Differentiation (1995) 58:281-290
Differentiation
0 Springer-Verlag 1995
Aromatase enzyme activity during gonadal sex differentiation
in alligator embryos
Craig A. Smith', Pam K. Elfz, Jeffrey W. Langz, Jean M. P. Jossl
I School of Biological Sciences, Macquarie University, N.S.W.2109, Australia
2
Department of Biology, the University of North Dakota, Grand Forks, N.D. 58202, USA
Accepted in revised form: 21 December 1994
Abstract. Sexual differentiation of the gonads in Alligator mississippiensis and many other oviparous reptiles is
controlled by egg incubation temperature. Estrogens are
thought to play a role in this process, and it has been hypothesized that estrogen production is thermosensitive in
species with temperature-dependent sex determination
(TSD). Using the tritiated water assay, we measured the
activity of the critical estrogen-synthesising enzyme,
P450 aromatase, in the gonad-adrenal-mesonephric kidney complex (GAM) of alligator embryos incubated at
male- and female-producing temperatures. Aromatase
activity increased in the GAM of developing embryos
incubated at 30" C (100% female-producing) and 34.5"
C (predominantly female-producing), while it remained
very low throughout development in embryos incubated
at the intermediate temperature of 33" C (100% maleproducing). However, it is unclear whether enhanced aromatase activity represents the initial signal for ovary
differentiation or whether it lies downstream in the female developmental pathway. For embryos incubated at
30" C (female-producing), there was no detectable increase in aromatase activity until developmental stage
24, which is after the temperature-sensitive period for
sex determination. This suggests that aromatase may be
a downstream component of the ovary-determining cascade. In female alligator hatchlings, most of the aromatase activity was localised in the ovary, activity being
low in the adrenal-mesonephros. Aromatase assays
carried out at 30" C and at 33" C indicated that, at viable
incubation temperatures, aromatase activity is not thermosensitive. This suggests that temperature directly or indirectly influences enzyme synthesis. Radioimmunoassay
of estradiol synthesised by the GAM during the aromatase assay confirmed increased enzyme activity during
female development but not during male development.
Increased aromatase activity and estrogen synthesis during female development were correlated with the timing
of ovary differentiation, particularly proliferation of the
gonadal cortex. These findings implicate aromatase in
Correspondence to: J.M.P. Joss
temperature-dependent gonadal sex differentiation in alligator embryos, higher enzyme activity being associated
with ovary development.
Introduction
Despite some major advances in recent years, the mechanisms underlying gonadal sex differentiation in vertebrates are still largely unknown. A role for steroid hormones in gonadogenesis was first suggested over 50
years ago [32, 57, 591. Many studies have subsequently
examined the effects of androgens and estrogens on the
development of the gonads [2,4]. The so-called hormonal theory of gonadal differentiation, however, has not
been completely supported. While the treatment of larval
fishes and amphibians with androgens or estrogens can
induce testis or ovary differentiation, respectively, [2, 58,
611, it is unclear whether sex steroids play a natural role
in gonadal morphogenesis in these vertebrates. Among
amniotes, a role for androgens in gonadal differentiation
has not been convincingly shown, whereas there is considerable evidence implicating estrogens in gonadogenesis, at least in birds and reptiles. In bird embryos, exogenous estrogens can transiently feminise the gonads of
genetic males, while the antiestrogen, tamoxifen, can
masculinise the gonads of females (see [40]). The gonads are more steroidogenically active in female chicken
embryos than in males [21], and gonadal estrogen levels
are higher in developing ovaries than in testes [40,60].
In many species of reptiles, sex is determined environmentally by egg incubation temperature ( [ 5 , 9, 18,
481, reviewed in [16, 24, 381). In the American alligator,
Alligator mississippiensis, high and low incubation temperatures produce mostly or exclusively females, while
intermediate temperatures produce males [30]. Estrogens
applied to eggs incubated at male-producing temperatures can induce ovary differentiation and female development in alligators and other reptiles with temperaturedependent sex determination (TSD) [6, 8, 14, 22, 35,
282
361; reviewed in [7] and [53]. Antiestrogens such as tamoxifen can masculinise the gonads (induce seminiferous
cord differentiation) at female-producing temperatures in
some species [ 141 (but tamoxifen may act as an estrogen
agonist in alligator [27]). Estrogens can also feminise
the embryonic gonads or induce complete ovarian differentiation among those reptiles in which sex is determined genotypically [6,38].
In species with TSD, the sex-determining mechanism
is sensitive to temperature during a critical period of embryogenesis. This temperature-sensitive period (TSP)
has been identified for each species by shifting eggs
from the male- to female-producing temperature, and
vice versa, throughout incubation and examining the sex
ratios of resulting hatchlings. The TSPs appear to coincide with the onset of gonadal sex differentiation [36,
43, 5 I]. Furthermore, sensitivity of gonadal differentiation to exogenous estrogens occurs prior to and/or
during the TSPs [6, 8, 23, 521. Estrogen is therefore implicated in the process of gonadal sex differentiation in
species with TSD.
Recent studies point to a key role for the estrogensynthesizing enzyme P450 aromatase in the ovary-determining pathway of birds and reptiles. Treatment of genetically female chicken embryos with a specific aromatase
inhibitor can induce sex reversal [15, 501. In reptiles with
TSD, aromatase inhibitors applied to eggs incubated at
the female-producing temperature can disrupt ovary differentiation [28] or induce male development at temperatures that otherwise yield both sexes [39, 551. It has been
hypothesised that aromatase enzyme synthesis is thermosensitive in species with TSD, higher levels being produced at female-determining temperatures, leading to enhanced estrogen production and ovary differentiation ([4,
lo], reviewed in [36]). The present study examines this
hypothesis by investigating aromatase enzyme activity in
the urogenital tissues of Alligator mississippiensis embryos during development at male- and female-producing
temperatures. Changes in aromatase activity are related
to the timing of gonadal sex differentiation.
Methods
Incubation of eggs and staging of embryos. Seven clutches comprising over 200 fertile eggs of Alligator mississippiensis were obtained soon after oviposition from wild nests located on the
Rockefeller Wildlife Refuge, southwestern Louisiana. The eggs
were transported by road to the University of North Dakota,
Grand Forks, where they were divided into three groups and incubated in foam box incubators [31] at either 30" C, 33" C or 34.5
"C. Incubation at 30" C and 33" C results in 100% female development and 100% male development, respectively [18, 26, 27,
301. Initial studies indicate that incubation at 34.5" C yields over
90% females ("high temperature females"), although embryonic
mortality increases at this temperature [30]. Embryos incubated at
the three temperatures were examined before (developmental
stage 19). during (stage 21) and after (stages 23, 24, 25 and 28)
the temperature-sensitive period (TSP) for sex determination [30].
All ambryos were weighed and then staged according to Ferguson
[ 171, prior to excision of the paired gonad-adrenal-mesonephros
complexes (GAMs). Aromatase assays were carried out on the entire GAM because of the difficulty in separating the three components until late in embryogenesis.
Ammatase assay. The tritiated water assay [ I , 471 was used to
measure aromatase activity in whole GAMs incubated in vitro, as
described previously for crocodile embryos [42]. Paired GAMs
dissected from individual embryos were incubated at least in triplicate in 400 pl HEPES-buffered RPMI- I640 culture medium, pH
7.2, supplemented with 3 mg/ml glutamine (Sigma, St. Louis,
USA) and 0.8 phf tritiated steroid substrate ([I P-3H]androstenedione, 24.1 Ci/mmol; New England Nuclear). For some stage-21 embryos, in which enzyme levels proved to be very low, pools of six
pairs of CAMS per replicate were also assayed. Tissues were incubated for 6 h in a shaking water bath set at the corresponding egg
incubation temperature. (For some developmental stages, the direct
effect of temperature on aromatase activity was examined by carrying out assays at different temperatures, as described in Results).
At the end of the incubation period, unreacted tritiated steroid was
removed by chloroform extraction and dextran-charcoal adsorption, as described previously [42]. Aromatase activity was calculated from the rate of tritium released from the lp-position during
the aromatisation of [ 1P-3H] androstenedione into estrogens. The
tritium, released as -'H,O, is recovered in the aqueous phase. Results are expressed as fmol/GAM/h or fmol/mg protein/h. The sensitivity of the assay was defined as twice the mean cpm of blank
tubes, -300 cpm, corresponding to an aromatase activity of 1.5
fmol/GAM/h. When assays were carried out in the presence of 100
pM 4-OH-androstenedione, a competitive steroidal inhibitor of aromatase, release of 3H was virtually abolished, demonstrating the
specificity of the assay for the aromatase enzyme.
Following the aromatase assay, one GAM from each pair was
processed for histology, as described below. The contralateral
GAM was snap-frozen in liquid nitrogen and stored at -70" C prior to assaying for total soluble protein. The bicinchoninic acid
(BCA) method (Pierce [45]) was used to measure protein content
following homogenisation of GAMs in phosphate-buffered saline
(PBS). Bovine serum albumin (BSA) was used as the standard.
Radioimmunoassay of estradiol produced during aromatase assays. To confirm the production of estrogens during the aromatase
assay, the chloroform and dextran-charcoal used to remove unreacted tritiated androstenedione were retained, extracted and assayed for any synthesised estradiol for stage-23, -25 and -28 embryos. (For technical reasons, samples from earlier stages were
not included). Firstly, the chloroform phases from triplicate tubes
were pooled and set aside. Secondly, the charcoal pellets from
these triplicates were extensively vortexed in the presence of fresh
chloroform and centrifuged at 2000 g. The chloroform layers were
removed, combined and then added to the pooled chloroform set
aside (above). The total volume of chloroform, containing pooled
steroids from triplicate tubes, was then dried down under nitrogen
and the steroids reconstituted in 1 ml 100%ethanol. Samples were
stored at 4" C prior to radioimmunoassay for estradiol- I7p. Samples were firstly dried under nitrogen and reconstituted in 550 pI
assay buffer (0.I % gelatin in PBS). The samples were allowed to
equilibrate overnight at 4" C, and duplicate 200-p1 aliquots were
then assayed for estradiol using modifications of a highly specific,
commercially available '251-estradiol-17P RIA kit (Diagnostic
Products Corp, Los Angeles, CA, USA) previously described
[46]. Separation of bound versus free steroid was accomplished by
the addition of 500 pI dextran-coated charcoal in assay buffer, incubation at room temperature for 10 min, followed by centrifugation at 3000 g, 4" C, for 15 min as described by Fivizzani et al.
[ 191. Supernatant aliquots of 500 pI were counted in a Beckman
5500 gamma counter. The sensitivity of this assay is 0.0625 pg/assay tube, and the interassay and intraassay coefficients of variation are 1 I .98% and 1 1.1 1 %, respectively. Results are expressed
as pg of estradiol produced/GAM during the 6-h aromatase assay
incubation period.
Gonadal histology. Unilateral CAMS used for measurement of aromatase activity were retained, fixed in glutaraldehyde, post-fixed
in osmium tetroxide and processed for high-resolution light microscopy, as described previously [41]. Semi-thin sections ( I pm)
283
were stained with 1% toluidine blue. In addition, some ultra-thin
sections of gonads from high-temperature female hatchlings
(34.5" C ) were stained with uranyl acetate and lead citrate and examined at the ultrastructural level.
Results
Aromatase activity
Aromatase enzyme activity was detectable in the GAM
throughout embryogenesis at both male- and female-producing temperatures. Detectable enzyme activity in the
GAM of female embryos incubated at either 30" C or
34.5" C increased from stage 24 to hatching, whereas aromatase activity remained very low throughout development in the GAM of male embryos incubated at33" C
(Fig. I). The increase in detectable aromatase activity in
presumptive female GAMs occurred after the temperature-sensitive period (TSP) for female sex determination
at 30°, as defined by Lang and Andrews [30] and shown
in Fig. 1. In younger embryos sampled earlier than stage
24, aromatase activity in the GAM was very low at 30",
-f 1
I
500
5
e-
33" and 34.5" C (approximtely 8 fmol/GAM/h) and was
not significantly different between the three temperatures (ANOVA, fi0.05 for stages 19, 21, and 23). E~~~
when six rather than two pairs of GAMs were pooled for
assay of aromatase, the overall activity was still very low
and there were no significant temperature-dependent differences in the levels of detectable aromatase activity
until after the TSP (i.e. after stage 23).
Table 1 shows the mean soluble Drotein content of the
GAM and aromatase activity/mg 'protein in the GAM
over development at 30" C, 33" C and 34.5" C. For each
stage examined, the protein content of the GAM was
similar at the three temperatures. Expressed per mg protein, aromatase activity actually decreased over stages 19
to 23 at each temperature. However, it increased after
stage 23 in presumptive females, but not in presumptive
males, consistent with the pattern of aromatase acivity/whole GAM (Fig. 1). Since aromatase activity in
the whole GAM was low but stable over stages 19 to 23,
decreasing activity/mg GAM protein was probably due
to the rapid growth of the mesonephric kidneys relative
to the gonads during the first half of incubation.
At stage 25, aromatase activity in the GAM of "hightemperture females" (34.5"C) was lower than that in
the GAM of "low-temperature females" (30" C), but by
stage 28 (hatching) levels were similar in both groups
(Fig. 1). (Histological analysis showed that the stage-
400-
:
e
Y
>
k
300
-
P
5
a
200-
W
v)
U
+
U
3 1000
K
a
-
I
0
I
18
20
.
I
.
22
I
.
24
1
26
.
1
20
STAGE
Fig. 1. Aromatase activity in the gonad-adrenal-mesonephros
complex (GAM) of alligator embryos during development at 30"
C (0)
33"
.C (H) and 34.5" C (0,females; 0males). The temperature-sensitive period for female sex determination at 30" C is
shown (TSP-30" C).Values are means*SEM
Table 1. Mean soluble protein
content (mg) and aromatase activity (fmolhg proteinh) in the
gonad-adrenal-mesonephros
(GAM) or alligator mississippiensis embryos over stages of development at 30" C, 33" C and
34.5" c
Stage
l9
21
23
25
28
FEMALES
MALES
Fig. 2. Aromatase activity in the gonad (H) and adrenal-mesonephros (0)of female and male Alligator mississippiensis hatchlings incubated at 34.5" C
Mean total protein (mg)
Mean aromatase activity
(fmolhg proteinh)
30" C
33" c
34.5" c
30" C
33" c
34.5" c
0.56
0.79
0.74
1.31
1.19
0.59
0.70
0.97
1.11
0.87
0.58
0.68
1.04
1.02
1.03
13.21
8.10
7.45
209.23
354.03
16.91
11.71
6.35
7.45
10.45
17.42
11.61
7.96
53.8
422.62
.
A
T T
I
/
0
2
33°C
30°C
T
100
4
Time
6
8
10
(hrs)
Fig. 3. Time course of aromatase activity in hatchling female alligator GAMs, showing increased aromatisation of steroid substrate
over time
2Y34.5" C embryos used for aromatase assay were indeed all females). Two of six embryos examined at
stage 2W34.5" C were males; aromatase activity in the
GAMs of these males was low and comparable to that
of males incubated at 33" C (Fig. 1). When the gonad
could be separated from the adrenal-mesonephros in
some female hatchlings incubated at 34.5" C, most aromatase activity was found to be in the gonad (ovary),
while the very low levels in males were similar in both
the gonad (testis) and the adrenal-mesonephros (Fig. 2).
Higher levels of aromatase activity appeared to be present in the adrenal-mesonephros of female hatchlings
than in males (Fig. 2), although this may have been due
to contamination of the female organs with ovarian tissue during dissection.
Time-course studies on aromatase activity were carried out using GAMs from female hatchlings that had
been incubated at 30" C. Aromatisation of tritiated steroid substrate by these GAMs was linear up to 9 h under
the conditions used here (Fig. 3), validating the 6-h incubation time chosen for routine assays.
To examine the direct effect of assay temperature on
aromatase activity, GAMs from stage-21 and stage-25
embryos were incubated .with steroid substrate at 30",
33" or 36" C. For stage-21 embryos, incubated at either
30" C or 33" C, aromatase activity in the GAM was very
low and not significantly different when assayed at 30"
C and 33" C (ANOVA; -0.05; Fig. 4A). However, for
embryos from both incubation temperatures, aromatase
activity was significantly reduced when assays were carried out at 36" C. Similar results were obtained when the
direct effect of temperature was examined in older embryos, in which clear sex differences in aromatase activity were apparent (stage 25). Aromatase activity in the
CAM was high in embryos incubated at 30" C (females)
and low in embryos incubated at 33" C (males) when assayed at either 30" C or 33" C (Fig. 4B). Aromatase ac-
30°C
33°C
EMBRYO INCUBATION TEMPERATURE
Fig. 4. Aromatase activity in the GAM of stage-21 ( A ) and stage25 (B)A. mississippiensis embryos incubated at 30" or 33" C and
assayed at 30" C (0).33" C (W) or 36" C
a)
23
25
28
STAGE
Fig. 5. Production of estradiol-17P by the CAM during the 6-h aromatase assay at stages 23, 25 and 28 for embryos incubated at
30" C (O), 33" C (W) and 34.5" C
Synthesised estradiol was
recovered from the charcoal and chloroform phases following the
aromatase assay, as described in the text
a).
tivity was again inhibited in both sexes when assays
were performed at 36" C (Fig. 4B).
Radioimmunoassays (RIAs) carried out on extracted
media following the 6-h aromatase assays confirmed
higher aromatase activity in the GAMs of female embry-
Fig. 6 a-f. Development of embryonic alligator gonads at 33" C
(a-c) and 30" C (d-f). All micrographs are of 1-pm longitudinal
resin sections. a Stage-21/33" gonad showing a well-defined cortex (C) and underlying medulla (M); insert, higher magnification
of the medulla, showing the first signs of the onset of testis differentiation. Small numbers of enlarged pre-Sertoli cells are present
in swollen medullary cords (arrows).b Stage-23/33" gonad showing pre-Sertoli cells differentiating into Sertoli cells as seminiferous cords become organised in the medulla (arrows).The cortex
(C)is still well developed. c Stage-25/33" gonad (testis) showing
well-developed seminiferous cords (e.g., S) and reduced cortex
(arrows).d Stage-21/30" gonad showing medulla (M) and cortex
(C). The gonad adjoins the mesonephric kidney (Ms);insert, highpower view of the medulla. Most cordal cells are small and irregular in shape (e.g., arrows),compare with a, insert. e Stage-23/30"
gonad showing germ cell proliferation in the cortex (arrows).The
medulla (M) shows signs of regression, marked by the fragmentation of medullary cords and the presence of many dark interstitial
cells. f Stage-25/30" gonad (ovary) showing a thickened cortex
( C ) and regressing medulla (M); bars in a, c, d and f, 50 pm, bars
in b and e, 20 pm, bars in inserts, 10 pm; toluidine blue staining
286
Fig. 7a-e. Gonad from stage-28 (hatchling) alligator incubated at
34.5" C. a Transverse section through the anterior pole of the gonad, showing ovarian characteristics. The cortex (C) is well developed, although with few germ cells, and the medulla (M) is fragmented. b Higher magnification of the medulla shown in a. The
medullary cords (Mc) are unorganised and separated by extensive
lacunae (L).c Transverse section through the posterior pole of the
same gonad, just anterior to the medullary rest. The cortex ( C ) is
still well developed, but the medulla (M) is not fragmented. d
Higher magnification of the medulla shown in c. The medullary
cords (Mc) are organised into structures resembling seminiferous
cords, with large polarised cells. The cortex (C)is distinct. e Electron micrograph of one enlarged medullary cord cell seen in d.
The polarised cell has ultrastructural characteristics of a Sertoli
cell, including a large ovoid nucleus ( N ) with its long axis perpendicular to the basement membrane (BM), prominent nucleolus (urTOW) and apical cytoplasm (Cy). Extensive cytoplasmic processes
(Cp), typical of seminiferous cords, are also present; burs in a and
c, 50 pm; burs in b and d, 20 pm; bur in e, 5 pm
287
0s than in males. For embryos incubated at both 30" C
(low temperature females) and 34.5" C (high temperature females) estradiol synthesis by the GAM in the
presence of tritiated androstenedione increased over
development at the stages tested; from 0.5-1 pg/GAM
at stage 23-22 to 25 pg/GAM at stage 28 (Fig. 5). In
contrast, very low levels of estradiol were measured after aromatase assay of male GAMs over stages 23-28
(Fig. 5).
Gonadal histology
At both 33" and 30" C, embryonic gonads were morphologically undifferentiated at stage 19, when aromatase
activity in the GAM was detectable but extremely low
and similar at both temperatures. By stage 21 at 33"
(male-producing), the gonad had a well developed cortex
and an underlying medulla comprising loosely organised
medullary cords together with darkly staining interstitial
cells (Fig. 6a). High magnification of the gonadal medulla showed the earliest signs of the onset of testis differentiation in these embryos, characterised by the appearance of small numbers of enlarged pre-Sertoli cells
within medullary cords (Fig. 6a, insert). By stage 23 at
33", enlarged pre-Sertoli cells were widespread and they
were differentiating into Sertoli cells as seminiferous
cords became organised in the medulla of the gonad
(Fig. 6b). At stage 25 at 33", differentiated testes were
present, with well-developed seminiferous cords and a
reduced cortex (Fig. 6c).
The gonads of stage-21 embryos incubated at 30"
(female-producing) had a well-defined cortex and medulla, as at 33" C (Fig. 6d). However, gonads at 30" did
not show the early signs of organisation in the medulla
as seen in presumptive male embryos; most cells in the
medullary cords were smaller, with relatively less cytoplasm and smaller nuclei (Fig. 6d, insert). By stage 23
at 30", when aromatase activity was still low and a low
level of in vitro estradiol synthesis was detected by
RIA, the gonad showed signs of ovarian differentiation.
Large pale germ cells and darker somatic cells were
proliferating in the cortex, while the medullary cords
were becoming more loosely organised (fragmented)
than at earlier stages (Fig. 6e). By stage 25 at 30°, when
aromatase activity in the GAM was high and synthesis
of estradiol in vitro was elevated, all gonads examined
were ovaries, characterised by a thickened cortex with
meiotic germ cells and an extensively fragmented medulla (Fig. 60.
Gonadal development in female embryos incubated at
34.5" C appeared similar to that. of low temperature females at stages 19, 21, 23 and 25, with cortical proliferation and regression of the medulla also evident at stage
23. However, close examination of the ovaries of three
female hatchlings (stage 28) incubated at 34.5" C revealed some apparent seminiferous cord development towards the caudal pole of the gonad. Anteriorly, the gonads had a typical ovarian structure. The cortex was well
developed (although with few germ cells in those sections examined), and the medulla was fragmented (Fig.
7a). In the medulla, epithelial cords were loosely arranged, and lacunae were present within cords (Fig. 7b).
At the posterior pole, just anterior to the so-called medullary rest (where regression of the medulla and proliferation of the cortex do not occur), the cortex was well developed, while the medullary cords were not fragmented.
Instead, the cords were organised into rounded structures
(Fig. 7c) that strongly resembled seminiferous cords
(Fig. 7d). Indeed, the cells within these cords showed ultrastructural characteristics of Sertoli cells. The typical
cell was elongated and polarised, with a basal nucleus
and apical cytoplasm. The long axis of the nucleus was
oriented perpendicular to the basement membrane, and
the nucleolus was prominent (Fig. 7e).
Discussion
The results of this study indicate that in alligator embryos aromatase activity in the GAM increases during development only at female-producing temperatures. This
increased aromatase activity accompanies mrophological
differentiation of the ovary. The level of aromatase activity found in this study, and the pattern of increase in
female alligator GAMs, is similar to that previously
reported for female crocodile (Crocodylus porosus) embryos [42].
It has been hypothesised that aromatase synthesis or
activity may be thermosensitive in embryonic reptiles
with TSD. This would provide a mechanism whereby
temperature might mediate estrogen synthesis, which in
turn influences gonadal sex differentiation ([ 10, 1 I , 121,
reviewed in [36]). In this study, temperature did not directly influence aromatase activity per se when assays
were carried out at 30" and 33" C (Fig. 4). For aromatase to be thermosensitive, temperature must directly or
indirectly regulate aromatase synthesis (or breakdown)
in embryonic alligator GAMs. Temperature-regulation
of aromatase synthesis has been proposed for Emys
orbicularis (a turtle with TSD), where gonadal aromatase activity also increases during female development
but not during male development [ 10, 111. The inhibition of aromatase activity at high assay temperature
(36" C) in A. mississippiensis is consistent with previous results reported for C. porosus embryos, where high
assay temperature (35"-37" C) also inhibited aromatase
activity [42]. This inhibition may represent a direct effect of temperature on enzyme kinetics, or an indirect
effect upon the GAM or the culture medium used in the
assay. In any case, such direct temperature inhibition of
aromatase activity is not relevant to sex determination
under natural conditions; constant temperatures above
35" C are incompatible with survival in embryonic alligators [30].
Since increased aromatase activity was not detectable until after the TSP for sex determination at 30" C
(Fig. l), aromatase may be a downstream component of
the female-determining pathway in alligator embryos.
Incubation temperature may control regulatory genes
which in turn control the differentiation of somatic steroidogenic cells and hence aromatase production. In an-
288
other crocodilian, C. porosus, increased aromatase activity in the GAM of female embryos was also undetectable until after the TSP [42]. Furthermore, in both A.
mississippiensis and C. porosus, morphological differentiation of the ovary appears to begin towards the end
of the TSP [41,42]. Lance and Bogart [29] have detected estradiol in the plasma of alligator embryos throughout incubation at male- and female-producing temperatures, but no clear sex differences were apparent. Altogether, these observations suggest that unidentified
female-determining genes may be activated prior to
morphological differentiation of the ovary and increased
estrogen synthesis. Thus, aromatase may be indirectly
controlled by temperature. In contrast, gonadal aromatase activity does increase during the TSP in embryos of
the turtle, Emys orbiculuris, incubated at the femaleproducing temperature [ 101, providing evidence that
aromatase lies close to the beginning of the ovary-determining pathway.
However, as in previous studies [lo, 121, this study
did detect low levels of aromatase enzyme activity
throughout the TSP in both presumptive sexes. We cannot exclude the possibility that small, local differences
in enzyme levels may exist at male- and female-producing temperatures, differences that were not detected by
the whole organ assay technique used here. Indeed, a
comparison of Fig. 1 and 5 shows that aromatase activity and estradiol synthesis in vitro (in the same tissues)
were not always correlated. Although there were no detectable sex differences in aromatase activity at stage
23, for example (the end of the TSP) (Fig. l), small but
significant sex differences in estradiol synthesis in vitro
were detected (Fig. 5). Other studies have used microsoma1 preparations rather than whole tissues to measure
aromatase activity [25] and this might be a more sensitive technique. However, previous studies using the turtle E. orbiculuris have relied exclusively on the whole
organ method [ 10, 11, 121 and this seemed sufficient to
reveal (albeit very small) sex differences in aromatase
activity.
The most likely site of aromatase activity in the
GAM of female embryos is the gonad, as suggested by
the high levels of enzyme activity in isolated hatchling
ovaries compared to those in the adrenal-mesonephros
(Fig. 2). The cell lineage in the developing ovary responsible for enhanced aromatase activity is, however,
not known. It seems unlikely to be the germ cell lineage, since, in bird embryos, female gonads can undergo somatic differentiation and synthesise estrogens
when the germ cell population has been significantly
reduced [33]. The somatic (pce-follicular) cells of the
cortex, medullary cord cells or interstitial cells of the
medulla are alternative sources. Pre-follicular cells begin proliferating and the medullary cords show signs of
regression between stages 22 and 23 (i.e., towards the
end of the TSP) in alligator embryos incubated at 30" C
[41], which corresponds with the onset of increased detectable estradiol production in female GAMs observed
here (Fig. 5 ) . We have recently attempted immunohistochemical localisation of aromatase in embryonic alligator gonads but have been unable to detect the enzyme,
using antibodies directed against either the mammalian
or the bird protein.
This study is the first to describe aromatase enzyme
activity throughout female development at both high
and low egg incubation temperatures in a reptile with
TSD. Since aromatase activity increased in the GAM of
female embryos incubated at both 30" and 34.5" C, it
might be argued that aromatase production is only temperature-dependent by virtue of its inhibition at the
male-producing temperature (33" C). Enhanced aromatase activity in the GAM may occur constitutively during alligator embryogenesis, playing a role in ovary differentiation, unless blocked by a factor activated at a
critical (male-determining) temperature (or small temperature range). This factor would directly or indirectly
divert development towards the male pathway. For alligators and other crocodilians, the factor would be variably active at 32" to 33" C (male-producing for all species) [30]; on either side of this temperature window,
production of the factor may be suboptimal or absent,
allowing aromatase synthesis, ovary differentiation and
female development. (In this context, no reptilian species has yet been shown to exhibit a male-female-male
pattern of TSD).
Evidence from this and other studies indicate that
increasing aromatase activity (and hence estrogen synthesis) is correlated with proliferation of the cortex and
regression of the medullary cords during ovary development in species with TSD [ 11, 28, 561. Lance and Bogart [28] found that administration of a non-steroidal aromatase inhibitor, Ciba Geigy 16949A (fadrozole) disrupted ovarian differentiation in alligator embryos incubated at 30°C. The cortex did not show normal germ
cell proliferation and/or medullary cords did not regress.
However, the gonads did not show any signs of testis
differentiation. In genetically female chicken embryos,
administration of the same aromatase inhibitor has been
shown to induce testis differentiation [ 151.
Aromatase enzyme activity may be inhibited by antiMullerian hormone (AMH) in alligator embryos incubated at the male-producing temperature. Bovine AMH can
inhibit aromatase activity in the ovaries of embryonic
rats [49], and in the embryonic ovaries of a turtle with
TSD [ 131. However, evidence from mammals indicates
that AMH is located downstream in the sex-determining
cascade [34]. Furthermore, studies on Mullerian duct
regression in male alligator embryos suggest that the
hormone is active towards the end of incubation in male
embryos [3]. Therefore, although AMH may negatively
regulate aromatase in male alligator embryos later in
development, it may not be the only putative inhibitor of
the enzyme and other factors may be involved earlier in
embryogenesis.
The development of some apparent seminiferous
cords and Sertoli cells in the ovaries of the high-temperature females examined indicates that both male- and female-specific tissue can develop within the same gonad
in alligator embryos. This may be related to the fact that
34.5" is close to the exclusively male-producing temperature of 33" C (indeed, small numbers of males are produced at 34.5" C). Female-producing temperatures close
289
to the male-producing temperature may allow some development of seminiferous cords, particularly near the
posterior pole of the gonad, where medullary cord regression and cortex proliferation do not normally occur
in female alligator embryos (the so-called “medullary
rest” described by Forbes [20]). It would be interesting
to examine the gonads of female embryos produced at
other temperatures close to the male-producing temperature (e.g., 32” C, which yields mainly males but some
famales [30]).
The widespread observation that exogenous estrogens
can “override” TSD (and GSD; reviewed in [38,53,54])
may be explained by the fact that both presumptive sexes
have the capacity to respond to estrogens. Uptake of radiolabelled estradiol by embryonic crocodile gonads
[44] and estrogen receptor in situ hybridisation studies
in turtle embryos [7] support this idea. Indeed, the fact
that exogenous estrogens can induce ovary differentiation at male-producing temperatures implies that pivotal
events downstream of estrogen production in the female
sex-determining pathway are not thermosensitive (at female vs. male temperatures). The focus is therefore
shifting to the (endogenous) synthesis of estrogens as
the thermosensitive mechanism involved in TSD. However, although aromatase inhibitors can induce male development at certain intermediate incubation temperatures in species with TSD [%], this does not necessarily
prove that aromatase is the initial signal for female development. (An example from human sex determination
may be pertinent. Many human females apparently carry
a normal copy of the SRY gene, which signals testis differentiation. These individuals are presumably sex-reversed due to a mutation downstream in the testis-determining pathway [37]). For species with TSD, the exact
role of aromatase in the sex-determining process will be
clarified by expression analysis of the gene during embryogenesis [25].
In conclusion, the data presented here show that ovary development in alligator embryos, at both high and
low incubation temperatures, is accompanied by increased aromatase enzyme activity (and estradiol synthesis in vitro). Proliferation of the gonadal cortex and
regression of the medulla during ovary formation may
depend upon this enhanced estrogen production. However, it is unclear whether enhanced aromatase activity is
the thermosensitive trigger for ovarian differentiation or
whether it lies further downstream in the female developmental pathway. The lack of detectable sex differences in aromatase activity until after the TSP, that is,
until after sex has been determined, suggest that aromatase may be a downstream component of the female-determining cascade in alligators..
Acknowledgements. We thank Ted Joanen, Larry McNease, Dr
Ruth Elsey and staff at the Rockefeller Wildlife Refuge, Grand
Chenier, Louisiana, for providing alligator eggs, and Turk Rhen
for transporting the eggs and monitoring incubation. We also
thank John Allsteadt and Dr. Alan Conley for their assistance in
various ways. Dr Albert Fivizzani generously provided reagents
and equipment for carrying out the aromatase assays and RIAs at
UND. Financial support was provided by an Australian Research
Council Grant to J.M.P.J.
References
1. Ackerman GE, Smith ME, Mendelson CR, McDonald PC,
Simpson ER (1981) Aromatization of androstenedione by human adipose tissue stromal cells in monolayer culture. J Clin
Endocrinol Metab 53:412417
2. Adkins-Regan E (1981) Hormones and sexual differentiation.
In: Norris DO, Jones RE (eds) Hormones and Reproduction in
Fishes, Amphibians, and Reptiles. Plenum Press, New York,
p. 1-29
3. Austin H (1989) Mullerian-duct regression in the American
alligator (Alligator misissippiensis): its morphology and testicular induction. J Exp Zoo1 251:329-338
4. Bogart MH (1987) Sex determination: a hypothesis based on
steroid ratios. J Ther Biol 128:349-357
5. Bull JJ (1980) Sex determination in reptiles. Q Rev Biol 55:
3-2 1
6. Bull JJ, Gutzke WHN, Crews D (1988) Sex reversal by estradi01 in three reptilian orders. Gen Comp Endocrinol70:425428
7. Crews D, Bergeron JM, Bull JJ, Flores D, Tousignant A, Skipper JK, Wibbels T (1994) Temperature-dependent sex determination in reptiles: proximate mechanisms, ultimate outcomes, and practical applications. Dev Genet 15:297-3 12
8. Crews D, Wibbels T, Gutzke WHN (1989) Action of sex steroid hormones on temperature-induced sex determination in
the snapping turtle (Chelydra serpentina). Gen Comp Endocrinol76:159-166
9. Deeming DC, Ferguson MWJ (1988) Environmental regulation of sex determination in reptiles. Phil Trans R Soc Lond
[B] 322~19-39
10. Desvages G,Pieau C (1992) Aromatase activity in gonads of
turtle embryos as a function of the incubation temperature of
eggs. J Steroid Biochem Molec Biol41:851-853
1 1. Desvages G,Pieau C (1992) Time required for temperatureinduced changes in gonadal aromatase activity and related gonadal structure in turtle embryos. Differentiation 52: 13-1 8
12. Desvages G,Girondot M, Pieau C (1993) Sensitive stages for
the effects of temperature on gonadal aromatase activity in
embryos of the marine turtle Dermochelys coriacea. Gen
Comp Endocrinol92:54-61
13. Di Clemente N, Ghaffari S, Pepinsky RB, Pieau C, Josso N,
Cate RL, Vigier B (1992) A quantitative and interspecific test
for biological activity of anti-Miillerian hormone: the fetal
ovary aromatase assay. Development 1 14:721-727
14. Dorizzi M, Mignot T-M, Guichard A, Desvages G,Pieau C
(1991) Involvement of oestrogens in sexual differentiation of
gonads as a function of temperature in turtles. Differentiation
47 :9- 17
15. Elbrecht A, Smith RG (1992) Aromatase enzyme activity and
sex determination in chickens. Science 255:467470
16. Ewert MA, Nelson CE (1991) Sex determination in turtles: patterns and some possible adaptive values. Copeia 1991:50-69
17. Ferguson MWJ (1985) Reproductive biology and embryology
of crocodilians. In: Gans C, Billet F, Maderson PFA (eds) Biology of the Reptilia. Vol XIV, Development A. J Wiley, New
York, pp. 329491
18. Ferguson MWJ, Joanen T (1982) Temperature of egg incubation determines sex in Alligator mississippiensis. Nature 296:
850-853
19. Fivizzani AJ, Colwell MA, Oring LW (1986) Plasma steroid
hormone levels in free-living Wilsons’s phalaropes, Phularopes tricolor. Gen Comp Endocrinol 62: 137-144
20. Forbes TR (1940) Studies on the reproductive system of the
alligator. IV. Observations on the development of the gonad,
the adrenal cortex and the Mullerian duct. Contrib Embryo1
28~129-156
21. Guichard A, Cedard L, Mignot T-M, Scheib D, Haffen K
(1977) Radioimmunoassay of steroids produced by cultured
chick embryonic gonads: differences according to age, sex,
and side. Gen Comp Endocrinol32:255-265
290
22. Gutzke WHN, Bull JJ (1986) Steroids hormones reverse sex in
turtles. Gen Comp Endocrinol64:368-372
23. Gutzke WHN, Chymiy DB (1988) Sensitive periods during
embryogeny for hormonally induced sex determination in turtles. Gen Comp Endocrinol71:265-267
24. Janzen FJ, Paukstis GL (1991) Environmental sex determination in reptiles: ecology, evolution, and experimental design.
Quart Rev Biol66:149-179
25. Jeyasauria P, Roosenburg WM, Place AR (1994) Role of aromatase in sex determination of the diamondback terrapin, Malaclemys terrapin. J Exp Zoo1 270:95-111
26. Joanen T, McNease L, Ferguson MWJ (1987) The effects of
egg incubation temperature on post-hatching growth of American alligators. In: Webb GJW, Manolis SC, Whitehead PJ
(eds) Wildlife Management: crocodiles and alligators. Surrey
Beatty, Sydney, pp 533-537
27. Lance VA, Bogart MH (1991) Tamoxifen “sex reverses” alligators at male producing temperature, but is an anti-estrogen
in female hatchlings. Experientia 47:263-266
28. Lance VA, Bogart MH (1992) Disruption of ovarian development in alligator embryos treated with an aromatase inhibitor.
Gen Comp Endocrinol86:59-7 1
29. Lance VA, Bogart MH (1994) Studies on sex determination in
the American alligator Alligator mississippiensis. J Exp Zool:
270:79-85
30. Lang JW, Andrews HV (1994) Temperature-dependentsex determination in crocodilians. J Exp Zoo1 270:28-44
31. Lang JW, Andrews H, Whitaker R (1989) Sex determination
and sex ratios in Crocodylus palustris. Am Zoo1 29:935-952
32. Lillie FR (1916) The theory of the free-martin.‘Science 43:
61 1-613
33. McCarrey JR, Abbott UK (1982) Functional differentiation of
chick gonads following depletion of primordial germ cells. J
Embryol Exp Morphol8:161-174
34. Munsterberg A, Lovell-Badge R (1991) Expression of the
mouse anti-Mullerian hormone gene suggests a role in both
male and female sexual differentiation. Development 13:
6 I 3-624
35. Pieau C (1974) DiffCrenciation du sexe en fonction de la
temp6rature chez les embryons d’Emys orbicularis L. (ChClonien); effets des hormones sexuelles. Ann Embryol Morphog
7: 365-394
36. Pieau C. Girondot M, Desvages G, Dorizzi M, Richard-Mercier N, Zaborski P (1994) Environmental control of gonadal
differentiation. In: Short RV, Balabou E (eds) The Differences
between the Sexes, Chap. 20. Cambridge University Press,
Cambridge, pp. 4 3 3 4 8
37. Pivnick EK, Wachtel S, Woods D, Simpson JL, Bishop CE
(1992) Mutations in the conserved domain of SRY are uncommon in XY gonadal dysgenesis. Hum Genet 90:308-310
38. Raynaud A, Pieau C (1985) Embryonic development of the
genital system. In: Gans C, Billet F (eds) Biology of the Reptilia. Vol XV, Development B. J Wiley, New York, pp. 149-300
39. Rhen T, Lang JW (1994) Temperature-dependent sex determination in the snapping turtle: manipulation of the embryonic
sex steroid environment. Gen Comp Endocrinol (in press)
40. Scheib D (1983) Effects and role of estrogens in avian gonadal differentiation. Differentiation 23 (Suppl):S87-92
41. Smith CA, Joss JMP (1993) Gonadal sex differentiation in Alligator mississippiensis, a species with temperature-dependent
sex determination. Cell Tiss Res 273:149-162 .
42. Smith CA, Joss JMP (1994) Steroidogenic enzyme activity
and ovarian differentiation in the saltwater crocodile, Cmcodylus porosus. Gen Comp Endocrinol93:232-245
43. Smith CA, Joss JMP (1994) Sertoli cell differentiation and gonadogenesis in Alligator mississippiensis. J Exp Zoo1 270:
57-70
44. Smith CA, Joss JMP (1994) Uptake of 3H-estradiol by the gonads during the period of sexual differentiation in crocodile
embryos. J Exp Zoo1 270 (in press)
45. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH,
Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk
DC (1985) Measurement of protein using bicinchoninic acid.
Anal Chem 150:76-85
46. Turzillo AM, Furtune JE (1990) Suppression of the secondary
FSH surge with bovine follicular fluid is associated with delayed ovarian follicular development in heifers. J Reprod Fert
. 89543-653
47. Thompson EA, Siiteri PK (1974) Utilization of oxygen and
reduced nicotinamide adenine dinucleotide phosphate by human placental microsomes during aromatization of androstenedione. J Biol Chem 2495364-5372
48. Viets BE, Ewert MA, Talent LG, Nelson CE (1994) Sex-determining mechanisms in squamate reptiles. J Exp Zoo1 270:
45-56
49. Vigier B, Forst MG, Eychenne B, BCzard J, Garrigou 0,
Robel P, Josso N (1989) Anti-Mullerian hormone produces
endocrine sex reversal of fetal ovaries. Proc Natl Acad Sci
USA 86:3684-3688
50. Wartenberg H, Lenz E, Schweikert H-U (1992) Sexual differentiation and the germ cell in sex reversed gonads after aromatase inhibition in the chicken embryo. Andrologia 24: 1-6
51. Wibbels T, Bull JJ, Crews D (1991) Chronology and morphology of temperature-dependent sex determination. J Exp Zoo1
260~371-38 1
52. Wibbels T, Bull JJ, Crews D (1991) Synergism between temperature and estradiol: a common pathway in turtle sex determination? J Exp Zoo1 260: 130-1 34
53. Wibbels T, Bull JJ, Crews D (1994) Temperature-dependent sex
determination: a mechanistic approach. J Exp Zoo1 270:71-78
54.Wibbels T, Crews D (1992) Specificity of steroid hormone-induced sex determination in a turtle. J Endocrinol 133: 121-129
55. Wibbels T, Crews D (1994) Putative aromatase inhibitor induces male sex determination in a female unisexual lizard and
in a turtle with temperature-dependent sex determination. J
Endocrinol 141 :295-299
56. Wibbels T, Gideon P, Bull JJ, Crews D (1993) Estrogen- and
temperature-induced medullary cord regression during gonadal differentiation in a turtle. Differentiation 53: 149-154
57. Willier BH (1939) The embryonic development of sex. In:
Danforth CH, Doisy EA (eds) Sex and Internal Secretions.
Williams and Wilkins, Baltimore, pp. 64-144
58. Witschi E (1939) Modification of the development of sex in
lower vertebrates. In: Danforth CH, Doisy EA (eds) Sex and
Internal Secretions. Williams and Wilkins, Baltimore, pp.
145-226
59. Wolff E, Ginglinger A (1935) Sur le transformation des poulets
males en intersexuts par injection d’hormone femelle (folliculine) aux embryons. Arch Anat Histol Embryol 20:219-278
60. Woods JE, Erton LH (1978) The synthesis of estrogens in the
gonads of the chick embryo. Gen Comp Endocrinol 36:
360-370
61. Yamamoto T (1962) Hormonic factors affecting sex differentiation in fish. Gen Comp Endocrinol I (Suppl): 341-345