Characterization of a novel telomerase-immortalized human endometrial stromal cell line, St-T1b

Primary cultures of human ESC were prepared from anonymized uterine biopsy samples obtained from premenopausal women (age 35–50) at the time of hysterectomy for benign gynecological disorders. Purified ESC were prepared as previously described [15] and maintained in ESC medium: phenolred-free Dulbecco's modified Eagle medium (DMEM)/Ham's F12 with 10% steroid-depleted dialyzed fetal bovine serum (FCS) (PAA Laboratories, Cölbe, Germany), 100 U/ml penicillin, 100 μg/ml streptomycin, and 0.2% Primocin (Invivogen, San Diego, CA), and supplemented with insulin (1 μg/ml) and 17β-estradiol (E2; 1 nM) (Sigma-Aldrich, Deisenhofen, Germany). Decidualization was performed in minimal medium 1 (MM1; ESC medium without insulin or E2) in the presence of 0.5 mM 8-Br-cAMP (Biolog, Bremen, Germany). The local ethics committee approved this study and patient consent was obtained before tissue collection.

Immortalization of primary ESC was performed in collaboration with Gordon Peters (Cancer Research UK London Research Institute, London, UK). ESC were isolated from a biopsy taken during the proliferative phase from a 30 year old patient with tubal disease awaiting IVF treatment. Briefly, the receptor for ecotropic retroviruses was first stably introduced into the primary ESC using neomycin selection, then the cells were infected with a retrovirus encoding hTERT and selected for hygromycin resistance. The pBABE retrovirus system was used in which the cDNA of interest is driven by the Moloney murine leukemia virus promoter in the viral long terminal repeat and the marker is driven by the SV40 promoter [16]. St-T1b cells were maintained and decidualized in the same media as described for ESC, but in the absence of Primocin. Stimulations with P4, MPA, dexamethasone (all from Sigma-Aldrich) and R5020 (Promegestone; PerkinElmer, Rodgau, Germany) were performed in MM1. With the exception of the long-term observation, all experiments were performed repeatedly between passage numbers 30 and 53, and representative experiments are shown.

The trophoblastic hybridoma cell line AC-1M88, a fusion of primary extravillous trophoblast from term placenta with a mutant of the JEG-3 choriocarcinoma cell line, was kindly provided by H.G. Frank and P. Kaufmann (University Hospital Aachen, Germany) [17, 18], and was maintained in DMEM/Ham's F12 with 10% complete FCS, 100 U/ml penicillin, and 100 μg/ml streptomycin.

Cells grown on 8-well chamber slides (BD Biosciences, Heidelberg, Germany) were fixed with 4% paraformaldehyde, blocked for endogenous peroxidase activity (0.75% H₂O₂ in methanol for 10 min at room temperature), and then immunostained with mouse monoclonal antibodies to vimentin (Clone V9, 1:1,000) (Dako, Hamburg, Germany), CD90 (Clone AS02, 1:100) (Dianova, Hamburg, Germany), or cytokeratin-7 (CK7) (Clone OV-TL, 1:200) (Dako). Immunoreactivity was visualized with 3,3'-diaminobenzidine tetrahydrochloride (DAB; brown) (Vectastain ABC-Elite-Universal:HRP Kit; Linaris).

Cultured cells were fixed with 4% paraformaldehyde for 10 min at room temperature, washed with PBS for 10 min, and permeabilized with 0.1% Triton X-100 in PBS for 10 min. Following three washes in PBS for 10 min each, cells were incubated in PBS/1% BSA for 25 min. Fluorescein-phalloidin (Invitrogen, Karlsruhe, Germany) stock solution (200 U/ml in methanol) was diluted 1:40 in PBS/1% BSA and placed on the monolayer for 20 min. After three washes in PBS, cells were mounted in ProLong Gold antifade reagent (Invitrogen).

RNA was extracted from cultured cells with peqGold RNApure reagent (Peqlab, Erlangen, Germany) according to the manufacturer's protocol, but the aqueous phase obtained after chloroform extraction was subjected to an additional purification step by phenol/chloroform/isoamylalcohol extraction. One μg of RNA was used for oligo(dT)-primed cDNA synthesis with the ImProm-II Reverse Transcription System (Promega, Mannheim, Germany). Of the resulting 20 μl of cDNA, 0.5 μl were used per semi-quantitative PCR reaction. Unless indicated otherwise, 20 μl PCR reaction mix contained 1 × Taq buffer (5PRIME; VWR International, Darmstadt, Germany), 0.2 mM dNTPs, 2 pmol sense and antisense primers, 1 M betain, and 0.2 μl 5PRIME Taq DNA polymerase (VWR International); for amplification of steroid hormone receptor cDNAs, 0.1 μl BioTherm Taq DNA polymerase (Genecraft, Lüdinghausen, Germany) was used. All programs started with a denaturation step at 95°C for 4 min and terminated with an elongation step at 72°C for 10 min. For amplification of dPRL cDNA, Hot Start Taq DNA polymerase (0.1 μl) was used with the corresponding buffer (Qiagen, Hilden, Germany) and a denaturation step at 95°C for 15 min. Specific amplification conditions and primer sequences are given in Table 1. PCR products were resolved in 2% agarose gels, stained with SYBR Gold (Molecular Probes, Invitrogen) and visualized in a Typhoon 8600 Imager (Amersham Biosciences, Freiburg, Germany).

Whole cells extracts were prepared in RIPA buffer (10 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% sodium deoxycholate, 1% Triton X-100, 0.1% SDS) with Complete protease inhibitors (Roche Applied Science, Mannheim, Germany). Proteins were electrophoresed on 10% SDS-polyacrylamide gels (NuPage Bis-Tris; Invitrogen) and transferred by tank-blotting onto polyvinylidene difluoride Immobilon membranes (Millipore, Eschborn, Germany). For electrophoresis and transfer under non-reducing conditions, RIPA lysates were added to an equal volume of loading buffer without β-mercaptoethanol (10 mM Tris, 10% SDS, 25% glycerol, 0.01% bromophenol blue), and antioxidant was omitted from the NuPAGE transfer buffer (Invitrogen). For reducing conditions, 25% β-mercaptoethanol was included in the loading buffer, and electrotransfer was carried out in the presence of 0.25% antioxidant. Immunodetection was performed with the enhanced chemiluminescence system (SuperSignal; Pierce, Bonn, Germany). KAI1 was detected under non-reducing conditions (antibody TS82b, 1:500) (Diaclone, Besançon, France). GAPDH (6C5, 1:10,000) (HyTest, Turku, Finland) and p53 (DO-1, 1:1,000) (Calbiochem, Darmstadt, Germany) were detected under reducing conditions. For analysis of FOXO1, cells were directly extracted into hot loading buffer, and equal volumes loaded per lane for reducing gel electrophoresis. The rabbit monoclonal FOXO1 antibody C29H4 (Cell Signaling Technology, Danvers, MA) was used at 1:1,000.

St-T1b cells were plated at a density of 4 × 10⁵ cells per well in 6-well-plates. Two days later, media were changed, and cells received stimuli (8-Br-cAMP, P4 or MPA alone or in combination) or MM1 alone for 4 d. Media were changed again, all cells received stimuli, and 3 d later supernatants were harvested for measurement of IGFBP-1 secretion, and total RNA was extracted from the monolayers. Every treatment was done in triplicate wells. Secreted IGFBP-1 was measured by ELISA (Mediagnost, Reutlingen, Germany) (detection range: 0.1 – 8 ng/ml) and normalized to total RNA.

St-T1b cells were initially characterized approximately 30 passages after immortalization with hTERT. First, we examined cell type-specific antigen expression of these fibroblast-derived cells. Vimentin and CD90 were used as fibroblast markers and contrasted to cytokeratin-7 (CK7) as an epithelioid marker (Fig. 1). St-T1b cells were found to display the appropriate phenotype being vimentin⁺/CD90⁺/CK7^-. The trophoblast-derived cell line AC-1M88 served as a positive control for CK7 staining and lack of vimentin and CD90 immunoreactivity.

St-T1b cells were treated with 8-Br-cAMP, a widely established stimulus for decidualization [1]. Morphological decidualization was evident by phase contrast microscopy and upon staining of the actin cytoskeleton, showing the same reorganization from elongated to polygonal shape in St-T1b cells as seen in primary ESC (Fig. 2A). St-T1b cells were also compared to primary ESC for their ability to express typical marker genes of decidualization, namely those coding for decidual PRL (dPRL) and IGFBP-1. Maximum induction of dPRL and IGFBP-1 transcripts was seen after 3 d of cAMP treatment, and induction was maintained for at least 7 d in both cell systems (Fig. 2B). Furthermore, treatment with 8-Br-cAMP over a 7-day time-course induced a gradual increase in both p53 and KAI1 proteins but not mRNA levels (Fig. 3). In summary, the cAMP responses of St-T1b cells closely reflected those of primary ESC.

We wondered if the cells retained their ability to decidualize in extended culture. Cells obtained at passage number 30 were therefore maintained in continuous culture for more than 113 additional passages. At intervals, a 3-day decidualization experiment was performed, measuring the relative induction of dPRL, IGFBP-1 and FOXO1 transcripts as the endpoints (Fig. 4A). The induction of dPRL mRNA was clearly detectable up to passage 82 and then began to weaken. IGFBP-1 transcripts were less robustly induced and faded around passage number 72. St-T1b displayed a cAMP-dependent induction of FOXO1 message levels throughout the entire study period with a decline at later passages. This was reflected by a decrease in cAMP-induced FOXO1 protein levels when comparing passage numbers 44 and 114. The response to cAMP was not altered by the addition of P4 or P4 plus E2 over 7 d (Fig. 4B). Morphological features of decidualization, and vimentin expression, were still maintained at passage 114 (Fig. 4C).

The doubling time of St-T1b cells was determined to be approximately 24 h. No growth crisis was apparent throughout the entire period of observation.

We next assessed the expression of the nuclear receptors for ovarian steroid hormones, namely progesterone receptor (PR) and estrogen receptor (ER). The expression of PR was tested in the absence and presence of P4, as it has been reported that liganded PR activates its own promoter in human ESC [19]. Using primers to various regions of the PR, namely to the N-terminus specific to the full length PR-B, to a region common to PR-B and the N-terminally truncated PR-A isoform, and to the ligand-binding domain (LBD) of PR [20], expression of PR was readily detectable in both St-T1b cells and ESC. Likewise, ERα transcripts were amplified from both cell types. Transcript levels were not altered by P4 treatment (Fig. 5A). ERβ transcripts were below the limit of detection in both cell systems (data not shown).

To further explore the P4 responsiveness of St-T1b cells, we turned to a panel of established genes under the transcriptional control of the activated PR isoforms, PR-A, PR-B or both. These included tissue factor (TF; gene F3) and integrin α6 (ITGA6) as PR-B target genes [21], decidual protein induced by progesterone (DEPP; C10orf10) as a PR-B/PR-A target gene [22, 23], HEF-1 (NEDD9) as a PR-A target gene [21], and promyelocytic leukemia zinc finger protein (PLZF; ZBTB16) as a product rapidly induced by P4 in ESC [24]. RNA was harvested after 4 and 24 h of treatment with P4, the synthetic progestin R5020 (Promegestone), or the glucocorticoid receptor (GR) agonist dexamethasone, and subjected to semiquantitative RT-PCR for the above targets (Fig. 5B). None of the transcripts was induced in response to P4 or R5020. In contrast, dexamethasone weakly up-regulated HEF-1 mRNA after 4 h but had a strong and sustained effect on PLZF mRNA expression. This is consistent with our previous observation that PLZF expression is regulated not only by P4 but also by glucocorticoids [24]. The absence of a transcriptional response to P4 prompted us to examine the expression of PR isoforms in St-T1b cells at the protein level. Neither PR-B nor PR-A were detectable by Western blotting in this cell line (data not shown). Thus, undifferentiated St-T1b cells are responsive to glucocorticoid but not to P4 as they seem to lack PR protein. In agreement, P4 treatment, in combination with E2, failed to induce morphological changes associated with decidualization in St-T1b cells (data not shown).

At physiologic concentrations, P4 has a half-life of only 2–4 h in cells, and even at 1 μM, the steroid is entirely metabolized within 18 h, whereas synthetic progestins including MPA are not metabolized [25]. Owing to its higher stability as compared to P4, MPA is frequently used in the extended treatments of ESC to induce the decidual reaction. We therefore compared the two progestins, P4 and MPA, alone and in combination with cAMP analog. St-T1b cells were treated for 3 or 7 d before supernatants and RNA were harvested (Fig. 6A). RT-PCR analysis of dPRL, IGFBP-1, FOXO1 and DEPP transcript levels revealed induction of all targets by cAMP after 3 d but not in response to progestin treatment alone for 3 or 7 days (Fig. 6B). However, different patterns of responses were observed for combined treatments. FOXO1 and dPRL mRNA were induced by cAMP after 3 and 7 d to similar degrees; this was not altered by the presence of P4 but minutely enhanced by MPA. For DEPP mRNA, on the other hand, the cAMP response seen after 3 d was reduced after 7 d, but recovered both by addition of P4 and MPA. Induction of IGFBP-1 in response to cAMP stimulation was transient, characterized by lower expression levels after 7 compared to 3 days of stimulation (see also Fig. 2). The decline in the expression of this decidual marker gene after 7 days of cAMP stimulation was entirely reversed upon co-treatment with P4 and even more so with MPA. These findings indicate a synergy between cAMP and MPA at both time points, and most importantly, the acquisition of P4-responsiveness after 7 days of exposure to cAMP. The supernatants of the above cultures were also assayed for secreted IGFBP-1. Remarkably, the expression profile seen at the mRNA level was precisely reflected at the protein level and clearly underpinned the significant reduction in cAMP-stimulated IGFPB-1 expression after 7 d, and a reversal of this reduction by added P4 at the later time point. Again, the most effective combination was that of cAMP with MPA (Fig. 6C). Secreted PRL levels were below the level of detection of the ELISA.

Taken together, cAMP-primed St-T1b cells are responsive to the synthetic progestin MPA and, upon extended induction of the decidual process, acquire responsiveness to P4.

Further experiments were undertaken to assess the role of E2. Over a 7 d stimulation, the induction of IGFPB-1, dPRL and FOXO1 expression obtained with cAMP alone, cAMP plus P4, or cAMP plus MPA was not enhanced by E2. No induction was obtained with steroids alone. Moreover, morphology of the cells was not altered by the addition of E2. Western blot analysis of the samples shown in Fig. 4B revealed no induction of PR protein in decidualized cells by E2 (data not shown).