| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mechanisms of Hormone Action |
a Department of Physiology, University of Maryland School of Medicine, Baltimore, Maryland 21201
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ERß. The former is a well-characterized mediator of estrogen-induced transcription, but the function of the latter is unclear. Recent in vitro studies suggest that both splicing forms of ERß expressed in rat tissues, ß1 and ß2, may function as inhibitors of ER transcriptional activity. To gain insight into the role of ERß in estrogen action, we examined the effects of estrogen and relaxin, a ligand-independent activator of ERs, on the expression of ERß1 and ERß2 mRNA in the uterus in vivo. Eighteen-day-old female rats were ovariectomized and, after recovery, treated with 17ß-estradiol, relaxin, or vehicle. Quantitative reverse transcription-polymerase chain reaction analyses of uterine RNA from estrogen-treated animals revealed marked decreases in the steady-state levels of the mRNAs for both ERß1 and ERß2 at 3, 6, and 24 h after treatment. Relaxin induced a similar effect. Neither hormone had any significant effect on ER mRNA levels. To determine if endogenous estrogen exerts this effect, we examined the expression of ERßs in the uterus during the estrous cycle. Levels of both isoforms were highest at diestrus (low estrogen), were significantly lower at early proestrus (rising estrogen), reached a nadir during late proestrus (peak estrogen), and rebounded at estrus (declining estrogen). These data suggest that down-regulation of ERß expression may be required for estrogen to exert its full trophic effects on the uterus.
estradiol, estradiol receptor, gene regulation, relaxin, uterus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
), was discovered [5–7]. A further layer of complexity was added when a second common splicing variant of ERß was identified in tissues of both the rat and mouse [8–12]. Designated ERß2 (with the original form becoming ERß1), it contains a 54-base pair (bp)/18-aa insert within the ligand-binding domain.
The physiological function of ERß is presently unclear, but its discovery has necessitated a complete reevaluation of estrogen signaling pathways in target cells. The cyclic growth and differentiation of the uterine endometrium and other tissues of the reproductive tract are controlled in large part by estrogen [2, 4, 13]. While the predominant ER in the uterus is ER, numerous investigators have reported that ERß is also expressed there, albeit at a markedly lower level [14–20]. ERß2, the longer splice variant present in rodents, has been reported to coexist with ER and ERß1 in all tissues examined, including the uterus, prostate, and ovary [10], and the level of its expression is similar to, in some cases greater than, that of ERß1 [8, 9]. ERß protein has also been detected in the rodent uterus by immunohistochemistry [19, 20]. Thus, a substantial body of evidence indicates that both ERß mRNA and protein are expressed in the uterus.
With regard to function, it has been repeatedly demonstrated in transfection studies that ERß1 is capable of mediating estrogen-induced reporter gene transcription in vitro, but it is significantly less active than ER [7, 21]. Furthermore, it remains to be definitively shown that endogenous ERß1 acts as a transcriptional activator of endogenous genes at physiological estrogen concentrations in vivo. ERß2 binds estrogen with still lower affinity than ERß1 and is a much weaker inducer of transcriptional activity [10–12]. On the other hand, there is mounting evidence, also from in vitro transfection studies, that both isoforms of ERß could actually serve to inhibit ER signaling in target tissues [10, 12, 21]. ERß2 was first shown to inhibit estrogen-dependent, ER-mediated expression of estrogen response element (ERE)-containing reporter genes in COS-1 cells [10]. Subsequently, Hall and McDonnell [21] demonstrated that human ERß (homologous to rat/mouse ERß1) also repressed ER activity in HepG2 cells in response to subsaturating concentrations of 17ß-estradiol, which is physiologically relevant because estrogen's uterotrophic effects require only 5–20% receptor occupancy [13]. The inhibitory effect of ERßs could involve formation of heterodimers with ER [11, 12, 21–23], an interaction that is enhanced by estrogen [11, 21]. In addition, ERß1, unlike ER, appears to interact with the ERE in the absence of ligand [21], suggesting that it may also competitively inhibit ER binding to DNA. Such interactions would, of course, require that ER and ERß be coexpressed in cells. Several studies indicate that both are expressed in uterine epithelial and stromal cells [16, 17, 19, 20, 24, 25]. That ERßs normally oppose estrogen's uterotrophic actions in vivo can also be inferred from the recent observation that ERß knock-out (ßERKO) mice show exaggerated uterine responsiveness to estrogen, such as greater edema and increased rates of cell proliferation [24]. This hypersensitivity may be detrimental to uterine function because ßERKO females that become pregnant are reported to exhibit an abnormally high incidence of fetal death [7, 24].
Resolving the question of ERß's physiological roles will depend in part on defining where and when the various ER subtypes are expressed relative to each other and how their expression is regulated in vivo. There is relatively little known about the regulation of ERß expression in target tissues at this time. In the ovary, ERß is down-regulated by gonadotropin treatment [26], while estrogen has been reported to reduce ERß expression in some areas of the brain and in the pituitary [27–29]. In contrast with the latter findings, it was recently reported that estrogen has no effect on ERß expression in the uterus at 24 h [30]. In the present study, we examined the effects of estrogen and relaxin, a peptide hormone that activates ERs independently of ligand [31], on the expression of both ERß1 and ERß2 at earlier time points. Furthermore, we determined their pattern of expression during the normal estrous cycle.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Animal studies were conducted in accordance with mandated standards of humane care as described in the Guide for the Care and Use of Laboratory Animals (National Research Council) and approved by the Institutional Animal Care and Use Committee of the University of Maryland School of Medicine.
To examine the effect of estrogen on ERß1, ERß2, ER, and VEGF (vascular endothelial growth factor) expression in the uterus, female Sprague-Dawley rats (Crl:CD[SD] BR strain; Charles River, Wilmington, MA) were ovariectomized at 18 days of age and allowed to recover for 7–10 days. For each experiment, animals were randomly divided into treatment groups. Treatments were administered as single i.p. injections. For the majority of the experiments, rats were treated with 17ß-estradiol (Sigma Chemical Co., St. Louis, MO), 50 ng/g body weight (BW) in 200-µl vehicle (ethanol:PBS, 1:500), or vehicle alone. At 3 and 6 h or 6 and 24 h following injection, animals from each treatment group were killed by cervical dislocation and uteri were removed for RNA extraction (below). In an additional experiment, the effects of different doses of 17ß-estradiol (0, 0.5, 5, and 50 ng/g BW) were compared at 6 h.
To compare the effect of relaxin, a ligand-independent activator of ERs, with that of estrogen, animals were treated with either 17ß-estradiol (50 ng/g BW), purified porcine relaxin (0.85 µg/g BW in 200 µl PBS; [32]), or vehicle alone. Animals were killed 6 h later and their uteri collected.
To examine the regulation of ERß1 and ERß2 expression in the ovary by gonadotropins, intact, 27-day-old female Sprague-Dawley rats were separated into four groups (three animals per group). Group 1 (controls) received a single injection (i.p.) of PBS vehicle. Groups 2–4 received single injections of eCG (8 IU/200 µl PBS/animal; National Hormone and Pituitary Program, NIDDK). Forty-eight hours following eCG treatment, animals in groups 3 and 4 received a single i.p. injection of hCG (5 IU/200 µl PBS/animal; Sigma), while animals in group 2 were killed by cervical dislocation and the ovaries removed. Six hours following hCG injection (preovulation), animals in group 3 were killed, followed at 20 h (postovulation) by animals in groups 1 and 4.
To measure uterine ERß mRNAs during the normal estrous cycle, intact 8- to 10-wk-old female Sprague-Dawley rats were used. To monitor cycles, vaginal cytology was examined daily. Only animals exhibiting two consecutive normal estrous cycles were used. Animals were killed and uteri collected either between 1000 and 1100 (early proestrus, estrus, or diestrus) or at 1700 (late proestrus) h.
RNA Extraction
In all experiments, uteri were dissected out immediately after rats were killed, trimmed to remove extraneous tissue, flash frozen in liquid nitrogen, and stored at -80°C for subsequent RNA extraction. Total RNA was extracted using the RNeasy RNA extraction kit (Qiagen, Inc., Valencia, CA). Tissues were first homogenized in buffer RLT using a bead mill (Mini Beadbeater; Biospec Products, Bartlesville, OK) and 1 mm zirconium silicate beads (30 sec at 5000 rpm). Homogenates were then sheared three times by passage through a 23-gauge needle. All subsequent purification steps were carried out according to the manufacturer's instructions. The concentration and purity of the RNA in each sample were determined by measurement of absorbance at 260 and 280 nm. The concentration of total RNA was then normalized for all samples to 1 µg/6 µl of water for reverse transcription (RT).
Reverse Transcription-Polymerase Chain Reaction
Relative levels of specific mRNAs/µg of total RNA were compared among treatment groups using quantitative reverse transcription-polymerase chain reaction (RT-PCR). This method, with modifications (described below), has been used previously for the comparison of relative levels of VEGF mRNA in the uterus after estrogen treatment [33] and in the ovary after gonadotropin stimulation [34]. This assay has been demonstrated to accurately measure the relative difference in specific mRNA concentration in standard solutions prepared by dilution [33, 34] and yields results similar to those obtained by Northern analysis [35, 36]. In the present study, relative levels of target mRNA PCR product in samples from each treatment group, amplified beginning with equal amounts of total RNA and under identical conditions (after determining the number of cycles for optimal exponential amplification for each target mRNA), were compared based on optical density of product bands rather than by comparing the extinction point of serially diluted samples. The accuracy of the densitometric analysis was validated using serial dilutions of a standard cDNA (i.e., a twofold difference in the amount of cDNA loaded on a gel yielded a twofold difference in optical density). To confirm that total RNA measurements were accurate and that approximately equal amounts of total RNA had been reverse transcribed, within each experiment, 18S ribosomal (r) RNA levels in samples from representative experiments were compared. 18S rRNA has been shown to be the least variable of several commonly used housekeeping genes [37]. There were no significant differences in the yield of 18S rRNA RT-PCR product from equal amounts of uterine total RNA among experimental groups, either when comparing uteri from control and estrogen-treated immature ovx females (P > 0.2 to 0.7) or from adult cycling females (P > 0.3 to 0.9). Normalization of target transcript levels to 18S rRNA levels had no effect on the results and, therefore, the results normalized to total RNA alone are shown. Further validation of this approach was the demonstration that the effects of estrogen on uterine VEGF mRNA levels (see Fig. 4) yielded results similar to those reported by us previously [33] as well as by others using Northern analysis [35].
|
Total RNA (1 µg/6 µl of water) was reverse-transcribed [33, 34] by incubation at 37°C for 60 min with the following reagents (from Life Technologies, Gaithersburg, MD, unless otherwise indicated): 1) 200 U (1 µl) cloned Moloney murine leukemia virus reverse transcriptase; 2) 5x reaction buffer (250 mM Tris-HCl [pH 8.3], 375 mM KCl, and 15 mM MgCl2), 4 µl; 3) bovine serum albumin (Boehringer Mannheim, Indianapolis, IN), 0.1% (wt/vol), 2 µl; 4) 100 mM dithiothreitol, 2 µl; 5) dNTPs (2.5 mM each of dATP, dCTP, dGTP, and dTTP), 4 µl; and 6) random primer oligodeoxyribonucleotides (500 ng/µl), 1 µl. Reverse transcribed samples were then placed on ice or frozen until used for PCR. For each RT run, a reagent blank was prepared in which an equal volume of water was substituted for RNA sample.
Oligonucleotide primers for PCR were synthesized by Life Technologies. Primers for each product were as follows, with the location of each primer in the target transcript (based on GenBank sequences) indicated in parentheses:
Primers were designed based on the sequences of the corresponding rat cDNAs (GenBank accession numbers: Y00102 for ER, AF042059 for ERß, AF215725 for VEGF, and X01117 K01593 for 18S rRNA). The ER, ERß1/ß2, and VEGF primer pairs all span at least one intron. For the other pairs, the possibility of amplification from genomic DNA was eliminated by pretreatment of the RNA with DNase (RNeasy RNA extraction kit, Qiagen). The ERß primer pair was located upstream of the alternative splice site and therefore did not differentiate between ERß1 and ERß2 (Fig. 1A); they yielded a single PCR product of 251 bp. The ERß1/ß2 primers [10] span the splicing site and yield two PCR products—one for each splicing form (246 and 300 bp for ERß1 and ERß2, respectively). The ERß2 primers, one of which was located within the 54-bp insert, recognize only that splicing form (yielding a 174-bp product).
|
PCR was performed as described previously [33, 34]. PCR incubations were carried out in a programmable thermal cycler (MJ Research, Cambridge, MA). A reagent blank prepared using the RT blank was included in all PCR runs. Following PCR, 8-µl aliquots of each sample were mixed with 2 µl of 5x loading buffer (30% Ficoll, 0.2 M EDTA, 0.25% bromphenol blue, and 0.25% xylene cyanole FF). These were fractionated by electrophoresis on 10 x 10-cm, 0.8-mm-thick, nondenaturing, 8% polyacrylamide gels at a constant 70 V. Double-stranded DNA standards were also run on each gel (50-bp DNA ladder; Life Technologies). Following fractionation, gels were stained for 30 min with ethidium bromide (0.5 µg/ml in distilled water; Sigma), destained for 15 min in distilled water, and examined on a 312-nm ultraviolet transilluminator. Gels were photographed as described previously [33].
Each cycle of PCR consisted of 1 min at 94°C, 0.5 min at 60°C, 1.5 min at 72°C, and a final extension of 5 min at 72°C. The optimal number of PCR cycles (the number of cycles yielding a readily detectable product but still within the linear range of amplification) was first determined for each target mRNA. The numbers of cycles were 20–22 for ER, 28–30 for ERß1/ß2, and 26–28 for VEGF.
Restriction Enzyme Analysis
Restriction enzyme analysis was used to verify the identity of PCR products. Restriction sites present in the amplified regions were mapped using the Baylor College of Medicine Search Launcher Sequence Utility (mbcr.bcm.tmc.edu/services.html). Digestions were carried out per manufacturer's instructions.
Densitometry and Statistical Analysis
The densities of gel bands for RT-PCR products were quantified by analyzing the film negatives using a Polaroid Digital Microscope Camera fitted with a Nikon AF Micro Nikkor 60-mm lens and Quantiscan for Windows (version 2.1; Biosoft, Inc., Ferguson, MO). Relative optical densities are expressed in arbitrary units. The data were analyzed using one-way ANOVA followed by the Fisher protected least significant difference test (StatView for Windows, Version 4.57; Abacus Concepts, Inc., Berkeley, CA). PCR product bands shown in the figures are taken from the photographic negatives and therefore are reverse images.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ERß mRNAs in the Rat Uterus by RT-PCR
An abundant ER PCR product of the expected size (419 bp) was generated after just 20–22 cycles from reverse transcribed uterine mRNA from immature ovx rats (Fig. 1B). The identity of this product was confirmed by restriction enzyme analysis (data not shown). Although there have been a few reports that ERß is not expressed in the rodent uterus, numerous other studies have shown that ERß transcripts are present in this organ, although at lower levels than ER mRNA [8, 10, 11, 14–16, 24, 25]. Consistent with this, ERß primers located upstream of the alternative splice site in the ERß sequence (primers 5'P1 and 3'P1; Fig. 1A) generated a single PCR product of the expected 251-bp size from uterine RNA after 28–30 PCR cycles (Fig. 1B). When primers that span the alternative splice site in the ERß gene (5'P2 and 3'P2; Fig. 1A) were used, two product bands were generated (Fig. 1B). Their sizes closely matched the 246- and 300-bp products expected for ERß1 and ERß2, respectively, with the difference corresponding to the 54-bp insertion in the ligand-binding domain of ERß2 [8–11]. Restriction enzyme analysis confirmed the identity of both the common ERß product and the products for the ß1 and ß2 splicing forms (data not shown). In agreement with previous reports that levels of ERß1 and ERß2 transcripts are often roughly equivalent [8–11], the yield of the two ERß splicing products was similar, with that for ERß1 being slightly more abundant. To further confirm that the larger product was that for ERß2, an antisense primer located within the 54-bp insert unique to that transcript was made (3'P3; Fig. 1A). Its use with primer 5'P2 resulted in a single PCR product of the expected size (174 bp; Fig. 1B); its identity was also confirmed by restriction enzyme analysis (data not shown). Thus, although ER mRNA was more abundant in the rat uterus, transcripts for both splicing forms of ERß were also readily detectable.
Effect of Estrogen on ER and ERß Expression> in the Rat Uterus
It was recently reported that estrogen treatment of immature ovx rats has no effect on ERß mRNA levels in the uterus [30]. However, that study did not look at expression until 24 h after estrogen treatment, although estrogen exerts direct effects on the expression of genes in the uterus within minutes to a few hours [33, 36]. In the present study, we examined the effects of estrogen treatment on ERß mRNA levels at much earlier time points. As shown in Figure 2A, 17ß-estradiol (50 ng/g BW) caused a significant reduction in the levels of both ERß1 and ERß2 PCR products generated from uterine RNA at both 3 and 6 h (by approximately 42% and 82% at 6 h, respectively). Thus, there is a rapid reduction in steady-state levels of ERß1 and ERß2 mRNAs in the uterus following estrogen treatment.
|
To determine if ERß levels rebound after this initial down-regulation, which might explain why Wang et al. [30] saw no down-regulation at 24 h, an experiment was carried out in which uteri were collected at 6 and 24 h post-estrogen treatment. As shown in Figure 2C, however, both, ERß1 and ERß2 levels were strongly depressed at both time points; the effect on ERß1 levels was even greater than in the first experiment, nearly matching the decrease in ERß2 levels.
In contrast with the effect of estrogen on ERß1 and ERß2, ER mRNA levels were unaffected by estrogen at 3, 6, or 24 h (Fig. 2, B and D).
In a third independent experiment, a 10-fold lower dose of 17ß-estradiol, 5 ng/g BW, was nearly as effective as 50 ng in reducing the yield of ERß1 and ERß2 RT-PCR products (-62% and -72% versus -71% and -80%, respectively; n = 3 per group; P > 0.1) compared with controls (n = 4; P < 0.05 for either dose) at 6 h. Again, there was no significant change in the level of ER mRNA in response to either dose of 17ß-estradiol at this time point.
Effect of Estrogen on VEGF Expression in the Rat Uterus
We have previously reported that estrogen rapidly up-regulates VEGF expression in the rat uterus [33]. The yield of PCR products for VEGF164 and VEGF120, the two predominant alternatively spliced VEGF transcripts in the uterus [33], from uterine RNA samples were strongly stimulated by estrogen at the same time points at which levels of ERß mRNAs were depressed (Fig. 3). Thus, the inhibitory effect of estrogen on ERß mRNA levels was specific and occurred at the same time that the levels of other estrogen-induced mRNAs were increasing.
|
Effect of Relaxin on ER and ERß Expression> in the Rat Uterus
The peptide hormone relaxin, which exerts acute trophic effects on the uterus similar to those induced by estrogen, does so in part through ligand-independent activation of ERs [31]. We asked, therefore, whether relaxin would also down-regulate uterine ERß levels. As shown in Figure 4, relaxin, like estrogen, caused a marked decrease in the steady-state levels of ERß1 and ERß2 mRNA at 6 h. The effect of relaxin was somewhat less than that of estrogen in the same experiment. This is consistent with our previous observation that the early uterotrophic effects of relaxin are somewhat weaker than those induced by estrogen [31].
Effect of hCG on ERß Expression in the Rat Ovary
It has been reported that LH or hCG rapidly down-regulate ERß expression in the rat ovary [26]. To determine if we could detect the same effect and to compare it with the effect of estrogen on uterine ERß levels, we examined ovaries from eCG-primed, immature animals after hCG treatment. As shown in Figure 5, hCG caused a significant decrease in ovarian ERß levels (to 60% of control) at 20 h. The magnitude of this effect was similar to that reported by Byers et al. [26], although it occurred later than in that study, which saw a marked decrease at 6 h after hCG.
|
Expression of ERß mRNA in the Rat Uterus During the Estrous Cycle
During the rat estrous cycle, serum 17ß-estradiol levels rise slowly during diestrus, increase rapidly to a peak during proestrus, and decline sharply to a nadir during estrus [38]. Based on the studies in immature animals, we hypothesized that ERß levels would be lowest during proestrus. Therefore, we examined the pattern of expression of ERßs in the uteri of adult rats at each stage of the estrous cycle. As in the immature ovx animals, both isoforms of ERß were readily detectable in the adult uterus at all stages of the cycle (Fig. 6, top). Levels of both isoforms were highest at diestrus, significantly lower at early proestrus, lowest at late proestrus (ERß1 22% and ERß2 24% of diestrus values; P < 0.05), and increased again at estrus. Thus, the increase in endogenous 17ß-estradiol at proestrus rapidly inhibits the expression of both ERß1 and ERß2 in the uterus, similar to the effect of exogenous 17ß-estradiol in immature ovx rats. Furthermore, these results indicate that ERß expression rebounds rapidly after estrogen levels decline during the estrous cycle; this differs from the situation observed in immature ovx mice treated with a high level of exogenous 17ß-estradiol.
|
ER mRNA levels during the cycle were less variable than those for ERß. Levels were highest during diestrus and early proestrus, fell by just over 40% during late proestrus (P < 0.05 compared with early proestrus), and increased again to near maximal levels during estrus (Fig. 6, bottom).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ERß are differentially regulated by estrogen in the rat uterus, further evidence that these two molecules play different roles there. Treatment of immature ovx rats with 17ß-estradiol rapidly down-regulated both ERß1 and ERß2 transcripts while leaving that for ER unchanged. This effect was also induced by relaxin, a ligand-independent activator of ERs [31], indicating that down-regulation does not require ligand per se but only activation of ERs (presumably ER). The physiological relevance of this rapid down-regulation is indicated by the fact that it also occurs during the normal estrous cycle: ERß mRNA levels decrease between early and late proestrus, when estrogen levels peak [38]. In the immature ovx rat, the inhibitory effect of estrogen was sustained for at least 24 h. During the cycle, however, ERß expression rapidly rebounded after proestrus. This difference most likely is due to a more sustained period of elevated estrogen in the immature ovx model. The decline in whole uterus probably represents the net result of greater and lesser changes in different uterine cell populations. This is supported by the observation that a decline in uterine ERß after 72 h of estrogen treatment occurred primarily in stromal cells [24]. This spatially links lower ERß expression and estrogen action because estrogen-induced epithelial cell proliferation is mediated by ERs in stromal cells, not those in the epithelial cells themselves [39]. The rapid decline in ERß mRNAs could result from decreased transcription, increased degradation, or both.
A possible implication of these results is that a reduction in ERß levels may be necessary for estrogen and other ER activators to exert their full trophic effect on the uterus. This assumes that a decrease in ERß mRNA is followed by a decline in protein. While quantitating low levels of ERß protein remains problematic, a limited number of studies support such a relationship [24, 40]. This pattern fits well with the mounting evidence that ERßs may negatively modulate ER transcriptional activity [10, 21, 41]. The rapid decrease in ERß mRNA levels precedes estrogen's late uterotrophic effects, such as increased DNA synthesis and cell proliferation, which occur only after 7–8 h of tight binding of hormone-ER complexes to nuclear sites [13, 42], a period during which ERß protein levels might likewise decline. Why this extended period of receptor association is necessary has never been explained. Our results suggest a possible explanation: the need to first down-regulate levels of ERß. Variation in the degree to which different genes are inhibited by ERß and the regulation of ERß levels by estrogen may be key mechanisms through which estrogen is able to progressively induce the expression of different genes over an extended period of time. Early response genes, such as VEGF [33], may be less subject to this inhibitory effect than those that are required later for DNA synthesis and endometrial hyperplasia.
The lower level of ERß expression in the uterus relative to that in the ovary or to that of ER in the uterus has led some to conclude that it could not play a significant role there. As a receptor as well as a transcription factor, ER levels in target cells must be high to enable the efficient detection of changes in estrogen concentration. If ERß plays different roles than ER, including some not dependent on ligand binding [21], the level of expression of the two genes or the concentration of the two proteins need not be similar. The relevant comparison may be between levels of ERß and liganded ER, a small fraction of total ER at physiological estrogen concentrations [13].
Based on our observations and numerous published studies, quantitative RT-PCR is the only method with sufficient sensitivity to reliably detect ERß expression in the rodent uterus and to quantitatively show a further reduction in levels following estrogen treatment. We chose to normalize expression to total RNA rather than a housekeeping gene because numerous studies have demonstrated that few, if any, such genes remain constant in cells or tissues undergoing rapid growth or remodeling in response to stimuli. One recent study found that even the least variable of these, 18S ribosomal RNA, still required normalization to total RNA before it could be used as a standard for normalization of other genes [37]. We confirmed this by measuring 18S rRNA in representative samples and found that the levels of this housekeeping gene did not differ significantly between experimental groups. Our previous studies and validation steps led us to a similar conclusion: RT-PCR product yield from equal amounts of total RNA under identical conditions and cycle number in the linear range of amplification is an accurate reflection of the relative amounts of target mRNA in the starting samples. This approach does not control for intraassay variation due to such things as pipeting error or slight differences in the efficiency of the reverse transcription or polymerization reactions. Because internal housekeeping genes are amplified in separate reactions and so are subject to the same random variables, normalizing to them also does not correct for such variability. Instead, we show the variability between replicates (the SEM), the major component of which is likely normal interanimal variation. As can be seen, the variation was modest and did not mask the differences resulting from estrogen treatment, which were highly reproducible and statistically significant. Finally, it should be remembered that, in the same samples in which ERß levels sharply declined, VEGF mRNA levels increased severalfold and ER mRNA levels were constant.
In contrast with our results, Wang et al. [30] reported that ERß mRNA levels in uteri of ovx adult rats were unaffected by 17ß-estradiol 24 h after treatment nor did they observe any significant change in levels during the estrous cycle [43]. Ovariectomy itself is reported to cause a marked decrease in ERß expression in the adult rat uterus [16], however, so a further decline in response to estrogen might not be observed. In agreement with our findings, Weihua et al. [24] found that daily injection of 17ß-estradiol for 3 days decreased both ERß mRNA and protein levels in the immature, intact mouse uterus; the magnitude of the change, however, was not quantitated. ER mRNA levels appeared to be somewhat higher than controls at the same time point. The authors speculated that the activity of ER may be suppressed by ERß during the prepubertal period but that estrogen secretion at puberty inhibits ERß expression, thereby making ER fully functional. We propose that this effect of ERß does not end at puberty but becomes cyclic with the level of ERß and the degree of suppression, varying inversely with estrogen levels (although some estrogen is apparently also required to maintain ERß expression [16]).
Estrogen down-regulation of ERß mRNA has also been observed in the brain, although these studies involved longer periods (1–2 wk) of treatment [27, 28]. ERß mRNA levels also decline in the pituitaries of a) ovx rats after treatment with estrogen for 3 days [29], b) intact female rats at puberty [44], and c) intact adult female rats during proestrus [29]. ERß expression also decreases in GnRH neurons in the hypothalamus at proestrus [45]. Most recently, estrogen (for 8 days) was shown to strongly inhibit ERß mRNA levels in mouse mesenchymal stem cells while at the same time up-regulating ER expression [46].
Similar to the effects of estrogen in the uterus, gonadotropin treatment rapidly down-regulates ERß mRNA [26] and protein [40] in the ovary. This might be due to rising estrogen levels. Consistent with this, treatment of immature rats with diethylstilbestrol for either 24 or 96 h also lowered granulosa cell ERß mRNA levels [47] and ERß mRNA in sheep granulosa cells varied inversely with estradiol production [48]. Studies of the ßERKO mouse indicate that ERß is essential for normal follicular development and function [4]. We propose that this may be because it protects ovarian cells from the effects of estrogen levels orders of magnitude higher than those to which peripheral tissues are exposed. Interestingly, ovarian ER transcripts did not change during the estrous cycle [26], similar to what we observe in the uterus.
In contrast with its effects on ERß, estrogen had little effect on ER mRNA levels in uteri of immature ovx rats. Only one other similar study has been done using this model [49]. As here, there was no difference in ER levels at 6 h, but there was a significant increase at 24 h, which we did not observe. The dose of estrogen, the route of administration, and the method of mRNA quantitation, however, differed between the two studies. Studies in adult ovx rats are complicated by the fact that ovariectomy itself causes an increase in uterine ER mRNA levels [16, 50, 51]. In one such study, estrogen further stimulated ER mRNA expression [50] but in another strongly repressed it [51]. Such differences may result from the fact that estrogen also stimulates the expression of many commonly used housekeeping genes, such as ß-actin, in the uterus [50, 52] and normalizing to such transcripts distorts results. Previous studies of ER expression in the uterus of the rat during the estrous cycle are also difficult to interpret. One group has reported highest levels at proestrus and a significantly lower level at metestrus [52, 53], but the magnitude of difference between the high and low points was substantially different in those two studies. They also did not examine late proestrus, and we did not look during metestrus.
In summary, these studies show that both estrogen and relaxin rapidly down-regulate ERß expression. We propose that this may be a requisite event for estrogen and other ER activators to exert their normal uterotrophic effects because ERßs may function in part as inhibitors of ER-mediated transcriptional activity. Because overstimulation of cell proliferation by estrogens is believed to contribute to the development of cancer in target tissues, inhibition of ERß expression by estrogen may contribute to that process [54]. Several recent studies have found that expression of ERß is lower in cancers compared with healthy tissues [55–57].
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Correspondence: Robert D. Koos, Department of Physiology, University of Maryland School of Medicine, 655 West Baltimore Street, Baltimore, MD 21201-1559. FAX: 410 706 8341; rkoos{at}umaryland.edu
3 S.B.P. and J.M.J. contributed equally and should be considered co-first authors
Received: 15 January 2002.
First decision: 4 February 2002.
Accepted: 1 July 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and ß in the female reproductive tract of the rat during the estrous cycle. Biol Reprod 2000 63:1331-1340 and ß in the female reproductive tract of the rat during the estrous cycle. Biol Reprod 2000 63:1331-1340This article has been cited by other articles:
|
|
 |
|
  A. M Davis, J. Mao, B. Naz, J. A Kohl, and C. S Rosenfeld Comparative effects of estradiol, methyl-piperidino-pyrazole, raloxifene, and ICI 182 780 on gene expression in the murine uterus J. Mol. Endocrinol., October 1, 2008; 41(4): 205 - 217. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Yan, J. Chen, A. A Wiley, B. D Crean-Harris, F. F Bartol, and C. A Bagnell Relaxin (RLX) and estrogen affect estrogen receptor {alpha}, vascular endothelial growth factor, and RLX receptor expression in the neonatal porcine uterus and cervix Reproduction, May 1, 2008; 135(5): 705 - 712. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Shao, B. Weijdegard, J. Fernandez-Rodriguez, E. Egecioglu, C. Zhu, N. Andersson, A. Thurin-Kjellberg, C. Bergh, and H. Billig Ciliated epithelial-specific and regional-specific expression and regulation of the estrogen receptor-beta2 in the fallopian tubes of immature rats: a possible mechanism for estrogen-mediated transport process in vivo Am J Physiol Endocrinol Metab, July 1, 2007; 293(1): E147 - E158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Durrer, K. Maerkel, M. Schlumpf, and W. Lichtensteiger Estrogen Target Gene Regulation and Coactivator Expression in Rat Uterus after Developmental Exposure to the Ultraviolet Filter 4-Methylbenzylidene Camphor Endocrinology, May 1, 2005; 146(5): 2130 - 2139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. D. Sherwood Relaxin's Physiological Roles and Other Diverse Actions Endocr. Rev., April 1, 2004; 25(2): 205 - 234. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Tena-Sempere, V.M. Navarro, A. Mayen, C. Bellido, and J.E. Sanchez-Criado Regulation of Estrogen Receptor (ER) Isoform Messenger RNA Expression by Different ER Ligands in Female Rat Pituitary Biol Reprod, March 1, 2004; 70(3): 671 - 678. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
P < 0.001 versus respective control value
