miRNA‐210‐3p regulates trophoblast proliferation and invasiveness through fibroblast growth factor 1 in selective intrauterine growth restriction

Abstract Selective intrauterine growth restriction (sIUGR), which affects approximately 10%‐15% of monochorionic (MC) twin pregnancies, is highly associated with intrauterine foetal death and neurological impairment in both twins. Data suggest that unequal sharing of the single placenta is the main contributor to birth weight discordance. While MC twins and their placenta derive from a single zygote and harbour almost identical genetic material, the underlying mechanisms of phenotypic discrepancies in MC twins remain unclear. MicroRNAs are small non‐coding RNA molecules that regulate gene expression but do not change the DNA sequence. Our preliminary study showed that microRNA‐210‐3p (miR‐210‐3p) was significantly upregulated in the placental share of the smaller sIUGR twin. Here, we investigate the potential role of miR‐210‐3p in placental dysplasia, which generally results from dysfunction of trophoblast cells. Functional analysis revealed that miR‐210‐3p, induced by hypoxia‐inducible factor 1α (HIF1α) under hypoxic conditions, suppressed the proliferation and invasiveness of trophoblast cell lines. Further RNA sequencing analysis and luciferase reporter assays were performed, revealing that fibroblast growth factor 1 (FGF1) is an influential target gene of miR‐210‐3p. Moreover, correlations among miR‐210‐3p levels, HIF1α and FGF1 expression and the smaller placental share were validated in sIUGR specimens. These findings suggest that upregulation of miR‐210‐3p may contribute to impaired placentation of the smaller twin by decreasing FGF1 expression in sIUGR.


| INTRODUC TI ON
Twin pregnancies carry a higher risk of maternal and foetal complications than singleton pregnancies. Approximately 30% of twin pregnancies are monochorionic (MC), 1 in which the foetuses derive from one fertilized zygote and share a single placenta. Selective intrauterine growth restriction (sIUGR), affecting approximately 10%-15% of MC twin pregnancies, is defined by an estimated foetal weight (EFW) of one foetus below the 10th percentile with or without a difference in EFW of greater than 25% between the foetuses 1,2 and is significantly associated with intrauterine foetal death and neurological impairment of both twins. 2,4,5 Accumulating evidence has revealed that unequal sharing of the single placenta is the main contributor to the birthweight discordance between MC twins. 6,7 We recently reported a higher rate of smaller placental volumes and less vessel branching in the smaller of the sIUGR twins. 8 Inadequate placental share and decreased perfusion lead to limitations in oxygen and nutrient transport to the foetus, resulting in intrauterine growth restriction in sIUGR pregnancies. However, the molecular mechanism underlying the pathophysiological processes of sIUGR has not been fully elucidated.
Monochorionic twins are fundamentally monozygotic, harbouring almost identical genetic material and sharing a similar intrauterine environment. Recently, epigenetic mechanisms have been widely studied in embryonic and placental development. 9 Epigenetics, which includes microRNAs (miRNAs), is the study of heritable changes in gene function that do not entail changes in the DNA sequence. MicroRNAs are short endogenous oligonucleotides (18-22 bp) that post-transcriptionally regulate gene expression in many biological processes by repressing mRNA translation or accelerating mRNA degradation. 10,11 More than 500 different miR-NAs are expressed in the human placenta, and some are involved in various pregnancy-related disorders. 12,13 In the first stage of pregnancy, trophoblast cells invade the maternal endometrium and spiral arteries to establish the maternal-foetal interaction. 15 This invasion process is essential for successful embryonic implantation, placentation and uteroplacental vasculature establishment. 16 Accumulating evidence has shown that aberrant miRNA expression can impair trophoblast function. However, the majority of studies are performed with singletons, in which the regulatory role of miRNAs might be affected by differences in genetic backgrounds among individuals. Thus, MC twins are considered an optimal model for investigating epigenome-mediated molecular mechanisms by using co-twins as internal controls.
In our preliminary study, we compared the miRNAs in the placental share between the smaller and larger twins of sIUGR pregnancies using RNA sequencing (RNA-seq) technology. Fifteen miRNAs were identified to be significantly upregulated in the placental share of the smaller twins relative to their expression in that of the larger twins. Among these miRNAs, microRNA-210-3p (miR-210-3p) was one of the most highly expressed in the smaller twins. MiR-210 is an important hypoxia-inducible factor (HIF) in many human diseases. 17 Hypoxia-inducible factor, a heterodimer consisting of an oxygen-sensitive subunit α and a constitutively active subunit β, mediates cellular responses to hypoxia by regulating transcriptional activation of a multitude of genes. 18,19 Under hypoxic conditions, hypoxia-inducible factor 1α (HIF1α) binds to the hypoxia-responsive element (HRE) in the miR-210 promoter and induces miR-210 expression. Previously, our group found that the level of HIF1α mRNA was significantly higher in the placental shares of sIUGR twins, especially of the growth-restricted foetuses, than in the placentas of non-sIUGR twins. 20 We propose that increased miR-210-3p expression is correlated with the relatively higher level of HIF1α in the sIUGR placenta, contributing to the growth discordance between monozygotic twins.
Fibroblast growth factor 1 (FGF1) is a member of the FGF family that exhibits various biological activities by controlling fundamental cellular functions, including proliferation, differentiation, adhesion and morphogenesis. 21 Studies have reported that FGF1, a ubiquitously expressed factor in different cell types, is involved in angiogenesis, wound healing, myocardial morphogenesis and some tumours. 22,23 During embryogenesis, FGFs play major roles in implantation, gastrulation, mesoderm induction and neural, circulatory and skeletal system formation. 25,26 However, few studies have investigated the pathological role of FGF1 in the setting of placental insufficiency and foetal growth restriction. In this study, we examined miR-210-3p expression in the placental shares of MC twins and investigated the potential role of miR-210-3p in the pathogenesis of sIUGR. Our results provide novel evidence that FGF1 is a direct target of hypoxia-induced miR-210-3p and is involved in the regulation of HTR-8/SVneo cell invasion and proliferation.

| Tissue specimens
This study was approved and supervized by the ethics committee of the First Affiliated Hospital of Sun Yat-Sen University (Guangzhou, China), and informed consent was obtained from all study participants. Placentas were obtained immediately after delivery by elective caesarean section from monochorionic diamniotic (MCDA) twin pregnancies (sIUGR group and control group, n = 18 per group) delivered in this hospital. After the maternal and foetal membranes were removed, placental tissues around the umbilical cord insertion of each twin were collected. Villous portions were harvested and washed in ice-cold saline (0.9% NaCl) three times to remove blood.
Samples for RNA extraction were transferred into sterile tubes and stored at -80°C.
All twin pairs included were diagnosed as MCDA by pre-natal ultrasound according to the American College of Obstetricians and Gynecologists guidelines. Monochorionic diamniotic twin pregnancies complicated by an EFW of one foetus below the 10th percentile and a growth discordance greater than 25% were recruited to the sIUGR group. Growth discordance was calculated as follows: ([weight of the larger twin − weight of the smaller twin]/weight of the larger twin). Normal MCDA twin pregnancies without sIUGR or other complications were included in the control group. Pregnancies complicated by twin-to-twin transfusion syndrome (TTTS), intrauterine foetal death, congenital anomalies, chromosomal abnormalities and maternal complications (such as hypertension, diabetes mellitus, cardiovascular disease, autoimmune disease or malignancies) were excluded.

| RNA extraction, reverse transcription and quantitative real-time PCR (qRT-PCR)
Total RNA was extracted from placental tissues or cultured cells using TRIzol (Invitrogen) following the manufacturer's instructions.
Then, relative protein levels were quantified by scanning densitometry, and the relative gray value of the protein bands was calculated by using α-tubulin as the internal control.

| Cell invasion assay
Cell invasion was assessed using a Transwell chamber assay (Corning Inc, New York, NY) with Matrigel-coated (BD Biosciences, Bedford, MA) membranes. First, cells were suspended in medium and seeded in the upper chambers of 24-well Transwell plates (1:3 dilution in medium, 50 μL/well) with a Matrigel-coated membrane. Then, RPMI 1640 medium (500 mL) containing 10% FBS was added to the lower chamber of the Transwell plates as a chemoattractant. After incubation for 24 hours (37°C, 5% CO 2 ), the upper surface of the insert was gently wiped with a cotton swab to remove non-invading cells. Cells invading into the lower chamber were fixed, stained, photographed and quantified by counting in five random 100× magnification fields.

| Luciferase reporter assay
For luciferase reporter assays, 2 × 10 4 cells were seeded in triplicate in 48-well plates and allowed to adhere for 24 hours. The indicated plasmids (wild-type or mutant) plus 1 ng of pRL-TK Renilla plasmid were cotransfected into the corresponding lentivirus-infected cells using Lipofectamine 3000 reagent (Life Technologies). Forty-eight hours after transfection, a dual luciferase reporter assay (Promega, Madison, WI) was performed according to the manufacturer's instructions.

| RNA-seq and bioinformatic analysis
Samples of total RNA from the placental tissues of six pairs of sIUGR twins were extracted for miRNA sequencing. MiRNA sequencing was carried out commercially by Annoroad Gene Technology Corporation following standard protocols. Briefly, small RNAs (18-30 nt) were separated and enriched using a 15% PAGE gel. Libraries were prepared using a TruSeq Small RNA Sample Preparation Kit and sequenced on an Illumina HiSeq 2500 (50SE). Clean reads were mapped to the human genome using miRDeep2, and heat map analysis was performed with the R programming language.
Samples of total RNA from HTR-8/SVneo cells overexpressing miR-210 or the corresponding vector control were extracted for mRNA sequencing. mRNA sequencing was carried out commercially by Vazyme Biotech Corporation following standard Illumina HiSeq protocols. Clean reads were mapped to the human genome using Hista2, and further analysis, including the generation of volcano plots and heat maps, was performed with the R programming language.

| Statistical analysis
Statistical analyses were carried out using spss version 21 (IBM Corporation, Armonk, NY) and GraphPad Prism 6. Data are expressed as the means ± SDs of at least three independent experiments. The Mann-Whitney test or Student's t test was used to compare two independent samples. One-way ANOVA and least significant difference (LSD) multiple comparison tests were used to compare the differences among groups. Spearman's correlation coefficients were used to examine the correlation between miR-210-3p expression and HIF1α and FGF1 expression. A p-value of <0.05 was considered to indicate statistical significance.

| MiR-210-3p expression is elevated in the placental share of the smaller foetus in sIUGR pregnancies
To screen for dysregulated miRNAs involved in sIUGR, we first  Table 1. No significant difference was found in maternal age between the sIUGR and control groups. The birth weights of both the smaller and larger foetuses were significantly lower in the sIUGR group than in the control group, and the birth weight discordance was significantly greater in the sIUGR group than in the control group (P < 0.01). Lower birth weight was closely associated with an earlier gestational age at delivery in the sIUGR group (average, 32 weeks) relative to that in the normal group (average, 35 weeks).
The placental share of the smaller sIUGR twins (29.7%) was significantly less than that of the smaller twins in the control group (45.0%) (P < 0.01).
Then, we examined whether the upregulation of miR-210-3p in the smaller twins was clinically relevant in sIUGR cases. A classification system based on the Doppler characteristics of the umbilical artery (UA) in the smaller sIUGR twin correlates well with distinct clinical evolution and outcomes. 29 Monochorionic twins in the sIUGR group were classified into three types: type I, normal UA Doppler; type II, persistent absent or reversed end diastolic velocity flow (AREDF); or type III, intermittent AREDF ( Table 1). The mean gestational age at delivery was earlier for type II and III twins than for type I twins. The placental share of the smaller type II (23.4%) and type III (28.4%) twins was smaller than that of the smaller type I twins (36.3%) (P < 0.01). Moreover, qRT-PCR revealed that the mean fold change in miR-210-3p expression in type II and III twins was relatively higher than that in type I twins ( Figure 1C).

| Identification of FGF1 as a target of miR-210-3p in human trophoblast cells
Because miRNAs regulate cellular processes by repressing target gene expression through interaction with the 3′-UTR of mRNAs, we next ex-   Therefore, we constructed luciferase reporter plasmids containing FGF1 3′-UTR sequences with the wild-type or corresponding mutated miR-210-3p binding site. The reporter assays showed that compared to the relative luciferase activity in control cells, the relative luciferase activity from the wild-type FGF1 3′-UTR was significantly decreased in miR-210-3p-overexpressing cells and increased in miR-210-3p-inhibited ( Figure 3F). However, when the predicted miR-210-3p binding sequence in the FGF1 3′-UTR was mutated, the luciferase activity was no longer affected by miR-210-3p ( Figure 3F). Collectively, these results validated that miR-210-3p directly suppressed FGF1 expression in HTR-8/SVneo trophoblast cell lines.
However, silencing FGF1 with two effective siRNAs in cells transfected with the miR-210-3p sponge partially reversed the increases in proliferation and invasiveness, further supporting the possibility that FGF1 is a direct functional target of miR-210-3p ( Figure 4B,D).
Taken together, these results demonstrated that FGF1 downregulation was functionally important for the inhibitory effect of miR-210-3p on HTR-8/SVneo cell proliferation and invasiveness.

| Correlation of miR-210-3p expression with HIF1α and FGF1 expression in placental tissues
As the in vitro experiments revealed that miR-210-3p expression was induced under hypoxia through HIF1α and inhibited FGF1 expression, we further examined the correlation between miR-210-3p expression and the protein levels of both FGF1 and HIF1α in the placental tissues of the smaller twins paired with those of the larger twins in sIUGR cases. As shown in Figure 6(A,B), significantly increased levels of HIF1α and miR-210-3p, and decreased levels of FGF1 were found in the smaller twins relative to those levels in the larger twins. The linear regression and correlation analyses showed that in the placental tissues, HIF1α expression was positively correlated with miR-210-3p expression, while FGF1 expression was negatively correlated with miR-210-3p expression, suggesting that the HIF1α/miR-210-3p/FGF1 axis is clinically relevant in sIUGR ( Figure 6C).

| D ISCUSS I ON
As a pre-requisite for the survival and development of the foetus, the placenta plays a critical role in the exchange of oxygen, nutrients and waste products between the mother and the foetus. Unequal division of the placenta is the most important cause of discordant growth in MC twins. 30 Typically, the smaller twin has a lower placental share, which reflects limited interaction with the maternal circulation and restriction of the supply of oxygen and nutrients.
We demonstrated that miR-210-3p expression was significantly upregulated in the placental share of the smaller foetuses and that the increased expression of miR-210-3p was strongly correlated with insufficient placentation by inhibiting the proliferation and invasive of trophoblast cells. In summary, our data show that miR-210-3p is upregulated by hypoxia and disrupts the proliferation and invasion of trophoblasts and identify for the first time that FGF1 is an influential target of miR-210-3p. Moreover, we showed that FGF1 expression was negatively correlated with miR-210-3p expression in the placental share of the smaller twin in sIUGR pregnancies. These results demonstrate that the interaction between miR-210-3p and FGF1 is clinically relevant in these MC twins, who harbour an almost identical genetic background. Our findings not only indicate that miR-210-3p plays important regulatory roles in placentation and foetal development but also provide a novel epigenetic mechanism through which miR-210-3p may be associated with phenotypic discrepancies in monozygotic twins.

ACK N OWLED G EM ENTS
This work was supported by a grant from the National Natural Science Foundation of China (No. 81671464).

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
Lin Li and Xuan Huang performed the experiments and drafted the manuscript. Zhiming He contributed to the collection of the samples. Yuanyan Xiong performed RNA-seq and analysed the data.
Qun Fang supervized the experiments and revised the manuscript.
All authors approved the final manuscript.