Department of Medicine University of Cambridge School of Clinical Medicine Cambridge CB2 2QQ, United Kingdom
Address all correspondence and requests for reprints to: Dr. J. E. Compston, Box 157, Department of Medicine, Addenbrookes Hospital, Cambridge CB2 2QQ, United Kingdom. E-mail: jec1001{at}cam.ac.uk.
Sex steroids play a central role in bone development and homeostasis. Androgens and estrogens are both required for normal skeletal development in women and men and are responsible for the sexual dimorphism of the mature skeleton. In age-related bone loss, however, there is increasing evidence that estrogen deficiency is the major determinant of bone loss both in women and in men. The overriding effect of estrogen on the skeleton is supported by the severe developmental failure of bone in males with deficient estrogen activity, as a result of estrogen receptor dysfunction or aromatase deficiency (1, 2), and the correlation between endogenous estradiol concentrations and both bone mineral density and bone loss in men (3, 4).
In recent years, there have been considerable advances in our understanding of receptor-mediated transcriptional effects of steroid hormones, particularly in the case of estrogen, and of nongenomic pathways. The concept of prereceptor regulation, in which factors present in the local environment modulate the amount of active hormone available, has also gained prominence, particularly in the context of aromatase-induced conversion of androgens to estrogens and the interconversion of cortisone and cortisol. However, the extent to which gonadal steroids can be synthesized by target tissues and the complexity of the pathways responsible is less well recognized. Labrie et al. (5) have used the term "intracrinology" to describe this process, in which active biosynthesis occurs in the same cells in which the steroids act without release of hormone into the circulation or extracellular space.
Local regulation of hormone activity in target tissues has obvious advantages. It enables fine-tuning of the intracellular concentration of active metabolites in a tissue in the presence of a wide range of circulating concentrations of hormones, thus providing mechanisms for tissue-specific responses in the absence of changes in systemic hormone production, and for the preservation of homeostasis in the face of alterations in hormonal status. Furthermore, the ability to modulate hormone activity within specific tissue microenvironments has obvious therapeutic potential for achieving selectivity of the effects of hormones, all of which have multiple and sometimes unwanted actions. It is clear from examples such as hypercortisolism and estrogen deficiency that these local pathways cannot always protect the skeleton against extremes of hormone secretion, but they may reduce adverse effects, and the interindividual variations in sensitivity to some hormones may, at least in part, be attributable to genetically determined differences in enzyme production.
A number of enzymes are implicated in skeletal sex steroid metabolism (Fig. 1); these include aromatase, 17ß-hydroxysteroid dehydrogenase (HSD; of which there are at least eight isoforms), 3ß-HSD, steroid sulfatase, and 5
-reductase, all of which have been demonstrated in osteoblastic cells. In addition, 7
-hydroxylase, which converts dehydroepiandrosterone (DHEA) to 7
-OH-DHEA, has recently been demonstrated in rat osteoblastic cells (6). The substrate for these enzymes is C19 precursors, which circulate in high concentrations and of which DHEA is the most abundant. From these compounds estrone and testosterone can be synthesized directly whereas estradiol and 5
-dihydrotestosterone (DHT) are formed from testosterone. While estrone and estradiol can be interconverted by 17ß-HSDs, the conversion pathway of testosterone to DHT is unidirectional and dependent on the presence of 5
-reductase enzymes (7, 8).
|
The presence of the 5-reductase enzymes in bone raises intriguing questions both about the relative roles of estrogen and androgen at different stages of skeletal maturation and the specific functions of testosterone and DHT in bone. Although estrogens are essential for normal skeletal development and maintenance in women and men, there is evidence that direct effects of androgens that are independent of aromatase activity also contribute. Thus, in the condition of complete androgen insensitivity in humans, in which mutations of the androgen receptor render target organs unresponsive to the actions of androgen, bone mineral density is decreased and total restitution of bone mass may not be achieved by the administration of estrogens (11). Testosterone prevents bone loss induced by orchidectomy in ERKO male mice (12), and the androgen receptor antagonist flutamide produces osteopenia in female rats (13) and also partially abrogates the protective effect of DHEA against bone loss in ovariectomized rats (14). Finally, both in vitro and in vivo, nonaromatizable androgens have similar effects on bone as testosterone. These data, thus, indicate that androgens can act directly on bone and suggest that these actions may be physiologically relevant. In particular, they may be responsible for the attainment of larger bone size in men as a result of periosteal apposition, a process that is stimulated by androgens but inhibited by estrogens (15, 16).
The relative roles of testosterone and DHT in the skeleton have not been resolved. Both have similar binding affinities to the androgen receptor and have direct effects on osteoblastic cells in vitro; the majority of studies have demonstrated stimulatory effects on both proliferation and differentiation, although these findings have not been universal (17). Furthermore, it has recently been shown that both DHT and 17ß-estradiol inhibit osteoblast apoptosis and stimulate osteoclast apoptosis by a nongenomic mechanism that, although receptor dependent, does not involve gene transcription but is mediated by the activation of extracellular signal-related kinases (18). In vivo, as noted above, nonaromatizable androgens at least partially abrogate bone loss in orchidectomized rats (19). Current evidence, thus, suggests that the two androgens have similar skeletal effects, but whether there are differential effects on gene expression remains to be established.
Most of the studies investigating effects of 5-reductase deficiency have focused on the type 2 isoenzyme, and models for deficiency of the type 1 isoenzyme have not been identified either in animals or man (type 1 knockout mice do not survive beyond mid-gestation). Consistent with the finding that the type 1 isoenzyme predominates in bone, administration of the type 2 5
-reductase inhibitor finasteride in men with benign prostatic hyperplasia does not appear to affect bone mineral density or biochemical markers of bone turnover (20). In 5
-reductase type 2 deficiency in humans, serum estrogen, testosterone, and gonadotropins are within the normal range for men whereas circulating DHT levels are reduced with an increase in the ratio of testosterone to DHT. Fisher et al. (21) described retardation of bone age in two children with the condition, although this was mild and no other skeletal abnormalities were noted. In another study, no significant reduction in mean bone mineral density was demonstrated in 16 affected patients, with Z-scores of -0.84 and 0.14 at the spine and femoral neck, respectively (22).
The evidence provided by Issa et al. (10) that the type 1 isoenzyme of 5-reductase is the predominant form in osteoblastic cells provides an important step forward in our understanding of the complex intracrinology of sex steroids in bone. It highlights the need for more research into the skeletal actions of testosterone and DHT and the local regulatory mechanisms controlling 5
-reductase production, in order that the functional significance of this enzyme pathway can be better understood. It is possible that the two androgens have specific actions, perhaps at different skeletal sites and/or at different stages of skeletal maturation. Alternatively, conversion of testosterone to DHT may be a mechanism for reducing estrogenic effects on bone or increasing androgenic actions in the skeleton; thus, whereas testosterone may act either as androgen or (after aromatization) estrogen, DHT is unambiguously androgenic. Although the preeminence of estrogens as regulators of skeletal homeostasis is undisputed, there is compelling evidence that direct androgenic effects also contribute, and these warrant further investigation.
Acknowledgments
Footnotes
Abbreviations: DHEA, Dehydroepiandrosterone; DHT, 5-dihydrotestosterone; HSD, hydroxysteroid dehydrogenase.
Received September 11, 2002.
Accepted September 15, 2002.
References