The operations of the animal protocol were ratified by the Institutional Animal Care and Use Committee of The International Peace Maternity and Child Health Hospital, School of Medicine, Shanghai Jiao Tong University (approval number: 20200815). Animal experiments were in full compliance with local, national, ethical, and regulatory principles and local licensing arrangements.

Human endometrial endothelial cells (hEEC) and EC cell lines (HEC-1-B, AN3-CA, KLE, HEC1-A, and Ishikawa) were provided by American Type Culture Collection (Manassas, USA). DMEM (Sigma-Aldrich, St Louis, MO, USA) added with 10% fetal bovine serum (FBS, Sigma-Aldrich, St Louis, MO, USA) was adopted for cell culture (Wang et al., 2020).

In Ishikawa cells, short hairpin RNA (sh-DLX6-AS1) and sh-NC were offered by Shanghai Integrated Biotechnology Solutions, Ltd. (Shanghai, China). MiR-374a-3p mimic, miR-374a-3p inhibitor, and their corresponding NCs were offered by RiboBio (Guangzhou, China); (Dai et al., 2020). The ZFX overexpression plasmid pcDNA3.1-ZFX and its corresponding negative control (NC) were procured from GenePharma (Shanghai, China) (Song et al., 2020). Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) was employed for construct transfection. The transfection efficiency was assessed by reverse transcription-quantitative polymerase chain reaction (RT-qPCR).

The groups were: sh-NC group, sh-DLX6-AS1 group, sh-DLX6-AS1 + inhibitor NC group, sh-DLX6-AS1 + miR-374a-3p inhibitor, mimic NC group, miR-374a-3p mimic group, miR-374a-3p mimic + pcDNA3.1 group, and miR-374a-3p mimic + pcDNA3.1-ZFX group.

Ishikawa cells (3 × 10³ cells/well) were incubated in 96-well culture plates for 0, 24, 48, and 72 h and incubated with MTT solution for 4 h. Then, dimethyl sulfoxide (DMSO) was appended, and cells were resuspended until the crystals were completely dissolved. Lastly, the optical density value was measured at 490 nm.

Cell migration and invasion capabilities were tested through the transwell assay in the presence or absence of Matrigel (Corning Incorporated, Corning, NY, USA). Cells were fixed by serum-free medium and incubation in the upward chamber, and DMEM with 10% FBS was filled into the bottom chamber. After hot water immersion for 24-h, cells were fixed to the bottom of the upward chamber, followed by ethanol treatment and 0.1% crystalline violet staining. A microscope was adopted for counting (Dai et al., 2020).

Cells were lysed using RIP buffer supplemented with RNAase inhibitors. The cell lysates were subjected to incubation with Protein A/G magnetic agarose beads (Thermo Fisher Scientific, Waltham, MA, USA) pre-coated with anti-Ago2 (Bio-Rad, Hercules, CA, USA) and anti-IgG (Abcam, Cambridge, UK) as the NC. Beads were harvested, and bound RNA was extracted to measure the levels of miR-374a-3p and ZFX mRNA (Li et al., 2020).

The DLX6-AS1 binding site to miR-374a-3p was queried by the RNA22 website. The TargetScan was implemented to detect the binding site of ZFX to miR-374a-3p. The miR-374a-3p binding sequence of DLX6-AS1 or ZFX was amplified and inserted into the pGL3 luciferase reporter vector. The recombinant vectors were called DLX6-AS1 wild type (DLX6-AS1-WT) or ZFX-WT. After that, the mutant binding site matching miR-374a-3p of DLX6-AS1 or ZFX was cloned into the pGL3 luciferase reporter vector to generate DLX6-AS1-mutant (MUT) or ZFX-MUT. Ishikawa cells were transfected with the above plasmids (mimic NC and miR-374a-3p mimic), and after 48-h, luciferase activity was determined in different groups by a dual luciferase reporter gene assay system (Li et al., 2019b).

RNA was isolated with Trizol reagent (Life Technologies, Carlsbad, CA, USA). Complementary DNA (cDNA) was synthesized with PrimeScript RT master mix kit (Takara, Dalian, China) together with All-in-OneTM miRNA First strand cDNA Synthesis Kit (GeneCopoeia, Rockville, MD, USA). RT-qPCR reactions were conducted using a SYBR Premix Ex Taq II Kit (Takara) along with a CFX96 real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA, USA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and U6 served as internal references in this study. All the used primers are listed in Suppl. Table 1. The expression assessment was determined by the 2^−ΔΔCt method (He et al., 2020).

Proteins were subjected to denaturation in sample buffer spiked with protease inhibitor mixture, SDS-polyacrylamide gel electrophoresis, followed by transfer to nitrocellulose membranes. The membranes were treated with 1-h blocking with 5% milk and overnight immunoblotting with the primary antibodies (ZFX antibody: 1:1000; GAPDH antibody: 1:2000 (both Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) at 4°C. The membranes were subjected to 1-h of incubation with a corresponding secondary antibody (1:3000; Solarbio, Beijing, China). Western blots were visualized and quantified (Guo et al., 2015; Bai et al., 2019).

Female BALB/c nude mice were housed at 24 ± 1°C, 50% humidity, 12-h light-dark cycle, with water and food supply. Ishikawa cells infected with lentivirus containing DLX6-AS1 shRNA, ZGX shRNA, or corresponding negative control (NC) lentivirus. After one-week of acclimation, mice were randomly classified into four groups (n = 6/group): sh-DLX6-AS1 group, sh-NC group, sh-ZFX group, and sh-CTR group. Then, 1 × 107 cells stably expressing sh-DLX6-AS1, sh-NC, sh-ZFX, and sh-CTR were injected subcutaneously into the dorsal belly. Tumor volume was calculated every 7 d for 35 d using the formula [volume = (π × length × width2)/6]. All animals were euthanized, and tumor xenografts were obtained and weighed (Wang et al., 2020). Tumor tissues were fixed with 4% paraformaldehyde for subsequent immunohistochemical staining.

All groups of tumor tissues were dehydrated, paraffin-embedded, and sectioned (4 μm). The tissue sections were pre-treated in a microwave oven for 20 min and then incubated with 3% H₂O₂ for 30 min to block the endogenous peroxidase activity in the sections. Next, the sections were incubated at room temperature with primary antibody against Ki-67 (1 μg/mL, ab15580, Abcam, Cambridge, UK) and Cleaved-caspase-3 (1:200, Cell Signaling Technology, Danvers, MA, USA) for 15 min and subsequently treated with the corresponding secondary antibody (1:2000, ab205718, Abcam, Cambridge, UK) for 30 min at room temperature. After that, the sections were developed with DAB (3,3’-diaminoben-zidine) solution (P0203, Beyotime, Shanghai, China) for 10 min, then counterstained with hematoxylin staining solution (C0107, Beyotime), followed by DPX blocking, and photographing under a microscope. The positive expression rates of Ki-67 and cleaved-caspase-3 in various tumor tissues were also calculated (Luo et al., 2021).

The SPSS 21.0 (SPSS, Inc., NY, USA) was implemented for data analysis. The t-test or one-way ANOVA was used for comparison between groups, as well as Tukey’s test. p < 0.05 was considered statistically significant.

The determination of DLX6-AS1 and miR-374a-3p levels in EC cell lines disclosed that DLX6-AS1 levels were high in EC cells (Fig. 1A), and miR-374a-3p levels were reduced in EC cell lines (Fig. 1B) compared to those in hEEC cells. These data evidenced the oncogenic character of DLX6-AS1, whereas the tumor suppressive role of miR-374a-3p in EC. Then, Ishikawa cells were screened for follow-up experiments.

The starBase revealed the binding sites of DLX6-AS1 and miR-374a-3p (Fig. 1C), whose relation was confirmed by the luciferase activity assay. MiR-374a-3p-mimic co-transfected with DLX6-AS1-WT showed reduced luciferase activity in Ishikawa cells (Fig. 1D). To unravel the correlation between DLX6-AS1 and miR-374a-3p, we measured DLX6-AS1 and miR-374a-3p levels in cells after sh-DLX6-AS1 transfection. It turned out that miR-374a-3p levels were enhanced after transfection of cells with sh-DLX6-AS1 (Fig. 1E). These data indicated the interaction between DLX6-AS1 and miR-374a-3p.

To determine any correlation between DLX6-AS1 and miR-374a-3p levels with EC cell progression, Ishikawa cells were further treated, and the DLX6-AS1 and miR-374a-3p levels reflected reduced DLX6-AS1 expression after transfection with sh-DLX6-AS1 and sh-DLX6-AS1 + anti-miR-374a-3p (Fig. 2A), indicating the successful transfection. MTT assay revealed that the decrease in DLX6-AS1 after DLX6-AS1 silencing prevented cell proliferation (Fig. 2B), and Transwell assay revealed restricted cell migratory and invasive capacities (Figs. 2C, 2D); However, miR-374a-3p depletion reversed the suppressive impacts of DLX6-AS1 silencing in EC cells growth. Collectively, these findings suggest that DLX6-AS1 depletion inhibited EC cell development by altering miR-374a-3p levels.

ZFX level is elevated in EC (Song et al., 2020). Hence, we examined the ZFX levels in EC cell lines, which suggested augmented ZFX expression in EC cells (Fig. 3A). TargetScan forecasted binding sites of miR-374a-3p and ZFX 3'UTR (Fig. 3B). Luciferase activity assay disclosed that miR-374a-3p could restrict the luciferase activity of ZFX-WT (Fig. 3C). RIP experiments indicated that ZFX and miR-374a-3p levels were high after treatment with Ago2, implying that ZFX targeted the RNA-caused depleted complex (Fig. 3D). The above discoveries demonstrated that miR-374a-3p negatively modulated ZFX in EC.

To probe whether miR-374a-3p regulated EC cell progression by targeting ZFX, we first examined miR-374a-3p and ZFX levels in EC cells upon further treatment with miR-374a-3p mimic and ZFX augmentation. The results revealed that miR-374a-3p levels were enhanced, while ZFX expression was reduced after miR-374a-3p mimic treatment (Fig. 4A). After miR-374a-3p augmentation, the growth of EC cells was repressed according to the findings from MTT and Transwell analyses, while ZFX augmentation offset the impacts of elevated miR-374a-3p on decelerating EC cell proliferative, migratory, and invasive capabilities (Figs. 4B–4D). The outcomes implied that miR-374a-3p controlled EC cell growth by targeting ZFX.

We then examined the ZFX levels by western blotting after further treatment. Inhibition of DLX6-AS1 decreased ZFX levels, while miR-374a-3p silencing reversed the impacts of silencing DLX6-AS1 on reducing ZFX levels; miR-374a-3p elevation reduced ZFX levels, and ZFX overexpression could reverse the impacts of miR-374a-3p elevation on reducing ZFX levels (Figs. 5A, 5B). These results thus evidenced that the DLX6-AS1/miR-374a-3p axis in EC cells regulated the abundance of ZFX.

A xenograft tumor model was constructed to address the impacts of DLX6-AS1 and ZFX in vivo. Sh-DLX6-AS1-or sh-ZFX-injected mice had diminished tumor volume and weight (Figs. 6A, 6B). In the tumor tissues of sh-DLX6-AS1-or sh-ZFX-injected mice positive expression rate of Ki-67 decreased, but the positive expression rate of cleaved caspase3 increased (Fig. 6C). In summary, the low expression of DLX6-AS1 and silencing of ZFX suppressed tumor growth and tumor tissue cell proliferation, and promoted tumor cell apoptosis in vivo.