Circular RNA VRK1 facilitates pre‐eclampsia progression via sponging miR‐221‐3P to regulate PTEN/Akt

Abstract Pre‐eclampsia (PE) is a worldwide pregnancy‐related disorder. It is mainly characterized by defect migration and invasion of trophoblast cells. Recently, circular RNAs (circRNAs) have been believed to play a vital role in PE. The expression patterns and the biological functions of circRNAs in PE remain elusive. Here, we performed a circRNA microarray to identify putative PE‐related circRNAs. Bioinformatics analyses were used to screen the circRNAs which have potential relationships with pre‐eclampsia, and we identified a novel circRNA (circVRK1) that was up‐regulated in PE placenta tissues. By using HTR‐8/SVneo cells, circVRK1 knockdown significantly enhanced cell migration and invasion abilities, as well as epithelial‐mesenchymal transition (EMT). Mechanistically, we found that circVRK1 and PTEN could function as the ceRNAs to miR‐221‐3p. Overexpression of miR‐221‐3p promoted cell migration, invasion and EMT via regulating PTEN. The cotransfection of miR‐221‐3p inhibitor or PTEN reversed the effect from circVRK1 knockdown. Moreover, the circVRK1/miR‐221‐3p/PTEN axis greatly regulated Akt phosphorylation. In general, circVRK1 suppresses trophoblast cell migration, invasion and EMT, by acting as a ceRNA to miR‐221‐3p to regulate PTEN, and further inhibit PI3K/Akt activation. The purpose of this paper is to open wide insights to investigate the onset of PE and provide new potential therapeutic targets in PE.

containing enhanced migration and invasion abilities. 7 So that trophoblast cells can anchor the placenta into maternal tissues and remodel the maternal spiral arterioles. 8 Therefore, the poor migration and invasion abilities of trophoblast cells were believed to be essential leading causes of PE. It is of great significance to explore the molecular mechanisms that drive the progression of preeclampsia, which may provide more effective clinical therapies. Circular RNAs (circRNAs) are a type of endogenous non-coding RNAs defined by continuous loops closed by covalent binding without 5′ to 3′ polarity and polyadenylated tails. 9,10 This type of RNA is formed by back-splicing events, resulting in enhanced stability relative to linear RNA. 11 In recent years, studies have proved that circRNAs participate in various diseases. 12,13 Competing endogenous RNA (ceRNA) is one of the functional roles of circRNAs which is involved in the metastasis of the diseases. [14][15][16] A circRNA functions as a miRNA sponge as it binds to the relevant miRNA and influences the generation of miRNAs that interact with their target mRNAs. 14,17 However, few studies demonstrated the regulation mechanism of circRNAs in PE. For example, Shen et al 18 found that hsa_circ_0006772 can inhibit trophoblast cell migration and EMT via the miR-762/Grhl2 pathway. Down-regulated hsa_circ_0088227 suppresses trophoblast cell proliferation and invasion by targeting miR-384. 19 Here, we believe that circRNAs play critical functional roles in PE.
In this study, we identified a circular RNA (circVRK1) by microarray from the placentas of PE and NP. It termed that circVRK1 was up-regulated in PE and could function as a ceRNA to PTEN to suppress trophoblast cell migration, invasion and EMT by targeting miR-221-3P. The circVRK1/miR-221-3p/PTEN also influenced the PI3K/ Akt signal pathway. Overall, our study demonstrated that circVRK1 serves as a potential biomarker for the diagnosis of PE, hoping to provides new insights to study the pathogenesis of PE. University, and all participants signed informed consent. PE is characterized by diastolic blood pressure (mm Hg ≥ 90) or systolic blood pressure (mm Hg ≥ 140) after 20 weeks of gestation (indicating new-onset hypertension), with new-onset proteinuria (≥0.3 g/24 h or ≥2 + examined using a dipstick test for two random specimens). In this study, we included patients that were diagnosed with severe pre-eclampsia which is defined with PE with severe features according to ACOG (systolic blood pressure of 160 mm Hg or higher, or diastolic blood pressure of 110 mm Hg or higher, any of thrombocytopenia, impaired liver function as indicated by abnormally elevated blood concentrations of liver enzymes and severe persistent right upper quadrant or epigastric   pain unresponsive to medication and not accounted for by alternative diagnoses, renal insufficiency, pulmonary oedema, new-onset   headache unresponsive to medication and not accounted for by alternate, visual disturbance), 1 to avoid the influence of nonplacental factors in mild pre-eclampsia. We excluded twin pregnancy, gestational diabetes, renal disease, chronic hypertension, acute or chronic hepatitis, hyperthyroidism, and hypothyroidism.

| Samples collection and patient enrolment
The participants were divided by early-onset PE (PE occurs before 34 gestational ages), late-onset PE (PE occurs after 34 gestational ages) and natural pregnancy. 20 The characteristics were provided in Table 1. Tissues from the maternal surface of the placentas were dissected, cleaned thrice with diethylpyrocarbonate-treated saline to wash out blood after which they were snap-frozen in liquid nitrogen. These procedures were completed within 15 minutes after the caesarean sections. Then, the tissues were stored at −80°C.

| Total RNA extraction
TRIzol reagent (Invitrogen) was used to extract total tissue RNA from PE and NP specimens following the instructions on the product manual. The purity and concentration of the RNA were evaluated based on OD 260/280 readings performed from a NanoDrop ND-1000 machine (Thermo Fisher Scientific).

| Microarray and analysis of circRNA expression profile
Total RNA was evaluated with a NanoDrop ND-1000 and processed using RNase R (Epicentre, Inc) for circRNA enrichment and linear RNA removal. This was followed by amplification and transcription of the enriched circRNAs to fluorescent cRNAs. The Arraystar Human circRNA Array V2 (8x15K, Arraystar) was used to hybridize the labelled cRNAs which were incubated at 65°C for 17 hours in an Agilent hybridization oven. After washing, slides were captured with an Agilent Scanner G2505C. The Agilent Feature Extraction software (version 11.0.1.1) was used for data analysis. Data processing and quantile normalization were carried out by the R software limma package. In our study, a P-value of ≤0.05 and a fold change of ≥2 were set as the cut-off for identifying differentially expressed circRNAs between the PE and NP groups. The t test was used to determine statistical significance. The Benjamini-Hochberg FDR method was applied to determine the false discovery rate to correct the P-values.
The human reference genome (hg19) was used to align all differentially expressed circRNAs. Hierarchical clustering was performed using MeV (Multiple Experiment Viewer) to identify circRNAs with vastly altered expression. The R package Bioconductor was utilized to carry out GO enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis for the differentially expressed circRNAs.

| RNase R treatment
The treatment was performed under the manufacturer's instructions.
Briefly, total RNA after extracting by Trizol reagent was incubated with 3 U/μg of RNase R (Epicentre Technologies) for 20 minutes at 37°C. Then, the treated RNA was purified by RNA clean kit (Tiangen) according to the guidance.

| Cell migration and invasion assay
Transwell assay was used to explore trophoblast cell migration and invasion abilities. HTR-8/SVneo cells (1 × 10 5 cells) were transfected and  Wound assay was performed to detect cell migration. HTR-8/ SVneo cells were seeded in six-well plates (1 × 10 6 cells); the singlecell layer was wound using 200 μL pipette tip and then washed three times with PBS to remove cell debris. RPMI-1640 without FBS was added to each well and incubate for 48 hours at 37°C with 5% CO 2 .
Images of the wound were obtained at 0 and 48 hour at the same scratch position by a microscope.

| Immunofluorescence analysis
HTR-8/SVneo cells were seeded in six-well plates at 37°C with 5% CO 2 and then fixed in 4% paraformaldehyde after washed with PBS two times. Cells were incubated with primary antibodies of Ecadherin and vimentin for 1 hour (Table S1). After corresponding with fluoresce-labelled secondary antibodies and DAPI for nuclear counterstaining, microscopy was used to catch the image.

| Western blot
Total protein was extracted by radioimmunoprecipitation assay lysis buffer (RIPA, Solarbio) and then separated by SDS-PAGE to polyvinylidene difluoride membranes (PVDF, Millipore). After blocking, the PVDF, membranes were incubated overnight at 4°C with primary antibody E-cadherin, N-cadherin, vimentin, twist1, snail, ZEB1, ZEB2, PTEN, p-Akt and Akt (Table S1). The PVDF membranes were incubated with the secondary antibody for 2 hours at room temperature after washing with TBST. The bands were visualized using a stand-

ard protocol for electrochemiluminescence (New Cell & Molecular
Biotech, Co). GAPDH was used as the internal standard.

| RNA pull-down assay
Biotin-labelled circ VRK1 probe and control probe (Sangon Biotech) were used for pull-down assay. In short, the probe was incubated with magnetic beads (Life Technologies) to generate probe-coated beads and then incubated with the lysed HTR-8/ SVneo cell samples overnight at 37°C. On the next day, after washing, the sample was incubated with lysis buffer and proteinase K buffer. Finally, TRIzol reagent was added into the complex for RNA extraction. qRT-PCR was performed to test the pulldown RNAs.

| RNA pull-down sequencing
The enriched miRNAs by RNA pull-down were taken to construct a library that will be suitable for the Illumina high-throughput sequencing platform with VAHTS Small RNA Library Prep Kit for Illumina. The miRNAs were, respectively, connected with universal adapters at 3′ and 5′, and then underwent reverse transcription, PCR expansion and magnetic bead purification to finally obtain a stepwise library suitable for the Illumina platform. At the same time, the stability and reproducibility of the library are improved through strict quality control of the library.

| Dual-luciferase reporter assay
A dual-luciferase reporter assay was performed to determine the in-

| Data analysis
Quantitative data are expressed as means ± SD. Statistics were carried out by SPSS 20.0 software (SPSS Inc) or Graph Pad Prism 8.0.2 (GraphPad Software). The differences between two groups were calculated by Student's t test. One-way ANOVA was used for the comparison of multiple groups. Mann-Whitney test was performed in the analysis of circVRK1 expression between NP and PE placenta tissues, 24-hour proteinuria quantification and proteinuria level.
Birth gender was tested using the chi-square test. P < 0.05 was represented as statistical significance.

| Differential circRNA expression profiles in preeclampsia and bioinformatical analysis
To explore the differentially expressed circRNAs in PE and NP tissues, microarray analysis was performed on pre-eclampsia placental tissues (PE, N = 5) and normal pregnancy tissues (NP, N = 5). The circRNA microarray profile data have been loaded to the GEO database (GSE137854). The data left generated or used during the study appear in the submitted article. The locations on human chromosomes and the differential expression levels of these circRNAs are presented in Figure 1A. As the profile result showed, 77% of the differentially expressed circRNAs were exonic, 2% were intergenic, 7% were sense overlapping, 12% were intronic, and 2% were antisense ( Figure 1B). The volcano plot showed the raw alterations in the circRNA expression pattern in the PE and NP placental tissues

| Identification of circVRK1 in PE
In the data of the circRNA microarray profile, according to the fold change and the P-value, circVRK1 (hsa_circ_0000566) has drawn our attention. It is formed from exon 2 to 11 of the vaccinia-related  Table 3). The circVRK1 expression level was remarkably higher in the PE group than the NP (Figure 2A). Moreover, we also tested the expression between early-onset PE and late-onset PE.
The results showed that the circVRK1 in the early-onset PE group is remarkedly higher than the late-onset PE group ( Figure 2B). Earlyonset PE is believed to be a placental origin complication, which is caused by trophoblast cell dysfunction and insufficient spiral artery remodelling. 27,28 It was supposed that circVRK1 may be associated with trophoblast cell behaviours. Afterwards, we identified the specific back-spliced circular construction of circVRK1. Firstly, we performed qPCR in HTR-8/SVneo cells and Sanger sequencing for the products; the sequence result could completely match with the backspliced region ( Figure 2C). Subsequently, we found that circVRK1 could resist the digestion of RNase R while VRK1 and GAPDH were digested by RNase R, and only amplified in cDNA by divergent primers but not in gDNA ( Figure 2D,E). Fluorescence in situ hybridization analysis was performed to further explore the location of circVRK1, it was found expressed in distributed in the cytoplasm ( Figure 2F).
Taken together, the data suggested that the aberrantly expressed circVRK1 is correlated with the progression of PE and the trophoblast cell biological function.

| Silencing circVRK1 promotes trophoblast cell migration, invasion and EMT
To investigate the potential roles of circVRK1. We successfully suppressed the circVRK1 expression in HTR-8/SVneo cells using siRNA and shRNA ( Figure S1A,B), which were specific to circVRK1 but not VRK1 ( Figure 3A,B). From the wound assay, the cell migration rate in the circVRK1 knockdown group is higher than the NC group, which indicated circVRK1 knockdown promoted the migration of HTR-8/ SVneo cells ( Figure 3C). Transwell assay illustrated that the silencing of circVRK1 significantly enhanced cell migration and invasion viabilities ( Figure 3D). The EMT-related proteins which were considered as EMT marker were detected by Western blot. [29][30][31][32][33] As shown in Figure 3E, the vimentin, twist1, snail, ZEB1, ZEB2 expression in both circVRK1 knockdown group were remarkably elevated, while the E-cadherin expression was down-regulated. Furthermore, immunofluorescence staining confirmed that circVRK1 knockdown increased vimentin expression and decreased E-cadherin expression, which further proved that circVRK1 knockdown promoted EMT of HTR-8/SVneo cells ( Figure 3F). Consequently, these data proved circVRK1 suppresses trophoblast cell migration, invasion and EMT.

| circVRK1 functions as ceRNA to miR-221-3p trophoblast cell
According to the ceRNA theory, circRNA can sponge to miRNA to suppress the mRNA which shares the same binding sites, and the following translation. 14

F I G U R E 2
The identification of circVRK1. A, qPCR examined circVRK1 expression between PE and NP. B, qPCR examined circVRK1 expression between early-onset PE, late-onset PE. C, Sanger sequence examined the circular splicing site of circVRK1. D, qPCR examined circVRK1 expression. VRK1 was digested by RNase R but circVRK1 was not. E, Using convergent or divergent primers, circVRK was amplified in cDNA but not in gDNA by using divergent primers. F, Fluorescence in situ hybridization showed circVRK1 was widely expressed in the cytoplasm of HTR-8/SVneo cells, the red represents circVRK1. Data are presented as means ± SD. **P < 0.01, ***P < 0.001

F I G U R E 3
Silencing circVRK1 promoted trophoblast cell migration, invasion and EMT. A and B, qPCR was used to term that the siRNA and the shRNA were specific to circVRK1 but not VRK1. C, Wound assay was used to detect cell migration ability after cells were transfected with circVRK1 knockdown and NC. D, Transwell assays were performed to detect cell migration and invasion abilities in HTR-8/ Svneo cells. E, Western blot assay was used to evaluate the EMT-related proteins in HTR-8/Svneo cells. F, Immunofluorescence analysis was used to detect E-cadherin and vimentin in HTR-8/Svneo cells. Data are presented as means ± SD. **P < 0.01, ***P < 0.001, ns: no significance | 1835 prediction, we selected a microRNA (hsa-miR-221-3p, miR-221-3p).
The result of RNA pull-down showed that circVRK1 was successfully pulled down by miR-221-3p ( Figure 4A,B). We further performed qPCR to find that the expression of miR-221-3p was remarkedly up-regulated by circVRK1 knockdown ( Figure 4C). RNA22 v2 and RNA hybrid were used to predict the binding relationship between circVRK1 and miR-221-3p ( Figure 4D). Dual-luciferase reporter assay showed that overexpressed miR-221-3p could significantly inhibit the luciferase activity of the wild-type of circVRK1 but not the mutant ( Figure 4E). Additionally, fluorescence in situ hybridization represented that circVRK1 and miR-221-3p were widely expressed in HTR-8/SVneo cells ( Figure 4F). Collectively, the results proved circVRK1 directly interacted with miR-221-3p in trophoblast cells.

| miR-221-3p regulates the migration, invasion and EMT of trophoblast cell via regulating PTEN
By using TargetScan, we found phosphatase and tensin homolog

| circVRK1 serves as ceRNA to inhibit trophoblast cell migration, invasion and EMT by sponging miR-221-3p to regulate PTEN
To further explore whether circVRK1 suppresses trophoblast cell Moreover, PTEN is a well-known tumour suppressor that has been proved closely related to the PI3K/Akt signal pathway. [34][35][36][37] So we further explore the potential regulation of circVRK1/miR-221-3p/PTEN axis in the PI3K/Akt activation. As the Western blot assay showed, the Akt expression showed no significant difference, circVRK1 knockdown showed higher phosphorylation levels of Akt. The cotransfection of miR-221-3p inhibitor or PTEN significantly weakened that effect ( Figure 5F). These data indicated that the circVRK1/miR-221-3p/PTEN axis plays an important role in the PI3K/Akt signal pathway in trophoblast cell progression. sponge to inhibit trophoblast cell migration, invasion and EMT via regulating PTEN, the circVRK1/miR-221 3P/PTEN axis is also revolved in PI3K/AKT pathway ( Figure 6). knockdown also significantly elevated the phosphorylation of Akt, which indicated the inhibition of circVRK1 in PI3K/Akt activation.

| D ISCUSS I ON
It is known that PI3K/Akt is widely involved in modulating cell biological behaviours such as migration and invasion abilities, as well as EMT. [60][61][62] The PI3K/Akt signal pathway plays an important role in PE.
PI3K/Akt activation, as well as the phosphorylation of Akt, remains at a low level during PE. The inhibition of Akt phosphorylation can lead to weakening migration and invasion viabilities. 21,63,64 In the research, the phosphorylation of Akt was remarkedly regulated by circVRK1/ miR-221-3p/PTEN axis, which termed that the regulation of the axis in trophoblast cell progression is involved in the PI3K/Akt signal pathway. However, there is little knowledge about that PI3K/Akt signal pathway modulating trophoblast cell EMT, and if the axis modulated the migration, invasion and EMT via the PI3K/Akt signal pathway in PE. It needs to deeper exploration in the future studies.
Back to circVRK1, it is spliced from VRK1 which is a Ser-Thr kinase located in the nucleus which participates in the cell cycle containing DNA repair processes and apoptosis to DNA damage. 65 It can specifically activate p53 phosphorylation to respond to DNA damage. 66 Moreover, p53 remains at an elevated phosphorylation level in PE. 67 It suggests that circVRK1 may also participate in the regulation of the trophoblast cell cycle. Next, we will further explore the influence on the trophoblast cell cycle from circVRK1 and VRK1.
In conclusion, this study demonstrates that circVRK1 can accelerate PE progression by suppressing trophoblast cell migration, invasion and EMT. It plays an inhibitory role by sponge miR-221-3p to regulate PTEN and the following PI3K/Akt signal pathway. Hoping this study can provide new therapeutic and diagnostic targets for PE.

ACK N OWLED G EM ENTS
This work was supported by the National Natural Science Foundation under Grant no. 81871173.

CO N FLI C T O F I NTE R E S T
There is no conflict of interest. Resources (lead).

DATA AVA I L A B I L I T Y S TAT E M E N T
The microarray profile data that support the findings of this study are openly available in the GEO database (GSE137854). The data not mentioned in the study that support the findings are available from the corresponding author upon reasonable request.
F I G U R E 5 circVRK1 modulate trophoblast cell migration, invasion and EMT via regulating miR-221-3p/PTEN/Akt. A and B, Wound assay was performed cell migration ability of HTR-8/Svneo cell. C and D, Transwell assays were used to detect cell migration and invasion abilities in HTR-8/Svneo cells. E, Western blot assay was used to evaluate EMT-related proteins in HTR-8/Svneo cells. F, Western blot assay was used to evaluate p-Akt and Akt in HTR-8/Svneo cells. Data are presented as means ± SD. **P < 0.01, ***P < 0.001, ns: no significance F I G U R E 6 circVRK1 acts as a ceRNA to miR-221-3p to modulate trophoblast cell migration, invasion, and EMT via regulating PTEN O RCI D Ziwei Li https://orcid.org/0000-0002-0912-4144