| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mechanisms of Hormone Action |
Department of Molecular Medicine,3 Karolinska Institutet, 171 76 Stockholm, Sweden
Division for Reproductive Endocrinology,4 Department of Woman and Child Health, Karolinska Institutet, 171 76 Stockholm, Sweden
Section for Obstetrics and Gynecology,5 Department of Clinical Science, Huddinge University Hospital, 171 76 Stockholm, Sweden
Division of Urology,6 Department of Surgery, Chang Gung Memorial Hospital, Tao Yuan 333, Taiwan
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-fodrin, CD24, immediate early response 5, insulin-like growth factor-binding protein 2, growth response protein CL-6 (INSIG-1), ladinin1, class I major histocompatibility complex heavy chain, lactadherin, ezrin, and Fas-activated serine/threonine kinase. Because of their function as regulators of proliferation and apoptosis, CD24, insulin-like growth factor-binding protein 2, and Fas/Fas ligand were examined further by immunohistochemical expression and tissue localization analysis. Our analysis confirms a contrasting regulation of these gene products by ovariectomy and estrogen treatment. The present study identifies novel mediators of estrogen actions in the uterus and provides genome-wide expression data from which novel hypotheses regarding uterine function can be generated.
estradiol, gene regulation, mechanisms of hormone action, uterus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Sex steroids influence a number of diseases of the female reproductive system, including endometrial cancer and leiomyomas. However, the exact role of sex steroid hormones in the development and progression of these diseases cannot be defined until their mechanism for controlling growth in normal tissues is understood. Ovariectomy in combination with hormone replacement is an established animal model for studying sex steroid regulation of target tissues. Studies concerning regulation of gene expression have shown that genes are not always reciprocally regulated in the uterus following ovariectomy and subsequent estrogen replacement [3, 4]. This may result from the different actions of estrogen and progesterone observed in the uterus. For example, in ovariectomized (OVX) rats, estradiol can induce DNA synthesis and mitosis in the uterus, whereas progesterone inhibits DNA synthesis only in the epithelium yet stimulates mitosis in the stromal cells [5]. Previously published data demonstrate that progesterone increases uterine weight in rats with low estradiol levels (ovariectomy with or without low-estradiol-concentration implants), whereas decreased uterine weight occurs at higher estradiol levels [6].
The OVX animal model is commonly used to study the regulatory actions of sex steroid hormones on various functions in the target tissues. Therefore, it is of great relevance to characterize this state in terms of changes in gene expression. Global gene expression profiles offer unique opportunities to understand the molecular events underlying the actions of sex steroids on reproductive tissues. In the present study, we have used DNA microarrays containing 3000 different rat cDNA clones to investigate the effect of ovariectomy in the rat uterus. Attempts were made to correlate the newly described variations in the expression patterns to the physiological changes in the uterus induced by estradiol treatment.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Animal experiments were approved by the Committee on Animal Care in Sweden and conducted in accordance with the Guiding Principles for the Care and Use of Research Animals promulgated by the Society for the Study of Reproduction. Adult female Sprague-Dawley rats (BK-Universal, Sollentuna, Sweden) weighing approximately 200 g were used. The animals were housed in a controlled environment at 20°C on a 12L:12D photoperiod. Rats had access to standard pellet food and water ad libitum. In experiment 1, rats were ovariectomized under light ether anesthesia and housed for 1, 4, or 8 days before they were killed. The rats in experiment 2 were treated with estradiol after ovariectomy and received a daily dose of 2.5 mg of 17ß-estradiol per rat for 1, 4, or 7 days. Intact rats in the estrous phase were used as controls. See Figure 1 for detailed information regarding the experimental design. Uteri from OVX rats and from estrogen-treated OVX rats were used for the microarray and immunohistochemistry studies. During anesthesia, the uteri were removed, stripped of fat and connective tissue, weighed (see Tables 1 and 2), and divided. Half the uterus was immersion fixed in 4% formaldehyde at 4°C and thereafter embedded in paraffin; the remaining tissue was immediately frozen in liquid nitrogen and stored at -70°C until used for RNA isolation as described below. The wet weights were analyzed with ANOVA and the Kruskal-Wallis test. Significances were calculated using the Dunn test. A level of P < 0.05 was regarded as significant.
|
|
|
Apoptosis Analysis (TUNEL Assay)
Cells with nuclear DNA fragmentation were visualized by the TUNEL method using the ApopTag Plus in situ apoptosis detection kit (Oncor, Gaithersburg, MD). The labeling procedure was performed according to supplied instructions, and the results were used to visualize nucleosome-sized DNA ends.
Deoxyribonuclease-pretreated sections in the kit were used as positive controls. A section incubated without terminal deoxynucleotidyl transferase served as a negative control.
RNA Preparation
Uteri collected from intact rats (in the estrous phase) at 1, 4, and 8 days after ovariectomy (denoted as C1, C4, and C8, respectively) and from the OVX rats treated with estradiol for 1, 4, and 7 days (denoted as T1, T4, and T7, respectively) were used for RNA extraction. Total RNA was isolated using TRIzol reagent according to the protocol supplied by the manufacturer (Life Technologies, Inc., Rockville, MD). The quality of RNA samples was examined on a denaturing agarose gel. Equal amounts of total RNA from four rats in the same experimental group were pooled before cDNA labeling.
Fabrication of cDNA Microarrays
The procedures for cDNA microarray fabrication were similar to previously published methods [7]. Approximately 3000 cDNA clones were selected from the TIGR Rat GENE Index (www.tigr.org) and from our own collection of rat cDNA libraries as previously described [8]. The group represents a set of nonredundant clones derived from 12 different cDNA libraries, including one obtained from uterus. The majority of clones represent genes with known function in rats or are orthologues to known mouse or human genes. Genes were not selected for specific functional categories. For the array fabrication, bacterial colonies were grown overnight in 1.5 ml of LB medium [7] in 96-well microplates (Beckman Coulter, Inc., Fullerton, CA). Plasmid minipreparations were followed by polymerase chain reaction (PCR) amplification of the inserts using the vector-specific primers T3 (5'-aattaaccctcactaaaggg-3') and T7 (5'-gtaatacgactcactatagggc-3'). Amplified inserts, each produced from pooling of two 100-µl PCR reactions, were purified by ethanol precipitation, dissolved in 40 µl of 3x SSC (1x SSC: 0.15 sodium chloride and 0.015 M sodium citrate), and purity verified on an agarose gel. Less than 10% of the array elements rendered more than a single band on PCR, and these probes were excluded from analysis. The sequences of all genes mentioned here by name were reverified by single-pass 3' sequencing. CMT GAPS Amino silane-coated slides (Corning, Inc., Corning, NY) were employed for printing, using a GMS 417 arrayer (Affymetrix, Santa Clara, CA). The slides were postprocessed and stored in a dust-free dark box until hybridization.
Microarray Hybridization
The protocol employed for probe labeling and purification has been described earlier [7] and was used in the present study with some modifications. Twenty micrograms of total RNA were used from each group of animals in the hybridization. Fluorescent-labeled cDNA was synthesized using oligo-dT primer (New England Biolabs, Inc., Beverly, MA) by reverse transcription (RT) using Superscipt II (Life Technologies, Inc.) in the presence of labeled nucleotides, Cy3-uridine 5'-triphosphate for control and Cy5-uridine 5'-triphosphate, for treatment mRNA (Amersham Pharmacia Biotech, Piscataway, NJ). Cy3- and Cy5-labeled cDNAs were pooled and purified using a Microcon 30 column (Millipore Corporation, Bedford, MA). The labeled and purified cDNA was added to the array at a final volume of 15 ml of hybridization buffer (5x SSC, 0.2% SDS, 10 µg of poly-A RNA, and 10 µg of yeast tRNA). Probes were heated at 100°C for 2 min before application onto the array, covered with a plastic 22- x 22-mm cover slip (Grace Biolabs, Bend, OR), and put into a sealed hybridization chamber (Corning). Following hybridization at 65°C for 15–18 h, the array was washed and dried. The array was then immediately scanned using a GMS 418 scanner (Affymetrix), and image analysis was performed using the GenePix Pro software (Axon Instruments, Foster City, CA). Measurement of experimental groups was performed twice (RNA samples within each comparison were reversed labeled to account for Dye-bias effects), and the results were expressed as the mean of two ratios.
Data Analysis
GenePix Pro software was used for the analysis of the image. The signal of each spot was calculated as the average intensity of the spot minus background. Spots with an intensity at least 2-fold above that of the background were included in the present study. The expression ratio calculated as Cy5:Cy3 signal was normalized using the LOWESS (Locally Weighted Scatter Plot Smoother) method in the SMA (Statistics for Microarrays Analysis) package [9]. The SMA package is an add-on library written in the public domain statistical language R [10]. The LOWESS algorithm performs a local fit to the data in an intensity-dependent manner. The significance of the expression ratios of both ovariectomy and estrogen replacement studies was estimated using the Significance Analysis of Microarray (SAM) technique [11]. A q value was assigned each detectable gene in the array. This value is similar to the familiar P value, measuring the lowest false-discovery rate (FDR) at which the gene is called significant. In the present study, genes with a q value of more than 2% in either set of data were excluded from further analysis. To the statistically based criteria, we have added a further requirement based on the absolute changes in expression ratios. Only genes with average changes of 100% in at least one of the experimental groups studied were listed as differentially expressed. This change in the level of expression has been shown in previous studies to be reproducible when other direct methods, such as Northern blot analysis, RNase-protection assay, and RT-PCR, are used to estimate gene expression [7, 12–15].
Immunohistochemistry
Proliferating cell nuclear antigen (PCNA) was chosen as the cell proliferation marker in the present study. To support the gene expression observations and to provide additional information regarding the cellular localization of some gene-encoded proteins, we analyzed the protein expression of insulin-like growth factor-binding protein 2 (IGFBP-2), CD24, Fas, and Fas ligand (Fas-L) in uterine samples from intact rats, OVX rats at 8 days, and OVX rats with 7 days of estradiol treatment. The biopsy samples were embedded in paraffin and cut in sections (thickness, 5 mm). The sections were dewaxed in Bioclear (Bio-Optica, Milan, Italy) and rehydrated in decreasing concentrations of ethanol. Sections were pretreated in 0.01 M citrate buffer in a microwave oven, and normal horse serum was used as a blocking agent. The sections were then incubated with the respective primary antibody. After washing in PBS containing 0.1% Tween-20, the sections were exposed to a secondary antibody. Biotinylated horse anti-rabbit (for PCNA, IGFBP-2, Fas, and Fas-L) and anti-mouse (for CD24) were used (Vector Laboratories, Burlingame, CA) as the secondary antibodies, after which the slides were incubated with horseradish peroxidase-avidin-biotin complex (Vectastain ABC Elite; Vector Laboratories). The presence of the enzyme was visualized by 3,3-diaminobenzidine (DAB-kit; Vector Laboratories). Sections were counterstained with hematoxylin and dehydrated before mounting with Pertex (Histolab, Gothenburg, Sweden). The different immunostaining procedures for each specific antibody are described below.
PCNA A polyclonal rabbit anti-human antibody was used for detection of PCNA (FL-261; Santa Cruz Biotechnology, Santa Cruz, CA). The primary antibody was diluted in normal horse serum (1:400) and incubated overnight at 4°C.
CD24 A monoclonal mouse anti-human antibody was used for detection of CD24 (Ab-2; Lab Vision Corporation, Fremont, CA). The primary antibody was diluted in 2% BSA (1:200) and incubated for 60 min at room temperature.
IGFBP-2 A polyclonal rabbit anti-mouse antibody was used for detection of IGFBP-2 (PAS1; GroPep, East Roseville, Australia). The primary antibody was diluted in normal horse serum (1:250) and incubated overnight at 4°C.
Fas A polyclonal rabbit anti-human antibody was used for detection of Fas (M-20, sc-716; Santa Cruz Biotechnology). The primary antibody was diluted in normal horse serum (1:200) and incubated overnight at 4°C.
Fas-L A polyclonal rabbit anti-human antibody was used for detection of Fas-L (C-178, sc-6237; Santa Cruz Biotechnology). The primary antibody was diluted in normal horse serum (1:200) and incubated overnight at 4°C.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
To better characterize the system in which the expression profiles were studied, the mean rat uterine weight, the expression of the PCNA proliferation marker, and the extension of apoptosis after ovariectomy and after estrogen treatment (Tables 1 and 2 and Fig. 2) were measured. As expected, uterine wet weight was reduced by ovariectomy in a time-dependent manner. This effect was completely reversed by estrogen treatment (Tables 1 and 2). Both PCNA-positive cells and apoptotic cells were confined to the myometrium and the endometrial epithelial and stromal cells (Fig. 2, a and b). Whereas the numbers of PCNA-immunopositive cells both in epithelial and stromal cells were lower 2 wk after ovariectomy (Fig. 2c), the number of apoptotic cells was higher when compared to rats in the estrous phase. Apoptosis-positive cells increased in epithelial, stromal, and smooth muscle cells after ovariectomy (Fig. 2d) and decreased after estradiol treatment (Fig. 2f) compared with intact rats in the estrous phase (Fig. 2b). After 7 days of estradiol treatment, the PCNA-positive population increased both in epithelial and stromal cells (Fig. 2e), whereas the number of apoptotic cells decreased (Fig. 2f).
|
Gene Expression Profiling
Of the 3000 arrayed probes, approximately 50% were reproducibly detected (signal 2-fold above background) in our experimental system. The statistical significance of changes in gene expression was analyzed using the SAM statistical technique and expressed as an estimated FDR for the subset of regulated genes. Using an FDR of less than 2%, 334 genes were found to be regulated by ovariectomy, whereas 440 genes were affected by estrogen treatment. The rest of the genes did not pass the statistical criteria or were expressed below the detection limit. We concentrated on analyzing those genes for which expression changes of more than 100% could be detected in at least one of the time points analyzed in either treatment. These conservative criteria identified 33 overexpressed and 47 underexpressed genes in OVX uterus compared to normal uterus. On the other hand, estrogen treatment of OVX rats induced the uterine expression of 142 genes and reduced the expression of 175. This represents approximately 20% of the total detectable genes and confirms the extensive effects caused by estrogen treatment in the uterus. Table 3"> shows the fold-changes in expression induced by ovariectomy and estrogen treatment for the set of genes in which significant changes could be recorded in both conditions. Expression information regarding all detectable genes, including the fraction affected only by ovariectomy or estradiol treatment, can be found as supplementary data. Nineteen genes were shown to respond oppositely to OVX and estradiol treatment. They include Thy-1 protein, secreted frizzled-related protein 2 (SFRP2), ribosomal protein S4, peptidylglycine -amidating monooxygenase (PAM), -fodrin, follistatin-related protein precursor (FRSP), carbonic anhydrase III (CA3), 3-hydroxyacyl-coenzyme A dehydrogenase, syndecan-1, CD24, lactadherin, complement component C3, ladinin-1, IGFBP-2, growth response protein CL-6, immediate early response 5, glutathione peroxidase 1, class I major histocompatibility complex heavy chain, ezrin, and Fas-activated serine/threonine kinase (FAST).
|
|
Immunohistochemistry and Real-Time PCR
To our knowledge, several of the genes showing opposite regulation by ovariectomy and estradiol treatment have not been previously described. We are particularly interested in novel genes with actions that may be related to apoptosis and proliferation, because estrogen is an important regulator of cell turnover in the uterus. CD24 encodes an extracellular glycoprotein that is bound to the cell membrane via a glycosyl-phosphatidylinositol anchor and has been shown to regulate apoptosis and to stimulate tyrosine kinase activity [16, 17]. The IGFBP-2 is a secreted protein that specifically binds insulin-like growth factor (IGF)-I and IGF-II with high affinity and regulates access of the IGF receptors to the ligand [18]. In addition, FAST, an intracellular mediator of the Fas/Fas-L-induced proapoptotic pathway [19], was up-regulated in the uterus after ovariectomy and down-regulated after estradiol treatment. This prompted us to study the regulation of the Fas/Fas-L pathway in the antiapoptotic actions of estrogen.
Immunostaining of IGFBP-2 was observed in uterine epithelial cells, stromal cells, and smooth muscle cells. The staining level was decreased after ovariectomy, especially in epithelial cells. The positive staining was regained after estradiol replacement (Fig. 3, b, f, and j). A typical membrane staining of CD24 was found not only in uterine epithelial cells in the intact rats but also in uterine stromal cells following estradiol treatment of OVX rats. However, the intensity of epithelial cell staining in this group appeared to be stronger than in the intact group of rats. In addition, the level of CD24 expression was decreased in the epithelium in the OVX groups, with no staining detected in the stroma (Fig. 3, a, e, and i). Finally, Fas and Fas-L were expressed mainly in the cytoplasm of endometrial epithelial cells and myometrial cells. In the estrogen-treated groups, Fas and Fas-L immunostaining was significantly lower than in the normal intact rats. The strongest immunostaining of Fas and Fas-L in epithelium and myometrium was observed in the rats killed 2 wk after ovariectomy (Fig. 3, c, d, g, h, k, and l). During all immunohistochemical experiments, no positive staining was observed in the negative control sections (Fig. 3, m and n).
|
Changes in gene expression induced on IGFBP-2, CD24, Fas, and Fas-L by ovariectomy and estrogen treatment were additionally confirmed by real-time PCR. The expression ratios calculated after RT-PCR analysis are in agreement with the protein changes detected by immunohistochemistry and microarray analysis.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The comparison between ovariectomy and estradiol replacement reveals interesting results. The majority of genes regulated by ovariectomy do not show opposite regulation by estradiol add-back treatment, which may reflect the importance of other ovary-derived factors for uterine homeostasis. This also implies that different pathways regulate uterine regression after ovariectomy and the proliferative response to estradiol treatment. Importantly, these results confirmed previous findings regarding ovariectomy and estrogen effects. Several genes found to be oppositely affected by ovariectomy and estradiol treatment in the present study have previously been reported to be estrogen dependent in the uterus, including complement component C3 [23], SFRP2 [24], PAM [25], CA3 [26], glutathione peroxidase [27], and syndecan-1 [28]. Moreover, the uterus depends on estrogen for its development, growth, and functions [29–31]. Recently, a number of studies have been published that make use of the DNA microarray to study the mouse and rat uterus [32–34]. In a related study, Watanabe et al. [34] analyzed changes in uterine gene expression of OVX mice 6 h after a single injection of 17ß-estradiol. Despite the differences in experimental and array designs, we were able to confirm the regulation of genes such as glutathione S-transferase, epoxide hydrolase, 3-hydroxy-3-methylglutaryl-coenzyme A synthase 2, and farnesyl pyrophosphate synthase. Our identification of several genes not previously reported as estrogen regulated stresses the complex nature of the uterotrophic effects of estradiol, depending on both the dose and the time of treatment.
The overall finding of the present study is that ovariectomy and estrogen treatment alter uterine gene expression extensively. Regulated genes are directly involved in signal transduction and include growth factors, growth factor receptors, and extracellular matrix proteins. Fewer genes related to carbohydrate and lipid metabolism or regulation of transcription were altered. When the effects of ovariectomy were compared with those of estradiol treatment, it was evident that only a few genes were oppositely regulated by these two treatments. It is beyond the scope of the present paper to discuss all data. Therefore, we will limit our discussion to those genes for which the expression is altered both by ovariectomy and by estradiol replacement. These genes may be important for uterine homeostasis, because their expression correlates with both the estrogen level and the antagonistic physiological processes of proliferation and tissue regression induced by estrogen/estrogen withdrawal. One should, however, be cautious in overinterpreting our results. Genes that do not appear to be regulated by estrogen treatment could, in fact, be affected at an earlier time-point transiently and therefore escape detection. The response to estrogen in OVX rats is known to be complex, with dramatic changes in expression of extracellular matrix proteins, transcription factors, oncogenes, growth factors and their receptors. These changes may be linked to the initial phase of mitogenesis, which involves a series of early hypertrophic responses that appear 4–6 hours after hormone treatment [29]. Estrogen also induces a late hyperplastic response that develops 24–30 h after hormone administration [30, 31]. However, opposite regulation by ovariectomy and estradiol treatment does not necessarily involve direct regulation via ER. Such conclusions would require a careful analysis of promoters of the regulated genes. For example, the regulation of SFRP2 by estrogen has previously been studied in mice and has shown data similar to those reported here. The reduction of SFRP2 by estrogen has been shown to be independent of ER and ERß [24].
Our results confirm previous findings regarding the regulation of syndecan-1. Both syndecan-1 and syndecan-3 are regulated by estrogen in mouse uterus [28, 35]. Syndecans are membrane-anchored heparan sulfate proteoglycans that act as coreceptors for both vascular endothelial growth factor and basic fibroblast growth factor [36, 37]. They are composed of a small protein core linked to multiple glycosaminoglycan chains. Similar to the expression of syndecan-1, the expression of ß-1,3 glucuronosyltransferase 3 (an enzyme central to the initial steps of proteoglycan synthesis) was also up-regulated by estrogen in our experiment (Table 3) [38, 39].
To our knowledge, the contrasting regulation of CD24, IGFBP-2, FRSP, and thymosin ß4 by ovariectomy and estradiol treatment has not been described previously. We observed that expression of the CD24 gene was increased after estradiol treatment. CD24 is an extracellular glycoprotein expressed in B cells, brain, bone marrow, kidney, liver, lung, human neoplastic cell lines, and human decidual stromal cells [40–42]. Although CD24 has no cytoplasmic portion, it can mediate intracellular signaling via a glycolipid-enriched membrane-dependent mechanism [16]. CD24 has been shown to associate with intracellular tyrosine kinases [17], and several results indicate that expression of CD24 correlates with cellular proliferation in different cell types [13, 16, 17]. A possible relationship between sex steroid hormones and CD24 is also observed in the ER-positive cell lines (MCF7 and T47D), in which CD24 is overexpressed in comparison to ER-negative cell lines [43]. We also found that estrogen induced CD24 in the rat uterus. In contrast, CD24 has been shown to induce apoptosis in Burkitt lymphoma cells by a mechanism that involves activation of the mitogen-activated protein kinase (MAPK) and tyrosine kinase pathways [16]. Whether this mechanism applies to endometrial cells, however, requires experimental confirmation. Growth factor activation of MAPK in endometrial cells has been linked to proliferation [44]. Therefore, CD24-mediated activation of this pathway may promote proliferation rather than apoptosis in uterine endometrium. Another interesting finding from the present study was that CD24 protein was found in the stroma of estradiol-treated OVX rats, but little staining was seen in the intact rats. This implies that progesterone or other factors may have an inhibitory action on stromal cells.
The IGFBP-2 was regulated in a manner similar to that of CD24. Like other IGFBPs, this protein can regulate the activity of IGFs by modulating the availability of free IGFs for their receptors. The IGFs function as stimulators of cell proliferation and inhibitors of cell apoptosis in uterine epithelial cells [45]. The regulation of IGFBP-2 expression is highly complex and influenced by multiple hormones and growth factors, including estradiol. The various actions of IGFBP-2 seem to have a component of tissue specificity. For example, in vascular smooth muscle cells, IGFBP-2 can inhibit IGF-stimulated growth by competing with receptors for ligand binding [46]. In contrast, a number of studies document a positive association between IGFBP-2 expression and cell proliferation [47]. Our data reveal that both gene and protein expression of IGFBP-2 changed dramatically following changes of estrogen levels. We found that the major changes of IGFBP-2 protein in uterus during the hormonal withdrawal and estradiol treatment were found in endometrial epithelial cells rather than in stromal and smooth muscle cells. This localization matches the sites where apoptosis and proliferation take place, suggesting direct involvement of IGFBP-2 in these processes. Our findings are also supported by in situ hybridization studies in which IGFBP-2 mRNA was localized in the luminal epithelium of the rat endometrium, being more abundant during proestrus and early estrus in comparison to other stages of the menstrual cycle [48].
Thymosin ß4 is a polar, 5-kDa peptide identified as a G actin-sequestering peptide with capacity to inhibit actin polymerization, leading to increased cell motility. This protein is highly expressed in multiple tissues and seems to be transcriptionally regulated in relation to cell proliferation [49]. Thymosin ß4 is also detected outside cells, where it inhibits inflammation and induces chemotaxis and angiogenesis [50, 51]. In the uterine endometrium, angiogenesis takes place actively to support the endometrial growth that occurs during transition from the diestrous to the estrous stage. This process is controlled by estrogen and can be mimicked by injections of estradiol to OVX mice [52]. Thymosin ß4 may mediate some of the actions of estrogen, because its expression correlates with changes in proliferation and apoptosis in the uterus.
Estrogen down-regulates the expression of FRSP, which belongs to a family of extracellular matrix-associated glycoproteins, including agrin, osteonectin, and follistatin [53]. The FRSP is most related to follistatin both in sequence and gene structure. Consequently, both proteins can bind and inhibit activin and its actions. Several studies have confirmed the actions of activin on FSH pituitary production [54]. Other reports also indicate a role for activin in decidualization of the uterus and in pregnancy [55]. More direct correlation between FRSP and estrogen has been demonstrated in a study using an osteoblastic cell line (CDO7F), supporting the concept that FRSP may be an estrogen-regulated gene [56].
A surface glycoprotein of 25–29 kDa, Thy-1 has a structure of a single variable-like immunoglobulin superfamily domain anchored to the plasma membrane through a glycosylphosphatidylinositol tail [57]. The Thy-1 protein was initially characterized in rodent thymocytes and neuronal cells, with possible function as a regulator of cell-cell interactions [58]. It was later shown that Thy-1 is expressed in a variety of tissues [59]. Triggering of Thy-1 can directly activate the very early induction of an apoptotic cell death program in the kidney [60]. We observed Thy-1 expression to be higher in the uterus after ovariectomy and to be reduced after the administration of estradiol, suggesting a possible link between Thy-1 protein induction and the apoptotic process in the uterus.
In the uteri of OVX rats, FAST is another gene significantly down-regulated by estrogen treatment. Although ovariectomy causes only a small increase in FAST expression, we considered it to be important to discuss its function, because its molecular actions seem to relate with a well-known apoptotic pathway activated through Fas. A serine/threonine kinase, FAST is activated during Fas-mediated apoptosis. It can be rapidly dephosphorylated, and it concomitantly activates TIA-1, a nuclear RNA-binding protein regulating alternative pre-mRNA splicing of the human apoptotic gene Fas [19, 61]. The Fas-L cell surface molecule binds to its receptor Fas, mediating receptor-triggered apoptosis. Some evidence points toward an involvement of these two proteins in the endocrine regulation of uterine functions. Estrogen influence in the control of Fas/Fas-L expression is suggested by an analysis of the menstrual cycle in humans that clearly shows increased expression in the late proliferatory phase and during the menstrual period in comparison to the early secretory phase [62]. We therefore investigated the expression of Fas and Fas-L following ovariectomy and estrogen treatment. Similar to the expression of the FAST gene, the differential protein expression of Fas and Fas-L in rat uterus coincides with the endometrial apoptosis occurring at the same time point and at the same cellular localization. Both Fas-L and Fas were also down-regulated by estrogen, further supporting their involvement in estrogen uterotrophic effects. It is important to note that Fas-deficient lpr and lprcg (a mutant lpr gene) mice do not exhibit ovariectomy-induced regression of vaginal epithelia, whereas uterine regression induced by ovariectomy remained intact [63]. It is possible that Fas/Fas-L mediate some estrogenic actions in intact animals and that some compensatory mechanism develops in lpr mice to mediate the cell death response. Another interpretation supported by our data is that several mechanisms coexist that may contribute to the apoptotic process. Speculatively, this redundancy may be a way to secure successful completion of reproduction despite possible inactivation of individual pathways.
The present study describes novel hormone-regulated genes that are oppositely affected by ovariectomy and estradiol treatment. These genes may constitute novel regulators of apoptosis and proliferation in the uterus and point toward the involvement of several pathways in these actions. The present study has also demonstrated that microarray technology is a powerful tool when investigating the reproductive system.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Amilcar Flores-Morales, Department of Molecular Medicine, Karolinska Hospital Building L8:01, 171 76 Stockholm, Sweden. FAX: 468 51776180; amilcar.flores{at}molmed.ki.se
Received: 19 January 2003.
First decision: 3 February 2003.
Accepted: 14 May 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and -ß. J Biol Chem 2000 275:28834-28842-amidating monooxygenase in the rat uterus: estrogen regulation. Proc Natl Acad Sci U S A 1998 95:7191-7196-ethynyl estradiol, bisphenol A, and genistein in the developing female reproductive system of the rat. Toxicol Sci 2002 68:184-199This article has been cited by other articles:
|
|
 |
|
  P.G. Groothuis, H.H.N.M. Dassen, A. Romano, and C. Punyadeera Estrogen and the endometrium: lessons learned from gene expression profiling in rodents and human Hum. Reprod. Update, July 1, 2007; 13(4): 405 - 417. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. O'Lone, K. Knorr, I. Z. Jaffe, M. E. Schaffer, P. G. V. Martini, R. H. Karas, J. Bienkowska, M. E. Mendelsohn, and U. Hansen Estrogen Receptors {alpha} and {beta} Mediate Distinct Pathways of Vascular Gene Expression, Including Genes Involved in Mitochondrial Electron Transport and Generation of Reactive Oxygen Species Mol. Endocrinol., June 1, 2007; 21(6): 1281 - 1296. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Naciff, G. J. Overmann, S. M. Torontali, G. J. Carr, Z. S. Khambatta, J. P. Tiesman, B. D. Richardson, and G. P. Daston Uterine Temporal Response to Acute Exposure to 17{alpha}-Ethinyl Estradiol in the Immature Rat Toxicol. Sci., June 1, 2007; 97(2): 467 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.A. Horcajadas, A. Pellicer, and C. Simon Wide genomic analysis of human endometrial receptivity: new times, new opportunities Hum. Reprod. Update, January 1, 2007; 13(1): 77 - 86. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Kwekel, L. D. Burgoon, J. W. Burt, J. R. Harkema, and T. R. Zacharewski A cross-species analysis of the rodent uterotrophic program: elucidation of conserved responses and targets of estrogen signaling Physiol Genomics, November 17, 2005; 23(3): 327 - 342. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A White and L. A Salamonsen A guide to issues in microarray analysis: application to endometrial biology Reproduction, July 1, 2005; 130(1): 1 - 13. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Rojek, J. Nielsen, H. L. Brooks, H. Gong, Y.-H. Kim, T.-H. Kwon, J. Frokiaer, and S. Nielsen Altered expression of selected genes in kidney of rats with lithium-induced NDI Am J Physiol Renal Physiol, June 1, 2005; 288(6): F1276 - F1289. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Dogan, D. A. Deshpande, M. S. Kannan, and T. F. Walseth Changes in CD38 Expression and ADP-Ribosyl Cyclase Activity in Rat Myometrium During Pregnancy: Influence of Sex Steroid Hormones Biol Reprod, July 1, 2004; 71(1): 97 - 103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Stahlberg, E. Rico-Bautista, R. M. Fisher, X. Wu, L. Cheung, A. Flores-Morales, G. Tybring, G. Norstedt, and P. Tollet-Egnell Female-Predominant Expression of Fatty Acid Translocase/CD36 in Rat and Human Liver Endocrinology, April 1, 2004; 145(4): 1972 - 1979. [Abstract] [Full Text] [PDF]
|
|
| |||||||
