Local Transformations of Androgens into Estradiol by Aromatase P450 Is Involved in the Regulation of Prolactin and the Proliferation of Pituitary Prolactin-Positive Cells

In previous studies we demonstrated the immunohistochemical expression of aromatase in pituitary cells. In order to determine whether pituitary aromatase is involved in the paracrine regulation of prolactin-producing pituitary cells and the physiological relevance of pituitary aromatase in the control of these cells, an in vivo and in vitro immunocytochemical and morphometric study of prolactin-positive pituitary cells was carried out on the pituitary glands of adult male rats treated with the aromatase antagonist fadrozole. Moreover, we analyzed the expression of mRNA for the enzyme in pituitary cells of male adult rats by in situ hybridization. The aromatase-mRNA was seen to be located in the cytoplasm of 41% of pituitary cells and was well correlated with the immunocytochemical staining. After in vivo treatment with fadrozole, the size (cellular and nuclear areas) of prolactin cells, as well as the percentage of prolactin-positive cells and the percentage of proliferating-prolactin cells, was significantly decreased. Moreover, fadrozole decreased serum prolactin levels. In vitro, treatment with fadrozole plus testosterone induced similar effects on prolactin-positive cells, inhibiting their cellular proliferation. Our results suggest that under physiological conditions aromatase P450 exerts a relevant control over male pituitary prolactin-cells, probably transforming testosterone to estradiol in the pituitary gland.


Introduction
Of the two pathways through which androgens are metabolized -reduction and aromatization-the latter depends on the presence of an enzyme (aromatase P450) belonging to the family of cytochrome P450.
There is growing awareness that androgens and estrogens play general metabolic roles that are not directly involved in reproductive processes. Estrogen is no longer considered solely an endocrine factor, but instead is produced in a number of extragonadal sites and acts locally at these sites in a paracrine and autocrine fashion. These sites include breast, bone, vasculature, and brain. Within these sites, the action of aromatase can generate high levels of estradiol locally, without significantly affecting circulating levels [1,2].
At our laboratory, we have demonstrated the expression of aromatase in the rat pituitary gland. This expression was related to the sex and age of the animals. Moreover, we reported a very strong correlation between the pituitary expression of aromatase and the incidence of spontaneous prolactinomas in old rats [3] and humans [4]. Similar results were later reported by different authors in non-tumoral pituitaries of humans and different animal species [5][6][7][8][9][10][11] and in human prolactinomas [12]. However, the physiological relevance of aromatase and the possible aromatization of testosterone to estradiol in the regulation of pituitary hormones are not well known.
Because estradiol is an important physiological regulator of pituitary prolactin secretion [13][14][15][16][17][18][19] and is involved in the control and proliferation of prolactin-producing cells [20], the aim of the present study was to analyze the relevance of the regulatory role of the local transformation of androgens into estradiol by aromatase in the maintenance of the population of prolactin cells in the male pituitary gland, through the pituitary aromatization of testosterone to estradiol. For this purpose, the in vivo or in vitro effects of aromatase on the pituitary glands of adult male rats were blocked by treatment with the aromatase antagonist fadrozole.

Animals
Animal experimentation was performed according to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (NIH Publication No. 85-23, revised 1996). All procedures were approved by the Committee for the Care and Use of Animals of the University of Salamanca, which ensures compliance with national and European legislation regarding the use of animals in research (Spanish RD  During the experiments, all groups were kept under standard stabling conditions (temperature 2162uC, relative humidity 5065%, controlled photoperiod of 14 h light/10 h darkness, food and water ad libitum with a balanced rat/mouse maintenance diet (Panlab).

Sample collection and processing
Animals were sacrificed between 10.00 and 11.00 h by decapitation after anaesthesia by isofluorane inhalation. The pituitary glands were carefully dissected out and immediately fixed in a solution of 15% saturated picric acid in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, for 24 h. Then, they were dehydrated in ethanol, cleared with xylene, and embedded in paraffin in order to obtain coronal serial sections of 5 mm thickness. These were placed on slides treated with gelatin-chrome alum and were then used for the immunohistochemical study. After sacrifice, blood samples were obtained to determine serum prolactin levels.

Pituitary cultures
Following anaesthesia with isoflurane, 5 male Wistar rats (175-200 g) were killed by decapitation and the anterior pituitary glands were removed and washed in Earle's balanced salt solution. Enzymatic dispersion was carried out by incubation for 15 minutes at 37uC in Hank's solution to which 0.15% MgCl 2 , 0.1% papain, 0.01% DNase and 0.1% of neutral protease had been added. Mechanical dispersion was achieved by passing the pituitaries through Pasteur pipettes and 20 to 22 gauge needles. After centrifugation, the supernatant was removed and the cells were suspended in an appropriate volume of Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% calf serum, 2.5% foetal calf serum, 2% L-glutamine, 1000 IU/ml of penicillin and 1000 IU/ml of streptomycin. The cells were seeded on 60 culture chamber slides (Nunc, 1 ml) at a final concentration of 2.5610 5 cells/ml and incubated at 37uC in a 5% CO 2 /95% air atmosphere for 7 days. On the 4th day of incubation the medium was changed by fresh medium. On the 7th day, cultures were treated with 10 26 M testosterone, 10 26 M Fadrozole and incubated for 1, 3, 6 or 12 hours (five chambers per treatment and time-point assayed). Previously, testosterone was diluted in ethanol, the final concentration of ethanol in the treated and control dishes being 0.0027%.
After treatment, the chambers were carefully washed with Dubelcco's sterile PBS and then the cells were fixed in Somogy's solution for 1 hour, followed by careful rinsing in PBS. In order to validate the results obtained, two pituitary cultures were made under identical experimental conditions.

Serum prolactin determination
From serum samples, prolactin was assayed by double antibody RIA using the Dipesarat hormone kit according to the manufacturer's instructions. The intra-and inter-assay coefficients of variation were lesser than 3% and 6% respectively.

Western blotting
For Western blotting, the pituitary glands were dissected from adult rats and immediately frozen. Tissues were then disrupted by homogenization in lysis buffer (137 mM NaCl, 10 mM TB, pH 7.4, 10% glycerol and 1% Triton X-100, containing a cocktail of protease inhibitors). Insoluble material was removed from the lysates by centrifugation at 10,000 rpm for 10 min. Protein concentrations were determined using the standard Bradford assay. Fifty mg of total protein from each rat sample was separated by 10% SDS-PAGE. Following electrophoresis, proteins were transferred to nitrocellulose and then blocked for 1 h with 5% non-fat dry milk in PBS. The nitrocellulose membranes were then incubated overnight with either preabsorbed anti-aromatase serum or anti rat aromatase P450 rabbit polyclonal serum (Rb-SG 872 Sigma), diluted 1:500. Blots were subjected to 3615 min washes with PBS and then incubated for 1 h with HRP-labelled secondary antibodies (1:10,000 in PBS). Following extensive washing, the blots were revealed by Amersham ECL western blotting detection reagents (Amersham). The average exposure time was 2 min. Western blotting revealed the existence of a protein with a molecular weight of around 50 kDa in the pituitary gland ( Figure 1a).
Single immunostaining was carried out for aromatase and prolactin: Primary antibodies were incubated with tissue sections at 4uC overnight. After washing, the slides were incubated for 45 min at room temperature with biotinylated-goat anti-rabbit or mouse IgG (Caltag, diluted at 1:100 in TBS) and then for 45 min at room temperature with streptavidin-peroxidase complex (Caltag, diluted at 1:150). The reaction was developed with freshly prepared 3-39-diaminobenzidine (Sigma, 0.024% in TB buffer plus 0.03% H 2 O 2 ). The washes and antibody dilutions were made in TBS: HCl-Trizma, 0.05 M, pH 7.4, plus 0.8% NaCl.
To study PCNA-positive cells and to determine the PCNA-Prolactin labelling index, a double labelling immunohistochemical method for PCNA and prolactin was developed. Endogenous peroxidase was blocked with H 2 O 2 in methanol and non-specific reactions of the secondary antibody by incubation in normal goat serum (Dako, diluted 1:30). Sections were incubated overnight at 4uC with mouse anti-PCNA monoclonal antibody (PC10, Dako, diluted 1:3000 in TBS). Biotinylated goat anti-mouse IgG (Dako, diluted 1:100) and Avidin-Biotinylated horseradish peroxidase complex (ABC kit, Dako, diluted 1:100) were successively applied at room temperature for 40 min and 30 min, respectively. The reaction was developed in freshly prepared 3,39-DAB (0.025% in TB buffer containing 0.03% of H 2 O 2 ). Following PCNA immunolabelling, the peroxidase-antiperoxidase (PAP) reaction was performed for the detection of prolactin, The reaction was allowed to progress by incubation for 40 min at room temperature with swine anti-rabbit IgG (Dako, diluted 1:100), and later with incubation with soluble peroxidase anti-peroxidase complex (Dako diluted 1:100) for 35 min at room temperature. The reaction was developed in freshly prepared 4-chloro-1-naphthol (1.7610 23 M in 3% absolute ethanol and TB-buffer containing 0.3% H 2 O 2 ).
Preabsorption tests with prolactin and tests replacing the specific serum by normal non-immune rabbit serum abolished the reaction. Using ELISA, the specificity of swine anti-rabbit IgG was lower than 1% for rat and mouse IgG and 100% for rabbit IgG. For the washes and dilutions of the sera, TB buffer (0.05 M, pH7.4) containing 0.8% NaCl was used. The reaction was developed in freshly prepared 4-chloro-1-naphthol (1.7610 23 M in 3% absolute ethanol and TB-buffer containing 0.3% H 2 O 2 ).

In situ hybridization
These studies were performed using a non-isotopic method involving the immunocytochemical detection of biotin using the streptavidin-biotin-peroxidase method. To accomplish in situ hybridization, the sense biotinylated oligonucleotide 59BIO-gag gat gac gtg att gac ggc tac ccg gtt aaa aag gga act aac atc att ctg aac atc gga, and antisense 59 Bio-ctc cta ctg cac taa ctg ccg atg ggc caa ttt ttc cct tga ttg tag taa gac ttg tag cca, 100% specific to rat aromatase P450 according to the GenBank data base (accession number, M33986), was used as probes [21].
Slides were prehybridized in Omnibuffer for 30 minutes at 37uC. Hybridization with the biotinylated-probe (100 pg/ml in Omnibuffer) was carried out using a Hybaid thermocycler overnight at 37uC. The reaction was stopped by washes in 1 x SSC at 54uC for 20 min, 1 x SSC at room temperature for 20 minutes, and for 20 minutes in PBS (0.01 M, pH 7.4, plus 0.8% NaCl). Biotin was detected using monoclonal anti-biotin antibodies (Roche, 1:250 in TBS: 0.05 M HCl-Trizma, pH 7.4, plus 0.8% NaCl) overnight at 4uC in a humidity chamber, followed by biotinylated goat anti-mouse (Caltag, 1:250 in TBS). The reaction was amplified using the tyramide amplification kit (Dako) according to the instructions of the manufacturers. The final reaction was developed with 3,39-diaminobenzidine (0.025 M, Sigma, in 0.05 M TB-HCl buffer, pH 7.4) to which 0.03% H 2 O 2 had been added. The slides were counterstained using Mayer's acid haematoxylin.
As controls, hybridization with sense probe, omission of the probe and pretreatment with RNase were performed, no reaction being observed in any case (Figures 1e and 1f). Quantification of aromatase mRNA-expressing cells The percentage of aromatase mRNA-expressing cells was quantified in each animal following the double-blind procedure. Briefly, eight thousand cells (with intact cellular and nuclear profiles) were counted from 20 sections, separated from one another by at least 50 mm (400 cells/section), chosen randomly from all parts of the gland, after which the percentage of reactive cells was calculated.

Morphometry
Using the Image J software (NIH), 500 randomly chosen prolactin-positive cells per animal, or 100 randomly chosen prolactin-positive cells per dish were measured for morphometric study (cells were considered when the nuclear and cellular profiles in the plane of section or focus were clearly distinguished). Cellular and nuclear areas were determined.

Statistical analyses
For each parameter evaluated, the values obtained, after each treatment, were processed statistically using the SPSS software and the differences observed among mean values were compared using analysis of variance, accepting p values of ,0.05 as significant for the Scheffé F test. Results are expressed as arithmetic means 6 standard error of the mean.

Aromatase P450 expression in the pituitary gland of male adult rats
Western blotting. The image in Figure 1 shows that the hypophysis of the adult male rat contains a protein of 52 kDa, which is detected with the Rb-872 antibody, specific for rat aromatase P450. The immunoblot did not appear after preabsorption (Figure 1a).
Inmunohistochemistry. As shown in Figure 1c, in many hypophyseal glandular cells of adult male rats the presence of the enzyme was detected by means of immunohistochemistry.
In situ hybridization. Figure 1d shows that in many hypophyseal cells of adult male rats it is possible to detect the presence of aromatase mRNA by in situ hybridization (controls: absence of the probe - Fig. 1e, and sense aromatase Fig. 1f).

In vivo effects of fadrozole
Prolactin levels. The untreated and control (saline-treated) animals showed similar basal prolactin serum levels (see Figure 2d). Treatment with fadrozole led to a clear and significant (p,0.01) decrease in serum prolactin levels, and significant differences were found among animals treated with one dose of fadrozole and those treated with 5 doses (p,0.05).
Morphological findings in prolactin-positive cells. The prolactin cells of the untreated animals were found scattered throughout the pituitary gland, with a predominance of polygonal and oval shapes, with scant but very reactive cytoplasm and a central nucleus. There were also cells with a polarized nucleus. (Figure 2a and arrows in Figure 2b). In both groups of treated animals (1 and 5 doses of fadrozole) the intensity of the cytoplasmic reaction of prolactin-positive cells was lower than in the untreated or control animals (arrows in Figure 2c Percentage of prolactin-proliferating cells. Prolactin-proliferating cells appeared with brown-stained nuclei and a dark blue-stained cytoplasm (Figures 3a, 3b). The percentage of prolactin-proliferating cells (Figure 3d

In vitro effects of fadrozole
Morphometric findings in prolactin-positive cells. No statistically significant variations in the cellular area of prolactinpositive cells were found following treatment with fadrozole or testosterone alone (Figure 4d). Treatment with fadrozole plus testosterone significantly decreased the size of the prolactinpositive cells with respect to the other treatment groups analyzed. This effect was statistically significant from 3 to 12 hours of treatment (p,0.01).
Percentage of prolactin-positive cells. The variations observed among the control cultures ( Figure 4a) and those treated with fadrozole or testosterone alone (Figure 4b) did not reach statistical significance (Figure 4e). In contrast, treatment with fadrozole plus testosterone (Figure 4c) significantly decreased the percentage of such cells as from 3 hours (p,0.01. Figure 4e).
Percentage of prolactin-proliferating cells. No significant differences were observed between the control cultures and those treated with fadrozole or testosterone alone (Figure 4f). The treatment with fadrozole plus testosterone (Figure 4c) significantly decreased the percentage of proliferating prolactin-positive cells as from the first hour of treatment (p,0.01. Figure 4f).

Discussion
Previous findings from our laboratory suggest an important role of aromatase in the regulation of pituitary hormones, because we observed that aromatase P450 is expressed in the rat pituitary gland; that this expression is modified by sex and age, and that it could be involved in the genesis of spontaneous prolactinomas in old rats [3] or humans [4][5][6][7][8][9][10][11][12]. Moreover, aromatase mRNA has been reported in pituitary cells using in situ hybridization in trout [5]. These findings have been confirmed by other laboratories, in the sense that aromatase is expressed in the pituitary gland of different animal species, including humans [6,10,22], and that variations in its expression are regulated by gonadal steroids [6] because testosterone increased the expression but estradiol decreased it.
However, the role of aromatase in the regulation of pituitary hormones or cells is not well established, although several lines of evidence suggest the possibility that estrogens generated locally by aromatization could be involved in the activation of anterior pituitary mitotic activity [23]. Although androgens have an inhibitory effect on prolactin, the percentage of prolactinproducing cells in the pituitary glands of adult male rats is high [24]. Because the expression of pituitary aromatase is higher in males than in females [25], the local aromatization of testosterone to estradiol in the pituitary of males could be involved in the regulation of the population of prolactin-producing cells. The findings observed in the present study seem to support such a notion.
Androgenic effects on prolactin cells and secretion of the hormone are different, depending on whether the androgen is susceptible to its aromatization. Non-aromatizable androgens are inhibitors or hardly exert effects on prolactin producing pituitary cells, while aromatizable androgens increase the release of prolactin [26] and the levels of prolactin mRNA [27]. These functional changes obtained with doses of testosterone similar to those used in our study (10 26   changes in cell growth [26]. These findings are consistent with those observed in our study, in which testosterone did not significantly modify the size or proliferation of prolactin-positive cells.
The in vitro inhibitory effects observed following treatment with fadrozole and testosterone jointly could be explained in terms of the androgenic effects of testosterone or because testosterone could be metabolized into non-aromatizable androgens.
Fadrozole is a potent non-steroidal inhibitor of aromatase P450, and it has been shown that a dose of 4 mg/Kg induces uterine atrophy without inducing adrenal hypertrophy [28]. In our in vivo study, we used a lower dose, within the ranges tested by other authors [29,30] in a similar protocol to that described in [31], although slightly higher than the one used by these authors, with the aim of obtaining security inhibition rates higher than 90% [28], but not to induce atrophy of the organs of the reproductive system.
Estradiol is an important regulator of pituitary prolactin. It is well documented in the literature that estradiol, through coupling of the estrogen-estrogen receptor complex to the specific response element to estrogens in the promoter of the gene encoding prolactin synthesis, induces increases in prolactin synthesis through a biphasic transcription that culminates in an increase in the prolactin mRNA; this has a direct action on the gene able to alter the transcription [32][33][34][35][36][37][38][39][40][41][42][43]. Moreover, estradiol stimulates the secretion of prolactin [14,19,[44][45][46].
In the present study, after in vivo treatment with the aromatase antagonist fadrozole, highly significant changes were observed in pituitary prolactin-positive cells: decreases in the intensity of the immunohistochemical cytoplasmic reaction, decreases in the percentage and proliferation of prolactin-positive cells, and decreases in the cellular and nuclear areas, accompanied by lower levels of serum prolactin. All these findings suggest a lower activity in prolactin cells of fadrozole-treated animals than in control or untreated animals. Similar findings were observed when the effect of fadrozole on prolactin-producing cellswas compared in vitro after testosterone treatment.
Differences in the intensity of the cytoplasmic reaction could be explained in terms of decreases in the intracellular amount of prolactin, as has been previously documented at our laboratory for pituitary prolactin-producing cells, and this is supported by the decreases in serum prolactin levels found in the present study.
Decreases in cellular and nuclear areas support the notion that the cellular activity and/or cellular proliferation of prolactin cells and hormonal secretion is decreased, as demonstrated in previous studies [59].
As may be deduced from the results of this study, and in accordance with previous studies, aromatase is clearly expressed in the pituitary of male rats, and pituitary aromatase is probably expressed in different endocrine cells, including prolactin-producing cells [6]. This observation is very relevant because in the male pituitary aromatizable androgens can be metabolized to estrogens by means of aromatase; thus, estrogens can modulate and regulate the population of prolactin-producing cells, accounting for the important percentages of these cells in the male pituitary [24].
Because estradiol induces effects opposite those of fadrozole on prolactin cells, and since the aromatization of testosterone to estradiol is mediated by aromatase, the effects of fadrozole observed in our study suggest that aromatase would be involved physiologically in the regulation of the pituitary prolactinproducing cells of male rats, transforming testosterone to estradiol at pituitary level.