Transwell invasion assay using Jar choriocarcinoma cells cultured in the presence of increasing concentrations of synthetic ELA peptide. Each dot represents the mean of three wells; error bars indicate SEM of three independent experiments. (I) Working model: mouse ELA, produced by ST cells, signals to APJ expressed on fetal endothelial cells (ECs) to facilitate normal placental angiogenesis. ELA also enters the maternal circulation, where it acts systemically to prevent symptoms of preeclampsia during pregnancy. Tb, trophoblast. Data are depicted as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001 of two-sample Student’s t test, unless otherwise stated.

It is noteworthy that overexpression of Apln (Fig. 3F), which is the alternate ligand for APJ, is not sufficient to rescue ElaΔ/Δ placentas, suggesting that these ligands elicit different signaling outcomes. Indeed, unlike ElaΔ/Δ mothers, AplnΔ/Δ mothers do not develop hypertension (fig. S5A) and in fact have significantly lower levels of proteinuria (fig. S5B). Similarly, AplnΔ/Δ placentas do not aberrantly up-regulate endothelial tip cell markers Esm1 and Igfbp3, as seen in ElaΔ/Δ placentas (fig. S5C). To further investigate the biological differences between APLN and ELA, we treated APJ-expressing primary allantoic cultures from somite stage embryos with equal concentrations of ELA, APLN, or both (fig. S5D). We found that ELA and APLN elicited opposite effects on the expression of Esm1 and several hypoxic response genes (fig. S5, D to E′). Lastly, we found that ELA could directly repress the expression of Apln in these allantoic explants (fig. S5F), raising the possibility that Apln derepression in the absence of ELA drives excessive and pathogenic tip cell differentiation. Indeed, we find that in a litter of ElaΔ/Δ null embryos, the most severely affected embryos are the ones expressing the highest levels of Apln (fig. S5G). Moreover, the two ligands display distinct spatiotemporal expression where Ela is highly concentrated and restricted to the developing heart, caudal neural tube and trophoblasts, whereas Apln is diffusely expressed and widespread in all embryonic and extraembryonic tissues (fig. S5, H to I′). Altogether, our data demonstrate that Ela and Apln are biologically distinct and elicit opposing effects on placental angiogenesis and symptoms of PE.

Because ELA appears to act as a systemic hormone during pregnancy, we asked whether administration of synthetic ELA during pregnancy may alleviate PE symptoms. ELA infusion did not affect BP and proteinuria parameters in pregnant WT mice (Fig. 4F), nor did it have noticeable side effects on embryogenesis, as measured by fetal weight, morphology, and subsequent postnatal development. We were able to normalize proteinuria (Fig. 4E) and BP (Fig. 4F) in ElaΔ/Δ pregnant mothers infused with recombinant ELA from GD 7.5 onwards, which is consistent with our model that ELA acts as a systemic hormone. Furthermore, infusion of ELA rescued the weight of ElaΔ/Δ fetuses (Fig. 4D) and glomerular endotheliosis in pregnant ElaΔ/Δ mothers, as assayed by periodic acid–Schiff and α-fibrinogen staining (fig. S6).

In humans, we find that ELA is predominantly expressed in villous cytotrophoblasts and STs of first-trimester placental tissue (8 + 3 weeks) (Fig. 4G) and term placentas (fig. S4, M and N). In PE, extravillous trophoblast invasion and subsequent spiral artery remodeling are frequently incomplete, leading to shallow and defective placentation (27). We therefore hypothesized that, in humans, ELA might contribute to trophoblast invasion. Indeed, addition of ELA to trophoblast-like JAR choriocarcinoma cells significantly increased their invasiveness in transwell invasion assays (Fig. 4H) (28). These data suggest that ELA, secreted from the ST layer, has a paracrine effect on trophoblasts differentiating into invasive extravillous trophoblasts. ELA activity potentiates invasion and might enhance subsequent spiral artery remodeling to prevent the development of PE during human pregnancy.

Pregnancy is a unique physiological state associated with increased cardiovascular demand and burden. Many processes work in concert to impart cardiovascular homeostasis in the pregnant female, although to date these are largely unknown. In the mouse, we propose that ELA produced by placental trophoblasts functions in two ways to prevent gestational hypertension (Fig. 4I). First, ELA exerts paracrine effects on fetal endothelial cells, where it curbs inappropriate differentiation of endothelial tip cells. This enables normal branching angiogenesis and the formation of an adequate labyrinth network required for proper perfusion. Second, ELA enters the maternal circulation to regulate cardiorenal function. We speculate that the latter role might have a direct effect on the maternal endothelium (e.g., through the stimulation of vasodilatory mechanisms such as nitric oxide production) (29) or by regulating diuresis (30). Although the PE-protective effects are presumably achieved through APJ signaling in endothelial cells, we do not rule out a possible contribution from as-yet-unidentified ELA receptors (1). ELA deficiency in the mouse leads to classical PE symptoms together with gross abnormalities in placental development. Similarly, ELA is expressed by trophoblasts in the chorionic villi of human placentas and potentiates trophoblast invasion in vitro. We speculate that in humans ELA might contribute to placentation by stimulating trophoblast migration and invasion, in addition to direct effects on the maternal endothelium, although these remain to be investigated.

In conclusion, we propose that ELA is a circulating hormone produced by the mammalian placenta to ensure cardiovascular integrity of both mother and fetus during pregnancy. Our results raise the possibility that transient enforcement of the ELABELA-nergic axis might therefore be beneficial for conditions that display hypertension, such as, but not limited to, preeclampsia.

Supplementary Materials

Materials and Methods

Figs. S1 to S6

Table S1

References (3136)

References and Notes

  1. Acknowledgments: We thank T. Quertermous (Stanford University, USA) for sharing the Apln and Apj knockout mice; T. Veenboer (AMC, Netherlands) for assistance with human placenta samples; D. Kalicharan for technical assistance with electron microscopy; members of the Advanced Molecular Pathology Laboratory, Institute of Molecular and Cell Biology, for assistance with histopathology; the Institute of Medical Biology Microscopy Unit for assistance with imaging; and the Genome Institute of Singapore sequencing facility for assistance with RNA-seq. RNA-seq data are deposited in the National Center for Biotechnology Information (NCBI) Sequence Read Archive under accession number PRJNA391361. The data reported in this manuscript are tabulated in the main paper and in the supplementary materials. The authors acknowledge financial support from the A*STAR Joint Council Organization (Young Researcher Collaborative Grant) and a Strategic Positioning Fund on Genetic Orphan Diseases from the Biomedical Research Council, A*STAR, Singapore. M.v.D. and S.B. are supported by a Ferring Research Institute Innovation Grant. B.R. is a fellow of the Branco Weiss Foundation, a recipient of the A*STAR Investigatorship, an EMBO Young Investigator, and an AAA fellow from AMC/VUmc. The authors declare no competing financial interests. B.R. and L.H. are inventors on patent application 10201605841S submitted by A*STAR, Singapore, which covers the use of ELABELA for the diagnosis and treatment of preeclampsia. The ELABELA-deficient mice and custom antibody are available from B.R. under a material transfer agreement with A*STAR, Singapore.

Stay Connected to Science

Navigate This Article