Immunocytochemical localization of type 5 17beta-hydroxysteroid dehydrogenase in human reproductive tissues

Endocrine and intracrine sources of androgens in women: inhibition of breast cancer and other roles of androgens and their precursor dehydroepiandrosterone

Serum androgens as well as their precursors and metabolites decrease from the age of 30-40 yr in women, thus suggesting that a more physiological hormone replacement therapy at menopause should contain an androgenic compound. It is important to consider, however, that most of the androgens in women, especially after menopause, are synthesized in peripheral intracrine tissues from the inactive precursors dehydroepiandrosterone (DHEA) and DHEA sulfate (DHEA-S) of adrenal origin. Much progress in this new area of endocrine physiology called intracrinology has followed the cloning and characterization of most of the enzymes responsible for the transformation of DHEA and DHEA-S into androgens and estrogens in peripheral target tissues, where the locally produced sex steroids are exerting their action in the same cells in which their synthesis takes place without significant diffusion into the circulation, thus seriously limiting the interpretation of serum levels of active sex steroids. The sex steroids made in peripheral tissues are then inactivated locally into more water-soluble compounds that diffuse into the general circulation where they can be measured. In a series of animal models, androgens and DHEA have been found to inhibit breast cancer development and growth and to stimulate bone formation. In clinical studies, DHEA has been found to increase bone mineral density and to stimulate vaginal maturation without affecting the endometrium, while improving well-being and libido with no significant side effects. The advantage of DHEA over other androgenic compounds is that DHEA, at physiological doses, is converted into androgens and/or estrogens only in the specific intracrine target tissues that possess the appropriate physiological enzymatic machinery, thus limiting the action of the sex steroids to those tissues possessing the tissue-specific profile of expression of the genes responsible for their formation, while leaving the other tissues unaffected and thus minimizing the potential side effects observed with androgens or estrogens administered systemically.

A cluster of eight hydroxysteroid dehydrogenase genes belonging to the aldo-keto reductase supergene family on mouse chromosome 13

A subclass of hydroxysteroid dehydrogenases (HSD) are NADP(H)-dependent oxidoreductases that belong to the aldo-keto reductase (AKR) superfamily. They are involved in prereceptor or intracrine steroid modulation, and also act as bile acid-binding proteins. The HSD family members characterized thus far in human and rat have a high degree of protein sequence similarity but exhibit distinct substrate specificity. Here we report the identification of nine murine AKR genes in a cluster on chromosome 13 by a combination of molecular cloning and in silico analysis of this region. These include four previously isolated mouse HSD genes (Akr1c18, Akr1c6, Akr1c12, Akr1c13), the more distantly related Akr1e1, and four novel HSD genes. These genes exhibit highly conserved exon/intron organization and protein sequence predictions indicate 75% amino acid similarity. The previously identified AKR protein active site residues are invariant among all nine proteins, but differences are observed in regions that have been implicated in determining substrate specificity. Differences also occur in tissue expression patterns, with expression of some genes restricted to specific tissues and others expressed at high levels in multiple tissues. Our findings dramatically expand the repertoire of AKR genes and identify unrecognized family members with potential roles in the regulation of steroid metabolism.

Selective reduction of AKR1C2 in prostate cancer and its role in DHT metabolism

BACKGROUND

As androgens play an essential role in prostate cancer, we sought to develop a real-time PCR to characterize mRNA expression profiles of human members of the Aldo-Keto Reductase (AKR) 1C gene family, as well as of 5 alpha-steroid reductase Type II (SRD5A2) in prostate cancer samples. Functional activity and regulation of AKR1C2, a 3 alpha-hydroxysteroid dehydrogenase (HSD) type III, was also assessed in prostate cancer cell lines.

METHODS

Gene specific PCR primers were established and relative gene expression of human AKR1C family members was determined in paired samples of cancerous and surrounding unaffected prostate tissue.

RESULTS

AKR1C2 preferentially reduces DHT to the weak metabolite 5 alpha-androstane-3 alpha,17 beta-diol (3 alpha-diol) without conversion of 3 alpha-diol to DHT in the PC-3 cell line, and its expression was increased by DHT treatment in LNCaP cells. Selectively reduced expression of AKR1C2 mRNA, but not AKR1C1 (97% sequence identity), was found in approximately half of the pairs whereas AKR1C3 relative expression was not significantly altered. No aberrant expression of AKR1C4 expression or significant differences in SRD5A2 gene expression were found.

CONCLUSIONS

AKR1C2 functions as a DHT reductase in prostate-derived cells lines and is regulated by DHT. Additional studies are needed to further define the significance of reduced AKR1C2 expression in prostate cancer and its potential role in modulating local availability of DHT.

Human 17beta-hydroxysteroid dehydrogenase type 5 is inhibited by dietary flavonoids

Phytoestrogens contained in a vegetarian diet are supposed to have beneficial effects on the development and progression of a variety of endocrine-related cancers. We have tested the effect of a variety of dietary phytoestrogens, especially flavonoids, on the activity of human 17beta-hydroxysteroid dehydrogenase type 5 (17beta-HSD 5), a key enzyme in the metabolism of estrogens and androgens. Our studies show that reductive and oxidative activity of the enzyme are inhibited by many compounds, especially zearalenone, coumestrol, quercetin and biochanin A. Among flavones, inhibitor potency is enhanced with increased degree of hydroxylation. The most effective inhibitors seem to bind to the hydrophilic cofactor binding pocket of the enzyme.

Cutaneous androgen metabolism: basic research and clinical perspectives

The skin, especially the pilosebaceous unit composed of sebaceous glands and hair follicles, can synthesize androgens de novo from cholesterol or by locally converting circulating weaker androgens to more potent ones. As in other classical steroidogenic organs, the same six major enzyme systems are involved in cutaneous androgen metabolism, namely steroid sulfatase, 3beta-hydroxy-steroid dehydrogenase, 17beta-hydroxysteroid dehydrogenase, steroid 5alpha-reductase, 3alpha-hydroxysteroid dehydrogenase, and aromatase. Steroid sulfatase, together with P450 side chain cleavage enzyme and P450 17-hydroxylase, was found to reside in the cytoplasm of sebocytes and keratinocytes. Strong steroid sulfatase immunoreactivity was observed in the lesional skin but not in unaffected skin of acne patients. 3beta-hydroxysteroid dehydrogenase has been mainly immunolocalized to sebaceous glands, with the type 1 being the key cutaneous isoenzyme. The type 2 17beta-hydroxysteroid dehydrogenase isoenzyme predominates in sebaceous glands and exhibits greater reductive activity in glands from facial areas compared with acne nonprone areas. In hair follicles, 17beta-hydroxysteroid dehydrogenase was identified mainly in outer root sheath cells. The type 1 5alpha-reductase mainly occurs in the sebaceous glands, whereby the type II isoenzyme seems to be localized in the hair follicles. 3alpha-hydroxysteroid dehydrogenase converts dihydrotestosterone to 3alpha-androstanediol, and the use of 3alpha-androstanediol glucuronide serum level to reflect the hyperandrogenic state in hirsute women may be a reliable parameter, especially for idiopathic hirsutism. In acne patients it is still controversial if 3alpha-androstanediol glucuronide or androsterone glucuronide could serve as suitable serum markers for measuring androgenicity. Aromatase, localized to sebaceous glands and to both outer as well as inner root sheath cells of anagen terminal hair follicles, may play a "detoxifying" role by removing excess androgens. Pharmacologic development of more potent specific isoenzyme antagonists may lead to better clinical treatment or even prevention of androgen-dependent dermatoses.

Androgenic and estrogenic 17beta-hydroxysteroid dehydrogenase/17-ketosteroid reductase in human ovarian epithelial tumors: evidence for the type 1, 2 and 5 isoforms

17beta-Hydroxysteroid dehydrogenase/17-ketosteroid reductases (17HSD/KSR) play a key role in regulating steroid receptor occupancy in normal tissues and tumors. Although 17HSD/KSR activity has been detected in ovarian epithelial tumors, our understanding of which isoforms are present and their potential for steroid metabolism is limited. In this investigation, 17HSD/KSR activity from a series of ovarian epithelial tumors was assayed in cytosol and microsomes under conditions which differentiate between isoforms. Inhibition studies were used to further characterize the steroid specificities of isoforms in the two subcellular fractions. Activity varied widely between tumors of the same histopathologic classification. The highest levels of activity were observed in mucinous tumors. Michaelis constants, maximum velocities, estradiol-17beta/testosterone (E(2)/T) activity ratios and inhibition patterns were consistent with a predominance of microsomal 17HSD/KSR2 and cytosolic 17HSD/KSR5, isoforms reactive with both E(2) and T, with evidence of estrogenic 17HSD/KSR1 in cytosol from some samples. In tumors where activity and mRNA expression were both characterized, Northern blots, PCR and sequence analysis indicated 17HSD/KSR5 was the predominant isoform. The presence of 17HSD/KSR5, which also has both 3alpha-HSD/KSR and 20alphaHSD/KSR activity, and 17HSD/KSR2 which also has 20alpha-HSD activity, could influence not only estrogen and androgen binding but progesterone receptor occupancy, as well, in receptor-containing tumors.

Androgens and mammary growth and neoplasia

OBJECTIVE

Evaluation of current clinical, experimental, genetic, and epidemiological data pertaining to the role of androgens in mammary growth and neoplasia.

DESIGN

Literature review.

SETTING

National Institutes of Health.

SUBJECT(S)

Recent, basic, clinical, and epidemiological studies.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

Effects of androgens on mammary epithelial proliferation and/or breast cancer incidence.

RESULT(S)

Experimental data derived from rodents and cell lines provide conflicting results that appear be strain- and cell line-dependent. Epidemiologic studies have significant methodological limitations and provide inconclusive results. The study of molecular defects involving androgenic pathways in breast cancer is in its infancy. Clinical and nonhuman primate studies, however, suggest that androgens inhibit mammary epithelial proliferation and breast growth and that conventional estrogen treatment suppresses endogenous androgens.

CONCLUSION(S)

Abundant clinical evidence suggests that androgens normally inhibit mammary epithelial proliferation and breast growth. Suppression of androgens by conventional estrogen treatment may thus enhance estrogenic breast stimulation and possibly breast cancer risk. Clinical trials to evaluate the impact of combined estrogen and androgen hormone replacement regimens on mammary gland homeostasis are needed to address this issue.

The biochemical basis for increased testosterone production in theca cells propagated from patients with polycystic ovary syndrome

Ovarian theca cells propagated from patients with polycystic ovary syndrome (PCOS) convert steroid precursors into T more efficiently than normal theca cells. To identify the basis for increased T production by PCOS theca cells, we examined type I-V 17 beta-hydroxysteroid dehydrogenase (17 beta HSD) isoform expression in long-term cultures of theca and granulosa cells isolated from normal and PCOS ovaries. RT-PCR analysis demonstrated that theca cells express type V 17 beta HSD a member of the aldo-keto reductase (AKR) superfamily (17 beta HSDV, AKR1C3), whereas expression of type I, II, and IV 17 beta HSD, which are members of the short-chain dehydrogenase/reductase superfamily, was limited to granulosa cells. Type III 17 beta HSD, the testicular isoform, was not detected in either granulosa or theca cells. Northern and real-time PCR analyses demonstrated that 17 beta HSDV transcripts were not significantly increased in PCOS theca cells compared with normal theca cells. RT-PCR analysis revealed that theca cells also express another AKR, 20 alpha HSD (AKR1C1). Both basal and forskolin-stimulated 20 alpha HSD mRNA levels were increased in PCOS theca cells compared with normal theca cells. However, 17 beta HSD enzyme activity per theca cell was not significantly increased in PCOS, suggesting that neither AKR1C3 nor AKR1C1 contributes to the formation of T in this condition. In contrast, 17 alpha-hydroxylase/C17,20 lyase and 3 beta HSD enzyme activities were elevated in PCOS theca cells, driving increased production of T precursors. These findings indicate that 1) increased T production in PCOS theca cells does not result from dysregulation of "androgenic" 17 beta HSD activity or altered expression of AKRs that may express 17 beta HSD activity; and 2) increased synthesis of T precursors is the primary factor driving enhanced T secretion in PCOS.

Ontogenesis of oxidative reaction of 17beta-hydroxysteroid dehydrogenase and 11beta-hydroxysteroid dehydrogenase in rat Leydig cells, a histochemical study

The enzyme 17beta-hydroxysteroid dehydrogenase is required for the synthesis and 11beta-hydroxysteroid dehydrogenase for the regulation of androgens in rat Leydig cells. This histochemical study describes ontogenetic changes in distribution and intensity of these enzymes in Leydig cells from postnatal day (pnd) 1-90. Using NAD or NADP as the cofactor, 17beta-hydroxysteroid dehydrogenase (substrate: 5-androstene-3beta,17beta-diol) peaks were observed on pnd 16 for fetal Leydig cells and on pnd 19 and 37 for adult Leydig cells. Between pnd 13 and 25 the fetal cells showed a higher intensity for the 17beta-enzyme than the adult cells; more fetal Leydig cells were stained with NADP, whereas more adult cells were positive with NAD on pnd 13 and 16. A nearly identical distribution of 11beta-hydroxysteroid dehydrogenase (substrate: corticosterone) was observed with NAD or NADP as the cofactor; the reaction was present from pnd 31 onwards, first in a few adult Leydig cells and later in almost all these cells homogeneously. The ontogenetic curves of the two enzymes show an inverse relationship. To conclude: (1) Generally, a stronger reaction for 17beta-hydroxysteroid dehydrogenase is shown with NAD as cofactor than with NADP; using NADP, fetal Leydig cells show a stronger staining than adult Leydig cells. (2) The data possibly support the notion of a new isoform of 11beta-hydroxysteroid dehydrogenase in addition to types 1 and 2.

The androgen-conjugating uridine diphosphoglucuronosyltransferase-2B enzymes are differentially expressed temporally and spatially in the monkey follicle throughout the menstrual cycle

UDP-glucuronosyltransferase (UGT) enzymes enhance the polarity of steroid hormones by catalyzing their conjugation with the sugar group from UDP-glucuronic acid. Previous results have shown that the monkey is a suitable animal model to study steroid glucuronidation in steroid target tissues. In humans, as in the monkey, the main androgen metabolites found in the circulation are 5alpha-androstane-3alpha,17beta-diol-glucuronide and androsterone glucuronide, and high levels of androsterone glucuronide were also measured in human follicular fluid. Ovarian androgens play a significant role as precursors for estrogens and may modulate the recruitment and growth of follicles. To analyze the expression pattern of UGT2B enzymes involved in androgen metabolism throughout the menstrual cycle, cynomolgus monkey ovaries were collected during the mid and late follicular and luteal phases. Microsomal proteins and total RNA were analyzed for UGT2B expression in the whole ovary. Western blot and specific RT-PCR analyses demonstrated no significant changes in the expression of UGT2B protein or transcripts during the menstrual cycle. Immunocytochemistry analysis showed that UGT2B proteins are expressed in the cytoplasm of thecal and granulosa cells of growing follicles. Interestingly, the thecal cells of secondary follicles and of corpus luteum were extensively stained, whereas luteal granulosa cells were not labeled. These results suggest an important regulation of cell type-specific UGT2B expression during follicular development. Previous results demonstrated similar changes in the expression of the androgen receptor. The colocalization of the androgen receptor and UGT2B enzymes in the same cell types of the ovary provide evidence for a potential role of glucuronidation as a modulator of the intracellular androgen response during follicular development.

Human types 1 and 3 3 alpha-hydroxysteroid dehydrogenases: differential lability and tissue distribution

3 alpha-Hydroxysteroid dehydrogenases (3 alpha-HSDs) catalyze the conversion of 3-ketosteroids to 3 alpha-hydroxy compounds. The best known 3 alpha-HSD activity is the transformation of the most potent natural androgen, dihydrotestosterone, into 5 alpha-androstan-3 alpha,17 beta-diol (3 alpha-diol), a compound having much lower activity. Previous reports show that 3 alpha-HSDs are involved in the metabolism of glucocorticoids, progestins, prostaglandins, bile acid precursors, and xenobiotics. 3 alpha-HSDs could, thus, play a crucial role in the control of a series of active steroid levels in target tissues. In the human, type 1 3 alpha-HSD was first identified as human chlordecone reductase. Recently, we have isolated and characterized type 3 3 alpha-HSD that shares 81.7% identity with human type 1 3 alpha-HSD. The transfection of vectors expressing types 1 and 3 3 alpha-HSD in transformed human embryonic kidney (HEK-293) cells indicates that both enzymes efficiently catalyze the transformation of dihydrotestosterone into 3 alpha-diol in intact cells. However, when the cells are broken, the activity of type 3 3 alpha-HSD is rapidly lost, whereas the type 1 3 alpha-HSD activity remains stable. We have previously found that human type 5 17 beta-HSD which possesses 84% and 86% identity with types 1 and 3 3 alpha-HSD, respectively, is also labile, whereas rodent enzymes such as mouse type 5 17 beta-HSD and rat 3 alpha-HSD are stable after homogenization of the cells. The variable stability of different enzymatic activities in broken cell preparations renders the comparison of different enzymes difficult. RNA expression analysis indicates that human type 1 3 alpha-HSD is expressed exclusively in the liver, whereas type 3 is more widely expressed and is found in the liver, adrenal, testis, brain, prostate, and HaCaT keratinocytes. Based on enzymatic characteristics and sequence homology, it is suggested that type 1 3 alpha-HSD is an ortholog of rat 3 alpha-HSD while type 3 3 alpha-HSD, which must have diverged recently, seems unique to human and is probably more involved in intracrine activity.

Type 5 17beta-hydroxysteroid dehydrogenase: its role in the formation of androgens in women

Type 5 17beta-HSD, one of the seven types of 17beta-hydroxysteroid dehydrogenase (17beta-HSD) so far characterized in humans, catalyzes the transformation of 4-androstenedione (4-dione) into testosterone (T). This reaction is also catalyzed by type 3 17beta-HSD which is responsible for pseudohermaphroditism in deficient man but is asymptomatic in deficient women. Since type 3 17beta-HSD is not found in the ovary, whereas type 5 is, it is suggested that the latter is involved in the conversion of 4-androstenedione to testosterone in the ovary. The comparison of type 5 17beta-HSD of different species shows that the human enzyme shares 95 and 78% identity with the monkey and mouse enzymes respectively. In addition, the human and monkey enzymes are labile and are destroyed upon homogenization of the transfected cells, whereas the mouse enzyme is not. Human type 5 17beta-HSD also possesses a high 20alpha-HSD activity that inactivates progesterone, whereas the monkey and mouse enzymes do not have such high 20alpha-HSD activity. Since the endocrine ovary is composed of two types of cells, one producing androgens (theca cells) and the other producing progesterone and estrogens (granulosa cells), it is tempting to suggest that the role of the high 20alpha-HSD activity of type 5 17beta-HSD is to protect the theca cells against the progesterone produced by the granulosa cells. The double activity of type 5 17beta-HSD in the female reproductive tissues is probably necessary to the control of the optimal level of progesterone and sex steroids.

Phytoestrogens inhibit human 17beta-hydroxysteroid dehydrogenase type 5

The 17beta-hydroxysteroid dehydrogenase type 5 (17beta-HSD 5) is involved in estrogen and androgen metabolism. In our study we tested the influence of environmental hormones, such as phytoestrogens (flavonoids, coumarins, coumestans), on reductive and oxidative 17beta-HSD activity of the human 17beta-hydroxysteroid dehydrogenase type 5 (17beta-HSD 5). These dietary substances were shown to be potent inhibitors of aromatase, different 17beta-HSDs and seem to play an important role in delay of development of hormone dependent cancers. Our studies show that reductive and oxidative activity of the enzyme are inhibited by many dietary compounds, especially zearalenone, coumestrol, quercetin and biochanin A. Among the group of flavones inhibitor potency is growing with increasing number of hydroxylations. We suggest that these substances are bound to the hydrophilic cofactor-binding pocket of the enzyme. An interesting inhibition pattern is observed for 18beta-glycyrrhetinic acid, which has no influence on the oxidative but only on the reductive reaction. This indicates that this substrate binds to pH- and cofactor-depending sites at the active center of the enzyme.

Analysis and characteristics of multiple types of human 17beta-hydroxysteroid dehydrogenase

Androgens and estrogens are not only synthesized in the gonads but also in peripheral target tissues. Accordingly, recent molecular cloning has allowed us to identify multiple types of 17beta-hydroxysteroid dehydrogenases (17beta-HSD), the key and exclusive enzymes involved in the formation and inactivation of sex steroids. However, only one form, namely, type 3 17beta-HSD, is responsible for pseudohermaphroditism in deficient boys. To date, seven human 17beta-HSDs have been isolated and characterized. Although they catalyze substrates having a similar structure, 17beta-HSDs have very low homology. In intact cells in culture, these enzymes catalyze the reaction in a unidirectional way - types 1, 3, 5 and 7 catalyze the reductive reaction, while types 2, 4 and 8 catalyze the oxidative reaction. It is noteworthy that rat type 6 17beta-HSD also catalyzes the reaction in the oxidative direction. In this report, we analyze the different characteristics of the multiple types of human 17beta-HSD.

Unique features of the basal cells of human prostate epithelium

The prostate gland is globally composed of epithelium and stroma. The epithelium plays an important role in the development of both benign and malignant disorders while the stroma is involved in benign prostatic hyperplasia. While the prostatic epithelium of the majority of laboratory animals is well recognized as a pseudostratified columnar, the classification of the human prostatic epithelium is controversial. Moreover, the role of the basal cells of the human prostatic epithelium is still uncertain. These cells have been described as undifferentiated cells, precursors of luminal cells, reserve and myoepithelial cells. The objective of the present study was to assess the similarities and/or differences between the epithelium of the human prostate and that of other laboratory animals and thus derive information about the potential functions of basal cells in the human prostate. In the human, basal cells form a continuous layer of cells resting on the basement membrane and upon which rests a layer of luminal cells. This results in a stratified columnar epithelium of two layers of cells, unlike the sporadic appearance of basal cells observed in other species where it results in a pseudostratified epithelium. In addition, the ratio of basal to luminal cells in the human is about 1:1, while the average ratio in the other animal species examined is about 1:7. Furthermore, the gap junctional proteins connexin 26 and 43, are present between basal and luminal cells in the human, thus suggesting that these cells communicate directly with each other. In addition, the ultrastructure of the human basal cells shows morphological evidence of differentiated but not of undifferentiated cells. Moreover, the presence of junction-like structures between adjacent basal cells suggests that these cells form a blood-prostate barrier. In this way, basal cells could prevent substances derived from the blood from directly coming in contact with the luminal cells. Human basal cells could thus regulate functions of the luminal cells by being part of a two-cell mechanism somewhat analogous to thecal and granulosa cells in the ovary.

Role of 17 beta-hydroxysteroid dehydrogenases in sex steroid formation in peripheral intracrine tissues

In postmenopausal women, almost 100% of active sex steroids are synthesized in peripheral target tissues from inactive steroid precursors and, in adult men, approximately 50% of androgens are made locally in target tissues. This new field of endocrinology has been called intracrinology. The last and key step in the formation of all estrogens and androgens is catalyzed by a series of substrate-specific, cell-specific and unidirectional 17 beta-hydroxysteroid dehydrogenases (17 beta-HSDs). To date, seven human 17 beta-HSDs have been cloned, sequenced and characterized. The 17 beta-HSDs provide each cell with the means of precisely controlling the intracellular concentration of each sex steroid according to local needs.

Localization of 17beta-hydroxysteroid dehydrogenase/17-ketosteroid reductase isoform expression in the developing mouse testis--androstenedione is the major androgen secreted by fetal/neonatal leydig cells

The final step in the biosynthesis of testosterone is reduction of androstenedione by the enzyme 17beta-hydroxysteroid dehydrogenase/ 17-ketosteroid reductase (17betaHSD/17KSR). In this study, we have examined expression of the four known reductive isoforms of 17betaHSD/ 17KSR (types 1, 3, 5, and 7) in the developing mouse testis and have determined changes in the localization of isoform expression and testosterone secretion during development. Using RT-PCR isoforms 1, 3, and 7 were shown to be expressed in the seminiferous tubules of neonatal testis, whereas isoforms 3 and 7 were expressed in the interstitial tissue of the adult testis. The type 7 isoform is unlikely to be involved in androgen synthesis and further study concentrated on the type 3 isoform. Developmentally, isoform type 3 was expressed in the seminiferous tubules up to day 10, showed little or no expression on day 20 and from day 30 was confined to the interstitial tissue. In situ hybridization confirmed that the type 3 isoform was expressed only in the seminiferous tubules in fetal testes and in the interstitial tissue in adult testes. In accordance with the localization of enzyme messenger RNA expression 17-ketosteroid reductase enzyme activity was very low in isolated interstitial tissue from neonatal testes while interstitial tissue from adult testes showed high activity. Seminiferous tubules from both neonatal and adult testes showed high levels of enzyme activity. The major androgen secreted by the interstitial tissue of prepubertal animals was androstenedione up to day 20 while 5alpha-androstanediol and/or testosterone were the major androgens secreted from day 30 onwards. These results show that fetal Leydig cells do not express significant levels of a reductive isoform of 17betaHSD/ 17KSR and that androstenedione is the major androgen secreted by these cells. Production of testosterone up until puberty is dependent upon 17betaHSD/17KSR activity in the seminiferous tubules--a "two cell" requirement for testosterone synthesis. Expression of the 17betaHSD/17KSR type 3 isoform (the main reductive isoform in the testis) declines in the seminiferous tubules before puberty but then reappears in the developing adult Leydig cell population.