| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Section of Integrative Biology, University of Texas, Austin, Texas 78712
Address all correspondence and requests for reprints to: Dr. David Crews, Section of Integrative Biology, Patterson Hall, University of Texas, Austin, Texas 78712. E-mail: crews{at}mail.utexas.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We examined gonadal SF-1 expression in embryos 1) incubating at three different temperatures, 2) after treating eggs with estrogen at a male-producing temperature, and 3) after inhibition of estrogen biosynthesis at a female-producing temperature. Our findings suggest that both temperature and estrogen lie upstream of SF-1 in a sex-determining regulatory hierarchy in red-eared slider turtles and that estrogen directly or indirectly modulates the regulation of SF-1 expression.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
/ß estrogen
receptors (5), and exogenous estrogen feminizes the gonad
in genetic male marsupials (6). GSD also occurs in birds
where females are ZW heterogametic. In birds, SRY is apparently not
present in either sex, and an ovary-determining mechanism is as yet
unknown, but estrogen does appear essential for female sex
determination. In chickens, for example, inhibiting aromatase, the
enzyme that converts testosterone to estrogen, causes genotypic females
to develop as phenotypic males (7), and exogenous estrogen
causes transient feminization of genotypic males (8). Temperature-dependent sex determination (TSD) occurs in all crocodilians, most turtles, and some lizards (9). These reptiles are genetically bipotential (10) and no functional SRY has been identified (11, 12). Estrogen is implicated in female sex determination of reptiles and specifically in conveying the effect of temperature in TSD reptiles (1, 2, 13, 14). In the red-eared slider turtle (Trachemys scripta elegans), a TSD species, application of 17ß-estradiol (E2) on eggs incubating at an all male-producing temperature results in 100% female hatchlings (9). This effect of E2 treatment is both time- and dose-dependent. The period of E2 sensitivity overlaps the temperature-sensitive period (TSP) and is synergistic with the effect of temperature. Hatchlings that result after E2 treatment at an all male-producing temperature are morphologically and histologically equivalent to those incubated at a temperature that produces only females. Application of aromatase inhibitor (AI) to eggs incubating at a female-biased temperature results in 100% male hatchlings. Higher doses of AI are required to produce this effect at increasingly female-biased temperatures. Male hatchlings from this treatment group are morphologically and histologically comparable to normal, temperature-derived males. Throughout the TSP, estrogen receptor is expressed in gonads of both putative males and females (15). Despite much effort, however, the initial endogenous source of estrogen in TSD reptiles remains unclear (13). Furthermore, little is known about molecular mechanisms underlying endogenous or manipulated sex determination in TSD species. Despite the apparent lability of triggers that initiate sex determination among amniotes, subsequent gonadogenesis and the underlying molecular components identified to date are largely conserved.
Steroidogenic factor 1 (SF-1) (16), also named Ad4BP (17), is known in mammals to regulate many genes in the reproductive axis. In steroidogenic tissues, SF-1 regulates the expression of most enzymes required to synthesize estrogen from cholesterol: cytochrome P450 steroid hydroxylases, including aromatase, and 3ßHSD (18, 19). In addition, SF-1 is preferentially expressed in developing testes compared with ovaries. SF-1 is, therefore, of clear interest in the study of sex determination in birds and reptiles. It has been cloned and its expression examined in chickens (20, 21), alligators (22), and red-eared slider turtles (14, 23). Regulation of SF-1 itself has not yet been fully characterized in any organism.
In this study we asked whether manipulating sex determination by shifting the incubation temperature, by treatment with E2, or by treatment with AI to inhibit estrogen biosynthesis would affect the expression of SF-1 in red-eared slider turtles. Adrenal-kidney-gonad complexes in red-eared slider turtles cannot be separated until late in gonadogenesis, and SF-1 is strongly expressed in adrenal as well as gonad (23). To quantify SF-1 message in gonad alone we therefore used in situ hybridization analysis. Our findings support the involvement of SF-1 in the regulatory hierarchy of TSD in red-eared slider turtles and show that incubation temperature and estrogen directly or indirectly modulate the regulation of SF-1 expression.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Probe preparation
Cowan and Wibbels (14) provided a 457-bp
clone of red-eared slider turtle SF-1 that we subcloned into the pCRII
vector (Invitrogen, San Diego, CA) to eliminate most
sequences in common with an alternate transcript, ELP1, as defined in
mouse gene structure (25). Before performing in
situ hybridizations, a Southern blot analysis was conducted to
determine the binding specificity of probes derived from this 330-bp
subclone (Fig. 1
). Genomic DNA was
isolated according to the method of Sambrook et al.
(26) from multiple red-eared slider turtle embryos and
digested with BamHI, EcoRI, or HindIII
(New England Biolabs, Inc., Beverly, MA). Three sets of
the three digests were electrophoresed (10 µg/lane) simultaneously on
a 0.7% agarose gel and transferred to a BrightStar Plus nylon membrane
(Ambion, Inc., Austin, TX). The membrane was baked for 30
min at 80 C, then cut into thirds. Each of the resulting equivalent
blots was hybridized to one of three DNA probes (Fig. 1A). Probes were
synthesized by linear amplification using PCR Strip EZ (Ambion, Inc.) and [-32P]deoxy-ATP (3000
Ci/mmol, 10 mCi/ml; NEN Life Science Products-DuPont,
Boston, MA) to a specific activity of more than
109 cpm/µg and were used at a concentration of
106 cpm/ml UltraHyb hybridization solution
(Ambion, Inc.). Blots were hybridized overnight at 43 C,
then washed for 40 min (with one solution change) at each of three
stringencies: 0.2 x SSC (standard saline citrate)/0.1% SDS at
room temperature, 0.1 x SSC/0.1% SDS at 43 C, and 0.1 x
SSC/0.1% SDS at 68 C (26) (Ambion, Inc.).
Bound probe was visualized by PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, CA; Fig. 1B) at each level of stringency (final
wash shown). Southern blot findings combined with previous Northern
blot analyses (23) indicate that the 330-nucleotide probe
binds a single copy of SF-1 only.
|
-35S]CTP from NEN Life Science Products-DuPont. Riboprobe was synthesized to a specific
activity of 8 x 108 cpm/µg and
used at a concentration of 0.3 µg probe x length (kb)/ml
hybridization solution.
In situ hybridization
All eggs were initially incubated at 26, 29.4, or 31 C. At stage
17 they were randomly divided into treatment and control groups. Eggs
in the continuous temperature groups, which also served as controls for
manipulations below, were retained at their starting temperatures. At
stages 15, 17–19, 21, and 23, a subset of these eggs was taken from
each temperature, and their embryos were quickly frozen on dry ice,
then stored at -80 C for later analysis.
Four other groups were treated at stage 17. A group of eggs
incubating at the all male-producing temperature of 26 C was shifted to
the all female-producing temperature of 31 C. Another group was shifted
from 31 to 26 C. A group of eggs incubating at 26 C was treated with 5
µg E2 (Sigma, St. Louis, MO) in 5
µl 95% ethanol, as previously described (9).
Carrier controls were treated with ethanol only at 26 C. Finally, a
group of eggs incubating at 29.4 C was treated with 100 µg (in 5 µl
95% ethanol) of the aromatase inhibitor fadrozole (CGS16949A,
Ciba-Geigy, Summit, NJ). Carrier controls were spotted
with ethanol only at 29.4 C. Subsets of eggs from each treatment group
were taken at stages 18, 19, 21, and 23, and the embryos were
frozen at -80 C until use. A subset of individuals from all groups
above were maintained until hatching, then sexed (Table 1). Two in situ hybridizations
were performed to accommodate the large number of individuals assayed.
To determine whether statistical comparisons could be made between the
two assays, consecutive slides of 10 29.4 C untreated individuals from
3 different stages were assayed in each experiment. Using the JMP
program (version 3.2.2, SAS Institute, Inc., Chicago, IL),
a multivariate ANOVA was performed on resulting SF-1 expression values.
No significant difference was found between assays (F = 0.71;
P = 0.42).
|
Darkfield quantification of silver grains in specifically labeled
cells, defined as having a density of silver grains at least three
times that of background, has been previously described (15, 28). Briefly, slides were computer coded and randomized. The
Grains Counting Program (University of Washington) then selected 45 of
the most densely labeled clusters, each cluster approximating the size
of a single cell, from the middle of a gonad in each individual (5
clusters, from 3 representative fields in each of 3 sections) and
counted the silver grains in each cluster. Average grains per cluster
were corrected for background in each individual by subtracting average
grains per cluster measured in the adjacent kidney tissue. Using the
JMP program, the corrected individual measures were used to calculate
means and SEs and to perform two-way ANOVAs (Figs. 2-5). Post-hoc comparisons
were made using the Dunn-Sidak method to provide a significance level
of ' = 1 - (1 - 0.05)1/k = 0.0012,
where k is the number of individual comparisons for an experimentwise
= 0.05 (29).
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) (30, 31). Using semiquantitative
in situ hybridization, we compared SF-1 gonadal expression
in embryos incubating at these temperatures from stages spanning the
TSP and found distinct male and female patterns. SF-1 expression was
significantly higher at the male- than the female-producing temperature
in all stages examined (Fig. 2A). These data expand our earlier
descriptive evidence for differential expression of SF-1
(23).
At stages 15 and 17, SF-1 messenger RNA (mRNA) appeared evenly
dispersed in the bipotential gonads of embryos at both male- and
female-producing temperatures. From stages 18–23, as early sex tubules
proliferate in putative testes (30), message became
localized in cells of the tubules and, to a lesser extent, the
interstitial space (Fig. 2B). In putative ovaries during the same
period, when immature sex cords regress and the cortical region
proliferates, SF-1 expression became faint and uniform throughout (Fig. 2C). In some ovaries there appeared to be a slightly greater
concentration of message in the cortical than in the medullary region.
Results from sense probe hybridizations and other validation controls
were negative. There was no apparent difference in SF-1 expression in
adrenal glands at male- compared with female-producing temperatures or
as a result of the treatments described below.
Gonadal SF-1 expression at an intermediate incubation
temperature
In red-eared slider turtles, incubation between 28.6–29.6 C
results in varying ratios of male to female hatchlings with no
intersexes. As temperature increases within this range, the male/female
sex ratio decreases. We incubated eggs at 29.4 C, a temperature that
produces an approximately 20:80 female-biased hatchling ratio (Table 1)
(9), and examined SF-1 expression at stages spanning the
TSP. The resulting curve lies intermediate to the all male (26 C)- and
all female (31 C)-producing temperature curves (Fig. 3A). We then sexed, by histological
analysis, all stage 21 and 23 individuals from the 29.4 C incubation
temperature and found three groups of individuals: those with
developing testes or ovaries, and a third group that still appeared
histologically bipotential. Of the sex-identified individuals,
approximately 80% were female, and 20% were male, as hatching data
would predict. High or low levels of SF-1 message corresponded to
identified males or females, respectively (Fig. 3B). This pattern along
with the corresponding distribution of message agree with the profiles
of 26 C males vs. 31 C females described above. Intermediate
levels of SF-1 message were associated with the histologically
bipotential individuals at both stages 21 and 23. At the younger stages
assayed, gonadal sex could not be distinguished histologically in most
individuals (there is no established marker for gonadal sex in
red-eared slider turtles), but effects of both sexes may be present in
the SF-1 expression data of these stages as well. The intermediate
position of the 29.4 C expression curve is, therefore, the result of
pooled male and female effects and does not represent an intermediate,
graded response by all individuals to the intermediate temperature.
|
) (30). After such a
shift, we found that gonadal SF-1 expression correspondingly changed
from a male to a female pattern (Fig. 4A). One stage after treatment, gonadal
SF-1 expression was already significantly lower in the shifted embryos
than in 26 C male controls. There was no significant difference between
shifted embryos and 31 C female controls at any stage assayed.
Distribution of SF-1 mRNA within gonads also followed the female
pattern (Fig. 4B compared with Fig. 2C).
|
), as commitment to female gonadal sex is already
seen in approximately 20% of individuals by this stage
(30). With this shift, average gonadal SF-1 expression was
significantly higher in shifted embryos than in 31 C female controls by
one stage after treatment (Fig. 4A). Shifted values initially surpassed
those of 26 C male controls as well, but by stages 21 and 23 they had
declined and were statistically the same as the male controls. All
stage 21 and 23 individuals assayed were sexed histologically, and an
approximately 80:20 male-biased ratio was found, as predicted by
hatching data. SF-1 expression was high in identified males and low in
identified females (data not shown).
Estrogen treatment down-regulates SF-1 expression to the
female pattern
When red-eared slider turtle eggs incubating continuously at an
all male-producing temperature are treated with
E2 during the TSP, the sex-determining effect of
temperature is overridden, and 100% female hatchlings result (Table 1)
(9). To determine whether this treatment alters SF-1
expression as well, we spotted eggs incubating at 26 C with
E2 at stage 17 and assayed four subsequent
developmental stages for SF-1 message. One stage after treatment and in
all other stages examined, the level of gonadal SF-1 expression was
significantly lower in E2-treated embryos
compared with 26 C male controls (Fig. 5A). The effect of treatment between
E2-treated embryos and 31 C female controls was
not significant, and distribution of message was equivalent in these
two groups (Fig. 5B compared with Fig. 2C). No significant effect of
vehicle alone was found on SF-1 expression or hatchling sex ratios
(data not shown).
|
)
(9). To determine whether this treatment also alters SF-1
expression, at stage 17 we spotted eggs incubating at 29.4 C with a
nonsteroidal AI at the minimum dosage required to produce all males at
this temperature. We found that inhibition of estrogen biosynthesis
directly or indirectly caused an up-regulation of SF-1 expression (Fig. 5A) and the male-like pattern of distribution (Fig. 5C compared with
Fig. 2B). A significant increase over 29.4 C controls occurred, but not
until three stages after treatment. At stages 19, 21, and 23, SF-1
gonadal expression in AI-treated males was statistically the same as
that in 26 C males and showed no similarity to that in 31 C females. No
significant effect of carrier alone was found on SF-1 expression or
hatchling sex ratios (data not shown). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Gonadal SF-1 expression was first examined in red-eared slider turtle embryos that were continuously incubated at an all male-producing (26 C), an all female-producing (31 C), or an intermediate (29.4 C) temperature. SF-1 message was significantly higher in 26 C males than in 31 C females. This difference was observed from the earliest stage examined, stage 15, which is during the indifferent phase of gonadal development and marks the approximate beginning of the TSP (30). Morphological distinction of testes from ovaries cannot be detected in red-eared slider turtles until three or four developmental stages later (approximately 10.5–14 days at 26 C and 3–4 days at 31 C). In mammals, SF-1 is expressed without sex bias in males and females during the indifferent phase and may participate in the retention of primary steroidogenic tissue during this time (18). Although a similar function could be served by SF-1 in red-eared slider turtles, it is not clear why its expression appears sex biased at this time.
After the indifferent phase, SF-1 expression is higher in differentiating testes than ovaries in both mammals (18, 19) and red-eared slider turtles, which may indicate homologous SF-1 function. In developing mammalian testes, SF-1 functions include regulation of Mullerian inhibiting substance (MIS), which causes regression of the Mullerian ducts. This gene regulation is quantitative; increased levels of SF-1 cause increased MIS expression (32), although the high physiological level of MIS appears unnecessary to effect known functions in mammals. Interestingly, high levels of MIS, which can inhibit expression of aromatase (33), may be critical in developing TSD males, where exogenous estrogen causes female development at male-producing incubation temperatures (9). MIS has been cloned in red-eared slider turtles and is expressed developmentally in the adrenal-kidney-gonad complex of putative males at a time comparable to its expression in mammals (14); SF-1 message is predominantly localized in the sex cords, the site of MIS expression in vertebrates.
Transcriptional targets of SF-1 during differentiation of ovaries have not been identified in mammals. In at least some adult tissues, SF-1 regulates aromatase expression (18, 19), and there is evidence that a lower level of SF-1 may be sufficient to regulate aromatase compared with other P450 steroid hydroxylases (34). In TSD turtles, gonadal aromatase activity has been measured during the TSP (2), when it is thought to catalyze the synthesis of estrogens involved in female sex development. SF-1 expression in red-eared slider turtles could regulate aromatase expression during ovarian differentiation.
In red-eared slider turtle embryos that were continuously incubated at the intermediate temperature of 29.4 C, which typically produces about 20% male and 80% female hatchlings (9), gonadal SF-1 expression in most individuals became statistically comparable over developmental time to either 26 C males or 31 C females. Data from this temperature group indicated a close temporal association between differential SF-1 expression and gonadal dimorphism, suggesting a threshold effect of SF-1. An XY human has recently been identified with streak-like gonads containing immature tubules, with consequent male to female secondary sex reversal, attributed to decreased levels of functional SF-1 protein (35). Although the formation of sex tubules in mammalian testes is initiated by SRY (36), continued maturation may require a male threshold level of SF-1. In red-eared slider turtles, immature sex tubules are present during the middle of the TSP at both male- and female-producing incubation temperatures (30). A male threshold level of SF-1 in testes may be important in subsequent sex tubule maturation in this species as well.
Having established a baseline for SF-1 expression in red-eared slider turtles during incubation at continuous temperatures, we then examined whether altering the course of gonadal sex outcome during development altered SF-1 expression as well. Sex outcome was first manipulated by changing the incubation temperature in the middle of the TSP. When embryos are shifted from an all male (26 C)- to an all female (31 C)-producing incubation temperature, they become female (30). Here, such a shift resulted in down-regulation of gonadal SF-1 expression to a female level, and the pattern of change suggested either a drop to basal expression or active repression of SF-l expression. On the other hand, when embryos are treated by a female to male (31 to 26 C) temperature shift, approximately 80% become male (30). The remaining 20% appear already committed to develop as females despite the change in temperature cue. After this treatment, the mean level of SF-1 expression increased rapidly, initially surpassing even the 26 C male controls before falling back to that level, suggesting at least a two-phase regulation. Within this group, however, those embryos that could be histologically identified as female had equivalent SF-1 expression as 31 C female controls. Prior commitment to ovarian development appeared to disallow modulation of SF-1 expression regardless of the change in incubation temperature. Taken together, data from temperature treatments indicate that SF-1 expression is directly or indirectly modulated, at least before commitment, by the temperature trigger, that it lies in the TSD molecular path, and that its sex-specific levels are tightly controlled.
Sex outcome was next manipulated, and resulting SF-1 expression assayed, by altering the estrogen content of red-eared slider turtle eggs. When embryos incubating at an all male-producing temperature are treated with E2, they develop as females (9). Here, after such a treatment, gonadal SF-1 expression was down-regulated, becoming statistically and histologically indistinguishable from that in 31 C female controls. On the other hand, when embryos incubating at a female-biased temperature are treated with AI to block biosynthesis of estrogen, they develop as males (9). After this treatment, SF-1 message increased to that of 26 C male controls, but this effect was delayed compared with temperature shift and E2 treatments. Although AI inhibits new synthesis of estrogen, preexisting endogenous estrogen presumably remains active for some time and may account for this delay. Our experimental design does not allow conclusions about how direct the effect of estrogen is on SF-1 expression, as the first posttreatment measurement was taken one developmental stage (3 days) after application of E2. Two lines of evidence suggest, however, that changes in SF-1 regulation were not simply an aftereffect of ovary formation. First, treatments were performed just as sex-specific gonadal development begins. Gonadal sex was still not clear in most individuals assayed one stage after treatment, but SF-1 expression had already changed significantly. Secondly, as noted above, a highly significant difference in SF-1 expression was found between 26 C males and 31 C females as early as three or four developmental stages before morphological or histological distinction of gonadal sex. Interestingly, treating pregnant rats with the synthetic estrogen diethylstilbestrol causes decreased levels of SF-1 in genotypic male embryos (37), although complete sex reversal has not been reported with this treatment in mammals.
Although direct measures of endogenous estrogen in TSD organisms have been conflicting or inconclusive, a large body of indirect evidence indicates that a critical estrogen effect lies downstream from the temperature cue in primary female TSD (1, 2, 13, 14). Our data provide support at a molecular level for this model. Estrogen and AI treatments that override temperature cues modulate SF-1 expression. A female-producing temperature and exogenous E2 both result in low levels of SF-1 message; a male-producing temperature and decreased endogenous estrogen (AI treatment) both result in high levels of SF-1 expression. Finally, the level of SF-1 mRNA in developing red-eared slider turtle adrenal glands does not appear to be sex specific (23) and is unaffected by any of these treatments (Fleming, A., and D. Crews, unpublished data). Estrogen receptor is expressed in developing gonads, but not in adrenal tissue (15).
Taken together, the findings of this study suggest that at female-producing temperatures endogenous estrogen modulates the down-regulation of gonadal SF-1 early in the TSP. Yet a female-specific increase in gonadal aromatase activity has not been detected until midway through (2) or after (38) the TSP in any TSD species, indicating that estrogen biosynthesis may initially be extragonadal. Evidence for early estrogen synthesis has been found in the brain of the olive Ridley sea turtle and the diamondback terrapin (39, 40). Alternatively, the initial source of steroid hormones may be maternal and found in the yolk (41). Finally, it is possible that the differential endogenous expression of gonadal SF-1 is initiated by ligand-independent estrogen receptor or by other factors altogether and is maintained by an estrogen-involved mechanism of feedback on SF-1 in the gonad.
Curiously, in alligators (which are also TSD reptiles) and chickens, the pattern of SF-1 gonadal expression is reversed during differentiation: expression is higher in females than in males (21, 22). The effect, if any, of exogenous E2 on SF-1 expression has not been described in either group. E2 treatment of genotypic male chicken embryos results in up-regulation of aromatase and down-regulation of MIS, whereas AI treatment of genotypic females has the opposite affect (42). We would predict similar results in red-eared slider turtles, with SF-1 at the center of this regulation, but it is unclear how reversed endogenous levels of SF-1 in red-eared slider turtles compared with alligators and chickens might result in similar regulation of both aromatase and MIS. Although considerable conservation of genes and their function in gonadal sex development exists among amniotes, a surprising number of evolutionary changes in regulation of these genes, as seen here in the regulation of SF-1, is becoming apparent.
The expression data presented here indicate that the functions of SF-1 during gonadogenesis may be largely conserved between red-eared slider turtles and mammals. Moreover, regulation of SF-1 in red-eared slider turtles is sex-specifically modulated by both the incubation temperature and the estrogen content of the developing egg. These data provide support at a molecular level for a critical endogenous effect of estrogen after a female temperature cue and indicate that SF-1, in turn, is one of the genes directly or indirectly affected by estrogen in female TSD. Finally, the relative amount of SF-1 mRNA expressed, high in developing testes compared with ovaries, appears pivotal within TSD and may indicate a threshold mechanism involving SF-1.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received July 24, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylase/C17–20 lyase, and P450
aromatase in fetal mouse gonads. Endocrinology 135:262–268[Abstract]
and ß. Science 286:2328–2331This article has been cited by other articles:
 |
|
 |
|
  S Kuntz, D Chardard, C-I Ko, H Dumond, M Ducatez, M Callier, S Flament, and A Chesnel Female-enriched and thermosensitive expression of steroidogenic factor-1 during gonadal differentiation in Pleurodeles waltl J. Mol. Endocrinol., February 1, 2006; 36(1): 175 - 186. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. von Hofsten, J. Karlsson, I. Jones, and P.-E. Olsson Expression and Regulation of Fushi Tarazu Factor-1 and Steroidogenic Genes During Reproduction in Arctic Char (Salvelinus alpinus) Biol Reprod, October 1, 2002; 67(4): 1297 - 1304. [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 |
, testes; 
, ovaries; UD,
histologically bipotential.
