Epidermal growth factor (EGF) is one of growth factors that are thought to mediate the stimulatory effects of estrogen on the proliferation of uterine epithelial cells. The present study was attempted to obtain direct evidence for the mitogenic effects of EGF on uterine epithelial cells, and to prove that EGF and EGF receptors are expressed in these cells. Mouse uterine epithelial cells were isolated from immature female mice and cultured with or without EGF for 5 days. EGF (1 to 100 ng/ml) significantly increased the number of uterine epithelial cells, and the maximal growth (141.9± 8.3% of controls) was obtained at a dose of 10 ng/ml. In addition, EGF (0.1 to 100 ng/ml) increased the number of DNA-synthesizing cells immunocytochemically detected by bromodeoxyuridine uptake to the nucleus. Northern blot analysis revealed that the uterine epithelial cells expressed both EGF mRNA (4.7 kb) and EGF receptor mRNAs (10.5, 6.6, and 2.7 kb) These results suggest that the proliferation of uterine epithelial cells is regulated by the paracrine and/or autocrine action of EGF. Our previous study demonstrated the mitogenic effect of IGF-I on uterine epithelial cells. To examine whether the EGF- and IGF-I signaling act at the same level in the regulation of the proliferation of uterine epithelial cells, the cultured cells were simultaneously treated with IGF-I and EGF. IGF-I was found to additively stimulate the mitogenic effects of EGF, suggesting that the EGF-induced growth of uterine epithelial cells is distinct from IGF-I-induced growth.
Estrogens and progestin regulate the growth of uterine endometrial cells (Huet-Hudson et al., 1989). Estrogen stimulates the growth of uterine epithelial cells in vivo but not in vitro (Fukamachi and McLachlan, 1991; Astrahantseff and Morris, 1994). It has been proposed that the estrogen-induced proliferation of uterine epithelial cells is mediated by several growth factors in an autocrine and/or paracrine manner (DiAugustine et al., 1988; Brigstock, 1991; Nelson et al., 1991; Borgundvaag et al., 1992; Das et al., 1994; Wang et al., 1994; Zhang et al., 1994).
EGF is thought to be a mediator of estrogen action, as EGF stimulates DNA synthesis in mouse uterine epithelial cells both in vitro and in vivo (Tomooka et al., 1986; Ghosh et al., 1991; Nelson et al., 1991), and EGF-specific antibody blocks the estrogen-induced mitogenesis of uterine epithelial cells (Nelson et al., 1991). Moreover, EGF is detected by immunocytochemistry in the borders of uterine luminal and glandular epithelial cells in immature mice, and estrogen treatment increases the amount of prepro-EGF mRNA levels in mouse uteri (DiAugustine et al., 1988).
In adult mice, EGF mRNA is detected only during late proestrus and estrus, and early on day 1 of pregnancy (Huet-Hudson et al., 1990), whereas immunoreactive EGF has not been detected in the luminal epithelium in estrogen-treated ovariectomized mice (Falck and Forsberg, 1996). Thus, the localization of EGF-synthesizing cells within the endometrium is still not clear. EGF receptors and EGF receptor mRNA have been detected in rat and mouse uteri (Mukku and Stancel, 1985; Iguchi et al., 1993), and EGF-high affinity binding sites are localized in the uterine cells (Mukku and Stancel, 1985; Tomooka et al., 1986; Nelson et al., 1991; Iguchi et al., 1993). However, Das et al. (1994) and Tong et al. (1996) have reported that EGF receptors are not found in the luminal and glandular epithelium in the adult mouse uteri. Thus, it is still unclear whether EGF is synthesized and EGF receptors are expressed in the uterine epithelial cells. Therefore, the aim of the present study of mouse uterine epithelial cells was to examine the expression of EGF mRNA and EGF receptor mRNA, and to clarify the effects of EGF on DNA replication in these cells in primary serum-free culture. Because we have found evidence of the mitogenic action of IGF-I on uterine epithelial cells (Shiraga et al., 1997), the effects of IGF-I on EGF action with respect to DNA replication in uterine epithelial cells were also studied.
MATERIALS AND METHODS
Immature female mice (21–23 days of age) of the ICR strain (CLEA Japan Inc., Osaka, Japan) were used in the present study. They were maintained in a temperature-controlled room and were fed with a commercial diet and tap water ad libitum. All animal experiments were performed in accordance with the Guidelines for Animal Experimentation, Faculty of Science, Okayama University, Japan, and the NIH Guide for the Care and Use of Laboratory Animals.
Isolation of uterine epithelial cells
The method for the isolation and culture of mouse uterine epithelial cells has been previously described (Shiraga et al., 1997). Briefly, uteri were dissected out, and the uterine horn was longitudinally cut to expose the endometrial surface. The tissue fragments were kept in 0.1% trypsin (Sigma, MO, USA) in Hanks' solution containing 20 mM HEPES and 0.3% bovine serum albumin at 4°C for 1 hr, and then at 37°C for 50 min. Trypsin was inactivated by soybean trypsin inhibitor (Sigma) at 37°C for 10 min. The epithelial tissues were separated from stromal tissues with forceps under a stereoscopic microscope. The epithelial tissue fragments were collected using a self-generated Percoll gradient (Pharmacia, Uppsala, Sweden). The collected epithelial fragments were further divided into smaller fragments by gentle pipetting. The cell viability was assessed by the trypan blue exclusion test, and was found to be usually more than 90% in each assay.
Preparation of collagen gel substratum for culture
Type I collagen solution was prepared from rat tail collagen fibers (Imagawa et al., 1984). Neutralized and osmolarity-adjusted collagen solution (0.16%) was added to a 12-well culture plate (Falcon, NJ, USA) and allowed to form a gel at 37°C.
Serum-free culture of isolated uterine epithelial cells
A 1:1 mixture of Dulbecco's modified Eagle's medium (DMEM) and Ham's F12 medium without phenol red and calcium was prepared and then supplemented with CaCl2 (0.1 mM), bovine serum albumin (fraction V, 1,000 mg/l, Sigma), hydrocortisone (100 μg/l), triiodothyronine (400 ng/l), transferrin (10 mg/l), glucagon (10 ng/l), parathormone (200 ng/l), and sodium selenite (5 μg/l).
The isolated epithelial cells were seeded at a cell density of 2.5× 104 cells/cm2 in collagen gel-coated 12-well culture plates or 9 cm-well plates for RNA extraction, and incubated in a humidified atmosphere of 5% CO2 in air at 37°C. After a 1-day preculture, EGF (Sigma) and IGF-I (Amersham, Buckinghamshire, UK) were added to the culture media at various concentrations. The culture medium was changed every 2 days.
Determination of the number of uterine epithelial cells
The number of cultured epithelial cells was determined by the tetrazolium assay using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) as a substrate (MTT assay; Alley et al., 1988). The amount of formazan after 4-hr incubation at 37°C was spectrophotometrically determined by measuring the absorbance at 540 nm. The amount of formazan produced correlated well with the number of cells in culture (Shiraga et al., 1997); hence, the MTT method was used to estimate cell numbers in the culture wells. The cell number was expressed as a percentage, with the cell number for the controls regarded to be 100%.
Determination of the proliferating cells
DNA-synthesizing cells were immunocytochemically detected using a Cell proliferation kit (Amersham). Epithelial cells were seeded on collagen-coated slide glass. The collagen-coated slide glass was prepared as follows: collagen solution (0.07%) was placed on slide glass at a concentration of 100 μg/mm2 and dried under a laminar in sterile air. Bromodeoxyuridine (BrdU) was added to the culture medium (3 mg/ml), and the cultures were continued for 5 hr. The percentage of BrdU-labeled cells to total cells was determined.
Northern blot hybridization
Total RNA was extracted from the uterine epithelial cells and the submaxillary glands of adult male mice using the single-step method (Chomczynski and Sacchi, 1987). The samples were electrophoresed on 1% formaldehyde-agarose gel and transferred to nylon membranes (Hybond N+, Amersham, UK). Hybridization was carried out with a 32P-labeled cDNA probe for 16 hr at 42°C. After hybridization, the membranes were washed under stringent conditions for 15 min at 42°C in 5 x SSPE (20 x SSPE=3 M NaCl, 0.1 M NaH2PO4·H2O and 0.02 M EDTA), for 30 min at 42°C in 1 x SSPE with 0.1% sodium dodecyl sulfate (SDS), and finally for 30 min at 42°C in 0.5 x SSPE with 0.1% SDS. Hybridized signals were detected by autoradiography.
Probes for Northern blot hybridization
A 960-bp mouse EGF cDNA clone (pmEGF-26F12) obtained from the American Type culture Collection (ATCC, MD, USA) and a 452-bp mouse EGF receptor cDNA clone (pcrmEGFR-NK1) prepared and verified by sequencing in our laboratory were used. For the Northern blot hybridization, the cDNAs were labeled with deoxycytidine-5′-[α- 32P] triphosphate by the random-primed labeling method using the Random Primer DNA Labeling Kit (Takara, Otsu, Japan).
Immunocytochemical analysis of uterine epithelial cells
Uterine epithelial cells were cultured on the collagen-coated glass slides in medium containing insulin (100 ng/ml) and EGF (10 ng/ml). The cells were fixed by pre-chilled methanol (–20°C) for 30 min and dried overnight at room temperature. For the detection of cytokeratin, a specific marker for epithelial cells, the cells were treated with 0.05% trypsin in PBS for 15 min at 37°C, then incubated with pkk-1 mouse anti-cytokeratin antibody (Labsystems, Finland) for 72 hr at 4°C. Cytokeratin immunoreactivity was visualized using an ABC kit (Vector Laboratories, CA, USA).
Observation of cultured uterine epithelial cells
Uterine endometrial epithelial cells were easily separated from endometrial stromal cells by treatment with trypsin and gentle pipetting. Most epithelial cells attached to a collagen gel substratum within a few hours after seeding. Phase-contrast microscopy demonstrated that the cells were flattened, forming polygonal sheets, a characteristic of epithelial cells in culture. Immunocytochemical analysis showed that the cultured cells expressed cytokeratin (Fig. 1). These observations indicate that the cultured cells were derived from the uterine endometrial epithelial cells alone.
Northern blot analysis of EGF mRNA and EGF receptor mRNA
To determine whether the endometrial epithelial cells synthesize EGF and EGF receptors, Northern blot analysis was performed with the labeled EGF- and EGF receptor-cDNA probes. Male submaxillary glands contained high levels of approximately 4.7-kb EGF mRNA transcripts, and the uterine epithelial cells cultured for 6 days contained EGF mRNA transcripts of the same size (Fig. 2). The size of these transcripts is consistent with the sizes previously observed in mouse uteri (DiAugustine et al., 1988).
EGF receptor mRNA transcripts (10.5, 6.6 and 2.7 kb in size) were detected in the cultured uterine epithelial cells. The major transcripts of 2.7 and 6.6 kb correspond to the truncated, and full-length receptor mRNAs, respectively, as reported by Das et al. (1994).
Effect of EGF on the growth of uterine epithelial cells
The number of uterine epithelial cells was determined by the MTT method. EGF increased the epithelial cell number in a dose-dependent manner (1 to 100 ng/ml) on day 5 of culture (Fig. 3). Maximal growth was obtained at a dose of 10 ng/ml (141.9± 8.3% compared with controls). The epithelial cells continued to grow until day 5 of culture with EGF (10 ng/ml) treatment (Fig. 4). DNA-replicating cells were detected by the immunocytochemical observation of BrdU uptake to the cell nucleus. EGF (10 ng/ml) increased the percentage of BrdU- labeled cells in a dose-dependent manner (Fig. 5).
Interaction between EGF- and IGF-I induced proliferation of mouse uterine epithelial cells
We found in our previous study that IGF-I stimulates the proliferation of uterine epithelial cells in a dose-dependent manner (Shiraga et al., 1997). In the present study, the uterine epithelial cells were cultured for 5 days with EGF (0.1 ng/ml, low), EGF (10 ng/ml, high), IGF-I (0.1 ng/ml, low), IGF-I (10 ng/ml, high), and combinations of these doses (Fig. 6). EGF and IGF-I stimulated the proliferation of uterine epithelial cells. Combination treatment with high doses of EGF (10 ng/ml) and IGF-I (10 ng/ml) increased the cell number more than a single treatment with EGF (10 ng/ml) or IGF-I (10 ng/ml). IGF-I enhanced the EGF-induced growth in an additive manner.
We have established a primary culture system for mouse uterine epithelial cells (Shiraga et al., 1997). Most of the cultured cells express cytokeratin-immunoreactivity and show a characteristic morphology of epithelial cells in culture. These observations indicate that the cultured uterine cells in our system consist entirely of epithelial cells. Using this culture system, we have found that EGF treatment increases the total number of cultured cells as well as the number of DNA-replicating cells in a dose-dependent manner, which is consistent with the results reported by Tomooka et al. (1986). Therefore, these findings lead us to conclude that EGF directly stimulates the proliferation of epithelial cells through EGF receptors, since uterine epithelial cells alone are cultured under serum-free conditions in our system.
In the present study, Northern blot analysis demonstrated that uterine epithelial cells from the immature mice synthesized a 4.7-kb transcript of EGF mRNA identical in size to the EGF mRNA transcript from the mouse submaxillary gland. In contrast, Huet-Hudson et al. (1990) have reported that adult uteri synthesize a 2.4-kb but not a 4.8-kb EGF mRNA transcript, with this 4.8 kb transcript being designated as a 4.7-kb transcript in the present study. The reason for this discrepancy is unclear, but we assume that there may be an age-related difference in the regulatory mechanism of EGF gene expression: for example, a 4.7-kb transcript in immature mouse uteri and a 2.4-kb transcript in adult mouse uteri.
DiAugustine et al. (1988) have found immunocytochemically that EGF is located at the border of luminal and glandular epithelial cells in immature and adult mice. Furthermore, in an in situ hybridization analysis of adult mouse uteri, EGF mRNA has been detected in the luminal and glandular epithelial cells (Huet-Hudson et al., 1990). These findings together with the findings of our Northern blot analysis indicate that EGF is produced in uterine epithelial cells, and hence that the secreted EGF very likely regulates the proliferation of uterine epithelial cells in an autocrine manner.
Estrogen treatment increases the EGF mRNA levels in both immature mouse uteri (DiAugustine et al., 1988) and adult ovariectomized mouse uteri (Huet-Hudson et al., 1990). EGF expression changes with the estrous cycle in adult uteri (Huet-Hudson et al., 1990); estrogen may therefore be required for EGF synthesis in the adult uterus, and changes in uterine EGF expression in adult female mice are considered to be caused by cyclic changes in blood estrogen levels.
Two distinct forms of EGF receptors are found in the mouse uterus, a full-length form (170 kDa), which is a functional receptor, and a truncated form (95 kDa), which is a secreted protein; these forms correspond to the 6.5-kb and 2.7-kb transcripts (Das et al., 1994; Tong et al., 1996). The truncated EGF receptors are thought to bind to EGF or EGF-related growth factors, or to interact with the full-length EGF receptors located in the cell membrane, resulting in the diminished action of EGF or EGF-related growth factors. We have demonstrated that both forms of EGF receptor transcripts are expressed in the cultured uterine epithelial cells from immature mice. Thus, the present study provides the first evidence at the gene-transcription level that cultured uterine epithelial cells of immature mice express both the full-length and truncated EGF receptors. Because EGF can bind to both full-length and truncated receptors, the growth-promoting activity of EGF in uterine epithelial cells depends upon the balance between the numbers of both types of EGF receptors expressed.
Das et al. (1994) have reported that the uterine epithelial cells in the adult mice lack the transcripts of full-length EGF receptors, indicating that functional EGF receptors are not present in the epithelial cells. The reason for the discrepancy with respect to EGF-receptor expression between the previous study and the present study is not clear. Tong et al. (1996) hypothesize that in adult mouse uteri, the EGF receptors expressed by blastocysts rather than the uterine epithelial cells are the direct physiological target site for EGF. Thus, there may be age differences in the gene expression of EGF receptors.
The presence of EGF binding sites has been demonstrated in the uterine epithelial cells of immature mice (Tomooka et al., 1986; Iguchi et al., 1993). A receptor assay has indicated that the dissociation constants (Kd) of EGF receptors in mouse uterine epithelial cells range from 0.06 to 1.8 nM (Mukku and Stancel, 1985; Tomooka et al., 1986; Nelson et al., 1992). In the present study, the range of Kd in the EGF receptors was within the concentrations of EGF that significantly increased the number of epithelial cells. Hence, the EGF-EGF receptor system in the uterine epithelial cells is thought to be involved in the regulation of uterine epithelial cell function.
It is known that IGF-I stimulates DNA synthesis in epithelial cells (Shiraga et al., 1997). In addition, it has previously been demonstrated that uterine epithelial cells synthesize IGF-I (Murphy et al., 1987; Ghahary et al., 1990; Carlsson and Billig, 1991; Kapur et al., 1992; Shiraga et al., 1997). The present and these previous findings support the hypothesis that IGF-I synthesized in the epithelial cells regulates the growth and function of these cells. In the present study, IGF-I acted additively with EGF to increase the cell number, suggesting that the signal pathways mediated by these growth factors may be separate, although we cannot eliminate the possibility that EGF-responsive cells and IGF-I-responsive cells are different. The interaction of EGF-and IGF-I-expression has been previously reported in the mouse uterus based on a semiquantitative RT-PCR assay (Hana and Murphy, 1994). It was found that EGF increases IGF-I expression in the stromal cells and, in contrast, that IGF-I increases EGF expression in the epithelial cells. Based on these findings, it was concluded that these paracrine actions between epithelial and stromal cells may facilitate endometrial cell proliferation. Consequently, it appears that the autocrine and paracrine action of EGF and IGF-I produced in the endometrial cells may be important for the regulation of epithelial cell growth. Further study is needed to clarify the interactions between EGF- and IGF-I-system in epithelial cells.
In conclusion, it was found that EGF stimulates the growth of mouse uterine epithelial cells in primary serum-free culture. EGF mRNA and EGF receptor mRNA (full-length and truncated forms) were detected using Northern analysis in the mouse uterine epithelial cells in vitro. Our findings suggest that EGF action on epithelial cells is mediated through EGF receptors located in the epithelial cells. In immature mice, uterine epithelial EGF may stimulate the proliferation of epithelial cells. EGF and IGF-I were found to additively stimulate uterine epithelial cell growth, indicating that the EGF-induced growth of uterine epithelial cells may be distinct from IGF-I-induced growth.
This study was supported in part by Grants-in-aid for Scientific Research from the Ministry of Education, Science, Sports and Culture, Japan to S. Takahashi.