Regulation of Expression of Epidermal Growth Factor Receptors in ...

0013-7227/97/$03.00/0 Endocrinology Copyright © 1997 by The Endocrine Society

Vol. 138, No. 12 Printed in U.S.A.

Regulation of Expression of Epidermal Growth Factor Receptors in Gonadotropes by Epidermal Growth Factor and Estradiol: Studies in Cycling Female Rats* JENNIFER L. ARMSTRONG

AND

GWEN V. CHILDS

Department of Anatomy and Neurosciences, University of Texas Medical Branch, Galveston, Texas 77555 ABSTRACT Changes in expression of epidermal growth factor (EGF) receptors by gonadotropes parallel those of GnRH receptors. Gonadotropes increase their expression of EGF receptors (EGFR) during diestrus to reach a peak on the morning of proestrus. This is followed by a decline in expression to reach a nadir by estrus. We hypothesized that regulatory factors that stimulate changes in GnRH receptors might mediate the same changes in EGFR. To test this hypothesis, pituitary cells were collected from cycling rats and grown overnight in media with or without serum, 100 pM estradiol, or 60 ng/ml activin. On the next day, some of the cultures were further stimulated with 1 nM GnRH (4 h). The cells were then dual-labeled for EGFR and LHb or FSHb antigens and analyzed for their content of EGFR and gonadotropins. Neither activin nor estradiol increased percentages of cells with gonadotropin antigens and EGFR. Estradiol decreased percentages of cells with EGFR and LH in proestrous rats and those with EGFR and FSH in diestrous rats. The estradiol-mediated decline in EGFR expression during proestrus is similar to that seen when GnRH receptors are studied. Serum containing media alone increased percentages of LH and FSH cells with EGFR in populations from estrous

E

PIDERMAL growth factor (EGF) may be an important regulatory factor in the pituitary. It stimulates TSH (1) PRL (2), ACTH (3, 4) or GH (3) release. In addition, it appears to be involved with the regulation of the reproductive system. Miyake et al. (5) found that perifusion of pituitary tissue with EGF stimulates the release of LH. Their results showed that EGF regulates the secretion of gonadotropins both by stimulating the hypothalamus and by increasing pituitary responsiveness to estradiol. Radford et al. (6) demonstrated direct effects of EGF on LH release. EGF may regulate pituitary cells via paracrine or autocrine mechanisms. Previously, we demonstrated that EGF messenger RNA (mRNA) is produced by subsets of all of the pituitary cell types, including gonadotropes (7). Another group has shown that subsets of all pituitary cells will release EGF in reverse hemolytic plaque assays (8, 9). In a study of immature pituitary cells, Mouihate et al. (10) showed that most (72%) of the EGF-secreting cells were LH antigen-bearReceived June 6, 1997. Address all correspondence and requests for reprints to: Jennifer L. Armstrong, Department of Anatomy and Neurosciences, MRB 10-104, 303 University Boulevard, University of Texas Medical Branch, Galveston, Texas 77555. E-mail: [email protected]. * The work was completed using Grant R01-HD-33915 and a developmental grant from the Sealy Smith Foundation, UTMB. The work is submitted in partial fulfillment of the Ph.D. Dissertation requirements for Jennifer L. Armstrong.

or metestrous rats. Therefore, further experiments were conducted to learn if serum factors or EGF might be a regulator. Removal of serum from the growth media did not prevent the increase in percentages of LH cells with EGFR over the 18-h growth period. However, removal of serum did prevent the increased percentages of FSH cells with EGFR. Similarly, adding 1:100 anti-EGF to the serum containing media did not affect expression of EGFR by LH cells. However, it did cause a 27% decrease in percentages of FSH cells with EGFR. Finally, when 10 ng/ml EGF was added to metestrous populations in serumfree media there was a 1.4- 1.5-fold increase in percentages of LH or FSH cells with EGFR. Collectively, these studies show that EGF receptors are not stimulated in gonadotropes by the same hormones that up-regulate GnRH receptors. Furthermore, EGF itself may be among the factors that up regulate EGFR in gonadotropes. EGF receptors may be down-regulated by estradiol during proestrus, but the effect is limited to LH cells. Finally, EGF’s differential effects on LH and FSH cells suggests that it may selectively act on monohormonal gonadotropes. EGF receptors may be a marker for a unique subset of developing gonadotropes. (Endocrinology 138: 5434 –5441, 1997)

ing gonadotropes and that the secretion of EGF could be enhanced by 10 nm GnRH. Collectively, these data suggest that EGF is a paracrine or autocrine mediator that may be regulated by key reproductive hormones. To further support the hypothesis for autocrine mechanisms, Mouihate and Lestage reported data that showed that most (73%) of the EGF secreting cells also express EGF receptors (11). Recent studies in this laboratory have characterized the distribution of GnRH or EGF receptors and changes in their expression during the estrous cycle (7, 12–15). Dual-labeling studies showed that subsets of all pituitary cell types expressed GnRH or EGF receptors (7, 12–16). Furthermore, the changes in EGF receptor expression by gonadotropes were similar to those reported previously for GnRH (13–16). There is an increase in gonadotropes that bind EGF (12) or GnRH (13–16) from metestrus to the morning of proestrus. The increase in GnRH receptive cells agreed with results from radioreceptor assays (reviewed in Refs. 13–16). After the proestrous peak, there is a decline in EGF or GnRH receptor expression by gonadotropes to reach a nadir by estrus (12–16). These parallel changes suggested that common regulatory factors might control expression of both sets of receptors. We have already shown that the increase in GnRH receptors could be partially mimicked by treating cells from metestrous or diestrous rats with estradiol for 18 –24 h (13).

5434

EGFR EXPRESSION REGULATION IN GONADOTROPES

It could be completely mimicked by treating diestrous rat pituitary cells with activin (15). Similarly, the decline seen during the afternoon of proestrus could be mimicked by estradiol (13) or inhibin (16) pretreatment of cells taken during the morning of proestrus (13). Thus, we hypothesized that the same regulatory factors might be involved in the expression of EGF receptors. The present studies applied the same experimental design used for the studies of GnRH receptors (13–15) to learn if estradiol or activin would affect expression of EGF receptors. The results of the first set of experiments showed that neither estradiol nor activin stimulated expression of EGF receptors by gonadotropes from metestrous or diestrous rats. This study presents these results along with evidence that EGF itself may be one of the regulators for increased expression of EGF receptors by gonadotropes. Materials and Methods Collection and dispersion of pituitaries Female Sprague-Dawley rats were housed three per cage under artificial illumination between 0600 h and 1800 h and given food and water ad libitum. They were acclimated for one week before the start of daily vaginal smearing performed to determine the stage of the cycle. The females underwent at least two complete estrous cycles before they were used. The animal care and use protocol was approved annually by the Institutional Review Committee.

Collection and growth of pituitary cells The anterior pituitary glands were collected in the morning and rapidly placed in cold DMEM (JRH Biosciences, Lenexa, KS) containing 0.3% BSA (Sigma Chemical Co., St. Louis, MO), 1.8 g/500 ml HEPES (Sigma), and 24.65 ml/500 ml sodium bicarbonate (JRH Biosciences). In order to prevent bacterial growth, 1 ml/100 ml gentamicin (Sigma) was used. The dissociation protocol was performed as reported previously (7). Cells were tested for viability by the trypan-blue dye exclusion test. The protocol produced 2–3 million cells/pituitary that were 98% viable. Cells were suspended in DMEM containing 0.005 mg/ml insulin (Sigma), 0.05 mg/ml transferrin (Sigma), and 0.001 mm sodium selenite (Johnson Matthey Chemical Ltd., New York, NY).

Experimental design Group 1: This group of experiments was designed to test the hypothesis that estradiol or activin might be among the regulatory factors that affect the expression of EGFR by gonadotropes. In addition, because estradiol or activin might work indirectly, by first increasing GnRH receptors, we added experiments that included exposure to GnRH during the last 4 h of treatment. Cells were plated on glass coverslips (Thomas Scientific, Suredesboro, NJ) that had been coated with poly-d-lysine (Sigma) in a 24-well tray at a density of approximately 40,000 –50,000 cells/50 ml/well. They were placed at 37 C for 1 h and then fed with the same DME containing 10% FBS (JRH Sciences). The cells were grown overnight (18 h) in this media with or without 100 pm b-estradiol 3-benzoate (Sigma). A second set of cultures were grown with or without 60 ng/ml activin (Genentech, San Francisco, CA). After the overnight plating, one half of each set of cultures was stimulated with 1 nm GnRH (Peninsula Laboratories, Inc., Belmont, CA) for 4 h at 37 C. Group 2: This experiment was designed after the results from the study of activin, estradiol and GnRH showed that none of the regulatory hormones stimulated increases in EGF receptor expression by gonadotropes. However, serum-containing media alone increased percentages of cells with EGF and gonadotropins in estrous and metestrous rats. Therefore, the second group of experiments was designed to test the hypothesis that EGF or other factors in serum might promote the expression of EGF receptors. Cells from metestrous rats were grown 18 h in DME with or without serum. Another group of cultures also received

5435

media with serum and a 1:100 dilution of anti-EGF (rabbit antimouse, Upstate Biotechnology Inc., Lake Placid, NY). Finally, one-half of the cultures grown in serum-free media received 10 ng/ml EGF during the culture period. This dose was chosen following dose-response tests. The cultures were then fixed in 2% glutaraldehyde for 30 minutes followed by washing for 1 h with 4.5% sucrose. Cells were then stored in the refrigerator until used for immunocytochemistry.

Immunocytochemistry Single and dual-immunolabeling with EGF receptor and LH or FSH was performed on the coverslips, as previously described (7, 12–15). Briefly, cells were exposed to a 1:1000 dilution of anti-EGFR overnight at room temperature. The antibody binding sites were detected by the DAKO (Carpinteria, CA) rapid immunoperoxidase kit using a biotinylated horse antimouse IgG, streptavidin peroxidase, and nickel-intensified diaminobenzidine. This gave a blue-black reaction product. The pituitary gonadotropins were each detected in separate protocols. The cells were incubated for 2 h at 37 degrees with 1:30 K antibovine LH-beta or 1:10 K antihuman FSHb. Heterologous antisera were used to prevent reactions from contaminating antibodies against a-subunits, which are unique when different species are compared. The gonadotropin antibodies were detected with the DAKO rapid kits which utilized a biotinylated goat antirabbit IgG, streptavidin peroxidase and orange-amber diaminobenzidine as the detection systems. The EGF receptor antibody was a monoclonal antibody (E3138, Sigma) raised against the intracellular domain of the EGF receptor. Specificity tests and controls have been reported previously (7). These were done both by preabsorption of the antiserum with the intracellular domain fragment as well as tests of reactivity of the antiserum in a dot blot assay. Preabsorption of the anti-EGFR antibody abolished labeling. Anti-bovine LHb was a gift from Dr. J. G. Pierce and antihuman FSHb was provided by Dr. A. F. Parlowe and the Pituitary Hormone Distribution Program (NIH). Both antisera have been tested repeatedly in tests that involved preabsorption of the antisera with either LH or FSH. The homologous antigens neutralized labeling with the antibody. Crossabsorbing each antiserum with the other gonadotropin had no affect on immunolabeling.

Analysis of immunolabeled cell populations One to two rats per stage of the estrous cycle were collected for each experimental group. The experiments were repeated until 6 –9 rats were collected. Therefore, each experiment was repeated at least three times. The final averages reported are of data from each of the repeated experiments (n 5 3–5). Data points for each experiment were collected by counting cells plated on three coverslips. The counts were done at 403. Fields were randomly selected and the first 100 –250 cells encountered were counted and analyzed. The raw counts of cells with single or dual labels were inserted into an Excel template which automatically calculated the percentage of cells labeled for each antigen as well as the percentage of dual labeled cells. This allowed us to constantly monitor the dual labeling to ensure that it did not interfere with the detection of either antigen. The percentages from the three coverslips provided an average for that experiment. These values were then averaged to produce the final data point (n 5 3–5 experiments). To determine significance in a group of experiments, 13 ANOVA or 23 ANOVA tests were performed. If the f value was significant (P , 0.05), individual differences were then detected by the Fisher’s least significant difference (LSD) post-hoc test.

Results Up-regulation of EGF receptor expression in gonadotropegonadotropes

The 18-h culture experiments were a continuation of studies that compared EGFR expression in all pituitary cell types during the different stages of the estrous cycle (12). In the experiments described in the recent report (12), the cells had been cultured only 1 h after removal from the rat. Therefore, the initial tests in the present studies compared data from the

5436


1 h cultures with those from the 18-h cultures to learn if the pituitary gonadotropes changed their expression of EGFR. We had already shown that the culture period did not affect gonadotrope expression of GnRH receptors over an 18- 36-h period (13). In the present studies, the percentages of proestrous pituitary cells with LH or FSH and EGFR remained the same over the 18-h culture period. However, the serum containing media alone caused a 1.8- to 2.25-fold increase in percentages of metestrous or estrous cells with EGFR and LH. Similarly, serum-containing media increased percentages of cells with FSH and EGFR 2 fold in estrous, metestrous or diestrous rats. [Note: As previously reported (12), 1 h after culture, LH cells with EGFR were 2.66 6 0.6% (metestrus), 2.7 6 0.6% (estrus), and 5.9 6 0.9% (diestrus) of pituitary cells, and FSH cells with EGFR were 4 6 0.6% (metestrus), 3.8 6 0.8% (estrus), and 3.8 6 0.6% (diestrus) of pituitary cells.] Because of the changes during the culture period, the 18-h cultures showed no significant differences in expression of EGF receptors by gonadotropes when the different stages of the cycle were compared (Figs. 1 and 2). Also, analysis of the percentages of LH or FSH cells showed that the 18-h culture period did not increase percentages of gonadotropes. They were within the range reported in previous studies (17–21). Figures 1 and 2 also show that GnRH did not affect percentages of pituitary cells with LH and EGFR if it was given during the last 4 h of culture. It did stimulate percentages of cells with FSH and EGFR in estrous rats (Fig. 2) when compared with 1 h cultures (12). However, the values were not different from those in the control cultures. Figures 3 and 4 show the effects of estradiol on expression

FIG. 1. Analysis of dual-labeled fields showing the effects of culture conditions on the percentages of pituitary cells with LHb antigens and EGF receptors. The stars show experimental groups in which growth overnight in serum-containing media (18 h) stimulated an increase in percentages of LH cells with EGFR. The data were compared with those reported in Ref. 12. The values for the 1-h experiments are cited in the text. As a result of these increases, there were no changes in percentages of pituitary cells with LH and EGFR when different stages of the cycle were compared. Furthermore, GnRH added during the last 4 h of culture had no stimulatory effect on expression of EGFR by LH cells.

Endo • 1997 Vol 138 • No 12

FIG. 2. Analysis of dual-labeled fields showing the effects of culture conditions on the percentages of pituitary cells with FSHb antigens and EGF receptors. The stars show experimental groups in which growth in serum-containing media for 18 h caused an increase in percentages of pituitary cells with of FSH and EGFR. The increased expression was similar to peak proestrous values reported in previous studies (12). Therefore, the cultures showed no changes in percentages of pituitary cells with FSH and EGFR with the stage of the estrous cycle. GnRH treatment caused significantly more cells from estrous rats to produce FSH and EGFR, but the values were not different from those seen after 18 h in media alone. GnRH had no effects on percentages of cells with FSH and EGFR from other stages of the cycle.

FIG. 3. Analysis of fields dual labeled for LHb and EGF receptors after estradiol treatment. Estradiol had no stimulatory effects on percentages of pituitary cells with LH and EGFR. Estradiol reduced percentages of pituitary cells with LH and EGF receptors if they came from proestrous rats (as shown by the stars, P , 0.05). GnRH did not promote recovery of the losses mediated by estradiol nor did it stimulate expression of EGFR by LH cells from other experimental groups.

of EGFR by LH or FSH gonadotropes. Estradiol had no stimulatory effects on percentages of pituitary cells with EGFR and gonadotropins during any stage of the cycle. This is in striking contrast to estradiol’s stimulatory effects on


5437

FIG. 4. Analysis of fields dual labeled for FSHb and EGF receptors after estradiol treatment. Estradiol reduced percentages of pituitary cells with FSH and EGFR from diestrous rats (as shown by the stars, P , 0.05). It had no effects on FSH cells from other stages of the cycle. GnRH did not promote recovery of the losses mediated by estradiol nor did it increase expression by FSH cells from other stages of the cycle.

expression of GnRH receptors by cells from metestrous or diestrous rats (13). Furthermore, in spite of the fact that cells with GnRH receptors are elevated following estradiol treatment of metestrous and diestrous populations (13–15), GnRH treatment during the last 4 h of culture did not stimulate more LH or FSH gonadotropes with EGF receptors. Finally, parallel tests of activin modeled after recently published studies (15) showed that it also did not stimulate expression of EGF receptors when added with or without GnRH (data not shown). Two-way ANOVA tests showed no significant differences among activin treated and untreated groups (P , 0.05) in all stages of the cycle studied. Downregulation of EGF receptors in gonadotropes

Figure 3 shows that estradiol decreased the percentages of pituitary cells with LH and EGFR in proestrous rats. This is consistent with its known effects on GnRH receptors in this same group (13). It did not affect percentages of cells with FSH and EGFR in proestrous rats (Fig. 4). However, estradiol caused a significant reduction in cells with FSH and EGFR in diestrous rats (Fig. 4). Regulatory effects of EGF on expression of EGFR by gonadotropes

Because of the media effects on gonadotropes early in the cycle (Figs. 1 and 2), the second group of experiments focused on metestrous rat populations. The experiments were designed to test the hypothesis that EGF might be one of the regulatory factors responsible for up-regulation of EGFR by gonadotropes. Figure 5 shows that removal of serum from the media did not block the increase in percentage of cells with LH and EGFR over the 18-h culture period. As shown in the previous experiments (Fig. 1), serum caused a 2.4-fold increase in LH cells with EGFR when compared with the 1-h cultures (2.66 6

FIG. 5. Analysis of cells dual-labeled for LHb or FSHb and EGFR. Metestrous rat cell populations were grown in media with (DME 1 serum) or without (DME only) serum for 18 h. A 1:100 dilution of anti-EGF was added to some of the cultures grown in serum containing media (serum 1 anti-EGF). Finally, 10 ng/ml EGF were added to some of the cultures grown in serum-free media (DME 1 EGF). As found in the first group of experiments (Figs. 1 and 2), growth overnight in serum containing media increased percentages of pituitary cells with LH and FSH and EGFR (see text for cited numbers). Removal of serum (DME only) significantly reduced only the percentages of pituitary cells with FSH and EGFR (P , 0.05). Adding anti-EGF serum to the media also reduced only the percentages of FSH cells with EGFR to values similar to those seen after 18 h of culture (single star 5 value significantly different from DME 1 serum values, P , 0.05). Exposure to EGF stimulated significant increases in percentages of pituitary cells with LH or FSH and EGFR (double stars 5 values significantly different from cells exposed to DME only, P , 0.05).

0.6% of pituitary cells have LH and EGFR in the 1 h cultures, 12). However, Fig. 5 shows that, even if serum was removed, there was still a 2.2-fold increase in percentages of pituitary cells with LH antigens and EGFR after 18 h of culture. In contrast, removal of serum resulted in lowered percentages of pituitary cells with FSH and EGFR. The resulting percentages were similar to levels seen after 1 h of culture (4 6 0.6% of pituitary cells have FSH and EGFR, 12). When anti-EGF serum was added to the serum-containing media, there were no changes in the percentages of LH cells with EGFR. However, anti-EGF serum caused a reduction in percentages of FSH cells with EGFR to levels seen in serum-free conditions. Finally, Fig. 5 also shows that adding EGF to metestrous cells for 18 h increased pituitary cells with EGFR and LH or FSH 1.4- to 1.5-fold. The resulting percentages are similar to those seen during peak expression periods in proestrous rats (Figs. 1 and 2). Counts of immunolabeled LH and FSH cells showed that this change did not coincide with increased percentages of LH- or FSH-bearing gonadotropes.

5438


Figures 6 – 8 illustrate the effects of EGF on EGF receptor expression by the population of LH cells from metestrous rats. Figure 6 and 7 are lower magnifications to show the changes overall in the cells populations. The insets show examples of both single and dual-labeled cells. The most striking morphological change seen in the EGF treated cells is the extension of processes. Even the normally round or ovoid gonadotropes have extended processes and become more stellate. Higher magnifications are shown in Fig. 8, a and b, to illustrate the dual labeling more clearly. The EGF receptor labeling is in patches or a linear pattern on the cells (seen as dense blue-black patches in color), whereas the labeling for the antigens is more diffuse and gray (seen as orange-amber in color). Discussion

EGFR have been found on all pituitary cell types (7, 12). This study focused on regulation of changes in their expression by gonadotropes during the estrous cycle. In previous studies, we showed that expression of EGF receptor proteins changes following cold stress or with the stage of the cycle (7, 12). Gonadotropes produce more EGF receptor proteins during diestrus and proestrus (12). Furthermore, the changes in EGFR expression by gonadotropes (12) paralleled those seen for GnRH receptors (13–16). This led us to hypothesize that the same regulatory factors controlling the expression of GnRH receptors during diestrus might also control gonadotrope expression of EGF receptors.

Endo • 1997 Vol 138 • No 12

Because estradiol and activin appear to be primary stimulators of the expression of GnRH receptors, these studies used the same experimental design that showed estradiol (13) and activin (15) up-regulation of GnRH receptors during metestrus and diestrus. Estradiol pretreatment also mimicked the decreases in GnRH receptive cells seen during proestrus (13). Experiments with GnRH were performed (during the last 4 h of the experiment) to learn if the changes in EGF receptor expression might be secondary to the up- or down-regulation of GnRH receptors. The results of the first group of experiments showed that the up-regulation of EGF receptors in gonadotropes appears to be mediated by factors other than those that increase expression of GnRH receptors. Neither activin nor estradiol increased percentages of pituitary cells with LH or FSH and EGFR. GnRH also had no stimulatory effect in most experimental groups, even if cells were stimulated to promote expression of GnRH receptors. The one group that showed a GnRH-mediated change were the FSH cells from estrous cell populations. GnRH stimulated an increase in percentages of cells with FSH and EGFR above the values seen after only 1 h (12). The values were not greater than those after 18 h in media alone, however. The estradiol mediated decline in percentages of pituitary cells with LH and EGFR seen in proestrous populations was similar to that of GnRH receptors. In recent studies of inhibin-treated populations, we have also shown that inhibin reduces the expression of GnRH receptors by proestrous rat

FIG. 6. Photograph illustrating control fields grown in DME overnight and dual-labeled for LHb and EGF receptors. The arrows indicate the dual-labeled cells. Magnification, 10303; bar, 10 mm.


5439

FIG. 7. Photograph illustrating fields grown in DME overnight and stimulated with 10 ng/ml EGF and dual-labeled for LHb and EGF receptors. Note the increase in EGF receptive cells, including those with LHb antigens. Also, all labeling patterns are represented in these fields. Arrowheads indicate the dual-labeled cells; solid arrows point to cells labeled only for EGF receptors and the open arrows point to cells labeled only for LH. The insets show higher magnifications of these cells. Note also that the population has become more stellate or angular. Magnification, 8253; bar, 10 mm. Magnification of the insets, 13403, bar, 10 mm.

gonadotropes (16). Future studies are needed to test possible effects of inhibin on gonadotrope expression of EGF receptors. In contrast, estradiol had no effects on percentages of cells with FSH and EGFR in proestrous populations. However, estradiol did reduce percentages of pituitary cells with FSH and EGFR in diestrous rats. These data are difficult to explain in light of the general positive feedback effects of estradiol seen normally during diestrus (13). Differential expression of EGFR by LH and FSH cells

A common theme found in this and the previous (7, 12) studies is the nonparallel expression of EGFR when LH and FSH gonadotropes are compared. In the first study (7), we showed that cold stress increased expression of EGFR by FSH cells only. It had no effect on LH cells. The more recent studies (12) showed a similar increase in expression of EGFR in antigen-bearing LH or FSH cells from estrous to proestrous. However, the increase in percentages of cells with FSH and EGFR occurred later than that with LH and EGFR. Furthermore, when LHb and FSHb mRNAs were detected by in situ hybridization in metestrous populations, proportionately more cells with EGFR and LHb mRNA were found. (Fifty percent of total metestrous LH cells contained EGFR compared to only 25% of total FSH cells.) Collectively,

these data suggested that EGFR may be produced earlier in the cycle by LH cells. Furthermore, we suggested that EGFR was produced initially by a population of monohormonal LH cells (12). The present studies find more examples of non-parallel expression of EGFR by LH and FSH gonadotropes. In diestrous populations, serum containing media increased the percentages of pituitary cells with FSH and EGFR while having no effects on cells with LH and EGFR. Removal of serum or adding anti-EGF sera to the media abolished the serum-mediated increase in EGFR expression by metestrous FSH cells, whereas it had little effect on EGFR expression by metestrous LH cells. These data suggest that different factors in serum may regulate increases in EGFR in FSH cells. Finally, as stated in the previous section, estradiol also differentially regulated EGFR in LH and FSH cells. It had potent effects on expression of EGFR by LH cells in proestrous rats and FSH cells in diestrous populations. Collectively, these data point to EGFR expression by a subset of monohormonal gonadotropes. These represent 25– 50% of the gonadotrope population depending on the stage of the cycle. They are also more numerous in estrus and metestrus (17–21). Often these cells are small and considered developmentally immature. However, sometimes they can be found among the largest gonadotropes (21). The presence

5440


FIG. 8. A and B, Higher magnification of cell dual labeled for LHb and EGF receptors. Note the changes in the shape and labeling pattern of cells after stimulation with EGF. Field a is a control field and field b shows a field stimulated overnight with EGF. Magnification: a, 23463, bar, 5 mm; b, 28353; bar 5 5 mm.

of EGFR may be a marker for monohormonal gonadotropes that are either immature or subserving a different function in the population. They may have different sensitivities to regulatory factors, like estradiol or EGF. Further tests are needed to learn their significance. Finally, we recognize that other factors, in addition to EGF could be involved in stimulating the expression of EGF receptors during early maturation. This could include factors produced by pituitary cells themselves. Further studies would be needed to identify them. Functions for EGF in gonadotrope regulation

The changes in expression of EGF receptors during the cycle suggests that they may be used by gonadotropes to modulate or mediate functional responses. There is evidence

Endo • 1997 Vol 138 • No 12

in the literature that EGF may be both a differentiating factor and a secretagogue for gonadotropes. Przylipiak et al. (22) reported that EGF enhanced the secretion of LH in dispersed monolayers of pituitary cells as well as in superfusion columns. The stimulatory effects were both dose and time dependent. Miyake et al. (5) found an increase in LH secretion when EGF was administered in a sequential double chamber perifusion system. They used the medial basal hypothalamus in one chamber in series with another chamber containing anterior pituitary quarters. The group reported increased LH release when EGF was added with estradiol. As stated in the introduction, EGF may actually function via autocrine or paracrine mechanisms. Studies by Mouihate et al. reported EGF release by gonadotropes and other cell types in a number of physiological states related to the reproductive system (11). Three sets of findings suggest that EGF may work via autocrine mechanisms. First, in lactating rats, nearly 3⁄4ths of EGF-secreting cells also expressed EGF receptors. Second, in immature female rats, over 70% of EGF secreting cells were LH gonadotropes (9). Third, GnRH stimulated more EGF secretion from individual cells which points once again to a gonadotrope source. Mouihate et al. (8) also reported differences in EGF secretion with the stage of the cycle (8) that correlate with changes in expression of EGFR in our studies (12). Cells from proestrous rats secreted higher amounts of EGF than those from metestrous rats. However, unlike our studies of EGFR, estradiol stimulated increases in the metestrous group to levels similar to those from proestrous populations (8). These data provide clues to an EGF regulatory circuit. This may begin with an increase in estradiol receptors during diestrus (23, 24). This is followed by estradiol stimulation of the production of GnRH receptors (reviewed in Ref. 13). Both estradiol and GnRH may stimulate secretion of EGF (8, 9). Evidence from the present study suggests that EGF itself may increase expression of EGF receptors. Thus, estradiol and GnRH may work indirectly through EGF to elevate EGF receptors. To summarize, these data suggest that EGF may stimulate the production of its own receptor by gonadotropes early in the cycle. GnRH appears to have little if any direct effect on expression of the EGF receptors by either LH or FSH cells. Estradiol has a negative feedback effect on EGF receptor expression by proestrous LH gonadotropes. The fact that estradiol can inhibit expression of EGFR during proestrus points to it as a potential regulatory hormone for the receptor later in the cycle. However, the same feedback effects are not seen on FSH gonadotropes. This supports the hypothesis that the target cells are monohormonal LH cells. Nonparallel changes in the expression of EGF receptors by LH and FSH cells seen in this and the previous studies (7, 12) suggest expression of EGFR by monohormonal subtypes that have different periods of sensitivity to EGF stimulation or estradiol regulation. In past studies, the monohormonal subtypes have been characterized as small and relatively immature (17–21). They may comprise a population of gonadotropes that are being readied for a future cycle. Alternatively, they may serve a different function on the population. Perhaps the EGFR modulate their development or responsiveness to secretagogues.

EGFR EXPRESSION REGULATION IN GONADOTROPES Acknowledgments The authors wish to thank Ms. Geda Unabia and Ms. Diana Rougeau for their technical assistance, and Drs. J. G. Pierce and A. F. Parlowe for the LHb and FSHb antibodies, respectively. We also thank the Hormone Distribution Program for the pituitary hormone antigens.

References 1. Altschuler L, Parisi M, Cageao L, Chiocchio S, Fernandez-Pol J, Zaninovich A 1993 Epidermal growth factor stimulates thyrotropin secretion in the rat. Neuroendocrinology 57:23–27 2. Aanestad M, Rotnes J, Torjesen P, Haug E, Sand O, Bjoro T 1993 Epidermal growth factor stimulates the prolactin synthesis and secretion in rat pituitary cells in culture (GH4C1 cells) by increasing the intracellular concentration of free calcium. Acta Endocrinol 128:361–366 3. Fisher D, Lakshamanan J 1990 Metabolism and effects of epidermal growth factor and related growth factors in mammals. Endocr Rev 11:418 – 442 4. Childs GV, Patterson J, Unabia G, Rougeau D, Wu P 1991 Epidermal growth factor enhances ACTH secretion and expression of POMC mRNA by corticotropes in mixed and enriched cultures. Mol Cell Neurosci 2:235–243 5. Miyake A, Tasaka K, Otsuka S, Kohmura H, Wakimoto H, Aono T 1985 Epidermal growth factor stimulates secretion of rat pituitary luteinizing hormone in vitro. Acta Endocrinol 108:175–178 6. Radford H, Panaretto B, Avenell J, Turnbull K 1987 Effect of mouse epidermal growth factor on plasma concentrations of FSH, LH and progesterone and on oestrus, ovulation and ovulation rate in Merino ewes. J Reprod Fertil 80:383–393 7. Fan X, Childs G 1995 Epidermal growth factor and transforming growth factor-a messenger ribonucleic acids and their receptors in the rat anterior pituitary: localization and regulation. Endocrinology 136:2284 –2293 8. Mouihate A, Lestage J 1995 Estrogen increases the release of epidermal growth factor from individual pituitary cells in female rats. J Endocrinol 146:495–500 9. Mouihate A, Verrier D, Lestage J 1996 Identification of epidermal growth factor-secreting cells in the anterior pituitary of lactating female rats. J Endocrinol 148:319 –324 10. Mouihate A, Verrier D, Lestage J 1996 EGF release by rat gonadotroph cells: Characteristics and effects of LHRH. Life Sci 58:107–114 11. Mouihate A, Lestage J 1995 Epidermal growth factor: a potential paracrine and autocrine system within the pituitary. Neuro Report 6:1273–1276

5441

12. Armstrong J, Childs GV 1997 Changes in expression of epidermal growth factor receptors by anterior pituitary cells during the estrous cycle: cyclic expression by gonadotropes. Endocrinology 138:1903–1908 13. Lloyd J, Childs G 1988 Changes in the number of GnRH-receptive cells during the rat estrous cycle: biphasic effects of estradiol. Neuroendocrinol 48:138 –146 14. Childs GV, Unabia G, Miller BT (1994b) Cytochemical detection of GnRH binding sites on rat pituitary cells with LH, FSH and GH antigens during diestrous upregulation. Endocrinology 134:1943–1956 15. Childs GV, Unabia G, Cytochemical studies of the effects of activin on gonadotropin releasing hormone (GnRH) binding by pituitary gonadotropes and growth hormone cells. J Histochem Cytochem, in press 16. Childs GV, Miller BT, Miller WL (1997) Differential effects of inhibin on gonadotropin stores and gonadotropin releasing hormone binding to pituitary cells from cycling female rats. Endocrinology 138:1577–1584 17. Childs GV, Unabia G, Tibolt R, Lloyd JM 1987 Cytological factors that support non-parallel secretion of LH and FSH during the estrous cycle. Endocrinology 121:1801–1813 18. Childs GV, Unabia G, Lloyd J 1992 Recruitment and maturation of small subsets of luteinizing hormone (LH) gonadotropes during the estrous cycle. Endocrinology 130:335–345 19. Childs GV, Unabia G, Lloyd JM 1992 Maturation of FSH gonadotropes during the rat estrous cycle. Endocrinology 131:29 –36 20. Childs G 1986 Functional ultrastructure of gonadotropes: a review. Curr Topics. Neuroendocrinol 7:49 –97 21. Childs GV, Hyde C, Naor Z, Catt K 1983 Heterogeneous luteinizing hormone and follicle-stimulating hormone storage patterns in subtypes of gonadotropes separated by centrifugal elutriation. Endocrinology 113:2120 –2128 22. Przylipiak A, Kiesel L, Rabe T, Helm K, Przylipiak M, Runnebaum B 1988 Epidermal growth factor stimulates luteinizing hormone and arachidonic acid release in rat pituitary cells. Mol Cell Endocrinol 57:157–162 23. Greeley Jr GH, Muldoon TG, Maheshi VB 1975 Correlative aspects of luteinizing hormone-releasing hormone sensitivity and cytoplasmic estrogen receptor concentration in the anterior pituitary and hypothalamus of the cycling rat. Biol Reprod 13:505–512 24. Kikuta T, Yamamoto K, Nmaiki H, Hayashi S 1993 Immunocytochemical localization of estrogen receptor in various anterior pituitary hormone cells of adult male and female rats. Acta Histochem Cytochem 26:609 – 613