| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Embryo |
a Murdoch Childrens Research Institute and Department of Paediatrics, The University of Melbourne, Royal Children's Hospital, Melbourne, Victoria 3052, Australia
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
developmental biology, embryo, estradiol, testis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although gonadal sex differentiation in birds closely resembles that in mammals, the sex-determining mechanism differs. In birds, the heterogametic sex is female (designated ZW), and the homogametic sex is male (designated ZZ). The sex chromosomes of birds are not homologous to those of mammals, and no SRY gene has been identified (reviewed in [3]). At present, the mechanism of avian sex determination remains unknown. It may involve a dominant ovary-determining gene carried on the W sex chromosome, or it may depend upon Z-chromosome dosage (one dose for female, two for male) [3]. A third possibility is that both mechanisms operate [4]. It was originally suggested that no dosage compensation of sex (Z)-linked genes occurs in birds [5]. However, results of recent studies of a larger number of Z-linked genes indicate that expression equalization between the sexes does occur for many genes [6]. If so, such dosage compensation appears not to depend upon Z-chromosome inactivation in males, because transcription has been observed from both Z chromosomes [7]. Compensation may therefore involve greater mRNA synthesis from the single Z-linked genes in females [8] or on posttranslational equalization. The issue of dosage compensation in birds is particularly relevant to sex determination. In the absence of compensation, a dosage-based sex-determining system is favored, whereby two doses of a gene product in ZZ embryos induce male development and one dose induces female development.
The chicken Z-linked gene, doublesex and Mab-3-related transcription factor 1 (DMRT1) has received considerable attention recently as a candidate testis-determining factor in birds [9–11]. DMRT1 belongs to a family of known or putative transcription factors sharing a conserved DNA-binding motif, the DM domain [12]. DMRT1 is conserved across vertebrates and is expressed in the embryonic gonads of fishes, reptiles, birds, and mammals [10, 13–15]. In all of these groups, expression is higher in developing male than in developing female gonads. Chicken DMRT1 (cDMRT1) is expressed exclusively in the embryonic urogenital system. Expression is higher in male (ZZ) gonads than in female (ZW) gonads prior to and throughout the period of gonadal sex differentiation [14]. Its location on the Z sex chromosome and its sexually dimorphic expression profile make DMRT1 a strong candidate testis-determining gene in the chicken embryo.
However, a role for DMRT1 in chicken gonadal development has not yet been demonstrated. The higher level of DMRT1 expression in ZZ embryos may be necessary for testis formation. Alternatively, because DMRT1 is located on the Z sex chromosome, its higher expression may reflect a lack of dosage compensation for this gene, and its sexual dimorphism may be unrelated to sex determination. A potential role for DMRT1 in chicken testis differentiation can be clarified by inducing experimental sex reversal, which is achieved in birds by blocking estrogen synthesis. Estrogen production is critical for ovarian differentiation in birds and other nonmammalian vertebrates [16]. The estrogen-synthesising enzyme P450 aromatase is expressed in female but not male gonads at the onset of gonadal sex differentiation [17, 18], and the administration of aromatase inhibitors results in testis differentiation in genetic females [19–21]. Conversely, administration of estrogen to genetic males causes transient feminization of the gonads [16]. In the present study, we examined the effects of aromatase inhibition and female-to-male sex reversal on the gonadal expression of DMRT1 at both the mRNA and protein levels. If DMRT1 plays an active role in testis differentiation, requiring higher expression in the male than in the female, then its expression is predicted to increase during female-to-male sex reversal despite the presence of only one Z chromosome. However, if higher expression in normal males simply reflects the presence of two Z chromosomes, then expression during female-to-male sex reversal should remain unchanged. Our results indicate that DMRT1 is upregulated during female-to-male sex reversal in ZW embryos, supporting an active role for this gene in chicken testis differentiation. In contrast, another candidate testis-regulatory gene, SOX9, shows delayed activation in sex-reversed embryos, suggesting that it may lie downstream in the avian sex-determination pathway.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Fertile chicken (Gallus domesticus) eggs were obtained from a commercial supplier and incubated under humid conditions at 37.8°C. On Day 4 of incubation (Hamburger and Hamilton stage 24) [22], eggs were treated with a single injection of fadrozole, a specific nonsteroidal aromatase inhibitor (AI) (Novartis, Basel, Switzerland). The fadrozole was dissolved in PBS at a concentration of 10 mg/ml, and 0.1 ml (1 mg) was injected into the albumen at the pointed end of each egg. Controls were treated with 0.1 ml of PBS alone. Holes were sealed with tape, and incubation was allowed to proceed. Paired urogenital systems were then harvested from embryos at stage 29/30 (Day 6–6.5), stage 32 (Day 7–7.5), stage 34/35 (Day 8–8.5), and stage 36 (Day 10.5). For some experiments, Day 14.5 (stage 40) embryos were also harvested. In the chicken embryo, the onset of gonadal sex differentiation is first apparent histologically at stage 30 (Day 6.5). Tissues were fixed and processed for either whole mount in situ hybridization or tissue section immunoflourescence. Some specimens were fixed in Bouin solution, embedded in paraffin, and processed for standard hematoxylin and eosin staining or immunohistochemical detection of aromatase protein.
Polymerase Chain Reaction for Sexing of Embryos
All embryos were sexed by polymerase chain reaction (PCR) amplification of a W-linked (female-specific) XhoI repeat sequence. At the time of dissection, a small piece of limb tissue was taken for sexing. Tissues were digested in 50 µl of PCR-compatible buffer at 55°C for 1 h [23] and then heated to 95°C for 10 min, and 3 µl was taken for PCR amplification. PCR was carried out using W-linked (female-specific) XhoI primers as described previously [17], except that chicken glyceraldehyde phosphate dehydrogenase (GAPDH) rather than ß-actin primers were used as internal controls. The GAPDH primers were cGAPDH/1 (5' CAG-ATC-AGT-TTC-TAT-CAG-C 3') and cGAPDH/2 (5' TGT-GAC-TTC-AAT-GGT-GAC-A 3'), amplifying a 600-base pair (bp) genomic fragment. Both primer pairs were included in a duplex reaction. By this rapid sexing method, only ZW (female) embryos showed amplification of the XhoI fragment.
Whole Mount In Situ Hybridization
Gonadal cDMRT1 transcripts were detected in control and AI-treated embryos using a sensitive whole mount in situ hybridization protocol. Urogenital systems were dissected from embryos at various time points following treatment and were fixed overnight at 4°C in 4% paraformaldehyde in diethyl pyrocarbonate-treated PBS. Tissues were pooled by sex and divided into groups for whole mount analysis of DMRT1 and SOX9 mRNA expression. Chicken SOX9 (cSOX9) is only expressed in male gonads at the onset of testicular differentiation on Day 6.5 [24, 25]. Whole mount in situ hybridization was carried out as described previously [26]. Tissues were dehydrated through graded methanols, rehydrated into PBS containing 0.1% Triton X-100, permeablized with proteinase K (10 µg/ml), washed, and placed in prehybridization solution overnight at 65°C. Digoxygenin (DIG)-labeled antisense and sense cRNA probes were synthesized from cloned cDNA templates using a DIG-labeling kit (Roche, Indianapolis, IN). For cDMRT1, a specific 366-bp fragment downstream of the DM domain was used. For cSOX9, a 388-bp fragment from the 3' untranslated region was used. These probes were the same as those used previously to detect cDMRT1 and cSOX9 mRNA by whole mount in situ hybridization [14, 24]. Following overnight hybridization at 65°C, tissues were washed in 2x saline-sodium citrate (SSC)/0.5% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) at 65°C and then in 0.2x SSC/0.5% CHAPS at 65°C, blocked, and incubated overnight in alkaline phosphatase-conjugated DIG antibody (Roche). Tissues were then extensively washed in Tris-Triton X-100 buffer prior to incubation in 5-bromocresyl-3-indolylphosphate/nitroblue tetrazolium chromogen. All tissues were left in chromogen solution for the same length of time (3 h) and photographed under the same conditions. At least four urogenital systems were used for each treatment at each time point.
Relative Quantitative Reverse Transcription PCR
DMRT1 gene expression was examined by relative quantitative reverse transcription PCR (RQ RT-PCR), using SYBR Green detection and real time PCR. Fertile eggs were treated with fadrozole or PBS control on Day 3 as described above, and gonadal pairs were harvested on Days 6, 7, 8, and 10. Embryos were sexed by PCR and pooled by sex (five to eight gonadal pairs per treatment). Total RNA was extracted from the pooled samples using the guanidinium thiocynate method, and reverse transcription (RT) was carried out using random hexamers plus oligo d(T)16 primers in the presence of murine leukemia virus reverse transcriptase (42°C for 30 min). To control for genomic DNA contamination, some samples were included that lacked reverse transcriptase. Successful RT was confirmed for all samples by then carrying out PCR amplification of the housekeeping gene, GAPDH. GAPDH cDNA bands were seen in all samples except for the RT-negative controls.
Specific DMRT1 cDNA amplification was then carried out on duplicate samples using Taq DNA polymerase and SYBR Green I Mastermix (Applied Biosystems, Foster City, CA). SYBR Green dye binds nonspecifically to double-stranded DNA. DMRT1 primers were designed such that one spanned the junction of two exons, giving a single cDNA PCR product and precluding amplification of genomic DNA. Similarly, specific chicken ß-actin PCR primers were designed for use as normalization controls. In pilot experiments using these primer pairs, single DMRT1 and ß-actin cDNA PCR products were obtained, as confirmed by sequence analysis. Real time PCR was carried out in an Prism 7700 thermocycler (Applied Biosystems) at a 60°C annealing temperature for 40 cycles according to the manufacturer's recommendations. During PCR, fluorescence accumulation resulting from DNA amplification was recorded using the Sequence Detector software (Applied Biosystems). PCR products were resolved on an agarose gel to confirm amplification. No PCR products were seen in the RT- or water control lanes. Cycle threshold (CT) values were obtained from the exponential phase of PCR amplification, and DMRT1 expression was corrected against ß-actin expression, generating a 
CT value (CT = DMRT1 CT - ß-actin CT). Relative expression was then expressed according to the equation 2-CT.
Polyclonal Antibody Production
Polyclonal anti-chicken cDMRT1 and aromatase antibodies were raised in rabbits using 16-amino acid peptides as antigens conjugated to diphtheria toxoid (synthesized by Mimotopes, Clayton, Australia). For DMRT1, the peptide CPSIPSRGHLESTSDL was used (the cysteine at the N terminus is nonnative, for conjugation). This peptide is outside the conserved DM domain and was predicted by computer modeling to lie at the surface of the protein (Mimotopes). For chicken aromatase, a sequence at the C terminus, CEMVFTPRSPNKNQSD, was used. This antibody was raised to monitor aromatase protein expression in normal and AI-treated embryos, using immunohistochemistry on paraffin-embedded sections with an alkaline phosphatase-conjugated anti-rabbit secondary antibody. Previous studies indicated that AI treatment leads to dramatic reduction in aromatase expression [21, 27]. For each peptide, one rabbit was immunized. An ELISA of serum revealed a strong immune response, and the antibodies were purified by protein A sepharose chromatography.
Western blot analysis was carried out to determine the specificity of the DMRT1 antibody. Mixed-sex embryonic chicken gonads were homogenized in 100 µl of single detergent lysis buffer containing 50 mM Tris-HCl (pH 8), 150 mM NaCl, 1% NP40, and protease inhibitor cocktail (Sigma, St. Louis, MO) using a dounce homogenizer. Protein concentration was determined using a bicinchoninic acid assay (BCA; Pierce, Rockford, IL). A 12.5- to 50-µg sample of gonad lysate was loaded into each lane with a prestained protein size marker, electrophoresed through a 10% SDS-polyacrylamide gel, and electroblotted onto a polyvinylidene difluoride membrane using a wet electroblotter. Membranes were blocked in skim milk powder in Tris-buffered saline with Tween (TBST; 5% skim milk powder, 25 mM Tris-HCl, 137 mM NaCl, 2.7 mM KCl, and 0.2% Tween 20, pH 7.4) for 1 h at room temperature. After blocking, membranes were incubated with either purified anti-DMRT1 antibody or anti-DMRT1 antibody preadsorbed against 100x molar excess antigen (5 µg/ml) at 4°C for 16 h in 0.5% skim milk powder/TBST solution. Blots were washed and incubated with peroxidase-labeled goat anti-rabbit antiserum (1:15 000), and signal was detected using an ECL Plus kit (Amersham Pharmacia, Piscataway, NJ) and X-OMAT film (Kodak, Rochester, NY).
Immunofluorescent Detection of DMRT1 Protein
Frozen sections were used for immunofluorescence. Urogenital systems from control and AI-treated embryos were briefly fixed in 4% PFA (paraformaldehyde)/PBS at room temperature for 10 min and then cryoprotected in 20% sucrose/PBS overnight at 4°C. Tissues were transferred to embedding compound, snap frozen in isopentane precooled in liquid nitrogen, and stored at -70°C. Frozen sections (10 µm) were thaw mounted onto silane-treated slides and permeablized in 1% Triton X-100/PBS for 10 min at room temperature. Following PBS washes, sections were blocked in 1% BSA in PBS for 2 h at room temperature and then incubated with primary antibody (1 µg/ml) overnight at 4°C. Sections were washed, incubated with fluorescently labeled goat anti-rabbit secondary antibody (ALEXA 488; Molecular Probes, Eugene, OR), mounted in Fluorosave (Calbiochem, Bad Soden, Germany), and viewed with fluorescence optics. At least three urogenital systems were sectioned for each treatment at each time point. For both cDMRT1 and aromatase, specific staining was abolished following preadsorption of the antibodies with the relevant unconjugated peptide. Staining was also abolished when primary antibody was replaced with preimmune serum.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In chicken embryos, the gonads appear as mesodermal thickenings on the ventromedial surface of the mesonephric kidneys at Day 3.5 (stages 21 and 22). The undifferentiated gonads comprise an outer epithelial cell layer (cortex) and underlying medulla. In the medulla, cords of epithelium-like cells (medullary cords) are interspersed with mesenchyme. In both sexes, the germ cells are scattered throughout the gonad but are more concentrated in the cortex. The gonads remain undifferentiated until Day 6.5 (stage 30), when the first histological signs of sexual differentiation become apparent. By Day 10.5 (stage 36), differentiation is advanced. In ZW female embryos examined here, the gonads showed their typical asymmetric development. The left gonad had become a large ovary, characterized by a thickened surface epithelium (cortex) containing proliferating germ and somatic cells (the site of subsequent folliculogenesis). In this left ovary, the medullary cords were thin and elongated, comprising flattened epithelial cells, and were separated by lacunae (luminal spaces; Fig. 1A). The right gonad of female embryos was significantly smaller than the left gonad and did not develop as an ovary. The medulla was lacunar, but the cortex did not develop (Fig. 1B). In ZZ male embryos, bilateral testes were evident by Day 10.5 (stage 36). Sertoli cells had differentiated within the medullary cords, giving rise to seminiferous cords, and the surface epithelium was reduced (Fig. 1F). The interstitium was well developed.
|
In comparison to control female gonads, those of AI-treated ZW embryos were strongly masculinized (Fig. 1, C and D). By Day 10.5, the left gonad was significantly smaller than that of PBS-treated control females, and the cortex failed to develop in most individuals (Fig. 1C). The medulla showed poorly organized cords, some with lacunae (female type) and some arranged as seminiferous (male type) cords (Fig. 1C). These left gonads were considered strongly masculinized. The right gonads of AI-treated ZW embryos were more masculinized and were considered testicular. Seminiferous cords were present throughout the medulla, and the cortex was reduced to an epithelial monolayer, as in control males (Fig. 1D). However, fewer seminiferous cords were present than in normal males, and the interstitium was unusually extensive (compare Fig. 1D and Fig. 1F). High magnification of the cords from the right gonads of AI-treated embryos showed histologically recognizable Sertoli cells (highly polarized cells with basal nuclei and apical cytoplasm encircled by basal lamina; Fig. 1E). Germ cells could also be seen within these apparent seminiferous cords (Fig. 1E), as occurs in normal males. Immunohistochemical staining for aromatase protein revealed strong expression in the medullary cords of both left and right control female gonads (Fig. 2, A and B) but dramatically reduced expression in AI-treated females (Fig. 2, C and D). Although there was some variation between individual AI-treated embryos, aromatase protein expression was always more strongly inhibited in the right than in the left gonad. Aromatase was not detected in control or AI-treated male tissues (not shown).
|
DMRT1 Expression in Control and AI-Treated Embryos: Whole Mount In Situ Hybridization
Whole mount in situ hybridization revealed DMRT1 gene expression in the gonads and Müllerian ducts, with stronger expression in control male gonads than in control female gonads. AI-treated female embryos showed a malelike pattern of DMRT1 expression, i.e., stronger expression than in control females (Fig. 3). The first time point examined after AI treatment on Day 3.5 was Day 6.5 (stage 30), which corresponds to the beginning of aromatase expression in normal females and the onset of morphological sexual differentiation in both sexes. At this time, DMRT1 was expressed in the gonads of both sexes, but expression was higher in male (ZZ) gonads than in both control and AI-treated embryos (not shown). However, at Day 7.5 (stage 32), DMRT1 expression in AI-treated female embryos was clearly stronger than that in control females and was comparable to that of control males (Fig. 3, top panels). At this stage, the pattern of DMRT1 mRNA expression became punctate or granular in the gonads of male and AI-treated embryos. The malelike pattern of DMRT1 expression in AI-treated ZW embryos was maintained up to at least Day 10.5 (stage 36; Fig. 3, bottom panels). At Day 10.5, the gonads of AI-treated embryos were morphologically more similar to those of control males than to those of control females (Fig. 3). In the left and right gonads of both control males and AI-treated females, DMRT1 showed a strong punctate expression pattern. In contrast, a weaker, more diffuse expression pattern was seen in the left ovary and regressing right gonad of control females (Fig. 3, bottom panels). For all of these developmental stages, AI-treated male embryos showed DMRT1 expression that was indistinguishable from that of control (PBS treated) males.
|
Relative Quantitative RT-PCR
The RQ RT-PCR analysis showed that DMRT1 expression was higher in male than in female control gonads from Day 6 to Day 10 (Fig. 4). This sexually dimorphic expression pattern confirmed the results of the whole mount in situ hybridization. The higher level of expression in males represented a >2-fold difference, implying active upregulation in males or downregulation in females. At Day 6, DMRT1 expression in AI-treated ZW females was low and similar to that of control females. However, by Day 7 expression had increased to male levels, a trend that was maintained during subsequent development (Fig. 4). DMRT1 expression in AI-treated male tissues was similar to that of control males at Days 8 and 10 but was actually higher at Days 6 and 7.
|
DMRT1 Immunofluorescence
The stimulatory effect of fadrozole treatment on DMRT1 expression in ZW embryos was studied in more detail at the protein level, using a specific antibody. The polyclonal chicken DMRT1 antibody raised in rabbit revealed a band of approximtely 34 kDa by Western blotting (Fig. 5A). This band was abolished following preadsoprtion of the antibody with 100x excess antigen (cDMRT1 16-amino acid peptide). Indirect immunofluorescence with this antibody showed staining specifically in embryonic chicken gonads, with stronger staining in males than in females. Specific immunofluorescence was abolished following preadsorption of the antibody with excess antigen (Fig. 5B).
|
In embryos examined at Day 6.5 (stage 30), DMRT1 protein was detected by immunofluorescence in the nuclei of the medullary cords in both genetic sexes (Fig. 6, A–C). At this stage, expression was already higher in male gonads than in both control and AI-treated female gonads. This time point corresponds to the onset of gonadal sex differentiation and aromatase expression in normal females [17], and the effects of aromatase inhibition are expected to appear at or after this time. At Day 7.5 (stage 32), DMRT1 protein in AI-treated embryos had become elevated (Fig. 6, F and G). In control females, DMRT1 expression remained low in the medullary cords (Fig. 6, D and E). In contrast, stronger staining for DMRT1 protein was detectable in Sertoli cells organizing into seminiferous cords of control males (Fig. 6H). A similar level of DMRT1 expression and malelike pattern of cord organization was seen in both gonads of AI-treated ZW embryos at this stage (Fig. 6, F and G). This pattern of elevated DMRT1 protein in AI-treated female embryos was maintained at Day 8.5 (stage 34), when the right gonad in particular showed a malelike pattern of cord organization and DMRT1 expression in solid male-type cords (Fig. 6, I–M).
|
At Day 10.5 (stage 36), the left and right testes of control males showed strong expression of DMRT1 protein in Sertoli and germ cells within seminiferous cords but not in the interstitium (Fig. 6R). Similarly, in AI-treated females, strong staining for DMRT1 was detectable in the seminiferous cords of both left and right sex-reversed gonads. In the left gonads of these embryos, both male-type (seminiferous) and female-type (lacunar) cords were present (Fig. 6P). The female-type cords with lacunae comprised either flattened or flattened and thickened epithelial cells. Strongly DMRT1-positive cells were observed in the male-type cords and in the thickened but not flattened epithelial cells of the female-type cords (Fig. 6P). In the right testes of AI-treated embryos, virtually all cords were of the male type (solid, lacking lacunae), and all stained strongly for DMRT1 protein (Fig. 6Q). Control female embryos continued to show lower levels of DMRT1 protein in the medulla at Day 10.5 (Fig. 6, N and O). However, strong expression was detectable in the germ cell population at this stage in normal females; in the left gonad, positive germ cells were concentrated in the proliferating gonadal cortex (Fig. 6N), and in the right gonad these cells were occasionally seen scattered through the medulla (Fig. 6O). As in the whole mount studies, AI-treated male embryos showed DMRT1 protein expression that was similar to that of normal control males.
SOX9 Expression in Control and AI-Treated Embryos
Chicken SOX9 is normally expressed only in male gonads at the onset of gonadal sex differentiation (Day 6.5, stage 30). This gene has been considered a candidate testis-determining gene in the chicken embryo. In the embryos studied here, whole mount in situ hybridization analysis showed no detectable expression of SOX9 in control females at any of the days tested (Days 8.5–10.5, stages 34–35; Fig. 7, A and D). In contrast, strong SOX9 expression was first detectable in male gonads at Day 6.5 (stage 30; not shown). Over Days 8.5–10.5, strong expresson was maintained in males (Fig. 7, C and F). In the AI-treated ZW (sex reversed) embryos, SOX9 expression was weakly apparent in the gonads at Day 8.5 (Fig. 7B) and more strongly apparent at Day 10.5 (Fig. 7E). In these embryos, expression appeared stronger in the right gonad, consistent with its greater degree of masculinization.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
These data indicate that DMRT1 expression is correlated with testis development in the chicken embryo and that inhibition of aromatase activity allows upregulation of DMRT1 to occur during female-to-male sex reversal. Elevation of DMRT1 expression in sex-reversed ZW female embryos occurs despite the presence of only one Z chromosome. Higher levels of DMRT1 expression can therefore be correlated with testicular differentiation per se and do not solely reflect a lack of dosage compensation for this Z-linked gene. DMRT1 protein is expressed in the nuclei of medullary cord cells (Figs. 5 and 6), which give rise to Sertoli cells and seminiferous cords. Several lines of evidence indicate that testicular morphogenesis in vertebrates begins with and depends upon Sertoli cell differentiation [28–30]. The results described here show a correlation between seminiferous cord formation and elevated DMRT1 expression in normal and sex-reversed chicken gonads.
The RQ RT-PCR studies indicated that the higher levels of DMRT1 expression in normal males versus females is unlikely to be attributable to gene dosage differences alone. If it assumed that DMRT1 escapes dosage compensation, males (ZZ) would show twice the level of expression of females (ZW). However, DMRT1 expression in male gonads was >2-fold higher than that in female gonads (Fig. 4). This finding implies active upregulation of gene expression in males, which also occurs in ZW sex-reversed embryos (Fig. 4). Male embryos treated with AI also showed elevated DMRT1 expression at Days 6 and 7 (Fig. 4), although this elevation was less marked at Days 8 and 10. The reasons for this increased expression are unclear, but fadrozole treatment might stimulate DMRT1 expression in males by a mechanism unrelated to aromatase inhibition (aromatase is not expressed in males).
In normal chicken embryos, DMRT1 is expressed in the gonads prior to and during the onset of sexual differentiation (from at least as early as Day 4.5, stage 26). At all stages examined, expression is higher in males than in females [10]. In contrast, aromatase gene expression begins at Day 6–6.5 (stages 29 and 30) and is only seen in the medulla of female gonads. In the embryos studied here, DMRT1 expression was similar in control and AI-treated embryos until the normal time of aromatase gene activation (Day 6.5), following which (at Day 7.5) DMRT1 expression became elevated in treated females (Figs. 4 and 6). This finding suggests that the absence of normal aromatase expression at Day 6.5 allowed subsequent upregulation of DMRT1 expression from the single Z chromosome of ZW females. In normal ZW embryos, aromatase gene expression may therefore contribute to the downregulation or inhibition of DMRT1, either directly or indirectly. One of the functions of estrogen synthesis in ZW embryos may be to negatively regulate DMRT1 from Day 6.5, the onset of ovarian development.
However, because DMRT1 is already expressed at lower levels in ZW female embryos prior to the onset of aromatase gene expression [14, 31], aromatase/estrogen synthesis cannot be responsible for the initial sexual dimorphism in DMRT1 expression. This dimorphism could be explained by 1) higher expression in males, if not dosage compensated, 2) positive regulation of the DMRT1 locus only in males, or 3) negative regulation of the locus in females. This last alternative implies inhibition by a female-specific (W-linked) factor. The two Z chromosomes are heavily methylated in male chickens, whereas the single Z of females is less methylated [32]. The so-called male hypermethylated region (MHM) is transcribed into high molecular weight noncoding RNA only in females, and this RNA accumulates at the site of transcription very close to the DMRT1 locus. Using triploid (ZZZ and ZZW) chicken cells, Teranishi et al. [32] showed that the W chromosome in females induces hypomethylation of the MHM on the Z chromosome, allowing transcription of the noncoding RNA. By analogy with Xist in mammals, this RNA would inhibit nearby genes (such as DMRT1). A factor on the W chromosome could therefore downregulate DMRT1 in ZW embryos by reducing methylation in the region and allowing inhibitory RNA to bind at the DMRT1 locus. In ZZ embryos, the region would remain hypermethylated, with no RNA transcribed and DMRT1 not inhibited. This process could generate a mechanism whereby DMRT1 is sexually dimorphic between the sexes. The studies reported here suggest that by blocking aromatase, and hence estrogen synthesis, this putative inhibition of DMRT1 is overcome in ZW females. Because no estrogen receptor 
(ER) has been detected in the right gonad [33], this action of estrogen may be mediated in the right gonad by another mechanism.
Aromatase Inhibition Causes Upregulation of Gonadal DMRT1 Expression
Fadrozole has previously been shown to be a potent and specific inhibitor or aromatase enzyme in embryonic chicken gonads, inducing testicular differentiation and female-to-male sex reversal [19–21, 27, 34] Although fadrozole inhibits the enzyme itself, expression of aromatase protein (Fig. 2) and mRNA [27] are both significantly reduced, which indicates that aromatase gene expression is being blocked. Therefore, estrogen (or a target of estrogen) must be necessary for maintenance of aromatase expression, and inhibition of estrogen synthesis by fadrozole blocks this function. Chicken ER is expressed in the left but not the right female gonad (at least from Day 7) [26, 33]. However, in the right gonad, aromatase is expressed at this time, and this expression is inhibited by fadrozole [27]. Thus, either estrogen is acting on the aromatase gene in the right gonad via a different receptor (such as ERß) or it is acting indirectly. Expression of ERß has not been studied in the chicken embryo.
The morphological effects of fadrozole treatment reported here are consistent with previous findings. Vaillant et al. [21, 34] described masculinization of embryonic and post-hatching chicken gonads with fadrozole. They found that medullary cords of AI-treated ZW embryos progressively thickened and differentiated into seminiferous cords, as found here. In addition to seminiferous cords, Vaillant et al. [34] noted two types of lacunar medullary cords in the sex-reversed females: cords with flattened epithelium (typical of females) and composite cords with both flattened and thickened cells. Studying embryos at Day 14, they found that AMH, SF1, and SOX9 were all expressed in the seminiferous cords of AI-treated females and in the thickened but not flattened regions of the composite cords. Although SF1 and AMH are also expressed in normal female gonads, SOX9 expression is only seen in developing male (Sertoli) cells. Vaillant et al. [34] therefore postulated that the composite cords of AI-treated females represented lacunar medullary cords transdifferentiating into (male) seminiferous cords, expressing the Sertoli cell marker SOX9. Such composite medullary cords were also noted in the present study, in addition to male-type seminiferous cords in the left gonads of AI-treated ZW embryos. Strong DMRT1 expression was seen in the thickened but not the flattened cells of composite cords (Fig. 6P). In the right gonads, virtually all cords were male type (seminiferous), and DMRT1 was strongly expressed in the cells of these cords, as in normal males. The results of Vaillant et al. and those of the present study suggest that differentiation of seminiferous cords in ZW embryos involves upregulation of DMRT1 in the medullary cells.
Stronger masculinization of the right than of the left gonad of AI-treated embryos can be correlated with the greater degree of aromatase inhibition. Aromatase was more strongly inhibited in the right gonad (see Fig. 2), and residual aromatase expression in the left gonad may account for the less complete and more variable masculinization. Thus, in the left gonad of AI-treated females, some female-type (lacunar) and composite cords were still present. The presence of residual aromatase activity may prevent differentiation of all medullary cords into seminiferous cords. Under this scenario, aromatase expression and estrogen synthesis are correlated with the development of lacunar female-type medullary cords [34], whereas a lack of aromatase and an upregulation of DMRT1 are correlated with the development of male-type (seminiferous) cords. The results reported here and those described previously suggest that aromatase expression in the medullary cords of ZW embryos plays a role in ovarian differentiation, whereas elevated DMRT1 expression in the cords of ZZ embryos is involved in testicular differentiation.
SOX9 Expression in Sex-Reversed ZW Embryos
In male embryos, DMRT1 alone is unlikely to be responsible for initiating Sertoli cell differentiation. High DMRT1 expression in ZZ embryos (as early as Day 4.5) precedes the onset of Sertoli cell differentiation (from Day 6.5). Thus, another factor may be involved. One candidate is the transcription factor SOX9, which is only expressed in male medullary cords from Day 6.5 [24]. Male-specific SOX9 expression is highly conserved and has been considered a marker of Sertoli cell differentiation in chicken embryos and in humans, mice, and other vertebrates (reviewed in [35]). Vaillant et al. [34] found that SOX9 is activated in the transdifferentiating medullary cords of AI-treated chicken embryos undergoing female-to-male sex reversal. The embryos in that study were first examined at Day 14, when gonadal sex differentiation is complete. In the sex-reversed ZW embryos studies here, SOX9 expression was not seen at Day 6.5, but weak expression was first detected by whole mount in situ hybridization at Day 8.5 (stage 34), becoming stronger by Day 10.5 (stage 36) (Fig. 7). (More sensitive RT-PCR might show that SOX9 expression begins earlier in AI-reated embryos.) Increased expression may reflect increasing differentiation of seminferous cords during female-to-male sex reversal. Expression of SOX9 in AI-treated embryos was never as strong as that seen in normal males, however (Fig. 7), consistent with the smaller number of seminiferous cords observed in the AI-treated embryos (Fig. 1).
The studies described here show that testis differentiation is correlated with an increase in DMRT1 expression in normal and sex-reversed chicken embryos. Testis formation in ZW (female) embryos shows that two copies of the DMRT1 gene are not necessary for testicular differentiation. However, upregulation of DMRT1 gene expression is correlated with differentiation. This finding provides some insight into the mechanism of avian sex determination. The 5' regulatory region of cDMRT1 should be examined to gain an understanding of how this gene is regulated during gonadal development. Induction of female-to-male sex reversal by the overexpression of DMRT1 in ZW embryos would also be informative.
|
FOOTNOTES |
|---|
2 Correspondence. FAX: 61 3 9345 6000; smithc{at}cryptic.rch.unimelb.edu.au
Received: 16 May 2002.
First decision: 5 June 2002.
Accepted: 22 August 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Ishimaru, T. Komatsu, M. Kasahara, Y. Katoh-Fukui, H. Ogawa, Y. Toyama, M. Maekawa, K. Toshimori, R. A. S. Chandraratna, K.-i. Morohashi, et al. Mechanism of asymmetric ovarian development in chick embryos Development, February 15, 2008; 135(4): 677 - 685. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Liu, F. Wu, B. Jiao, X. Zhang, C. Hu, B. Huang, L. Zhou, X. Huang, Z. Wang, Y. Zhang, et al. Molecular cloning of doublesex and mab-3-related transcription factor 1, forkhead transcription factor gene 2, and two types of cytochrome P450 aromatase in Southern catfish and their possible roles in sex differentiation J. Endocrinol., July 1, 2007; 194(1): 223 - 241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. V. Prasanth and D. L. Spector Eukaryotic regulatory RNAs: an answer to the 'genome complexity' conundrum Genes & Dev., January 1, 2007; 21(1): 11 - 42. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. J. Vallender and B. T. Lahn Multiple independent origins of sex chromosomes in amniotes PNAS, November 28, 2006; 103(48): 18031 - 18032. [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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS |
