| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Pathology and Laboratory Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
female reproductive tract, steroid hormone receptors, uterus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Toward an improved understanding of such interactions, we have chosen the human endometrium, a dynamic tissue that undergoes cyclical phases of proliferation and differentiation in response to ovarian steroid hormones. Both tissue-recombinant experiments in vivo and coculture studies in vitro have demonstrated that the responses of endometrial epithelial cells to estrogen (e.g., increase in proliferation) [3, 4] and progesterone (e.g., decrease in DNA synthesis) [5] are independent of their hormone-receptor status but are dependent on signaling from stromal cells that are hormone receptor-positive.
Our laboratory recently reported a coculture model that allows in vitro study of the molecular mechanisms that contribute to stromal cell control of proliferation in the uterus [6, 7]. Briefly, adherent endometrial stromal cells are grown on multiwell plates into which membrane inserts supporting epithelial cells are suspended in the culture medium; reconstituted basement membrane complex may be either added or omitted as a function of the experimental protocol [7]. This methodology allows the two cell types to share growth medium and the soluble paracrine factors in the medium to be analyzed.
We previously demonstrated that Ishikawa endometrial adenocarcinoma cells grown under coculture conditions become responsive to stromal control of proliferation [7]. In that study, the stromal cell source was derived from primary cultures obtained from endometrial biopsy samples. The procurement and use of primary cultures of normal stromal cells for experiments in vitro often present challenges. First, normal endometrial stromal cells senesce predictably as they are passaged in culture [8]; the initially robust hormonal response typically is lost by the third to sixth passage. Second, the need to replenish continuously our supply of stromal cells with fresh biopsy samples often is limited by clinical availability. Third, patient-to-patient variability is an ongoing concern for experimental interpretation.
To circumvent these difficulties, we attempted to construct an immortalized stromal cell line that would respond to hormones in a manner like of normal stromal cells. Previous efforts in our laboratory have resulted in the generation of immortalized endometrial stromal cells by transfection with an origin-defective SV40 construct [9]. The stromal cells transfected with SV40 large T antigen have been useful for the study of transformation, but as is typical of SV40-transformed cells, the immortalization process results in genetic instability leading to numerous chromosomal aberrations [10, 11]. Furthermore, the SV40 large T antigen interacts with the CREB-binding protein [12], an essential component of the estrogen-receptor complex that permits the genomic effects of estrogen [13]. These facts preclude the use of SV40-derived cells for studies regarding normal mediation of hormonal signaling. Numerous laboratories have immortalized fibroblasts successfully by exogenously expressing the gene coding for the human telomerase reverse transcriptase (TERT) subunit [14, 15].
Telomerase is a ribonucleoprotein expressed in 85–90% of all human cancers, and its primary role is to preserve the integrity of the telomeres of replicating chromosomes. It is expressed differentially in embryonic cells and sporadically in germ line cells, stem cells, and rapidly dividing cells, including proliferating endometrial epithelial cells [16, 17]. However, telomerase is silenced in most differentiated cells because of inactivation of the reverse transcriptase subunit of the enzyme. Although cells expressing TERT are immortalized, they lack characteristics of transformed cells, such as serum independence, lack of contact inhibition, and anchorage-independent growth [18–20]. Because fibroblasts and stromal cells are both of mesenchymal lineage, we hypothesized that exogenous expression of human telomerase in stromal cells would increase their life span while preserving normal hormone-signaling pathways. An identical strategy was adopted by Krikun et al. [21], who first demonstrated that telomerized endometrial stromal cells maintain their ability to decidualize in response to progesterone stimulation.
In the present study, we address additional aspects of hormonal regulation and show directly that stromal expression of the two estrogen receptors as well as progesterone receptor (PGR) isoform A is not changed by the expression of telomerase, that parental and telomerized strains have a very similar pattern of basal gene expression, and that all three receptors appear to function normally in mediating the proliferative response to hormones in the coculture system. Furthermore, in agreement with the results of Krikun et al. [21], we show that our telomerized stromal cell line has the capacity to express the widely accepted decidualization markers prolactin [22] and plasminogen-activator inhibitor type-1 (PAI1; http://www.gene.ucl.ac.uk/nomenclature/data/gdlw_columndef.html-gd_app_symSERPINE1) [23]. We extend their finding by characterizing the TERT-expressing stromal cells, and we explore the function of these cells in regulating homeostasis in a reconstructed endometrial tissue.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Primary human endometrial stromal cells were obtained from biopsies following our published protocol [6]. Endometrial tissue was obtained in accordance with National Institutes of Health guidelines on the use of human subjects and approved by the University of North Carolina at Chapel Hill Committee on the Protection of the Rights of Human Subjects (School of Medicine Institutional Review Board). Stromal cells and Ishikawa endometrial adenocarcinoma cells, an established endometrial epithelial cancer cell line, were maintained in SM medium consisting of a 1: 1 mixture of Ham F12 (Gibco, Invitrogen Corp., Carlsbad, CA) and M199 basic medium (Sigma, St. Louis, MO) supplemented with 5% bovine calf serum (BCS; Hyclone, Logan, UT), 0.1% Mitoplus (BD Biosciences, Becton Dickinson, Franklin Lakes, NJ), and 2 µg/ml of insulin (Sigma) and antibiotic/antimycotic solution diluted to yield 100 units/ml penicillin G sodium, 100 µg/ml streptomycin sulfate and 250 ng/ml amphotericin B (ABAM; Gibco). This and all other media were phenol red-free. Except for the decidualization study, cells were transferred to steroid-free medium (JAC4) containing 1:1 Ham F12:M199, ABAM, 4% charcoal-stripped fetal bovine serum (Hyclone), 0.25% insulin-transferrin-selenium plus lipoic acid (ITS+; Becton Dickinson), 0.1 mM phosphorylethanolamine (Sigma), and 2 mM glutamine (Gibco) before hormonal stimulation. All cultures described in the present report were maintained at 37°C in a humidified atmosphere of 5% CO2. Karyotype analysis was performed using a traditional Giemsa stain-based protocol [24] on TERT-expressing stromal cells at passage 21.
Decidualization
Duplicate cultures of SHT290 cells at passage 33 were maintained for 2 wk in Dulbecco modified Eagle medium (DMEM) with 2% fetal bovine serum (Hyclone), 20 ng/ml of epidermal growth factor (EGF; Becton Dickinson), 2 mM glutamine, and ABAM and then stimulated with 10–8 M 17ß-estradiol (E2; Sigma) in the same medium for 1 wk followed by a 2-wk exposure to 10–8 M E2 and 10–6 M 6
-methyl-17-hydroxy-progesterone acetate (MPA; Sigma). Control cells were grown for the same amount of time in the presence of E2 only.
Detection of Prolactin Message by Reverse Transcriptase-Polymerase Chain Reaction
Total RNA was prepared from these cells using an RNA purification kit from Promega (Madison, WI). A single-tube protocol was adopted to estimate the amount of prolactin mRNA in each cell population: the reverse primers for both the housekeeping-gene hypoxanthine phosphoribosyl transferase 1 (Lesch-Nyhan syndrome; HPRT1) and the prolactin gene messages were included in the same reverse-transcription reaction. Thus, only these two specific messages were reverse transcribed to cDNA. Samples were normalized by adjusting the amount of RNA included in the reverse-transcription step to obtain a similar amount of HPRT1 amplicon in the subsequent polymerase chain reaction (PCR) step. The normalized cDNA sample was then used to PCR-amplify the prolactin fragment. The PCR products were loaded on a 2% agarose gel along with a DNA-ladder standard and stained with ethidium bromide. For negative controls, cDNA was replaced by water, and the absence of genomic DNA in the RNA samples was verified by showing that direct PCR amplification of the RNA samples without a reverse-transcription step did not yield any product. Primers for the amplification of HPRT1 were designed in-house. Prolactin message was amplified by a nested-PCR procedure as described by Phelps et al. [25].
Detection of PAI1 by Western Blot Analysis
The procedure for cell lysis and immunodetection is described below. The primary antibody for the detection of PAI1 was obtained from American Diagnostica, Inc. (Greenwich, CT). Recombinant PAI1 protein was a gift from Dr. Frank Church (University of North Carolina, Chapel Hill, NC).
Assessment of Cell Proliferation
Stromal cells grown in JAC4 were seeded in 12-well cluster culture plates (Becton Dickinson) at 50 000 cells per well. The next day, Ishikawa cells grown in JAC4 were seeded onto cell culture inserts with 0.4-µm porosity (Becton Dickinson) at a density of 4000 cells/insert. The inserts were suspended above the adherent stromal cells. Cocultures were treated with hormones in JAC4 for 2 days, then switched to a serum-free medium formulated as F12/M199 mixture, 0.25% ITS+, 0.25% BSA (Sigma), 20 ng/ml of EGF, ABAM, 2 mM glutamine, and appropriate steroids for the last 2 days of culture. To assess proliferation, Ishikawa cells were detached from the inserts using a 1:1 trypsin:EDTA mix and versene (Gibco), thoroughly mixed to obtain single cell suspensions, and enumerated with a Coulter Counter (Beckman Coulter, Inc., Fullerton, CA).
To measure the effects of estrogen on proliferation, stromal cells were mixed with 0.4 ml of growth factor-reduced, phenol red-free Matrigel (BD Biosciences) as previously described [6]. Cocultures were equilibrated in steroid-free JAC4 for 48 h before the addition of 10–8 M E2. To measure the effects of progestin, cocultures were equilibrated in JAC4 plus 10–8 M E2 for 48 h before exposure to the pure progestin ORG 2058 or the antiprogestin ORG 31710 (10–7 and 10–6 M, respectively; generous gifts of Organon, Rotterdam, The Netherlands). Each treatment was performed on a total of six cocultures per experiment, and each experiment was repeated at least twice. A Student t-test (homeoscedastic) was applied to ascertain the significance (P < 0.05) of apparent differences in cell numbers between treatments.
Construction of Telomerase-Expressing Stromal Cells and Telomeric Repeat Amplification Protocol
The plasmid pDSWK-8 encodes a version of the retroviral pBABE-hygro-TERT construct of Dr. Robert A. Weinberg (Whitehead Institute for Biomedical Research, Cambridge, MA) modified to express puromycin resistance instead of hygromycin resistance and kindly donated to us by Dr. William K. Kaufmann (University of North Carolina, Chapel Hill, NC) [26]. Vesicular stomatitis virus-G pseudotyped, replication-incompetent retroviruses were produced by transient three-plasmid transfection into host cells (HEK-293T, or human embryonic kidney cells) [27]. Normal stromal cells, designated as NS290 and derived from a patient with normal endometrium, were plated as 50 000 cells at passage 1 and infected with the pDSWK-8 retrovirus in the presence of 8 µg/ml of hexadimethrine bromide (Sigma) for 14 h. Starting the next day, these cells and a mock-infected control were treated with 200 ng/ml of puromycin (Gibco) for 10 days. Following this interval, the control uninfected NS290 cells died, and survivor cells in the infected culture were amplified and passaged twice in the presence of puromycin. The resulting telomerized cells were designated as SHT290 (stromal human TERT).
Telomerase activity was assayed with a telomeric repeat amplification protocol (TRAP) using PCR-based methodology (TRAPeze telomerase detection kit; Chemicon International, Temecula, CA). The manufacturer's instructions were strictly followed. Amplified products were separated on a 10% nondenaturing polyacrylamide gel, stained with Vista Green (Amersham Biosciences, Piscataway, NJ), and visualized using a PhosphorImager with blue fluorescence capacity (Storm 840; Amersham).
Epifluorescence Localization of Cytoskeletal Proteins and DNA
The following protocol and reagents were used to test the localization of specific cytoskeletal proteins that are established markers of stromal versus epithelial cell identity. Ishikawa cells were used as a positive control for epithelial expression markers and were maintained in EPI medium, a 1:1 mixture of Ham F12 and M199 with 0.1% Mitoplus (BD Biosciences), bovine pituitary extract (2 ml/L; BD Biosciences), 0.1% ITS+ (Becton Dickinson), and ABAM. Briefly, NS290 and SHT290 cells at passage 9 and Ishikawa cells were seeded as monocultures onto acid alcohol-cleaned, sterilized, 22- x 22-mm2, no.-1 glass coverslips (Corning, Corning, NY) placed in each well of a sterile six-well cluster culture plate (BD Biosciences). Cells were plated at 1 x 103 cells/well and incubated for 7 days, ensuring active stromal cell division. Media tested for SHT290 cell maintenance included either conventional SM (5% BCS content) or a low-serum medium that was higher in glucose content (DMEM high glucose supplemented with ABAM, 2% charcoal-stripped serum, 2 mM glutamine, and 20 ng/ml of EGF). Culture medium was changed completely every 48 h; the final change into fresh medium was 14 h before fixation. On the morning of cell labeling, coverslips were removed and rinsed three times with sterile, room-temperature PBS (pH 7.3) before undisturbed fixation with 3.7% formaldehyde in PBS for 30 min at 37°C. Cells were rinsed three times for 5 min in PBS, permeabilized with 1% Triton X-100 in PBS for 1 min, and rinsed, and then free aldehyde groups were blocked with 150 mM glycine/5% BSA in PBS for 30 min at room temperature. Cells on coverslips were single- or triple-labeled for detection of cytoskeletal and nuclear markers (i.e., vimentin/cytokeratin/DNA or vimentin/ filamentous actin [F-actin]/DNA). Primary-antibody sources for cytoskeletal localization were mouse monoclonal antibodies against vimentin, cytokeratin 13, and cytokeratin 18 (Novocastra Laboratories, Newcastle upon Tyne, U.K.) used at 1:50 dilutions for 1.5–2 h at 37°C. Controls for immunolabeling were samples that had the primary antibody omitted and substituted with PBS during the initial incubation step. All goat anti-mouse immunoglobulin G secondary antibodies for microscopy were from Molecular Probes, Inc. (Logan, UT), conjugated to either Alexa Fluor 488 or rhodamine 123, and used at 1:50 dilutions for 30 min at 37°C. The F-actin cytoskeleton also was evaluated using rhodamine-phalloidin (1:50 dilution, 50 min at 25°C; Molecular Probes). Hoechst 33342 (Sigma) was used to detect A-T base pairs of DNA for assessment of nuclear morphology. After fluoroprobe labeling, each coverslip was mounted cell-side down into approximately 20 µl of fresh antifade medium (2% propyl gallate in 50% glycerol) placed on an acid alcohol-cleaned, 3- x 1-inch glass Gold Seal slide. Microscopy was performed using a Zeiss LSM 510 Pascal laser-scanning confocal microscope in conventional (nonlaser) mode with 10 x 0.30 Plan Neofluar and 20 x 0.50 Plan Neofluar differential interference contrast objective lenses equipped with appropriate filters for rhodamine, fluorescein isothiocyanate, and 4',6'-diamidino-2-phenylindole (DAPI) epifluorescence (Carl Zeiss, Inc., Thornwood, NY). Companion images were recorded at identical camera image settings in identical focal planes except when noted. Digital images postcapture were archived using Adobe Photoshop software (Adobe Systems, Inc., San Jose, CA).
Immunodetection of Hormone Receptors
Cultures were washed in cold saline buffer, scraped in 52 µl/cm2 of modified RIPA buffer (50 mM Tris-HCl, pH 7.4; 150 mM NaCl; 1 mM EDTA; 1% NP-40 detergent; 0.25% sodium deoxycholate; 1 mM PMSF; 1 µg/ml each of aprotinin, leupeptin, and pepstatin; 1 mM sodium orthovanadate; and 1 mM sodium fluoride). The resuspended cells were transferred to a 1.5-ml microtube and incubated in ice for 30 min, then centrifuged at 13 000 x g for 30 min. The supernatants were assayed for total protein concentration using a bicinchoninic acid assay (Pierce Chemical Co., Rockford, IL). Proteins in the cell-free extracts were separated by SDS-PAGE in 10% polyacrylamide and electrotransferred onto a nitrocellulose membrane. As buffer, TBS/T (20 mM Tris-HCl, pH 7.6; 73 mM NaCl; and 0.1% Tween-20) was used for all steps of the immunodetection, and each step was preceded by three 5-min washes at room temperature. Blocking was conducted in 5% nonfat dry milk for 1 h at room temperature. The blot was incubated overnight at 4°C with primary antibody diluted in 5% BSA-TBS/T, then exposed to secondary antibody linked to horseradish peroxidase (1:2000 in 5% nonfat dry milk in TBS/T; Amersham) for 1 h at room temperature. For protein detection, the blot was incubated in a luminol substrate (Pierce) for 5 min, covered in plastic wrap, and exposed to x-ray film. The primary antibodies used were anti-estrogen receptor-1 rabbit polyclonal H-184 (1:1000 dilution; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), anti-estrogen receptor-2 mouse clone 9.88 (1:2000 dilution; Oncogene, Boston, MA), and anti-PGR mouse monoclonal antibody clone PR-2C5 (1:500 dilution; Zymed Laboratories, Inc., South San Francisco, CA).
Complementary DNA Microarray Analysis of Gene Expression
We isolated mRNA from cultures of normal parental and TERT-expressing stromal cells using a Fast Track kit according to manufacturer instructions (Invitrogen). To perform two-color microarray experiments, Cy3- and Cy5-labeled DNA probes were produced from both samples (Amersham reagents). The Cy3- and Cy5-labeled cDNAs were hybridized to two human 1-cDNA microarrays (Agilent Technologies, Palo Alto, CA) containing 12 814 unique human genes. The arrays were scanned with an Axon scanner (Molecular Devices Corporation, Union City, CA). We analyzed the resulting images with GenePix software (Molecular Devices). Differentially expressed genes were selected using the automated functions of the UNC Microarray Database (https://genome.unc.edu). Only genes that changed by greater than twofold on more than one array were considered to be differentially expressed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
First-passage stromal cell cultures were infected with a retrovirus encoding both the human TERT subunit and puromycin resistance under the control of the same cytomegalovirus (CMV) promoter. The two genes are linked by an internal ribosomal entry sequence so that the TERT gene must be transcribed for puromycin resistance to be expressed. The transductants were grown in medium containing puromycin for two passages. This puromycin-resistant population was designated as SHT290, and it is routinely grown in the absence of puromycin. We subsequently plated the SHT290 cells at very low density for cloning. Twelve clones were amplified to approximately 1 x 106 cells; this represents 20 population-doublings, a number of doublings that normal endometrial stromal cell populations typically cannot attain without senescing. Selected clones and the SHT290 population were shown to express a heat-inactivatable telomerase activity by a TRAP assay, whereas the normal stromal cell parental strain did not (Fig. 1). We conclude from the successful cloning of stromal cells expressing TERT that, like fibroblasts, endometrial stromal cells can have their life span greatly extended by exogenously expressing telomerase alone. For the remainder of the present study, we used the uncloned stromal population expressing TERT, because the normal cells we wished to compare them with were not clonal. The TERT-expressing population has exceeded 100 population-doublings to date, suggesting that they are immortalized. Karyotype analysis of telomerized SHT290 cells at passage 21 indicated that the cells were diploid without chromosomal anomalies (data not shown).
|
Comparative Cell Morphology
The cellular morphology of the TERT-expressing and normal stromal cells grown as monocultures in DMEM-based medium were similar. Both cell sources exhibited analogous growth patterns; uniform bipolar, fibroblast-like shape; and often, prominent stress fibers with active lamellipodia (Fig. 2). As seen with NS290 cells, telomerized SHT290 cells showed negligible labeling for cytokeratins 18 and 13, conventional markers of glandular and lumenal epithelium, and they had compartmentalization of DNA within intact nuclear membranes. Both NS290 and SHT290 cell populations demonstrated strong expression of vimentin and its colocalization with F-actin to a level typically associated with differentiated stromal cells of mesenchymal origin. Distinct preferences in medium composition were observed between the two cell types based on comparative parameters of cytoskeletal architecture. Although normal stromal cells typically exhibited similar morphology and F-actin labeling across a variety of low- and high-serum media tested, the SHT290 cells proliferated faster and exhibited more conventional stress-fiber cytoarchitecture in DMEM high-glucose medium supplemented with EGF (data not shown).
|
Comparison of the Pattern of Basal Gene Expression in Normal and TERT-Expressing Stromal Cells
To assess the impact of exogenous expression of TERT on the overall pattern of basal gene expression, we compared normal and telomerized stromal strains by DNA microarray analysis (Table 1). The mRNA from the SHT290 cell line was used to synthesize Cy5-labeled cDNA. Normal parental-strain NS290 was used to synthesize a Cy3-labeled cDNA probe. The two probes were cohybridized to commercially available cDNA microarrays. The ratio of red to green for each spot on the array was evaluated to identify differences in gene expression. The correlation between the duplicate experiments was high, with a Pearson correlation coefficient of 0.82. Table 1 lists the genes with an expression level that changed by more than twofold on more than one array. An R:G ratio of greater than two indicates that the gene was upregulated at least twofold in the SHT290 strain relative to parental strain. Out of 12 814 unique human genes on the array, seven genes were downregulated, and eight genes were upregulated. Among those genes that did change, the magnitude of the fold-change was low (less than threefold) in all cases. Thus, exogenous expression of the TERT subunit in normal endometrial stromal cells caused limited changes in gene expression.
|
Comparison of Hormone-Receptor Expression in Normal and TERT-Expressing Stromal Cells
To assess the expression of the estrogen receptors 1 and 2 as well as the PGR, we grew the parental and telomerized stromal cells in a steroid-free medium to which we added either 10–8 M E2 or 10–6 M MPA for 5 days. After this priming period, the cells were exposed to an inducing dose of the opposite hormone (i.e., the one with which they had not been primed). Thus, cells grown for 5 days in E2 were induced with 10–6 M MPA, and cells grown in MPA for 5 days were induced with 10–8 M E2. In this way, we hoped to evaluate the long-term and short-term responses of endometrial stromal cells to hormones using Western blot analysis.
No overall differences in hormone-receptor expression were found between the parental cells and the telomerized derivative (Fig. 3). Both express estrogen receptor 1 in a constitutive manner, and both show a slight induction of estrogen receptor 2 by short-term exposure to estrogen after priming with MPA (Fig. 3A). Note that because of the limited amount of early passage normal stromal cells, only half the amount of total proteins could be loaded for lanes 4–6 of the gel probed for expression of estrogen receptor 2. Taking into account this difference in protein loading, the telomerized cell strain is nearly identical to the parent. Under our growth conditions, the most marked variation in protein levels in both cell strains was for the A isoform of PGR. Long-term growth in MPA suppresses the expression of PGR isoform A, and at least 48 h of exposure to estrogen is required to restore expression (data not shown). Long-term growth in estrogen, on the other hand, allows lasting expression of PGR isoform A. Recent use of a different antibody source suggests that the SHT290 cells express a low level of PGR isoform B (data not shown).
|
Expression of Decidualization Markers Following Long-Term Treatment with E2 and Progesterone
Prolonged exposure to E2 and progesterone induces expression of prolactin (assessed by reverse transcriptase-PCR) and of PAI1 (assessed by Western blot analysis), as shown in Figure 4. The prolactin amplicon derived by PCR was identical to the corresponding cDNA sequence as reported by the BLAST sequence comparison (Fig. 4A). As shown in Figure 4B, the PAI1 molecule detected by Western blot analysis in the cell-free extracts of control and test cells migrated slightly slower than the recombinant unglycosylated PAI1 protein. We attribute this difference in migration to the glycosylation of native PAI1 [28]. These results indicate that the telomerized stromal cells are capable of expressing decidualization-specific markers when treated appropriately with hormones. Similar to the results obtained by Krikun et al. [21], we did not observe the typical enlarged and cuboidal morphology of uterine decidual cells.
|
Paracrine Regulation of Ishikawa Cell Proliferation in Coculture
We previously demonstrated that normal stromal cells stimulate the proliferation of Ishikawa cells in the presence of estrogen [7]. This result has now been reproduced using the telomerized stromal cells in the coculture assays, suggesting that the estrogen receptors are functional in SHT290 (Fig. 5A). To assess the functionality of the PGRs expressed in the telomerized cells, we compared the effect of adding the pure progestin ORG 2058 to estrogen-treated cocultures of Ishikawa cells and stromal cells (normal or telomerized). The cocultures were primed with 10–8 M E2 for 48 h to maximize PGR expression, after which they received a mixture of 10–8 M E2 and 10–7 M ORG 2058 for 4 days. Enumeration of Ishikawa cells at the end of this treatment demonstrated that progestin inhibits the estrogen-driven proliferation only if stromal cells are present in the cocultures. Importantly, the telomerized SHT290 cells retained the ability to inhibit proliferation of Ishikawa cells to the same extent as NS290 cells. In subsequent experiments, we modified our growth conditions to include a serum-free version of the JAC medium that allows maintenance of both Ishikawa cells and normal stromal cells for relatively short intervals (48 h; data not shown). Early passage normal stromal cells were no longer available over time, yet substitution with the telomerized cells showed that the estrogen-driven proliferation of Ishikawa cells was inhibited by 34% in response to the addition of ORG 2058. Furthermore, this effect of ORG 2058 could be abrogated by adding 10–6 M of the antiprogestin ORG 31710 (Fig. 5B), confirming that this effect of the progestin is mediated through the PGR.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our ability to propagate clonal cultures of the stromal cells expressing telomerase strongly suggests that they have been immortalized. We draw this conclusion because it is difficult to clone normal endometrial stromal cells because of their rapid senescence. Instability of hormone-receptor expression in cell cultures often is reported, and a review of the literature reveals that experiments probing hormone effects on endometrial stromal cells are always carried out with cells at early passage. Our telomerized stromal cells at passages 9–33 are as responsive to progesterone or estrogen as their normal parental cells are. The diploid karyotype without chromosomal anomalies observed for our TERT stromal cells supports other reports of karyotype stability, even under conditions of ionizing radiation, for human TERT-immortalized fibroblasts [31]. It is our hope that the availability of the SHT290 cells will facilitate the study of hormonal regulation of stromal cell physiology as well as stromal-epithelial interactions in human endometrial tissue.
Interestingly, the telomerized stromal cells had requirements that were more stringent regarding the type of growth medium that they would tolerate. The preferred requirement for a high-glucose medium could be the result of an increase in the fraction of metabolically active cells present at any one time in the population, or it could reflect more profound changes in the physiology of those cells. We have observed that in the routine SM medium used to grow normal stromal cells, the SHT290 telomerized stromal cells often adopt an enlarged or hypertrophic morphology, accompanied by an increase in the length and width of stress-fiber bundles. This phenomenon may be a response to the expression of an active telomerase; to date, it has been observed after transduction of TERT into two additional stromal strains originating from different patients. The molecular basis for this phenomenon is obscure at present. It must be noted that many laboratories already grow endometrial stromal cells in high-glucose DMEM, so normal stromal cells will not be compromised by being grown under conditions that might favor a normal phenotype in the telomerized strain. Further in-depth studies of the SHT290 cells will be needed to explain the cell physiology that may be responsible for this medium-dependent phenotype.
The overall pattern of basal gene expression in the TERT-expressing endometrial stromal cells showed few differences from that in normal parental cells. Future studies will address the impact of these changes in gene expression on cell signaling and behavior. As we generate more telomerized cell strains, it will become clearer which of these observed changes are intrinsic characteristics of repeated passages of telomerized cells in artificial medium in vitro or a consequence of telomerase expression in human endometrial stromal cells. Of particular importance will be the assessment of changes in the expression of extracellular matrix proteins (Table 1) that could influence, either directly or indirectly, homeostasis of the stromal cells as well as their interaction with epithelial cells. Our results strongly suggest that expression of an active telomerase has no marked effect on regulation of hormone-receptor expression, nor does it appear to affect the functionality of selected hormone receptors in our coculture assay.
Assessing the normality of the pattern of hormone-receptor expression in our stromal cells in vitro can be difficult. Uncertainties about the interpretation of hormone-receptor expression in vivo often are a function of different methodologies used to assess receptor activity and function. For example, regulation of uterine steroid hormone-receptor expression as studied with immunocytochemistry of human tissue sections often is reported as representative of hormonal regulation; however, many additional factors operate in human tissue. Such differences may explain why contradictory results often are reported when comparing regulation of hormone-receptor expression using methods that measure exogenous expression versus a reporter gene-promoter fusion in a functional assay [32–34].
Paracrine signaling between stromal and epithelial cells in normal human endometrium is dependent on the appropriate regulation of stromal cell hormone-receptor expression. It generally is accepted that in vivo, the magnitude of variation in hormone-receptor expression is relatively low in the stromal compartment compared with either the native endometrial epithelium or the hormone-responsive epithelial cell lines. Expression of estrogen receptors peaks during the late proliferative/early secretory phase of the cycle and then declines, reaching a minimum in late secretory phase [33, 35]. We could not reproduce this pattern of regulation of estrogen-receptor expression in vitro by switching the cultures from estrogen to progesterone, or vice versa. It is not unexpected that additional factors other than the presence of a single hormone are necessary for regulating expression of the estrogen receptors in human endometrial stromal cells. Studies of patients undergoing sequential hormone-replacement therapy have shown a distinct heterogeneity in the degree of differentiation found in adjacent glands during the pseudoluteal phase. The less differentiated epithelial glands maintain high levels of estrogen-receptor expression; such glands are surrounded by stromal cells that also express relatively high levels of estrogen receptors [36]. An additional requirement may exist for the presence of low-estrogen concentration together with the dominant progesterone concentration when mimicking the secretory phase.
We did observe hormone-induced regulation of the expression of the progesterone receptor A after several days of exposure. This suggests that these are indirect but endogenous and/or autocrine effects requiring the synthesis of other signaling molecules. Overall, our results are consistent with those of the immunocytochemical studies previously cited [33, 35], which found the level of progesterone-receptor expression to be highest in the late proliferative phase and to decrease slowly during the secretory phase. We found that estrogen stimulates the expression of the PGR isoform A and that progesterone inhibits it, but the magnitude of the effect of progesterone seems to be amplified in our system. Our finding is consistent with those of others who have shown that the PGR isoform A is expressed predominantly in stromal cells [37, 38]. Although we were able to detect a very low level of PGR isoform B in the SHT290 cells using a different antibody, the normal parental strain was no longer available for comparison. Another possible explanation for the differences between our results and those of previous studies could be, at least in part, individual genetic variation between patients.
Our important finding is that expression of an active telomerase does not significantly alter the pattern of hormone-receptor expression. Telomerized stromal cells are as efficient as normal stromal cells in regulating the growth of epithelial cells in response to ovarian steroid hormones. We also verified that our telomerized stromal strain is capable of expressing markers of decidualization following culture in the presence of estrogen and progesterone for 2 wk. These results are similar to those of Krikun et al. [21], who compared normal and telomerized endometrial stromal cells and concluded that expression of telomerase does not affect the functionality of the PGR in the decidualization response. Taken together, these data indicate that the estrogen receptors as well as the PGRs function normally in the telomerized cell population.
In summary, the telomerized stromal cells are an appropriate substitute for normal stromal cells in our coculture assay. The availability of a consistent source of stromal cells with normal characteristics now allows the study of paracrine interactions that regulate hormone responses, expands the applications of our coculture system, and provides large populations of stromal cells that can be grown for genomic and proteomic studies. The SHT290 cells will permit future genetic modifications and coculture applications to carefully dissect the cell signaling and cell functions that underlie stromal-epithelial interactions in the regulation of normal, preneoplastic, and neoplastic human endometrium.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: David G. Kaufman, Department of Pathology and Laboratory Medicine, Brinkhous-Bullitt Building Room 620, CB 7525, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599. FAX: 919 966 5046; uncdgk{at}med.unc.edu
Received: 9 August 2004.
First decision: 13 September 2004.
Accepted: 10 March 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Barbier, H. J. Kloosterboer, and D. G. Kaufman Effects of Tibolone Metabolites on Human Endometrial Cell Lines in Co-culture Reproductive Sciences, January 1, 2008; 15(1): 75 - 82. [Abstract] [PDF]
|
|
|
|
 |
|
 P. Chapdelaine, J. Kang, S. Boucher-Kovalik, N. Caron, J. P. Tremblay, and M. A. Fortier Decidualization and maintenance of a functional prostaglandin system in human endometrial cell lines following transformation with SV40 large T antigen Mol. Hum. Reprod., May 1, 2006; 12(5): 309 - 319. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Tang, A. Mikhailik, I. Pauli, L. C. Giudice, A. T. Fazelabas, S. Tulac, D. D. Carson, D. G. Kaufman, C. Barbier, J. W. M. Creemers, et al. Decidual Differentiation of Stromal Cells Promotes Proprotein Convertase 5/6 Expression and Lefty Processing Endocrinology, December 1, 2005; 146(12): 5313 - 5320. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
