Visualization of X-chromosome inactivation mosaicism of Tfm gene in XTfm/X+ heterozygous female mice

The testicular feminization (Tfm) locus, which produces a deficiency in androgen receptors, is located on the X-chromosome. Steroid autoradiographic techniques were used to demonstrate the mosaicism of the X-chromosome inactivation in two androgen target tissues of XTfm/X+ heterozygous female mice. In the mesenchyme of urogenital sinuses of wild-type female fetuses (X+/X+), more than 95% of the cells were androgen-receptor positive (labelled with [3H]testosterone) while in that of heterozygous fetuses (XTfm/X+), about half of the cells were receptor positive (Tfm gene inactive). Statistical analysis of coherent clone size was applied to the heterozygous mesenchyme of the urogenital sinus and the coherent clone size of receptor-positive cells was estimated to be two or three cells per clone. This small clone size suggests that considerable cell mixing occurred in the tissue during embryonic development. Androgen binding in the mammary gland rudiments was restricted to the mesenchymal cells only in close vicinity to the epithelial mammary bud. In the wild-type rudiments most of the mesenchymal cells beneath the epithelium were receptor positive, while in heterozygous rudiments, receptor-positive and -negative cells intermingled. This observation suggests that in the wild-type mammary gland rudiments the epithelial bud may induce the formation of androgen receptors in adjacent mesenchymal cells rather than attract pre-existing receptor-rich mesenchymal cells.     

Autoradiographic studies of androgen-binding sites in the rat urogenital sinus and postnatal prostate

Binding sites of [3H]testosterone and [3H]dihydrotestosterone in the rat fetal urogenital sinus and postnatal prostate and vagina grown in vitro were examined by steroid autoradiography. Distinct nuclear incorporation of both androgens appeared between 14.5 and 16.5 days of gestation in rat fetuses. Nuclear labelling in the sinus was restricted to the mesenchyme surrounding the epithelium which showed no nuclear labelling. A similar distribution of labelled cells was observed in male and female sinuses up to 18.5 days of gestation. By 20.5 days of gestation, the labelling in the ventral mesenchyme of female urogenital sinuses became less intense but persisted in the mesenchyme of the dorsal sinus wall from which the vagina is formed. In the postnatal prostate, the epithelium showed nuclear [3H]testosterone labelling at 10 days coinciding with the onset of its functional differentiation. Epithelial labelling became more intensive at 4 weeks post partum while that of the mesenchyme declined. The results suggest two phases of androgen action: formation of the prostatic buds mediated by the androgen-activated mesenchyme of the fetal urogenital sinus and the differentiation of the postnatal prostatic epithelium directly stimulated by androgens.     

Analysis of prostatic bud induction by brief androgen treatment in the fetal rat urogenital sinus

The androgen dependency of prostatic bud formation in fetal rat urogenital sinuses was studied using brief treatments with androgen, and the incorporation of androgens by the sinus mesenchyme was followed by steroid autoradiography. Urogenital sinuses from 16.5-day fetuses of both sexes were grown in organ culture and treated with androgens for periods ranging from 4 to 72 h and then transferred to control medium. A minimum treatment of 24 h was required to induce prostatic buds in male sinuses and of 36 h in all female sinuses. This difference in response disappeared after more prolonged treatment. In both sexes the number of prostatic buds increased with the time of exposure to androgens. Prostatic bud formation continued for 24-36 h after transfer to control medium. Steroid autoradiographic analysis showed that the labelled androgen was concentrated in the mesenchymal nuclei. The rate of incorporation rose steeply during the first 12 h and then more slowly. After transfer to control medium the amount of labelled androgen decreased rapidly to half within 12 h and then decreased more slowly. In the competition experiments a 200-fold excess of unlabelled testosterone or dihydrotestosterone in the labelling medium greatly reduced the nuclear labelling with [3H]testosterone.     

Inability of Tfm (testicular feminization) epithelial cells to express androgen-dependent seminal vesicle secretory proteins in chimeric tissue recombinants

To assess the role of androgen receptors (ARs) in the expression of androgen-dependent seminal vesicle (SV) secretory proteins, tissue recombinants were prepared with rat seminal vesicle mesenchyme plus ureter epithelium of wild-type or Tfm mice (rat SVM plus wild-type mouse URE and rat SVM plus Tfm mouse URE, respectively). After growth in male hosts, both the wild-type and Tfm ureter epithelia were induced by SVM to differentiate into a simple columnar epithelium exhibiting the complex folded morphology characteristic of the SV. In SVM plus wild-type mouse URE recombinants, epithelial ARs were induced, and the epithelium expressed the full spectrum of SV secretory proteins. By contrast, in SVM plus Tfm mouse URE recombinants, the Tfm epithelium was genetically incapable of producing functional ARs and failed to produce SV secretory proteins. These data demonstrate in vivo that the induction of SV secretory proteins by androgens is an event requiring intraepithelial ARs. In contrast, androgen-dependent epithelial morphogenesis, columnar cytodifferentiation, and probably also proliferation can be expressed in Tfm epithelium grown in association with wild-type mesenchyme, strongly suggesting that these events are indirect effects on the epithelium mediated by mesenchymal ARs.     

Androgen receptors in thymic epithelium modulate thymus size and thymocyte development

Castration of normal male rodents results in significant enlargement of the thymus, and androgen replacement reverses these changes. Androgen-resistant testicular feminization (Tfm) mice also show significant thymus enlargement, which suggests that these changes are mediated by the androgen receptor (AR). The cellular targets of androgen action in the thymus are not known, but may include the lymphoid cells (thymocytes) as well as nonlymphoid epithelial cells, both of which have been believed to express AR. In the present study immunohistochemical analysis and hormone binding assays were used to demonstrate the presence of AR in thymic epithelial cells. The physiological significance of this epithelial cell AR expression was defined by further studies performed in vivo using chimeric mice, produced by bone marrow transplantation, in which AR expression was limited to either lymphoid or epithelial components of the thymus. Chimeric C57 mice engrafted with Tfm bone marrow cells (AR(+) epithelium and AR(-) thymocytes) had thymuses of normal size and showed the normal involutional response to androgens, whereas chimeric Tfm mice engrafted with C57 bone marrow cells (AR(-) epithelium and AR(+) thymocytes) showed thymus enlargement and androgen insensitivity. Furthermore, phenotypic analyses of lymphocytes in mice with AR(-) thymic epithelium showed abrogation of the normal responses to androgens. These data suggest that AR expressed by thymic epithelium are important modulators of thymocyte development.     

Tissue interaction in androgen response of embryonic mammary rudiment of mouse: identification of target tissue for testosterone.

In the androgen response of the embryonic mammary rudiment of the mouse, both gland epithelium and surrounding mesenchyme are visibly involved. The question whether this is due to a direct action of testosterone on both tissues was investigated in experimental combination of mammary epithelium and mammary mesenchyme, derived either from normal or from androgen-insensitive (XTfm/Y) embryos. A typical androgen response occurred in combinations of androgen-insensitive epithelium with normal mesenchyme, whereas all combinations of normal epithelium with androgen-insensitive mesenchyme failed to respond. It is therefore concluded that only the mesenchyme of the mammary rudiment is the target tissue for testosterone, and that all changes in the gland epithelium, including its necrosis, are secondarily caused by testosterone-activated mesenchymal cells.     

Promotion of murine hepatocarcinogenesis by testosterone is androgen receptor-dependent but not cell autonomous.

Tfm (testicular feminization) mutant mice lack functional androgen receptors. By studying liver tumor development in Tfm mice, we have shown that the greater susceptibility of male mice relative to female mice for liver tumor induction by N,N-diethylnitrosamine is androgen receptor-dependent. C57BL/6J normal and Tfm mutant mice were injected at 12 days of age with N,N-diethylnitrosamine (0.2 mumol/g, i.p.), and liver tumors were enumerated in 50-week-old animals. Normal males averaged 20 liver tumors per animal; Tfm males, 0.7; normal females, 0.6; and Tfm/+ heterozygous females, 1.5. The androgen receptor gene and the Tfm mutation are X chromosome linked. Because of random X chromosome inactivation, hepatocytes from Tfm/+ heterozygous female mice are mosaic with respect to the expression of mutant or wild-type receptors. To determine if testosterone acts directly as a liver tumor promoter, through the androgen receptor in preneoplastic hepatocytes, or by an indirect mechanism, we chronically treated these mosaic female mice with testosterone and measured the androgen receptor content of the resulting tumors. B6C3F1 Tfm/+ mosaic and +/+ wild-type female mice were injected i.p. at 12 days of age with N,N-diethylnitrosamine (0.1 mumol/g) and ovariectomized at 8 weeks of age. Half of the mice of each group subsequently received biweekly s.c. injections of testosterone (0.15 mg per mouse) for 30 weeks. Tumor multiplicity was the same for wild-type and Tfm/+ mosaic females treated with testosterone (31-32 tumors per animal at 38 weeks of age) and was increased relative to females not treated with testosterone (13-17 tumors per animal at 50 weeks of age). Testosterone treatment did not significantly increase the percentage of androgen receptor-positive tumors in Tfm/+ mosaic females: 58% of the tumors from Tfm/+ mosaic females treated with testosterone were receptor positive compared to 48% in Tfm/+ females not treated with testosterone and 92% in wild-type females treated with testosterone. Finally, the number of androgen receptors in the majority of liver tumors examined was greatly decreased relative to the surrounding normal liver tissue. We conclude that liver tumor promotion by testosterone requires a functional androgen receptor in the intact animal. However, this promotion is not cell autonomous; that is, the response of the preneoplastic hepatocyte is not dependent on the expression of functional receptor in the target cell.     

Metabolism of testosterone by the epithelium and mesenchyme of the rat urogenital sinus

Testosterone metabolism was measured in separated epithelium and mesenchyme from the urogenital sinuses of 17- and 19-day-old male and female rat embryos and compared with testosterone metabolism in the intact sinus. Both the epithelium and the mesenchyme converted testosterone to 5 alpha-dihydrotestosterone. The epithelium produced much more androstanedione and androsterone but less 3 alpha, 17 beta-androstanediol than did the mesenchyme. The whole sinus synthesized all four metabolites, but in different proportions, producing relatively more androsterone than either of its two component tissues. These data suggest that androsterone is formed by the joint action of epithelium and mesenchyme. Metabolism of testosterone did not differ with sex or foetal age in either of the separated tissues or in the intact sinus, implying that the failure of urogenital mesenchyme from 19-day-old female foetuses to induce prostatic morphogenesis is not due to the loss of 5 alpha-reductase. It is suggested that this lack of inductive capacity may be attributable to a decline in androgen levels with age in female mesenchyme.     

Androgen dependence of growth and epithelial morphogenesis in neonatal mouse bulbourethral glands

The early development of the mouse bulbourethral gland (BUG) and the role of testosterone (T) in the normal growth and epithelial morphogenesis of this male accessory sex gland were examined. The mouse BUG differentiates from the urogenital sinus on day 17 of gestation (vaginal plug = day 0; birth = day 19), and initially consists of a solid epithelial rudiment encased in a large condensed capsular mesenchyme. The epithelium begins to branch and canalize on day 1 postnatally, and the branches enlarge and become more numerous on days 2 and 3. On day 4, secondary branches appear, and by day 6, the epithelium has become extensively arborized and almost fills the mesenchymal capsule. The BUG increases 3.9-fold in DNA content from day 0 (day of birth) to day 6 postnatally; the epithelium grows proportionately more than the mesenchyme during this period (12-fold vs. 2.3-fold). Growth of BUGs in mice castrated at birth or castrated and then treated with cyproterone acetate, an antiandrogen, over the first 6 days of life was reduced by 80%, but not abolished. Thus, the growth of the BUG is partially independent of androgens during early neonatal life. However, morphogenesis of the BUG epithelium is totally abolished in neonatally castrated mice. T replacement given to neonatally castrated mice during days 0-6 restored development to normal. T injections also reinitiated growth and morphogenesis in developmentally retarded BUGs from 6-day-old neonatally castrated mice. The partial dependence of the neonatal BUG on androgens for growth is similar to that seen in the prostate, which is also derived from the urogenital sinus. In contrast to the prostate, where neonatal castration reduces but does not abolish epithelial morphogenesis, androgen deprivation completely abolished epithelial morphogenesis in the neonatal BUG. (Endocrinology 121: 2153-2160, 1987).     

17Beta-HSD type 5 expression and the emergence of differentiated epithelial Type II cells in fetal lung: a novel role for androgen during the surge of surfactant

Lung maturation is delayed in male fetuses compared to female fetuses. This has been attributed to higher levels of androgens in the male lung. We previously showed that the genes encoding for the 17beta-hydroxysteroid dehydrogenase (HSD) type 5 (conversion of androstenedione in testosterone) and type 2 (the opposite reaction) are, respectively, expressed in the human epithelial Type II (PTII)-like A549 cells and in human lung fibroblasts. Here, we aim to explain the physiological relevance of androgen synthesis by PTII cells. We showed that 17beta-HSD type 2 and type 5 genes are both up-regulated in correlation with the emergence of mature PTII cells in both male and female developing lungs of the fetal mouse. In contrast, the androgen receptor gene is expressed at similar levels in both sexes with no temporal regulation. In conclusion, the expression profile of the 17beta-HSD type 5 gene does not explain the presence of higher levels of androgens in the male fetal lung but that androgen synthesis must be a normal characteristic of mature PTII cells for both sexes. The production of androgens after the emergence of mature PTII cells should negatively regulate PTII cell maturation and thus, a novel and normal role for androgens in cell reprogramming is proposed.     

Effect of human recombinant mullerian inhibiting substance on isolated epithelial and mesenchymal cells during mullerian duct regression in the rat.

The effect of human recombinant Mullerian Inhibiting Substance (MIS) on the regression of the Mullerian duct (MD) of female rat fetuses was examined in vitro to determine whether MIS acts on MD epithelium and/or mesenchyme at the critical periods of sexual differentiation. Urogenital ridges (URs) of female rat fetuses at 14.5- to 18.5-days of gestation (plug day = 0) were cultured for 3 days with or without recombinant human MIS in CMRL 1066 medium with 10% female fetal calf serum. In URs from 14.5- and 15.5-day-old fetuses, the cranial portion of the MD regressed almost completely during the 3-day culture period in the presence of MIS, whereas the caudal half to third of the MD remained intact but tapered to a fine point cranially. MDs survived in URs from 16.5-day-old fetuses cultured in the presence of MIS except that the cranial portion of the MDs was deformed. MIS did not elicit regression of MDs in URs obtained from 17.5- and 18.5-day-old fetuses, but instead caused the MD epithelium to form bulges projecting into the mesenchyme. MD epithelium at 15.5-days of gestation was separated from the surrounding UR mesenchyme, and both components (MD epithelium and mesenchyme) were cultured separately for 3 days in the presence or absence of MIS. Both epithelial and mesenchymal cells survived in the presence or absence of MIS. MD epithelium formed typical epithelial colonies, whereas UR mesenchyme spread as fibroblastic cells. Analysis of labeling index after incorporation of [3H] thymidine demonstrated that MD epithelial DNA synthesis was not influenced by MIS. In contrast, mesenchymal labeling index was reduced significantly by MIS. This effect of MIS on UR mesenchyme in conjunction with earlier histological observations of mesenchymal condensation during MD regression and an absence of direct effects of MIS on the epithelium suggests that MIS elicits its effect on the MD epithelium via the surrounding mesenchyme.     

Induction of androgen receptor formation by epithelium-mesenchyme interaction in embryonic mouse mammary gland.

The role of tissue interaction in the development of hormone responsiveness was studied in the embryonic mammary gland of the mouse, which becomes sensitive to testosterone on day 14. Previously, the mesenchyme had been identified as the sole target tissue for the hormone, although it was also demonstrated that its response to testosterone required the presence of mammary epithelium. Using autoradiography, we now show that [3H]testosterone or [3H]5 alpha-dihydrotestosterone is bound only by those mesenchymal cells closest to the epithelial mammary bud. When mammary epithelia were experimentally associated with mesenchyme of the mammary region and cultured together for 3 days in vitro, they also became surrounded by several layers of [3H]testosterone-binding mesenchymal cells. Correspondingly, this tissue association was accompanied by a substantial increase of androgen-binding sites in the explants. No hormone-building mesenchymal cells were seen in combinations with epidermis or pancreas epithelium; only salivary epithelium showed a weak positive effect. From these results we conclude that mammary epithelium induces the formation of androgen receptors in adjacent mesenchyme and thereby controls the development of androgen responsiveness in this tissue.     

Role of androgens in proliferation and differentiation of mouse mammary epithelial cell line HC11

Androgens have been found in mammary epithelium and in milk throughout the cycle of the mammary gland in vivo. The aim of this study was to investigate the possible role of these substances in mammary epithelial growth and differentiation in the mouse HC11 cell line. Cells were stimulated with testosterone, dihydrotestosterone, androstenedione and 5alpha-androstane-3alpha,17beta-diol at concentrations ranging between 0.3 nM and 30 nM. Cyproterone acetate or flutamide, androgen receptor antagonists, (3 microM) were used to block specific androgen effects. Proliferative effects were measured by an MTT (tetrazolium blue) conversion test and [(3)H]thymidine uptake. HC11 cells were transfected with pbetacCAT, a chimeric rat beta-casein gene promoter-chloramphenicol acetyl transferase (CAT) gene construct and CAT ELISA was used to determine gene expression. RT-PCR was performed to detect androgen receptor expression. After 24, 48 and 72 h androgens significantly (P<0.05) increased proliferation. Androgen antagonists significantly (P<0.05) reduced the proliferative effects. Furthermore androgens potentiated the lactogenic effect of prolactin, insulin and dexamethasone (P<0.05). Finally, the androgen receptor gene was expressed in both proliferating and differentiated HC11 cells. These observations lead us to hypothesize an activity of this class of steroids in mammary physiology. In particular, androgens stimulate cell proliferation and beta-casein gene expression; this influence appears to be mediated by androgen receptors.     

Androgen receptor expression in the developing rat prostate is not altered by castration, flutamide, or suppression of the adrenal axis

Mesenchymal epithelial interactions are believed to be important to the growth and development of the neonatal prostate. Prior studies in the rat ventral prostate, using autoradiography and tritiated dihydrotestosterone, indicate that androgen receptors are present in the prostatic stroma on day 3 and are detected in the epithelium by the tenth postnatal day. These findings suggested that androgen stimulation of the prostatic mesenchyme is a crucial step in the growth and development of the prostate. We have examined this developmental program directly using polyclonal antibodies that recognize specific epitopes of the androgen receptor to examine the pattern of androgen receptor expression in intact and neonatally castrate animals. In keeping with previous studies, androgen receptors are present in the prostate stroma at birth and subsequently appear in the prostatic epithelium by the 10th postnatal day. Development of androgen receptor expression in the epithelium was not changed when the animals were castrated at birth, castrated and blocked by flutamide, or castrated and given hydrocortisone to suppress the production of adrenal androgens. These findings suggest that the appearance of androgen receptors in the prostatic epithelium is programmed by androgens before birth or that factors other than testicular or adrenal androgens control the development of epithelial androgen receptors.     

A link between lung androgen metabolism and the emergence of mature epithelial type II cells

Lung maturation is delayed in male fetuses compared with female fetuses, which has been attributed to higher levels of androgens in the male lung. Our previous studies demonstrated that the genes encoding for the 17 beta-hydroxysteroid dehydrogenase (17 beta-HSD) type 5 (androstenedione --> testosterone) and type 2 (the opposite reaction) are expressed in human epithelial type II (PTII)-like A549 cells and in human lung fibroblasts, respectively. Here, we aim to explain the physiological relevance of androgen synthesis by PTII cells. We showed that both 17 beta-HSD type 2 and type 5 genes are upregulated in correlation with the emergence of mature PTII cells in both male and female developing lungs of the mouse. In contrast, the androgen receptor gene is expressed equally in both sexes with no temporal regulation. We conclude that the expression profile of the 17 beta-HSD type 5 gene does not explain the presence of higher levels of androgen in the male fetal lung, but that androgen synthesis must be a normal feature of mature PTII cells for both sexes. The production of androgens after the emergence of mature PTII cells should negatively regulate PTII cell maturation and, thus, a role for androgens in cell reprogramming is suggested.     

Altered expression of genes involved in regulation of vitamin A metabolism, solute transportation, and cytoskeletal function in the androgen-insensitive tfm mouse testis

Androgens are essential for the development and maintenance of spermatogenesis, but the underlying mechanisms of androgen action in the testis remain unclear. To help clarify these mechanisms, gene expression was measured in testes of pubertal (20 d old), androgen-insensitive, testicular feminized (Tfm) mice and in normal controls. Using microarrays (Affymetrix chips 430A and 430B), initial data identified a large number of genes down-regulated in the Tfm testis (>4700). These genes were largely of germ cell origin, reflecting the arrest of spermatogenesis that is apparent in the 20-d-old Tfm testis. Subsequent screening in vitro and in silico of this gene set identified 20 genes of a somatic tubular origin that were significantly down-regulated in the Tfm testis and six genes that were significantly up-regulated. Altered expression of these genes was confirmed by real-time PCR, and genes down-regulated in the Tfm testis were shown to be up-regulated in testes of hypogonadal (hpg) mice treated with androgen. In a developmental study using real-time PCR most of the regulated genes showed normal expression during fetal and neonatal development and deviated from control only between 10 and 20 d. In all cases, expression was also reduced in the adult, although interpretation is more complex because of the inherent cryptorchidism in the adult Tfm mouse. Of the total number of somatic genes showing differential expression in the Tfm testis, 50% were associated with three separate groups of genes involved in regulation of vitamin A metabolism, solute transportation, and cytoskeletal function. Thus, effects of androgens on tubular function and spermatogenesis may be mediated in part through regulation of the tubular environment and control of retinoic acid concentrations.     

Expression of PGP9.5 in ductal cells of the rat pancreas during development and regeneration: can it be a marker for pancreatic progenitor cells?

The expression of protein gene product 9.5 (PGP9.5), a known neuron marker, was immunohistochemically investigated in rat pancreas. In fetal pancreas, a cluster of cells expressed PGP9.5 among the initial epithelial buds at embryonic day 11.5 (E 11.5). At E 13.5, PGP9.5 appeared among elongated and branching epithelial cells as well as along nerve fibers in the mesenchyme. On E 17.5, tubular cells became ductal cells with lumen, which strongly expressed PGP9.5. In newborn rats, ductal cells of the common bile duct (CBD) to the centroacinar cells and islet cells expressed PGP9.5. Ten days after birth, the number of the ductal cells expressing PGP9.5 was reduced, and PGP9.5-negative cells appeared in half of the duct cells. On day 21, all centroacinar cells and intercalated ductal cells became PGP9.5-negative, but some CBD and interlobular ductal cells remained positive for PGP9.5. On day 28 and thereafter, PGP9.5 was no longer detected. In a pancreatic duct ligation model, acinar cells changed to cells with duct-like structure after duct ligation. These cells strongly expressed PGP9.5 on the fifth day after duct ligation. Three to four weeks after ligation, the cells with duct-like structure changed to acinar cells, islets of Langerhans and ductal cells, but the ductal cells were PGP9.5-negative at this point. These results suggested that PGP9.5 is expressed in ductal cells that possess a potential for differentiation to pancreatic endocrine cells, and therefore can serve as a marker for the progenitor of pancreatic endocrine cells.     

Androgen receptor expression in developing male reproductive organs

The distribution of androgen receptors (AR) in developing male BALB/c mouse reproductive organs was determined by 3H-dihydrotestosterone steroid autoradiography. The efferent ductules, urogenital sinus (UGS) and Wolffian ducts, and their derivatives, the epididymis, ductus deferens, seminal vesicles, coagulating glands, prostate and bulbouretheral glands, were examined in mice from 13-days fetal (gestation = 19-20 days) to 10 days postnatal. All organs contained AR in their mesenchymal/stromal cells at all times examined. The Wolffian ducts and UGS did not contain epithelial AR on days 13-14 or 16 of gestation. The efferent ductule was the first site of epithelial AR expression in the male tract during development; this organ had epithelial AR on day 16 and at all subsequent times. The epididymis and ductus deferens contained epithelial AR beginning on day 19 of gestation. Seminal vesicle and coagulating gland epithelium was AR- at birth, became weakly AR+ on day 1, and was strongly AR+ on day 2 and subsequently. Prostatic epithelium was AR- up to day 4, when some positive epithelial cells were seen; the prostatic epithelium was strongly AR+ on day 6 and subsequently. The last organ to begin expressing epithelial AR was the bulbouretheral gland; this epithelium did not become clearly AR+ until day 8 postnatally. In summary, these results indicate that initial epithelial AR expression in the male reproductive tract occurs in a clear temporal sequence and proceeds in a cranial-caudal direction. Epithelial AR first appear in the efferent ductules, followed by initial epithelial AR expression in Wolffian-derived organs and finally in the UGS-derived organs. The factors controlling initial epithelial AR expression are unclear, but mesenchyme may be involved.