No abstract

Aromatization and 5 alpha-reduction are known to be required for the full expression of testosterone actions in neuroendocrine tissues. Although aromatase and 5 alpha-reductase activities in brain and pituitary can be experimentally manipulated by castration and steroid replacement, naturally occurring variations during seasonal reproductive cycles have not been examined in any species. Goldfish (Carassius auratus) were selected for study because they exhibit exceptionally high levels of aromatase in both brain and pituitary, although 5 alpha-reductase levels resemble the vertebrate norm. Four animals of each sex were tested monthly through three breeding seasons (2.5 yr). Using previously validated techniques, the enzymes were assayed by product formation from [3H]androstenedione in homogenates of anterior hypothalamus-preoptic area (AHPOA), remaining telencephalon (TEL), whole pituitary, ovary, and testis. Seasonal variations in aromatase were most dramatic in the AHPOA of female fish, exhibiting a peak in April and May that was 6-fold higher than the nadir in July. As judged by changes in the appearance and weight of the gonads, maximal aromatase coincided with the spawning season, whereas low enzyme levels corresponded to reproductive inactivity. Seasonal variations were similar but of a lesser magnitude in the TEL of females and in the AHPOA and TEL of males (2- to 3-fold, peak to nadir). Both ovarian and testicular aromatase showed cyclic changes; however, activity was much lower than that in brain at all times of the year (4.5, 1.2, and 47.0 pmol/mg protein, maximal values in ovaries, testes, and AHPOA, respectively). Pituitary aromatase varied from 5-22 pmol/mg protein, but was not consistently correlated with season. Cyclic changes in 5 alpha-reductase were distinctly different from those in aromatase, with maximal values in both brain and pituitary occurring when fish were reproductively inactive. In general, circulating sex steroids were high when aromatase was high and low when reductase was maximal; however, there was no apparent causal relationship suggested by temporal changes in a given steroid. Variations in testosterone metabolism, by regulating the quantity and quality of active hormone in close proximity to receptor sites, may be responsible for the changes in feedback sensitivity and behavioral responsiveness that are known to occur in seasonal breeders.

Testosterone (T) exerts its actions in brain directly via androgen receptors or, after aromatization to estradiol, via estrogen receptors. Brain aromatase activity in teleost fish is 100-1000 times greater than in mammals and would be expected to significantly reduce the quantity of androgen available for receptor binding. Experiments were carried out on the goldfish Carassius auratus to determine if androgen receptors are present in teleost brain and whether their physicochemical properties reflect elevated aromatase. Cytosolic and nuclear extracts were assayed with the use of [3H]T and charcoal, Sephadex LH-20, or DNA-cellulose chromatography to separate bound and free steroids. Binding activity was saturable and had an equally high affinity for T and 5 alpha-dihydrotestosterone (Kd, approximately 2.4 X 10(-9) M). Although mibolerone was a relatively weak competitor, the putative teleost androgen 11-ketotestosterone, methyltrienolone (R1881), estradiol, progesterone, and cortisol were poor ligands. Characteristics that distinguish this receptor from a steroid-binding protein in goldfish serum are the presence of binding activity in both nuclear and cytosolic extracts, a low rate of ligand-receptor dissociation, electrophoretic mobility, sedimentation properties in low vs. high salt, and tissue distribution (forebrain greater than or equal to pituitary greater than mid-/hindbrain). DNA cellulose-adhering and nonadhering forms were detected, but these did not differ in other variables measured. Although goldfish androgen receptors resembled those of mammals in all important physicochemical characteristics, they were unusually abundant (5-68 pmol/g tissue) compared to levels in rat brain, but comparable to levels in prostate and other male sex hormone target organs. Moreover, there were seasonal variations in total receptors, with a peak at spawning (April) 4- to 5-fold higher than values in reproductively inactive fish (July/August). This temporal pattern and magnitude of change corresponded to previously reported changes in brain aromatase. Thus, both phylogenetic and physiological correlates point to a functional interdependence between androgen receptors and aromatase in the brain. These studies in goldfish indicate that brain androgen receptors have a long evolutionary history and have been highly conserved through the vertebrate series.

Estrogen formation in brain and pituitary mediates certain androgen actions in central targets. Goldfish (Carassius auratus) and quail (Coturnix coturnix japonica) have been advantageous for studying the role of locally formed estrogen in autoregulating aromatization and in controlling estrogen receptor occupancy, androgen receptor levels, and behavioral expression. Data from these two experimental models reveal a molecular basis for androgen-estrogen synergism in neuroendocrine tissues and for alterations in androgen sensitivity/responsiveness. These mechanisms are essential components of seasonal reproduction in the test species and may have wider relevance for cyclicity in other vertebrates, including mammals. BRAIN AROMATASE: AN OVERVIEWFollowing development of high specific activity radiolabeled steroids in the late '~O S , pathways of steroidogenesis in glandular tissues were intensively investigated (for review, Engel, '75). During this era it was established that androstenedione and testosterone (T) are the immediate and obligatory precursors of estrone and estradiol (E2), respectively. The enzyme catalyzing this reaction was termed aromatase, a name describing transformation of the A4, 3-ketone-A ring of androgens to the aromatic benzene ring of natural estrogens. Although initial studies focused primarily on human placenta and human or porcine ovary, tissues exceptionally rich in aromatase, it was soon determined that estrogen is synthesized in testis, adrenal, and a variety of non-glandular tissues such as fat and muscle (Longcope et al., '78). Owing to their large mass, peripheral sites of estrogen synthesis make a substantial contribution to the circulating estrogen pool when androgen substrates are available in the peripheral blood.In 1971, the first identification of aromatase was made in human fetal brain and shortly thereafter in rat, rabbit, and mouse brain (Naftolin et al., '75). Significantly, enzyme activity was restricted to regions with E2-target cells and known to control reproduction and sex behavior (hypothalamus = HTH, preoptic area = POA, amygdala).These findings led to proposal of the "aromatization hypothesis": namely, that aromatization of 0 1990 WILEY-LISS, INC.circulating androgen in the brain itself is required for the full expression of androgen actions in central targets. Although this hypothesis explained the well-known "paradoxical effects" of sex steroids on vertebrate behaviors (Young, ,611, it met with some skepticism at first, due in part to extremely low levels of estrogen actually formed in brain. Proof that even small quantities of active hormone may be physiologically important when synthesized in close proximity to sites of action was obtained by experiments showing that peripheral administration of [3H]T to gonadectomized-adrenalectomized rats resulted in nuclear-bound [3HlE2 in brain regions with aromatase activity but not in other areas (Lieberburg and McEwen, '77). It is noteworthy here that unmetabolized T and 5a-dihydrotestosterone (DHT), another T m...

No abstract