Activating transcription factor 3 promotes embryo attachment via up-regulation of leukemia inhibitory factor in vitro

The patients enrolled in this study were recruited from the in vitro fertilization (IVF) unit of the Reproductive Center of the Affiliated Drum Tower Hospital of Nanjing University Medical School. All of the participants were 22 - 38 years old, with a body mass index (BMI) of 18 - 26 kg/m² and regular menses. Among the participants from which secretory phase endometria were obtained, the endometrium thickness was 8 - 14 mm. A total of 15 recurrent implantation failure (RIF) patients, who had failed to achieve pregnancy after at least four good-quality embryo transfers over a minimum of three fresh or frozen cycles, were enrolled. Thirteen age-matched fertile controls with once embryo transplantation were also enrolled (Table 1). Secretory phase endometria were obtained from the RIF patients (n = 15) and fertile controls (n = 13) within days 19 to 23 of menstruation by diagnostic uterine curettage (Additional file 1: Table S1). A total of 10 proliferative phase endometria were obtained from patients undergoing endometrial biopsy within day 5 to day 12 of menstruation with the single diagnosis of fallopian tube jam, excluding endometrial disease such as endometriosis, uterine fibroid, intrauterine adhesion, and endometrial hyperplasia. Ten secretory phase endometrial tissues were obtained from age-matched fertile controls within days 19 to 23 of menstruation via diagnostic uterine curettage (Table 2).

This study was approved by the institutional review board of the Affiliated Drum Tower Hospital of Nanjing University Medical School on December 5, 2013 (2013–081–01), and signed informed consent was obtained from all patients.

The Ishikawa cells and BeWo cells used in this study were donated by Dr. Yali Hu from Department of Gynecology and Obstetrics in the Affiliated Drum Tower Hospital of Nanjing University Medical School. The cells were cultured in Dulbecco’s modified minimum essential medium (HyClone, Thermo Scientific, South Logan, UT, USA) containing 10% fetal bovine serum (FBS, Gibco BRL/Invitrogen, Carlsbad, CA, USA) and 1% penicillin/streptomycin (HyClone). To simulate physiological conditions, the Ishikawa cells were cultured at the concentration of 1 × 10⁴ cells/cm², and treated or not treated with 10 nM E₂ and 1 μM MPA (Sigma, St. Louis, MO, USA) for 0.5, 1, 2, 3, 4, 6 and 9 h with 2% charcoal/dextran-treated FBS and 1% penicillin/streptomycin in DMEM/F12 at 37 °C. ICI182780 (V900926, Sigma) and mifepristone (M8046, Sigma) were used at a concentration of 1 μM for 24 h to block estrogen receptors (ESR) and progesterone receptors (PR) in Ishikawa cells.

Human endometrial stromal cells (hESCs) were isolated using standard procedures [18]. Secretory endometrial tissues were minced and enzymatically digested with 0.1% collagenase (Worthington, Freehold, NJ, USA) for 30 min at 37 °C. Stromal cells were separated from intact glands by filtration of the digested tissue through 40 μm gauze. After centrifugation at 800 g for 5 min, the pellet was resuspended in DMEM/F12 (Gibco BRL/Invitrogen, Carlsbad, CA, USA) containing 10% charcoal/dextran-treated FBS (HyClone; Thermo Scientific, South Logan, UT, USA). The endometrial stromal cells were then maintained in DMEM/F12 supplemented with 10% charcoal/dextran-treated FBS and 1% penicillin/streptomycin at 37 °C. The cultured stromal cells were 95% pure, as judged by vimentin staining [19].

Primary cultures of human endometrial epithelial cells (hEECs) were prepared as described previously [20]. Secretory endometrial tissues were minced and enzymatically digested with 0.1% collagenase (Worthington, Freehold, NJ, USA) for 30 min at 37 °C. After incubation at above condition, the cell suspension was passed through 40 μm gauze to separate the single cells from undigested tissue. The unstrained tissues were resuspended in DMEM/F12 containing 10% charcoal/dextran-treated FBS and 1% penicillin/streptomycin and were transferred to dishes for 30 min to allow the adherence of contaminating stromal cells. Non-adherent cells/glands were filtered through 40 μm gauze 3 times to wash off the single cells, and the non-filtered pellets were resuspended and cultured. The endometrial epithelial cells were then maintained in DMEM/F12 supplemented with 10% charcoal/dextran-treated FBS and 1% penicillin/streptomycin at 37 °C. The cultured stromal cells were 90% pure, as judged by cytokeratin staining [19].

Total RNA was extracted from Ishikawa cells using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. A 1 μg aliquot of purified total RNA was reverse transcribed into cDNA using a Prime Script RT reagent kit (Takara-Bio, Otsu, Japan). qRT-PCR was performed using a SYBR green kit (Bio-Rad Laboratories, Hercules, CA, USA). The specific primers were as follows: human ATF3, 5′- TCGGAGAAGCTGGAAAGTGT-3′ and 5′-TCTGGAGTCCTCCCATTCTG-3′; human LIF, 5′-AGTCGTGACCTTGGCACCTC-3′ and 5′-GTTGACAGCCCAGCTTCTTC-3′; and human 18S rRNA, 5′-CGGCTACCACATCCAAGGAA-3′ and 5′-CTGGAATTACCGCGGCT-3′. The reactions were performed using a MyiQ Single-Color Real-time PCR Detection System (Bio-Rad). The PCR conditions were as follows: 95 °C for 15 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s. The fold change in the expression of each gene was normalized to the expression of the endogenous control (18S rRNA), and the data were analyzed using the 2^-ΔΔCT method.

Protein extracts were prepared from endometria or cells that were washed with ice-cold phosphate-buffered saline (PBS) and lysed in whole cell lysis buffer (50 mM Tris–HCl [pH 7.6], 150 mM NaCl and 1.0% NP-40) containing a protease inhibitor cocktail and phosphatase inhibitor cocktails 2 and 3 (Sigma). Equal amounts of total protein (30 μg) were separated on 12% (w/v) SDS-polyacrylamide gel and transferred to PVDF membranes (Millipore, Billerica, MA, USA). The membranes were blocked in Tris-buffered saline solution containing 5% nonfat milk for 1 h and then exposed to the following primary antibodies: ATF3 (1:1000; HPA001562, Sigma, 22 kD), LIF (1:500; L0669, Sigma; 38 kD), STAT3 (1:1000; #12640, Cell Signaling Technology, Danvers, MA, USA, 86 kD), p-STAT3 (1:1000; #9145, Cell Signaling Technology, 90 ~ kD), His-Tag antibody (1:2000; M30111, Abmart, Shanghai, China), and GAPDH (1:10000; AP0063, Bioworld, Nanjing, China; 42 kD). Immunodetection was accomplished using goat anti-rabbit or donkey anti-mouse secondary antibodies and an enhanced chemiluminescence detection kit (Millipore). The density of each band was scanned and measured by ImageJ software with target protein that had been normalized to GAPDH.

After the endometrial samples were dewaxed, endogenous peroxidase activity was blocked using freshly prepared PBS containing 0.3% hydrogen peroxide for 15 min. Antigen retrieval was conducted by autoclaving the samples at 121 °C for 15 min in the presence of EDTA (pH = 9.0). The sections were washed with PBS and then incubated with antibodies against ATF3 (1:1000; HPA001562, Sigma) overnight at 4 °C in a humidified chamber. The sections were subsequently rinsed with PBS and incubated with a horseradish peroxidase (HRP)-conjugated goat anti-rabbit secondary antibody at 37 °C for 30 min. HRP activity was detected using diaminobenzidine (Invitrogen), and the sections were counterstained with hematoxylin. Nonspecific rabbit serum (Boster, Wuhan, China) was used as a negative control.

Ishikawa cells plated on 18 mm micro-cover glasses were subjected to E₂ and MPA for 6 h. Treated or untreated Ishikawa cells were fixed with 4% paraformaldehyde in PBS for 30 min at room temperature. After washing with PBS, the cells were permeabilized with 0.2% Triton X-100 in PBS for 15 min at room temperature. After blocking with 1% bovine serum albumin (BSA) in PBS, the cells were probed for ATF3 and F-actin and then incubated at 4 °C overnight with anti-ATF3 polyclonal antibody (1:1000; HPA001562, Sigma) and F-actin (1:300, P5282, Sigma), with or without Alexa Fluor 594-conjugated goat anti-rabbit IgG (1:200; Invitrogen). Nuclei were stained with 4′,6-diamidine-2-phenylindole (DAPI), which was included in the Vectashield Mounting Medium for Fluorescence with DAPI kit. Images were visualized using a confocal microscope (Leica).

Ishikawa cells (70% confluence) were infected with Ad- ATF3-His (abm, Canada) or Ad-LacZ (Clontech, Japan) for 48 h. Cells were then washed with PBS and crosslinked with 1% formaldehyde for 15 min at room temperature. Crosslinking was stopped with the addition of glycine (0.125 M final concentration) for 10 min. Cells were washed twice with cold PBS, harvested in lysis buffer (20 mM Tris–HCl [pH 8.0], 85 mM KCl, 1 mM EDTA, 0.5 mM EGTA, 0.5% Nonidet P40, and protease inhibitor cocktail [Sigma]), and pelleted by centrifugation. Cell pellets were then lysed in nuclear lysis buffer (50 mM Tris–HCl [pH 8.0], 10 mM EDTA, 1% SDS, and protease inhibitor cocktail) and sonicated on ice to yield genomic DNA fragments with sizes of approximately 500–2000 base pairs. Next, precleared sonicates were immunoprecipitated using ATF3 antibody (HPA001562, Sigma) and nonspecific IgG as a technical control. Beads were collected and washed extensively. Immune complexes were eluted by incubation with fresh elution buffer (1% SDS and 0.1 M NaHCO3) at 65 °C for 30 min and then incubated at room temperature for 15 min. Crosslinks were reversed by incubation with NaCl at a final concentration of 0.3 M for 5 h at 65 °C. The eluates were incubated with proteinase K, and DNA was purified by phenol-chloroform extraction and ethanol precipitation. Finally, the purified DNA fragments were used as a template for PCR amplification. The specific primers that were used to amplify the LIF promoter DNA fragments containing an ATF3 binding sequence were 5′-GGCCTAGTAACCTCTGCTC-3′ and 5′-CGGCTCCCTCATGGAAG-3′ (spanning 232 bp;−2756 to−2497) as well as 5′-GCGAGGCTGCAAGAGCTC-3′ and 5′-CATGCCGTCCCTAAAGCTGC-3′ (spanning 150 bp; −1331 to−1181).

The wild-type human LIF promoter sequence (−2756 to−2497,−1331 to−1181,−1023 to 0) was amplified by PCR using Ishikawa cell genomic DNA with the primers 5′-GCATGGTACCGGCCTAGTAACCTCTGCTC-3′ and 5′-GCGCCTCGAGCGGCTCCCTCATGGAAG-3′, 5′-GCATGGTACCGCGAGGCTGCAAGAGCTC-3′ and 5′-GCGCCTCGAGCATGCCGTCCCTAAAGCTGC-3′ and 5′-GCATGGTACCTGGAGGTGTCCCTGTGCTC-3′ and 5′-GCGCCTCGAGGGATCCCCAGTCCAGGAAG-3′. PCR product was cloned into the pGL3-basic luciferase reporter plasmid (Promega). Preconfluent (60%) Ishikawa cells were transfected with the indicated plasmids using Lipofectamine 2000 (Invitrogen). The cells were incubated for an additional 48 h and then harvested for preparation of cell extracts. Luciferase activity was measured with a Luciferase Assay System (Promega, Madison, USA) in which Renilla luciferase plasmids were cotransfected as controls to standardize the transcription efficiency. These assays were performed using a Centro XS3 LB 960 luminometer (Berthold Technologies, BW, Germany) according to the manufacturer’s instructions.

According to our standard laboratory protocol [21], we used multicellular spheroids of human choriocarcinoma BeWo cells with endometrial Ishikawa cells as an in vitro model of attachment. BeWo cells were detached with 0.25% trypsin (Gibco) after reaching 80% confluence. The BeWo cell suspensions were then placed in 35 mm² dishes coated with an anti-adhesive polymer, poly-2-hydroxyethyl methacrylate (polyHEMA, Sigma), to induce the formation of BeWo spheroids that were 150–200 μm in diameter after 48 h of culture. Simultaneously, the confluent monolayer Ishikawa cells grown to 50% confluency in 24-well plates were infected with Ad-ATF3-His and Ad-LacZ (at a multiplicity of infection (MOI) of 0 or 40) for 48 h or were transfected with si-ATF3 and si-CTL (0, 50, 100 nM) for 48 h. A functional blocking antibody against LIF was also used in the attachment assay as the adenovirus-treated cells were cultured in the presence of a mouse monoclonal antibody against LIF (L0669, Sigma) or the mouse preimmune IgG antibody (Abcam, Cambridge, MA, USA) at a concentration of 0.5 μg/mL for 1 h (Additional file 2: Figure S1) before the transfer of the BeWo spheroids onto the surface of a confluent monolayer of Ishikawa cells. After incubation at 37 °C for 2 h, any unattached spheroids were removed by washing the cells with PBS containing Ca²⁺ (0.1 mg/L) and Mg²⁺ (0.1 mg/L). Then, attached spheroids were counted under a light microscope, and the attachment rate was expressed as the percentage of the total number of spheroids added to the Ishikawa monolayer. All of the attachment assays were performed in triplicate.

Unless stated otherwise, the numerical data are presented as the mean ± SD of at least three independent experiments. Student’s t-test was used to test for differences between two groups. ANOVA with Bonferroni correction was employed for multiple comparisons. Pearson correlation analysis was used to assess the relationship between ATF3 and LIF. Differences were considered significant at p < 0.05.

Secretory phase endometria from normal cycling women who had not been treated with hormones in the last three months were used to examine ATF3 expression in the human endometrium. As shown in Fig. 1A, we observed specific staining for the ATF3 protein in the human endometrium, both in the epithelial cells and stromal cells, compared with that of the negative control. Endogenous ATF3 was mainly localized in the nucleus in Ishikawa cells (Fig. 1B-a) as well as human endometrial stromal cells (Fig. 1B-b), as observed with immunofluorescence staining. ATF3 expression was greater in the secretive phase (n = 10) than in the proliferative phase of the endometria (n = 10) (*p < 0.05, Fig. 1C, Additional file 1: Table S2).

ATF3 mRNA expression was enhanced five-fold after 2 h by either E₂ or MPA in a time-dependent manner in Ishikawa cells (Fig. 2a). When stimulated with both E₂ and MPA, the ATF3 mRNA level in Ishikawa cells had increased 7-fold at 0.5 h and over 10-fold at 2 h (Fig. 2a). ATF3 protein expression began to increase from 1 h and reached a maximum (approximately 5-fold) increase at 6 h (*p < 0.05, Fig. 2b). ATF3 protein expression decreased following exposure to pretreatment with ICI182780 and mifepristone (Fig. 2c). We also determined that ATF3 mRNA level was declined with pre-treatment of ICI182780 and mifepristone were performed before sex hormones were perfomed (Additional file 2: Figure S2).  Immunofluorescence staining demonstrated that ATF3 accumulated predominantly in the nucleus upon E₂ and MPA treatment (Fig. 2d).

As shown in Fig. 3a, overexpression of ATF3 in Ishikawa cells increased the ratio of BeWo spheroid adhesion by 1.6-fold compared with that of the Ad-LacZ group (25.6 ± 4.19% vs. 42.3 ± 2.33%, **p < 0.01). In contrast, siRNA-mediated knockdown of endogenous ATF3 protein expression in Ishikawa cells reversed the facilitating effect of E₂ and MPA on BeWo spheroid attachment: 48.5 ± 15.78% (si-CTL) vs. 29.2 ± 4.31% (si-ATF3) (*p < 0.05, Fig. 3b). These data suggest that ATF3 contributes to embryo attachment in vitro.

Next, we discovered that ATF3 distinctly promoted the expression of LIF, consequently activating the phosphorylation of the STAT3 pathway (Fig. 4A-a, B) in Ishikawa cells. Up-regulation of LIF expression induced by ATF3 was also found in primary human epithelial cells (Fig. 4A-b) and stromal cells (Fig. 4A-c). In contrast, hormone-induced LIF expression was significantly reduced by knocking down ATF3 in Ishikawa cells (Fig. 4C).

ATF3 binds to a consensus DNA sequence (TGACGTCA) and forms selective heterodimers via the leucine zipper region [22]. We visually scanned the wild-type LIF (–3500 to +65) and found the specific ATF3 binding sites in the promoters of LIF (–2756 to–2497,–1331 to–1181). We generated 3 LIF-Luc reporter constructs, namely F1-Luc, F2-Luc, and F3-Luc (Fig. 4D). The reporter gene assay revealed a marked increase in luciferase activity of the fragment of−1331 to–1181 but no change in the fragments coding−2756 to–2497 and−1023 to 0. Conventional ChIP-PCR analysis was used to investigate whether the LIF promoter is a direct target for ATF3 in Ishikawa cells. As shown in Fig. 4E, the F2 promoter region (−1331 to−1181) was effectively recovered from immunoprecipitates of ATF3 proteins, but it was not recovered from those of the LacZ control and F1 promoter region (−2756 to−2497) of the LIF gene.

An antibody-blocking assay further demonstrated that the ATF3-mediated promotion of BeWo spheroid adhesion to Ishikawa cells [attachment rate: 45.3 ± 6.03% (lane b. ATF3) vs. 21.7 ± 2.08% (lane a. LacZ), **p < 0.01] was more significantly inhibited by pretreatment with an antibody specific to LIF than with control mouse IgG [attachment rate: 40.7 ± 12.01% (lane d. control IgG) vs. 26.0 ± 2.00% (lane c. LIF antibody), #p < 0.05; Fig. 4 F]. These results suggest that ATF3 regulates human endometrial receptivity and embryo attachment in vitro via up-regulation of leukemia inhibitory factor.

It was previously reported that compared with control patients, RIF patients exhibit deregulated LIF expression during the receptive phase [23]. Here, we discovered a trend for reduced ATF3 protein expression (Fig. 5a and b) as well as reduced LIF mRNA content (Fig. 5c) in the endometria of RIF patients (n = 15) compared with that of fertile females who underwent embryo transplantation once (FER) (n = 13) (Table 1), and a positive correlation was established between ATF3 and LIF expression (r = 0.461, p = 0.013; Fig. 4E). The IHC results also demonstrated that ATF3 was aberrantly expressed at low levels in both the epithelial cells and the stromal cells of the RIF group (Fig. 4D), indicating that decreased ATF3 expression in the endometrium was associated with impaired endometrial receptivity.