Sex-specific Occurrence of Androgen Receptors in Rat Brain*

The metabolism and binding of [1,2,6,7-3H]testosterone in male and female rat brain has been studied in an attempt to find an explanation for the relative androgen unresponsiveness characterizing the female hypothalamo-pituitary axis involved in regulation of hepatic steroid metabolism. The most significant sex differences in the pattern of [SH]testosterone metabolites recovered from several brain regions (including pituitary, pineal gland, and hypothalamus) after intraperitoneal administration of [‘HItestosterone were the predominance of testosterone and androstenedione in male brain compared to the quantitative importance of 5u-androstane-3a,l7fi-diol, 5a-androstane-3P,17P-diol, epitestosterone, and dihydroepitestosterone in female brain. One possible explanation for the androgen unresponsiveness of female rats is, therefore, the faster metabolism of testosterone to inactive compounds in female brain. Experiments both in uiuo and in uitro showed the presence of high affinity, low capacity binding sites for [SH]testosterone in male pituitary, pineal gland, and hypothalamus (& values in the region of 1 x lo-“’ to 1 x 1O-9 M and number of binding sites 1.0 to 1.4 x 10-l’ mol per mg of protein). The

From the Department of Germfree Research and Department of Chemistry, Karolinska Institutet, S-104 01 Stockholm 60, Sweden The metabolism and binding of [1,2,6,7-3H]testosterone in male and female rat brain has been studied in an attempt to find an explanation for the relative androgen unresponsiveness characterizing the female hypothalamo-pituitary axis involved in regulation of hepatic steroid metabolism. The most significant sex differences in the pattern of [SH]testosterone metabolites recovered from several brain regions (including pituitary, pineal gland, and hypothalamus) after intraperitoneal administration of ['HItestosterone were the predominance of testosterone and androstenedione in male brain compared to the quantitative importance of 5u-androstane-3a,l7fi-diol, 5a-androstane-3P,17P-diol, epitestosterone, and dihydroepitestosterone in female brain. One possible explanation for the androgen unresponsiveness of female rats is, therefore, the faster metabolism of testosterone to inactive compounds in female brain.
Experiments both in uiuo and in uitro showed the presence of high affinity, low capacity binding sites for [SH]testosterone in male pituitary, pineal gland, and hypothalamus (& values in the region of 1 x lo-"' to 1 x 1O-9 M and number of binding sites 1.0 to 1.4 x 10-l' mol per mg of protein). The steroid.macromolecular complexes generally had a p1 of 5.1, were excluded from Sephadex G-200, were heat-labile, and were sensitive to protease. Competition experiments indicated the following order of ligand affinities: testosterone > 5a-dihydrotestosterone and estradiol > androstenedione >> corticosterone. No steroid-binding proteins of similar nature were found in pituitary, pineal gland, or hypothalamus from female rats. On the basis of these results it is suggested that the androgen unresponsiveness of female rats referred to above relates to the absence of receptor protein for androgens in female rat brain. In support of this hypothesis, 2%day-old female rats, which are known to be affected by androgens with regard to liver enzyme activities, were shown to contain receptor proteins for androgen in the brain.
In conclusion, the relative androgen unresponsiveness of the female hypothalamo-pituitary axis is probably explained by the absence of receptor proteins for androgen in female hypothalamus and pituitary. The fast metabolism of testosterone in female rat brain also serves to decrease the availability of active androgen to potential receptor sites. It may be speculated that the presence of androgen receptors in male brain is the result of neonatal programming ("imprinting") by testicular androgen.
Administration of androgen to castrated rats results in significant changes in activities of hepatic enzymes active on steroid hormones (1, 2). In these experiments, male rats respond much better to androgen than female rats, and it has been shown that the greater androgen responsiveness in male rats is due to neonatal imprinting by androgens (3). Recently, results in our own laboratory have shown that an intact hypothalamo-pituitary axis is necessary for androgen regulation of hepatic enzyme activities (4, 5), and it seems reasonable to believe that androgen action on liver enzymes is mediated via central mechanisms. In view of this, it appears adequate to search for the explanation for the relative androgen unresponsiveness of female rats, as regards liver enzymes, in a sex difference in androgen action in brain. Information on this point in literature is relatively scanty and confusing. Whereas *This work was supported by grants from the Swedish Medical Research Council (03X-2819) and LEO Research Foundation. some authors deny the existence of specific binding sites for androgen in rat brain (6-8), other groups have described the occurrence of androgen receptors in pituitary (g-11), hypothalamus (9, ll), pineal gland (12), and cortex (11) of male rats. Furthermore, diverging opinions exist on the nature of the active metabolites of testosterone in rat brain, and both 5a-dihydrotestosterone (10,11,13) and estrogens (14) have been suggested as mediators of androgen action.
Considering the present confusion in the field of mechanism of action of androgens in the brain we have undertaken a combined in vivo and in vitro study aimed at finding an explanation for the relative androgen unresponsiveness in female rats. In view of the uncertainty as to the nature of the androgen metabolites present in brain tissue we have characterized the metabolites of intraperitoneally administered [1,2,6,7-SH]testosterone in several brain regions showing androgen uptake. Furthermore, isoelectric focusing has been used as a sensitive method to detect androgen-binding proteins in cytosol from some brain regions. and for identification of radioactive metabolites by thin layer and radiogas chromatography. This type of experiment was performed three times with each sex. In one series of experiments castrated male and female rats were given an intraperitoneal injection of 250 nCi of [1,2,6,7-sH]testosterone in 120 ~1 of acetone. Thirty minutes later the animals were anesthetized with ether and the brain perfused with cold 0.9% NaCl solution (saline) via the aortic artery. The brain was quickly removed, chilled on ice, and dissected.

MATERIALS AND METHODS
The pituitary and the pineal gland were first removed.
The olfactory bulbs were transected just in front of the frontal pole and removed. Transections were made behind and in front of the cerebellum leaving the pons and medulla oblongata after removal of the cerebellum. The brain was then turned upside down, and an incision, approximately 2 mm deep, was made transversely through the pyriform lobe at the same level as the caudal border of the mammillary bodies. A similar incision was made half-way between the fissura chorioidea and sulcus rhinalis running parallel to these grooves. The brain was then divided into two parts by a sagittal section through the midline.
Each of the halves were transected according to Fig. 1. The section in front of the genu of the corpus callosum and the section through the third ventricle and the anterior commissure left a slice from which a piece of the cerebral cortex was taken. The septal area together with the nucleus tractus diagonalis were separated from the rest of this slice by the corpus callosum, the lateral ventricle, and a section from the lateral ventricle to a point at the inferior surface 2 mm lateral to the midline.
The midbrain was removed by a transection from behind the mammillary body to in front of the anterior collicle. A horizontal section separated the thalamus and the preoptic area, so that the latter region could be taken out bounded laterally by the fissura chorioidea. four to six castrated male and female rats were each given 250 nCi of [1,2,6,7-*HItestosterone as described above and were anesthetized 30 min later. Blood was collected by the ophthalmic venous plexus method (16). The brain was perfused, taken out, and dissected as described above. Brain specimens from one rat were pooled with corresponding specimens from other rats of the same sex. Ten volumes of acetone/ethanol, l/l (v/v) were added to each pool of samples, the tissue was homogenized, and the h.omogenate was kept in a shaking water bath at 37" overnight.
The homogenate was centrifuged at 20,000 x g, and the supernatant was evaporated to dryness i n uacuo. The residue was dissolved in distilled water and passed through a 10-g XAD2 column as previously described (17 Fig. 3 shows representative radiogas chromatograms.
In most regions of male rat brain, unmetabolized testosterone constituted the major radioactive compound. Another metabolite of great quantitative importance in male brain was androstenedione and in the pituitary, pineal gland, and hypothalamus, testosterone and androstenedione made up more than 80% of the total radioactivity. Small amounts of 5a-dihydrotestosterone, 5a-androstane-3a, 17P-diol, and 5aandrostane-3&17P-diol were also identified in male rat brain. The pattern of [1,2,6,7-3H]testosterone metabolites in female rat brain was quite different from that of male. Thus, testosterone and androstenedione were of much smaller quantitative importance whereas 5a-androstane-3a,l7@-diol and 5cu-androstane-3&17fl-diol occurred in relatively large amounts. The most striking difference, however, was the sex-specific presence of epitestosterone and 5a-dihydroepitestosterone in female rat brain.
These metabolites were present in practically all regions of the brain and constituted 20 to 40% of the total radioactivity.
one single radioactive peak (p1 5.1) was seen. Displacement macromolecule in male pituitary cytosol is a protein. experiments where the effect of various unlabeled steroids on When pituitary cytosol from castrated female rats given high affinity binding of [1,2,6,7-*H]testosterone to pituitary [1,2,6,7-'HItestosterone 30 min before death was analyzed by cytosol was investigated (Table II) showed that corticosterone gel filtration on Sephadex G-25, the void volume always did not compete for the binding sites whereas androstenedione, contained less than 20% of radioactivity per mg of protein when 5Lu-dihydrotestosterone, and estradiol competed relatively effi-compared to the corresponding fraction from male pituitary ciently, although not to the same degree as unlabeled testoster-cytosol. Isoelectric focusing of the macromolecular-bound fracone. The competition experiment with 5a-dihydrotestosterone tion of female pituitary cytosol did not show any focused indicates that the binding site on the macromolecule binds radioactive peak. When cytosol from female pituitaries were testosterone with somewhat higher affinity than 5cu-dihydrotes-incubated with [1,2,6,7-'HItestosterone under the same conditosterone.
tions as in incubations with male preparations, only small The [aH]testosterone. macromolecular complex formed in amounts of radioactivity bound to macromolecules (correvitro by incubation of male pituitary cytosol was unstable to sponding to about 10% of the amount, calculated per mg of heat and showed a rapid dissociation at 50" (Table III). protein, normally bound in incubations with male cytosol). Furthermore, the complex was sensitive to treatment with Furthermore, it was not possible to displace the radioactivity protease. These results indicate that the testosterone-binding bound by adding excess amounts of unlabeled testosterone to the testosterone-binding protein in male pineal gland had a similar ligand specificity as that present in male pituitary (Table II) except that estradiol seemed to have a somewhat higher affinity to the androgen receptor in pineal gland. No high affinity, low capacity binding of testosterone could be detected in female pineal gland cytosol, either in uiuo or in uitro.
Protein Binding of [I,2,6,7-JH]Testosterone in Hypothalamus In Viuo and In Vitro-Isoelectric focusing of cytosol from male hypothalami labeled in uiuo with [1,2,6,7-'HItestosterone showed a reproducible pattern of two radioactive peaks, a major peak with a p1 of 5.1 and a minor one with a p1 of 4.0 (Fig. 6). These peaks were shown to contain only [3H]testosterone. A similar pattern of peaks was obtained following electrofocusing of male hypothalamic cytosol labeled in uitro with [1,2,6,7-8H]testosterone.
The K,, for the binding of [1,2,6,7-'HItestosterone to the hypothalamic receptor protein, as determined by Scatchard analysis, was 1.1 x lo-" M and the number of binding sites 1.4 x 10-l' mol per mg of protein. Competition experiments indicated that the testosterone receptor in male hypothalamus had a similar ligand specificity as those in male pituitary and pineal gland. Female hypothalamic cytosol Cytosol incubated with [1,2,6,7-3H]testosterone for 30 min at 37" was chromatographed on Sephadex G-25. The void volume was divided into aliquots which were incubated for 10 min at 0, 37, 50, 60, and 80". The incubation at 0" was a control for the experiment. appeared to be devoid of testosterone receptors, as judged from both in vivo and in vitro studies.
Because of the relatively large amounts of [sH]androstenedione found in male hypothalamus following administration of [1,2,6,7-3H Itestosterone, the protein binding of [1,2,6,7-3H]androstenedione was also investigated. This steroid was found to bind to high affinity, low capacity sites with a Kd of 0.9 x lo-lo M; the number of binding sites was 1.4 x lo-l5 mol per mg of protein.
In order to investigate the presence of testosterone receptors in immature rats, in vitro incubations were performed of varying concentrations of [1,2,6,7-3H]testosterone with cytosol from hypothalamic tissue from male and female 2%day-old rats castrated 14 to 16 h before death. When the results were analyzed according to Scatchard it was evident that both male and female hypothalamic tissue contained high affinity, low capacity binding sites for testosterone. The K, was calculated to be 0.5 x 10-l" M and the number of binding sites 0.9 x lo-'* mol per mg of protein for both sexes. These results indicate that the lack of testosterone receptors in adult female hypothalamic tissue does not develop until after puberty.

Protein
Binding of [1,2,6,7-'HITestosterone in Brain Cortex-Analysis by isoelectric focusing of cytosol from male  brain cortex labeled in viuo with [1,2,6,7-3H]testosterone showed a complex pattern of radioactive peaks that was not reproducible in different experiments. In uitro incubations of [1,2,6,7-SH]testosterone with cytosol from male brain cortex showed the presence of high affinity, low capacity binding sites. K, was calculated to 1.0 x 10-l' M, and the number of binding sites to 1 x lo-l6 mol per mg of protein. The data obtained in the present study on the binding characterisitcs of different ligands in various brain regions are summarized in Table IV. Protein Binding of Androgen in Rat Blood-Isoelectric focusing of serum from male and female rats given [1,2,6,7-'HItestosterone did not show any focused radioactive peak. Binding studies in vitro gave no indication for the presence of any specific androgen-binding protein in serum. In view of these findings it may be concluded that the steroid-binding proteins measured in male brain do not originate from blood.

DISCUSSION
The present investigation has demonstrated a marked sexual difference in concentration of high affinity, low capacity binding sites in rat pituitary, hypothalamus, and pineal gland. Whereas these tissues in male rats showed the presence of a receptor protein for testosterone, it was not possible to detect a macromolecule of similar nature in corresponding brain regions from female rats. According to current concepts of mechanism of action of steroid hormones, these compounds are bound to specific receptor proteins in the cytosol of target organs and the steroid. receptor complex is transported into the cell nucleus where transcription of mRNA is facilitated. Absence of receptor protein from a steroid target tissue is combined with unresponsiveness to the hormone in question (22). Probable primary target sites for androgen action on the liver enzymes are the hypothalamus or the pituitary (cf. above), or both, and the present findings therefore offer one possible explanation for the relative androgen unresponsiveness of female liver enzyme activities.
A further indication that testosterone receptor concentration in the brain may determine responsiveness of liver enzymes to androgen was obtained from experiments on 28-day-old male and female rats showing presence of quantifiable amounts of testosterone receptor in pituitary tissue from both sexes. These results correlate well with previous findings on androgen responsiveness of liver enzyme activities in developing rats; at 28 days of age male and female rats respond equally well to treatment with androgen (23).
In a previous study we have shown that androgen responsiveness in male rats is developed as a result of neonatal imprinting by testicular androgen (24). The present investigation has indicated that the underlying mechanism may be irreversible induction of a receptor for testosterone in brain. Neonatal androgen imprinting is known to occur in the hypothalamus (25); whether the androgen receptor in male pituitary and pineal gland is induced as a result of direct imprinting on these tissues or whether it is regulated by the hypothalamus remains to be shown.
Irreversible programming of sex steroid receptor concentration in adult life by androgen action in the neonatal period may be a general phenomenon.
Treatment of female rats with testosterone propionate shortly after birth decreased the uptake of estradiol in hypothalamus and uterus on Day 60 of life (26). On the other hand, feminization of newborn male rats with cyproterone resulted in increased estradiol binding in the hypothalamus (27). When guinea pigs were treated with cyproterone acetate pre-and postnatally and were castrated early in life they developed a relative androgen unresponsiveness compared to postpubertally castrated control animals with respect to testosterone stimulation of prostate, seminal vesicle, and preputial gland growth (28). Insufficient production of androgen during the neonatal period has been suggested to play a role in the etiology of testicular feminization in the human and in animals, a disease associated with generalized absence of androgen receptors in all organs (22). Again the question may be asked whether receptor concentration in peripheral target organs is regulated by direct androgen imprinting or whether it is regulated by the imprinted hypothalamo-hypophyseal system. In addition to the differences in specific protein binding, sex differences were also observed in the pattern of [3H]testosterone metabolites recovered from several brain regions including pituitary, pineal gland, and hypothalamus. The most conspic-uous differences were the predominance of testosterone and androstenedione in male brain compared to the quantitative importance of 5a-androstane-3a, 17&diol, 5n-androstane-3@,17@-diol, epitestosterone, and dihydroepitestosterone in female brain. It may be suggested that an additional explanation for the androgen unresponsiveness in female rats is the faster metabolism of testosterone to inactive compounds in female brain. On the other hand, the reason why testosterone appears to be more protected from metabolism in male brain may be binding of testosterone to specific high affinity protein-binding sites.
The sex-specific occurrence of epitestosterone and dihydroepitestosterone in female brain is probably related to the sex-specific presence of 17a-hydroxysteroid reductase in female brain. Epitestosterone is known to be inactive as androgen, and in confirmation of this we did not find any change in liver enzyme activities in castrated male or female rats when these were given epitestosterone propionate.' In spite of this, epitestosterone was found to bind to high affinity, low capacity sites in both male and female brain tissue. It is possible that conversion of testosterone to epitestosterone and dihydroepitestosterone in female brain constitutes a process of physiological significance yet to be elucidated.
The physical characteristics and ligand affinities and specificities of the testosterone receptors in male pituitary, pineal gland, and hypothalamus were similar and also agree well with reports from other laboratories (9-12). It may well be that the testosterone receptors studied in various brain regions are identical. The physiological ligand for the receptor appears to be testosterone rather than 5a-dihydrotestosterone, as suggested by some authors (10,13). This conclusion is in agreement with previous results indicating that testosterone itself may be the predominant active androgen principle in uiuo in most androgen target organs and that conversion to 5a-dihydrotestosterone generally is not a prerequisite for androgen activity (15).