MicroRNA-16 inhibits feto-maternal angiogenesis and causes recurrent spontaneous abortion by targeting vascular endothelial growth factor

Recurrent spontaneous abortion (RSA) is a common health problem that affects women of reproductive age. Recent studies have indicated that microRNAs are important factors in miscarriage. This study investigated the role of miR-16 in regulating vascular endothelial growth factor (VEGF) expression and the pathogenesis of RSA. In this report, clinical samples revealed that miR-16 expression was significantly elevated in the villi and decidua of RSA patients. In vitro, miR-16 upregulation inhibited human umbilical vein endothelial cell proliferation, migration and tube formation. Conversely, the downregulation of miR-16 reversed these effects. In vivo, we demonstrated that abnormal miR-16 levels affect the weights of the placenta and embryo and the number of progeny and microvascular density, as well as cause recurrent abortions by controlling VEGF expression in pregnant mice. VEGF, a potential target gene of miR-16, was inversely correlated with miR-16 expression in the decidua of clinical samples. Furthermore, the luciferase reporter system demonstrated that miR-16 was found to directly downregulate the expression of VEGF by binding a specific sequence of its 3′-untranslated region (3′UTR). Collectively, these data strongly suggest that miR-16 regulates placental angiogenesis and development by targeting VEGF expression and is involved in the pathogenesis of RSA.

normal uteroplacental development 11 . Abnormal angiogenesis is considered one of the most important causes of RSA. Identifying a possible mechanism for the regulation of placental angiogenesis might lead to a treatment to combat RSA.
miRNAs are a class of conservative, 19-to 24-nucleotide non-protein-coding RNAs that can regulate gene expression by targeting messenger RNAs (mRNAs) via binding to specific sites in the 3′ UTR, resulting in translational repression, cleavage or destabilization 12,13 . miRNAs elicit critical changes in gene expression programs that underlie diverse aspects of biology, such as embryogenesis and regulation of cellular differentiation, metabolism, proliferation and apoptosis 14 . An increasing number of studies have demonstrated that miRNAs are expressed abundantly in the human placenta, and dysregulation of miRNAs has been associated with RSA pathogenesis 7,15 .
The objectives of the current study were to investigate the expression level of miR-16 in RSA, to evaluate the role of miR-16 on placental angiogenesis both in vitro and in vivo, to provide evidence for the regulation of VEGF expression by miR-16 and to reveal the role this regulation plays in the pathogenesis of RSA (see Supplementary  Fig. S1).

Results
The abundance of mir-16 in the villi and deciduais elevated in RSA patients. Immunohistochemistry revealed that VEGF expression was remarkably lower in RSA patients compared with control pregnant women (Fig. 1A). To investigate whether a specific miRNA regulates VEGF expression in RSA, we detected a total of 8 potential miRNAs that targeted the VEGF gene. The expression of these 8 miRNAs was demonstrated by qPCR. We found that the expression of miR-16 in the decidua of the RSA group was significantly greater than that in the decidua of the control group (Fig. 1B). The higher expression of miR-16 was confirmed in the villi of RSA cases (Fig. 1C). Therefore, these results suggested that miR-16 is overexpressed in RSA.

mir-16 inhibits placentation and angiogenesis in vitro and in vivo.
Considering the inhibitory effect of miR-16 on VEGF expression, which is related to placental angiogenesis, maintenance and stabilization, we assessed the anti-angiogenic activity of miR-16 by transfecting Human Umbilical Vein Endothelial Cells (HUVECs) with a miR-16 mimic/inhibitor in vitro. qPCR analysis demonstrated that transfecting HUVECs with the miR-16 mimic increased the miR-16 levels by approximately 14-fold over the endogenous transcription level. In contrast, transfecting HUVECs with a miR-16 inhibitor significantly decreased the miR-16 levels ( Fig. 2A). It is known that angiogenesis demands the proliferation and migration of endothelial cells. The effects of miR-16 on the VEGF-induced proliferation and migration of HUVECs were evaluated. The overexpression of miR-16 in HUVECs significantly inhibited HUVEC proliferation and migration, whereas the reduction of miR-16 expression promoted HUVEC proliferation and migration ( Fig. 2B-D). In tube formation assays, the length of formed tubules was calculated using inverted phase contrast microscopy, which directly revealed the ability of the HUVECs to form tubular structures. Moreover, we discovered that miR-16 overexpression significantly repressed HUVEC tube formation, whereas transfection with the miR-16 inhibitor increased tube formation (Fig. 2E,F).
To identify the effect of miR-16 on placental angiogenesis in vivo, cholesterol-conjugated miRNAs were locally injected into the placenta using a percutaneous ultrasound-guided approach on day 7.5 of pregnancy. Mice fetuses and placentas were collected 7 days after injection (Fig. 3A). Although miR-16 overexpression did not affect the number of embryos, the number of progeny was significantly reduced after miR-16 injection (Fig. 3B,C). Moreover, the weights of the placenta and embryo were significantly decreased in the miR-16 overexpression group compared with the control group (Fig. 3D,E). However, the weight ratio of fetus/placenta at 14.5 days of embryogenesis did not differ between the two groups (Fig. 3F). Subsequent histologic analysis revealed that the total placental vasculature and the number of capillaries were significantly inhibited by miR-16 mimic injection but significantly promoted by miR-16 inhibitor injection (Fig. 3G,H, Table 1). Immunohistochemistry also showed similar results in which miR-16 overexpression inhibited placental angiogenesis, whereas the inhibition of miR-16 expression induced placental angiogenesis (Fig. 3I,J). VEGF is a direct target of mir-16. miRNAs exert their function mainly by targeting the 3′ UTR of target mRNAs. To demonstrate whether miR-16 affects the expression of VEGF, we first evaluated the correlation between VEGF expression and the miR-16 level in placentas of pregnant mice. Immunohistochemistry and western blotting confirmed that the VEGF protein expression level was significantly decreased by miR-16 mimic injection but significantly elevated by miR-16 inhibitor injection ( Fig. 4A-D). In addition, overexpression due to injection of the placenta with miR-16 mimics significantly decreased the protein expression of Proliferating Cell Nuclear Antigen (PCNA). However, injection with miR-16 inhibitors significantly upregulated the PCNA protein levels (Fig. 4E,F).
In the clinical specimens, the VEGF levels in the decidua with high miR-16 expression were significantly lower than those in the decidua with low miR-16 expression (Fig. 5A). To verify whether miR-16 directly suppresses VEGF expression, we analyzed the potential miR-16 seed sequence in the 3′ UTR of VEGF using the online publically available algorithms (TargetScan andmiRanda) and cloned a wild-type or mutant fragment containing the binding sequence (376 bp) into the luciferase reporter gene system (Fig. 5B). Western blot analysis revealed that VEGF expression was significantly downregulated by miR-16 mimic transfection in HTR-8/SVNEO and JEG-3 cells. However, the inhibition of miR-16 upregulated VEGF levels (Fig. 5C,D). The miR-16 mimic/inhibitor and the redesigned luciferase reporter plasmid were then co-transfected into HTR-8/SVNEO and JEG-3 cells. The results showed that overexpression of miR-16 in both cell lines led to significantly reduced luciferase activity for the wild-type, whereas the miR-16 knockdown increased wild-type luciferase activity. In contrast, the activity of the luciferase reporter gene linked to the 3′ UTR of mutant VEGF did not change in the presence of the miR-16 mimic/inhibitor (Fig. 5E,F). Quantification of the percentage of the distance migrated was achieved by measuring these distances in 5 highpower fields. (E,F) Representative images of tube formation (scale bar = 50 μ m). Quantification of the vascular tube structure was conducted by measuring the total tubule length in 5 high-power fields. Data are expressed as the mean ± SD. *P < 0.05 vs. control mimic/inhibitor group, n = 5.

Discussion
The identification and validation of novel biomarkers for RSA is a high priority not only for the diagnosis and clinical follow-up of RSA but also for defining novel therapeutic strategies. During the past few years, although there have been substantial advances in the awareness of RSA, major limitations still exist in managing RSA. Endometrial angiogenesis and decidualization are prerequisites for a successful implantation and a good outcome of pregnancy. In the uterus during pregnancy, critical angiogenic signals likely are produced by the decidualizing endometrial cells acting on the endothelial cells to promote their proliferation, migration and differentiation 16 . Accumulating evidence shows that placental angiodysplasia and endothelial dysfunction may contribute to cases of RSA 17,18 . A majority of pregnancies that result in miscarriage are believed to be caused by defective placentation associated with an absence of physiological changes in maternal spiral arteries 19 .
miRNAs are believed to play important roles in a variety of gynecological reproductive diseases including leiomyoma, endometriosis, and RSA 20 . Indeed, there is a differential expression of miRNAs in the villi and decidua of RSA patients, and all significantly differentially expressed miRNAs affect pregnancy by regulating adhesion. Furthermore, a few of the miRNAs can lead to RSA by regulating the expression of angiogenesis-related genes. All of these miRNAs act together to regulate a variety of physiological functions, leading to RSA. One study suggested that miR-16 inhibits angiogenesis, trophoblast cell proliferation, and migration, and inhibits angiogenesis through VEGF suppression 21 . These data imply that miR-16 may play a critical role in placental angiogenesis.
In the present study, we initially detected the expression level of miR-16 in samples of RSA patients and control women. Importantly, quantification of the data indicated that miR-16 was expressed at significantly higher levels in the villi and decidua of the RSA patients than that in the controls. These results indicate that the status of miR-16 expression should be determined in RSA patients. In vitro, miR-16 overexpression in HUVECs suppresses proliferation, migration and tube formation, suggesting that miR-16 may be a novel suppressor of and may play a critical role in angiogenesis. Moreover, injection with miR-16 in the placenta of pregnant mice caused remarkably decreased placental vasculature and microvascular density, leading to aberrant placentation and spontaneous abortion. A recent study indicated that miR-16 inhibits the proliferation, migration and angiogenesis-regulating potential of mesenchymal stem cells 22 . Our data are consistent with a previous study demonstrating that miR-16 exerts inhibitory biological features during placental angiogenesis.
In addition, we validated VEGF as a direct target gene for miR-16, and we found that miR-16 negatively regulates VEGF expression by directly targeting the 3′ UTR of VEGF mRNA in HTR-8/SVNEO and JEG-3 cells. VEGF, acting as a key angiogenesis promoter, plays a significant role in physiological and pathophysiological vascular development and maintenance 23 . It has been shown that VEGF expression is found in the decidua cells of early pregnancy, and it participates in physiological placental angiogenesis during embryogenesis and reproductive functions during the initial stages of pregnancy 24 . Dysregulation of VEGF expression has been linked to the onset of placental pathologies including preeclampsia, early pregnancy loss and intrauterine growth restriction and is deeply implicated in multiple steps of RSA occurrence and development 25,26 . Previous studies have indicated that early pregnancy loss is associated with low VEGF expression and that altered VEGF expression is likely to contribute to the etiology of RSA 27,28 . In this study, we demonstrated that miR-16 mediates the reduction of VEGF and found that normal VEGF expression is a notable feature and may be one of the many critical events that occur in placental angiogenesis. Aberrant angiogenesis may be partially attributed to overexpressed miR-16 in the placenta of RSA patients.
Overall, the current study found that miR-16 expression is upregulated in the villi and decidua of RSA patients. Our data revealed that the important molecular mechanism by which miR-16 exerts its negative effects on placental angiogenesis occurs via VEGF suppression. However, our work is merely a starting point for the study of miRNAs in RSA, and the exact regulatory mechanism and related signaling pathways of miR-16in RSA still require more experimental evidence to verify. Taken together, the identification of miR-16 provides a potential diagnostic marker and therapeutic target for RSA patients.  (Table 2). Placental villi and decidua tissue were also obtained from cases of induced abortion between 6 and 12 gestational weeks. Medical abortion was performed as previously described 29 . All samples were examined to exclude chromosomal, anatomical, hormonal and infectious pathologies. All tissue samples were snap-frozen in liquid nitrogen for real-time polymerase chain reaction (qPCR) analyses or were fixed in 10% formalin and embedded in paraffin for immunohistochemical analyses. The institutional ethics committee of the Ningxia Medical University approved the study protocol, and written informed consent was obtained from participants. The human experiments were performed in accordance with the relevant guidelines, including any relevant details. miRNA expression assay. Total RNAs were extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. cDNA was synthesized from total mRNA (5 μ g) with specific reverse transcription primers. The expression of miRNAs was quantified using a miRNA-specific TaqMan miRNA Assay Kit (Applied Biosystems, Foster City, CA, USA). qPCR was performed with a FastStart Universal SYBR Green Master (Roche Diagnostics, Mannheim, Germany). Quantitative PCR reactions utilized TaqMan    was continued in fresh medium. Images of the cells were obtained with an inverted phase contrast microscope (Olympus, Tokyo, Japan) at 0 h and 24 h after scratching. The migration capacity was calculated as the percentage of wound closure with the initial wound width defined as 100%.

Patients and Tissue Samples. A prospective case-control study was conducted between
Tube formation assay. The formation of HUVECs into capillary-like structures was conducted as described previously 30 . At least 30 min before the experiment, 96-well plates were coated with Matrigel (BD, Biosciences, Bedford, MA, USA). HUVECs (5 × 10 3 cells/well) were cultured in Ham's F12K basal medium (Gibco, Carlsbad, CA, USA) with 200 ml tumor cell-conditioned medium. Tumor cell-conditioned medium was prepared from 1 × 10 6 JEG-3 cells transfected with a miR-16 mimic or inhibitor, which were cultured in the same conditions for  48 h. Images showing the formation of capillary-like structures were obtained after 12 h with an inverted microscope (Olympus, Tokyo, Japan) at 50 × magnification. Tubular structures were assessed by measuring the total tubule length in each field of view using Image-Pro Plus software.
Animals. Wild-type C57BL/6 mice (8 to 12 weeks old) were maintained on an autoclaved chow diet in filter-topped cages in a specific pathogen-free (SPF) animal room (controlled temperature (24 ± 2 °C), with approximately 40% humidity and a 12/12 h light/dark cycle. The presence of a vaginal plug was designated as day 0.5 of pregnancy. To deliver cholesterol-conjugated miRNAs (Austin, TX, USA), 10 nmol miRNAs in 50 μl saline buffer was locally injected into the placenta using apercutaneous ultrasound-guided approach on day 7.5 of pregnancy. All placentas in a single litter received the same miRNA (control or miR-16 mimic/inhibitor) injection. Pregnant females were killed on day 14.5 of pregnancy, and placenta samples were collected. All animal experiments were approved by the Ethics Review Board of the Ningxia Medical University and conducted following institutional guidelines.
Western blot assays. Protein extraction and western blotting were performed as previously described in more detail 32 . Briefly, proteins (50 μ g) were separated by 10% sodium dodecyl sulphate-polyacrylamide gel (SDS-PAGE) and transferred to a nitrocellulose membrane (Amersham Bioscience, Buckinghamshire, U.K.). Membranes were blotted with an anti-VEGF antibody (sc-152, 1/1000 dilution, Santa Cruz Biotechnology, Santa Cruz, CA) overnight at 4 °C and then incubated with horseradish peroxidase-conjugated secondary antibody (Transduction Laboratories, Lexington, UK). GAPDH antibody (1/1000 dilution, Cell Signaling Technology, Boston, MA, USA) was used as a control. The bands were detected with Enhanced Chemiluminescence (Amersham Biosciences, Piscataway, NJ) and visualized with the ChemiDoc XRS System (Bio-Rad, Hercules, CA, USA).
Luciferase reporter assays. The cells (HTR-8/SVNEO and JEG-3) were grown to 70-80% confluence in 24-well plates and co-transfected with the luciferase reporter vector described above (200 ng) and the appropriate miRNA (50 nM) using Lipofectamine TM 2000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. After 48 h, cells were washed and lysed with passive lysis buffer (Promega, Madison, WI, USA), and luciferase activity was analyzed using dual luciferase assays (Promega, Madison, WI, USA). The relative reporter activity was obtained by normalization to the Renilla luciferase activity. Statistical analysis. SPSS13.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analyses.
Values are expressed as the mean ± standard deviation (SD). Significant differences between groups were analyzed using Student's t-test. Differences were considered statistically significant at P < 0.05.