Estrogen Deficiency in Men Is a Challenge for Both the Hypothalamus and Pituitary

Dirk Vanderschueren1 and Roger Bouillon

Laboratory for Experimental Medicine and Endocrinology (Legendo) Catholic University Leuven 3000 Leuven, Belgium

Address correspondence and requests for reprints to: Roger Bouillon, Laboratory for Experimental Medicine and Endocrinology (Legendo), Katholieke Universiteit Leuven, Herestraat 49, B-3000 Leuven, Belgium. E-mail: Roger.Bouillon{at}med.kuleuven.ac.bc


    Introduction
 Top
 Introduction
 References
 
There is little doubt that androgens are crucial for phenotypic expression of gender differences. That men and women should be politically and socially equal despite small phenotypically differences due to separate hormones was once applauded in the French parliament as "vive la petite différence." Indeed, androgens are not only the driving force for masculinization of both internal and external genitalia but are probably also responsible for sexual dimorphism of brain, bone, skin, hair, and lipid metabolism. Androgens and estrogens were, for a long time, considered as two opponents using two different receptors with mainly antagonistic effects. However, are these hormones so different from each other? Androgens, indeed, also have the puzzling feature of being prohormones for estrogens. The aromatase enzyme complex (CYP19) has the unique ability to convert androgens into estrogens via a multistep enzymatic process (1). Therefore, the relatively low serum concentration of estrogens in men may not accurately reflect their intracrine and paracrine actions of estrogens. Our understanding of the clinical implications of estrogen deficiency in men has recently been updated by the description of similar phenotypes of men suffering from mutations resulting in either loss of function of the estrogen receptor (ER) {alpha} or the cytochrome P450 aromatase enzyme (2). An important role of estrogen was revealed in skeletal development, male fertility, carbohydrate and lipid metabolism, and in regulation of gonadotropin secretion. Furthermore, the characterization of mice made deficient by targeted disruption of the ER (ER-{alpha}, ß, or combined deficiency) or of the aromatase gene confirmed the crucial role of estrogens for male skeletal development, reproductive function, and gonadropin regulation by the hypothalamic-pituitary axis (3). The terminology "hypothalamic-pituitary axis," however, often masks our poor understanding of the precise site and mode of action of gonadotropin regulation by sex steroids. The hypothalamic decapeptide GnRH was discovered more than 30 yr ago. Yet, the precise regulation of the oscillating GnRH pulse generator by sex steroids remains difficult to investigate because of the inability to directly measure GnRH in humans due to the poorly accessible sites of production and action and its short half-life.

In this issue of the journal, Hayes et al. (4) applied an intelligent approach, described earlier by the same group, to circumvent this problem. The effects of selective estrogen deficiency (by aromatase inhibition) on GnRH secretion was evaluated in a tandem study of either normal men or men with idiopathic hypogonadotrophic hypogonadism (IHH) treated with fixed GnRH pulses. Their results confirmed the already well established inhibitory effects of estrogens on LH and FSH secretion in men. However, the interest of the present study lies in the comparison of the effect of estrogen deficiency on gonadotropin secretion in normal men with free oscillating GnRH vs. IHH men with a hypothalamic GnRH clamp. Indeed, despite similar sex steroid hormone concentrations, the response of LH both in terms of frequency and amplitude of pulses was significantly greater in normal men compared with IHH men. Therefore, at least part of the difference in LH output is due to differences in GnRH secretion. Hayes et al. (4) further dissected out, again indirectly, whether this increase of GnRH output was due to an increase of GnRH amplitude or differences in pituitary sensitivity.

The acute administration of a submaximal dose of a specific GnRH antagonist resulted in a similar decrease of LH output in normal men whether estrogen replete (without aromatase inhibitor) or estrogen deplete (with aromatase inhibitor) conditions, implying that the hypothalamic GnRH pulses are of equal amplitude in both groups. The increased pulse frequency in estrogen-deficient males is, thus, due to a direct hypothalamic effect on GnRH pulse frequency, whereas the increased amplitude of LH would be due to decreased negative feedback at the pituitary level. This increased sensitivity of the pituitary gland in response to sex steroids was, indeed, already documented by earlier experiments in IHH men during a similar hypothalamic clamp (5, 6). Therefore, the hypothalamus is now another target tissue where the effects of androgens are mediated via the ER after prior aromatization.

The study by Hayes et al. (4) is a clear example of a well-designed clinical study to answer an at first sight unapproachable question in men. The major end point, the frequency of LH pulses as marker for GnRH pulse frequency has been evaluated by an established technique but could have been improved by deconvolution analysis.

Many questions, however, remain unanswered.

First, what are the target cells for estrogen action and their mode of action on the GnRH network: the GnRH neuron itself (which seems to lack functional ER during adult life) or the surrounding glia cells or afferent neurons? (7). Moreover, which ER ({alpha} or ß) is involved and are the effects generated via the classical genomic mode of action or mediated via rapid nongenomic mechanisms? Indeed, in animal experiments a rapid (<2 h) inhibitory effect of estrogens on GnRH release suggested a nongenomic mode of action.

Second, where are the estrogens responsible for the action on the hypothalamic GnRH secretion produced? Are these estrogens produced locally by the aromatase activity within the GnRH network or are systemic estrogens the key messengers? Indeed, the aromatase activity is found in many tissues and cells in higher primates, whereas in most other vertebrates aromatase activity is rather limited to the gonades and the brain. In humans, CYP19 is, however, also expressed in the placenta, adipose tissue, hair follicle, muscle, bone, and liver (8). In the brain CYP19 activity is found in various places, including the anterior and mediobasal hypothalamus. It is presently unclear whether GnRH neurons or other cells of the GnRH network express high levels of aromatase activity (7). If produced locally then the regulation of the CYP19 gene should be taken into account. Indeed, many cytokines including those found in life-threatening stress (e.g. tumor necrosis factor{alpha} and interleukin-1) or prostaglandins have potent effects on aromatase expression in other cells and, thus, may potentially mediate stress effects on the hypothalamic-pituitary gonadotrophic axis (9).

Third, is there still a place for direct action of androgens (with or without prior 5{alpha}-reduction) via the androgen receptor (AR) or all are actions on the GnRH network mediated via the ER? Because the GnRH network has a gender-specific pattern, it would at first sign seem to be directly androgen responsive (7). The masculinization of the GnRH neurons originates, indeed, from their exposure to an androgen pulse during the early neonatal period and eliminates forever the estrogen-mediated positive feedback on the GnRH network, which is so typical for the female ovulatory LH surge. However, exposure to an estrogen pulse during the first 5 days of life also masculinizes the GnRH neurons, suggesting that aromatization of testosterone and the ER are the pathways involved. In contrast, nonaromatizable androgens are unable to inhibit GnRH secretion in animals and men (10). Other arguments seem to indicate that the AR activation nevertheless has a direct role on GnRH/LH/FSH secretion. Indeed, gonadotropin secretion in both animals and men with an inactivating AR mutation is frankly increased despite high circulating estrogen levels pointing toward a separate direct negative effect of androgens (10, 11, 12). These effects, however, could be due to direct androgen action at the pituitary level (see following paragraph).

Fourth, is the inhibitory action and mode of action of estrogens at the pituitary level different from their effects on the hypothalamus? Again, the same questions can be raised regarding the origin (local or systemic) of estrogens, their use of the ER {alpha} and ß, and the choice between genomic/nongenomic activation pathways. Moreover, the negative effects could be located at different levels because estrogens can decrease the number of GnRH receptors on gonadotrophs (13). The GnRH peptidase activity also fluctuates. Therefore, genomic effects at the transcriptional level are the most likely mechanism. In contrast to the GnRH network, several arguments point toward direct effects of testosterone on LH secretion via the AR. Indeed, prior conversion into dihydrotestosterone by the abundant pituitary 5{alpha}-reduction was found to be essential for testosterone action on GnRH-induced calcium signaling in the rat gonadotroph (14).

It is now increasingly clear that androgens and estrogens are not always hormones with counteracting effects. In fact, in many situations both hormones use the same final pathway, after local transformation and activating the same receptor and postreceptor mechanisms. The hypothalamic regulation of GnRH secretion is now a new example. This was already previously well documented for the gonadotroph in the pituitary (5, 6), the growth plate chondrocyte, bone cells, and testicular efferent ductuli (2, 3, 15, 16, 17). Indeed, the pubertal growth plate does not close in men with either an ER or aromatase mutation. Therefore, such men show continuous linear growth and develop eunuchoid characteristics. Administration of estrogens to aromatase-deficient men, however, result in mineralization and closure of the growth plate (2). Moreover, bone cells and especially osteoblast-mediated osteoclastogenesis are sensitive to both estrogens and androgens, and aromatase inhibition causes similar degree of osteoporosis in male animals as orchidectomy (18). The role of estrogens for male reproduction below the level of GnRH and LH/FSH regulation is not yet solved. Male {alpha}-ERKO mice show an apparently normal development of the reproductive tract but are nevertheless infertile, with decreased sperm quality resulting in the inability to fertilize wild-type oocytes (3, 17). This failure of spermatogenesis in {alpha}-ERKO mice is caused by severely impaired fluid absorption, luminal distension and atrophy of the seminiferous tubuli. No such abnormalities are found in ß-ERKO mice. Aromatase mutations in male rats and men also have variable effects on fertility. However, failure of spermatogenesis in aromatase knockout mice is not caused by a primary defect in seminiferous tubuli fluid reabsorption (such as described in {alpha}-ERKO mice), but seems secondary to direct failure of germ cell differentiation (15).

The role of prior aromatization of androgens on the male brain (beside GnRH secretion) and especially on male behavior is also not settled (19). Indeed, in ER and aromatase mutant men, sexual behavior seemed to be grossly normal. {alpha}-ERKO male mice nevertheless show decreased mounting attempts with an especially low rate of intromissions and ejaculations. {alpha}-ERKO male mice also have a marked deficit in all male aggressive behavioral indices, which may, therefore, be estrogen dependent (3). AR-deficient XY mice and men (Tfm), in contrast, do not display male-like gender behavior. Therefore, most observations point toward a dominant role for the androgen rather than ER in male gender role behavior.

Of course, not all androgen actions are mediated via prior aromatization and the ER (see Fig. 1Go). Sometimes, the AR and ER also regulate genes independently in the same (e.g., osteoblast-derived cytokines) or opposite direction (e.g. lipoproteins). Moreover, a number of crucial gender-specific characteristics, such as growth, differentiation of prostate or hair follicles, and muscle development, directly depend on the activation of the AR, with or without prior 5{alpha}-reduction of testosterone.



View larger version (16K):
[in this window]
[in a new window]
 
Figure 1. The multiple mode of action of androgens on target tissues in men. Testosterone-activated AR (with or without prior 5{alpha}-reduction) is essential for its effects on the skin, male accessory sexual glands, muscle, and gender behavior. Aromatization of androgens is, however, required for its action on skeletal development, hypothalamic-hypophyseal secretion of GnRH-LH-FSH, and male reproductive function.

 
The older concept in endocrinology of a single hormone activating a single receptor and specific transduction system is increasingly replaced by a more complex mode of action. Indeed, the peripheral transformation of the original hormonal signal in the target tissue is becoming increasingly the rule rather than the exception. Therefore, the hormonal activity will ultimately also depend on intracellular enzyme activities, receptor concentration and function, and postreceptor cross-talk by other signals to modulate the fine-tuning of the biological response. Examples include the dejodinase enzymes (T4/T3), 25-OHD-1{alpha}-hydroxylase (vitamin D action in nonclassical target tissue), cortisol inactivation in renal tubular cell (for selectivity of aldosterone action), and 17ß-steroid dehydrogenase activity for E1/E2 conversion. The functional activity of the aromatase enzyme can now be added to this list as being important for the functional translation of the testosterone signal in an increasing number of target tissues, and the hypothalamus is now clearly as such a tissue (4).


    Acknowledgments
 
We thank Claire Dignef for secretarial assistance.


    Footnotes
 
1 Senior Clinical Investigator for the Fund for Scientific Research-Flanders, Belgium (F.W.O.-Vlaanderen). Back

Received July 17, 2000.

Accepted July 17, 2000.


    References
 Top
 Introduction
 References
 

  1. Simpson E, Rubin G, Clyne C, et al. 2000 The role of local estrogen biosynthesis in males and females. Trends Endocrinol Metab. 11:184–188.[CrossRef][Medline]
  2. Grumbach MM, Auchus RJ. 1999 Estrogen: consequences and implications of human mutations in synthesis and action. J Clin Endocrinol Metab. 84:4677–4694.[Abstract/Free Full Text]
  3. Couse JF, Korach KS. 1999 Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev. 20:358–417.[Abstract/Free Full Text]
  4. Hayes FJ, Seminara SB, DeCruz S, Boepple PA, Crowley Jr WF. 2000 Aromatase inhibition in the human male reveals a hypothalamic site of estrogen feedback. J Clin Endocrinol Metab. 85:3027–3035.[Abstract/Free Full Text]
  5. Finkelstein JS, Whitcomb RW, O’Dea LstL, Longcope C, Schoenfeld DA, Crowley Jr WF. 1991 Sex steroid control of gonadotropin secretion in the human male. I. Effects of testosterone administration in normal and gonadotropin-releasing hormone-deficient men. J Clin Endocrinol Metab. 73:609–620.[Abstract]
  6. Finkelstein JS, O’Dea LstL, Whitcomb RW, Crowley Jr WF. 1991 Sex steroid control of gonadotropin secretion in the human male. II. Effects of estradiol administration in normal and gonadotropin-releasing hormone-deficient men. J Clin Endocrinol Metab. 73:621–628.[Abstract]
  7. Herbison AE. 1998 Multimodal influence of estrogen upon gonadotropin-releasing hormone neurons. Endocr Rev. 19:302–330.[Abstract/Free Full Text]
  8. Sharpe RM. 1998 The roles of oestrogen in the male. Trends Endocrinol Metab. 9:371–377.[CrossRef]
  9. Van den Berghe G, de Zegher F, Bouillon R. 1998 Clinical review 95: acute and prolonged critical illness as different neuroendocrine paradigms. J Clin Endocrinol Metab. 83:1827–1834.[Free Full Text]
  10. Quigley CA. 1998 The androgen receptor: physiology and pathophysiology. In: Nieschlag E, Behre HM, eds. Testosterone: action-deficiency-substitution, 2nd ed. Heidelberg: Springer-Verlag; 33–106.
  11. Vanderschueren D, Van Herck E, Suiker AMH. 1993 Bone and mineral metabolism in the androgen-resistant (testicular feminized) male rat. J Bone Miner Res. 8:801–809.[Medline]
  12. Purvis K, Hang E, Clausen OPF, Naess O, Hansson V. 1977 Endocrine status of the testicular feminized male (Tfm) rat. Mol Cell Endocrinol. 8:317–334.[CrossRef][Medline]
  13. Kaiser UB, Conn PM, Chin WW. 1997 Studies of gonadotropin-releasing hormone (GnRH) action using GnRH receptor-expressing pituitary cell lines. Endocr Rev. 18:46–70.[Abstract/Free Full Text]
  14. Tobin VA, Canny BJ. 1998 The regulation of gonadotropin-releasing hormone-induced calcium signals in male rat gonadotrophs by testosterone is mediated by dihydrotestosterone. Endocrinology. 139:1038–1045.[Abstract/Free Full Text]
  15. Jones ME, Robertson KM, O’Donnell L, Boon W, Simpson ER. 2000 The aromatase knockout (ArKO) mouse. Re-evaluating a role for estrogen in male physiology. In: Francavilla F, Francavilla S, Forti G, eds. Andrology 2000. Proceedings of the 1st European Congress of Andrology. Italy: Collana di Studi Abruzzesi; p. 299–324.
  16. Rochira V, Balestrieri A, Madeo B, Granata ARM, Carani C. 2000 Clinical aspects of congenital estrogen defect and physiological role of estrogens in the human male. In: Francavilla F, Francavilla S, Forti G, eds. Andrology 2000. Proceedings of the 1st European Congress of Andrology. Italy: Collana di Studi Abruzzesi; p. 325–344.
  17. Hess RA. 2000 Estrogen and the male reproductive tract. In: Francavilla F, Francavilla S, Forti G, eds. Andrology 2000. Proceedings of the 1st European Congress of Andrology. Italy: Collana di Studi Abruzzesi; p. 279–298.
  18. Vanderschueren D, Boonen S, Bouillon R. 1998 Mini-review: action of androgens versus estrogens in male skeletal homeostasis. Bone 23:391–394.
  19. Wilson JD. 1999 The role of androgens in male gender role behavior. Endocr Rev. 20:726–737. [Abstract/Free Full Text]