| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Section of Integrative Biology, University of Texas at Austin, Austin, Texas 78712
Address all correspondence and requests for reprints to: David Crews, West 24th and Speedway, Patterson Labs, Austin, Texas 78705. E-mail: crews{at}mail.utexas.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Genes involved in GSD vertebrate gonadogenesis include FoxL2, Wnt4, Dax1, Dmrt1, and Mis. While the expression of some of these factors has been examined in organisms with TSD (1, 2, 3, 4, 5, 6, 7, 8, 9), their ability to respond to changes in temperature has not been investigated. Although experiments at constant incubation temperatures can correlate temporal and spatial patterns of gene expression to sexual fate, sex-reversing temperature-shift experiments represent an important functional manipulation in an organism lacking typical genetic techniques. Rapid change in gonadal expression of these genes after a temperature shift confirms a role in the formation of the gonad, whereas the timing of this response provides an indication of its hierarchical placement within the molecular cascade governing organ development. In this study, we examine the ability of temperature to regulate the expression of FoxL2, Wnt4, Dax1, Dmrt1, and Mis. The regulation of Sox9 by temperature was demonstrated previously, and here we extend those results (10).
We examined the molecular pathway underlying TSD in the red-eared slider turtle, Trachemys scripta, a species that exhibits TSD. The slider turtle is sensitive to the effect of temperature from approximately Greenbaum’s stage 14 through stage 19 at a female-producing temperature (FPT) and through stage 20 at a male-producing temperature (MPT) (11, 12). Cooler incubation temperatures (25–27 C) produce all male hatchlings and warmer temperatures (31–35 C) result in all female hatchlings, with varying sex ratios produced by temperatures in between (11, 13). Shifting eggs during the temperature-sensitive period (TSP) from one end of the temperature spectrum to the other (i.e. from 26 C to 31 C or vice versa) redirects gonadal development, resulting in 100% sex reversal (14, 15, 16).
The process of forming an ovary is less well understood than testicular differentiation (for review, see Ref. 17). Ovarian development is characterized by a proliferation of cells in the gonad’s cortical region concurrent with medullary regression. Within the thickened cortex, granulosa cells organize to surround germ cells, whereas steroidogenic theca cells remain interstitial between developing follicles. One of the few ovarian-specific factors is FoxL2 (Forkhead box protein L2), a single-exon transcription factor with a conserved winged-helix forkhead domain, mutations in which cause ovarian failure in humans with blepharophimosis/ptosis/epicanthus inversus syndrome (BPES) type I disease (18, 19). FoxL2 is expressed in the developing ovary of mouse and chick embryos (7, 20) and is involved in postnatal mammalian granulosa cell differentiation (21). Furthermore, FoxL2–/– XX adult mice exhibit marked up-regulation of several testis-specific markers, including Sox9, Dhh, and Fgf9, indicating that FoxL2 repression of testis-determining genes may occur well past mammalian embryonic sex determination (22). In T. scripta, FoxL2 expression is detected in both MPT and FPT gonads and is later restricted to the developing ovary (7).
Wnt4 (wingless-type MMTV integration site family member 4) signaling has been implicated in the development of the vertebrate reproductive system in both sexes (see Ref. 23), mediating the initial formation of the Müllerian ducts and regulating the migration of steroidogenic mesonephric cells into the developing gonad (24, 25). In the ovary, Wnt4 acts via Follistatin to prevent the formation of a testis-specific coelomic blood vessel (26). In Wnt4–/– XY mice, gonadal expression of Sox9, Mis, Dhh, and Sf1 is decreased but can be rescued by ectopic Wnt4, suggesting an active role for Wnt4 in testicular development as well (27). To our knowledge, the gene expression patterns or functional role of Wnt4 has not been examined in an organism with TSD.
Dax1 (Dosage-sensitive sex-reversal, Adrenal hypoplasia congenital on the X chromosome 1, also known as Nr0b1 and Ahch) is a novel orphan member of the nuclear hormone receptor superfamily and plays a role in both mammalian sex determination and adrenal function (see28). Its early ovary- and late testis-specific expression pattern in the mouse suggests a complex role in the development of the gonad in both sexes (29). Similarly, it is expressed in both sexes in chick and two species with TSD, the American alligator and the Olive Ridley sea turtle (2, 5, 6). Originally thought to have an ovarian-determining function (30, 31), undisrupted embryonic gonadogenesis but abnormal postnatal formation of follicles in Dax1–/– XX mice led to the suggestion of an adult Dax1 role in the ovary (32). In testicular development, Dax1 may control availability of estrogens by regulating aromatase transcription (33, 34, 35). Furthermore, studies in mouse suggest that Dax1 may be up-regulated by Wnt4 signaling (36, 37).
The development of the vertebrate testis has been well studied and requires the action of Dmrt1, Mis, and Sox9, among other factors (for review, see Ref. 38). Elimination of Dmrt1 (Doublesex mab3-related transcription factor 1) is thought to be responsible for male-to-female sex reversal seen in XY humans with chromosome 9 deletions (39). Although it seems to play a downstream role in testis differentiation in mammals (40), it has been proposed to be a master sex-determining gene in both chicken and medaka (41, 42). Extensive studies in mouse of the Sry-related gene Sox9 (SRY-like HMG-box 9) have shown it to be both necessary and sufficient to cause the determination and differentiation of a testis (43). In both humans and mice, SOX9 interacts directly with SF1 (Steroidogenic factor 1) to up-regulate the expression of Müllerian-inhibiting substance (Mis or anti-Müllerian hormone, Amh) (44, 45). Mis, a member of the TGF-ß superfamily, is one of the early factors secreted by differentiated Sertoli cells in the testis and causes the regression of the Müllerian ducts, anlagen that otherwise develop into the uterus, cervix, and fallopian tubes in females (46).
To identify possible roles in an organism with TSD, we report cloning Wnt4 and Dax1 in the red-eared slider turtle. We analyze the expression patterns of these two genes as well as FoxL2 throughout six stages of gonadogenesis encompassing the temperature-sensitive sex-determining period. To examine the early factors directing a bipotential gonad down an ovary or a testis developmental trajectory, we extend our previous findings on the expression patterns of three testis-specific factors, Dmrt1, Mis, and Sox9 to earlier time points within the TSP. To investigate placement of these factors in a temporal hierarchy regulating the development of the turtle gonad, we analyze their ability to rapidly respond to a shift in temperature during the sex-determining period. For the first time, we report a response of FoxL2, Wnt4, Dmrt1, and Mis to a change in temperature, confirming their role in the molecular pathway underlying TSD. As well, we provide evidence that Dax1 is involved in gonadogenesis in both sexes, and we extend previous findings on the expression of Sox9 in a species with TSD.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Examining a New Stage of Development
The TSP in T. scripta lasts from approximately stage 14 through stage 20, and shifting embryos between constant temperatures (26 and 31 C) during this time causes complete sex reversal (11). We used this lability to assess the involvement of various candidate genes in TSD gonadogenesis. After development progressed along one trajectory (MPT or FPT) for several stages during the sex-determining period, eggs were shifted at stage 16 to the opposite temperature. We assessed the ability of six candidate sex-determining genes to rapidly respond to this change in temperature at two subsequent time points. Turtle embryonic development is prolonged, and the period from stage 16 to stage 17 can last up to 2 d at FPT (31 C) or 4 d at MPT (26 C). Therefore, to examine the effects of the temperature shift on a finer temporal scale, we define a stage halfway between stage 16 and stage 17, termed stage 16.5. We distinguish this stage as having morphological characteristics midway between those associated with stages 16 and 17, as described in Fig. 1
.
|
, A and B). At the end of the TSP, the spatial expression pattern of FoxL2 in the developing ovary is localized primarily to the cortical region, although faint expression is seen in primitive sex cords that later degenerate (stage 19, Fig. 2C). Cortical expression in the ovary continues through differentiation (stage 23, Fig. 2E). Expression is also seen faintly at MPT at the close of the TSP and through differentiation in putative testicular seminiferous tubules (stages 19 and 23, Fig. 2, D and F). At all stages examined, FoxL2 expression was detected in both sexes in the dorsal metanephric tubules of the future kidney, although more strongly at FPT (Fig. 2, G and H). Furthermore, expression of FoxL2 occurs in epithelial cells lining the interior of the developing Müllerian duct of embryos at stages 19 and 23 (Fig. 2I, data not shown).
|
). However, at the end of the TSP and through ovarian differentiation, FPT expression levels rise significantly above levels at MPT (stages 19, 21, and 23, P < 0.0015, Fig. 3A). Temperature-shifting embryos at stage 16 causes gonadal expression of FoxL2 to rise, regardless of the direction of the shift (MPT
FPT or FPTMPT, Fig. 3A'). However, this rise was significant in only one comparison: FoxL2 expression in stage 16.5 MPTFPT-shifted gonads is significantly higher than expression in unshifted stage 16.5 MPT gonads (P < 0.009, Fig. 3A').
|
In situ hybridization detected Wnt4 mRNA throughout the bipotential gonad during the TSP (stages 17 and 19, Fig. 2
, K–N). This expression becomes primarily localized to the cortical region of the ovary and the seminiferous tubules of the testis during gonad differentiation and is also retained between sex cords in the testis (stage 23, Fig. 2, O and P). At all stages examined, strong Wnt4 expression in both sexes is observed in the mesonephros and dorsal metanephric tissues destined to become the adult kidney (Fig. 2, Q and R). Expression in the developing Müllerian ducts is found at stages 19, 21, and 23 in both sexes; in FPT embryos, Wnt4 expression increases as the duct grows, whereas at MPT, expression decreases with duct regression (Fig. 2, S and T).
Throughout the TSP, Wnt4 is expressed at comparable levels as measured by qPCR between gonads developing at FPT and MPT (stages 16–19, Fig. 3B). However, during ovarian differentiation, expression at FPT significantly rises above MPT by up to 4-fold (stage 21, P = 0.015; stage 23, P < 0.0001; Fig. 3B). In response to a FPTMPT temperature shift, expression levels one stage later (stage 17) are increased from unshifted stage 17 FPT levels, although not significantly (P = 0.0508, Fig. 3B'). In gonads from the converse shift (MPTFPT), Wnt4 expression is not significantly different from unshifted levels at either time point.
Dax1 Is Expressed throughout the Gonad and Müllerian and Wolffian Ducts
To examine its expression in the slider turtle, we cloned a 297-bp fragment of Dax1. BLAST analysis identified 96% nucleotide sequence homology to Dax1 in Lepidochelys olivacae, 92% to Alligator mississippiensis, 87% to G. gallus, and 78% homology to H. sapiens. The translated 98-amino-acid sequence (MacVector) corresponds to the C-terminal ligand-binding domain of the consensus vertebrate protein (see supplemental Figs. 3 and 4, published as supplemental data on The Endocrine Society’s Journals Online web site at http://mend.endojournals.org) and is 95% homologous to L. olivacea, 93% to A. mississippiensis, 92% to G. gallus, and 76% to H. sapiens Dax1.
In situ hybridization during the TSP reveals a diffuse pattern of Dax1 expression in gonads developing at FPT and MPT (stage 17, Fig. 2, U and V). At the end of the TSP, spatial expression patterns begin to organize (stage 19, Fig. 2, W and X), and during gonad differentiation, expression is concentrated in the cortex of ovaries and in a striped pattern indicative of sex cords in testes (stage 23, Fig. 2, Y and Z). At all stages examined, Dax1 is expressed in dorsal metanephric tissue in both sexes, typified by the staining seen at stage 23 (Fig. 2, AA and BB). In both developing Müllerian and underlying Wolffian ducts, Dax1 expression occurs in the epithelial cells lining the interior of the ducts (Fig. 2, CC and DD). Expression in Müllerian ducts is apparent at stages 19 (data not shown) and 23, whereas Wolffian duct expression is detected at stage 23. Some sections showed punctate Dax1 expression in a mesonephric region from which adrenal tissue is expected to develop (data not shown).
Throughout gonadogenesis, Dax1 expression levels measured by qPCR were not significantly different between gonads developing at MPT and FPT (Fig. 3C). However, within FPT, a significant change in expression between stages is observed. In the developing ovary, Dax1 levels are significantly greater at stages 16, 16.5, and 17 than at stages 21 and 23 (P < 0.009 for all comparisons). In response to a shift in temperature, expression levels by stage 17 have changed in sex-typical directions, although not significantly (Fig. 3C'). Thus, after a MPTFPT shift, Dax1 expression levels are FPT-typical by stage 17, whereas after a FPTMPT shift, they are MPT-typical.
Testis-Specific Genes Show a Rapid Response to Temperature Shifts
Previously, we examined the spatial and temporal expression patterns of Dmrt1, Mis, and Sox9 throughout gonadogenesis in the slider turtle (9). Here, we report an extension of those results in three independent qPCR samples, reveal dimorphic expression levels earlier in the TSP than we previously described, and demonstrate that Dmrt1 and Mis expression is regulated by temperature.
Beginning early in the TSP, expression of both Dmrt1 and Mis is significantly higher at MPT as compared with FPT (Dmrt1: stage 16.5, P < 0.027, Fig. 4A; Mis: stage 16, P < 0.007, Fig. 4B). This pattern becomes substantially more dimorphic through testicular differentiation (stages 17, 19, 21, and 23, P < 0.0001 for both genes). In response to a FPTMPT temperature shift, Dmrt1 expression between stages 16.5 and 17 rises significantly above unshifted gonads remaining at FPT (P = 0.0003) and is not different from MPT expression levels (Fig. 4A'). Conversely, one stage after a MPTFPT shift, expression levels have begun to drop but are not yet significantly lower than unshifted MPT levels. Thus, one stage after a temperature shift to MPT, expression of Dmrt1 has been significantly up-regulated, but one stage after a shift to FPT, expression has not yet been significantly down-regulated (Fig. 4A').
|
FPT shift is significantly decreased from unshifted MPT levels (P < 0.01) and is not significantly different from FPT levels (stage 17, Fig. 4B'). After a FPTMPT shift, expression has not significantly changed by stage 17 from unshifted FPT levels. Thus, expression of Mis has been significantly down-regulated one stage after a shift to FPT but has not yet been significantly up-regulated after a shift to MPT (Fig. 4B').
Although Dmrt1 and Mis show significantly dimorphic expression levels early in the TSP, Sox9 expression levels are similar between gonads at MPT and FPT (stages 16, 16.5, and 17, Fig. 4C). Sox9 expression in the putative turtle testis becomes significantly higher than in the developing ovary at the end of the sex-determining period (stage 19, P < 0.0001) and continues to be significantly higher through testicular differentiation (stages 21 and 23, P < 0.0001). Intriguingly, after a MPTFPT shift, gonadal expression of Sox9 increases significantly above unshifted MPT levels by one half-stage later (stage 16.5, P < 0.006) but are again indistinguishable from MPT by stage 17 (Fig. 4C'). After a FPTMPT shift, expression is not different from unshifted FPT levels at either subsequent time point.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
It should be noted that patterns of gene expression as analyzed by qPCR in this study reflect transcript abundance across the gonad as a whole and cannot take into account cell type differences in contribution to the total transcript pool. Germ cells and the various somatic cells that make up the gonad may be, and in fact are likely to be, expressing these gene products at different levels. Thus, if a temperature shift causes proliferation or loss of a particular cell type that is contributing to or causing an expression pattern, this subtlety may be lost in our analysis. Separating cell types of the gonad for quantitative analysis of gene expression would resolve this in future studies.
FoxL2 Expression Is Consistent with a Role in Ovarian Development
FoxL2 is one of the few ovarian-specific genes identified thus far, and a proposed function in mice is to repress male-specific gene expression postnatally (22). Although its expression in mouse during embryonic gonadogenesis is solely in the developing ovary, a sex-determining role during this period has not been identified. Here we confirm and extend a previous report of its expression pattern in T. scripta (7). In situ hybridization reveals localization of expression during differentiation in the cortex of the developing ovary, a pattern that is similar to other vertebrates. This pattern may reflect expression in the granulosa cell lineage, which is known to express FoxL2 in mammals and localizes to the cortex, surrounding germ cells (22). Through quantitative real-time PCR, we are able to temporally define when expression at FPT significantly rises above MPT, which occurs at the close of the TSP. This pattern supports a role of FoxL2 in ovarian development and suggests it may be involved in differentiation. Because expression is similar between the sexes early in the TSP, future temperature-shift experiments should extend past this period. However, regulation of FoxL2 by temperature is observed, and taken together, our data provide evidence that FoxL2 plays a role in the development of the ovary in an organism with TSD.
Although downstream target genes of FoxL2 remain unknown, there is evidence for regulation of aromatase expression by FoxL2 in chicken, trout, tilapia, and polled goats (47, 48, 49, 50). Aromatase encodes the enzyme that converts androgens to estrogens. Hudson et al. (48) showed that inhibition of aromatase in the developing chick gonad leads to reduced, but not abolished, FoxL2 expression. This is suggested to occur by a positive feedback mechanism in which FoxL2 up-regulates aromatase, which then enhances FoxL2 activation. Furthermore, it was recently shown by Ramsey et al. (51) that T. scripta gonadal aromatase expression is dimorphic between FPT and MPT embryos early in the TSP; expression at FPT is organized in circular clusters of expressing cells surrounding nonexpressing cells, putatively pregranulosa cells surrounding germ cells, whereas at MPT, expression remains unorganized. If a positive feedback loop exists between FoxL2 and aromatase in an organism with TSD, exogenous aromatase inhibitor in turtle embryos should also cause decreased FoxL2 expression.
Expression of Wnt4 Is Conserved across GSD and TSD Vertebrates and Is Consistent with a Role in Both Ovary and Testis Development and Duct and Kidney Formation
To our knowledge, examination of Wnt4 in the slider turtle represents the first analysis of this signaling molecule in an organism with TSD. Here we find Wnt4 is expressed in turtle gonads developing at both MPT and FPT throughout the periods of sex determination and gonad differentiation, suggesting that it functions in both the developing testis and ovary. Expression at MPT begins to localize in a striped pattern indicative of developing sex cords as early as the end of the TSP, although expression in interstitial cells between cords is retained. Expression in the ovary becomes concentrated in the cortical region, and during differentiation, quantitative real-time PCR shows expression levels at FPT rise significantly above MPT. It is possible that this up-regulation reflects a shift in Wnt4function during gonad differentiation to a later female-specific role.
Mouse studies support a complex role for Wnt4 in both male and female development, and the regulation of Wnt4 signaling in a bipotential gonad presents an interesting problem. It was recently suggested that antagonistic action between Wnt4 and Fgf9 signaling pathways regulate the switch between ovarian and testicular development in the mammalian bipotential gonad (52). However, Wnt4 may have several sex-specific roles throughout gonadogenesis. For example, Wnt4–/– XY mice retain normal Sry expression but have decreased levels of Sox9 and Dhh expression, disrupted Sertoli cell differentiation, and a substantial overpopulation of steroidogenic cells migrating into the gonad from the mesonephros (27). Thus, an important role for Wnt4 in the developing mammalian testis includes regulating the migration of these cells, whereas wholly preventing this migration into developing ovaries. The expression of Wnt4 in the turtle in the mesonephros of both sexes as well as in developing sex cords and interstitial cells of the testis may reflect a similar role. Mesonephric expression of Wnt4 is also important in the development of other vertebrate organs and is required for the cellular mesenchymal-to-epithelial transition that occurs during mammalian kidney tubule formation (53). Strong turtle Wnt4 expression in developing mesonephric and metanephric tissues destined to become the kidney is consistent with a similar function in an organism with TSD.
A testis-specific medullary coelomic blood vessel forms during mammalian gonadogenesis. It was shown that Wnt4–/– XX mice ectopically develop this blood vessel, whereas overexpression of Wnt4 in XY mice results in a disorganized coelomic blood vessel (25, 54, 55). If a similar coelomic blood vessel forms in MPT organisms with TSD, a possible role for Wnt4 expression seen in the medullary region of the developing turtle ovary may be repressing its formation at FPT.
Finally, the role of Wnt4 in the formation of the Müllerian ducts, anlagen that initially form in both sexes and then develop into the cervix, uterus, and oviducts in females, has been examined extensively in the mouse. It is suggested that Wnt9b up-regulates Wnt4, which leads to initial Müllerian duct formation in both sexes as well as inhibition of Wnt7a and Pax8 (24, 53, 56). In males, Mis then causes the degradation of these ducts. Lack of either Wnt9b or Wnt4 prevents the initial formation of Müllerian ducts, whereas a lack of Wnt7a results in persistent ducts and prevents downstream organs from forming (53, 56). In the turtle, the Müllerian duct develops in both sexes in a rostral to caudal wave and appears along the anterior/posterior axis near the gonad at the end of the TSP (stage 18/19) (57). Müllerian duct primordia develop in both sexes for two stages and then enlarge at FPT and degrade at MPT (57). Our data show that before Müllerian ducts form, Wnt4 is expressed in the mesonephric tissue from which they will derive. Furthermore, Wnt4 expression is detected within primordial Müllerian ducts from the first stage they appear. This early expression pattern is consistent with a role for Wnt4 signaling in the embryonic development of these ducts.
Dax1 Expression Is Consistent with a Role in Gonad and Adrenal Development
The complexity of the role of Dax1 in vertebrate sexual development has thus far prevented a clear understanding of its function. Expression is detected early in the development of the vertebrate gonad, and in mammals, this expression becomes female specific through time, whereas in chick, expression continues in both the testis and ovary (2, 31). To examine its role in an organism with TSD, we cloned a partial Dax1 sequence in the slider turtle and find a similar expression pattern to chick. Dax1 is gonadally expressed throughout sex determination and differentiation at levels that are not significantly different between MPT and FPT gonads, although expression at FPT does decrease significantly through time. Furthermore, Dax1 mRNA first appears dispersed throughout the bipotential gonad and then becomes progressively restricted to the ovarian cortex and testicular sex cords. Dax1 is involved in the formation of the human adrenal cortex (28, 58). Here we find Dax1 expression in the turtle in a region of the mesonephros from which adrenal gland develops, indicating conservation of expression pattern. Interestingly, Dax1 is the only gene examined in this study by in situ hybridization expressed in the developing Wolffian ducts, structures that eventually degrade in females and become functional tubules in males.
To explain its complex role in both developing mammalian testes and ovaries, a dosage-dependent mechanism has been proposed in which one copy of X-linked Dax1 plays a role in activating male development, whereas two copies can inhibit Sry action (34, 35, 59). Because organisms with TSD are thought not to have functional sex chromosomes, dosage-dependent regulation in these systems could be conserved if regulated by a different mechanism. There are presumably two Dax1 gene copies in every diploid turtle embryo, and thus functional Dax1 differences must be regulated at transcriptional or translational levels. Furthermore, if the role of DAX1 is dose dependent, this regulation must at some point be correlated with incubation temperature. In this study, we find that expression levels between MPT and FPT are not significantly different, and consequently, any dimorphic DAX1 function must be posttranscriptionally regulated in the turtle. During the preparation of this manuscript, a model of dosage-dependent sex determination in organisms exhibiting TSD was proposed in which effects of a Z-linked male determinant depend on both temperature and copy number, and our data are not inconsistent with this mechanism (60).
A regulatory relationship between Wnt4 and Dax1 has been previously postulated, in which Wnt4 signaling leads to activation of Dax1 in a sex-specific manner (37). Wnt4–/– XX mice show decreased Dax1 expression at 11.5 d postcoitum, whereas no change in expression level is observed in 11.5-d postcoitum Wnt4–/– XY mice (37). The general colocalization of expression of these factors in the developing turtle gonad is consistent with regulation of Dax1 by Wnt4 in an organism with TSD, and additional experiments are required to explore this possibility.
Response to Temperature Suggests an Early Role for Dmrt1 and a Later Role for Mis in Testicular Development
As expected from our previous work (9), expression levels of Dmrt1, Mis, and Sox9 are strongly sexually dimorphic during gonadogenesis. By quantitative real-time PCR, both Dmrt1 and Mis are expressed at significantly higher levels at MPT as compared with FPT beginning early in the TSP. For both genes, these levels become increasingly disparate through differentiation. In contrast, Sox9 is expressed in both sexes during the TSP and become significantly greater at MPT at the end of the TSP and during differentiation.
For the first time, we report a transcriptional response of Dmrt1 and Mis to a shift in embryonic temperature. Dmrt1 expression in the gonad shows a rapid response when embryos are shifted from FPT to MPT; between one half and one stage later, expression is up-regulated to a level significantly greater than unshifted FPT levels and is not different from MPT levels. The converse shift from MPT to FPT does not show a significant change one stage later, although Dmrt1 expression has begun to drop toward FPT levels. Taken together, these data indicate that testicular development requires the rapid up-regulation of Dmrt1 and places the action of Dmrt1 downstream of temperature. Furthermore, it suggests an early role for Dmrt1 function in the developing testis and an upstream position in the temporal hierarchy of genes underlying development of the gonad. Delayed response in gonads shifted toward FPT suggest that repression of this gene may be a later event in the molecular cascade governing ovarian development.
In contrast to the case with Dmrt1, Mis shows a different type of rapid response to temperature. After a shift from FPT to MPT, gonadal Mis expression one stage later has begun to increase but is not yet statistically higher than unshifted FPT levels. This suggests that the role of Mis is downstream of Dmrt1 in the molecular hierarchy governing testicular development, given the assumption that Mis has not yet been up-regulated by upstream factors. Conversely, response of Mis expression to a shift from MPT to FPT shows a dramatic change one stage later. Expression is significantly down-regulated from unshifted MPT levels and is not distinguishable from FPT levels. One testis-specific function of Mis in mammals is to induce regression of the Müllerian ducts, which otherwise differentiate into female-specific organs. The repression of this hormone is thus critical in mammalian ovaries, and our data suggest the same may be true in organisms with TSD. The extremely low levels of Mis observed in FPT gonads at all stages, as well as the rapid down-regulation after a shift to an FPT, support this hypothesis.
Although the temporal response of gene expression to a temperature shift can be used to infer placement of gene action in the molecular hierarchy governing gonadogenesis, we cannot rule out the possibility that this response does not occur under constant conditions. Thus, it should be considered that a novel environmental temperature may not engender the same molecular response as in an embryo experiencing the same temperature in nonshift conditions.
Here we demonstrate that Sox9 expression levels in the turtle gonad are statistically similar in both sexes early in the TSP. By qPCR, we temporally define when Sox9 expression becomes significantly male specific, which occurs at the end of the TSP. However, as is true for any monomorphic gene expression pattern, SOX9 protein may be differentially functional at an earlier time point than when dimorphic expression occurs. It was shown that in response to a MPTFPT temperature shift, SOX9 protein levels decreased in cultured gonad explants within the TSP (10). Furthermore, SOX9 remains cytoplasmic in cells of the mammalian bipotential XX gonad but is translocated to the nucleus in XY gonads where it can regulate downstream gene transcription (61, 62). Thus, although our quantitative expression data supports a later role for Sox9 in the development of the testis, differential Sox9 function may occur earlier via translational regulation of SOX9 or subcellular localization of the Sox9 mRNA or protein.
Examining the patterns of up-regulation of these testicular genes may also reveal something about their role in development. Dmrt1 and Sox9 expression levels rise sharply early in development, showing the greatest activation in expression in the middle of the TSP, between stages 16 and 19. In contrast, Mis expression increases most dramatically at the end of the TSP and during differentiation, from stage 19 to stage 23, implying a later role and suggesting up-regulation by earlier testis-specific genes. A candidate for activating Mis is Sox9, because it is known to directly target Mis in mammals (44, 45). As discussed, differential SOX9 protein translation or localization provides a mechanism by which equal levels of Sox9 expression could differentially regulate Mis activation in the turtle and, although not tested in the current study, warrants future attention.
In this study, we examine the spatial and temporal response of six candidate sex-determining genes to a sex-reversing change in temperature. This functional manipulation in a nonmodel system allows an analysis of their hierarchical placement within the molecular network governing gonadogenesis. By studying the molecular cascade regulating the development of a testis and an ovary from a bipotential gonad in an organism with TSD, we begin to identify components of the pathway that may be functionally conserved through diverse modes of sex determination.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Staging
Progression of development was monitored by staging external morphological characteristics of a sampling of individuals according to Greenbaum’s staging series (12). Morphological characteristics of a stage 16 slider turtle include the initial formation of a beak, a white line that extends ventrally from the pupil of the eye, and a large digital plate with a smooth periphery with slight indications of digital ridges (Fig. 1A). By stage 17, a white caruncle, or egg tooth, has formed on the upper jaw, the white line of the eye is absent or interrupted as pigmented cells reorganize, and the periphery of the digital plate is serrated with well-developed digital ridges (Fig. 1C). We defined stage 16.5 for the purposes of this study as midway between stages 16 and 17, characterized by a caruncle that is just visible and slightly below the epidermis of the beak, a narrowing white eye line, and slight serrations around the periphery of the digital plate (Fig. 1B). For morphological characteristics representative of other stages, see Greenbaum (12).
Cloning of Turtle Gene Homologs
Total RNA was extracted from pooled AKG complexes from a variety of sexes and stages and reverse-transcribed using oligo(dT) primers with Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA). Primers for Dax1 and FoxL2 were designed based on published mouse, human, alligator, and chicken sequences where available and were Dax1, 5'-AGT GCT GGA GCC TGG ACA TCG-3' and 5'-CAT CTC CAG CAG CAT GTC ATC C-3', and FoxL2, 5'-TCC GGC ATC TAC CAG TA-3' and 5'-TTG CCG GGC TGG AAG TG-3'. Wnt4 primers amplifying a 390-bp fragment based on mammalian sequences were a gift from Pascal Bernard and were 5'-GTA CCT GGC CAA GCT GTC-3' and 5'-AGC ATC CTG ACC ACT GGA AGC-3'. Amplified fragments were ligated into pCR4-TOPO vector according to the manufacturer’s protocol (Invitrogen) and sequenced using M13F and M13R primers. The sequence of the 261-bp FoxL2 clone was subsequently found to correspond to bp 73–330 of a previously published T. scripta FoxL2 sequence (7) (accession no. AY155535). Sequences for Dax1 and Wnt4 were submitted to GenBank (accession nos. EF591056 and EF591055, respectively).
In Situ Hybridization
AKG complexes were dissected from embryos, immediately frozen in OCT embedding medium, and stored at –80 C. They were subsequently serially sectioned in four series on a 2800 Reichert-Jung cryostat at 20 µm and thaw-mounted onto SuperFrost Plus slides (Erie Sciientific Co., Portsmouth, NH). Sections were fixed in cold 4% paraformaldehyde, washed in PBS, and incubated in 0.25% acetic anhydride/triethanolamine. After washes in 2x standard saline citrate (SSC), slides were dehydrated in an increasing ethanol series, air dried, and stored at –80 C. Riboprobes were reverse-transcribed in the presence or absence of digoxigenin (DIG)-labeled UTP (Roche, Indianapolis, IN) using a T3/T7 Megascript in vitro transcription kit (Ambion, Austin, TX) to produce antisense or sense DIG-labeled or unlabeled riboprobes. Slides were warmed to room temperature, air dried, and preequilibrated in hybridization buffer (50% formamide, 5x SSC, 5x Denhardt’s solution, 125 µg/ml Baker’s yeast tRNA, 250 µg/ml denatured herring sperm DNA) for 2 h at either 50 C (Wnt4, FoxL2) or 45 C (Dax1). Sections were incubated in riboprobe overnight at the same temperatures, respectively. Experimental slides were exposed to antisense DIG-labeled probe, whereas control slides were incubated in either sense DIG-labeled probe or a competitive series of antisense DIG-labeled/unlabeled probes. After RNase A treatment at 37 C, sections were washed in a decreasing series of SSC and equilibrated in 150 mM NaCl/100 mM Tris (pH 7.5) at room temperature before incubation in 1:5000 anti-DIG-alkaline phosphatase Fab fragments (Roche) in 0.5% Tween 20/PBS for 2 h at room temperature. Sections were washed in 100 mM Tris (pH 7.5) and incubated in 5 mM levamisole [in 100 mM Tris (pH 9.5), 100 mM NaCl, 50 mM MgCl2 buffer]. Chromogenic product was formed using nitroblue tetrazolium/5-bromo-4-chloro-3-indolyl-phosphate chromogen in 100 mM Tris (pH 8.0) (Roche) at 30 C until desired darkness was achieved and was terminated simultaneously for all slides within a gene. Sections were dehydrated, delipidated, and coverslipped under Permount (Fisher). Sets of gonads from five embryos per sex/stage were analyzed.
qPCR
Embryos were removed from the egg, staged, and killed by rapid decapitation. AKG complexes were immediately dissected and floated in sterile PBS. Pairs of gonads were cut away from underlying mesonephric tissues using fine scissors and a dissecting microscope from 12–20 embryos per independent sample (n = 3 at each sex/stage), pooled, immediately placed in RNA denaturing solution (Promega, Madison, WI), vortexed to dissociate, and stored at –80 C. A maximum of 5 min elapsed from decapitation to placement in denaturing solution. Total RNA was extracted using the RNAgents Total RNA Isolation kit (Promega) and treated with DNA-Free DNase I (Ambion, Austin, TX). cDNA was reverse-transcribed using the SuperScript First-Strand Synthesis for RT-PCR system (Invitrogen) with both oligo-(dT) and random hexamer primers. Relative gene expression levels were quantified using SYBR Green I dye (Invitrogen) and an ABI PRISM 7900HT real-time PCR cycler (ABI SDS 2.2.1 software). Samples were each run in triplicate, and the median value was used for analysis. PCR efficiencies were calculated from gene-specific standard curves. Relative transcript abundance was normalized to expression of protein phosphatase 1 (PP1), a constitutively expressed transcript across both stage and sex selected previously (9). A modified 
CT method that allows for correction of differential gene PCR efficiencies was used, where mean normalized expression (MNE) is calculated as follows: MNE = mean [(Eref– Ctref)/(Etarget– Cttarget)], where E is gene-specific PCR efficiency, Ct is cycle threshold from each independent sample, target is gene of interest, and ref is constitutively expressed reference gene (63, 64, 65). To examine expression fold changes within a gene, MNE values were calibrated to the lower value at stage 23 (either MPT or FPT). Primers used to assay gene expression were designed across exon boundaries for Wnt4 and Dax1 using MacVector; this was not possible for FoxL2, which contains a single exon. Primer specificity was verified by agarose gel electrophoresis, and for PP1, Dmrt1, Mis, and Sox9 were as previously described (9). qPCR primers for FoxL2, Wnt4, and Dax1 were as follows: FoxL2, forward 5'-TGG CAG AAC AGC ATC CGC-3' and reverse 5'-GGG TCC AGC GTC CAG TAG TTG-3'; Wnt4, forward 5'-CCG TAA CCG TCG CTG GAA C-3' and reverse 5'-GGA GGA GAT GGC ATA CAC AAA AGC-3'; Dax1, forward 5'-GGA CTG TGC TCT TCA ACC CG-3' and reverse 5'-GCT TGC TGT GCT TCC CTC TG-3'.
Stastical Analysis
For each gene, expression values measured by qPCR relative to the housekeeping gene (PP1) were plotted to test normality. Wnt4 and Dax1 data were found to have normal distributions, whereas FoxL2, Dmrt1, Mis, and Sox9 values were log-transformed to meet the normality assumption. A two-factor ANOVA (mixed data) on all MNE values within a gene tested for sex, stage, and sex by stage interaction effects (SAS software program; SAS Institute, Inc., Cary, NC). Post hoc pairwise comparison tests of all possible comparisons within a gene were analyzed (every combination of two MNEs), correcting for multiple tests using Tukey’s honestly significant difference (HSD).
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online August 7, 2007
Abbreviations: AKG, Adrenal-kidney-gonad; DIG, digoxigenin; FPT, female-producing temperature; GSD, genotypic sex determination; HSD, honestly significant difference; MNE, mean normalized expression; MPT, male-producing temperature; qPCR, quantitative real-time RT-PCR; SSC, standard saline citrate; TSD, temperature-dependent sex determination; TSP, temperature-sensitive period.
Received for publication May 21, 2007. Accepted for publication August 1, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  G.-C. Wu and C.-F. Chang wnt4 Is Associated with the Development of Ovarian Tissue in the Protandrous Black Porgy, Acanthopagrus schlegeli Biol Reprod, December 1, 2009; 81(6): 1073 - 1082. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. G. Dias and D. Crews Regulation of Pseudosexual Behavior in the Parthenogenetic Whiptail Lizard, Cnemidophorus uniparens Endocrinology, September 1, 2008; 149(9): 4622 - 4631. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
