Insulin‐like growth factor 1 ameliorates pre‐eclampsia by inhibiting zinc finger E‐box binding homeobox 1 by up‐regulation of microRNA‐183

Abstract As a common hypertensive complication of pregnancy, preeclampsia (PE) remains one of the leading causes of maternal and fetal with high morbidity and mortality worldwide. Much research has identified the vital functions of insulin‐like growth factor 1 (IGF‐1) in PE treatment. However, the combined roles and molecular mechanism of IGF‐1 and microRNAs (miRNAs) underlying PE remain unclear. Therefore, we first measured the expression of IGF‐1, zinc finger E‐box binding homeobox 1 (ZEB1) and microRNA‐183 (miR‐183) expression in the placenta tissues of patients with PE by Western blot analysis and RT‐qPCR. Interactions among IGF‐1, ZEB1 and miR‐183 were assessed by Western blot analysis, ChIP‐PCR and dual‐luciferase reporter gene assay. The effect of IGF‐1 on the biological characteristics of trophoblast cells was investigated by CCK‐8, colony formation assay and in vitro angiogenesis experiments after cells were transfected with si‐IGF‐1. Finally, a mouse eclampsia model induced by knockdown of IGF‐1 was established to confirm the in vitro effect of IGF‐1 on PE. We found that IGF‐1, ZEB1 and miR‐183 were highly expressed in the placental tissues of patients with PE. The knockdown of IGF‐1 resulted in reduced proliferation and invasion of trophoblast cells and was accompanied by inhibited angiogenesis. ZEB1 was positively regulated by IGF‐1 via ERK/MAPK pathway, which in turn inhibited miR‐153 expression by binding to the miR‐183 promoter. The in vitro experiments further confirmed that IGF‐1 knockdown could induce PE. To sum up, IGF‐1 knockdown elevated expression of miR‐183 by downregulating ZEB1, thereby promoting deterioration of PE.


| INTRODUC TI ON
Preeclampsia (PE) is a heterogeneous and common human-specific pregnancy complication related to impaired placental development, which can jeopardize maternal and foetal lives if proper measures are not taken. [1][2][3] The prevalence of PE has been reported to be 3%-5%. 4 Delayed maternal age, overweight and vascular diseases have been implicated as important risk factors for PE. 5 The current clinical management strategies available for PE are delivery of the placenta or preterm-fetus. 6,7 However, present understandings of the intricate relationships among respective PE disease pathways are limited, which hinders us from developing better methods for successful management and prevention of PE. 8 Therefore, this study was designed to explore a promising new therapeutic target to alleviate PE at the molecular level.
As a mitogenic hormone, insulin-like growth factor 1 (IGF-1) has been indicated to be closely related to various biological processes such as growth, metabolism, angiogenesis and differentiation. 9 Previous research has confirmed that IGF1 was involved in the compensation for placental deficits by enhancing the proliferation of smooth chorion extravillous cytotrophoblasts. 10 Besides, another recent study has demonstrated that IGFs are important biomarkers in the prediction of PE. 11 Moreover, IGF1 has been reported to be associated with PE by mediating the expression of storkhead box 1 in extravillous trophoblasts. 12 Zinc finger E-box binding homeobox 1 (ZEB1) is a molecule that is predominantly involved in the epithelial-to-mesenchymal transition (EMT) of various cancers. 13 Nevertheless, ZEB1 has been found to be a crucial regulator in the progression of placental maturation. 14 In addition, it has also been confirmed that IGF-1 can repress the EMT in melanoma-initiating cells and thus affect the downstream expression of ZEB1. 15 MicroRNAs (miRNAs) are a group of small noncoding RNAs that have been reported have an association with the pathogenesis of many diseases, including cancers, gestational diabetes and placental miRNAs dysregulation, which can induce pregnancy complications. [16][17][18] Besides, miRNAs have also been identified as the vital regulatory molecules participating in the progression of PE. 19 For instance, miR-126 has been identified as a potential biomarker for PE pathogenesis. 20 Furthermore, miR-558 has been indicated to exert a promoting effect on the invasion and migration of trophoblasts, thereby participating in the progression of PE. 21 More importantly, miR-183 has also been confirmed to be a serum biomarker for PE. 22 Meanwhile, a previous report has illustrated that miR-183 can be inhibited by ZEB1. 23 Based on the above-mentioned literature, we speculated that IGF1 could mediate miR-183 expression by regulating ZEB1 expression. However, it is still largely uncertain whether the same effects will be exerted in the context of PE. Hence, the present study was conducted to explore the regulatory role of IGF1 in the molecular mechanism underlying PE and ultimately provide a novel therapeutic target for the treatment of PE.

| ETHICS STATEMENT
This study was approved by the Obstetrics Ethics Committee of Second Xiangya Hospital, Central South University, and written informed consent was obtained from each participant. Human researches were conducted in strict accordance with the Declaration of Helsinki. All animal experiments were performed under the approval of the Animal Ethics Committee of Second Xiangya Hospital, Central South University. 36

| Immunohistochemistry (IHC)
Paraffin sections were dewaxed with xylene conventionally, hydrated with gradient alcohol (anhydrous ethanol, 95% ethanol and 75% ethanol for 3 min each), treated in citrate repair solution and blocked with TBS solution containing 10% normal serum and 1% bovine serum albumin (BSA) for 2 h. Then, the sections were added with normal serum of rabbit polyclonal antibody to insulin-like growth factor 1 (IGF-1, ab9572, 1: 1000, Abcam,) dropwise as negative control (NC) and incubated in a refrigerator at 4°C overnight. Next, 50 μl of 3% H 2 O 2 were added into each section and incubated for 20 min to eliminate endogenous peroxidase activity. In addition 50 μl of polymer enhancer and 50 μl of goat anti-rabbit secondary antibody (ab205718, 1: 2000; Abcam) were added into section and incubated for 30 min at 37°C, respectively. Finally, each section was added with 100 μl freshly prepared diaminobenzidine (DAB) developing solution and observed   under the microscope for 3-10 min, with positive tissues presenting   as brown colour. Then, the sections were rinsed with distilled water,   counterstained with haematoxylin, dehydrated with gradient alcohol   (75% ethanol, 95% ethanol and anhydrous ethanol), sealed with neutral resins and observed again under the microscope.

| Western blot analysis
Total proteins in tissues or cells were extracted using radioimmunoprecipitation assay (RIPA) lysis buffer containing phenylmethylsulfonyl fluoride (PMSF). The total protein concentration was measured using the BCA kit (P0012, Beyotime Biotechnology Co., Ltd.,). Then, 50 μg of proteins were dissolved in 2 × sodium dodecyl sulphate (SDS) loading buffer after boiled for 10 min, and then subjected to SDS-polyacrylamide gel electrophoresis (PAGE) gel electrophoresis.

| RNA isolation and quantitation
The total RNAs in tissues or cells were extracted in strict accordance to the manufacturer's instructions of the TRIzol kit (cat: 15596018; Invitrogen), whereas the RNA concentration was determined using RT-qPCR. The cDNA was synthesized using the ImProm-II™ reverse transcription system kit. The reverse-transcribed cDNA was diluted to 50 ng/μl for subsequent fluorescent quantitative PCR. The reaction amplification system was employed according to the manufacturer's protocols of Synergy Brands (SYBR) premixed ExTaq II kit (Takara). β-actin was taken as an internal reference. The primer sequences used for RT-qPCR are listed in Table 1.

| Chromatin Immunoprecipitation (ChIP) assay
After treatment with 4% formaldehyde (final concentration 1%), the collected HTR-8/SVneo cells were sonicated by ultrasonic, followed by the addition of the ZEB1 antibody and fully contacted with the miR-183 promoter. Protein A Agarose/SaLmon Sperm DNA was added to the cells to allow binding to ZEB1 antibody-PBX3-OTX1 promoter complexes. The complexes were then precipitated and rinsed to remove the non-specific binding and the enriched promoter complexes were obtained. RT-qPCR was then conducted after the enriched promoter complexes were de-crosslinked and purified.

| Mouse model of PE construction
A total of 20 specific pathogen-free grade CD-1 mice (aged 6-8 weeks, After the mice were euthanized, the placenta tissues were collected and immediately frozen and store at −80°C for future use. The HIV-I based self-extinction lentiviral vector plasmid pLV-EGFP was prepared as described previously 24  If the systolic blood pressure and proteinuria on Day 18 of pregnancy were dramatically increased (110.12 ± 2.73 mmHg), the establishment of the mouse PE model was considered successful. The mice with successful modelling were divided into two groups: sh-NC and sh-IGF-1, with 10 mice in each group.

| Blood pressure measurement
Mice were tested for blood pressure on Days 6, 12 and 18 of pregnancy using a rat tail blood pressure meter (Kent International USA, Inc.,). The measurement was conducted at 8 AM. after mice were deprived of water for 2 h. When measuring blood pressure, the room temperature was kept at about 25°C, and the incubator was preheated so that body temperature remained at 37°C.
Mice were placed in the incubator 10 min prior to the measurement. The test clamp was placed at the end of the mice tail and the blood pressure was measured after the heartbeat was stabilized.
Each mouse was measured three times, and the average value was obtained.

| Automated biochemical analysis
The urine of mice in each group was collected on the E18.5. The level of urine protein was determined by an automatic biochemical analysis. Urine was centrifuged at 1006.2 g for 10 min at 4°C, and the level of urine protein was determined using biprenore reagent (HPBIO-R1253, Pengpai Technology company,). On the 21 st (final) day of pregnancy, the mice were not delivered, but the placenta was taken by caesarean section, part of which was used for extraction and total protein and RNA detection, whereas the other samples were fixed for subsequent experiments.

| Statistical analysis
The SPSS 21.0 statistical software (IBM Corp,) was employed for statistical analysis. All measurement data were expressed as mean ± standard derivation. A comparison between the two groups was analysed by independent t-test. Comparisons among multiple groups were analysed by one-way analysis of variance (anova) and followed by Tukey's post-hoc test. Comparison among groups at different time points was analysed by repeated-measures anova and followed by Bonferroni's post-hoc test. Enumeration data were expressed as a percentage or ratio. Correlations between data were analysed by Pearson correlation analysis. A p < 0.05 was considered as statistically significant.

| IGF-1 was poorly expressed in placenta tissues of preeclamptic patients
The clinical features of participants, including 18 women with PE and 18 in normal pregnancy (NP) are presented in Table 2. There was no significant difference in age between the subjects in the NP group and PE group, but systolic blood pressure, diastolic blood pressure and urine protein content increased remarkably in the PE group.
The expression of IGF-1 in placental tissues from the PE and NP groups was determined by IHC, showing that the expression of IGF-1 in placental tissues of the PE group was lower than that in the NP group ( Figure 1A). RNA and protein levels of IGF-1 were determined by RT-qPCR and Western blot analysis, showing that RNA and protein expressions of IGF-1 were decreased dramatically in the PE group ( Figure 1B, C). Hence, low expression of IGF-1 was present in the placental tissues of PE. In addition, the invasive ability of cells was detected by Transwell assay, which showed the invasive rate of cells was significantly reduced after transfection with si-IGF-1 ( Figure 2D). HUVEC cells were tested for angiogenesis in vitro and the number and length of newly formed blood tubes were measured. Based on these observations, the number of angiogenesis was decreased and the length of newly formed blood tubes became shorter in cells transfected with si-IGF-1 compared with cells transfected with si-NC, thus suggesting that angiogenesis was inhibited after downregulation of IGF-1 ( Figure 2E). According to these findings, knockdown of IGF-1 could inhibit the proliferation and invasion of trophoblast cells, as well as angiogenesis in vitro.

| The proliferation and invasion of trophoblast cells and angiogenesis were enhanced by ZEB1 through inhibiting miR-183
The effect of ZEB1 knockdown is presented in Figure 4A. Results from the ChIP-PCR assay revealed that ZEB1 was capable of binding to the miR-183 promoter and was reduced in cells transfected with si-ZEB1 ( Figure 4B). Dual-luciferase reporter gene assay also confirmed that the transcription of miR-183 was regulated by ZEB1 ( Figure 4C). RT-qPCR was conducted to investigate whether ZEB1 could affect the expression of miR-183, and the results illustrated that the expression of miR-183 was notably increased in HTR-8/SVneo cells after being transfected with si-ZEB1, whereas expression was significantly decreased after transfected with oe-ZEB1 ( Figure 4D). These results indicated that ZEB1 could act as a transcriptional repressor to regulate the expression of miR-183.
Next, we examined whether miR-183 affects the biology of trophoblast cells. HTR-8/SVneo cells were transfected with miR-183 mimic and cell proliferation was detected by CCK-8 assay and colony formation assay. As shown in Figure 4E invasion and angiogenesis, and that miR-183 expression was regulated by ZEB1, which may affect the early development of the placenta.

| IGF-1 downregulation inhibited proliferation and invasion of trophoblast cells and angiogenesis in vitro could be reversed by silencing of miR-183
We further examined whether the expression of miR-183 was reg- Results indicated decreased expression after IGF-1 knockdown, whereas an increased VEGF expression was observed after the silencing of miR-183 ( Figure 5E). In vitro angiogenesis experiments also confirmed that inhibition of miR-183 reversed the repressed angiogenesis caused by si-IGF-1 ( Figure 5F). In summary, these findings suggest that IGF-1 could regulate miR-183 by regulating ZEB1 expression, which affected the biological functions of trophoblast cells, ultimately involving in the pathogenesis of PE.

| Silencing of IGF-1 could induce the pathogenesis of PE
To investigate the function of IGF-1 in mice, we established a lentiviral-mediated placental-specific knockdown model and confirmed the model establishment by RT-qPCR ( Figure 6A). To determine whether sh-IGF-1 induces PE in mice, we measured blood pressure on embryonic days E0, E8.5, E10.5, E12.5, E14.5, E16.5 and E18.5. Results showed that systolic and diastolic blood pressure in the sh-IGF-1 group increased significantly on E16.5, and continued to increase during the rest of the pregnancy (p < 0.05, Figure 6B).
Notably, the blood pressure returned to normal after delivery, which is similar to the phenomenon of human postpartum recovery. The urine protein levels measured at E18.5 were remarkably higher in the sh-IGF-1 group than in the NC group ( Figure 6C). These findings indicated that sh-IGF-1 could induce a human-like PE model, in which mouse body weight was significantly lower than the normal group ( Figure 6D). Placental sections were subjected to HE-staining, which showed that the structure of the placenta was disordered after sh-IGF-1 treatment ( Figure 6E). Staining with CD34 antibodies also revealed a significant decrease in vascular density and branching, All experiments were repeated three times. Data were measurement data and expressed as mean ± standard deviation. The independent sample t-test was used for comparison between two groups. Comparisons among multiple groups were analysed by one-way anova and followed by Tukey's post-hoc test. * p < 0.05 vs. control group and NC group. The cell experiment was repeated three times. ZEB1, zinc finger E-box binding homeobox 1; ERK1/2, extracellular signal-regulated kinase 1/2; p-ERK1/2, phosphorylated-ERK1/2; NC, negative control indicating that IGF-1 plays an important role in mouse trophoblast cell proliferation and invasion, and the formation of blood vessels in the placenta ( Figure 6F). In short, these results suggested that the silencing of IGF-1 may implicate in the pathogenesis of PE.

| DISCUSS ION
Preeclampsia (PE) is a common complication of pregnancy, which threatens maternal and fetal health. 25 Currently, placental delivery is the only radical solution for PE, whereas antihypertensive drugs were only effective for patients in a stable condition. However, eclampsia or lethal complications including stroke and pulmonary edema can occur if patients do not receive proper or timely treatments. 26 The IGF system is thought to play a role in abnormal placentation, whereby IGF-I activity is blocked and trophoblast invasion is reduced. Therefore, it has been suggested that pregnant women who later develop preeclampsia may have a lower concentration of IGF-I in early pregnancy compared with other pregnant women. 27 In the current study, the potential role of IGF-1 in PE and the molecular All experiments were repeated three times. Data were measurement data and expressed as mean ± standard deviation. The independent sample t-test was used for comparison between two groups. Comparisons among multiple groups were analysed by one-way anova and followed by Tukey's post-hoc test. * p < 0.05 vs. the si-NC group, si-NC + inhibitor-NC and si-IGF-1 + inhibitor-NC group. The cell experiment was repeated three times The present study further illustrated that IGF-1 was able to upregulate ZEB1 expression through ERK/MAPK pathway. ZEB1 mediates a diverse array of processes including mesoderm-derived cell differentiation, eye development, neural development and lymphopoiesis, as well as cell proliferation and senescence. 35 In line with our findings, ZEB1 has been confirmed as a key transcriptional regulator of EMT, and its expression was elevated by IGF-1 in prostate cancer cells. 36 Moreover, it has been previously demonstrated that IGF-1 could protect renal tubular epithelial cells from renal injuries via ERK/MAPK pathway activation. 37 Besides, the ERK/MAPK pathway has also been demonstrated to be a crucial player in the appropriate development, differentiation and morphogenesis of the placenta. 38 In the meantime, the nuclear localization of green fluorescent protein-ZEB1 fusion clones has been identified to be disrupted by IGF-1-induced MEK/ERK phosphorylation. 35 Interestingly, the present study revealed that the elevation of IGF-1-mediated ZEB1 resulted in reduced miR-183 expression, thereby promoting the proliferation, invasion and angiogenesis of trophoblast cells, and thus alleviated the progression of PE in the mouse model. The downregulation of ZEB1 has been illustrated to be an inducer of PE by weakening the migration and invasive capacity of trophoblast cells. 39 Moreover, the elevated miR-20a levels in PE tissues have also been identified to inhibit the proliferation and invasion of trophoblast cells. 40 Furthermore, miR-15b proved to be an inhibitor of trophoblast cell invasion and angiogenesis of endothelial cells. 41 Previous studies have also validated the relationship between aberrant expression of miR-183 and the pathogenesis of PE. 22,42 Till date, several studies have also certified the negative regulatory effect of ZEB1 on the expression of miR-183. 43,44 In addition, miR-183-5p has been confirmed to be suppressed by ZEB1, which was associated with the development of cardiac hypertrophy and heart failure. 45 Moreover, miR-431 inhibited the migration and invasion of trophoblastic cells by targeting ZEB1, which might itself give rise to the onset of PE. 39 IGF-1 has also been confirmed to be repressed by miR-30a-3p in the placenta tissues of patients with PE, as well as repressed the invasive F I G U R E 6 Silencing of IGF-1 induces PE pathogenesis. A, The expression of IGF-1 after transfection of sh-IGF-1 determined by RT-qPCR. B, Systolic blood pressure and diastolic blood pressure in cells transfected with sh-IGF-1 group or sh-NC at embryonic days E0, E8.5, E10.5, E12.5, E14.5, E16.5 and E18.5. C, Urine protein content in cells transfected with sh-IGF-1 of mice at E18.5. D, Weight of mice embryo. E, The structure of the placenta observed by HE staining (20 ×). F, The structure of the placenta observed by CD34 staining (20 ×). Data were measurement data and expressed as mean ± standard deviation. An independent t-test was used for comparison between two groups. * p < 0.05 vs. that of cells transfected with sh-NC. The cell experiment was repeated three times. HE, haematoxylin-eosin

| CON CLUS ION
The results of the present study revealed that IGF-1 protected against PE by decreasing miR-183 expression through elevating ZEB1 expression (Figure 7). Results of this study thus support a model in which an IGF-1-mediated IGF-1/ZEB1/miR-183 axis modulates in the progression of PE, suggesting promising new therapeutic targets for PE. However, we expect that other available molecular mechanisms may also contribute to better clinical outcomes of PE management.

CO N FLI C T S O F I NTE R E S T S
The authors declare that they have no conflicts of interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
All the data obtained and/or analyzed during the current study were available from the corresponding author on reasonable request.