| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Institute of Anatomy, Department of Experimental Embryology, University of Tübingen, D-72074 Tübingen, Germany
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
androgen receptor, embryo, Müllerian ducts, testosterone, vagina
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
According to textbooks of embryology, the vagina arises from paired evaginations of the urogenital sinus, the sinovaginal bulbs, which by fusion form the vaginal plate. The latter unites cranially with the Müllerian ducts. The terms sinovaginal bulbs and vaginal plate were coined by Koff [1], who studied a series of human embryos from the Carnegie collection. Koff explicitly rejected earlier studies by Hart [2] and Mijsberg [3], in which the bulbs were identified as the lower ends of the Wolffian ducts and were designated as "Wolffian bulbs." Witschi [4] reexamined the same human embryos and confirmed the old observation that the vaginal bud is formed by the lower ends of Wolffian and Müllerian ducts. He demonstrated that the lower end of the vagina slides down along the urethra to its separate opening into the vestibulum. Nevertheless, upward growth of sinovaginal bulbs and formation of a vaginal plate remained the favorite interpretation in textbooks and recent publications on the development of the sex organs.
Despite the uncertainties over the derivation of the vagina, and independent from the unsettled questions regarding the developmental mechanism, today it is generally held that the lower third of the definitive vagina stems from the urogenital sinus, and the upper two thirds from the Müllerian ducts. This old tenet seems to be confirmed by the presence of a shortened "sinus vagina" in androgen-insensitive Tfm (i.e., testicular feminization) mice, which must have resisted the anti-Müllerian effect of AMH (anti-Müllerian hormone) produced by the embryonic testes and thus should not be derived from the Müllerian ducts [5]. As will be shown below, this reasoning is not correct.
Testicular Feminization Syndrome
The X-linked mutation for testicular feminization (Tfm) in mice [6] causes complete androgen insensitivity and thus corresponds to the respective syndrome in humans. The molecular basis was shown to be a defect of the androgen receptor protein [7, 8]. The Tfm mutation in mice is a single base deletion of exon 1 of the androgen receptor gene, which leads to a frame shift and premature termination of translation downstream of codon 412 in exon 1 [9].
Previous Experiments
Our interest in the formation of the vagina was raised by the analysis of intersex mice produced by sex reversal of Tfm heterozygotes [10]. In heavily feminized intersexes with a high proportion of androgen-insensitive Tfm cells, the Wolffian ducts and the excretory ducts of the vesicular glands run downward behind the vagina and join the urethra, together with the latter at the penile bulb. The most reasonable explanation was descent of all genital ducts during formation of the vagina.
The effect of androgens on this morphogenetic process was studied in organ culture [11]. We explanted the junctional zone in which Wolffian and Müllerian ducts contact each other and merge with the urogenital sinus. Treatment with testosterone induced differentiation of the Wolffian ducts, leading to a separation of the Müllerian ducts from the latter. Without testosterone, a broad fusion zone between Wolffian and Müllerian ducts developed.
In addition, we studied vaginal development by 3D reconstruction of male, female, and Tfm mice from the indifferent stage up to 8 days after birth [12]. In mice, the structures in the fusion zone were the same as those depicted in human embryos. The reconstruction clearly indicated that the vagina was formed by down growth of the genital ducts. Nevertheless, the derivation of the "sinovaginal bulbs" from the urogenital sinus or the Wolffian ducts, and whether sinovaginal bulbs bud off cranially to form a vaginal plate, as maintained by Koff [1], or whether the Wolffian ducts descend caudally and participate in vagina formation, as described by Witschi [4], remained an open question.
Rationale of the Experiment
The present experiment was designed to unequivocally prove downward migration of genital ducts during formation of the vagina, and androgen sensitivity in this process. To this end, pregnant carrier females of the Tfm mutation were treated with testosterone. In the offspring, androgen-insensitive Tfm embryos were compared with androgen-treated X/X embryos. Vaginal development was studied by 3D reconstruction of urethra and genital ducts, including the symphysis as an external landmark.
In Tfm/Y embryos, downward growth of genital ducts and formation of a vagina from the Müllerian ducts occurred as it did in normal females, although due to the AMH from the embryonic testes, the upper part of the newly formed Müllerian vagina secondarily regressed. In testosterone-treated X/X females, downward growth of genital ducts and formation of a vagina was halted far above the symphysis.
In contrast to a previous study [12], in the present experiment, embryos of the same litter in exactly the same stage of development are compared, and downward growth of the genital ducts during formation of the vagina is quantitatively determined. The comparison of androgen-treated females, representing the male phenotype, with androgen-insensitive Tfm embryos representing the female phenotype, allows the conclusion that the difference in the level of the vaginal bud is androgen-dependent.
In addition to the genetic approach, we describe the immunohistochemical localization of the androgen receptor (AR) during vaginal development in normal male and female embryos. A systematic description of AR along the entire embryonic genital tract with reference to the respective literature was given elsewhere [13]. Here we show that ARs are differentially expressed in Wolffian-derived or Müllerian-derived structures and identify the morphogenetically active Wolffian mesenchyme as the target tissue of androgens for the negative control of vaginal development.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In order to identify the embryos, inspect the gonads, and determine the effect of the X-linked coat color marker, Tabby (Ta) was used as outlined [14]. Carrier females (XTfm+Blo/X+Ta+) were mated with Tabby males (X+Ta+/Y). In the offspring, in Tabby homozygous (X+Ta+/X+Ta+) and hemizygous (X+Ta+/Y) embryos, vibrissae were heavily affected, in Tfm heterozygotes (XTfm+Blo/X+Ta+) single vibrissae were missing, whereas in Tfm embryos (XTfm+Blo/Y), the vibrissae pattern was normal.
Vaginal development was studied at Day 20 of pregnancy, shortly before birth. The lower half of embryos were fixed in Bouin solution, embedded in paraffin (paraplast), serially sectioned (5 µm), and stained with hematoxylin and periodic acid-Sciff to accentuate the basal membranes. From the serial sections, computer based 3D reconstructions were prepared after manually segmenting urethra and genital ducts using the MOP Videoplan equipment of Kontron (Eching/Munich, Germany). Contour lines of every fourth section were used and oriented along the course of the urethra. The distance between contour lines corresponds to 20 µm. For documentation, the hidden line representation of stacks of contour lines was shifted 45° in the sagittal plane and viewed from above at an angle of 50°.
Immunohistochemistry of AR was carried out as described [15]. NMRI mice between Days 17 and 18 of pregnancy were used. A total of seven male and eight female embryos were studied. After 4% paraformaldehyde fixation and paraffin embedding, serial sections (7 µm) were mounted on slides coated with aminopropyltriethoxysilane (APES; A3648; Sigma Chemical Company, St. Louis, MO). For antigen retrieval, the slides were microwaved for 30 min at 900 W under 0.01 M citrate-buffer pH 6.0 [16], cooled to room temperature, and rinsed with PBS pH 7.4. Endogenous peroxidase activity was blocked with 0.1% H2O2 in methanol for 15 min. The following steps were performed in droplets on the slides in a moist chamber: preincubation with normal goat serum for 20 min, incubation with the polyclonal antibody (clon PG-21) raised against human AR in rabbits [17] (Fa. Dianova, Hamburg, Germany) in a dilution of 1:40 for 1 h at room temperature after washing the incubation with biotinylated anti-rabbit immunoglobulin G (Unitect-ABC-Kit, Fa. Oncogene Science/Dianova, Hamburg, Germany) for 45 min at room temperature, incubation with avidin-biotin-peroxidase complex reagent for 45 min at room temperature, and development with diaminobenzidine for 5–10 min. As a negative control on each slide, one section was incubated with normal rabbit immunoglobulin G instead of the first antibody.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Pregnant carrier females heterozygous for the X-linked Tfm mutation were treated with a single dose of Testoviron at Day 18 of pregnancy, which had a depot effect and led to high testosterone levels over several days. Development of the vagina was studied 2 days later at Day 20 of pregnancy, shortly before birth. Tfm embryos were compared with testosterone-treated female littermates. Tfm embryos were insensitive to the testosterone treatment and, therefore, in this experiment they represented uninhibited (female) vaginal development. Females were sensitive to testosterone and exhibited the effect of testosterone on vaginal development, thus representing the male phenotype.
The stage of development before onset of testosterone treatment is schematically depicted in Figure 1. The respective real reconstruction at this stage, including Tfm embryos, is represented in [12]. In females, the Wolffian ducts have already regressed cranially, leaving only the caudal ends of the Wolffian ducts in place, corresponding to the Wolffian bulbs [3] or the sinovaginal bulbs [1]. Due to androgen insensitivity, the same is true for Tfm embryos. In Tfm mice, in addition, the Müllerian ducts have disappeared cranially due to the AMH produced in Tfm testes (not depicted in Fig. 1). The arrows indicate the outcome of the experiment as documented in Figure 2. The short arrow represents incipient descent of genital ducts in the female, which occurred before testosterone injection. The long arrow indicates descent of genital ducts in the Tfm animal, which was insensitive to the effect of testosterone.
|
|
Figure 2, a and b shows the reconstruction of urethra and vaginal anlage in a pair of Tfm and testosterone-affected X/X embryos at Day 20. Because both embryos were from the same litter, the stage of development was exactly the same. Sections were aligned along the urethra, assuming a straight course of the urethra in the reconstructed area. A standardized plane was taken to measure distance in the first section, in which the urethra and fusing genital ducts formed a coherent structure (Z1). The section in which right and left horns of the symphysis fused in the midline was taken as an external reference and designated as the zero plane for the measurements (Z0). In the testosterone-affected female (Fig. 2a), the fusion zone was located 420 µm above, and in the Tfm littermate (Fig. 2b) 120 µm below the symphysis, so that a testosterone-dependent difference of 540 µm had developed within 2 days. In the female embryo, rudimentary vesicular glands appeared at the base of the Wolffian ducts, and rudimentary prostate buds appeared in the regular male location beneath the bladder. The Müllerian ducts were fused in the midline, as occurs in a normal female. In the Tfm embryo, the Müllerian ducts had fused as they do in female development, and the cranial ends had regressed due to the presence of AMH from the Tfm testes. The cranial Wolffian ducts had regressed due to insensitivity to the trophic effects of androgens.
A total of three testosterone-treated pregnancies were studied, and in two of these, pairs of Tfm and testosterone-affected female embryos were encountered. In the second pair, in the female (Fig. 2c), the junction was 500 µm above, and in the Tfm (Fig. 2d), it was 420 µm below the symphysis, indicating that the fusion zone had migrated 920 µm in a caudal direction. The morphology of prostate buds and Wolffian ducts in the female indicated an earlier onset of the testosterone effect, and the more advanced regression of Wolffian ducts in the Tfm animal indicated a more advanced development at the time of reconstruction compared with the couple documented in Figure 2, a and b.
Immunohistochemistry of AR in Normal Mice
The genetic experiment using the Tfm mutation was supplemented by the immmunohistochemical localization of ARs in the fusion zone in normal NMRI male and female embryos.
In the sagittal section of a Day 18 male (Fig. 3), the transition was visible between the cranially differentiated Wolffian duct and its yet-undifferentiated thickened caudal end. Differentiation of the cranial part, which is induced by endogenous testosterone, was characterized by an increase of AR staining in the epithelial cell nuclei and a decrease of AR content in the mesenchymal wall layers. In contrast, the thickened caudal end of the Wolffian duct exhibited only a faint staining, but in the adjacent mesenchyme, AR staining was prominent. The vesicular glands arose cranially from the thickened end of the Wolffian duct. Caudally, the duct fused with the sinus ridge, which as the urethra, exhibited a high AR content. Figure 4 shows the fusion zone in a more medially located section from the same embryo, in which the caudal end of the Müllerian duct appeared. The Müllerian duct was AR-negative, but directly in contact with the AR-positive mesenchyme of the fusion zone. Due to the clear differences of AR expression, the epithelial components of the fusion zone could be clearly distinguished. Due to their heavy AR expression, some mesenchymal cells were observed in the fusion zone squeezing in between Wolffian and Müllerian epithelium, indicating the cranial emergence of separated ducts.
|
In the Day 17.5 female, the pattern of AR expression was in principle the same as in the male, although the intensity was in general much lower. Figure 5 shows a cross-section through the fusion zone. As in the male, totally negative Müllerian ducts, faint nuclear staining of Wolffian ducts, prominent expression of AR in urethra and sinus ridges, and relatively heavy expression in the mesenchyme were observed. In the female, the Wolffian ducts had disappeared above the fusion zone so that at this level, only AR-negative, well-developed Müllerian ducts with their also negative wall layers were present. Cranial remnants of AR-positive Wolffian ducts and positive mesenchyme in the Wolffian body appeared at the level of the ovary (not shown). Figures 6 and 7 depict oblique cross-sections from a Day 18 NMRI female corresponding to stage Day 19 as reconstructed in [12]. The cranial section (Fig. 6) showed AR expression in the female urethra and in the surrounding mesenchymal layers, but no expression in the fused Müllerian ducts that formed the definitive vagina. The caudal section (Fig. 6) is shown through the fusion zone between the remnants of the Wolffian ducts and the sinus ridges below the Müllerian ducts. The right side shows the opening of the Wolffian duct into the sinus ridge. Urethra and sinus ridges expressed AR. Prominent AR immunoreactivity was present in the surrounding mesenchyme, in particular in the lateral mesenchymal ridges, which marked the longitudinal line of separation between urethra and vaginal anlage, and formed a whorl of mesenchyme indentation between the caudal crura of the fusing sinus ridges.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
At the end of the indifferent stage of development, the upper parts of the Wolffian and Müllerian ducts have developed as separate entities which, with the rise of AMH and testosterone, differentiate into the male and female genital tracts, respectively. The caudal ends of the genital ducts are still in a state of morphogenesis, which during male development, comes under the control of testosterone, which precludes female development in the fusion zone.
The Müllerian ducts, which arose as an invagination of the celomic epithelium and had grown downward inside the basal lamina of the Wolffian ducts [18, 19], reach the fusion zone of Wolffian ducts with the urogenital sinus. The tips of the Müllerian ducts are still located inside the basal lamina of the Wolffian ducts. The Wolffian ducts, in turn, are also in the process of downward migration along the dorsal wall of the sinus, which followed the separation of the ureters. The surge of androgens secreted from the embryonic testes induces male morphogenesis via the AR-positive mesenchyme, which entails a stop signal for further downward migration of the genital ducts. Without testosterone, Wolffian and Müllerian ducts continue their descent along the sinus ridges. As in the indifferent stage of development, separate male and female ducts with mesenchymal wall layers continuously arise from the fusion zone.
AMH and Androgen
In the male, Müllerian ducts regress under the influence of AMH, which acts in the mesenchyme and leads to contraction of the surrounding wall layers. In females, the Wolffian ducts undergo apoptosis when the trophic effect of testosterone fails to occur. The trophic effect is elicited via ARs in the Wolffian epithelium. Thus, in both cases, as a prerequisite, regression needs fully developed ducts composed of an epithelial tube surrounded by mesenchymal wall layers [20]. In Tfm animals, the Wolffian ducts degenerate due to androgen insensitivity. In addition, the cranial portions of the Müllerian ducts have disappeared due to the AMH produced by the testes. Nevertheless, a vagina develops. The Müllerian-derived vagina secondarily degenerates following the degeneration of the Wolffian ducts in a considerable distance. This process finally determines the size of the highly variable vaginal pocket in the adult Tfm phenotype [12].
Effects of Androgens on Morphogenesis
In the caudal, thickened ends of the Wolffian ducts corresponding to Koff's sinovaginal bulbs, AR expression is low in the epithelium and high in the mesenchyme, whereas cranially, with the onset of differentiation, AR expression is high in the epithelium of the Wolffian duct and it disappears from the adjacent mesenchymal layers. This difference in AR expression in the lower Wolffian ducts was not noted in other autoradiographic and immunohistochemical studies, and might be the basis for the morphological separation as sinovaginal bulbs from the Wolffian ducts by Koff.
The mesenchymal AR expression around the thickened ends of the Wolffian ducts is the basis for the morphogenetic effects of androgens. Cranially, not only the Müllerian ducts, but also their wall layers, are always AR negative. In the fusion zone, the AR-negative Müllerian epithelium comes into direct contact with the morphogenetically active AR-positive mesenchyme. A functional explanation for the participation of the Wolffian ducts in the formation of the vagina is mediation of the stop signal from the AR-positive Wolffian to the AR-negative Müllerian structures.
The negative effect of androgens on vagina formation is mediated by the AR-positive Wolffian ducts and adjacent morphogenetically active mesenchyme, thus explaining their participation in development, as was already suspected by Witschi [4] on the basis of teratological evidence of his time and the virilization of the duct system in the female human fetus after prenatal exposure to androgens. The teratogenic effect of prenatal androgens was systematically studied in rats [21]. The rather low dose of 1 mg of testosterone propionate administered daily to the mother was sufficient to cause complete vaginal atresia.
Recently [22], the immunohistochemical localization of AR was described in vaginal development of female human fetuses. Those authors state, "Androgen receptors were expressed in the epithelium and the stroma of the urogenital sinus, sinovaginal bulbs, and mullerian and wolffian ducts." The notion that, in contrast to the data for mice, AR is expressed in the Müllerian ducts, needs a more detailed analysis because in the figures presented in the study, a Wolffian duct is shown cranially from a Müllerian duct and separate from an evagination of the urogenital sinus, which is termed sinovaginal bulb.
Role of Sinus Ridges
The Wolffian ducts are connected to the dorsal wall of the urethra by the sinus ridges. The sinus ridges are vestigial structures that occlude the originally patent opening of the Wolffian ducts (as excretory ducts of the mesonephros) and serve as guiding structures for further downward migration of the Wolffian ducts. Separation of the vaginal canal from the urethra occurs by abscission of the sinus ridges via formation of longitudinal ridges of condensed mesenchyme, preceding the downward migration of the genital ducts. AR expression in the mesenchyme indicates that the abscission of sinus ridges is also under negative control of androgens. AR expression in the whole sinus epithelium indicates its plasticity to form alternatively the respective male or female phenotype.
A discrepancy seems to exist between the present observations on the derivation of the vagina and a tissue recombination study demonstrating that the lower sinus-derived portion of the vagina of the newborn mouse, in contrast to the Müllerian-derived upper portion, has the competence to form the prostate gland [23]. As shown in sagittal sections and 3D reconstruction, downward migration of genital ducts continues until Postnatal Day 4 [12], so that in newborn mice, the nonpatent, lower portion of the vagina still comprises the fusion zone of Wolffian ducts and sinus ridges (compare Fig. 2b). This explains the competence of the tissue to form prostate.
Evolutionary Aspects
The complicated development of the vagina and the dominant role of androgens can be interpreted within evolutionary aspects. Formation of a vagina is a late event in phylogeny and occurs at the transition stage in mammals together with the establishment of embryonic development inside the female genital tract. In the female, it results in complete anatomical separation of the birth canal from the urethra, whereas in the male, a common urogenital canal persists. In marsupial and eutherian mammals, different solutions for the separation of the birth canal evolved. In marsupials, two lateral vaginas and a temporary median birth canal are present [24]. In both branches of mammals, the indifferent stage of the genital ducts developing in the embryonic period is similar [25]. In marsupials, further sexual differentiation occurs after birth during pouch life [26], and expression of AR and secretion of testicular androgens are not synchronized [27]. In eutherian mammals, sexual differentiation continues inside the female genital tract and extends into the fetal period. With the prolongation of intrauterine life, estrogens were eliminated as a regulatory factor in female sexual morphogenesis, probably because of interference with maternal steroid hormones, and androgens acquired a dominant role. As shown here, downward growth of genital ducts and consecutive formation of a Müllerian vagina came under negative control of androgens.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received: 12 December 2001.
First decision: 13 January 2002.
Accepted: 26 April 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Basha, S. Chang, E. M. Smolock, R. S. Moreland, A. J. Wein, and S. Chacko Regional differences in myosin heavy chain isoform expression and maximal shortening velocity of the rat vaginal wall smooth muscle Am J Physiol Regulatory Integrative Comp Physiol, October 1, 2006; 291(4): R1076 - R1084. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Yin, C. Lin, and L. Ma Msx2 Promotes Vaginal Epithelial Differentiation and Wolffian Duct Regression and Dampens the Vaginal Response to Diethylstilbestrol Mol. Endocrinol., July 1, 2006; 20(7): 1535 - 1546. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kobayashi, K.-M. Kwan, T. J. Carroll, A. P. McMahon, C. L. Mendelsohn, and R. R. Behringer Distinct and sequential tissue-specific activities of the LIM-class homeobox gene Lim1 for tubular morphogenesis during kidney development Development, June 15, 2005; 132(12): 2809 - 2823. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
