Individuals with ovarian endometriosis were chosen from the Obstetrics and Gynecology Department at Liuzhou Maternal and Child Health Hospital. Post-surgery, the endometrial tissues were categorized into eutopic endometrial (EU) and ectopic endometrial (EC). EU samples were sourced from the endometrial lining within the uterus, while EC samples were taken from the ovarian tissue. We included women aged 18–45 with confirmed ovarian endometriosis and excluded those with malignancies, autoimmune diseases, other gynecological conditions, recent hormone therapy (within six months), or systemic diseases. To control for age, we focused on women aged 18–45 and ensured all samples were collected during the proliferative phase of the menstrual cycle without recent hormone treatment to minimize hormonal and systemic effects. Normal endometrium (normal control; n = 8) was obtained during the surgical operation. Human tissue samples were obtained with patients’ or their families/legal guardians’ written informed consent, in accordance with the Declaration of Helsinki guidelines. The research received approval from the Ethics Committee of Liuzhou Maternal and Child Health Hospital (KS-KY-2020-073).

Human endometrial stromal cells (HESCs) were sourced from Otwo Biotech (HTX2487-2, Shenzhen, China) and cultured with DMEM as previously [29]. The culture medium was enriched with 1.5 g/L sodium bicarbonate (144-55-8, Sigma-Aldrich, St. Louis, MO, USA), 1% recombinant Human Insulin (ITS)+ Premix (354352, Corning, Corning, NY, USA), 500 ng/mL puromycin, and 10% charcoal-stripped fetal bovine serum (F6765, Sigma-Aldrich), and kept in a humidified incubator at 37 °C in a 5% CO₂ environment. All cell lines were validated by short tandem repeat (STR) profiling and tested negative for mycoplasma. 5-Azacytidine (5-aza-dC) (HY-10586) and Suberoylanilide Hydroxamic Acid (SAHA) (HY-10221) were obtained from MCE (Middlesex County, NJ, USA), while sodium butyrate (NaB) (S1999) was acquired from Selleck. The expression of TERT in HESC cells was assessed by real-time quantitative PCR (qPCR) after treatment for 12 hours with SAHA (5 µM) or NaB (5 mM), or with 5 µM 5-Aza-dC for 48 hours.

To overexpress TERT and METTL3 in HESCs, human TERT and METTL3 were synthesized and cloned into the pLVX plasmid vector (General Biosystems, Anhui, China). For the knockdown of TERT and METTL3, lentiviruses containing small interfering RNAs (siRNAs) targeting TERT (5^′-AAAAATGTGGGGTTCTTCCAA-3^′), METTL3 (5^′-TCTAACTCAGGATCTGTAGCT-3^′), or a nonsense siRNA (5^′-CAGCCATCAACTCAGATTGTT-3^′) were constructed by General Biosystems (Anhui, China). Recombinant lentiviral vectors or control vectors, along with two auxiliary packaging plasmids (pSPAX2 and pMD2G), were co-transfected into HEK293T cells to produce the lentiviruses, as provided by General Biosystems (Anhui, China). After 48 hours, the supernatant was collected and used to infect HESCs with an initial multiplicity of infection (MOI) of 50. Transduction efficiency was assessed 48 hours post-infection using Western blot analysis (Supplementary Fig. 1).

The retrieved tissue specimens were promptly preserved in 4% buffered formalin and subsequently prepared for immunohistochemical (IHC) analysis by Servicebio Biotechnology Co., Ltd. (Wuhan, China). Paraffin-embedded slices were subjected to staining with a primary antibody targeting METTL3 (1:200, Proteintech, 15073-1-AP, San Diego, CA, USA) according to the manufacturer’s instructions.

To assess m6A levels in total RNA samples, an EpiQuik m6A RNA Methylation Quantification Kit (Epigentek, P-9005-48, Farmingdale, NY, USA) was utilized. Initially, the RNA was applied to wells pre-coated with RNA high binding solution, then capture and detection antibody solutions were introduced. The m6A levels were determined by measuring absorbance at 450 nm and calculating relative quantification.

Cell proliferation was evaluated through the Cell Counting Kit-8 (CCK-8) assay (MCE, HY-K0301). Cells were seeded in a 96-well plate and treated as needed. After treatment, 10 µL of CCK-8 solution was added to each well and incubated at 37 ℃ for 1–4 hours. Absorbance at 450 nm was then measured using a microplate reader to assess cell proliferation and viability. For 5-ethynyl-2’ -deoxyuridine (EdU) staining, the culture medium was supplemented with 10 µM EdU reagent (Beyotime Biotechnology, C0078S, Shanghai, China) and incubated for 3 hours at 37 °C. Cells were then stained with Hoechst 33342 as per the manufacturer’s instructions and examined under a fluorescence microscope.

Cell invasion and migration were assessed using Transwell chambers (Corning, Corning, NY, USA). For invasion assays, Matrigel (Corning, 354248) was mixed with fetal bovine serum (FBS)-free DMEM/F12 medium at a 1:3 ratio and applied to the upper chamber. For migration assays, Matrigel was not used. Cells (8 $\times$ 10⁴) were resuspended in 200 µL of serum-free DMEM/F12 and placed in the upper chamber, while the lower chamber contained 600 µL of medium with 20% FBS. After 24 hours, cells that adhered to the membrane’s underside were fixed, stained, and then quantified using a microscope.

Once cells reached 90–100% confluence, a sterilized pipette tip (200 µL) was used to create scratches. Following PBS washes to remove detached cells and debris, wound healing was monitored and photographed at 0 and 24 hours post-scratching. The wound area at each time point was compared to the initial size at 0 hours, and the percentage of area closure was calculated.

The RIP assay was conducted with the Magna™ RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore, 17-700, Burlington, MA, USA). Cells were lysed using a RIP lysis buffer, and magnetic beads were conjugated with human antibodies against YTH N6-methyladenosine RNA binding protein 2 (YTHDF2) (Proteintech, 24744-1-AP) or m6A (Synaptic Systems, 202-003, Gottingen, Germany). Normal mouse Immunoglobulin G (IgG) antibody (Millipore, 12-370) was used as a negative control. RNA obtained with the RIP assay was assessed by RT-qPCR and normalized to the input.

Total protein was extracted with RIPA buffer (Beyotime, P0013B). Following protein quantification, SDS-PAGE separation, and transfer onto polyvinylidene difluoride (PVDF) membranes were performed. The membranes were treated with 5% skim milk to prevent non-specific binding, and then incubated with the specified antibodies: TERT (1:1000, Thermo, MA5-16034, Waltham, MA, USA), METTL3 (1:2000, Cell Signaling Technology, #86132, Danvers, MA, USA.), YTHDF2 (1:2000, Abcam, ab246514, Cambridge, UK), YTHDF3 (1:2000, Abcam, ab220161), IGF2BP1 (1:2000, Abcam, ab290736), IGF2BP2 (1:2000, Abcam, ab124930), IGF2BP3 (1:2000, Abcam, ab177477), and GAPDH (1:6000, Proteintech, 60004-1-Ig) overnight at 4 °C. Subsequent incubation with secondary antibodies was followed by detection of protein bands using ECL chemiluminescent reagent (Beyotime, P0018S) and densitometry analysis using ImageJ (version 1.41, LOCI, University of Wisconsin, Madison, WI, USA).

The luciferase reporter assay was carried out utilizing the Dual-Luciferase Reporter Assay System. Cells were transfected with a pmiGLO-based luciferase vector carrying either wild-type or mutated TERT, or a control vector, employing Lipofectamine 3000 reagent (Invitrogen, Waltham, MA, USA). After 24 hours, the luciferase and Renilla activities were measured according to the manufacturer’s guidelines. For further analysis, pmirGLO-TERT-3^′UTR-WT and pmirGLO-TERT-3^′UTR-Mut constructs were introduced into control or shMETTL3 cells for 24 hours, and F-luc and R-luc activities were evaluated.

To measure RNA stability, cells were treated with 5 µg/mL actinomycin D (Act-D, Sigma-Aldric, #A9415). RNA samples were collected at various time intervals post-treatment, and real-time PCR was performed to analyze RNA levels. The TERT mRNA half-life (t_1/2) was calculated with the formula ln2/slope, using GAPDH as the reference control.

RNA was extracted with Trizol reagent (Invitrogen, A33250). For cDNA synthesis, 500 ng of mRNA was reverse transcribed with the PrimeScript RT Master Kit (Takara, Dalian, China). qPCR was performed using TB Green Premix Ex Taq II (TaKaRa, RR820Q), and transcript levels were determined based on the threshold cycle number. GAPDH served as the normalization control, and relative expression was calculated using the 2^{-Δ⁢Δ⁢Ct} method. The primer sequences are listed in Supplementary Table 1.

Results are expressed as mean $\pm$ SEM. The Shapiro-Wilk test was employed to evaluate data normality. For comparing two groups, an unpaired Student’s t-test was used for normally distributed data, while the Mann-Whitney test was applied for data that did not follow a normal distribution. For multiple group comparisons, ANOVA was performed with subsequent Bonferroni correction to adjust for multiple tests. A significance level of p 0.05 was adopted, and all experiments were performed independently. Statistical analyses were performed using GraphPad Prism software (version 9.1.4, Dotmatics, Boston, MA, USA).

To explore the involvement of TERT in endometriosis, we first evaluated its expression levels using Western blotting and qRT-PCR across normal endometrial (NC), EU, and EC. Our analysis revealed a notable elevation in TERT expression in EC relative to both EU and NC (Fig. 1A,B). No significant variation in expression levels was detected between EU and NC. To confirm these findings, we conducted immunohistochemistry (IHC) to assess TERT expression across NC, EU, and EC samples. The IHC results verified higher TERT expression in EC relative to NC and EU (Fig. 1C), supporting the results obtained from qRT-PCR and Western blot analyses. This indicates that increased TERT levels may play a role in the progression of endometriosis.

Fig. 1.

TERT overexpression in endometriosis. (A,B) Protein (A) and mRNA (B) expression levels of TERT in NC, EU, and EC tissue. (C) Immunohistochemistry of TERT expression in the endometriosis and control groups. Scale bar: 200 µm or 80 µm. Error bars represent the mean $\pm$ SD of results from triplicate biological experiments. **p 0.01. TERT, telomerase reverse transcriptase; NC, normal endometrial; EU, eutopic endometrial; EC, ectopic endometrial.

To explore the potential biological function of TERT in HESCs, lentiviruses were used to knockdown (shTERT) or overexpress TERT. EdU assay and CCK-8 assays were then performed to evaluate cell viability following TERT knockdown or overexpression. TERT knockdown significantly reduced the proliferation of HESCs, whereas TERT overexpression had the opposite effect (Fig. 2A,B).

Fig. 2.

The impact of TERT on HESC proliferation, migration, and invasion. (A,B) The influence of TERT silencing or overexpression on HESC proliferation was determined by CCK-8 assay (A) and 5-ethynyl-2’ -deoxyuridine (EdU) incorporation (B) Scale bar: 100 µm. (C) Transwell assays were conducted to evaluate migration (without matrigel) and invasion (with matrigel) 24 hours post-TERT manipulation. Representative images and quantification are displayed. Scale bar: 100 µm. (D) The wound healing assay was employed to measure cell motility at 0 and 24 hours, with representative images and quantitative data provided. Scale bar: 100 µm. Error bars denote the mean $\pm$ SD from 3 independent experiments. *p 0.05, **p 0.01. HESC, Human endometrial stromal cell; CCK-8, Cell Counting Kit-8; shNC, short hairpin negative control.

Next, the migratory and invasive abilities of HESCs with altered TERT expression were investigated using wound healing and transwell assays. TERT overexpression was found to promote cell migration and invasion, whereas TERT knockdown inhibited these processes (Fig. 2C,D). These findings indicate that TERT plays a crucial role in promoting cell proliferation and migration during endometriosis.

Recent studies demonstrated the involvement of epigenetic mechanisms in gene dysregulation. We therefore investigated whether epigenetic factors contribute to the upregulation of TERT in endometriosis. TERT expression remained unchanged following the treatment of HESCs with a DNA methyltransferase inhibitor (5-aza-dC), suggesting that DNA methylation does not play a role in regulating TERT. Furthermore, the application of broad-spectrum histone deacetylase (HDAC) inhibitors SAHA and NaB did not affect TERT expression, indicating that histone acetylation is not involved in TERT regulation in HESCs (Supplementary Fig. 2).

m6A modification in endometriosis was investigated by measuring m6A levels in NC, EU, and EC using a colorimetric method. The results showed that m6A levels were lower in EC compared to NC and EU. No significant difference in the m6A level was observed between EU and NC (Fig. 3A). The expression of m6A-associated genes in NC, EU, and EC was subsequently evaluated in order to identify the key molecules responsible for reduced m6A in endometriosis. Both the mRNA and protein levels of METTL3 were found to be downregulated in EC relative to NC and EU (Fig. 3B–D). These findings were consistent with the lower m6A levels, suggesting a potential role for METTL3 in m6A modification in endometriosis. Moreover, overexpression of METTL3 was found to significantly reduce TERT mRNA and protein expression, while reducing METTL3 levels led to an increase in TERT expression relative to normal controls (Fig. 3E,F).

Fig. 3.

METTL3 decreases m6A levels in endometriosis. (A) m6A content in NC, EU, and EC tissue. (B) mRNA expression of m6A-related genes measured by quantitative Reverse Transcription PCR (qRT-PCR). (C) METTL3 protein levels assessed through Western blotting. (D) METTL3 localization in tissues identified via Immunohistochemistry (IHC). Scale bar: 200 µm or 80 µm. (E,F) TERT mRNA levels evaluated by qRT-PCR (E) and Western blotting (F). Error bars indicate the mean $\pm$ SD from three independent experiments. *p 0.05, **p 0.01. METTL3, methyltransferase-like 3; m6A, N6-methyladenosine.

We investigated how TERT influences METTL3-driven cell proliferation, migration, and invasion in HESCs by introducing a METTL3-overexpressing lentivirus into cells with TERT overexpression. It was observed that elevated METTL3 levels markedly reduced cell proliferation, migration, and invasion. In contrast, TERT overexpression counteracted these suppressive effects (Fig. 4A–D). These findings indicate that TERT plays a role in modulating the METTL3-mediated inhibition of HESC proliferation, migration, and invasion.

Fig. 4.

METTL3 regulates HESC proliferation, migration, and invasion via TERT. (A,B) The EdU (A) and CCK-8 assay (B) were used to evaluate the effect of negative control and METTL3-overexpressing vector on HESC proliferation, both in the absence or presence of TERT overexpression. Scale bar: 100 µm. (C) Transwell assays were performed 24 h after transfection. Scale bar: 100 µm. (D) Changes in cell motility were assessed through wound healing assays at 0 and 24 h. Representative images and quantification are presented. Scale bar: 100 µm. Error bars represent the mean $\pm$ SD of triplicate biological experiments. **p 0.01.

Emerging evidence suggests that m6A influences various aspects of RNA fate [30, 31, 32]. We next investigated the role of m6A modification in TERT upregulation. Data from the online RNA modification resource (RMBase, http://rna.sysu.edu.cn/rmbase/) indicate that multiple m6A sites are present within the TERT transcript (Supplementary Fig. 3). In order to validate whether TERT is a target of METTL3-mediated m6A modification, we performed MeRIP-qPCR in HESCs. A remarkable decrease in m6A modification of TERT was observed when METTL3 was knocked down (Fig. 5A). Furthermore, the distribution of m6A methylation in TERT mRNA was investigated using fragmented RNA isolated from HESCs. The highest level of m6A methylation was observed in the 3^′ untranslated region (UTR) of TERT (Fig. 5B). Two m6A modification sites were identified within the 3^′UTR of TERT (Fig. 5C). To investigate the potential role of m6A in the 3^′UTR, luciferase reporters containing the wild-type or mutant TERT m6A methylation sites were generated (Fig. 5D). A firefly luciferase reporter assay demonstrated that, in HESCs with METTL3 knockdown, the luciferase activity of the reporter with the wild-type site was markedly elevated compared to that observed in control cells. However, mutations in either one or both of the TERT m6A motifs abolished the increased luciferase activity observed in cells with METTL3 knockdown (Fig. 5E). These findings suggest that m6A methylation in the TERT 3^′UTR is responsible for mRNA stability. Additionally, TERT mRNA expression was higher in cells co-transfected with METTL3 and mutant TERT-3^′UTR sites (TERT-3^′UTR-MUTs) compared to cells with the wild-type site (TERT-3^′UTR-WT) (Fig. 5F). Furthermore, the mRNA stability of TERT-3^′UTR-WT was greater in control cells than in cells with METTL3 overexpression, and the mutation in TERT-3^′UTR abolished this difference (Fig. 5G,H). These results indicate that m6A in the TERT 3^′UTR is crucial for mRNA stability, possibly due to its impact on the secondary structure of TERT mRNA.

Fig. 5.

METTL3-driven m6A modification regulates TERT expression. (A,B) Methylated RNA immunoprecipitation (MeRIP)-qPCR analysis of relative enrichment of m6A on TERT. (C) A diagram illustrates the locations and mutations of m6A sites within the TERT 3^′UTR, with “mut” denoting mutant and “chr” referring to chromosome. (D) A schematic shows the mutations introduced in the 3^′UTR to examine the impact of m6A on TERT expression. (E) The relative luciferase activity was measured for in control or shMETTL3 HESCs. (F) TERT-3^′UTR-WT or TERT-3^′UTR-DMut was transfected into control or shMETTL3 HESCs for 24 hours, followed by qPCR analysis to determine TERT mRNA expression levels. (G,H) Following 24-hour transfections of TERT-3^′UTR-WT (G) or TERT-3^′UTR-DMut (H) into control or shMETTL3 cells, the cells were treated with actinomycin D (5 µg/mL) for various durations. TERT mRNA levels were subsequently measured by qPCR. Error bars indicate the mean $\pm$ SD from triplicate experiments. **p 0.01; ns. indicates not significant. IgG, Immunoglobulin G; WT, wild type.

To regulate gene expression, m6A modification is performed by m6A writers and requires recognition by m6A readers [33, 34]. m6A readers, such as YTHDF2, YTHDF3, and IGF2BP1~3, can modulate mRNA stability [35, 36]. The present study found that YTHDF2, but not YTHDF1, YTHDF3, or IGF2BP1/2/3, showed significant binding to TERT mRNA in HESCs (Fig. 6A). Furthermore, the interaction between YTHDF2 and TERT was reduced in HESCs with METTL3 knockdown relative to control cells (Fig. 6B). Stability assays of mRNA indicated that TERT mRNA had an extended half-life in cells with YTHDF2 knockdown, while it was notably shorter in cells with YTHDF2 overexpression (Fig. 6C,D). Consistent with these findings, YTHDF2 negatively regulated TERT expression in HESCs (Fig. 6E,F). Additionally, the expression level of TERT mRNA in cells co-transfected with YTHDF2 and mutant TERT-3^′UTR sites (TERT-3^′UTR-MUTs) was higher than in cells co-transfected with YTHDF2 and the wild-type site (TERT-3^′UTR-WT) (Fig. 6G). These findings demonstrate that YTHDF2 recognizes METTL3-methylated TERT mRNA and facilitates its decay (Fig. 7).

Fig. 6.

YTHDF2 regulates the decay of TERT mRNA in an m6A-dependent fashion. (A) TERT mRNA was examined using RIP-qPCR with antibodies specific to YTHDF2, YTHDF3, and IGF2BP1~3. (B) The binding of YTHDF2 to TERT mRNA was assessed in control and shMETTL3 cells via RIP-qPCR. (C,D) The degradation rate of TERT mRNA was analyzed at various time points after treating with actinomycin D (5 µg/mL). (E,F) The levels of TERT protein (E) and TERT mRNA (F) were measured following the suppression or overexpression of YTHDF2 in HESCs. (G) TERT-3^′UTR-WT or TERT-3^′UTR-DMut was transfected into control or YTHDF2-overexpressing HESCs for 24 hours, after which TERT mRNA expression was quantified using qPCR. Error bars denote the mean $\pm$ SD from triplicate biological experiments. **p 0.01. YTHDF2, YTH N6-methyladenosine RNA binding protein 2.

Fig. 7.

Graphical representation of the roles of METTL3 and TERT in endometriosis. METTL3 modifies TERT mRNA through m6A, while YTHDF2 mediates the decay of TERT mRNA in an m6A-dependent manner.