Dorsomorphin dihydrochloride, an AMPK inhibitor and 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR), an AMPK activator, were obtained from Tocris Bioscience and puromycin dihydrochloride was purchased from InvivoGen, Inc.

Human endometrial tissues were obtained from fertile participants in the proliferative [9-11 menstrual cycle days (mcd); n=11; mean age, 36.7±2.8 years] and secretory (20–24 mcd; n=9; mean age, 37.8±2.6 years) phases between April and June 2019 at the Konyang University Hospital. Samples from patients with infertility were collected from the secretory phase (20–22 mcd; n=7; mean age, 38.9±3.5 years) between April 2018 and April 2020 at the MizMedi Hospital (Seoul, South Korea). Patients with infertility did not receive any medication, including hormone treatments, during the sampling period. Endometrial sampling was performed using a disposable uterine sampler (Rampipella; Ri.mos. (S.r.l.). The menstrual stages of the samples were determined using the Noyes criteria by an experienced gynecological pathologist (18). The present study was approved by the Bioethics Committee of Konyang University Hospital [Institutional Review Board (IRB) file no. KYUH 2018-11-007] and MizMedi Hospital (IRB file no. MMIRB 2018-3). Signed informed consent was obtained from each patient.

Human endometrial cell lines (RL95-2 and AN3CA) were acquired from the American Type Culture Collection. RL95-2 and AN3CA cells were cultured in DMEM/F-12 and MEM (HyClone; Cytiva) supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin (HyClone; Cytiva). Mycoplasma contamination tests were performed with MycoStrip (InvivoGen, Inc.).

Human choriocarcinoma JAR cells were obtained from the South Korean Cell Line Bank and cultured in RPMI-1640 medium (HyClone; Cytiva) supplemented with 10% heat-inactivated FBS and 1% penicillin/streptomycin. All cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂. RL95-2 cells were treated with dorsomorphin dihydrochloride for 24 h. AN3CA cells were treated with AICAR (AMPK activators; Tocris Bioscience) for 24 h.

For the generation of lentivirus particles, 293T cells were transfected with 400 ng of either a control (pLKO1; SHC016) or a short hairpin (sh)RNA targeting adiponectin receptor 1 (shADIPOR1; sense: 5′-CGTCTATTGTCATTCAGAGAA-3′, antisense: 5′-TTCTCTGAATGACAATAGACG-3′) expression vector (MilliporeSigma) alongside 400 ng pLP1, 100 ng pLP2 and 300 ng pLP/VSVG vectors (all from Thermo Fisher Scientific, Inc.) based on a 3rd generation system using Lipofectamine^® 3,000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's recommended protocol. Lentiviral particles targeting ADIPOR1 and control lentiviral particles were collected from media following transfection for 48 h. Subsequently, RL95-2 cells were transduced with 10 MOI control lentiviral particles or those carrying shRNA against the ADIPOR1 mRNA. RL95-2 cells with ADIPOR1 knockdown were isolated and selected using 0.5 µg/ml puromycin dihydrochloride.

Total RNA extraction from both 1×10⁶ AN3CA or 2×10⁶ RL95-2 cells and endometrial tissues was performed using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) following the manufacturer's protocol. Complementary DNA (cDNA) was synthesized from 2 µg RNA for cells and 5 µg for tissues using Moloney Murine Leukemia Virus reverse transcriptase (Promega Corporation) according to the manufacturer's recommended protocol.

RT-qPCR was performed to analyze mRNA expression using iQ SYBR Green Supermix (Bio-Rad Laboratories, Inc.) on a CFX Connect Real-time PCR Detection System (Bio-Rad Laboratories, Inc.) according to the manufacturer's recommended protocol. Specific gene primers and their annealing temperatures are listed in Table I. The following amplification conditions were used for select genes encoding ADIPOR1, ADIPOR2, ITGB5, L-Selectin, E-Selectin, E-cadherin, AMPK and β-actin: An initial denaturation step at 95°C for 3 min, followed by 49 cycles of denaturation at 95°C for 30 sec, annealing at 58°C for 15 sec and extension at 72°C for 15 sec. Gene expression levels were normalized to those of β-actin, an internal control. Relative expression was calculated using the 2^−ΔΔCq method (19) and fold change was evaluated compared with the control.

Cells were lysed using ice-cold radioimmunoprecipitation assay (RIPA) buffer on ice (JuBiotech), supplemented with protease and phosphatase inhibitors (Roche Diagnostics GmbH), and protein concentration was quantified using the bicinchoninic acid assay (Thermo Fisher Scientific, Inc.). Subsequently, 50 µg (for p-AMPK, AMPK, E-cadherin, integrin β5 and β-actin detection) or 70 µg (for ADIPOR1, ADIPOR2, L-selectin, E-selectin, E-cadherin and β-actin detection) proteins were separated by sodium dodecyl sulphate-polyacrylamide gel electrophoresis on 8 or 10% gels and transferred onto polyvinylidene difluoride membranes. The blots were blocked with 5% skimmed milk (Difco; BD Biosciences) for 2 h at room temperature, washed with 1X TBS containing Tween 20 buffer (Biosesang) and then incubated with primary antibodies against ADIPOR1 (1:500; cat. no. ab70362; Abcam), ADIPOR2 (1:500; cat. no. ab77612; Abcam), L-selectin (1:1,000; cat. no. PAA086Hu01; Cloud-Clone Corp.), E-selectin (1:1,000; cat. no. ab18981; Abcam), integrin β5 (1:1,000; cat. no. ab184312; Abcam), Phospho-AMPKα (Thr172) (1:1,000; cat. no. 2531; Cell Signaling Technology, Inc.), AMPKα (1:1,000; cat. no. 2532; Cell Signaling Technology, Inc.), E-cadherin (1:1,000; cat. no. 3195; Cell Signaling Technology, Inc.) and β-actin (1:3,000; cat. no. 4967; Cell Signaling Technology, Inc.) were incubated overnight at 4°C. The following day, blots were probed with horseradish peroxidase-conjugated secondary antibodies (1:3,000; cat. no. AP132P; MilliporeSigma) for 2 h at room temperature and bands were visualized using an Enhanced Chemiluminescence Kit (Thermo Fisher Scientific, Inc.). Quantification of western blot images was performed using the ImageJ software 1.50b (National Institutes of Health).

JAr cell spheroids were prepared by seeding onto a V-bottom microplate (Greiner Bio One Ltd.) and incubating in DMEM supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin (HyClone; Cytiva) for 24 h in a humidified atmosphere with 5% CO₂. Endometrial cells were cultured in 24-well plates until they reached 90% confluency. The harvested spheroids were co-cultured with either RL95-2 cell monolayers (for 1 h) or AN3CA cell monolayers (for 2 h), These cells were pre-treated with either dorsomorphin, AICAR, ADIPOR1 shRNA vector, or respective controls. The attached spheroids were quantified by inverting the microplate and centrifuging at 10 × g for 10 min. The attached spheroids were then counted under a microscope (Olympus Corporation). The percentage of spheroid attachment was calculated by dividing the number of attached spheroids after centrifugation by the total number of spheroids. The implantation assay was conducted at least three times for validation.

Data are presented as mean ± standard error of the mean. The results were analysed using unpaired Student's t-test or one-way ANOVA with Tukey's post hoc test with GraphPad Prism 5 (Dotmatics). P<0.05 was considered to indicate a statistically significant difference.

To investigate the role of ADIPOR1 in the endometrial cycle, the present study first quantified its expression in the endometria during the proliferative and secretory phases of fertile women as well as the secretory phase of infertile women. The mRNA levels of ADIPOR1 were markedly higher in the secretory phase compared with the proliferative phase in fertile women. ADIPOR1 expression was markedly reduced during the secretory phase in infertile women compared with that in fertile women (Fig. 1A). Based on these results, to explore the function of ADIPOR1 in the endometrial receptivity process, the receptive RL95-2 and non-receptive AN3CA cell lines were selected. Endometrial receptivity between RL95-2 and AN3CA cell lines were confirmed using an in vitro JAr spheroid attachment model. The frequency of spheroid attachment was higher in receptive RL95-2 cells compared with non-receptive AN3CA cells, consistent with previous data (20) (Fig. 1B). For further validation, the expression patterns of proteins that were biomarkers of endometrial receptivity were examined. The expression levels of E-selectin, L-selectin and ITGB5 were significantly higher in the receptive RL95-2 cells compared with the non-receptive AN3CA cells (data not shown). In particular, the expression of E-cadherin was significantly higher in receptive RL95-2 cells compared with non-receptive AN3CA cells (Fig. 1C). Based on these results, the expression pattern of ADIPOR1 in cells with different receptivity were investigated. Lower mRNA levels of ADIPOR1 were observed in non-receptive AN3CA cells compared with receptive RL95-2 cells (Fig. 1D). ADIPOR1 protein levels were reduced in AN3CA cells (Fig. 1E). By contrast, ADIPOR2 was more highly expressed in the endometria of infertile women compared with those of fertile women and in non-receptive epithelial cells compared with receptive epithelial cells in the present study (Fig. S1). These results suggested that differences in ADIPOR1 expression may be associated with endometrial receptivity.

Due to the differences in ADIPOR1 expression in cells with different receptivity, the changes in endometrial receptivity caused by ADIPOR1 downregulation were investigated in relatively upregulated recipient RL95-2 cells. Stable cells with downregulated ADIPOR1 (shADIPOR1) were established using shRNA-mediated gene silencing. The expression levels of ADIPOR1 were examined to verify the establishment of stable cell lines. The mRNA levels of ADIPOR1 and corresponding protein levels were significantly lower in shADIPOR1 cells compared with the pLKO1 control cells (Fig. 2A and B). Next, an in vitro implantation assay using the JAr spheroid attachment method was performed to investigate the relationship between the changes in ADIPOR1 expression and embryo implantation. As shown in Fig. 2C, ADIPOR1 downregulation attenuated spheroid attachment in JAr cells compared with that in the control cells. The E-cadherin protein is important for successful embryo implantation (21). Therefore, the effect of ADIPOR1 knockdown on E-cadherin expression was examined. E-cadherin mRNA expression was reduced in shADIPOR1 cells compared with that in the control cells (Fig. 2D). Consistent with the changes in mRNA expression, E-cadherin protein levels were reduced in shADIPOR1-treated cells (Fig. 2E). E-selectin, L-selectin and ITGB5 expression was reduced following ADIPOR1 knockdown (Fig. S2). These results suggest that ADIPOR1 positively regulates E-cadherin expression and may play an important role in the regulation of endometrial receptivity during embryo implantation.

The present study examined the AMPK phosphorylation state in receptive and non-receptive cells showing differential ADIPOR1 expression to investigate the role of the association between ADIPOR1 and AMPK activity in regulating endometrial receptivity. AMPK phosphorylation was lower in non-receptive AN3CA cells compared with receptive RL95-2 cells (Fig. 3A). Knockdown of ADIPOR1 induced decreased AMPK phosphorylation in RL95-2 cells (Fig. 3B). These results indicated that ADIPOR1-mediated AMPK regulation may be associated with the receptivity of endometrial epithelial cells.

The present study used dorsomorphin, which inhibits AMPK activity, to investigate whether the AMPK pathway regulates the receptivity of endometrial epithelial cells. Dorsomorphin did not affect cell viability in cytotoxicity tests. The inhibition of AMPK phosphorylation was also confirmed via western blotting in RL95-2 cells treated with various concentrations (Fig. S3A). AMPK phosphorylation was reduced upon dorsomorphin treatment in a dose-dependent manner (Fig. 4A). Next, JAr cell spheroid attachment-mediated endometrial receptivity in RL95-2 cells following dorsomorphin treatment was investigated. Dorsomorphin-induced reduction in spheroid attachment was proportional to the inhibition of AMPK phosphorylation (Fig. 4B). The expression patterns of biomarkers associated with endometrial receptivity was confirmed. Silencing AMPKa1 triggers the disruption of cell-cell adhesion via the Twist/E-cadherin pathway in breast cancer (22). Therefore, E-cadherin expression in receptive RL95-2 cells treated with dorsomorphin was examined. The mRNA expression of E-cadherin decreased in the cells treated with dorsomorphin (Fig. 4C). Dorsomorphin treatment also reduced E-cadherin protein expression in a dose-dependent manner (Fig. 4D). These results suggest that AMPK inhibition is closely associated with decreased endometrial receptivity.

The AMPK activator AICAR was used to further investigate the relationship between AMPK activity and endometrial receptivity. The toxicity following AICAR treatment was assessed using the MTT assay for cell viability and the phosphorylation state of AMPK was verified via immunoblot analysis at different concentrations in non-receptive AN3CA cells (Fig. S3B). The phosphorylation level of AMPK increased proportionally with AICAR (250 and 500 µM) (Fig. 5A). The rate of JAr cell-spheroid attachment was significantly increased following AICAR treatment, consistent with enhanced AMPK phosphorylation (Fig. 5B). Next, E-cadherin expression in non-receptive AN3CA cells following AN3CA treatment was examined. AICAR treatment significantly increased the mRNA expression of E-cadherin in a dose-dependent manner (Fig. 5C). E-cadherin protein levels increased in a dose-dependent manner in the AICAR-treated cells (Fig. 5D). These results strongly suggested that AMPK activity affects endometrial receptivity by regulating E-cadherin expression.