Brain angiotensin II (Ang II) inhibits pituitary prolactin release by an indirect mechanism requiring stimulation of dopamine formation and release. We report that [125I]Sar1–Ang II binding to AT1 receptors and AT1A receptor mRNA expression increase selectively in the dorsomedial arcuate nucleus of 17β-estradiol-primed ovariectomized rats after treatment with progesterone. In hormone-treated rats, arcuate nucleus AT1Areceptor mRNA expression is associated with tyrosine hydroxylase-positive neurons. No AT1A receptor mRNA was detected in tyrosine hydroxylase-positive cells of the arcuate nucleus of intact male rats. Conversely, in the anterior pituitary, where local or circulating Ang II stimulates prolactin release, [125I]Sar1–Ang II binding to AT1 receptors and AT1B receptor mRNA expression are decreased in 17β-estradiol/progesterone-treated ovariectomized rats.
Thus, AT1A receptors in the dorsal arcuate nucleus and AT1B receptors in the anterior pituitary are regulated inversely by estrogen/progesterone treatment, supporting the hypothesis of a dual role for brain and pituitary Ang II on prolactin release. The colocalization of AT1A receptor mRNA and tyrosine hydroxylase in neurons of the arcuate nucleus furthermore indicates that within this area central Ang II acts directly on dopaminergic neurons. These results support the hypothesis that central Ang II inhibits pituitary prolactin release indirectly via modulation of dopaminergic activity in the arcuate nucleus.
- angiotensin II receptors
- tyrosine hydroxylase
- in situ hybridization
- anterior pituitary
Angiotensin II (Ang II) is produced by many tissues, including the brain and pituitary gland, where the peptide plays a role in the control of reproductive hormones (Saavedra, 1992). In the anterior pituitary, Ang II is synthesized locally and stimulates prolactin release (Steele et al., 1981; Aguilera et al., 1982; Schramme and Denef, 1983; Ganong et al., 1989; Jones et al., 1990; Steele and Myers, 1990; Thomas and Sernia, 1990). Ang II also participates in brain mechanisms controlling pituitary prolactin release, but here its role is inhibitory rather than stimulatory (Myers and Steele, 1989, 1991).
There is evidence that the regulation of pituitary and brain Ang II receptors is important in the control of prolactin secretion and that the Ang II receptor expression in specific brain areas and in the anterior pituitary is under the control of reproductive hormones. Treatment of estrogen-primed ovariectomized (OVX) rats with progesterone upregulates Ang II receptors in selective brain areas involved in the inhibition of pituitary prolactin release, whereas estrogen treatment downregulates the expression of Ang II receptors in the anterior pituitary (Chen and Printz, 1983; Carriére et al., 1986; Seltzer et al., 1992, 1993).
Ang II receptors have been classified pharmacologically into two subtypes, AT1 and AT2 receptors (Timmermans et al., 1995). On the basis of quantitative autoradiography with displacement of [125I]Sar1–Ang II binding by AT1 but not AT2receptor-selective ligands, we reported the presence of AT1receptors in the anterior pituitary and in all brain areas related to the central control of pituitary function (Tsutsumi and Saavedra, 1991a,b). AT1 receptor subtypes have been subdivided further by the cloning of AT1A and AT1Breceptors, which are highly homologous and encoded by two distinct genes (Elton et al., 1992; Iwai and Inagami, 1992; Kakar et al., 1992;Sandberg et al., 1992). Because it is not possible to differentiate between AT1A and AT1B receptors by using binding studies (Timmermans et al., 1995), we subcloned fragments from the 3′ noncoding regions of AT1A, AT1B, and AT2 receptor cDNAs to produce receptor subtype-specific riboprobes (Jöhren et al., 1995b). No significant homology exists among the noncoding regions of Ang II receptor cDNAs (Iwai and Inagami, 1992; Inagami et al., 1994). In situ hybridization experiments using these probes showed that in adult rats the AT1A receptor subtype predominates in brain areas involved in pituitary function, whereas the AT1Breceptor subtype is expressed in the anterior pituitary (Jöhren et al., 1995b; Jöhren and Saavedra, 1996).
We asked the question of whether reproductive hormones controlled not only the expression of the Ang II receptor protein but also the expression of receptor mRNA. We studied this question with quantitative receptor autoradiography and in situ hybridization histochemistry in ovariectomized (OVX) rats treated with estrogen and progesterone. Because central Ang II has been proposed to inhibit prolactin secretion by stimulating the synthesis and/or release of dopamine (Steele, 1992), we attempted to clarify how Ang II receptors within the dorsal arcuate nucleus are associated with dopamine-producing neurons.
MATERIALS AND METHODS
Animals and tissue preparation. All animal procedures were approved by the National Institute of Mental Health Animal Care and Use Committee. Intact male (200–250 gm) and ovariectomized (OVX) female (150–200 gm) Sprague Dawley rats were obtained from Zivic-Miller (Zelienople, PA). Rats were provided with standard food and water ad libitum and kept under a 12 hr/12 hr light/dark cycle. Two weeks after surgery female rats were divided into two groups (OVX plus estradiol/progesterone replacement and OVX plus placebo). Hormones were given in the form of slow-release pellets designed to deliver a constant dose of hormone over at least 21 d (Innovative Research of America, Sarasota, FL). In the first group each rat received one pellet of 17β-estradiol (0.05 mg/pellet) at 10:00 A.M. (Day 0) placed subcutaneously in the interscapular region. This replacement dose of estrogen (0.05 mg/pellet) has been shown to result in 17β-estradiol levels of <50 pg/ml (Seltzer et al., 1992). In the second group each rat received one placebo pellet (Innovative Research of America). On day 2 at 10:00 A.M., each rat of the estradiol-treated group received one pellet of progesterone (Innovative Research of America; 50 mg/pellet, 21 d release, s.c.), and each rat of the placebo group received a second placebo pellet. The selected dose for 17β-estradiol and progesterone results in serum hormone levels within the physiological range (Butcher et al., 1974; Barron et al., 1986;Arbogast and Voogt, 1993; Brann et al., 1993; Michels et al., 1993). All pellets were implanted under ketamine anesthesia (150 mg/kg ketamine-HCl and 15 mg/kg azepromazine maleate, i.p.). Rats were decapitated between 3:00 and 4:00 P.M. on day 3. It has been shown before that this timing of hormone administration decreases plasma prolactin levels and increases arcuate nucleus dopamine turnover on the afternoon of day 3 (Rance et al., 1981).
Brains and pituitary glands were removed immediately, frozen by immersion in isopentane at −30°C, and stored at −80°C. For colocalization studies, rats were anesthetized with ketamine and perfused transcardially with 100 ml of saline, followed by 200 ml of 4% paraformaldehyde in PBS. Brains were removed, post-fixed for 24 hr in 4% paraformaldehyde/PBS at 4°C, and incubated overnight at 4°C in PBS containing 18% sucrose. Brains were frozen by immersion in isopentane at −30°C and stored at −80°C.
Consecutive 16-μm-thick coronal sections of brains and pituitaries were cut at −20°C in a cryostat. Alternate sections were collected for receptor binding and in situ hybridization. For receptor autoradiography, sections were thaw-mounted on gelatin-coated glass slides, dried overnight in a desiccator at 4°C, and stored at −80°C. For in situ hybridization and immunohistochemistry, sections were thaw-mounted on silanated glass slides (Digene Diagnostics, Beltsville, MD) and stored at −80°C.
Quantitative receptor autoradiography.Sar1–Ang II (Peninsula Laboratories, Belmont, CA) was iodinated by New England Nuclear (Boston, MA) to a specific activity of 2200 Ci/mmol. Receptor binding was performed as described (Tsutsumi and Saavedra, 1991a). Briefly, sections were preincubated for 15 min at 22°C in 10 mm sodium phosphate buffer, pH 7.4, containing 120 mm NaCl, 5 mmNa2EDTA, 0.005% bacitracin, and 0.2% proteinase-free BSA, and then incubated for 120 min in fresh buffer containing 5 × 10−10 m[125I]Sar1–Ang II (total binding). Nonspecific binding was determined by incubation in the presence of 5 × 10−6 m unlabeled Ang II. To characterize Ang II receptor subtypes, we incubated consecutive sections with [125I]Sar1–Ang II in the presence of 10−5 m losartan (DuPont Merck, Wilmington, DE), an AT1 receptor antagonist, or 10−7 m CGP 42112 (Neosystem SA, Strasbourg, France), an AT2 receptor-selective ligand. The number of AT1 receptors was determined as the specific binding displaced by unlabeled losartan. The number of AT2receptors was determined as the specific binding displaced by unlabeled CGP 42112. After incubation, slides were washed four times in ice-cold 50 mm Tris-HCl, pH 7.6, dried, and exposed for several days to Hyperfilm-3H (Amersham, Arlington Heights, IL). [125I]Sar1–Ang II binding was quantified by measuring optical densities on a Macintosh computer, using the public domain National Institutes of Health Image program (developed at National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/). Optical densities were transformed to corresponding values of femtomoles per milligram of protein (Nazarali et al., 1989).
In situ hybridization histochemistry. We used AT1A, AT1B, and AT2receptor subtype-specific riboprobes (Jöhren et al., 1995b, 1996;Jöhren and Saavedra, 1996). Therefore, fragments from the 3′ noncoding regions of the rat AT1A receptor cDNA (corresponding to nucleotides 1316–1684; Murphy et al., 1991), rat AT1B receptor cDNA (corresponding to nucleotides 1445–1841; Sandberg et al., 1992), and rat AT2 receptor cDNA (corresponding to nucleotides 1467–1838; Kambayashi et al., 1993) were subcloned into the polylinker site of the pBluescript II KS(+) vector (Stratagene, La Jolla, CA).
Antisense and sense (control) probes were labeled to a specific activity of 1.28 × 109 dpm/μg by in vitro transcription, using the TransProbe T Kit (Pharmacia Biotech, Piscataway, NJ). Transcription was performed according to the manufacturer’s protocol in the presence of 200 μCi of [35S]UTP (800 Ci/mmol, Amersham), 1 μg of linearized plasmid, and T3 or T7 RNA polymerase. Labeled riboprobes were separated from unincorporated [35S]UTP by centrifugation through Nick spin columns (Pharmacia).
In situ hybridization was performed as described previously (Jöhren et al., 1996). Sections of rat adrenal gland were included in the hybridization experiments as a positive control, because all antisense riboprobes (AT1A, AT1B, and AT2) used in this study have been shown to hybridize specifically to sections of rat adrenal gland (Jöhren et al., 1995b). Sections were fixed in 4% paraformaldehyde for 10 min, rinsed twice in PBS, acetylated for 10 min in 0.1 m triethanolamine HCl, pH 8.0, containing 0.25% acetic anhydride, dehydrated in alcohols, and air-dried. Each section was covered with 50 μl of hybridization buffer containing 50% formamide, 0.3 m NaCl, 2 mm EDTA, 20 mm Tris, pH 8.0, 1× Denhardt’s solution, 10% dextran sulfate, 100 μg/ml salmon sperm DNA, 250 μg/ml yeast RNA, 250 μg/ml yeast tRNA, 100 mm DTT, 0.1% SDS, and 10 ng/ml sense or antisense probe. After hybridization for 18 hr at 54°C, sections were rinsed several times in 4× SSC (1× SSC is 0.15m sodium chloride and 0.015 m sodium citrate). Nonhybridized probes were digested by incubation with 40 μg/ml RNase A (Sigma, St. Louis, MO) for 30 min. After a final high-stringency wash in 0.1× SSC at 65°C for 60 min, sections were dehydrated in alcohols containing 0.3 m ammonium acetate and air-dried.
Sections were exposed to Hyperfilm-3H (Amersham) for 8–14 d. Films were developed in D-19 developer (Eastman Kodak, Rochester, NY) for 4 min at 0°C and fixed in rapid fixer (Eastman Kodak) for 4 min at 22°C. Slides were dipped in Kodak NTB2 photo emulsion, exposed for 4–6 weeks, developed in Kodak D-19 developer for 3 min at 15°C, fixed for 4 min, and counterstained with toluidine blue. The location of silver grains was analyzed by microscopic examination of the sections. For semiquantitative analysis of placebo-treated OVX rats and estrogen/progesterone-treated OVX rats, silver grains were counted at high (1000×) magnification. A cell was considered positive if, after hybridization with AT1A, AT1B, and AT2 receptor antisense probes, it contained more than two times the number of silver grains than those present over cells after hybridization with nonspecific sense probes (background). The background counted in adjacent sections was less than four silver grains per cell. In the arcuate nucleus, silver grains were counted in all positive neurons present in each section, usually 10–20 neurons. In the anterior pituitary, microscopic fields were preselected randomly within each section at low (25×) magnification. Silver grains over all positive cells within the preselected field were counted at high magnification (1000×). All groups were coded for blind analysis. Comparison of mean values was performed by ANOVA, followed by the Tukey–Kramer multiple comparisons test.
Colocalization of AT1A receptor mRNA and tyrosine hydroxylase. For colocalization studies, in situhybridization for AT1A receptors was combined with immunohistochemical staining for tyrosine hydroxylase, the first and rate-limiting enzyme in the biosynthesis of catecholamine neurotransmitters. Sections of perfusion-fixed brains were washed for 5 min in 2× SSC, incubated for 20 min in 1 μg/ml proteinase K (Boehringer Mannheim, Indianapolis, IN), and post-fixed for 5 min in 4% paraformaldehyde. The labeling of probes and the hybridization procedure were identical to those described above. After the last washing step in 0.1× SSC, sections were processed for immunohistochemistry as described before (Jöhren et al., 1995a).
To detect tyrosine hydroxylase-positive cells, we used a monoclonal antibody (clone TH-2) purchased from Sigma. The primary antibody was diluted 1:1000 in Tris-buffered saline, pH 7.6, containing 0.1% BSA, and sections were incubated with the primary antibody for 60 min at 37°C. Positive staining was detected by the avidin–biotin complex (ABC) method (Hsu et al., 1981) with a biotinylated secondary antibody and peroxidase-conjugated streptavidin, using the DAKO LSAB Kit (DAKO, Carpinteria, CA) according to the manufacturer’s protocol. The chromagen was 3,3′-diaminobenzidine tetrahydrochloride (DAB, Sigma). As a negative control, sections were processed for immunostaining in the absence of the first antibody. In these control experiments no neuronal staining was observed in the arcuate nucleus (see Fig. 10 C). Furthermore, the specificity of the tyrosine hydroxylase antibody was evaluated by the specific localization of tyrosine hydroxylase-positive neurons in catecholaminergic areas like the A1, A2, A11, A12 (arcuate nucleus), and A13 (zona incerta) cell groups (zona incerta is shown in Fig. 9 A).
Regulation of [125I]Sar1–Ang II binding by estrogen/progesterone
Quantitative receptor autoradiography of sections from the hypothalamus and pituitary of placebo-treated OVX rats revealed specific binding of [125I]Sar1–Ang II in the median eminence, lateral hypothalamus (Fig.1), and anterior pituitary (Fig.2). In the dorsomedial arcuate nucleus of placebo-treated OVX rats, binding of [125I]Sar1–Ang II was very low and not detected in other ventral or lateral parts of the arcuate nucleus (Fig. 1 A). In all brain areas, as well as in the pituitary gland, specific [125I]Sar1–Ang II binding was totally displaced by the AT1 receptor antagonist losartan (Figs. 1 C, 2 C), but not by the AT2receptor ligand CGP 42112 (Figs. 1 E,2 E). Thus, all specific [125I]Sar1–Ang II binding can be assigned to AT1 receptors (Fig.3).
Estrogen/progesterone treatment of OVX rats produced a fivefold increase in [125I]Sar1–Ang II binding in the dorsomedial arcuate nucleus, when compared with placebo-treated OVX rats (Figs. 1 B, 3). As in placebo-treated OVX rats, all [125I]Sar1–Ang II binding in the dorsomedial arcuate nucleus of estrogen/progesterone-treated OVX rats was displaced by losartan (Fig. 1 D), but not by CGP 42112 (Fig. 1 F). No difference in [125I]Sar1–Ang II binding was detected in the median eminence and the lateral hypothalamus between placebo-treated OVX rats and estrogen/progesterone-treated OVX rats (Fig. 3). In the anterior pituitary, estrogen/progesterone treatment of OVX rats decreased [125I]Sar1–Ang II binding by 60% (Figs. 2 B, 3). Again, all [125I]Sar1–Ang II binding in the anterior pituitary of estrogen/progesterone-treated OVX rats was displaced by losartan (Fig. 2 D), but not by CGP 42112 (Fig. 2 F), and therefore represents binding to AT1 receptors (Fig. 3).
Regulation of AT1 receptor mRNA expression by estrogen/progesterone
To identify the AT1 receptor subtype (AT1Aor AT1B) involved and to address the question of whether estrogen/progesterone treatment affects AT1receptor mRNA levels, we performed in situ hybridization in sections taken from brain and pituitary adjacent to the ones used for the binding experiments. Sections of placebo-treated OVX rats and estrogen/progesterone-treated OVX rats were hybridized with AT1A and AT1B receptor antisense probes or with AT1A receptor sense probes that served as a negative control. In all brain areas studied and in the anterior pituitary, hybridization with AT1A receptor sense probes resulted in a low background signal (Figs.4 D,6 D). The background signal was usually less than four silver grains per cell (Fig. 5).
In the arcuate nucleus or the median eminence of placebo-treated OVX rats, the hybridization signal obtained with AT1A or AT1B receptor antisense probes was not significantly higher than the signal found after hybridization with nonspecific sense probes (Fig. 5). Significant AT1A receptor mRNA expression, however, was detected in scattered cells of the lateral hypothalamic area (Fig. 5). In the anterior pituitary we detected AT1Breceptor mRNA (Fig. 6 C) but no AT1A receptor mRNA (Fig. 6 B). Although present in control adrenal sections, no AT2 receptor mRNA expression was detected in the brain areas studied or in the anterior pituitary of placebo-treated OVX rats.
Marked alterations in the expression of both AT1A and AT1B receptor mRNA levels occurred after treatment of OVX rats with estrogen/progesterone. In the arcuate nucleus of estrogen/progesterone-treated OVX rats, in situhybridization with AT1A receptor antisense probes resulted in intense labeling of cells (Fig. 4 B), indicating the induction of AT1A receptor gene expression in this area (Fig. 5). Conversely, in the anterior pituitary, estrogen/progesterone treatment decreased AT1B receptor mRNA expression by 70% (Fig. 5). The expression of AT1A receptor mRNA in cells of the lateral hypothalamic area was not affected by hormone treatment (Fig. 5).
As it was the case in placebo-treated OVX rats, no AT1A receptor mRNA expression was detected in the median eminence of estrogen/progesterone-treated OVX rats, and no AT1B receptor mRNA was detected in any of the brain areas studied (Figs. 4 C, 5). In the anterior pituitary no significant AT1A receptor mRNA expression over background was detectable in hormone-treated OVX rats (Fig. 5). As in placebo-treated OVX rats, no AT2 receptor mRNA was detected in any brain area studied or in the pituitary of estrogen/progesterone-treated OVX rats.
Colocalization of AT1A receptor mRNA with tyrosine hydroxylase in the arcuate nucleus
To clarify whether the expression of AT1Areceptor mRNA found in cells of the arcuate nucleus of estrogen/progesterone-treated OVX rats was associated with dopaminergic neurons, we double-labeled sections by combining in situhybridization for AT1A receptors with immunological staining of tyrosine hydroxylase-expressing neurons. The localization of AT1A receptor mRNA in cells of the dorsomedial arcuate nucleus of estrogen/progesterone-treated OVX rats correlated well with the localization of tyrosine hydroxylase-immunoreactive neurons (Fig. 7 C,D). Although tyrosine hydroxylase-positive neurons were present in the arcuate nucleus of placebo-treated rats, no AT1A receptor mRNA was detected here (Fig. 7 A,B). Microscopic examination at high magnification of the arcuate nucleus of estrogen/progesterone-treated OVX rats revealed the cellular colocalization of AT1Areceptor mRNA and tyrosine hydroxylase (Fig.8). Most of the tyrosine hydroxylase-positive neurons in the arcuate nucleus expressed AT1A receptor mRNA. These neurons were characterized by the brown reaction product from the immunological staining in the cytoplasm and by an accumulation of silver grains surrounding and overlaying their nucleus (Fig. 8).
The association of AT1A receptor mRNA with tyrosine hydroxylase-positive neurons was selective for the arcuate nucleus because in the zona incerta of estrogen/progesterone-treated OVX rats tyrosine hydroxylase-positive neurons were present in the A13 cell group (Fig. 9 A), but no positive AT1A receptor hybridization signal was detected (Fig. 9 B). Conversely, in the lateral hypothalamic area, where AT1A receptor mRNA was found in scattered cells (Fig.9 D), no tyrosine hydroxylase-positive cells were detected (Fig. 9 C).
The regulation of pituitary prolactin release by Ang II has been proposed to be mediated by brain and pituitary mechanisms operating in opposing directions (Steele, 1992). In the brain, Ang II has indirect inhibitory control of prolactin release. Brain Ang II may function to limit the magnitude of the prolactin secretion under certain conditions, such as after concurrent estrogen/progesterone administration to OVX rats (Myers and Steele, 1989) or after stress in male rats (Myers and Steele, 1991). This indirect inhibitory effect of Ang II probably is attributable to its stimulation of dopamine release, the predominant inhibitor of prolactin secretion (Steele et al., 1982;Inoue and Negro-Vilar, 1989).
A likely site for the central inhibitory control of prolactin secretion by Ang II is the arcuate nucleus. The dorsomedial arcuate nucleus contains small tyrosine hydroxylase-immunopositive dopaminergic neurons (Everitt et al., 1992), and Ang II- immunoreactive nerve fibers are in close proximity to these neurons (Mounzih et al., 1994). The dorsal arcuate nucleus of the female rat expresses AT1, but not AT2, receptors (Seltzer et al., 1993). We hypothesized that it is via stimulation of these receptors that brain Ang II exerts its inhibitory effect on prolactin release by stimulating dopamine formation and its release into the portal circulation. Indeed, Ang II has been shown to selectively regulate dopamine levels in the arcuate nucleus (Steele et al., 1982) and to facilitate dopamine release from the hypothalamus (Inoue and Negro-Vilar, 1989). This hypothesis has been supported by the blockade of Ang II-induced inhibition of prolactin secretion by dopamine receptor antagonists (Steele et al., 1982).
We recently have provided additional support for this hypothesis, showing that the expression of AT1 receptors in the dorsomedial arcuate nucleus is high only during the estrus phase of the estrous cycle (Seltzer et al., 1993). Thus, a parallelism between the variations in portal plasma dopamine levels, which reach the highest levels at estrus (Ben-Jonathan et al., 1977), and the expression of AT1 receptors exists during the estrous cycle. We also have shown that estrogen/progesterone replacement in OVX rats results in a marked, and anatomically selective, upregulation of AT1receptors in the arcuate nucleus (Seltzer et al., 1993), a report confirmed here. The timing we used for the estrogen/progesterone treatment was the same as that described by Rance et al. (1981), who have shown that in the afternoon of day 3, 24 hr after progesterone pellet implantation, plasma prolactin concentrations were reduced significantly. The prolactin decrease was accompanied by an increase of dopamine turnover in the arcuate nucleus (Rance et al., 1981). Higher numbers of AT1 receptors could possibly magnify the effect of endogenous Ang II. This may explain the enhancement of the estrogen/progesterone-caused inhibition of prolactin release by intraventricular injection of Ang II (Myers and Steele, 1989). Progesterone increases the secretion of dopamine into the hypophyseal portal blood (Cramer et al., 1979), an effect that could be mediated via a regulation of the number of AT1 receptors in the dorsomedial arcuate nucleus. In support of this hypothesis is the presence of estrogen and progesterone receptors in dopaminergic neurons of the arcuate nucleus (Sar, 1984, 1988; Simerly et al., 1990; Kohama et al., 1992), the colocalization of estrogen and progesterone receptors in this area (Warembourg et al., 1989), and the regulation of progesterone receptors by estrogen (Romano et al., 1989).
Our present in situ hybridization study demonstrates that the increased binding of [125I]Sar1Ang II to AT1 receptors in the arcuate nucleus of estrogen/progesterone-treated OVX rats is paralleled with an expression of AT1A receptor mRNA. Because in the arcuate nucleus AT1A receptor mRNA is expressed in significant amounts only after hormonal treatment, our results suggest an induction of AT1A receptor gene expression in this area by reproductive hormones. Furthermore, our results indicate that the expression of AT1A receptor mRNA in the arcuate nucleus differs between intact male rats and female rats treated with estrogen/progesterone, because no receptor mRNA could be detected in intact male rats. In accordance with this observation, Myers and Steele (1991) found no effect of centrally injected Ang II receptor antagonists on prolactin levels in male rats. Because this is also the case in untreated OVX rats (Myers and Steele, 1989), it is possible that endogenous Ang II affects prolactin release only under certain conditions like estrogen treatment. Indeed, during restraint stress in male rats, when prolactin levels are increased, blocking of central Ang II receptors further arguments the prolactin increase (Myers and Steele, 1991). Thus, central Ang II may function to limit the magnitude of stress-induced prolactin response (Steele, 1992). It is conceivable that stress as well as estrogen/progesterone treatment induces the expression of arcuate nucleus AT1 receptors in male rats.
The hypothesis of a direct regulation of dopamine formation and release by Ang II prompted us to determine whether AT1A receptors were associated with the dopamine-forming, tyrosine hydroxylase-containing neurons. Our present results show that this is indeed the case, because we were able to selectively colocalize AT1A receptor mRNA and tyrosine hydroxylase immunoreactivity in neurons of the arcuate nucleus. It has been shown recently that Ang II can increase tyrosine hydroxylase activity and tyrosine hydroxylase mRNA levels (Yang et al., 1996; Yu et al., 1996). Our present data demonstrate for the first time that tyrosine hydroxylase-producing neurons can express AT1A receptors and support the hypothesis of the involvement of Ang II in the regulation of neuronal catecholamine synthesis. On the basis of our observations, we propose that Ang II acts on AT1A receptors produced by dopaminergic neurons of the arcuate nucleus to stimulate directly dopamine formation and/or its release to the portal circulation.
In the anterior pituitary, Ang II is produced locally in gonadotrophs and acts in a paracrine manner on lactotrophs to stimulate prolactin release (Aguilera et al., 1982; Paglin et al., 1984; Ganong et al., 1989; Thomas and Sernia, 1990), an effect mediated by AT1receptors (Becú-Villalobos et al., 1994; Moreau et al., 1994). Estrogen treatment of OVX rats downregulates the number of AT1 receptors in the anterior pituitary (Chen and Printz, 1983; Carriére et al., 1986; Seltzer et al., 1992). Our present results indicate that a combined estrogen and progesterone treatment of OVX rats also results in a downregulation of AT1 receptors in the anterior pituitary. In addition, our present in situhybridization study shows that estrogen/progesterone replacement in OVX rats decreases the expression of AT1B receptor mRNA and confirms earlier reports on the presence of AT1B, but not AT1A or AT2, receptor mRNA in the anterior pituitary gland of the rat (Kakar et al., 1992;Jöhren and Saavedra, 1996).
In conclusion, we demonstrate for the first time that the expression of the AT1A receptor mRNA is induced in dopaminergic neurons of the arcuate nucleus in OVX rats after estrogen/progesterone treatment. Thus, the effect of reproductive hormones on receptor expression is likely to represent transcriptional events rather than changes in the receptor turnover or availability for binding. Our finding that both brain AT1A receptor and pituitary AT1B receptor expression is inversely regulated by reproductive hormones confirms the hypothesis of a dual role of Ang II on the regulation of prolactin release: brain Ang II inhibits and pituitary Ang II stimulates the release of the hormone. Such regulation can be considered part of delicate feedback mechanisms that fine-tune the hormonal responses and their central regulations in conditions of health and disease.
G.L.S. was supported by Grant 200794/94 from the National Council of Scientific and Technological Development of Brazil. We thank Ricardo Dreyfuss, Medical Arts and Photographic Branch at National Institutes of Health, for the expert microphotography, and we thank the National Institutes of Health Scientific Computer Resource Center for image processing.
Correspondence should be addressed to Dr. Olaf Jöhren, Section on Pharmacology, National Institute of Mental Health, 10 Center Drive, MSC 1514, Building 10, Room 2D-57, Bethesda, MD 20892.