|Home | About | Journals | Submit | Contact Us | Français|
A decline in dehydroepiandrosterone (DHEA) and GH levels with aging may be associated with frailty and morbidity. Little is known about the direct effects of DHEA on somatotropes. We recently reported that 17β-estradiol (E2), a DHEA metabolite, stimulates the expression of GH in vitro in young female rats. To test the hypothesis that DHEA restores function in aging somatotropes, dispersed anterior pituitary (AP) cells from middle-aged (12–14 months) or young (3–4 months) female rats were cultured in vitro with or without DHEA or E2 and fixed for immunolabeling or in situ hybridization. E2 increased the percentage of AP cells with GH protein or mRNA in the aged rats to young levels. DHEA increased the percentages of somatotropes (detected by GH protein or mRNA) from 14–16 ± 2% to 29–31 ± 3% (P ≤0.05) and of GH mRNA (detected by quantitative RT-PCR) only in aging rats. To test DHEA’s in vivo effects, 18-month-old female rats were injected with DHEA or vehicle for 2.5 d, followed by a bolus of GHRH 1 h before death. DHEA treatment increased serum GH 1.8-fold (7 ± 0.5 to 12 ± 1.3 ng/ml; P = 0.02, by RIA) along with a similar increase (P = 0.02) in GH immunolabel. GHRH target cells also increased from 11 ± 1% to 19 ± 2% (P = 0.03). Neither GH nor GHRH receptor mRNAs levels were changed. To test the mechanisms behind DHEA’s actions, AP cells from aging rats were treated with DHEA with or without inhibitors of DHEA metabolism. Trilostane, aminogluthemide, or ICI 182,780 completely blocked the stimulatory effects of DHEA, suggesting that DHEA metabolites may stimulate aging somatotropes via estrogen receptors.
GH IS AN IMPORTANT hormone that affects metabolism and body composition throughout life, in addition to playing a crucial role in promoting growth during puberty. There is a decline in GH (somatopause) with progressive aging in humans (1–3). GH replacement in the elderly has shown favorable effects on bone mineral strength, increase in lean body mass, decrease in adiposity, improvement in cardiac function, and quality of life (4–9). However, the high cost (10) and attendant risks and inconvenience of exogenous sc GH therapy (11, 12) warrant a search for other regulators that could maintain GH cell functions.
After peak expression in the early 20s, there is also a decline in levels of dehydroepiandrosterone (DHEA; adrenopause) with aging in humans (13). Studies looking at a possible stimulatory effect of DHEA on serum GH/IGF-I levels show variable results (increased or no change) in clinical trials on the elderly population (14–20).
The decline in these hormones during aging may be interrelated (3), especially because reports in the literature suggest that GH levels in women of reproductive age are influenced by cyclical changes in estrogen levels (21–23). Similarly, in young female rats, levels of GH synthesis and secretion vary with estrogen levels during the cycle (24, 25), and recent studies from our laboratory show that expression of GH mRNA and proteins are stimulated by low concentrations of estrogens (26). Because DHEA serves as an adrenal precursor hormone for estrogen, the decline in this potential source of estrogen may also influence the expression of GH.
At the level of the pituitary, the decline in GH appears to be related to a loss in numbers of identifiable GH cells. In humans, this appears during middle age (27, 28). In middle-aged rats (10–14 months), Takahashi et al. detected a progressive decline in pituitary GH and mean plasma GH (29) along with a decline in GH mRNA levels (30, 31). Similarly, Jurado et al. (32) reported a reduction in the density of immunoreactive GH cells by 20 months in female rats. Mechanisms behind the decline are not known, although it could also be related to changes in the expression or activity of hypothalamic GHRH and somatostatin (33–39).
Previous reports describing DHEA administration in aging rats (18 months old) have shown a reversal of age-related changes in various tissues, including the hypothalamus and pituitary (40–42). In a study of young animals, female rats implanted with DHEA (100-mg pellet) showed a significant increase in serum GH levels after 1 wk (43). These studies suggested that DHEA may have some functions in the pituitary; therefore, we hypothesized that DHEA may restore the loss in age-related GH gene expression in the pituitary of middle-aged female rats.
The first objective of this study was to determine whether DHEA acts directly on pituitary cells to restore losses in GH cells. After evidence for restoration was found, the study was expanded to learn whether DHEA acted on somatotropes in vivo. When DHEA’s positive actions were detected in vivo and in vitro, the study then focused on differential inhibitors of DHEA metabolism to learn whether it acted by itself via membrane receptors (44, 45) or if its action depended on metabolism to estrogens. This report presents the results of these studies, showing that DHEA may have a favorable effect on certain aspects of GH expression and that its effects may be mediated via estrogen receptors (ERs).
Young female rats (3–4 months old; weight, 200–250 g) obtained from Harlan Sprague Dawley (Indianapolis, IN) were used in this study. The animal care protocol was approved by the institutional animal care and use committee, annually. Animals were housed three or four per cage with a 12-h light, 12-h dark cycle (lights on at 0600 h) at a constant room temperature of 70 F. A standard pellet chow diet (rodent diet 8640, Harlan Teklad, Madison, WI) and water were available ad libitum. Animals were acclimated for approximately 2 wk before vaginal smears were started and were not used until two successful cycles were detected. The animals were raised in-house until they were 12–14 months old. The vaginal smears showed that all of these animals had stopped cycling and were in persistent estrus.
For the in vitro study, pituitaries from diestrous (3–4 month) and middle-aged rats (12–14 months; 220–300 g) were collected as described previously (26). Subsequent in vivo studies were focused on older rats (18 months). For the in vivo study, the animals were aged at Harlan Sprague Dawley, and they were 16 months of age when they arrived. They were acclimated for approximately 2 months before the start of the study. The animals were divided into two groups, A and B, and injected according to the protocol described by Givalois et al. (40). Group A was injected sc once every 12 h with 100 μl vehicle (absolute ethanol) for 2.5 d. Group B was injected on the same schedule with DHEA dissolved in the same amount of vehicle (Sigma-Aldrich Corp., St. Louis, MO) at a dose of 12 mg/kg body weight at 12-h intervals sc for 2.5 d. Two hours after the last DHEA injection, the animals were sc injected with GHRH (Sigma-Aldrich Corp.; 1 mg/kg body weight). One hour after the GHRH injection, they were anesthetized with ip injections of sodium pentobarbital (25 mg/kg or 0.5 ml/250 g rat) and then killed by guillotine.
Pituitaries from female rats (both diestrous and aged rats) were rapidly removed and dispersed into single-cell suspensions as described previously (26). These methods had been shown to preserve the hormone content and percentages of cells for at least 1 wk (compared with freshly dispersed cells or cells in tissue sections). The cells were resuspended in DMEM supplemented with insulin, transferrin, sodium selenite, and BSA (ITS; Sigma-Aldrich Corp.). They were plated by pipetting 20 μl cell suspension onto each of the polylysine-coated 13-mm coverslips (Thomas Instruments, Charlottesville, VA) and were allowed to adhere for 1 h (providing 40,000–50,000 cells/well). For the collection of mRNAs, 200 μl cell suspension was plated onto polylysine-coated tissue culture dishes and allowed to adhere for about 1 h. Additional DMEM plus ITS (400 μl) was added after 1–2 h in both the 24-well tissue culture trays and culture dishes and was left for incubation in a CO2 incubator at 37 C.
After an additional 1–2 h of incubation in DMEM plus ITS, the cells were exposed to various doses (0–250 nM) of DHEA or E2. DHEA was dissolved in absolute ethanol to make a stock of 0.1 mM solution. The E2 (water soluble; Sigma-Aldrich Corp.) was dissolved in DMEM to make a stock of 0.1 M solution. DMEM plus supplements were used to further dilute DHEA or E2 to the appropriate doses needed for stimulation. DMEM plus supplements with a final concentration of 0.25% alcohol or less (depending on the dose of DHEA) was used as vehicle. DMEM only was the vehicle for E2, because this compound was water soluble (Sigma-Aldrich Corp.). The cells were incubated for 24 h at 37 C.
For the in vitro studies, testing the mechanism of DHEA action, the anterior pituitary (AP) cells from three aged animals (killed at the same time) were incubated with vehicle or inhibitors for 30 min before adding DHEA. Trilostane (Sanofi Synthelabo, Paris, France), aminoglutethemide (Sigma-Aldrich Corp.), and ICI 182,780 (Sigma-Aldrich Corp.) were used to determine the mechanisms behind the DHEA actions. In previous tests, ICI 182,780 had blocked estrogen’s stimulatory actions on GH cells (data not shown). Stock solutions (0.01 M) of these inhibitors were diluted with DMEM plus ITS before adding to the cells. After 30-min incubation in the 5% CO2 incubator, DHEA was added to the cells (without removing the existing medium containing vehicle or inhibitors), so that the final concentration in the wells for both DHEA and inhibitors was 10 nM. The 24-plate trays were gently rotated to ensure that there was uniform concentrations of DHEA and inhibitors within the well. The trays were incubated for 24 h at 37 C in a 5% CO2 incubator until they were ready to be fixed.
After 24 h in culture with or without DHEA, the cells in 24-well tissue culture trays were fixed with 2% glutaraldehyde diluted in 0.1 M phosphate buffer (pH 7.4–7.6) for 30 min at room temperature. Fixation was followed by four washes for 15 min each with 0.1 M phosphate buffer containing 4.5% sucrose and 0.15% glycine before immunolabeling or in situ hybridization.
Fixed pituitary monolayer cells were pretreated with a 3% hydrogen peroxide solution for 10 min. After bleaching, the cells were washed with a 0.1 M PBS (pH 7.2–7.4) buffer three times. The cells were blocked with 0.1 M PBS diluent containing 1% BSA, 0.1% nonfat dry milk, and 5% normal goat serum. The cells were then incubated with the primary antibody generated against rat GH (National Hormone and Peptide Program) diluted 1:83,500 (in the blocking solution) for 30 min at 37 C, and washed with 0.1 M PBS buffer three times. Cells were then treated with a 1:200 dilution of biotinylated goat antirabbit IgG (Bio-Ganti, Vector Laboratories, Inc., Burlingame, CA) diluted in blocking solution for 30 min at room temperature. Each of the wells was washed three times after application of Bio-Ganti. The cells were then exposed to 1:200 streptavidin (DakoCytomation, Carpinteria, CA) for 30 min at room temperature and washed twice with 0.1 M PBS buffer and twice with acetate buffer. The peroxidase was detected with nickel-intensified diaminobenzidine solution as previously described (24).
The specificity of these antisera was proved by neutralization of the immunoreaction after absorption with 10–1000 ng/ml rat GH. Method controls included omission of the primary antibody or the secondary antibody, either of which prevented any labeling.
Coverslips with AP cells treated with biotinylated GHRH (1 nM, 10 min before fixation) were labeled with the avidin-biotin peroxidase complex (ABC Elite kit, Vector Laboratories, Inc., Burlingame, CA) and nickel-intensified diaminobenzidine (Black DAB). This involved treatment with hydrogen peroxide (3%) for 5 min, washing with 0.05 PBS (three times), and a blocking step (0.05 M PBS and 0.1% BSA, then exposure to avidin-biotin complex, made 30 min before use). The peroxidase was detected by nickel-intensified diaminobenzidine. After washing with 0.05 M PBS three times, the cells were left in blocking solution (0.05 PBS containing 1% BSA, 0.1% nonfat dry milk and 5% normal goat serum) for 30 min. The cells were then immunolabeled for GH as described previously (22), and the peroxidase was detected by diaminobenzidine solution (one tablet of diaminobenzidine and one tablet of Tris-HCl dissolved in 15 ml water with 12 μl 30% hydrogen peroxide; Sigma-Aldrich Corp.). A set of three coverslips per experimental group was then dehydrated and mounted. The number of GHRH-binding cells detected by black labeling was counted manually under a ×100 oil immersion (200 cells/coverslip, with at least three coverslips per animal), and the percentage of GHRH-bound cells was calculated.
After fixation in 2% glutaraldehyde for 30 min, followed by four 15-min washes in phosphate buffer containing 4.5% sucrose and 0.15% glycine (pH 7.4–7.6), the cells were used for in situ hybridization as previously described (26, 46). The antisense and sense oligonucleotide probes for GH mRNA were made by GeneDetect.com. The antisense probe hybridizes against nucleotides 64–101 of the rat GH gene mRNA (accession no. U62779). The sequence for the antisense probe is CTGAAGGTCAGGAGCCAGGGAGTCTGAGAGTCTGCAGC. Controls were described in a recent study (26) and included use of the sense sequence of GH mRNA or vehicle instead of the biotinylated antisense sequence. In addition, controls were run in which ×83 unlabeled probe was added with the labeled probe to successfully compete for mRNA sites and abolish labeling.
At the end of 24-h incubation, cells cultured in DMEM plus additives in tissue culture dishes were placed on ice, gently scraped using a cell scraper, and collected in a centrifuge tube by centrifugation at high speed for 10 min. Total cellular RNA was extracted using the RNeasy Mini Kit (Qiagen, Valencia, CA), which included the on-column DNA digestion step to ensure removal of any contaminating genomic DNA, following the manufacturer’s protocol. Assays for GHRH receptor (GHRH R; accession no. L01407) mRNA were performed on APs, which were immediately processed for mRNA extraction after decapitation by the same methods described above.
Total RNA (0.5–2 μg) was reverse transcribed using SuperScript II reverse transcriptase (Invitrogen Life Technologies, Inc., Carlsbad, CA) and 10 pM poly(deoxythymidine) primer in a 10-μl total volume reaction. The first cDNA synthesis reaction was heated to 55 C for 5 min to inactivate the reverse transcriptase and then diluted with 10 mM Tris-HCl (pH 7.5) to a final volume of 100–300 μl. An aliquot (4 μl) of cDNA was used for quantification of gene expression.
Standards for each gene for QRT-PCR were prepared according to the method of Zhou et al. (47). The cDNA fragments that were used as standard for each gene were amplified from a plasmid cDNA clone of the corresponding gene or from the first cDNA pool and purified from the gel using a QIAquick column (Qiagen). The molar concentration of the DNA was calculated based on the molecular weight of each DNA fragment, and the DNA concentration was measured by reading absorbance at 260 nm in a spectrophotometer. Each DNA fragment was diluted to 4.15 amol/μl, from which eight 10-fold serial dilutions were made, to be used as real-time PCR standards, (e.g. 107–104 molecules/4 μl). The real-time QRT-PCRs were carried out using a LightCycler Instrument and LightCycler FAST-START DNA Master SYBR Green I enzyme mix (Roche, Indianapolis, IN) with primers and optimized MgCl2 concentrations that specifically amplified the transcripts of rat genes encoding GH, ribosomal protein S9 (RPS9), and hypoxanthine guanine phosphoribosyltransferase (HPRT). The latter two genes served as internal controls for normalizing the expression of each target gene in different samples. We used deoxyribonuclease digestion after RNeasy silica membrane-based RNA purification to remove any trace of DNA. In addition, the forward and reverse primers for rat GH (accession no. V01239) anneals to exons 3 and 5, respectively. For rat HPRT (accession no. XM_217584) and RPS9 (accession no. BC004686) genes, the primers were designed over introns based on the exon/intron gene structures from mouse HPRT and RPS9, because complete genomic sequences in the GenBank for rat HPRT and RPS9 are not known. The primer sequence for GH and housekeeping genes are: GH: forward primer, 5′-GCAGAGAACTGACATGGAATTG-3′; reverse primer, 5′-TTTTGAGCAGAGCGTCATC-3′; GHRH R: forward primer, 5′-TGCCCTTGGAACTGCTAAC-3′; reverse primer, 5′-TGCCCTTGGAACTGCTAAC-3′; RPS9: forward primer, 5′-TTTGTCGCAAAACCTATGTGACC-3′; reverse primer, 5′-TTCTCGTCCAGCGTCAACAG-3′; and HPRT: forward primer, 5′-AGGGCATATCCAACAACAAAC-3′; reverse primer, 5′-GTCAACGGGGGACATAAAAG-3′.
A typical reaction involved an initial denaturation at 95 C for 1 min, followed by 45 cycles of 95 C for 0 sec, 60 C (HPRT)/62 C (RPS9)/64 C (GH and GHRH R) for 5 sec, and elongation at 72 C for 10–15 sec. Fluorescence was measured at 84 C for GH, GHRH R, and HPRT and at 82 C for RPS9. A standard melting curve cycle was used to check the quality of amplification and to confirm the absence of primer dimer formation. To ensure accuracy, each reaction was repeated two to five times per sample. The mean value of the repeat was used for each gene per sample to calculate the ratio between the mean value of the target gene and the internal control gene. The relative level of each gene was expressed by the ratio of the number of transcripts of the gene to an internal control gene.
Three coverslips per treatment group per animal with uniformly plated cells, immunolabeled for GH, were analyzed to measure the integrated optical density (IOD). The pituitary cells from all animals (n = 3–4) were immunolabeled using the same solutions and incubation periods to avoid any interexperimental variation. Before the image analysis program was applied, 12–15 high resolution (1600 × 2400) images/treatment group/animal were taken. The photographs from each treatment group (vehicle/vehicle, vehicle/DHEA, inhibitor/vehicle, or inhibitor/DHEA groups) were digitized with the use of a SPOT camera, the ×40 objective, and the same lighting and condenser settings. The fields were chosen systematically and included one in each of four quadrants to the upper and lower left and right of center. The camera field display in the microscope oculars was used to prevent overlap of fields. The entire groups of experiments were photographed in the same time period with identical settings. The BioQuant NovaPrime Image analysis system designed for use with Windows XP was used to detect area and the IOD of label on these photographs as previously described (26). One topographical field per photograph was analyzed (12–15 fields/treatment/animal). However, when the total cell areas measured per topographical field were averaged, the differences in total cell areas between the animals were not significant. Thus, the analysis of cell areas showed that any gains or losses in label area were not due to overall changes in cell area or population density. The sum of the IODs from the 12–15 fields from individual animals was averaged to get the final value for IOD for GH immunolabeling.
Serum samples from both vehicle- and DHEA-injected animals were sent to Linco Diagnostics, Inc. (St. Charles, MO), for estimation of serum GH using an RIA kit with a sensitivity range of 0.5 ng/ml GH.
Fields that were single labeled were counted under a ×40 objective to determine the percentage of GH-containing cells. The first 200 cells encountered in a random scan of the field on one coverslip were analyzed. There were three coverslips per experiment, and the experiment was repeated at least three times. Each experiment collected cells from at least one aged or diestrous rats, and the experiments were repeated so that a total of at least three rats were used per group. The experiment included three coverslips per animal, with 200 cells counted per coverslip. The values from each experiment were averaged. Each experiment was replicated three times, and the final n represents the average number of at least three replicates. Either Student’s t test or one-way ANOVA was used to detect significant differences (P < 0.05), depending upon the number of experimental groups. Fisher’s least significant differences test identified the groups that were different.
The first set of studies was designed to determine whether aging changed the number of cells with GH and GHRH R. In agreement with studies of humans, middle-aged rats showed a significant decrease in the percentage of GH-expressing cells compared with young diestrous rats. Figure 1A shows that the percentage of somatotropes in middle-aged rats was significantly lower [P = 0.01 (protein); P = 0.01 (mRNA)] than those in young female rats when detected by their protein (immunocytochemistry) or mRNA (in situ hybridization). The values for the young diestrous group were similar to those published in recent studies (24, 26); this period of the cycle represents a period of peak expression of GH proteins or mRNA.
The lower expression of GH in AP in aging rats may be associated with a change in the sensitivity of the pituitary to GHRH. To test this, we detected GHRH R mRNA levels by real-time QRT-PCR from pituitary lysates prepared with no intervening culture period. Figure 1B shows that GHRH R mRNA declined significantly (P = 0.02) in middle-aged rats, which correlates well with the losses in GH cells seen in the counts.
This first set of studies showed that in vitro exposure to DHEA in middle-aged rat pituitary cells (Fig. 2B) restored the percentage of GH cells to levels similar to those in young rats (Fig. 2A). These significant (P < 0.05) increases were seen with concentrations of 0.1 nM (P = 0.04), 1 nM (P = 0.009), and 10 nM DHEA (P = 0.006). In contrast, DHEA had no effect on the percentage of GH-expressing cells in young pituitary cultures, as shown in Fig. 2A, even when a full range of concentrations was tested.
Representative photographs of these fields are shown in Fig. 2, C–F. Figure 2C illustrates GH labeling in young diestrous rats. Figure 2D illustrates one of the controls showing the effect of GH absorption on labeling of fields from young rats with anti-GH; no labeling is evident. Omission of the anti-GH or antirabbit IgG resulted in no labeling (not shown). Figure 2E illustrates the decline in GH cells in fields from middle-aged rats. The DHEA-mediated increase in percentages of immunolabeled GH cells was evident when vehicle-treated cultures from aging rats (Fig. 2E) were compared with DHEA-treated cultures (Fig. 2F).
To determine whether the changes in cells expressing GH antigens reflected changes in the expression of GH mRNA, pituitary cultures were labeled for mRNA by in situ hybridization. There were no DHEA-mediated changes in the percentage of GH mRNA-bearing cells in young animals (Fig. 3A). In contrast, Fig. 3B shows that in middle-aged animals, DHEA stimulated an increase in the expression of GH mRNA, as detected by counts of labeled cells, restoring it to levels not different from those in young animals. The DHEA-mediated increase in the percentage of somatotropes detected by GH mRNA was significant (1 nM, P = 0.05; 10 nM, P = 0.02) compared with the vehicle controls. This profile of GH mRNA expression in individual somatotropes paralleled the expression seen when the population was detected by their protein content.
Figure 3, C–H, shows photographs of labeling for GH mRNA from young rats (Fig. 3C), including control fields treated with biotinylated sense GH mRNA (Fig. 3D) or a solution containing 100 ng/ml biotinylated antisense GH probe and ×83 unlabeled antisense probe (Fig. 3E). Labeling was evident only in the field exposed to biotinylated antisense probe. It was completely absent in the two control fields. Figure 3F illustrates the loss in labeling for GH mRNA in the middle-aged rat. Figure 3, G and H, also illustrates the DHEA-mediated increase in labeling for GH mRNA after exposure to 1 nM (Fig. 3G) or 10 nM (Fig. 3H) to levels seen in young animals (Fig. 3C).
GH gene expression in pituitary cells was also assayed by QRT-PCR to correlate with the in situ hybridization evidence. The assays showed that middle-aged rats had 30–60% of the GH mRNA levels found in young rats when normalized with either of two housekeeping genes, RPS9 (Fig. 4) or HPRT (data not shown). These losses compared favorably to those seen when GHRH R mRNA was assayed (Fig. 1B). When the cells were treated in vitro with 1 nM DHEA, GH mRNA levels were restored to levels similar to those in young rats (P = 0.03). In contrast, GH mRNA levels in the young diestrous rats did not change with DHEA treatment.
GHRH R mRNA was also assayed after 24-h treatment with DHEA or vehicle. There were, however, significant losses in the basal expression of GHRH R mRNA during the treatment period. Hence, studies of DHEA regulation of GHRH R mRNA were studied in vivo. Additional work is needed to achieve optimal culture conditions for retention of GHRH R mRNA during a 24-h treatment period.
The restorative effects of DHEA in vitro suggested that DHEA or its metabolites could have positive effects on the expression of GH mRNA or proteins. To learn whether these effects could be seen in vivo, 18-month-old animals were given a brief series of injections of DHEA. The pituitaries were then removed and processed for immunolabeling, affinity cytochemistry for biotinylated GHRH binding, and GH mRNA. Serum was collected for GH RIA.
GH RIA showed that serum GH (Linco Diagnostics, Inc) in response to a bolus of GHRH (1 mg/kg, sc) was 2-fold higher in rats that had received DHEA (P = 0.02) than in rats that had received vehicle only (Fig. 5). These data correlate well with the two additional assays.
The analysis of GH expression in this set of experiments was quantified by an image analysis system, IOD, that integrated changes in both the density of stores and the number of cells. When this analysis was applied to GH cells, DHEA treatment showed a significant increase in IOD (P = 0.02), as shown in Fig. 6. The in vivo study analyzed GHRH R with biotinylated analogs of GHRH and found that aging rats had about 8–10% AP cells expressing GHRH R compared with the 26–30% seen in young animals during most stages of the cycle (26). DHEA injections stimulated a significant increase (P = 0.03) in the percentage of cells with GHRH R to 20%, which is still lower than that normally seen in young rats (Fig. 7, A and B). However, DHEA in vivo did not stimulate the expression of GHRH R mRNAs (Fig. 7C). The GH mRNA detected by in situ hybridization and quantified by IOD (Fig. 8A) and QRT-PCR (Fig. 8B) also did not change with DHEA treatment in vivo.
The foregoing studies showed that DHEA may selectively increase the expression of GH in aging pituitary cells. However, additional work was needed to discover the mechanisms behind its actions. Before postulating that it acted via ERs, the in vitro studies were expanded to include a study of the effects of estrogen on aging rat pituitaries. The design was similar to that of the study recently reported (26).
Figure 9A showed that estrogen does have a potent restorative effect on GH cells for the pituitaries of aging rats detected by GH proteins in lower concentrations than those used for DHEA. These concentrations ranged from 0.01–1 nM [0.01 nM (P = 0.01), 0.1 nM (P = 0.006), and 1 nM (P = 0.008)]. Estrogen also restored the number of GH cells detected by GH mRNA in the aging pituitary in concentrations of 0.1, 1, and 10 nM (P = 0.004, P = 0.006, and P = 0.01, respectively) as shown in Fig. 9B.
To learn whether DHEA’s actions were mediated by estrogens, a series of inhibitors was used, which selectively blocked metabolite formation at each step in the pathway leading to estrogens. This study also focused on the analysis of changes in GH antigens quantified by IOD, which was a reliable and consistent measurement of responses to DHEA.
The first blocker was trilostane [3β-hydroxysteroid dehydrogenase (3β-HSD) enzyme blocker], which blocked additional metabolite formation and tested whether DHEA’s actions were mediated by its own receptor. Trilostane showed that DHEA’s actions were dependent on additional metabolite formation (Fig. 10A; P = 0.01). Aminoglutethemide (an aromatase inhibitor) was then used to learn whether the actions of DHEA were dependent on aromatization to estrogens. Figure 10A shows that aminoglutethemide completely blocked DHEA’s enhancing effects (P = 0.03) on GH expression.
After these studies showed that aromatization was needed for DHEA actions, the study concluded with tests of ICI 182,780 (an ER antagonist). Figure 10A shows that this inhibitor also completely blocked DHEA effects (P = 0.01), suggesting that the pathway for restoration of somatotropes required ERs. Parallel studies of this antagonist have shown that it also blocks estrogen’s enhancing actions on GH cells (data not shown). The inhibitors by themselves did not significantly affect GH expression by aging somatotropes (Fig. 10B). Representative fields of vehicle- and DHEA-treated cells with or without inhibitors are shown in Fig. 10C.
The first objective in this study was to learn whether the reduction in GH proteins (29) or mRNA (30, 31) reported by other workers in rats and humans was due to losses in GH antigen-bearing or GH mRNA-bearing cells. Our studies agree with those of previous workers who showed that pituitary GH protein levels decline to about 50% in middle-aged female rats (29). This decline in pituitary GH is associated with a decline in serum GH levels (29, 48), which, in turn, affects GH-dependent gene expression in other organs in the aging rat (48, 49) and eventually results in a decline in GH cell number by 20 months of age (32). The new findings in our study include the fact that this decline in GH cell number can be seen as early as 12–14 months of age. We also report a 50% drop in the expression of GHRH R mRNA and a 60–70% drop in target cells for GHRH R. Middle age in female rats corresponds to the perimenopausal age group in women, when earlier reports had shown a maximal loss of somatotropes in human pituitaries (27, 28).
After confirmation that reduced GH functions reported by others were caused in part by fewer GH antigen- or mRNA-bearing cells, the second objective in our study was to learn whether DHEA could reverse these aging effects by direct actions on pituitary cells. DHEA treatment in vitro significantly increased the percentages of immunolabeled soma-totropes only in aging animals and restored percentages of GH-bearing cells to levels seen in young animals.
Once the in vitro data showed restorative effects of DHEA on aging GH cells, the studies sought in vivo evidence that this adrenal hormone had similar effects. The objective was to learn whether a brief exposure period to DHEA would stimulate the aging GH cell population so that they would be better able to respond to a bolus of GHRH in vivo. The GHRH bolus was given because of the pulsatile nature of GHRH secretion and the recognition that endogenous pulses of GHRH could mask the results in either DHEA- or vehicle-treated animals. This study reports that DHEA administration for 2.5 d significantly enhanced serum GH (1 h after GHRH exposure), and this correlated well with the significantly increased expression of GH proteins and GHRH receptivity (detected by binding to biotinylated GHRH). Our in vivo studies agree with a recent report by Suarez et al. (50), who showed that DHEA potentiated GHRH-induced GH release (not basal GH release) by modulating cAMP production in cultured young rat AP cells.
DHEA did not enhance GH mRNA levels in vivo. This may reflect in vivo influences by other factors that affect GH cells, including GHRH, somatostatin, ghrelin, IGF-I, and sex steroids (including metabolites of DHEA) (36). The endogenous levels of the some of these factors, the expression of their specific receptors (51), and the sensitivity of the pituitary to them may also change during aging (52, 53). The interpretation of an in vivo study must recognize that DHEA treatment may also affect each of these above-mentioned factors either directly or indirectly.
The tests of DHEA actions on GHRH R expression were limited to the in vivo study. The reason for this was because of the rapid loss in transcripts with time in culture in these defined medium. In this context, it should be noted that tests of stability of the GH transcript in culture were made in parallel to those in which GHRH R mRNA was tested. There was an approximately 15% loss in GH transcripts detected during the first 3 h of culture, with no significant losses during the subsequent 24-h culture period.
Thus, the first sets of studies agreed that DHEA restored GH cell numbers in vivo and in vitro, especially if they were detected by immunolabeling. Under the experimental conditions of our study, DHEA did not affect the percentages of somatotropes from young diestrous animals in the wide dose range tested, which agrees with one in vivo study (54). Halmy et al. (54) reported that a 30-d implantation of 80 mg DHEA in mature female rats produced no change or a decrease in GH content.
However, two studies have reported enhancing effects of DHEA on the young pituitary population (43, 55). Suarez et al. (43) showed that either DHEA or estrogen given to young female rats in vivo caused hyperplasia in the pituitary. Simard et al. (55) also reported DHEA enhancement of GH protein expression. This could be explained by different times of treatment and also the fact that the diestrous rats in our study were already expressing the highest levels of GH normally seen during the cycle (24) and might be considered to be in an estrogen-rich environment. Simard et al. (55) used a group of mixed cycling female rats, which might include metestrous rats, which express about 50–70% of the GH proteins and mRNA seen in the other stages (24). Collectively, these studies suggest that DHEA may have a more potent enhancing effect on GH in rats when administered in a low estrogen state. The reason our study compared aging animals with diestrous animals was because diestrus is a stage when ER expression is maximal (56, 57). This hypothesis is supported by recent reports in humans that showed that responses to DHEA varied with the estrogenic environment (58). In this study DHEA enhanced GH secretion only in patients who were not receiving estrogen replacement therapy. Thus, it is possible that the effects of DHEA would be seen more often in estrogen-deprived conditions.
Once the in vivo and in vitro studies showed DHEA enhancement, we began tests designed to learn whether DHEA worked directly through its own receptor or after metabolism to estrogens. In the first of these studies, we tested estrogen directly on the expression of GH proteins and mRNA by aging pituitary cells. Our studies showed a potent restorative effect of estrogens on aging GH cells. This agrees with a previous study (55) in young female rats that showed that an antiestrogen, LY156758, prevented DHEA enhancement of GH protein expression in vitro after 72 h. Comparing DHEA and estrogen showed that the latter steroid was effective at lower doses. However, DHEA was effective at concentrations comparable to physiological circulating levels of DHEA in the rat (~1 nM) (59). The series of inhibitors tested supported the hypothesis that DHEA did not work directly on its receptor (44, 45) and that it must be metabolized to estrogens before it restores the expression of GH cell proteins. The enzymes necessary for DHEA conversion to downstream metabolites [3β-HSD (60) and 17β-HSD (61)] have been previously reported to be expressed in the pituitaries of young female rats. Furthermore, aromatase enzyme, which mediates the final conversion of any of the DHEA metabolites to estrogens, is expressed in both pituitary cells and endothelial cells of the pituitary of young (62) and aged rats (63).
The results of these studies of estrogenic effects correlate well with those of other studies that show estrogen involvement in GH gene expression. Aromatase knockout mice have decreased expression of GH, GHRH R, and Pit-1 mRNA, and estrogen replacement enhanced the expression of all of these genes (64). There is a coincident surge in GH along with the LH surge at ovulation in sheep, indicating that rising estrogen levels enhance GH secretion (65). Studies from our laboratory in young cycling female rats across the estrous cycle showed an increase in GH mRNA during diestrus and proestrus, when serum estrogen is rising (24).
Thus, collectively, our data suggest that DHEA metabolites are aromatized and act on ERs. However, there are reports suggesting that DHEA could act on ERs without conversion to estrogens. Nephew et al. (66) have shown that DHEA can directly bind to ERs in vivo (even inhibiting estrogen binding to its own receptor) and cause dimerization of ERs, which are functionally active in yeast. However, these actions required approximately 1000 times higher concentrations of DHEA than E2. Another study has shown direct binding by DHEA to ERβ in transfected human embryonic kidney 293 cells, which do not have the required converting enzymes for its bioconversion to estrogen (67). Other studies show direct binding of DHEA to ERα (68) and DHEA mediating ER-estrogen response element-dependent transcriptional activity, independent of its conversion to estradiol (69). Estrogen has been shown to be stimulatory on the 5′-promoter activity of GH gene using the MtT/S rat pure somatotrope cell line (70). Somatotropes are known to express both ERα and ERβ (57). Because ICI 182,780 blocks both ERα and ERβ, it cannot be inferred from this study whether one or both of these ER subtypes are involved in DHEA actions. Because our studies used mixed cultures of pituitary cells, the possibility of DHEA acting through ERs expressed in gonadotropes, lactotropes, or other cell types to mediate these changes in a paracrine fashion must also be considered.
Finally, these data do not rule out conversion to androgens, because DHEA-mediated restoration of somatotropes could occur with minimal conversion of DHEA to estrogen, and estrogen mediates the same effect at doses that are 1/10th or 1/100th the concentration of DHEA. Although androgen receptors have been shown to be expressed in rat AP (71), the role of androgens in GH expression in females has not been studied extensively. A 72-h incubation with dihydrotestosterone (0.01–50 nM) and testosterone (0.01–1000 nM) did not seem to affect spontaneous GH release in pituitary cultures from pituitaries of cycling female rats taken from random stages of the cycle (55). Some of the metabolites of DHEA, such as androstenediol (byproduct of DHEA by 17β-HSD), 3α-androstenediol, and 3β-androstenediol (derivatives from dihydrotestosterone), have weak estrogenic activity (72, 73). Thus, although we cannot rule out actions mediated by androgen receptors at this point, the total blocking effects of the aromatase inhibitor in our study strongly supports the hypothesis that the enhancing actions of DHEA on aging somatotropes are dependent on its aromatization to estrogen metabolites.
Circulating levels of DHEA are approximately 1 nM in rats (59). DHEA may be synthesized de novo in various regions of the brain as a neurosteroid. DHEA levels are three times higher in the hypothalamus than in the cortical regions of the brain (74), which might provide a route for DHEA-mediated regulation of the pituitary. It is interesting to note that the precursor for DHEA, pregnenolone sulfate, declines in the aging rodent brain (75), and it is not known whether this decline plays a role in the decline of GH gene expression in aging.
In summary, this is the first study in female rats to demonstrate a potential pathway within the aging pituitary by which DHEA or its estrogen metabolite may restore some of the expression of GH in aging somatotropes to levels seen during peak periods of the cycle in young animals. The results indicate that both estrogen and its precursor, DHEA, positively influence GH gene expression in aging. Collectively, this study shows a therapeutic potential for DHEA in the restoration of GH cell functions that might benefit frail elderly with low GH levels.
We acknowledge the Hormone Distribution Program, National Institutes of Health, and Dr. A. F. Parlow for the antiserum to rat GH. We thank Ms. Wendy Chang (Sanofi Synthelabo) for the generous gift of trilostane, and Dr. Kelly Mayo (Northwestern University, Chicago, IL) for the GHRH receptor clone. We appreciate the technical assistance of Ms. Evita Asumugha and Dr. Noor Akhter.
A section of this paper was presented in poster format at the 86th Annual Meeting of The Endocrine Society in 2004, and a second section was presented in a poster at the 87th Annual Meeting of The Endocrine Society in 2005. This presentation was also submitted in partial fulfillment for a Ph.D. degree, December 2004.
This work was supported by National Institutes of Health Grant R01-HD-33915-01 and the Committee for Allocation of Graduate Student Research Fund) from the Graduate School of Biomedical Sciences, University of Arkansas for Medical Sciences.