|Home | About | Journals | Submit | Contact Us | Français|
The general perspective of ovary organogenesis is that the ovary is the default organ which develops in the absence of testis-promoting factors. Testis formation, on the other hand, is a male-specific event promoted by active components that override the default ovarian process. However, when comparing the sex determination mechanism among different vertebrate species, it is apparent that this default view of ovary formation can only be applied to mammals. In species such as reptiles and birds, ovary formation is an active process stimulated by estrogen. Remnants of this estrogen-dominant pathway are still present in marsupials, a close relative of eutherian mammals, like humans and mice. Although initial formation of the mammalian ovary has become strictly regulated by genetic components and is therefore independent of estrogen, the feminizing effect of estrogen regains its command in adult ovaries. When estrogen production, or its signaling, is inhibited, transdifferentiation of ovarian tissues to testis structures occur in adult females. Taken together, these observations prompt us to reconsider the process of ovary organogenesis as the default organ and question if testis development is actually the default pathway.
Bisexual development of the gonads is a universal phenomenon in vertebrates. The common gonadal primordium, a precursor for both testis and ovary, arises on the surface of the mesonephros. Before dimorphic differentiation occurs, the gonadal primordium is morphologically indistinguishable in male and female embryos and is pliable in terms of its developmental potential. Although initial formation of the gonadal primordium is conserved across species, diverse mechanisms have evolved to trigger its dimorphic differentiation. In most reptilian species, where embryos develop outside the mother, egg incubation temperature determines the fate of the gonad. Birds, a close relative to reptiles, have evolved a genetic mechanism where chromosomal composition becomes the sex-determining factor. Although reptiles and birds use different mechanisms to trigger gonadal differentiation (temperature vs. chromosome), they both require estrogen to facilitate the formation of the ovary. However, in mammals such as marsupials and eutherians, estrogen becomes dispensable for ovary formation as genetic regulation takes full control. In this review, we provide a comparative view on the mechanisms of ovary organogenesis in four vertebrate species: reptiles, birds, marsupials and mammals. Our focus is primarily on the regulatory mechanisms and their relation to the morphological changes in ovary development.
Reptiles, such as crocodilians and most turtles, lack sex chromosomes. In these animals, the sex of the embryo is adaptable to environmental changes, particularly egg incubation temperature.1 This type of sex determining mechanism is referred to as temperaturedependent sex determination (TSD) and is used by many species of oviparous reptiles.2 Although the mechanisms of TSD are similar among reptiles, ranges of temperature that trigger gonadal differentiation differ tremendously among species. We will focus on two species, the American alligator, (Alligator mississippiensis) and the red-eared slider turtle (Trachemys scripta).
In the American alligator, development of the ovary occurs exclusively at egg-incubation temperatures below 30°C and above 34.5°C, whereas incubation temperatures between 32.5°C and 33°C yield only testes.3 In the red-eared slider turtle, incubation of eggs at high temperatures (31°C) results in ovarian development, while those incubated at lower temperatures (26°C) produce testes (Table 1). Eggs incubated at intermediate temperatures produce a mixed ratio of sexes.4 While the sex-determining temperatures differ between these reptiles, they both possess a window of time when sex is susceptible to temperature change. This period is referred to as the thermosensitive period. The thermosensitive period was determined by the classic temperature-shifting experiments where eggs incubated at the female-producing temperature were switched at different time points during development to the male-producing temperature or vice versa. Shifting eggs from the female-producing temperature to the maleproducing temperature early in the thermosensitive period produced 100% male offspring. However, as the incubation period progressed and eggs were switched from the female to the maleproducing temperature, fewer offspring became male.3,4 These results indicate that gonadal differentiation is pliable early in the thermo- sensitive period, but sex becomes irreversibly committed as the thermo- sensitive period progresses.
The mechanism describing how temperature elicits its effects on ovarian organogenesis has gradually emerged. Estrogen and aromatase, the key enzyme for the conversion of androgen to estrogen, have been implicated as effectors downstream of the temperature stimulus to trigger ovarian differentiation (Table 1). In the American alligator, estrogen administration to eggs incubated at the male-producing temperature resulted in 100% female offspring.5 Similar results were seen in the red-eared slider turtle; estrogen feminized embryos incubated at the male-producing temperature.6 On the other hand, blocking estrogen synthesis by aromatase inhibitors caused female to male sex reversal at the female-producing temperature in both the alligator and turtle5,7 (Fig. 1). These studies demonstrate that estrogen has the ability to promote ovarian development while overriding the testicular pathway. Although the site and timing of estrogen production in female alligator embryos remains controversial, it is clear that estrogen acts on the cortical and medullary regions of the developing ovary.2,8
In these reptiles, both male and female embryos initially develop primitive sex cords in the gonads. In the male gonad, these sex cords become prominent in the medullary region of the gonad, engulf the developing germ cells, and eventually develop into seminiferous tubules. In contrast, sex cords in the female gonad degenerate and become fragmented during the thermosensitive period due to a decrease in medullary cell proliferation.9,10 The cortex of the ovary thickens as a result of proliferation of germ cells and pre-granulosa cells11 (Fig. 1). Estrogen treatment in alligators and turtles at the male-producing temperature induced ovarian transformation, including proliferation and entry of germ cells in the cortex into meiosis, and at the same time, degeneration of sex cords in the medullary region.12,13 At the male-producing temperature (no estrogen production) or when estrogen production is inhibited at the female-producing temperature, the gonadal cortex fails to differentiate and sex cords are allowed to expand, producing testis characteristics (reviewed in ref. 2) (Fig. 1). Estrogen appears to promote the ovarian fate by stimulating the expansion of the cortex while inhibiting the maintenance of sex cords.
The earliest form of sex determination, TSD, still remains a successful mechanism of ovarian development. However, genetic sex determination has evolved in higher vertebrates as an alternate form of sex determination. While it is not known exactly how or why the sex chromosomes evolved, speculation exists that it may have occurred as a means to overcome environmental changes. Animals utilizing TSD are susceptible to increases in temperature, as well as exposure to chemicals such as synthetic estrogens. These factors have the potential to skew sex ratios. It is possible that controlling sex at the chromosomal level evolved as a protective mechanism in order to shield embryos from the changing environment.
Birds and marsupials are unique because sex is determined by the composition of the sex chromosomes although the embryo remains sensitive to the effects of estrogen. We focus specifically on the domestic chicken, (Gallus domesticus), and tammar wallaby (Macropus eugenii), the most well characterized models for avian and marsupial species, respectively.
In birds, in contrast to mammals, males are homogametic (ZZ) and females are heterogametic (ZW) (reviewed in ref. 14) (Table 1). Whether the dosage of the Z chromosome, or the presence of the W chromosome is a testis or ovary-promoting factor in avian sex determination remains controversial. In mammals, chromosomal aneuploidy is a valuable model to determine the role of specific chromosomes. Regardless of the amount of X chromosomes present, an individual with a Y chromosome becomes a male. For example, individuals with an XXY composition are male, while XO animals are female. Unfortunately, data to analyze this type of sex chromosome aneuploidy in birds are lacking. However, triploid chickens (ZZW) have provided some insight into avian sex determination. These birds initially develop as intersexes with a right testis and temporary left ovotestis which regresses in adulthood.15 Unfortunately, this aneuploidy model does not provide solid evidence to support either theory. Several novel genes have been implicated in avian sex determination. Drosophila Doublesex and C. elegans Mab-3 Related Transcription factor 1 (DMRT1) has been linked to testis development due to its Z-linked characteristic.16 Two W-linked genes, Avian Sex-specific W-linked (ASW) and Female-Expressed Transcript 1 (FET1), were proposed to be the ovary-determining genes in birds.17,18 However, there is no evidence to support the functional roles of these genes. It is worth-noting that no Sry gene has been identified in the bird.14
While the primary signal that triggers avian sex determination remains unidentified, the effectors downstream of the genetic components are well documented, particularly for the organogenesis of the ovary. Similar to reptiles, estrogen is the driving force for ovarian differentiation.19 Steroidogenic enzymes including P450 Scc, 3β-HSD, P450c17 and 17β-HSD are already present in chick gonads of both sexes before sex determination commences.20–22 However, aromatase expression is only detectable in female gonads with expression starting at day 5–6 of incubation, when sex determination occurs.20–26 No aromatase has been detected in male gonads at similar times. The significance of estrogen in ovarian development was further proven by in ovo hormone manipulation. Aromatase inhibitors were able to masculinize ZW, or female embryos, before sex determination occurred.27,28 On the other hand, in ovo administration of estrogen before, or at the time of sex determination, temporarily feminized genetic males.29 (Table 1). This feminization is not permanent since these genetic males revert to phenotypic males within the first year of life (reviewed in ref. 5).
In the bird, morphological changes in gonads resemble those in reptiles. Primitive sex cords are present in the bipotential gonad. In the male gonads, sex cords enlarge due to Sertoli cell proliferation and become prominent in the medullary region. In the female embryo, asymmetrical gonadal development arises as the left gonad differentiates into a functional ovary and the right gonad regresses. In the left female gonad, the ovarian cortex expands as a result of proliferation of somatic cells and germ cells while primitive sex cords in the medulla regress to form the lacunae.14 Cortical expansion does not occur in the right gonad and a lack of estrogen receptor (ER) expression has been attributed to the degeneration of the right gonad.
Although birds diverged from reptiles with the evolution of a genetic mechanism for sex determination, the ovary-determining action of estrogen remains present in the bird despite the fact that these species diverged from reptiles approximately 245 million years ago.30 This phenomenon is also found in marsupials, a close relative to eutherian mammals.31 Most marsupials, such as the tammar wallaby (Macropus eugenii), are born underdeveloped with a mixture of fetal and neonatal characteristics. Gonadal development in the marsupial is similar to that of mice and humans except that the development of marsupial fetuses occurs outside of the uterus.32 Marsupial gonads are sexually indifferent at birth and gonadal differentiation commences immediately after birth.32,33 By postnatal day 7, ovarian differentiation has progressed to the point where the gonad is morphologically distinguishable.33 In the tammar wallaby, unlike birds and reptiles, the ovary does not produce estrogen at the time of gonadal differentiation. In fact, ovarian steroid production does not commence until almost 200 days after birth.33 Based on this fact, it appears that estrogen is not necessary for the differentiation of the ovary as it is in birds and reptiles. However, administration of estrogen to male embryos immediately after birth can induce complete ovarian development, similar to reptiles and birds.34
The ability of estrogen to feminize the male gonad in the marsupial suggests that estrogen can override the genetic components derived from the XY mechanism (Table 1). Marsupials have an XX/XY sex-determining mechanism identical to that of eutherian mammals.35 Although the Sry gene was identified in marsupials and is expressed at the correct time and in the proper male tissues, it is not known whether Sry triggers testicular development.36,37 If Sry is indeed the testis-determining gene in marsupials, it is not sufficient to overcome the feminizing effects of estrogen and thus sustain testis development. Intriguingly, ovary organogenesis in marsupials does not require estrogen since it is not produced at the time of gonadal differentiation. This deviation from reptiles and birds places marsupials in a unique evolutionary position suggesting that organogenesis of the marsupial ovary has become strictly dependent upon genetic regulation similar to eutherian mammals.
Mammals evolved from lower vertebrates approximately 80 million years ago.30 Accompanying that recent evolution was a new form of sex determination, genetic sex determination (GSD) that utilizes only chromosomal composition to determine sex. Unlike birds and marsupials that use GSD but remain sensitive to steroids, eutherian mammals evolved a mechanism whereby sex determination is completely resistant to steroids.
Mammalian males are heterogametic (XY) and females are homogametic (XX). The composition of the sex chromosomes is determined at the time of fertilization when sperm carrying either an X or Y chromosome fertilize the oocyte. The presence of the Sry gene, (Sex-determining Region of the Y chromosome), on the Y chromosome directs the gonadal primordium to form a testis. Ovaries arise in the absence of the Sry gene. Ovarian development is not dictated by the dosage of the X chromosome as evident by the fact that sex determination occurs normally in XXY and XO individuals. The lack of candidate genes for ovary determination has led to the speculation that ovarian development is a default process.
The ovary-determining gene, the counterpart of the Sry gene, has yet to be identified. An alternative “Z” theory has been proposed based on the phenomenon that testes develop in some XX individuals without Sry. It was proposed that a “Z” gene is present in the XX gonad to suppress testis development while promoting ovarian development.38 In the XY gonad, Sry inhibits the expression or activity of the Z gene, and therefore allows for the emergence of the testis pathway. Several genes that show female-specific expression patterns have been postulated as candidates for the ovary-determining gene or Z factor. Dax1 (dosage sensitive sex-reversal-adrenal hypoplasia congenital-critical region of the X chromosome gene1) was initially proposed to be the ovary-determining gene based on its ovary-specific expression profile and location on the X chromosome. However, genetic analysis in mice and humans revealed that Dax1 is not required for normal ovarian development and, ironically, is crucial for testicular development.39,40 Therefore, Dax1 was eliminated as a potential ovary-determining gene.39
Another gene with an ovary-specific pattern is Wnt4. Original findings on Wnt4 null embryos revealed the appearance of ectopic Leydig cells in the ovary leading to the proposal that Wnt4 could be the elusive Z factor.41 However, further analysis demonstrated that the ectopic steroid-producing cells in Wnt4 null ovaries were not Leydig cells but in fact adrenal cells misplaced in the gonadal primordium.42,43 Although it appears that Wnt4 is not responsible for inhibiting Leydig cell development in the ovary, it has been shown conclusively that Wnt4, along with its downstream target follistatin, antagonizes a specific aspect of testis development: the formation of testis-specific vasculature. Deletion of Wnt4 or follistatin caused the appearance of a testis-specific vessel, the coelomic vessel, in the XX gonad.44,45 These findings demonstrate that Wnt4 and follistatin form a novel signaling cascade that inhibit a portion of the testis pathway to allow proper ovarian development to occur. Interestingly, ectopic expression of Wnt4 in the XY mouse did not prevent the establishment of testis vasculature, indicating that multiple or redundant pathways must be present in the XY gonad to ensure formation of testis-specific vasculature.
Foxl2, a forkhead transcription factor that is linked to the Sry-independent XX to XY sex reversal in goats with Polled/Intersex syndrome, has emerged as a potential Z gene candidate (reviewed in refs. 46–51). The conserved ovary-specific expression of Foxl2 in chickens, turtles, fish and mammals suggests that Foxl2 could play a critical role in ovary development.52,53 Furthermore, Foxl2 was identified to be the genetic component for human Blepharophimosis Ptosis Epicanthus inversus Syndrome (BPES) as well as premature ovarian failure.52,54–58 However, null mutations of Foxl2 in mice do not affect initial ovarian formation, and in humans, no cases of XX to XY sex reversal have been reported in individuals with BPES.59, 60 These findings indicate that at least in mice and humans, Foxl2 is not a candidate for the ovary-determining gene or Z factor at the time of sex determination. Interestingly, ovaries of the Foxl2 null neonates exhibited sex-reversal phenotypes with the expression of testis-specific markers (i.e., Sox9 and anti-Müllerian hormone) and the appearance of testis cord-like structures.61 The exact mechanism for these phenotypes is unclear but it appears that Foxl2 plays a critical role in maintaining ovarian identity after birth.
The ovary to testis sex reversal phenotype seen in the Foxl2 null newborn ovary was also reported in Wnt4 null ovaries at birth.41 Furthermore, germ cells were lost at birth in these null ovaries. Female germ cells are known to be essential for organizing and maintaining the ovarian structure. When germ cells are absent from the XX gonad, ovarian follicles, which are the functional units in the ovary, never form.62 Additionally, if female germ cells are lost after formation of follicles, follicles rapidly degenerate.63–65 In contrast, testis development progresses normally in the absence of germ cells.66 This evidence indicates that proper maintenance of ovarian development requires a timely interaction between female germ cells and genetic components.
While the genetic component for ovary formation remains a mystery, much can be learned from the developmental process of the gonads to determine the stepwise organization of the ovary. The testis and ovary have completely different structures in adulthood, but surprisingly initiate as analogous tissues. In all vertebrates, gonads arise as bipotential organs that are initially indistinguishable in male and female embryos. In most mammals, somatic and germ cells are intermixed throughout the bipotential gonad prior to sex determination. In contrast, gonads of most reptiles and birds have a different arrangement of somatic and germ cells at early developmental stages. In reptiles, primitive sex cords composed of somatic cells are present throughout the gonad early in development while germ cells are initially restricted between somatic cells of the coelomic epithelium. 4,67 Despite the fact that structural differences in the undifferentiated gonad exist in reptiles and mammals, later establishment of ovarian compartments is remarkably similar between species. In the mammal and reptile, female germ cells populate the ovarian cortex, where folliculogenesis occurs, while the medullary region regresses (Fig. 1).
In reptiles, birds, and marsupials, estrogen is the common ovary-determining factor regardless of the initial signal (temperature or chromosomal composition). Eutherian mammals have evolved a mechanism where development is controlled completely by genetics. In fact, administration of estrogen to XY gonads did not induce formation of an ovary.68 Furthermore, blockage of estrogen production, or loss of estrogen receptors (ER) in XX embryos, did not lead to testis formation.65,69–71 Insensitivity to the feminizing effect of estrogen in eutherian embryos is a logical adaptation due to the fact that eutherian fetuses develop inside the mother and are constantly exposed to maternal-derived estrogens through the placenta. Eutherian mammals must have evolved mechanisms to protect the embryo from maternal estrogens. Unlike birds, reptiles, and marsupials, where ER is present in the gonadal primordium during a critical window of sex determination, gonads of eutherian embryos do not express ER until after sex determination. This observation indicates that while eutherian gonads are exposed to estrogen, they lack the ability to respond to it. Alpha-fetoprotein (AFP), which has a high affinity for sex steroids, is thought to be responsible for limiting embryonic exposure to estrogen.72 However, removal of AFP did not cause feminization of the XY gonad and sex determination appeared normal.73 This result further supports the notion that eutherian mammals probably developed multiple/redundant mechanisms to prevent the feminizing effects of estrogen.
While it appears that estrogen has no impact on primary sex determination in the developing eutherian embryo, a different situation arises after birth. In the adult mammalian female, estrogen plays a vital role in maintaining the ovary. In the aromatase knockout mouse, testicular cell types and structures arose in postpubertal ovaries.74 A similar phenomenon was observed in mice lacking both estrogen receptors alpha and beta (ERαβ KO), where Sertoli cells and seminiferous-like structures appeared in the ovaries after puberty65,69 (Table 1). These observations bring about an intriguing hypothesis that the primitive estrogen-induced mechanism of ovary development remains present in eutherian females. While the effects of estrogen are silenced in the embryo, the feminizing effect of estrogen re-emerges in the adult female. Without estrogen production and its signaling pathway, eutherian ovaries lose their sexual identity and are partially sex-reversed, reminiscent of the female to male sex reversal scenario present in avian, reptilian, and marsupial embryos when estrogen synthesis is blocked.
When comparing the different mechanisms of sex determination that have evolved between vertebrate species, it becomes clear that while the beginning (formation of the gonadal primordium) and end (production of an ovary or testis) are very similar, the triggers for ovarian differentiation are diverse. TSD is a popular mechanism of sex determination in reptiles, and species that utilize TSD also depend on estrogen for development of the ovary. Initially this mechanism of sex determination evolved in a colony where no sex chromosomes existed. As sex chromosomes (or GSD) emerged, effects of environmental factors, as well as estrogen, gradually waned. Little is known as to why GSD has evolved, but speculation exists that GSD is a protective mechanism for fetuses from environmental changes which can shift sex ratios in animals utilizing TSD. GSD may therefore have evolved as a way to prevent harmful exposure to external factors (temperature fluctuations) or internal factors (maternal hormones). Eutherian mammals have escaped the estrogen dependency that lower vertebrates depend on. However, the necessity for estrogen in eutherian mammals is still present but does not become a necessary factor until adulthood. This returning dependency is obvious in the ERαβ KO and aromatase KO mice where ovaries undergo transdifferentiation to produce testicular tissues and structures in an XX adult. This evolutionary evidence prompts us to review the dogma of sex determination that ovarian development is default. Further research is required to challenge this dogma and understand how genetic mechanisms have replaced estrogen dependency in mammalian ovary organogenesis.
We appreciate the funding supports from the Endocrine, Developmental, and Reproductive Toxicology Training Grant (A.C.D.), March of Dimes Birth Defects Foundation (H.H-C.Y), and National Institute of Health (HD46861 for H.H-C.Y).
Previously published online as an Organogenesis E-publication: http://www.landesbioscience.com/journals/organogenesis/abstract.php?id=2491