miR-16 inhibits the proliferation and angiogenesis-regulating potential of mesenchymal stem cells in severe pre-eclampsia

Authors

  • Yaping Wang,

    1. Immunology and Reproductive Biology Laboratory, Medical School & State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, China
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Hongye Fan,

    1. Immunology and Reproductive Biology Laboratory, Medical School & State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, China
    Search for more papers by this author
    • These authors contributed equally to this work.
  • Guangfeng Zhao,

    1. The Affiliated Drum Tower Hospital of Nanjing University Medical School, China
    Search for more papers by this author
  • Dan Liu,

    1. Immunology and Reproductive Biology Laboratory, Medical School & State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, China
    Search for more papers by this author
  • Leilei Du,

    1. Immunology and Reproductive Biology Laboratory, Medical School & State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, China
    Search for more papers by this author
  • Zhiqun Wang,

    1. The Affiliated Drum Tower Hospital of Nanjing University Medical School, China
    Search for more papers by this author
  • Yali Hu,

    Corresponding author
    1. The Affiliated Drum Tower Hospital of Nanjing University Medical School, China
    • Immunology and Reproductive Biology Laboratory, Medical School & State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, China
    Search for more papers by this author
  • Yayi Hou

    Corresponding author
    1. Jiangsu Key Laboratory of Molecular Medicine, Nanjing, China
    • Immunology and Reproductive Biology Laboratory, Medical School & State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, China
    Search for more papers by this author

Correspondence

Y. Hou, Immunology and Reproduction Biology Lab Medical School, Nanjing University, Nanjing 210093, China

Fax: +86 25 83686441

Tel: +86 25 83686441

E-mail: yayihou@nju.edu.cn

Y. Hu, The Affiliated Drum Tower Hospital of Nanjing University Medical School, Nanjing 210008, China

Tel: +86 25 83321980

E-mail: yali_hu@hotmail.com

Abstract

Pre-eclampsia is thought to be a systemic disease of maternal endothelial cell dysfunctions. miRNAs regulate various basic biological functions in cells, including stem cells. Mesenchymal stem cells exist in almost all tissues and are the key cellular source for tissue repair and regeneration. Our aims are to investigate whether miRNAs regulate MSCs in fetal–maternal interfaces to influence the pathogenesis of pre-eclampsia. The differential expression of miRNAs in decidua-derived mesenchymal stem cells of all patients with severe pre-eclampsia (= 20) and normal groups (= 20) was first screened by microarray analysis and validated by quantitative real-time PCR analysis. The integrated bioinformatics analysis showed that miR-16 showed the highest number of connections in the miRNA GO network and the miRNA gene network. Moreover, over-expressed miR-16 inhibited the proliferation and migration of decidua-derived mesenchymal stem cells and induced cell-cycle arrest by targeting cyclin E1. Interestingly, over-expression of miR-16 by decidua-derived mesenchymal stem cells reduced the ability of human umbilical vein endothelial cells to form blood vessels and reduced the migration of trophoblast cells. Furthermore, decidua-derived mesenchymal stem cell-expressed endothelial growth factor VEGF-A was involved in migration of trophoblast cells and human umbilical vein endothelial cells as well as tube and network formation. Importantly, the levels of cyclin E1 and VEGF-A were negatively correlated with the level of miR-16 expression in decidua-derived mesenchymal stem cells from the patients with severe pre-eclampsia. Together, these data suggest that the alteration of miR-16 expression in decidua-derived mesenchymal stem cells may be involved in the development of pre-eclampsia.

Abbreviations
CCNE1/D1

cyclin E1 and D1

CDK2/6

cyclin-dependent kinase 2 and 6

dMSCs

decidua-derived MSCs

HUVEC

human umbilical vein endothelial cells

MSCs

mesenchymal stem cells

pre-miR-16

miR-16 precursor

sPE

severe pre-eclampsia

VEGF-A

endothelial growth factor

Introduction

Pre-eclampsia (PE) is a relatively common pregnancy disorder that may threaten the survival of both mother and baby. Some evidence showed that PE is a systemic vascular disorder that may suggest later cardiovascular disease in the mothers [1]. Previous studies have highlighted the role of a disturbed angiogenic balance as one of the key features of this disease [2]. The endothelial growth factor VEGF-A, an angiogenesis-related factor, plays an important role in angiogenesis, thus we assume that VEGF may participate in the pathological mechanism of PE [3, 4].

Mesenchymal stem cells (MSCs) are capable of self-renewing and have the potential to differentiate into mesenchymal and non-mesenchymal tissues. The cells thus have great potential for treating various diseases [5]. The roles of MSCs in neoangiogenesis and their medical applications have been discussed widely [6-8]. Moreover, it was recently reported that the vascular niche harbors a pool of placenta-derived MSCs that contribute to vessel maturation and stabilization [9]. In addition, decidua-derived MSCs (dMSCs) from patients with PE and those with normal pregnancies showed different expression of cytokines [10]. However, little is known about the potential roles of dMSCs in PE development and progression.

miRNAs are integral elements in post-transcriptional control of gene expression. Many studies have shown that miRNAs regulate various basic biological functions in stem cells [11-14], and aberrant expression of miRNAs is involved in several diseases. Moreover, differential expression of miRNAs has been observed in placentas from patients with PE and those with normal pregnancies [15-18]. miR-210 and miR-155 play important roles in the pathogenesis of PE [19, 20]. Thus, in order to explain the pathogenesis of PE, it is necessary to explore the differential expression and functions of miRNAs in dMSCs.

In the present study, we compared the expression profile of miRNAs in dMSCs from pregnant women with severe PE (sPE) or normal pregnancies. We performed a global analysis of miRNA-related genes and their signaling pathways, and we found that miR-16 showed the highest number of connections in the miRNA GO network and miRNA gene network [20a]. Moreover, we found that high expression of miR-16 prevented the proliferation of dMSCs by inducing cell-cycle arrest, but did not affect apoptosis of dMSCs. Furthermore, over-expression of miR-16 reduced VEGF-A expression in dMSCs by directly targeting its 3′ UTR, which influenced the migration ability of human umbilical vein endothelial cells (HUVEC) and human trophoblast cells (HTR-8/SVneo cells). dMSCs with over-expressed miR-16 also reduced the ability of HUVEC to form blood vessels. All of these data suggest that alteration of miR-16 expression in dMSCs may be involved in the development of PE.

Results

miR-16 shows the highest number of connections in the intersection of miRNA gene network and miRNA GO network

Using the Agilent human miRNA microarray kit version 16.0, consisting of 940 miRNAs probes [20b], corresponding to the sanger center mirbase version 16.0, we assessed the miRNA expression profile of dMSCs from patients with normal pregnancies or sPE. In order to explore the relationship between miRNAs and gene properties, the miRNA GO network was constructed based on the relationship between significant GOs, genes and miRNAs (Fig. 1A). In this network, the degree represents the contribution of an individual miRNA or GO category to adjacent miRNAs or GO categories. In addition, we built an miRNA gene network of the differential expression of miRNAs (Fig. 1B). In this network, the degree represents the contribution of an individual miRNA or gene to adjacent miRNAs or genes. By analyzing the miRNA GO network and the miRNA gene network, we found that miR-16 showed the highest number of connections in the up-regulated miRNAs (Fig. 1C).

Figure 1.

miR-16 shows the intersection with the highest number of connections in the miRNA gene network and miRNA GO network. (A,B) miRNAs at the intersection of differentially expressed miRNAs from the miRNA GO network and miRNA gene network. (C) Among the highly expressed miRNAs, miR-16 shows the highest number of connections in both networks. In these networks, the term ‘degree’ represents the contribution of an individual miRNA to adjacent genes or GO categories.

Up-regulation of miR-16 in sPE dMSCs is validated by quantitative PCR analysis

The expression of aberrant miRNAs was determined by quantitative real-time PCR in dMSCs from 20 patients with sPE and 20 patients with normal pregnancies. We confirmed that miR-16, miR-29b and miR-30a were up-regulated in sPE dMSCs, is consistent with the microarray analysis, but miR-210 and miR-152 showed no obvious changes (Fig. 2).

Figure 2.

Quantitative real-time PCR analysis confirmed that miR-16 expression is up-regulated in dMSCs from patients with sPE. Quantitative real- time PCR was used to analyze the expression of miR-16, miR-29b, miR-30a, miR-210 and miR-152 in dMSCs from patients with sPE. Data indicate relative expression following normalization. Values are means ± SE (*< 0.05).

Over-expression of miR-16 reduces the proliferation of dMSCs but does not induce apoptosis

To explore the role of miR-16 in dMSCs, we examined the effect of over-expressed miR-16 on the viability and proliferation of dMSCs. The cells were transfected with various oligonucleotides miR-16 precursor (pre-miR-16), a precursor negative control (pre-miR control), miR-16 inhibitor (anti-miR-16) and an inhibitor negative control (anti-miR control). Forty-eight hours after transfection, dMSCs over-expressing miR-16 showed a significant reduction in their viability and proliferation activity compared with dMSCs expressing the pre-miR control (Fig. 3C–E). Consistent with these results, dMSCs transfected with anti-miR-16 showed increased viability compared with cells transfected with the anti-miR control. However, there were no obvious changes in the viability and proliferation activity of dMSCs after transfection with pre-miR-16. Over-expression of miR-16 neither caused significant alterations to the apoptosis rate nor changed the expression of phenotypic surface antigens of dMSCs (Fig. S1). MSCs derived from sPE deciduas show low proliferation activity compared with normal dMSCs (Fig. S2).

Figure 3.

Over-expression of miR-16 reduces dMSC proliferation. dMSCs were transfected with control oligonucleotides (pre-miR control or anti-miR control), pre-miR-16 or anti-miR-16. All oligonucleotides (30 pmol) were used for transfection in 12-well plates at a density of 8 × 104 cells per well. dMSCs were collected 24 and 48 h after transfection for the following experiments. (A,B) PCR and quantitative PCR were performed to analyze the expression of miR-16 in dMSCs after transfection for 24 h. (C) Morphology of dMSCs 48 h after transfection (magnification × 20). (D) dMSC proliferation was determined by cell count after transfection for 24, 36, 48 and 72 h. (E) dMSC viability was determined using Cell counting kit-8 (CCK8) after transfection for 24, 36, 48 and 72 h. Data are representative of three independent experiments. Values are means ± SEM. Asterisks indicate significant differences (*< 0.05, ***< 0.001).

Over-expression of miR-16 triggers an accumulation of dMSCs in the G0/G1 phase

To further explore the potential biological function of miR-16 in dMSCs, dMSCs were transfected with pre-miR-16 for 48 h. It was found that over-expression of miR-16 triggered an accumulation of dMSCs in the G0/G1 phase compared with the negative control (pre-miR control), while down-regulating the expression of miR-16 decreased the percentage of dMSCs in G0/G1 phase (Fig. 4).

Figure 4.

miR-16 impairs the proliferation of dMSCs by triggering an accumulation of cells in G0/G1 phase. dMSCs were transfected with control oligonucleotides (pre-miR control or anti-miR control), pre-miR-16 or anti-miR-16. All oligonucleotides (30 pmol) were used for transfection in 12-well plates at a density of 8 × 104 cells per well. dMSCs were collected for cell-cycle analysis 48 h after transfection. Transfection with pre-miR-16 triggered accumulation of cells in G0/G1 phase compared with the pre-miR control. Consistently, down-regulating the expression of miR-16 decreased the percentage of cells in G0/G1 phase. Data are representative of three independent experiments. Values are means ± SEM. Asterisks indicate significant differences (*< 0.05, **< 0.01).

miR-16-over-expressing dMSCs regulate angiogenesis

PE may be due to inappropriate modification of spiral arteries and shallow invasion of extravillous trophoblasts into the uterus [20c]. We determined whether dMSCs over-expressing miR-16 impair the migration capacity of HTR-8/SVneo cells and HUVEC, as well as capillary tube and network formation, by a Matrigel assay. dMSCs were transfected with various oligonucleotides (miR-16 precursor, pre-miR control, miR-16 inhibitor and anti-miR control) for 36 h, and culture supernatants (free of fetal bovine serum) were collected from various groups of dMSCs. HTR-8/SVneo cells and HUVEC with various densities were put into a 24-well culture plate containing 500 μl medium with 10% fetal bovine serum. After 8 h incubation, migration of HTR-8/SVneo cells and HUVEC was detected. The results showed that dMSCs transfected with the pre-miR-16 oligonucleotide significantly inhibited migration of HTR-8/SVneo cells and HUVEC compared with those in the pre-miR control group, while dMSCs transfected with the anti-miR-16 oligonucleotide significantly enhanced migration of HTR-8/SVneo cells and HUVEC compared with those in the anti-miR control group (Fig. 5A,B). Moreover, a tube formation assay was used to assess the effect of dMSCs over-expressing miR-16 on the angiogenesis process. Various densities of HUVEC were added to Matrigel, and it was found that supernatants from dMSCs over-expressing miR-16 could reduce the tube network, while supernatants from the cells in which miR-16 expression was down-regulated resulted in longer tubes (Fig. 5C). Together, these results suggest that dMSCs over-expressing miR-16 may alter some factors of the pregnancy microenvironment, influencing sPE-associated angiogenesis processes.

Figure 5.

dMSCs over-expressing miR-16 play an important role in regulating angiogenesis. We transfected dMSCs with control oligonucleotides, pre-miR-16 and anti-miR-16 for 36 h, and collected culture supernatants (free of fetal bovine serum) for the following experiments. (A,B) After incubation of various densities of cells for 8 h, the migration of HTR-8/SVneo cells and HUVEC was significantly inhibited by supernatants from dMSCs transfected with pre-miR-16 oligonucleotide, compared with the pre-miR control group, and their migration was enhanced by the supernatant from transfection with anti-miR-16 oligonucleotide compared with the anti-miR control group. (C) After 12 h incubation, supernatants of dMSCs showing high expression of miR-16 had an inhibitory effect on the tube network, while supernatants in which expression of miR-16 was down-regulated resulted in longer tubes. Representative images from three independent experiments are shown on the left, and a statistical analysis is shown on the right. Values are means ± SEM. Asterisks indicate significant differences (*< 0.05, **< 0.01).

CCNE1, which is the target of miR-16, is crucial to trigger the accumulation of dMSCs in G0/G1 phase

Cyclin E1 (CCNE1), cyclin D1 (CCND1), cyclin-dependent kinase 2 (CDK2) and CDK6 play important roles in accumulation of dMSCs in G0/G1 phase. We found that over-expression of miR-16 reduced the protein level of CCNE1, but did not change the levels of CCND1, CDK2 and CDK6 in dMSCs 48 h after transfection (Fig. 6A). Interestingly, CCNE1 was predicted to be a putative target of miR-16 by the miRNA gene network analysis and other target prediction programs [miRanda (http://diana.cslab.ece.ntua.gr/microT/), TargetScan (http://www.targetscan.org/) and PicTar (http://pictar.mdc-berlin.de/) algorithms). To test whether CCNE1 is the direct target of miR-16, the 3'-UTRs (untranslated regions) of CCNE1 mRNA, which contain the target sites for miR-16, were PCR-amplified and then introduced downstream of the luciferase reporter gene in the XbaI-cloning sites of the pGL3 control vector. Co-transfection experiments showed that, compared with the negative control, there was a significant reduction in relative luciferase activity in cells transfected with pre-miR-16, while the relative luciferase activity was increased in cells transfected with anti-miR-16 (Fig. 6B).

Figure 6.

miR-16-targeted CCNE1 is crucial for triggering the accumulation of G0/G1 phase dMSCs. (A) Western blot analysis of the levels of CDK6, CDK2, CCND1 and CCNE1 after treatment with control oligonucleotide (pre-miR control or anti-miR control), pre-miR-16 and anti-miR-16 for 48 h. (B) miR-16 down-regulated CCNE1 expression by directly targeting its 3′ UTR. This confirms the interaction between miR-16 and the human CCNE1 3′ UTR. (C) Western analysis of expression of CCNE1 in dMSCs treated with small interfering (si)RNA control oligonucleotide, si-CCNE1-1 or si-CCNE1-2. (D) The proliferation of dMSCs was determined using CCK8 48 h after transfection. (E,F) Transfection with si-CCNE1 oligonucleotides triggered accumulation of cells in G0/G1 phase compared with the negative control, while down-regulation of CCNE1 decreased the percentage of cells in S phase. Data are representative of three independent experiments. Values are means ± SEM. Asterisks indicate significant differences (*< 0.05, **< 0.01, ***< 0.01).

CCNE1 was reported to play a role in G0/G1 distribution [21]. To test whether miR-16 targets CCNE1 to trigger accumulation of dMSCs in G0/G1 phase, CCNE1 expression was down-regulated by siRNA. The results showed that both the mRNA and protein levels of CCNE1 were efficiently repressed, thereby triggering accumulation of dMSCs in G0/G1 phase (Fig. 6C–F).

dMSC-expressed VEGF-A is involved in the migration of HTR-8/SVneo cells and HUVEC, as well as tube and network formation

VEGF-A is one of the most common promoters of angiogenesis. It is also a putative target of miR-16 according to miRNA gene network analysis and other target prediction programs (miRanda, TargetScan and PicTar algorithms). Over-expression of miR-16 reduced the protein levels of VEGF-A in dMSCs 36 h after transfection (Fig. 7A,B). Reporter assays confirmed that miR-16 represses VEGF-A expression through its 3′ UTR in dMSCs (Fig. 7C). Strikingly, when the VEGF-A-neutralizing antibody Avastin (Roche, Basel, Switzerland) [22, 23] was used to neutralize VEGF-A in dMSC culture supernatants, migration of HTR-8/SVneo cells and HUVEC and tube formation were all markedly inhibited (Fig. 7D–F).

Figure 7.

dMSC-expressed VEGF-A is essential for migration of HTR-8/SVneo cells and HUVEC as well as tube and network formation. (A,B) Western blot and ELISA analysis of the levels of VEGF-A after treatment with control oligonucleotide (pre-miR control or anti-miR control), pre-miR-16 or anti-miR-16 for 36 h. (C) miR-16 down-regulates VEGF-A expression by directly targeting its 3′ UTR. This confirms the predicted interaction between miR-16 and human VEGF-A 3′ UTR, as determined by the database Targetscan. (D,E) VEGF-A neutralizing antibody (Avastin-beva) was used to neutralize VEGF-A in the culture supernatant of dMSCs, and tube formation and cell migration were observed in vitro. After neutralizing VEGF-A in the culture supernatant of dMSCs, the number of migrating HTR-8/SVneo cells and HUVEC obviously decreased. (F) The supernatant of dMSCs treated with Avastin inhibited tube and network formation of HUVEC. Representative images from three independent experiments are shown on the left panels, and a statistical analysis is shown on the right. Values are means ± SEM. Asterisks indicate significant differences (*< 0.05, **< 0.01, ***< 0.001).

miR-16 is negatively correlated with CCNE1 and VEGF-A in dMSCs from patients with sPE

As miR-16 was not only up-regulated in dMSCs from patients with sPE but also targeted CCNE1 and VEGF-A, we anticipated that CCNE1 and VEGF-A levels would be correlated with miR-16 expression in dMSCs from patients with sPE. Indeed, the results revealed that the mRNA levels of CCNE1 and VEGF-A were significantly decreased in dMSCs from patients with sPE compared with those with normal pregnancies (P < 0.05; Fig. 8A). Moreover, the levels of CCNE1 and VEGF-A were negatively correlated with the level of miR-16 expression in dMSCs from patients with sPE (Fig. 8B).

Figure 8.

miR-16 is negatively correlated with CCNE1 and VEGF-A in dMSCs from sPE. (A,B) Quantitative RT-PCR was used to analyze the mRNA expression for CCNE1 and VEGF-A in dMSCs from patients with sPE (= 18) and normal pregnancies (= 20). (C,D) Pearson's correlation scatter plot of the expression of miR-16 and its target mRNAs (CCNE1 and VEGF-A) in dMSCs from decidua of patients with severe pre-eclampsia (= 19) (*< 0.05).

Discussion

MSCs exist in almost all tissues [24]. It has been reported that miRNAs regulate most functions of MSCs [24a]. miR-204 has been shown to influence the differentiation of MSCs, and miR-21 plays an important role in their adipogenic differentiation [13, 14]. PE is a systemic disease including the dysfunction of maternal endothelial cell [25]. The decidua is the highly modified endometrium whose proper formation and function are required for normal pregnancy [26]. There are several indications that the pathophysiology of PE arises in the fetal–maternal interface [27, 28], our results showed that abnormal dMSCs may be involved in the development of PE. Differential expression of cytokines has been observed between dMSCs from patients with PE and those with normal pregnancies [10]. Moreover, differential expression of miRNAs has also been observed between the placenta of patients with PE and those with normal pregnancies [18]. In the present study, we profiled the expression of miRNAs in dMSCs from patients with sPE or normal pregnancies, and found abnormal expression of miRNAs in dMSCs from patients with PE. In addition, unpublished research in our laboratory has shown that MSCs ameliorate pre-eclampsia symptoms in mice, and alterations of expression of some miRNAs in MSCs suggests that the MSCs lost the ability to ameliorate such symptoms [29]. Together, these results suggest that differential expression of miRNAs in MSCs is involved in the development of PE.

miR-16 showed the highest number of connections in both the miRNA gene network and the miRNA GO network according to microarray analysis. miR-16 was not only up-regulated in dMSCs from sPE patients, but was also high-expressed in placenta tissue of sPE patients [18]. We found that dMSCs derived from sPE show low proliferation activity. Moreover, we also found that miR-16 inhibited the proliferation of dMSCs and triggered an accumulation of cells in G0/G1 phase. It has previously been reported that miR-16 regulates the proliferation of cells [30, 31] and controls cell-cycle progression by inducing cell-cycle arrest in G0/G1 phase [32, 33]. Liu et al. reported that the miR-16 induces cell-cycle arrest by regulating multiple cell-cycle genes [21]. We detected expression of CCNE1, CCND1, CDK2 and CDK6 in dMSCs. These proteins regulate the process of G1–S transition. However, only expression of CCNE1 was repressed by miR-16. This suggests that miR-16 may inhibit the proliferation of dMSCs and trigger accumulation of cells in the G0/G1 phase via targeting the cell-cycle gene CCNE1.

dMSC-expressed VEGF-A is involved in the migration of trophoblast cells and tube and network formation. VEGF-A is one of the most common promoters of angiogenesis. It is also one of the putative targets of miR-16 according to miRNA gene network analysis and other target prediction programs (miRanda, TargetScan and PicTar algorithms). We found that secretion of VEGF-A by dMSCs was inhibited by high expression of miR-16. In order to eliminate the effect of the number of cells, we transfected dMSCs with oligonucleotides for 36 h, at which time there were no obvious changes in cell numbers or vitality. We confirmed that miR-16 regulates the angiogenesis-regulating potential of dMSCs, and that VEGF-A is a crucial factor responsible for cell migration and tube and network formation [34]. Patients with PE show dysfunction of the endothelium or shallow invasion of trophoblast cells into the uterus [35]. Our results show that the levels of CCNE1 and VEGF-A are negatively correlated with the level of miR-16 expression in dMSCs from the same patients with sPE. Thus the alteration of miR-16 expression in dMSCs may be involved in the development of PE.

In conclusion, differential expression of miRNAs occurs in dMSCs from patients with PE. miR-16 showed the highest number of connections in the miRNA GO network and the miRNA gene network. Over-expressed miR-16 inhibited both the proliferation and migration of dMSCs, and induced cell-cycle arrest by targeting CCNE1. Moreover, over-expression of miR-16 by dMSCs reduced the ability of HUVEC to form blood vessels and the migration of the trophoblast cells, and this may be attributed to the decreased levels of VEGF-A produced by dMSCs. Importantly, the levels of CCNE1 and VEGF-A were negatively correlated with the level of miR-16 expression in dMSCs from patients with severe PE. This suggests that the alteration of miR-16 expression in dMSCs may be involved in the development of PE. This work suggests the existence of a new molecular event that may explain the pathogenesis of PE.

Experimental procedures

Decidua collection

Human decidua tissue from PE patients and age-matched normotensive controls was collected in a sterile manner at the time of delivery by cesarean section at the Department of Gynecology and Obstetrics of the Affiliated Drum Tower Hospital of Nanjing University Medical School. The hospital ethics committee approved the consent forms and the protocol for evaluation of the tissue. Written consent was received from the women prior to surgery. Decidua tissue was obtained from 20 patients with sPE and 20 patients with normal pregnancies. The relevant clinical details for the patients are shown in Table 1. PE was defined as the presence of hypertension and proteinuria beyond the 20th week of pregnancy. Blood pressure elevation with systolic blood pressure > 140 mmHg or diastolic pressure > 90 mmHg was considered hypertensive. sPE was defined as either severe hypertension (maternal blood pressure > 160/110 mmHg at two separate readings at least 6 h apart) or severe proteinuria (> 2 g protein over a 24 h period). Any complications of pregnancy such as multiple pregnancies such as twins, fetal structural or genetic anomalies, and the presence of maternal chronic hypertension, hemolysis, elevated liver enzymes, the HELLP syndrome, cardiovascular disease, renal disease, hepatic disease, diabetes or other infectious disease were criteria for exclusion.

Table 1. Clinical characteristic of the patients with a normal pregnancy or severe pre-eclampsia. Values are means ± SEM
VariableNormal pregnancy (= 20)Severe pre-eclampsia (= 20)P value
Maternal age (years)30.50 ± 0.758330.81 ± 0.73720.7694
Gestational age at delivery (weeks)38.06 ± 0.664137.89 ± 0.60980.8544
Birth weight (kg)3.150 ± 0.47242.017 ± 0.34990.0623
Systolic blood pressure (mmHg)112.3 ± 2.356160.5 ± 4.797< 0.001
Diastolic blood pressure (mmHg)80.50 ± 1.910110.3 ± 1.852< 0.001
Proteinuria (mg per 24 h)02108.5 ± 22.30< 0.001

Isolation and culture of MSCs from deciduas

The decidua tissues were cut into 1–2 mm3 fragments after washing with phosphate-buffered saline (NaCl/Pi) several times to remove excess blood. Then they were incubated in an enzyme cocktail (5 U·mL−1 hyaluronidase, 125 U·mL−1 collagenase and 50 U·mL−1 dispase; Sigma, St Louis, MO) for 90–120 min with gentle agitation at 37 °C. This tissue was then crushed with forceps to release individual cells, and large pieces of tissue were removed. The cells were pelleted by centrifugation at 250 g for 5 min, resuspended in fresh medium containing Dulbecco's modified Eagle's medium/F12 (Gibco, Grand Island, NY, USA) and 20% fetal bovine serum and transferred to six well plates. Cells were incubated at 37 °C in an incubator with 5% CO2 at saturating humidity. When cells reached 70–80% confluence or when numerous colonies were observed, the cells were detached using 0.25% trypsin/EDTA (Invitrogen, Carlsbad, CA, USA), and the trypsin was inactivated using Dulbecco's modified Eagle's medium/F12. The culture medium was replaced every 3 or 4 days. After the 2nd to 4th cell passages, the adherent cells were fibroblast-like, with phenotypic surface antigens of stem cells such as CD105+, CD73+, CD90+, HLA-ABC+, CD29+, CD44+, CD106−, HLA-DR−, CD19−, CD11b−, CD14−, CD34−, CD31− and CD45−.

Cell line and culture

HTR-8/SVneo, an immortalized human trophoblast cell line, established from first-trimester human cytotrophoblast cells, was kindly provided by Charles H. Graham (Queen's University, Kingston, ON, Canada). HTR-8/SVneo cells were cultured in RPMI-1640 medium (Gibco) supplemented with 10% fetal bovine serum, 100 IU·mL−1 penicillin and 100 g·mL−1 of streptomycin, and were incubated at 37 °C in a humidified atmosphere with 5% CO2. Human umbilical vein endothelial cells (HUVEC) were obtained from Shanghai Institute of Cell Biology (Shanghai, China). HUVEC were grown in RPMI-1640 medium supplemented with 10% fetal bovine serum, and were incubated in a humidified atmosphere with 5% CO2 at 37 °C.

miRNA microarray analysis, and miRNA gene network and miRNA GO network analysis

Ten samples of dMSCs, five from patients with normal pregnancies (control group) and five from patients with sPE (matched for mother's age and gestational age at delivery) were assayed using the Agilent Technologies (Santa Clara, CA, USA) Human miRNA microarray kit version 16.0 (CapitalBio Corporation, Beijing, China). For each miRNA, multiple probes were spotted on the array, and the mean intensity of these probes was calculated to represent the expression value of the miRNAs. In addition, multiple spots were included as negative controls. For each sample, 100 ng total RNA were hybridized with the miRNAs array and further processed in accordance with the manufacturer's instructions. The scanned images were processed using the Sanger Center miRBase version 16.0. The miRNA gene network was constructed based on the interactions of miRNAs and genes in Sanger miRNA database. The miRNA GO network was constructed based on the relationships of significant GO categories and genes/miRNAs.

Quantitative RT-PCR analysis

Total RNA, including miRNA, was extracted using Trizol reagent (Invitrogen) according to the manufacturer's instructions. The concentration of RNA was measured using a SmartSpec™ Plus spectrophotometer (Bio-Rad, Hercules, CA, USA), and the RNA quality was confirmed by agarose gel electrophoresis.

Total RNA was purified using an mirVana miRNA isolation kit (Ambion, Austin, TX, USA) to enrich the small RNA fraction. RNA (10 ng) was reverse-transcribed using MultiScribe reverse transcriptase, reverse transcriptase buffer, dNTPs, RNase inhibitor and miR-specific primers in the GeneAmp 9700 PCR system (all from Applied Biosystems, Foster City, CA, USA). The cDNA obtained was used for real-time PCR using miR-specific TaqMan primers (Applied Biosystems, Foster City, CA, USA). For relative quantification, the expression of U6snRNA (Applied Biosystems) was used as an endogenous control.

To quantify mRNA, 1 μg total RNA was reverse-transcribed using a Revert Aid first-strand cDNA synthesis kit (Fermentas, Burlington Ontario, Canada). Subsequently, the mRNA expression of the target gene was determined by SYBR Green assays (Bio-Rad, Hercules, CA, USA). SYBR Green QPCR SuperMix-UDG was purchased from Invitrogen. Quantitative PCR was performed using an Applied BioSystems 7300 sequence detection system. All experiments were performed in triplicate. The level of expression was calculated based on the PCR cycle number (Ct) and the relative gene expression level was determined using the ΔΔCt method [36].

Western blot analysis

Whole-cell lysates for western blotting were extracted with lysis buffer containing 50 mm Tris (pH 8), HEPES (pH7.5), 150 mm NaCl, 1.5 mm MgCl2, 1 mm EDTA, 0.1% Triton X-100, 0.25% sodium deoxycholate and protease inhibitor (Roche). Protein samples were resolved by 10% SDS/PAGE, and gels were transferred to poly(vinylidene difluoride) membranes (Roche). Membranes were blocked using 5% non-fat milk for 1–2 h at room temperature, and subsequently incubated overnight at 4 °C with diluted primary antibodies against CCNE1, CCND1, CDK6, CDK2, VEGF-A and GAPDH (1:2000 dilution) (Cell Signaling Technology, Danvers, MA, USA). Signals were detected using the appropriate horseradish peroxidase-conjugated secondary antibody (Cell Signaling Technology). The blots were visualized using an enhanced Immobilon Western chemiluminescent horseradish peroxidase substrate (Millipore, Billerica, MA), according to the manufacturer's instructions, and the relative intensity of the specific bands was quantified using the FluorChem FC2 system (Alpha Innotech Corporation, San Leandro, CA, USA).

ELISA analysis

Immunoreactive VEGF-A protein levels were measured using commercial ELISA tests in accordance with the manufacturer's instructions and validated for use with the culture supernatant (R&D Systems, Minneapolis, MN). Cells were transfected with negative control oligonucleotides, anti-miR-16 or pre-miR-16 as described previously. Cell growth media were collected 36 h after transfection, and the cells were harvested and counted. Supernatants were then centrifuged at 14 000 g for 5 min to remove cellular debris, and subsequently stored at −70 °C until needed. All standards and supernatants from experimental and control cultures were assayed in duplicate and the values were averaged.

Construction of plasmids

Computer-based programs were used to predict potential target genes of miR-16. Using ‘hsa-mir-16 ‘as a search term, we queried the miRanda, TargetScan and PicTar. Human genomic DNA was used as a template to amplify the CCNE1 3′ UTR and the VEGF-A 3′ UTR. The amplified PCR product was gel-purified and digested with XbaI (Takara, Shiga, Japan), and inserted into the XbaI sites of the PGL3 vector (Promega, Fitchburg, WI, USA), producing PGL3-CCNE1-3′UTR and PGL3-VEGF-A-3′UTR.

Transient transfection and reporter gene assay

The pre-miR precursor molecule for miR-16 (pre-miR-16), the anti-miR inhibitor for miR-16 (anti-miR-16) and their scrambled non-coding RNAs were obtained from Ambion. Three siRNAs (siCCNE1) are available from RIBO BIO (Guangzhou, China). During the experiment, we used two siRNAs with a more obvious interference effect (si-h-CCNE1_001; sense strand 5′-GGAUGUUGACUGCCUUGAAdTdT-3′ and antisense strand 3′-dTdTCCUACAACUGACGGAACUU-5′; si-h-CCNE1_002; sense strand 5′-GGACAAUAAUGCAGUCUGUdTdT-3′ and antisense strand 3′-dTdTCCUGUUAUUACGUCAGACA-5′).

Transfection of dMSCs with miRNA oligonucleotides was optimized utilizing Lipofectamine 2000 (Invitrogen). They were co-transfected with the luciferase reporter constructs described above (200 ng), pRL-CMV (20 ng; Promega) and the appropriate miRNA precursor using Lipofectamine 2000. After 36 h, cells were washed and lysed with passive lysis buffer (Promega), and firefly luciferase (f-luc) and Renilla luciferase activities were determined using the dual-luciferase reporter assay system (Promega) and a luminometer. The relative reporter activity was obtained by normalization to the Renilla luciferase activity.

Flow cytometry

The specific surface antigens of dMSCs in the cultures, after passages 2–4, were characterized by flow cytometry analysis. The following murine monoclonal antibodies, purified or directly conjugated with fluorescein isothiocyanate (FITC), phycoerythrin (PE) or allophycocyanin (APC), were used in fluorescence-activated cell sorting (FACS) analysis: anti-CD105, anti-CD73, anti-CD90, anti-HLA-ABC, anti-CD29, anti-CD44, anti-CD106, anti-HLA-DR, anti-CD19, anti-CD11b, anti-CD14, anti-CD34, anti-CD31, anti-CD45 and IgG/IgM isotype controls (all from BD Biosciences, San Jose, CA, USA). For fluorescence measurements only, data from 10 000 single cell events were collected using a standard FACScalibur™ flow cytometer (Immunocytometry Systems/Becton Dickinson, San Jose, CA, USA). Data were analyzed using CELLQuest™ (Becton Dickinson).

Cell proliferation assay

Cell proliferation was assessed by seeding appropriate numbers in 12-well plates (Corning, Lowell, MA, USA), culturing for various times, followed by harvesting and counting. A cell counting kit (Dojindo Laboratories Inc., Kumamoto, Japan) was used to obtain a qualitative index of cell viability. Briefly, treated groups of dMSCs were plated in 96-well plates in Dulbecco's modifed Eagle's medium/F12 (Gibco) supplemented with 10% fetal bovine serum at a density of 2 × 103 cells per well. The cells were cultured for ≤ 2 days. Ten microliters of a solution containing WST-8 (4-[3-(2-methoxy-4-nitrophenyl)-2-(4-nitrophenyl)–2H-5-tetrazolio]-1,3-benzene disulfonate sodium salt) was added to each well. Following incubation for an additional 4 h, the absorbance was measured at 450 nm using a multi-detection microplate reader (Hynergy™ HT; Bio-Tek, Winooski, USA).

Cell apoptosis assay

Annexin V-DY634 (5 mL; Immunostep, Salamanca, Spain) was added to 4 × 105 cells suspended in 400 μl binding buffer (see below). As a negative control, 2 × 105 cells were used without addition of Annexin V antibody. Cells were vortexed gently and incubated in the dark for 15 min at room temperature. Cells were centrifuged at 300 g for 5 min, and 1 ×  binding buffer (500 mL) (10 mm Hepes, 140 mm NaCl, 2.5 mm CaCl2 in double distilled water) and propidium iodide were added to each tube. The samples were analyzed using FACScan (BD Biosciences) with Cell Quest software (BD Biosciences).

Cell cycle assay

The dMSCs were harvested, washed once with NaCl/Pi for five minutes, and fixed in 70% ethanol overnight. Approximately 1 × 106 fixed cells were incubated with 50 mg·mL−1 propidium iodide and 1 mg·mL−1 RNase A for 30 min. Analysis was performed on a Cell-cycle modeling was performed using Modfit software version 3.0 (Verity Software House, Topsham, ME, USA).

Cell migration assay

For transwell migration assays, cells (1 × 104–2 × 104) were cultured in the upper well of transwell plates with an 8 mm pore size membrane (BD Biosciences). The transwell plates were incubated for 6–8 h in a 5% CO2 atmosphere saturated with H2O. At the end of the incubation, cells that had entered the lower surface of the filter membrane (migrant cells) were fixed with 90% ethanol for 30 min at room temperature, washed three times with distilled water, and stained with 0.1% crystal violet in 0.1 m borate and 2% ethanol for 30 min at room temperature. Cells remaining on the upper surface of the filter membrane (non-migrant) were scraped off gently using a cotton swab. Images of migrant cells were captured using a BX51 photomicroscope (Olympus, Tokyo, Japan). Cell migration was quantified by blind counting of the migrated cells on the lower surface of the membrane for five fields per chamber.

HUVEC capillary tube and network formation assay on Matrigel

Matrigel (BD Biosciences, San Jose, CA, USA) (300 μL) diluted 1 : 1 with serum-free medium was added to 24-well plates and incubated for at least 1 h for gelling (pre-solidified Matrigel). HUVEC (1 × 105) were added to the pre-solidified Matrigel. The cells started the process of forming capillary tubes and networks once seeded on Matrigel. Incubation for 8 h revealed the most pronounced difference between treated groups in terms of capillary tube and network morphology. Tube-like structures were defined as endothelial cord formations that were connected at both ends, and the mean tube length in five random fields per well was quantified. These results were analyzed using image-pro plus software (Media Cybernetics, Rockville, MD, USA).

Statistical analysis

The results are expressed as means ± SEM. All experiments were repeated at least three times. The statistical significance between groups was analyzed by one-way ANOVA followed by Student–Newman–Keuls multiple comparison tests. Differences with a P value < 0.05 were considered statistically significant. Graphics were created using GraphPad prism software (www.graphpad.com).

Acknowledgements

This work was supported by the grant from the National Natural Science Foundation (project numbers 81072410, 81101552 and 81270713) and the special grant for maternal–fetal medicine from the Jiangsu Province Health Department of China (project number 81070508).

Ancillary