0013-7227/91/1295-2423S03.00/0 Endocrinology Copyright © 1991 by The Endocrine Society

Vol. 129, No. 5 Printed in U.S.A.

Diethylstilbestrol Stimulates Persistent Phosphatidylinositol Lipid Turnover by an Estrogen Receptor-Mediated Mechanism in Immature Mouse Uterus DIANE M. IGNAR-TROWBRIDGE*, ARLENE R. HUGHESf, JAMES W. PUTNEY JR., JOHN A. McLACHLAN, AND KENNETH S. KORACH Laboratory of Reproductive and Developmental Toxicology (D.M.I.-T., J.A.M., K.S.K.) and Laboratory of Cellular and Molecular Pharmacology (A.R.H., J. W.P.), National Institute of Environmental Health Sciences, National Institutes of Health, Research Triangle Park, North Carolina 27709

ABSTRACT. The effect of estrogen on phosphoinositide (PI) metabolism was evaluated in the immature mouse uterus, a tissue which undergoes estrogen-induced proliferation. Uteri isolated from untreated mice or from mice injected ip with diethylstilbestrol (DES) were incubated with [3H]myo-inositol and assessed for incorporation of label into PI lipids or inositol phosphate generation. DES administration elicited a rapid increase in [3H]oryo-inositol incorporation, which persisted until at least 18 h post treatment. This effect could not be duplicated by incubation of uteri with DES in vitro, although [3H]myoinositol incorporation in uteri removed from DES-treated mice remained elevated for 3 h of in vitro incubation. Stimulation of PI lipid metabolism by DES was blocked by ICI 164,384, a

specific estrogen receptor antagonist. The effect of DES on PI metabolism consisted of a time-dependent increase in the specific activity of both phosphatidylinositol-4-phosphate and phosphatidylinositol-4,5-bisphosphate and a significant increase of inositol (l,4,5)-trisphosphate mass by 12 h post treatment. These changes occur before the onset of estrogen-induced DNA synthesis. The results indicate that estrogens rapidly modulate PI lipid turnover through an estrogen receptor-mediated mechanism. Since the metabolic products of PI lipids are important for signal transduction and cellular proliferation, altered metabolism of these lipids may play an integral role in estrogeninduced mitogenesis. {Endocrinology 129: 2423-2430,1991)

E

STROGENS elicit a myriad of rapid biochemical changes in the uterus which are thought to ultimately culminate in mitotic division. Understanding of the mechanisms controlling estrogen-induced growth will require knowledge of the earliest biochemical events that occur when mitogenesis is initiated by estrogens. One early effect of estrogens in the uterus is stimulation of the metabolism of phospholipids within 1 h after a single injection of estradiol as assessed by labeling of lipid pools by radioactive precursors (1, 2). Elevated [3H]myo-inositol incorporation into phosphoinositides (PI) has been observed in uteri removed from sheep which had been treated with estradiol (3) and in MCF-7 breast cancer cells incubated with estradiol for 24 h (4). Grove and Received July 8,1991. Address all correspondence and requests for reprints to: Dr. John A McLachlan, M.D. 17-02, National Institute of Environmental Health Sciences, P.O. Box 12233, Research Triangle Park, North Carolina 27709. * Supported by a PRAT fellowship from the National Institute of General Medical Sciences. t Present address: Department of Clinical Neurosciences, Burroughs Wellcome Company, Research Triangle Park, North Carolina 27709.

Korach (5) have previously demonstrated that diethylstilbestrol (DES) elevates the amount of [3H]myo-inositol incorporation into uterine phosphatidylinositol lipids. They further observed that Z,Z-dienestrol, a weak estrogen receptor (ER) agonist incapable of initiating DNA synthesis, was able to mimic the effect of DES. This result raised the possibility of direct effects of estrogen on PI metabolism that are not mediated by the ER. Evidence which demonstrates that steroid hormones, including estrogens, may induce effects that are nongenomic, i.e. are not mediated by nuclear estrogen receptors (6, 7), gives credence to this hypothesis. The metabolism of PI has generated much interest over recent years because the products of their breakdown are thought to be mediators of signal transduction and, therefore, may play a role in cellular proliferation. Many mitogens, including peptide growth factors, act through plasma, membrane receptors to elicit phospholipase C-mediated metabolism of phosphatidylinositol4,5-bisphosphate (PI-4,5-P2), which results in concomitant generation of both inositol (l,4,5)-trisphosphate [(1,4,5)IP3)] and diacylglycerol. (1,4,5)IP3 is responsible

2423

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 July 2015. at 11:58 For personal use only. No other uses without permission. . All rights reserved.

DES ELEVATES UTERINE PI LIPID METABOLISM

2424

for intracellular calcium release (8), whereas diacylglycerol activates protein kinase C (9) which is thought to modulate signal transduction, gene expression, and cell growth (10). Disruption of phospholipase C (PLC)-mediated breakdown of PIP 2 inhibits mitogen-induced growth in cells in culture (11, 12) and injection of PLC isoforms into NIH 3T3 cells stimulates DNA synthesis (13). However, Escobedo and Williams (14) demonstrated that platelet-derived growth factor (PDGF) receptor mutants that could stimulate PI turnover were incapable of inducing DNA synthesis. Thus, PI metabolism appears to be necessary, although not totally sufficient, for induction of DNA synthesis. Modulation of the metabolism of these lipids by estrogens could be necessary for estrogen-induced mitogenesis. Although the activation of PI metabolism by mitogens is well established in cell culture systems, there is a paucity of data concerning mitogen-induced proliferation in the whole animal. The uterus of the prepubescent mouse provides an excellent model for the study of estrogen-induced growth in vivo. All three major cellular compartments—epithelial, stromal, and myometrial layers—proliferate in response to in vivo estrogen treatment (15). In the present study, we have used this model to extend the findings of Grove and Korach (5) with emphasis on the following questions: 1) Do estrogens modulate PI lipid metabolism through a direct or ER-mediated mechanism? 2) Are phosphoinositide metabolism and (1,4,5)IP3 levels affected by estrogen treatment? and 3) Can the in vivo effect of estrogen on PI lipid turnover be mimicked by in vitro incubations of uterine tissue with estrogen?

Materials and Methods Animals Immature CD-I mice (17 or 18 days old) were housed with lactating dams (10 pups per dam) in an animal room with a fixed lighting schedule (0800-2000 h). Mice used for studies of the effect of in vivo DES treatment weighed between 8.0-9.0 g and were randomized among dams before each experiment. Mice received a single ip injection of either DES or vehicle (0.1% ethanol in saline). In the antagonist study, either ICI 164,384 or dimethylsulfoxide vehicle were administered ip 30 min before DES or its vehicle. pHJMyo-inositol labeling Uteri were removed and horns were slit longitudinally to expose the luminal surface. Uteri were labeled with [3H]myoinositol (0.4 mCi/ml) in phenol red-free Dulbecco's modified Eagle's medium (DMEM)/F-12 containing 1% BSA and 2 mM CaCl2 for 1.5 h (unless specified otherwise) in a Queue (Parkersburg, WV) tissue culture incubator (95% room air/5% CO2). Viability studies were performed on uteri incubated for 1, 2, or 3 h in this manner (data not shown). The tissues appeared

Endo«1991 Voll29«No5

normal upon histological examination, and lactate dehydrogenase release did not increase substantially during the incubation. After labeling, the uteri were washed and either processed for evaluation of lipids or further incubated for determination of inositol polyphosphate production. Tissue extraction Inositol phosphates were extracted by a modification of the method of Sugiya et al. (16). Uteri were placed in separate tubes and were gassed with 95% O2/5% CO2 in 0.5 ml DMEM/F-12 with 10 mM lithium chloride added 5 min before agonist addition to block the degradation of inositol monophosphates. After incubation with vehicle (1% ethanol in DMEM/F-12) or agonist, uteri were homogenized with a polytron (Brinkmann, Westbury, NY) on setting 6 for 5 sec in 3% final perchloric acid/1 mg/ml phytic acid. The samples were left on ice for 20 min and were centrifuged at 1000 x g for 5 min. The supernatants containing the [3H] inositol phosphates were removed and neutralized with 0.5 M KOH/9 mM Na2B4O7. Lipids were extracted from the tissue pellet with chloroform:methanol:2.4 M HC1 (1.5:2:1). Two extractions were performed, and the combined layers were washed once with methanol:l M HC1 (1:1) and either were allowed to evaporate in vials for total lipid counts or were evaporated under nitrogen in acid-washed tubes when further lipid processing was performed. This procedure extracted at least 94% of PI lipids as determined by extraction of [3H]PIP and PIP 2 standards added to uterine tissue homogenates. Anion exchange chromatography Neutralized acid extracts of uterine tissue were applied to 0.5 ml AG-1-X8 (formate form, 200-400 mesh) resin and, after 10 water washes (10 ml each) to remove free [3H]m/yo-inositol, inositol monophosphate (IP), inositol bisphosphate (IP2), and inositol trisphosphate (IP3) were separated as described (17). [3H]-(1,4,5)IP3, (1,4)IP2, and (1)IP standards were run on separate columns. Data are expressed as a percentage of the radioactivity of total inositol phosphates per total [3H]myoinositol-labeled inositol lipids. Determination of Pi-specific radioactivity Phospholipids were deacylated as previously described (18) and the free fatty acids extracted by addition of 1 ml water and 1.2 ml iV-butanokpetroleum ether:ethyl formate (4:2:1). Deacylated lipids were washed with 100 n\ 60% perchloric acid for 45 min at 170 C and were diluted with 330 n\ deionized H2O. An aliquot was removed to determine total disintegrations per minute in each sample. Total lipid phosphate was measured by a colorimetric method using malachite green (19). Data are expressed as disintegrations per minute per nmol of inorganic phosphate. The specific activity of polyphosphoinositides was determined by separation from other lipids by neomycin column chromatography (20). [3H]PIP and PIP 2 standards were run on separate columns. Approximately 20% of the radioactivity associated with the PIP standard was detected in the PIP 2 fraction, and 10% of PIP 2 radioactivity was detected in the PIP fraction. The lipids were deacylated and processed as described

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 July 2015. at 11:58 For personal use only. No other uses without permission. . All rights reserved.

DES ELEVATES UTERINE PI LIPID METABOLISM above for measurement of radioactivity and total phosphate in PIP or PIP 2 . (1,4,5)IPS Mass assay

Uteri were removed from mice treated with DES for 3 or 12 h or vehicle, and the tissue was immediately homogenized in 3% perchloric acid without phytate for extraction of inositol phosphates as described above. Approximately 15 sec elapsed between time of killing and homogenization. The (1,4,5)IP3 mass was quantitated in duplicate samples of the supernatant using a [3H]-(1,4,5)IP3 radioreceptor kit obtained from Amersham (Arlington Heights, IL). PHJthymidine incorporation DES-treated mice were injected ip with 10 /*Ci [3H]thymidine and were killed 1 h later. The uteri were removed, and one horn was frozen on dry ice for quantification of total DNA and trichloroacetic acid (TCA)-precipitable radioactivity. The other horn was placed in Bouin's fixative, and the tissue was embedded in paraffin for histological sectioning. The number of epithelial cells along the basement membranes was quantitated in sections that had been stained with hematoxylin and eosin. Other sections were deparaffinized and exposed to Kodak emulsion (Eastman Kodak Co., Rochester, NY). After 2 weeks, the slides were developed, and [3H]thymidine labeled vs. unlabeled nuclei in epithelium were quantitated. The amount of DNA in uterine horns was assessed by the method of Labarca and Paigan (21) with modifications (22). Hoescht Dye 33258 fluorescence was measured on a PerkinElmer (Norwalk, CT) 650-40 fluorescence spectre-photometer. Trichloroacetic acid (TCA)-precipitable radioactivity was determined by collection of the precipitate by vacuum filtration after the samples incubated 20 min in 10% TCA on ice. Ecoscint A scintillation fluid was added to dry filters and the samples were counted 24 h later. Source of chemicals [3H]M;yo-inositol (15 Ci/mmol) was obtained from American Radiolabeled Chemicals (St. Louis, MO). [3H]Phosphatidylinositol-4-phosphate, [3H]phosphatidylinositol-4,5-bisphosphate, [3H]inositol (l,4,5)-trisphosphate, [3H]inositol (1,4)bisphosphate, and [3H]inositol (l)-phosphate were purchased from New England Nuclear-DuPont (Wilmington, DE). DES, carbachol (carbamylcholine), neomycin, glyceryl glass beads, phytic acid, Hoescht dye 33258, and LDH assay kit were purchased from Sigma Chemical Co. (St. Louis, MO). Ag-1-X8 resin was obtained from Bio-Rad Laboratories (Rockville Centre, NY). ICI 164,384 was a gift from Dr. Alan Wakeling (ICI Pharmaceuticals, Cheshire, England). Statistics The data in this study were analyzed by one- or two-way analysis of variance. Duncan's multiple range test was used for post-hoc analysis. P values less than 0.05 were considered significant and were denoted by asterisks.

2425

Results Elevation of the amount of [3H]myo-inositol incorporation into PI lipids has been previously reported in adult ovariectomized (OVX) mice treated with DES (5). Figures 1 and 2 depict the time course and dose-response curve, respectively, of DES-induced inositol incorporation into uterine lipids in prepubescent mice. Although a significant increase in radiolabeling was observed at 1.25 Mg/kg and above, 5 fig/kg was administered in the rest of the experiments, because this dose gave the most reproducible response. A single dose of DES induced a persistent elevation of [3H]myo-inositol incorporation which occurred as early as one 1 h after DES administration. The data in both experiments are similar to the data obtained in adult OVX mice (5), except that the effect of DES is already significant by 1 h after DES administration in the immature mice. The elevation of [3H]m/yo-inositol labeling in immature mice persists until at least 18 h post treatment (data not shown). The effect of vehicle injection on [3H]myo-inositol incorporation was tested at several early and late time points. Since no effect was observed, for the remainder of the studies the saline injection was administered at the 12-h time point. In direct contrast to the effect observed only 1 h after in vivo DES treatment, incubation of uteri with 100 nM DES during [3H]myo-inositol labeling for 1, 2, or 3 h had no effect on the amount of label incorporated into inositol lipids (Fig. 3). Z,Z-Dienestrol, a very weak estrogen receptor agonist, was able to mimic the effect of DES on PI labeling in adult OVX mice (5). This finding raised the possibility that estrogens affect PI metabolism through a mechanism that is not mediated by the estrogen receptor. Therefore, it was important to determine if DES could directly stimulate the production of inositol phosphates. As shown in Fig. 4, 100 nM DES did not elevate total inositol phosphates (expressed as percentage of [3H] phosphoinositides) vs. control. A dose of 1 nM DES yielded the same results (data not shown). Conversely, carbachol, a muscarinic receptor agonist, which was used as a positive control for a membrane receptor agonist known to activate phospholipase C (23), stimulated the production of IP's in a time-dependent manner. All three uterine cellular compartments may participate in the response to carbachol. Varol et al. (24) demonstrated that muscarinic receptors in the myometrium modulate PI lipid turnover, and we have observed that isolated epithelial and stromal cells will respond to carbachol with increased IP generation (data not shown). The specific ER antagonist ICI 164,384 was used to ascertain whether the in vivo effect of DES is mediated through classical estrogen receptors. As shown in Fig. 5, the effects of both 3- and 12-h DES treatments were

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 July 2015. at 11:58 For personal use only. No other uses without permission. . All rights reserved.

DES ELEVATES UTERINE PI LIPID METABOLISM

2426

Endo«1991 Voll29«No5

275 FIG. 1. Time course of DES effect on [3H[myo-inositol incorporation into phosphoinositicles. Mice were killed at the same time after staggered injections of DES (5 jig/kg). DES data was compared to uteri from mice injected with vehicle 12 h before killing. Uteri were removed, incubated with [3H]myo-inositol for 1.5 h, and the lipids extracted. The incorporated radioactivity and inorganic phosphate were determined as described in Materials and Methods. Data are expressed as the mean ± SD as a percent of inositol lipid labeling in the vehicle control (9894 ± 1710 dpm/nmol lipid phosphate). F (4,25) = 17.3. *, Treatment group was different from control; n = 5 for all groups.

C3

"3

250

-

225

LU 00

200

Diethylstilbestrol stimulates persistent phosphatidylinositol lipid turnover by an estrogen receptor-mediated mechanism in immature mouse uterus.

The effect of estrogen on phosphoinositide (PI) metabolism was evaluated in the immature mouse uterus, a tissue which undergoes estrogen-induced proli...
1015KB Sizes 0 Downloads 0 Views