The human immortalized endometrial stroma telomerase-immortalized human endometrial stromal cells (T-HESCs) and HTR-8/SVneo (HTR8) trophoblast cells were maintained in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.), 1% glutamine, 1% sodium pyruvate and 1% penicillin/streptomycin under a humidified atmosphere containing 5% CO₂ at 37°C. The HTR8 cell line was established using extravillous cytotrophoblasts freshly isolated from a first-trimester placenta and transfected with a plasmid containing the simian virus 40 large T antigen (32). This cell line contains two populations, one of epithelial and one of mesenchymal origin (33). For triggering decidualization in vitro, T-HESCs were treated with 0.3 mM 8-Bromo-cAMP (cat. no. BML-CN115; Enzo Life Sciences) and 10 µM medroxyprogesterone 17-acetate (MPA) at 37°C for 72 h. To block centrosome amplification, cells were treated with 250 nM Centrinone (LCR-263, cat. no. HY-18682; MedChemExpress) at 37°C for 72 h. To activate p38, cells were treated with 1 µM sodium salicylate (cat. no. S3007; MilliporeSigma) at 37°C for 72 h.

A commercially available siRNA against autophagy-related 16-like 1 (ATG16L1; SMARTpool; GE Healthcare Dharmacon, Inc.), and miRNA mimics or inhibitors (miRIDIAN microRNA; GE Healthcare Dharmacon, Inc.) were purchased from Horizon Discovery. The siRNA and miRNA sequences were as follows: SMARTpool-siATG16L1 (5′-UGUGGAUGAUUAUCGAUUA-3′; 5′-GGCACACACUCACGGGACA-3′; 5′-GCAUUGGAUUAACGGAAAC-3′; 5′-GUUAUUGAUCUCCGAACAA-3′); miR-20b-5p mimic (5′-CAAAGUGCUCAUAGUGCAGGUAG-3′); miR-155-5p mimic (5′-UUAAUGCUAAUCGUGAUAGGGGU-3′); miR-718 inhibitor (5′-CUUCCGCCCCGCCGGGCGUCG-3′); and miR-155-5p inhibitor (5′-UUAAUGCUAAUCGUGAUAGGGGU-3′). The negative controls of miRNA mimics and inhibitors were commercially available. The miRNA mimic negative control (5′-UCACAACCUCCUAGAAAGAGUAGA-3′; cat. no. MIMAT0000039) was purchased from Horizon Discovery (GE Healthcare Dharmacon, Inc.). miRNA inhibitor negative control (5′-UUGUACUACACAAAAGUACUG-3′; cat. no. MIMAT0000295) was purchased from Horizon Discovery (GE Healthcare Dharmacon, Inc.). The scrambled siRNA with the target sequence 5′-GAUCAUACGUGCGAUCAGA-3′ was purchased from Sigma (MilliporeSigma).

Cells were transfected with 50 nM miRNA or siRNA using Lipofectamine 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C for 72 h. Samples were collected and analyzed immediately. To investigate the role of p38 in modulating primary cilia, SB203580 (working concentration, 10 µM; Abcam) was used to inhibit the function of p38 at 37°C for 72 h.

Cells were trypsinized and lysed with CelLytic™ MT cell lysis reagent (MilliporeSigma) containing protease inhibitor for 10 min on ice. Lysates were centrifuged at 13,000 × g for 10 min at 4°C and the supernatant was collected. The protein concentration was determined using the Bio-Rad protein quantity assay (Bio-Rad Laboratories, Inc.) by BioSpectrometer® (Eppendorf AG, Germany). The samples (50 µg/lane) were mixed with 2X sample buffer and heated to 100°C for 10 min. Equal amounts of proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis at 150 V for 90 min. Following gel separation, the samples were transferred to PVDF membranes at 20 V overnight. Next, the membrane was blocked with 3% BSA (MilliporeSigma) dissolved in TBST (0.005% Tween) at room temperature for 1 h and incubated with the following antibodies: Anti-Akt (cat. no. #4691; Cell Signaling Technology, Inc.), anti-phospho-Akt (Thr308; cat. no. #9275; Cell Signaling Technology, Inc.), anti-ATG16L1 (cat. no. #8089S; Cell Signaling Technology, Inc.), anti-cleaved-caspase 3 (cat. no. #9664; Cell Signaling Technology, Inc.), anti-cleaved-poly (ADP-ribose) polymerase (PARP; cat. no. #9542; Cell Signaling Technology, Inc.), anti-cyclin A (cat. no. ab38; Abcam), anti-cyclin D (cat. no. 2G3G5; Thermo Fisher Scientific, Inc.), anti-cyclin E (cat. no. #4132; Cell Signaling Technology, Inc.), anti-1A/1B light chain 3B (LC3A/B; cat. no. #12741; Cell Signaling Technology, Inc.), anti-matrix metalloproteinase 2 (MMP2; cat. no. #40994; Cell Signaling Technology, Inc.), anti-p38 (cat. no. #9212; Cell Signaling Technology, Inc.), anti-phospho-p38 (cat. no. 09-272; Merck KgaA), anti-p44/42 mitogen-activated protein kinase (ERK1/2; cat. no. #4695; Cell Signaling Technology, Inc.), anti-phospho-ERK1/2 (cat. no. #4370; Cell Signaling Technology, Inc.) and anti-actin (cat. no. ab8227; Abcam) in 1:10,000 dilutions at 4°C overnight. After washing with TBST for 30 min, the membrane was incubated with an HRP-conjugated secondary antibody (ab205718, Abcam) in 1:7,000 at room temperature for 1 h. Subsequently, the signals were detected by the ECL kit (MilliporeSigma) with actin as the loading control and analyzed by ImageJ (version 1.53k, National Institutes of Health, USA).

Cells were fixed and permeabilized in ice-cold pure methanol (MilliporeSigma) for 5 min at −20 °C. After permeabilization, cells were incubated in blocking buffer [containing 0.2% Triton X-100, 0.05% Tween-20 and normal goat serum (MilliporeSigma)] for 1 h at room temperature and incubated with the following antibodies: Anti-acetylated tubulin (cat. no. T6793; MilliporeSigma), anti-ADP-ribosylation factor-like 13B (ARL13B; cat. no. 17711-1-AP; Proteintech Group, Inc.), anti-centrosomal protein 164 (CEP164; cat. no. NBP1-81445; Novus Biologicals, LLC), anti-intraflagellar transport 88 (IFT88; cat. no.13967-1-AP; Proteintech Group, Inc.) and anti-γ-tubulin (cat. no. T6557; MilliporeSigma) in 1:700 dilutions at 4°C for overnight. After which, the cells were washed for 30 min with PBS three times and were then incubated with FITC- or Cy3-conjugated goat anti-mouse or anti-rabbit IgG secondary antibodies (cat. nos. ab175473 and ab150077; Abcam) in 1:700 dilutions in the presence of 6-diamino-2-phenylindole at room temperature for 1 h at room temperature. The cells were then further washed for 30 min with PBS three times. The coverslips were overlaid on 50% glycerol in PBS and the signals were observed through an Axio imager D2 fluorescence microscope and analyzed by ZEN 2 blue edition software (Zeiss AG).

Total RNA was extracted using TRI reagent (MilliporeSigma) and Direct-zol™ RNA miniprep (Zymo Research Corp.). RNA (1 µg) was then reverse transcribed into cDNA using an RT kit (sensiFAST; Bioline) according to manufacturer's protocol. Fast SYBR Green Master Mix (MilliporeSigma) was used to perform qPCR with a final concentration of 0.25 nM gene-specific primers. The sequences of the primers were as follows: E-cadherin forward: 5′-CCGAGAGCTACACGTTC-3′, reverse: 5′-TCTTCAAAATTCACTCTGCC-3′; fibronectin-1 forward: 5′-CCATAGCTGAGAAGTGTTTTG-3′, reverse 5′-CAAGTACAATCTACCATCATCC-3′; MMP2 forward: 5′-GTGATCTTGACCAGAATACC-3′, reverse: 5′-GCCAATGATCCTGTATGTG-3′; thrombospondin 1 (THBS1) forward: 5′-GTGACTGAAGAGAACAAAGAC-3′, reverse 5′-CAGCTATCAACAGTCCATTC-3′; and zinc-finger E-box-binding homeobox 1 (ZEB-1) forward: 5′-AAAGATGATGAATGCGAGTC-3′, reverse: 5′-TCCATTTTCATCATGACCAC-3′. PCR signals were compared among groups after normalization, with GAPDH forward: 5′-TCGGAGTCAACGGATTTG-3′, reverse: 5′-CAACAATATCCACTTTACCAGAG-3′) used as the internal reference, calculated using the 2^−ΔΔCq method (18,34,35). The thermocycling conditions were as follows: 95 °C for 10 min. Next, they were denatured at 95°C for 15 sec and annealing the primer with DNA template at 60°C for 1 min. These cycles were repeated for 40 times.

HTR8 cells were plated on 6-well plates and transfected with miR-155-5p mimic and/or inhibitor prior to the migration assay. A wound was created by scratching the cells with a sterile 200-µl pipette tip, and fresh medium (serum-free) was added to the cells after washing with PBS. The microscopy images of the cells migrating into the wound area were captured at ×10 magnification under an optical microscope at 0 and 12 h after scraping, in order to record the wound closure area. Wound closure was calculated using the following formula: Wound closure (%)=100-(initial area of wound-nth h area of the wound)/(initial area of the wound).

The cell invasion assay was performed using Falcon cell culture inserts (pore size, 8 µm) according to the manufacturer's protocol. Matrigel (1 mg/ml at 37°C; Corning, Inc.) was loaded onto the upper chambers of the inserts for 30 min, whereas the lower chamber contained a serum-loaded medium as an attractant. Cells were seeded and overlaid on the Matrigel at a density of 3×10⁴ cells/well in serum-free medium. Upon termination of the incubation at 37°C for 24 h, invasive cells were fixed in 100% methanol for 4 min and stained with Giemsa's azure eosin methylene blue solution (Merck KGaA) at room temperature for 2 h at room temperature and cell numbers were counted in six high-power fields per insert under an optical microscope. The experiments were performed in triplicate and repeated three times.

HTR8 cells were plated at a density of 1×10⁵ cells on a 6-well plate. After incubating cells for 24, 48 and 72 h, cells were trypsinized and resuspended with 1 ml PBS for cell counting with hemocytometer. All treatments were carried out in triplicate, with each experiment performed three times.

HTR8 cells were seeded on 25×25-mm coverslips in 6-well tissue culture plates and transfected with the miR-155-5p mimic and/or inhibitor. The cell proliferation ability was measured using Click-iT® EdU Imaging Kits (Thermo Fisher Scientific, Inc.). Briefly, 1 µl 10 µM EdU working solution (Component A) was added and incubated at 37°C for 1 h. After which, the cells were permeabilized in ice-cold methanol for 5 min. Subsequently, 1X Click-iT® EdU buffer was added, as per the manufacturer's instructions, and incubated at 37°C for 1 h. The cells were then washed for 30 min with PBS three times and the signals were observed under an Axio imager D2 fluorescence microscopy (Zeiss AG).

Supernatants from different experimental groups were analyzed using insulin-like growth factor binding protein-1 (IGFBP1; cat. no. ab233618, Abcam) and prolactin (cat. no. ab226901; Abcam) ELISA kits according to the manufacturer's instructions. Briefly, supernatants were added to each well followed by addition of 50 µl antibody cocktails and shaking at 30 × g at 37°C for 1 h. Then, each well was washed with 350 µl PBS three times. After washing, 50 µl TMB development solution was added to each well, followed by the termination of the reaction via the addition of a stop solution. The solution was measured at 450 nm.

The predicted target gene of miR-20b-5p was conducted by 2.0 of the miRNA Pathway Dictionary Database (miRPathDB), which was freely accessible at https://mpd.bioinf.uni-sb.de/. MiRPathDB 2.0 performed analysis of the underlying metabolic pathway and potential miRNA targets, and with customized application, generated the relation chart of predicted centrosome-associated target genes of miR-20b-5p.

Values from the experimental assays are expressed as the mean ± SEM of ≥3 independent experimental repeats. Differences between two groups or >2 groups were compared using unpaired Student's t-test or one-way ANOVA with Dunnett's post hoc test, respectively. Data were analyzed using GraphPad Prism version no.9 (Dotmatics). P<0.05 was considered to indicate a statistically significant difference. All experiments are repeated at least for three times.

Aberrant expression levels of miR-20b-5p, miR-155-5p and miR-718 have been observed in women who suffer from recurrent miscarriage (19). To investigate the underlying molecular mechanisms, uterine endometrial differentiation, and trophoblast growth and invasion, which are essential events for successful pregnancy, were examined. It was first checked whether aberrant expression levels of these miRNAs affected endometrial stromal cell growth. Overexpression of miR-20b-5p and miR-155-5p were used to mimic the upregulation of these two miRNAs, and transfection with a miR-718 inhibitor was used to induce downregulation of miR-718 in T-HESCs. T-HESCs were transfected with miR-20b-5p, miR-155-5p and miR-718 inhibitor (Fig. 1A-C), and cell numbers were counted. Transfection with all three miRNAs had no effects on T-HESC growth (Fig. 1D), suggesting that miR-20b-5p, miR-155-5p and miR-718 inhibitor did not affect endometrial stromal cell growth. After which, the ability of uterine endometrial stromal differentiation, known as decidualization, was examined. Decidualization was induced by treating T-HESCs with cAMP and MPA for 3 days, followed by examining the decidual markers IGFBP1 and prolactin. Prolactin and IGFBP1 levels were significantly increased upon cAMP and MPA co-treatment (Fig. 1E and F). After which, the effects of miR-20b-5p, miR-155-5p and miR-718 inhibitor on decidualization were examined. Transfection of cells with miR-20b-5p, miR-155-5p or miR-718 inhibitor did not affect the levels of prolactin and IGFBP1 (Fig. 1G and H), suggesting that these miRNAs did not disturb endometrial decidualization.

Trophoblast proliferation contributes to establishing the maternal-fetal interface. It was assessed whether miR-20b-5p, miR-155-5p and miR-718 inhibitor affected trophoblast cell growth. The immortalized human trophoblast HTR8 cell line was used in the present study. Overexpression of miR-20b-5p and inhibition of miR-718 did not affect HTR8 cell growth; however, when cells overexpressed miR-155-5p, the number of cells was significantly reduced and apoptotic bodies were observed, suggesting that miR-155-5p inhibited trophoblast cell growth (Fig. 2A and B). The levels of cyclins affect cell cycle progression; therefore, the expression levels of different cyclins were assessed. Overexpression of miR-155-5p did not affect the expression levels of cyclin A and cyclin D; however, the levels of cyclin E were reduced, and this phenotype was rescued by co-transfecting cells with the miR-155-5p inhibitor (Fig. 2D). The expression levels of miR-155-5p were also confirmed by performing RT-qPCR (Fig. 2C). These findings indicated that miR-155-5p reduced the expression of cyclin E. Cyclin E regulates S phase entry (23); therefore, whether miR-155-5p decreases the ability of HTR8 cells to enter the S phase was assessed by EdU incorporation assay. Overexpression of miR-155-5p reduced the proportion of EdU⁺ cells, and this phenotype was rescued by co-transfecting cells with the miR-155-5p inhibitor (Fig. 2E and F). Thus, miR-155-5p may reduce the S phase entry of HTR8 cells. The ability of cells to enter the M phase was next examined by assessing the mitotic index. Overexpression of miR-155-5p reduced the mitotic index of HTR8 cells, and this was rescued by co-transfecting cells with miR-155-5p inhibitor (Fig. 2G). Taken together, miR-155-5p inhibited trophoblast HTR8 cell growth by reducing S phase and M phase entry.

In addition to cell cycle progression, some apoptotic bodies were observed in miR-155-5p-overexpressing cells; therefore, cell apoptosis was assessed. Cleaved-caspase 3 and cleaved-PARP, which are markers of apoptosis, were examined in miR-155-5p-overexpressing cells. Overexpression of miR-155-5p increased the levels of cleaved-caspase 3 and cleaved-PARP (Fig. 3A and B), and this phenotype was rescued by co-transfection with the miR-155-5p inhibitor. These data suggested that miR-155-5p induced cell apoptosis.

Aberrant mitosis triggers cell apoptosis (36); therefore, whether miR-155-5p-induced apoptosis by aberrant mitosis was assessed. The mitotic spindle poles were examined by staining cells with γ-tubulin. Under normal conditions, cells contained two mitotic spindle poles and the chromosomes were aligned in the middle of the cells (Fig. 3C, left panel); however, when cells overexpressed miR-155-5p, multiple mitotic spindle poles (>2 γ-tubulin spots) and misaligned chromosomes were observed (Fig. 3C, right panel; and Fig. 3D). This phenotype was rescued when cells were co-transfected with the miR-155-5p inhibitor (Fig. 3D). Thus, these data suggested that miR-155-5p triggered aberrant mitosis. Aberrant mitosis is caused by centrosome amplification (cells with >3 centrosomes) during interphase (37); therefore, the centrosome copy numbers during interphase were next assessed. Under normal conditions, cells contained two centrosomes (Fig. 3E, left panel); however, overexpression of miR-155-5p induced centrosome amplification (Fig. 3E, right panel). Co-transfection with the miR-155-5p inhibitor alleviated miR-155-5p-induced centrosome amplification (Fig. 3F), supporting the hypothesis that miR-155-5p induced centrosome amplification. Whether centrosome amplification contributed to apoptosis upon miR-155-5p overexpression was next examined. Overexpression of miR-155-5p increased the levels of cleaved-caspase 3 and cleaved-PARP; however, inhibition of centrosome amplification by treating cells with the centrosome inhibitor centrinone rescued these phenotypes (Fig. 3H and I), suggesting that centrosome amplification promoted cell death. Using bioinformatics analysis, several genes, such as GSK3β, PLK3, or KIF3A, were predicted as the targets of miR-155-5p. GSK3β was identified as a putative candidate of miR-155-5p (Fig. 3G). A previous study showed that the downregulation of GSK3β led to the accumulation of β-catenin followed by induction of centrosome amplification (38,39). Thus, miR-155-5p may target GSK3β to induce centrosome amplification. Taken together, miR-155-5p induced centrosome amplification, thus triggering trophoblast cell apoptosis.

Trophoblast cell migration and invasion are critical for placenta formation during early pregnancy (30); thus whether overexpression of miR-155-5p impeded trophoblast migration and invasion was examined. Since miR-155-5p inhibited HTR8 cell growth, seeding cell numbers were adjusted according to the growth assay. A total of 5×10⁵ control and 7×10⁵ miR-155-5p-transsfected cells were seeded to perform the migration and invasion assays. First, the effects of miR-155-5p on trophoblast cell migration was assessed by performing a wound healing assay. Overexpression of miR-155-5p inhibited trophoblast cell migration, and this phenotype was rescued by co-transfecting cells with the miR-155-5p inhibitor (Fig. 4A and B), suggesting that upregulation of miR-155-5p reduced trophoblast cell migration. Next, trophoblast cell invasion was examined. Consistent with the results of the cell migration assay, the invasive ability of trophoblast cells was reduced by miR-155-5p overexpression (Fig. 4C and D), suggesting that miR-155-5p inhibited cell invasion. Collectively, miR-155-5p inhibited HTR8 trophoblast cell migration and invasion. EMT is key for trophoblast cell invasion during early pregnancy (18). Since miR-155-5p inhibited trophoblast cell invasion, the EMT was examined in the present study. In miR-155-5p-overexpressing cells, the epithelial marker E-cadherin was increased, whereas the mesenchymal markers ZEB-1, fibronectin-1 and THBS1 were significantly decreased, suggesting that miR-155-5p inhibited EMT (Fig. 5A-D). By contrast, co-transfection with the miR-155-5p inhibitor rescued the expression levels of the EMT markers. Taken together, miR-155-5p inhibited trophoblast cell invasion by reducing EMT.

Primary cilia promote trophoblast cell invasion; therefore, whether miR-155-5p affected primary cilia formation was assessed. HTR8 cells grew primary cilia, as evidenced by several markers, including markers of the axoneme (acetylated tubulin), ciliary membrane (ARL13B), intraflagellar transporter (IFT88) and basal body (CEP164; Fig. 6A-C). Overexpression of miR-155-5p reduced the proportions of ciliated cells, and this phenotype was rescued by co-transfection of cells with the miR-155-5p inhibitor (Fig. 6D and E). These data suggested that miR-155-5p reduced primary cilia formation. Next, which signaling contributed to primary cilia formation was assessed. AKT, ERK and p38 are known to regulate primary cilia formation; therefore it was next assessed whether these signaling pathways were activated and participated in miR-155-5p-mediated primary cilia formation. Overexpression of miR-155-5p did not affect AKT and ERK activation; however, phospho-p38 (active form) was reduced, and this phenotype was rescued by co-transfecting cells with the miR-155-5p inhibitor (Fig. 7A-C). Thus, miR-155-5p inhibited p38 activation. It was then examined whether p38 reduced primary cilia formation. HTR8 cells grew primary cilia; however, in response to treatment of HTR8 cells with the p38 inhibitor SB203580, the proportions of ciliated cells were decreased (Fig. 7D). It was also observed that the p38 inhibitor reduced invasive ability (Fig. 7E), suggesting that p38 activity was required for trophoblast invasion. To confirm this finding further, miR-155-5p-overexpressing cells were treated with the p38 activator and found activation of p38 restored the primary cilia formation and invasive ability (Fig. 7F and G). Taken together, miR-155-5p inactivated p38, thus reducing primary cilia formation and trophoblast invasion.

Autophagic homeostasis maintains trophoblast cell growth and invasion. It was then checked whether autophagy is mediated by miR-155-5p. The levels of LC3 II were increased in miR-155-5p-overexpressing cells, whereas this phenotype was rescued by co-transfection with the miR-155-5p inhibitor, suggesting that miR-155-5p induced autophagy (Fig. 8A). Next, whether autophagy contributed to cell death was assessed. ATG16L1, an important regulator of autophagy, was depleted by transfecting HTR8 cells with siRNA against ATG16L1 (Fig. 8B). In addition, the cell death and cell cycle pathways were examined. Depletion of ATG16L1 reduced cleaved-caspase 3 and cleaved-PARP, whereas the cyclin E expression was increased in ATG16L1-deficient cells (Fig. 8C and D). These data indicated that autophagy triggered cell death and inhibited cell cycle progression. Degradation of the extracellular matrix by MMP2 promotes cell invasion; therefore, the effect of miR-155-5p on MMP2 expression was examined. Overexpression of miR-155-5p reduced the mRNA and protein expression levels of MMP2, whereas co-transfection with the miR-155-5p inhibitor rescued the expression of MMP2 (Fig. 8E and F). Thus, miR-155-5p may inhibit MMP2 expression. Next, whether MMP2 was regulated by autophagy was assessed. The levels of MMP2 were increased in ATG16L1-deficient cells (Fig. 8D), suggesting that autophagy inhibited cell invasion. To confirm whether autophagy regulated trophoblast cell growth and invasion, autophagy was inhibited by treating cells with the selective inhibitor of autophagy, 3-methyladenine (3-MA). Treatment with 3-MA induced HTR8 cell invasion (Fig. 8G). These data suggested that autophagy is required for trophoblast cell growth and invasion. Taken together, miR-155-5p induced autophagy, reducing cell growth and invasion.