The present study was approved by the Ethics Committee of the West China Hospital of Sichuan University (Chengdu, China), and all patients signed written informed consent forms prior to the commencement of the study. The placenta tissues were obtained from pregnant women who were admitted to the West China Hospital of Sichuan University between January 2018 and January 2019. The diagnosis of patients with preeclampsia complied with The American College of Obstetricians and Gynecologists Practice Bulletin 2019 (3). The control group comprised women with normal pregnancy, and those with pregnancy diseases were excluded. The ages of pregnant women in the two groups ranged between 24-38 years, with a mean age of 28 years. There was no statistical difference in the age of pregnant women between the two groups. Placental tissues (1 cm³) and serum samples were collected from 30 pregnant women with preeclampsia and 30 normal parturients with the same gestational age (control group).

HTR-8 (cat. no. CRL-3271) and JEG-3 (cat. no. HTB-36) cells were obtained from the American Type Culture Collection and cultured in Dulbecco's modified Eagle's medium (cat. no. D5796; MilliporeSigma), containing 10% fetal bovine serum (cat. no. 16000-044; Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin (cat. no. C0222; Beyotime Institute of Biotechnology). The cells were incubated at 37°C in presence of 5% CO₂. The circRNA-0004904-targeting small interfering (si)RNA (siRNA-001 and siRNA-002), negative control siRNA (siNC), and lentivirus-mediated circRNA-0004904 overexpression (OE) vector with its corresponding negative control (OE-NC) were supplied by Shanghai GenePharma Co., Ltd. The sequences of the siRNAs were as follows: siRNA-001, 5′-GCA AAT TCC AGA GAG CAC GTT-3′; siRNA-002, 5′-AAA TTC CAG AGA GCA CGT TTT-3′; and siNC, 5′-UUC UUC GAA GGU GUC ACG UTT-3′. Plasmid expression vectors for the overexpression of fused in sarcoma (FUS) and autophagy-related 12 (ATG12) (accession nos. NM_004960 and NM_004707, respectively) were constructed by GeneCopoeia, Inc. In addition, siATG7 (5′-GGU CAA AGG ACG AAG AUA ACA-3′), siFUS-1 (5′-GGC UAU GGA ACU CAG UCA ACU-3′), siFUS-2 (5′-GGA CAG CAG CAA AGC UAU AAU-3′), hsa-miR-570-3p mimics (cat. no. HmiR-SN0669), mimics control (cat. no. CmiR-SN0001-SN), hsa-miR-570-3p inhibitor (cat. no. HmiR-AN0669-SN-10) and inhibitor control (cat. no. CmiR-AN0001-SN) were obtained from GeneCopoeia, Inc. For transfections, cells were seeded in a 6-well plate (1×10⁵ cells/well). When the cells were ~80% confluent, 2 µg plasmids were transfected using the EndoFectin™ MAX transfection reagent (cat. no. EFM1004-01; GeneCopoeia, Inc.) for 24 h at 37°C. The mRNA or protein levels were detected at 48 h post-transfection.

A circRNA-0004904 probe was used for RNA pull-down assay (cat. no. A04001; GenePharma Co., Ltd.) as previously described (24). Cells were lysed using RIPA Lysis Buffer (cat. no. P0013D, Beyotime Biotechnology). RNA pulldown assay was performed using Pierce™ Magnetic RNA-Protein Pull-Down Kit (cat. no. 20164; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions with 20 µl lysate per reaction. The miRNA level was detected by reverse transcription-quantitative (RT-q)PCR.

RT-qPCR was performed as previously described (24). Total RNA was extracted from cells and tissues using TRIzol^® reagent (Takara Biotechnology Co., Ltd.) according to the manufacturer's instructions. For miRNA analysis, cDNA was obtained using the All-in-One™ miRNA First-Strand cDNA Synthesis kit (cat. no. QP013; GeneCopoeia, Inc.), and RT-qPCR was performed using the All-in-One™ miRNA qRT-PCR Detection kit (cat. no. QP015; GeneCopoeia, Inc.). U6 small nuclear RNA was used as an endogenous control for normalization. For circRNA analysis, cDNA was synthesized from 1 µg total RNA using the All-in-One First Strand Synthesis kit (cat. no. AORT-0020; GeneCopoeia, Inc.). The real-time PCR kit was purchased from GeneCopoeia, Inc. qPCR was performed under the following conditions: Initial denaturation at 95°C for 10 min, followed by 35 cycles of 95°C for 10 sec and 60°C for 20 sec, and a final extension at 72°C for 10 sec. qPCR was conducted using the PCR System (Bio-Rad Laboratories, Inc.) according to the manufacturer's instructions. Each experiment was performed in triplicate. Gene expression analysis was performed using the 2^−ΔΔCq method (25). The circ-0004904 primers (5′-CCT CCG GTG ATA GGT TCT CA-3′ and 3′-TAA ATT TGG TGC AGG GTG GT-5′) were synthesized by Shanghai GenePharma Co., Ltd. Primers for ATG12 (cat. no. HQP055328), GAPDH (cat. no. HQP064347), miR-570 (cat. no. HmiRQP0669) and U6 (cat. no. HmiRQP9001) were purchased from GeneCopoeia, Inc.

Protein expression levels of ATG12, GAPDH, LC3-II, P62, FUS and vascular endothelial growth factor (VEGF) were detected by western blotting. The transfected HTR-8 and JEG-3 cells were cultured in 6-well plates and lysed at 48 h post-transfection with radio-immunoprecipitation assay lysis buffer (MilliporeSigma) containing phosphatase and protease inhibitors. Protein quantification was performed using BCA assay. Proteins (20 µg/lane) were separated by 10% SDS-PAGE and transferred to PVDF membranes. The membranes were blocked with 5% BSA for 1 h at room temperature and incubated overnight at 4°C with the following primary antibodies (dilution, 1:1,000): Rabbit monoclonal anti-ATG12 (cat. no. ab109491; Abcam), rabbit recombinant monoclonal GAPDH (cat. no. ab181602; Abcam), rabbit recombinant monoclonal LC3-II (cat. no. ab192890; Abcam), rabbit recombinant monoclonal anti-P62 (cat. no. ab109012; Abcam), rabbit recombinant monoclonal FUS (cat. no. ab124923; Abcam) and rabbit recombinant monoclonal VEGF (cat. no. AB1876-I; Sigma-Aldrich; Merck KGaA). Following washing for 3×10 min using the Western Wash Buffer (cat. no. P0023C3; Beyotime Institute of Biotechnology) at room temperature, HRP-conjugated secondary antibodies (1:2,000; cat. no. 7074; Cell Signaling Technology, Inc.) were applied to the membranes for 1 h at room temperature, and chemiluminescent signals of membranes were measured by using an ImageQuant LAS system (GE Healthcare). Densitometric analysis was performed using Image Lab (version 3.0, Bio-Rad Laboratories). All experiments were performed in triplicate and repeated three times.

Matrigel invasion assay was performed to assess the invasive ability of HTR8 or JEG3 cells as previously described (26,27). The upper chamber of the 24-well Transwell insert was pre-coated with 100 µl diluted Matrigel at 37°C for 5 h. A total of 1×10⁵ HTR8 or JEG3 cells resuspended in serum-free medium were seeded into the upper chamber. The lower chambers contained medium with 10% fetal bovine serum. The cells were incubated at 37°C for 24 h. Subsequently, cotton swabs were used to remove the non-invasive cells. The migrated cells attached to the lower surface of the membrane were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet for 15 min at room temperature. The cells that migrated to the lower chamber were observed under an optical microscope (×20 magnification). Finally, the crystal violet on the membrane was solubilized with 500 µl 33% acetic acid solution. The optical density of the plates was measured at 595 nm wavelength using a microplate reader.

The source of template was total RNA of circRNA0004904 (sequence, 5′-AGA GCA CGT TTT CAA TAT CAT AGG AGC ATT TGA TAT TCC ACG CTT TGT GTA CAA TTC AGA AAG AAA AAA ATT TCTT CC TCTGTTAATGAC CAA CCA CCC TGC ACC AAATTTATT TGG AAC ACC AAG AGA TAA AGC AGA GAT GTT TCG TGA GCG ATA TAC CAT TTT GCA CCA GAG GAC CCA CAG GCA TGA ATT ATT TAC TCC TCC GGT GAT AGG TTC TCA CCC TGA TGA AAG CGG AAG CAA ATT CCA G-3′). A Cy3-labeled probe against circRNA0004904 (sequence, 5′-CUC CUA UGA UAU UGA AAA CGU GCU CUC UGG AAU UUG CUUC CGC UUU CAU C-3′) and a FITC-labeled probe against miR-570 (sequence, 5′-CAA AGG UAA UUG CUG UUU UCG-3′) were synthesized by Shanghai GeneChem Co., Ltd. The FISH assay was conducted according to the manufacturer's instructions provided with a FISH kit (cat. no. F32201; Shanghai GeneChem Co., Ltd.) (28,29). HTR8 cells (2×10⁵) were fixed with absolute ethanol to in a 48-well plate at room temperature for 15 min, followed by treatment with 100 µl 0.1% Triton x-100 at room temperature for 15 min. Gradient ethanol dehydration (70, 85 and 100%). Subsequently, denaturation was performed at 73°C for 5 min. A total of 100 µl probe mixture was added and incubated overnight at 37°C. The next day, the probe mixture was removed, 100 µl 0.4X saline sodium citrate/0.3% Tween-20 was added, and the cells were washed twice at room temperature for 5 min. Finally, 100 µl Hoechst stain was added to each well for 15 min at room temperature. Confocal images were acquired using a laser-scanning confocal microscope (×400 magnification). A total of five fields were imaged per sample. All experiments were performed in triplicate and repeated three times.

Dual-luciferase receptor assay was performed using the Dual-Luciferase Reporter Assay system (Promega Corporation). In brief, 2.5×10⁵ HTR8 cells were harvested in a 24-well plate. A total of 2 µg wild-type or mutant circ-000490 reporter was constructed by Shanghai GeneChem Co., Ltd. and co-transfected with 100 nM miR-570 or miR-NC and the Renilla plasmid using EndoFectin™ MAX transfection reagent (cat. no. EFM1004-01; GeneCopoeia, Inc.) for 24 h at 37°C. Luciferase activity was detected at 48 h post-transfection. Firefly luciferase activity was normalized to that of Renilla luciferase. All experiments were performed in triplicate and repeated three times.

A total of 1×10⁵ JEG3 cells were co-transfected with mRFP-GFP-LC3 and siNC, siRNA-001 or siRNA-002 using the EndoFectin™ MAX transfection reagent for 24 h at 37°C and cultured for 22 h. Subsequently, the cells were treated with Earle's balanced salts solution for 2 h at 37°C. HTR8 cells were co-transfected with mRFP-GFP-LC3 and OE or OE-NC using the EndoFectin™ MAX transfection reagent for 24 h at 37°C and analyzed by confocal microscopy after 24 h. LC3 dots were quantified using the image pro-plus 6.0 software. All experiments were performed in triplicate and repeated three times. To evaluate the autophagic flux, 3-MA and chloroquine (CQ) were used. At 48 h post-transfection, 5 mM 3-MA or 50 µM CQ were added to the medium for 12 h at 37°C, and LC3 protein expression levels were detected by western blotting.

A total of 1×10⁵ HTR-8 cells were seeded into a 24-well plate and treated with 200 µl EdU (dilution, 1:1,000) in the dark for 1 h at 37°C using the BeyoClick™ EdU Cell Proliferation kit (Beyotime Institute of Biotechnology) according to the manufacturer's instructions. The cells were washed with PBS (5 min three times). Hoechst 33342 (Beyotime Institute of Biotechnology) was used for nuclear staining (30). Subsequently, the cells were visualized under a fluorescence microscope (×200 magnification; Olympus Corporation). A total of five fields were analyzed per sample.

The interaction between circ-0004904 and miRNAs was predicted using the CircInteractome database (https://circinteractome.nia.nih.gov/) (31). The binding site between ATG12 and miR-570 was predicted by TargetScan (32,33).

All statistical analyses were performed using SPSS software version 17.0 (SPSS, Inc.). Data are presented as the mean ± SD or SEM. Statistical analysis was performed by the unpaired Student's t-test or one-way ANOVA with Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

The expression levels of circ-0004904 were detected in the placental tissues and plasma samples of 30 patients with preeclampsia and 30 normal parturients with the same gestational age. The results demonstrated that the expression levels of circ-0004904 were upregulated in the placental tissues and plasma samples of patients with preeclampsia compared those in the control group (Fig. 1A and B).

The roles of circ-0004904 were investigated in HTR8 and JEG3 cells. First, the expression levels of circ-0004904 were detected in HTR8 and JEG3 cells. The results revealed that the expression levels of circ-0004904 in JEG3 cells were higher compared with those in HTR8 cells. Silencing or ectopic expression of circ-0004904 significantly inhibited or elevated the expression levels of circ-0004904, respectively (Fig. S1A and B). By contrast, silencing or ectopic expression of circ-0004904 did not affect apoptosis (data not shown). The results also demonstrated that silencing of circ-0004904 repressed LC3-II expression, and promoted the proliferation and migration of JEG3 cells (Figs. 2A, C and S2). By contrast, ectopic expression of circ-0004904 upregulated the levels of LC3-II expression, and inhibited the proliferation and migration of HTR8 cells (Figs. 2B, D and S2). In order to evaluate the effects of circ-0004904 on autophagic flux, 3-MA and CQ were used; the results demonstrated that autophagy mediated by circ-0004904 was inhibited by 3-MA, but not CQ in HTR8 cells (Fig. 2E and F), which suggested that circ-0004904 affected autophagosomes rather than autophagolysosomes. In addition, mRFP-GFP-LC3 was used to detect autophagic flux, and the results revealed that silencing of circ-0004904 inhibited autophagic flux in JEG3 cells (Fig. 2G), whereas overexpression of circ-0004904 promoted autophagic flux in HTR8 cells (Fig. 2H).

Since circ-0004904 promoted autophagic flux in HTR8 cells, the present study further assessed whether circ-0004904-mediated autophagy may regulate cell proliferation and invasion. The results demonstrated that ectopic expression of circ-0004904 inhibited the proliferation and invasion of HTR8 cells. However, following the silencing of ATG7 or the addition of 3-MA, cell proliferation and invasion was partially promoted (Fig. 3A-D). These results suggested that circ-0004904-mediated autophagy inhibited cell proliferation and invasion.

Bioinformatics analysis predicted that circ-0004904 interacted with a number of miRNAs, including miR-494, miR-513a-3p, miR-570, miR-572, miR-589 and miR-876-3p. The results of the RNA pull-down assays confirmed that circ-0004904 directly bound miR-570 (Fig. 4A). In addition, circ-0004904 was mainly located in the cytoplasm, and circ-0004904 and miR-570 were co-localized in the cytoplasm in HTR8 cells (Fig. 4B). Following silencing of circ-0004904, the expression levels of miR-570 were increased in JEG3 cells compared with those in the cells transfected with siNC (Fig. 4C). By contrast, ectopic expression of circ-0004904 inhibited the expression levels of miR-570 in HTR8 cells compared with those in cells transfected with the OE-NC vector (Fig. 4D). Co-transfection with miR-570 mimics and a wild-type circ-0004904 vector inhibited luciferase activity compared with that observed following transfection with the NC mimics in HTR8 cells (Fig. 4E).

Based on the results of the bioinformatics analysis, ATG12 was selected as a potential target of miR-570. Transfection with miR-570 inhibitors or mimics significantly inhibited or elevated the levels of miR-570 compared with those in the corresponding NC groups (Fig. S1C and D). The results also revealed that transfection with the miR-570 mimics downregulated the mRNA and protein expression levels of ATG12 in HTR8 cells compared with those in the NC group (Fig. 5A), whereas transfection with the miR-570 inhibitors upregulated the expression levels of ATG12 (Fig. 5B). The dual-luciferase reporter assay confirmed that ATG12 was the target of miR-570 in HTR8 cells (Fig. 5C). In addition, circ-0004904 positively regulated the expression levels of ATG12 in HTR8 cells (Fig. 5D and E). However, following transfection with the miR-570 inhibitors, this regulation was partly blocked in HTR8 cells (Fig. 5F and G).

The present study further attempted to determine whether circ-0004904 regulated autophagy via ATG12. Silencing or ectopic expression of ATG12 significantly inhibited or elevated, respectively, the mRNA and protein levels of ATG12 compared with those in the corresponding NC groups (Fig. S1E and F). Further results demonstrated that overexpression of circ-0004904 promoted the expression of LC3-II in HTR8 cells compared with that in the NC group, whereas following silencing of ATG12, this regulation was partly suppressed (Fig. 6A). In addition, silencing of circ-0004904 inhibited the expression of LC3-II in JEG3 cells compared with that in the cells transfected with siNC; however, following overexpression of ATG12, the downregulation was reversed (Fig. 6B).

The present study identified that circ-0004904 directly bound to the FUS protein by bioinformatics analysis. A previous study has reported that FUS inhibits VEGF expression (34); thus, we hypothesized that circ-0004904 may regulate the FUS/VEGF axis. RNA pull-down assay confirmed that circ-0004904 directly bound to the FUS protein (Fig. 7A). Additionally, following silencing of circ-0004904, the expression levels of FUS decreased, whereas the expression levels of VEGF increased in JEG3 cells compared with those in cells transfected with siNC (Fig. 7B). Following ectopic expression of circ-0004904 in HTR8 cells, the expression levels of FUS increased, whereas the levels of VEGF reduced compared with those in the NC group (Fig. 7C). The results also demonstrated that silencing or ectopic expression of FUS significantly inhibited or elevated, respectively, the mRNA and protein levels of FUS compared with those in the corresponding NC groups (Fig. S1G and H). In addition, silencing of FUS increased the expression levels of VEGF compared with those in the siNC group, whereas overexpression of FUS exerted an opposite effect (Fig. 7D and E). When FUS was silenced or overexpressed, this regulation was partially reversed, suggesting that the regulatory effect of circ-0004904 on VEGF protein was dependent on FUS (Fig. 7F and G).