| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Zoology, The University of Melbourne, Victoria 3010, Australia
ABSTRACT
Estrogens have a feminizing effect on gonadal differentiation in fish, amphibians, reptiles, and birds. However, the role of estrogen during gonadal differentiation in mammals is less clear. We investigated the effect of estrogen on gonadal differentiation of male tammar wallabies. Male pouch young were treated orally with estradiol benzoate or oil from the day of birth, before seminiferous cords develop, to Day 25 postpartum and were killed at Day 50 postpartum. In all estrogen-treated neonates, a decrease in gonadal volume, volume of the seminiferous cords, thickness of the tunica albuginea, and number of germ cells was found. The stage of treatment affected the magnitude of the response. Two of three male young born prematurely after 25 days of gestation and treated subsequently with estradiol had ovary-like gonads, with well-developed cortical and medullary regions and primordial follicle formation. Furthermore, at Day 50 postpartum, many (21%) of the germ cells in these sex-reversed ovaries were in the leptotene and zygotene stages of meiosis, similar to female germ cells at the same stage of development. In the other males born on Day 26 of gestation or later, estradiol treatment from the day of birth caused development of dysgenetic testes, with abnormal Sertoli cells, atrophy of the seminiferous tubules and tunica albuginea, and absence of meiotic germ cells. In this marsupial, therefore, estradiol can induce either partial or complete transformation of the male gonads into an ovary with meiotic germ cells. These results confirm that estrogen can inhibit early testicular development, and that testis determination occurs during a narrow window of time.
developmental biology, early development, estradiol, ovary, testis
INTRODUCTION
Mammalian gonadal differentiation is a complex process initiated during early fetal life. Sex is determined by the presence or absence of the testis-determining gene SRY, which directs the indifferent gonad to form a testis that produces testicular hormones. Mutations in genes in the sex-determining cascade can result in gonadal sex reversal in the human and mouse [1], but the occurrence of ovotestes in mammals is quite rare [2]. The timing and level of SRY expression is critical in transgenic mice [3], suggesting that the male pathway must be initiated in a narrow window of time during gonad development to suppress the female pathway [2, 4]. The supporting cell lineage is critically dependent on SRY expression, and testis formation appears to depend on sufficient numbers of Sertoli cells differentiating within this specific time frame. Otherwise, the supporting cells differentiate as prefollicular cells, and gonadal development proceeds along the female pathway [5]. Male differentiation can be induced in XX gonads in "sandwich" cultures with XY gonads, but only before 12.5 days postcoitum in the mouse, again supporting the idea of a brief period of sensitivity when testicular differentiation can be initiated [6].
Sex reversal is more readily achieved in nonmammalian vertebrates, and estrogen administration has profound effects on gonadal differentiation. Sex reversal with development of ovaries occurs after treatment with estradiol in male rainbow trout [7], reedfrog [8], reptiles [9], chicken [10], and zebra finch [11, 12]. Conversely, inhibition of estrogen production or action in fetuses that would otherwise develop as females results in testis development in fish [13, 14], reptiles [9], and birds [10–12]. These findings indicate that in lower vertebrates, estrogen has a pivotal physiological role in gonadal differentiation.
Equivalent treatment of eutherians is difficult, because all gonadal differentiation takes place in utero. The administration of estrogen to pregnant female eutherians did not prevent testicular differentiation of the male mouse fetus [15], although male newborn mice that had been treated from Day 11 to Day 17 of pregnancy had reduced seminiferous tubular formation, decreased number of Sertoli cells, decreased size of Leydig cells, and a reduction of the gonocyte degeneration that normally occurs at term [16]. However, because the maternally administered steroids had to cross the placenta, the amount of estrogen reaching the fetuses is unknown.
In marsupials, the onset of testicular differentiation occurs at approximately the time of birth, so the young are readily accessible in the pouch and do not need a complicated intrauterine surgical approach to manipulate the fetus. Marsupials have homologs of SRY on the Y chromosome and homologs of most genes in the eutherian sex-determining cascade [17]. These mammals provide a unique model with which test the interactions between genes and hormones during early gonadal development [18]. Many aspects of gonadal differentiation are similar between marsupials and eutherians [19], and ovarian sex reversal occurs in both [20]. In the tammar wallaby, seminiferous cord-like structures form after ovaries are either transplanted into male hosts or cultured with recombinant Müllerian-inhibiting substance (MIS) [21]. Exposure to MIS appears to be toxic to female germ cells, but all cases of XX gonadal sex reversal in mammals involve a loss of the germ cells, suggesting that XX germ cells may inhibit testicular differentiation [2, 5, 20, 21].
In three different marsupial species, administration of estrogen interferes with testicular differentiation, resulting in a range of phenotypes. In studies of the American opossum (Didelphis virginiana) by Burns [22–25] more than 60 years ago, treatment of males with estradiol dipropionate for 30 days postpartum induced development of an ovary-like structure, with well-developed cortical and medullary regions. Treatment of the male short-tailed gray opossum (Monodelphis domestica) with estrogen at Days 1 and 3 postpartum impaired testicular development, so that at 22 wk postpartum, these were indistinguishable from the testes of an animal at 1 day postpartum [26]. Similarly, estrogen treatment from birth up to 9 days postpartum prevented sex-cord development and impaired testicular development [27]. In contrast, treatment of tammar wallaby (Macropus eugenii) pouch young with estradiol benzoate for the first 25 days postpartum induced only partial sex reversal, in which the testes contained areas devoid of tubules, resembling the ovarian cortex [28].
In the distantly related American opossum, complete sex reversal was achieved only if the neonatal treatment was initiated in young that were born after the shortest possible gestational period (12.5 days), some 12 h before the average gestation of 13 days. The testes are undifferentiated in those young that are born early, whereas males with long gestations have identifiable seminiferous cords by the time of birth. In the gray short-tailed opossum, testicular differentiation is variable on the day of birth [27, 29, 30], probably due to variations in gestational lengths [31]. In the tammar wallaby, most newborn males have undifferentiated testes, but because the day of birth can range from 25 to 28 days of gestation (mean ± SEM, 26.5 ± 0.4) [32], the stage of development and the state of testicular differentiation also vary on the day of birth [31, 33–36]. Collectively, these observations suggest that the critical window of sensitivity for testicular development in the marsupial is around the time of birth.
Treatment of marsupial pouch young with estradiol also affects germ cells. In the ovary-like testes of estradiol-treated American opossums, male germ cells had entered prophase 1 of meiosis [22]. In the estradiol-treated tammar wallaby pouch young, both male and female germ cells were still in mitosis when the experiment was terminated at 25 days postpartum [28]. The current study, therefore, was designed to induce gonadal sex reversal of tammar male young by treatment with estradiol benzoate from the day of birth to Day 25 postpartum and to allow additional time for gonadal development to determine whether estrogen can induce complete gonadal sex reversal in testes. We also designed experiments to define the developmental stage critical for sex reversal. Treated young were allowed to grow to Day 50, when normal ovarian germ cells have entered meiosis [37], but when testicular germ cells are in mitotic arrest (unpublished observations), to determine whether the XY germ cells would behave as XX germ cells in sex-reversed testes
MATERIALS AND METHODS
Animals
Tammar wallabies of Kangaroo Island origin were held in open grassy yards at our breeding colony in Melbourne, Victoria, Australia. Food was supplemented with lucerne cubes, oats, and fresh vegetables. Tammar wallabies are seasonal breeders and have a postpartum estrus that results in a diapausing blastocyst in the uterus while a pouch young is suckled. Removal of the pouch young (RPY) causes reactivation of the quiescent blastocyst, and birth occurs after an active pregnancy of, on average, 26.5 ± 0.4 days [32]. Animals were checked for birth daily from Day 24 to Day 34 after RPY to provide information on the duration of "active" gestation, and the day of birth was termed Day 0 postpartum. The pouch young were examined with an otoscope to determine sex by the presence of scrotal bulges or mammary primordia [33].
Estrogen Treatment
Male pouch young were treated orally from Day 0 until Day 25 with estradiol benzoate (n = 11) in arachis oil (1.2–2.5 mg kg-1 day-1; Organon Labs, Morden, UK) or with arachis oil only (six males and four females) as described previously [28]. The amount of estradiol benzoate was adjusted every 5 days to accommodate the rapid growth at this stage of development (Table 1). At Day 50, the young were killed. One gonad was removed and fixed in Bouin fixative for 24 h, washed in three changes of 70% ethanol, paraffin embedded, serially sectioned at 5 µm, and stained with hematoxylin and eosin. The other gonad was removed, fixed in 4% paraformaldehyde, and embedded in embedding compound (Bayer) for other studies.
|
Gonadal Volume
The volume of the gonads and proportions of the interstitium, tunica albuginea, seminiferous cords, cortex, and medulla were calculated in histological sections by the point-counting procedure [38, 39] using a coherent square test system (1[p] = 10 mm, a[p] = 100 mm2) [40] on a Zeiss (Jena, Germany) Axiovet 35 M inverted microscope equipped with a Sony (Tokyo, Japan) high-resolution video system.
Germ Cell Populations
Primordial germ cells were identified based on the large size and ovoid shape, pale-staining cytoplasm, and vesicular nucleus [37, 41, 42]. Germ cells were classed as either mitotic or meiotic (i.e., prophase 1) using the criteria described by Alcorn [43]. Germ cells were counted in three fields of view on every 20th section on a Leitz (Stuttgart, Germany) photomicroscope using a 1.3 NA, 100x, oil-immersion objective with a 12.5;ts eyepiece. A counting grid was projected onto the section. The first field was chosen arbitrarily. After counting, the field of view was moved in a horizontal direction (0.5 mm) to a second position and, after counting again, was then moved in a vertical direction (0.5 mm) to a third position. Only those germ cells with a prominent, visible nucleolus were counted. Total germ cell population (GCPop) in the gonad was estimated by
where GCCount is the number of germ cells counted per gonad, SCount is the number of sections counted per gonad, VCount is the volume per section counted, and VGonad is the total volume of the gonad.
Statistics
Treatment and control groups were compared using two-sample t-tests. Volume fractions of the gonads were arcsine transformed before analysis [44]. Total germ cell population numbers were square-root transformed before analysis.
RESULTS
"Active" Gestational Length
Birth of male pouch young occurred after a gestation of 25 (n = 3), 26 (n = 4), 27 (n = 2), 28 (n = 1), and 29 days (n = 1) after RPY in treated animals and on Day 25 (n = 1), 26 (n = 2), and 27 (n = 3) in control animals (Table 2).
|
Morphology of Control Gonads
Ovaries of Day 50 females were well developed, and the cortex contained abundant germ cells clustered together (Figs. 1C and 2, A and C) as well as a surface epithelium of cuboidal and columnar cells (Fig. 2, A and C). Columns of granulosa cells, identified based on the characteristic dark-staining, elongated nuclei with a central nucleolus, separated the germ cell clusters into columns that radiated from the medulla (Fig. 2A). A layer of connective tissue separated the cortex and medulla, and strands of the connective tissue projected into the cortex. The rete cords of the ovarian medulla contained cells with dark-staining nuclei, and the cords were surrounded by thin bands of connective tissue. Ovarian germ cells were observed in clusters within the cortex, and primordial follicles were present close to the medullary region. Forty-four percent of germ cells were in prophase 1 of meiosis, with the remainder in interphase or mitotically dividing (Table 3).
|
|
|
The testes of Day 50 control males were larger than ovaries at the same age (Figs. 1 and 3B). The tunica albuginea of the testis consists of a thick band of connective tissue beneath a squamous surface epithelium (Fig. 4C). The seminiferous cords of the testes were well developed (Figs. 1A and 4, A and C), and Sertoli cells lined the seminiferous cords, with basal nuclei and strongly eosinophilic cytoplasm (Fig. 4, A and C). Most germ cells were in the center of the seminiferous cords, although some were between the cords. Most germ cells were in mitotic interphase, and a few were in mitosis (Table 3). None of the testicular germ cells was meiotic. Thin, elongated, peritubular myoid cells around the cords were oriented perpendicular to the Sertoli cells (Figs. 4, A and C, and 5A). Interstitial tissue between the cords was comprised of connective tissue, blood vessels, and Leydig cells, with characteristic cytoplasm.
|
Morphology of Estrogen-Treated Testes
The testes of estradiol-treated, Day 50 males had two distinct phenotypes, which were designated as dysgenetic testes (n = 9) or sex-reversed (i.e., ovary-like) testes (n = 2). Both neonates with sex-reversed testes were born on Day 25 after RPY, whereas eight of the nine neonates that developed dysgenetic testes were born on Day 26 or later (Table 2).
The nine estrogen-treated males with dysgenetic testes had disorganized seminiferous cords, with fibrous tissue in regions directly adjacent to the surface epithelium. Nuclei were not in the basal part of the Sertoli cells, as they were in the control testes (Fig. 4, B and D). In these disorganized areas, Sertoli cells and germ cells were scattered in the interstitium, and no peritubular myoid cells could be identified. In the remainder of the testis, the seminiferous cords were similar to those in controls. The tunica albuginea was undeveloped or absent, and the interstitium terminated at the surface epithelium (Fig. 4, B and D). The surface epithelium contained squamous, columnar, and cuboidal cells. Dysgenetic testes were smaller than the controls (Fig. 1B and 3B) due to a decreased seminiferous cord and tunica albuginea volume (Fig. 6A). In contrast, the volume of interstitial tissue remained the same.
|
The remaining estrogen-treated young were both born early on Day 25 after RPY. At Day 50, the gonads in these two males lacked seminiferous cords but had a characteristic, ovarian-like medulla surrounded by an ovarian-like cortex (Figs. 1D and 2B) that contained clusters of germ cells between columns of dark-staining cells with elongated nuclei similar to those of granulosa cells in control ovaries (Fig. 2, B and D). Some germ cells, close to the medulla, were surrounded by prefollicle or granulosa-like cells, and they appeared to form primordial follicles. The cortical cells of these transformed testes had dark-staining cytoplasm and elongated nuclei with central nucleoli, and these cells radiated from the medulla to the surface epithelium, dividing the germ cell clusters. The cortex was separated from the medulla by a thick band of connective tissue (Figs. 1D and 2B). The medullary cords (i.e., rete cords) in the center of the sex-reversed testis contained cells with dark-staining nuclei, and the interstitium between the medullary cords was small and contained thin bands of dark-staining cells. These sex-reversed gonads were similar in volume to the control ovaries (Fig. 3B). The volumes of cortex and medulla were also similar to those of control ovaries (Fig. 6B), and instead of a tunica albuginea, the capsule consisted of cells with dark-staining nuclei under a stratified, cuboidal surface epithelium (Fig. 2D).
|
Germ Cell Analysis
In the dysgenetic testes, most germ cells were located in the remaining seminiferous cords, but some were present in the interstitium. The total number of germ cells decreased (P < 0.05) (Fig. 4A), and all were in interphase or mitosis (Table 3 and Fig. 5). In contrast, 21% of the germ cells in the cortex of the completely sex-reversed testes were in prophase 1 of meiosis (Table 3 and Fig. 5A) based on the appearance of characteristic leptotene and zygotene features (Fig. 5). The remainder were in interphase or in mitosis (Table 3 and Fig. 5).
|
DISCUSSION
Male-to-female gonadal sex reversal can be induced in marsupials by treating neonatal males with estradiol. Depending on when treatment commences, the reversals range from partial suppression of seminiferous tubule development (i.e., dysgenetic testes) to formation of a morphologically normal ovary with cortex, medulla, and apparently healthy meiotic and mitotic germ cells (i.e., complete sex reversal). These results confirm those of early studies on the American opossum [22, 23], and they support the evidence found in eutherians from estrogen receptor and aromatase knockout (ERKO and ArKO, respectively) mice that estrogen plays a critical role in mammalian male gonadal development [45].
Although estrogens do not cause sex reversal in gonads of eutherian males, they do influence sexual development [46]. Double gene knockouts of estrogen receptors (ER
and ERß) cause the ovarian transdifferentiation in postpubertal female mice [47]. Estrogen is also essential in male development. Estrogen receptors are distributed widely among male reproductive tissue [45], and semen and rete testicular fluids contain significant concentrations of estrogen [48]. The ER and ERß are expressed in fetal testes, with ERß being predominant in spermatogonia and ER predominant in the Leydig cells in the mouse [49] and human [50]. Aromatase is also expressed in mouse testes [51], and in ArKO mice, spermatogenesis is disrupted, suggesting a direct action of estrogen in male germ cell development in the mouse [52]. Thus, estrogen appears to act locally as a hormone in the male [52], and it may have important roles in mammalian sexual differentiation and in germ cell development.
Estrogen treatment in the tammar wallaby reduces the volume of the testicular cords, tunica, interstitium, and germ cell components of the testis. Tubular breakdown may be the result from loss of interaction between the peritubular myoid and Sertoli cells. Separation between Sertoli and peritubular cells in culture by a filter inhibits tubule formation [53–55]. Furthermore, mesonephric cells contribute to the peritubular myoid cell population [56–58] and induce mesospheric cell migration into an XX gonad, resulting in cord formation [6]. Thus, interactions between Sertoli and peritubular cells are essential for the formation of the basal lamina of the seminiferous cords, and estrogen may interfere with this process. Both Sertoli and peritubular cells of mice express ER [59].
Morphological change of the testis when exposed to estrogen is influenced by the developmental stage of the young at the time of birth (and by the commencement of estrogen administration). Two of the three animals born on Day 25 of gestation had complete histological transformation of testes to ovaries, as compared to those born on Day 26 or later, which had dysgenetic testes. The low number of Day 25 births is a direct result of the natural variation in gestational length. Most births occur at Day 26 or later, but a very small number occur "prematurely" on Day 25 of active gestation [32, 60]. Our early study [28] induced partial feminization of the testis, but re-examination of our data indicates that none of these animals was born on Day 25. In addition, all those animals were examined at Day 25 after birth, whereas the animals in the present study were allowed to grow until Day 50.
In opossums, the developmental stage at the start of treatment is critical. The first sign of testicular development in the opossum is at stage 35 on Day 13 of gestation, when birth normally occurs [61]. Administration of estrogen to animals born on Day 12.5, at stage 34 (
12 h earlier), caused complete gonadal sex reversal, including formation of primary follicles. In contrast, estrogen administration that commenced 12 h later disrupted testicular development but did not induce ovary-like structures [23, 24]. Thus, estrogen-induced testicular sex reversal occurs when estrogen is administered before the first signs of testicular differentiation, namely, before the development of seminiferous cords. If given after the male pathway is initiated, a variable degree of disruption of testicular morphogenesis occurs.
The number of germ cells in the testes of treated young is reduced. Although male (XY) germ cells do not normally enter meiosis until puberty, in the estrogen-treated testes, many XY germ cells that were surrounded by follicle-like cells had entered meiosis, as female (XX) germ cells do at this same age (i.e., Day 50 postpartum) [37]. In the dysgenetic testes, the small number of germ cells in the cords remained in mitosis. This suggests that testicular cords inhibit the entry of germ cells into meiosis within them. Likewise, in mice, XY germ cells at ectopic locations, such as the adrenal, can enter meiosis, whereas those situated in the testis enter mitotic arrest [62, 63]. Mouse XY germ cells can enter meiosis when cultured in lung and mesonephric tissue, but not in the testis [64, 65]. In addition, XY mouse germ cells enter meiosis in the gonad of XX 
XY chimeras and XXSxr males in areas where tubular disruption is evident [66–68].
Estrogen appears to play a physiological role in gonadal differentiation in a variety of nonmammalian vertebrates [69–71]. Whereas administration of estrogens to reptiles with temperature-dependent sex determination at male-producing temperatures induces ovarian differentiation, administration of aromatase inhibitors at female-producing temperatures induces testicular differentiation [69–72]. Likewise, in birds, treatment of male embryos with estrogens causes ovarian development, whereas treatment of female embryos with aromatase inhibitors causes testicular development [10–12]. Aromatase is expressed in a sexually dimorphic pattern in the chicken [73], suggesting that in these species, estrogen is a key hormone in female gonadal development.
Estrogens also have profound effects on development and function of the gonads in eutherian mammals, as shown in ERKO and ArKO mice [52, 74]. Thus, interactions between estrogen, estrogen receptor, and the genes on the sex-determining pathway may be fundamental to vertebrate sexual differentiation in mammalian and nonmammalian vertebrates. The apparent ease with which marsupial testes can be sex reversed may be explained in two ways. First, differentiation of marsupial gonads may be more plastic than in eutherians. More likely, however, and as suggested by McLaren [5], is that commitment of the male gonad to form a testis occurs during a narrow window of time, and disturbance of the signals leading to testicular differentiation during this period can lead to ovarian development in males. In eutherian mammals, this narrow window occurs in the fetus in utero, buffered by the maternal system, but in marsupials, gonadal differentiation commences at approximately the time of birth. The treatment of neonatal marsupial young with exogenous estradiol during this window of sensitivity allows the observation of effects that are difficult to achieve in a eutherian fetus in utero. Because mice with ER and ERß knockouts develop ovaries that later transdifferentiate into testes, estrogen is not essential for initial formation of the ovary, as it is in reptiles and birds. In marsupials, no evidence has been found, to our knowledge, for the production of estrogen by either the ovary or the testis during gonadal differentiation [75], so the observed effects of estrogen in male gonadogenesis may be a result of interference with the normal testicular pathway of differentiation leading to ovarian development.
ACKNOWLEDGMENTS
We thank Patrick Jackson, Chris Nave, Deidre Mattiske, Cyrma Hearn, and Richard Moyle for help in animal handling. We also thank Drs. J.D. Wilson and Andrew Pask for helpful criticism of the manuscript. Animals were collected under permits from South Australian National Parks and Wildlife and were held under permit RP-95-088 of the Department of Natural Resources and Environment, Victoria, Australia.
FOOTNOTES
First decision: 24 January 2001.
1 Supported by grants from the Australian National Health and Medical Research Council. 
2 Correspondence. FAX: 61 3 9348 1719; m.renfree{at}zoology.unimelb.edu.au
Accepted: March 21, 2001.
Received: December 18, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  G.-C. Wu, S. Tomy, M. Nakamura, and C.-F. Chang Dual Roles of cyp19a1a in Gonadal Sex Differentiation and Development in the Protandrous Black Porgy, Acanthopagrus schlegeli Biol Reprod, December 1, 2008; 79(6): 1111 - 1120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Vizziano, D. Baron, G. Randuineau, S. Mahe, C. Cauty, and Y. Guiguen Rainbow Trout Gonadal Masculinization Induced by Inhibition of Estrogen Synthesis Is More Physiological Than Masculinization Induced by Androgen Supplementation Biol Reprod, May 1, 2008; 78(5): 939 - 946. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Pannetier, S. Fabre, F. Batista, A. Kocer, L. Renault, G. Jolivet, B. Mandon-Pepin, C. Cotinot, R. Veitia, and E. Pailhoux FOXL2 activates P450 aromatase gene transcription: towards a better characterization of the early steps of mammalian ovarian development. J. Mol. Endocrinol., June 1, 2006; 36(3): 399 - 413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. H-C Yao and B. Capel Temperature, Genes, and Sex: a Comparative View of Sex Determination in Trachemys scripta and Mus musculus J. Biochem., July 1, 2005; 138(1): 5 - 12. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. BRITT, J. KERR, L. O'DONNELL, M. E. E. JONES, A. E. DRUMMOND, S. R. DAVIS, E. R. SIMPSON, and J. K. FINDLAY Estrogen regulates development of the somatic cell phenotype in the eutherian ovary FASEB J, September 1, 2002; 16(11): 1389 - 1397. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
