GENtRAl.
AND
COMPARATIVE
ENDOCRINOLWJY
88,
(I9921
!(-I9
Adrenal-Kidney and Gonadal Steroidogenesis during Sexual Differentiation of a Reptile with Temperature-Dependent Sex Determination RICHARD B. WHITE’ AND PETER THOMAS The
University
of Texas
Marine
Science Institute, Port Aransas,
P.O. Box 1267, Texas 78373
The University
of Texas
at Austin,
Accepted February I, 1992 Adrenal-kidney and gonadal steroidogenesis were studied during early development in the red-eared slider turtle, Trachemys scripta, which exhibits temperature-dependent sex determination. In vitro steroid secretion by adrenal-kidney-gonad complexes (AKGs) incubated for 6 hr was determined by radioimmunoassay (RIA). AKCis from presumptive males and females secreted progesterone at developmental stages before (stage 151, during (stages 17 and 19), and after (stage 21) the temperature-sensitive period for sex determination, and progesterone secretion increased significantly throughout the period from stage 15 to 21. Presumptive male AKGs secreted significantly more progesterone than female AK& at stage 19. Corticosterone secretion by AKGs was observed at stage 17 in mates only, but in both sexes at stages 19 and 21. Testosterone, estradiol, androstenedione, and dehydroepiandrosterone secretion by AKGs was detected only at stage 21. Of the steroids measured, progesterone and corticosterone were consistently secreted at the highest levels. Although some sex differences were observed, no obvious patterns of sexually dimorphic steroid secretion from AKGs were apparent. Gonads from stage 21, stage 24, and M-day-old hatchlings from both presumptive sexes incubated with [7-3H]pregnenolone showed only weak precursor conversion, primarily to polar metabolites, in incubations as long as 24 hr. None of the steroids assessed by RIA of AKG incubates could be identified by TLC or HPLC analysis of the stage 21 and stage 24 gonadal incubates. However, progesterone was tentatively identified in incubates of IO-day post-hatch female gonads. For stage 21 females, AKGs were separated into gonadal and adrenal-kidney tissue (AK) components and incubated in vitro for 1, 3, and 18 hr. Secretion of progesterone, testosterone, esttadiol, and corticosterone from gonads was nondetectable by RIA, whereas secretion of progesterone and corticosterone by AKs was evident at all three time points and testosterone was detected in the media after 18 hr of incubation. Tissues from these incubations were extracted and assayed for progesterone and testosterone; neither of these steroids was detected in gonads and only progesterone was detected in AKs. These results indicate that the gonads are relatively quiescent, whereas adrenal-kidney tissue is steroidogenically active before, during, and after the temperature-sensitive period for sex determination in T. scripta.
0 1992 Academic Press. Inc
Although the physiology of sexual differentiation in mammals and birds has been well described, particularly with regard to the characterization of steroidogenesis and the morphogenetic effects of steroids (reviewed in Adkins-Regan, 1981; George and Wilson, 1988), research on sexual differen-
tiation in reptiles has been sparse. Previous work focused on localization of steroidogenie enzymes and the effects of exogenous steroids on sexual di@erentiation (reviewed in Raynaud and Pieau, 1985; Crews et al., 1987). While these studies have provided indirect evidence for the role of steroids during early development of reptiles, dl~rect evidence that steroids are present, and thus r Department of Cell and Structural Etiology, Stopford Building, University of Manchester, Manchester knowledge of sex-specific or ontogenetic patterns of steroidogenesis, is lacking. Ml3 9PT, UK. IO 0016~6480/92 $4.00 Copyright 8 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
ADRENAL
STEROIDOGENESIS
The hormonal theory of sexual differentiation holds that the genotype acts directly to produce testes or ovaries, and then the remainder of sexual development proceeds independently of direct genetic effects (Jost, 1953, 1970). The gonads direct the sexual differentiation of somatic cells by secreting hormonal signals (e.g., sex steroids and Miillerian inhibiting substance) at critical times in embryonic development. This theory has been generally supported in the mammal and bird species studied. However, contrary to the hormonal theory, studies on American and Australian marsupials (Renfree and Short, 1988; Shaw et al., 1990) have demonstrated that events once thought to be under gonadal (i.e., steroidal) control in mammalian sexual development are actually under direct genetic control. Many somatic sexual dimorphisms precede the first indication of testicular development (and presumably steroid secretion) in these species, thus suggesting that the hormonal theory of sexual differentiation is less pervasive than previously believed. A conspicuous problem in applying the hormonal theory to reptiles is the phenomenon of temperature-dependent sex determination (TSD), which is widespread in three of the four reptilian orders (Bull, 1980). In TSD, sex is established not by genotype but by egg incubation temperature. Clearly, the initial differentiation of testes or ovaries in reptiles with TSD cannot follow the hormonal theory. However, the question of whether the remainder of sexual differentiation in reptiles proceeds according to the hormonal theory remains unanswered. One of the implicit goals of the present study was to establish a framework for testing this in reptiles. A particular advantage of using a species with TSD to study sexual differentiation is that it allows u priori knowledge of the direction of sexual development. This facilitates the comparison of steroidogenic activity in the presumptive sexes prior to any visible signs of sexual differentiation. In
AND
TSD
11
general, this type of study is difficult or impossible in animals with genotypic ( = chromosomal) sex determination because they must manifest some diagnostic feature of sexual differentiation before the sexes can be separated for experimental comparisons. For the present study, we chose a reptile with TSD, the red-eared slider turtle (Trachemys scripta), which invariably produces 100% males at low egg incubation temperatures and 100% females at comparatively high egg incubation temperatures (Crews et al., 1991; Wibbels et al., 1991). The primary goal of the present study was to describe in vitro steroidogenesis during early development, particularly in relation to the temperature-sensitive period for sex determination. Highly specific and sensitive radioimmunoassays (RIAs) were used to assess the in vitro secretion of steroids. Pilot incubations of embryonic T. scripta adrenal-kidney tissues (AKs) and adrenal-kidney-gonad complexes (AKGs) with [7-3H]pregnenolone followed by TLC and HPLC separations of the radiolabeled products allowed us to tentatively identify (cf. Sandor and Idler, 1972) progesterone and corticosterone as major peaks (unpublished observations). Although androstenedione, dehydroepiandrosterone, and testosterone were not identified from these incubations, immunoreactivity to each of these steroids has been reported in young turtles (Emys orbicularis, Pieau et al., 1982). Furthermore, in adult T. scripta testes, testosterone has been positively identified by Bourne and Licht (1985), and androstenedione and dehydroepiandrosterone have been tentatively identified by Garstka et al. (1991). Estradiol has not previously been identified in T. scripta; however, it has been identified in other reptiles (e.g., Chrysemys picta, Callard et al., 1976; Naja nuja, Lance and Lofts, 1977). Estradiol is known to reverse gonadal sex in T. scripta (Crews et al., 1991) as well as in other reptiles (Pieau, 1974; Bull et al., 1988; Crews et al., 1989), thus suggesting its possible in-
12
WHITE
AND
volvement in normal sexual differentiation. In addition to the assessment of steroid secretion using RIA, incubations of gonads with [7-3H]pregnenolone followed by TLC separation of the radiolabeled products were also performed. We incubated those tissues most likely to be steroidogenically active during sexual differentiation: gonads, AKs, and AKGs. This approach allowed us to evaluate the relative in vitro steroidogenic activity of these tissues. By measuring tissue secretion at specific developmental stages before, during, and after the temperaturesensitive period for sex determination, we were able to describe the ontogeny of steroidogenesis. Possible sex differences in steroid secretion were examined by using embryos from both male-producing and female-producing temperatures. MATERIALS
AND METHODS
Treutment of animals. T. scripta eggs were acquired on the day of laying from a commercial supplier (R. Kliebert Farms) in Hammond, Louisiana. The eggs were transported to Port Aransas, Texas, and kept at 27 2 3” for approximately 10 days. Inviable eggs were discarded and the remaining eggs were placed in covered plastic trays containing moistened vermiculite (1 g vermiculite:0.75 g water). Containers of eggs were placed at either 26.0 ? 1.0” or 31 .O 2 1.0” prior to the temperature-sensitive period for sex determination. Embryos were staged throughout development by periodic examination of external morphology according to criteria previously described for the snapping turtle, Chelydra serpentina (Yntema, 1968). In T. scripta, the temperature-sensitive period begins at about stage 16 (embryo weight -0.17 g) and continues through stage 19 (embryo weight -0.55 g) (Wibbels et al., 1991) which comprises the middle one-third of the total egg incubation period. Hatching occurs at stage 26 (weight -6.5 g). Incubation at 26” invariably produces only males and 31” yields only females (Crews et al., 1991; Wibbels et al., 1991). Reagents. High-purity, glass-distilled ethyl acetate and hexane were obtained from Burdick and Jackson Laboratories (Muskegan, MI). [7-3H]Pregnenolone (13 Ci/mmol) was purchased from New England Nuclear (Boston, MA). [1,2,6,7-3H]Progesterone (85 Ci/mmol), [i,2,6,7-3H]androstenedione (97 Ci/mmol), [I ,2,6,73H]dehydroepiandrosterone (65 Ci/mmol), and [1,23H]corticosterone (53 Ci/mmol) were purchased from
THOMAS
Amersham (Arlington Heights, IL). ]1,2,6,7‘H]Testosterone (90 Ciimmol) and [2,4,6,7‘Hlestradiol (100 Ciimmol) were purchased from Research Products International (Mount Prospect, IL). Radioisotopes were purified by celite-propylene glycol column chromatography before use. Progesterone, androstenedione, and dehydroepiandrosterone, and corticosterone antisera were obtained from Endocrine Sciences (Calabasas Hills, CA). Testosterone and estradiol antisera were generated in rabbits in our laboratory to testosterone-3- and estradiol-3-o-carboxymethyloxime thyroglobulin conjugates, respectively. All other materials were obtained from general laboratory vendors. Tissue incubations for RIA. To describe the ontogeny of steroidogenesis in AK&, embryos from stages 15, 17, 19, and 21 were decapitated and AKGs were removed with the aid of a stereomicroscope. The AKGs were rinsed briefly in incubation medium (Dulbecco’s modified Eagle’s medium F-12 HAM, Hepes buffered (#D8900; Sigma St. Louis, MO), pH 7.6, supplemented with 60 mg penicillin and 100 mg streptomycin/liter), and then placed in 16 x 100 mm borosilicate glass tubes containing 1.0 ml incubation medium. AKGs from 10 animals (stage 15) or 6 animals (stages 17, 19, and 21) were pooled for each sample (n = 3 samples for stage 15; n = 4 samples for stages 17, 19, and 21). Total protein concentration was cl.6 mg (see below) per milliliter of medium for all incubations. The number of samples and tissue pieces per sample were chosen so that incubations comprising one group of replicates could be started ~1 hr after dissection. Tissues were incubated in an oxygen-enriched atmosphere at 28.5 +- 0.5”. a temperature precisely intermediate to the two egg incubation temperatures, to allow direct comparison of the steroid secretion rates by tissues from the two temperature groups. An assumption of this experimental protocol is that a 2.5 shift from egg incubation temperature to tissue incubation temperature would cause at most a minor shift in steroidogenic enzyme kinetics. At the end of the 6-hr incubation period, samples were frozen and stored at - 20” until assayed for steroid content. For comparing steroidogenesis by gonads versus AKs, the above procedures were used with the following modifications. AKGs were separated into gonadal and AK components and incubated separately for 1, 3, and 18 hr to give information on the time course of in vitro steroidogenesis. Tissue and media were analyzed for steroid content, as described below. Stage 21 female tissues were used for these incubations because preliminary work (data not shown) indicated that they were the most steroidogenic of those examined. Tissue extractions for RIA. Tissues were thawed and transferred to another glass tube containing 0.5 ml deionized water and 25 ~1 recovery tracer consisting of -500 dpm [3H]testosterone in borate-bovine serum al-
ADRENAL
STEROIDOGENESIS
bumin (BSA) buffer (0.4 g boric acid, 0.2 g BSA, and 0.1 g sodium azide!liter deionized water, pH 8.0). Samples were sonicated for 30 set and then extracted once with 5 ml hexane: ethyl acetate (1: 1, v/v). Extraction efficiency was determined for each sample and ranged from 81 to 90%. Because previous tests showed no significant difference between recovery of testosterone and progesterone tracers (data not shown), a single tracer was used to monitor the recovery of both steroids. After solvent extraction, protein content was determined by the method of Bradford (1976) using BSA as standard. The protein content of individual gonads, AKs, and AKGs from stage 21 females was 4.3 * 0.4, 118.6 +- 3.9, and 123.0 5 7.1 pg, respectively (mean * SE, n Z= 9 pools of 12 pieces of tissue for each value given). RIAs. Both progesterone and testosterone concentrations were determined from the same tissue extract. Extracts were dried at 45” under N,, reconstituted with 300 pl borate-BSA buffer, and vortexed gently for 5 min. The sample tubes were covered and incubated at 45” for 2 hr to increase reconstitution efficiency. Reconstituted samples were divided into five parts: two dilution aliquots each for measurement of testosterone (100 and 50 ~1) and progesterone (50 and 25 pl), and one aliquot for recovery determination (50 ~1). Aliquots for assay were incubated at 4” overnight with the appropriate antiserum (50 ~1) and tracer (50 pl, -8000 dpm). To increase sensitivity of the assays, antiserum concentrations were set to give 30-35% total binding. Total volumes were adjusted to 150 and 200 pl with borate-BSA buffer for the progesterone and testosterone assays, respectively. Bound steroid was separated from unbound steroid by adding 500 ul Dextran-coated charcoal (0.02 g Dextran T-70 and 0.2 g No&-A charcoaliliter borate-BSA buffer) to each tube, followed by centrifugation at 3000g. The supernatants were decanted into polyethylene scintillation vials, 5 ml of scintillation cocktail (4.0 g 2,5diphenyloxazole and 0.1 g 1,4-bis-[5-phenyl-2oxazolyl] benzene/liter toluene) was added, and the vials were shaken for 10 min to extract the steroids from the aqueous supernatant into the cocktail. Radioactivity was measured with a liquid scintillation spectrometer. Media were assayed as described above except for the following changes. Samples were not extracted before assay, but were assayed directly using duplicate aliquots of 100 pl of medium for testosterone and estradiol and 50 pl of medium for progesterone, corticosterone, androstenedione, and dehydroepiandrosterone. Total volumes were adjusted with ligand-free medium to 225 ~1 for testosterone and estradiol assays and 200 ul for progesterone, corticosterone, androstenedione, and dehydroepiandrosterone assays. Intraassay variation was determined on samples of ligand-free medium spiked with sufficient pure ligand
AND
TSD
13
(10-80 pg, depending on the assay) to give 50-75% total binding. The coefficient of variation (Chard, 1987) for each assay was
AND
COMPARATIVE
ENDOCRINOLWJY
88,
(I9921
!(-I9
Adrenal-Kidney and Gonadal Steroidogenesis during Sexual Differentiation of a Reptile with Temperature-Dependent Sex Determination RICHARD B. WHITE’ AND PETER THOMAS The
University
of Texas
Marine
Science Institute, Port Aransas,
P.O. Box 1267, Texas 78373
The University
of Texas
at Austin,
Accepted February I, 1992 Adrenal-kidney and gonadal steroidogenesis were studied during early development in the red-eared slider turtle, Trachemys scripta, which exhibits temperature-dependent sex determination. In vitro steroid secretion by adrenal-kidney-gonad complexes (AKGs) incubated for 6 hr was determined by radioimmunoassay (RIA). AKCis from presumptive males and females secreted progesterone at developmental stages before (stage 151, during (stages 17 and 19), and after (stage 21) the temperature-sensitive period for sex determination, and progesterone secretion increased significantly throughout the period from stage 15 to 21. Presumptive male AKGs secreted significantly more progesterone than female AK& at stage 19. Corticosterone secretion by AKGs was observed at stage 17 in mates only, but in both sexes at stages 19 and 21. Testosterone, estradiol, androstenedione, and dehydroepiandrosterone secretion by AKGs was detected only at stage 21. Of the steroids measured, progesterone and corticosterone were consistently secreted at the highest levels. Although some sex differences were observed, no obvious patterns of sexually dimorphic steroid secretion from AKGs were apparent. Gonads from stage 21, stage 24, and M-day-old hatchlings from both presumptive sexes incubated with [7-3H]pregnenolone showed only weak precursor conversion, primarily to polar metabolites, in incubations as long as 24 hr. None of the steroids assessed by RIA of AKG incubates could be identified by TLC or HPLC analysis of the stage 21 and stage 24 gonadal incubates. However, progesterone was tentatively identified in incubates of IO-day post-hatch female gonads. For stage 21 females, AKGs were separated into gonadal and adrenal-kidney tissue (AK) components and incubated in vitro for 1, 3, and 18 hr. Secretion of progesterone, testosterone, esttadiol, and corticosterone from gonads was nondetectable by RIA, whereas secretion of progesterone and corticosterone by AKs was evident at all three time points and testosterone was detected in the media after 18 hr of incubation. Tissues from these incubations were extracted and assayed for progesterone and testosterone; neither of these steroids was detected in gonads and only progesterone was detected in AKs. These results indicate that the gonads are relatively quiescent, whereas adrenal-kidney tissue is steroidogenically active before, during, and after the temperature-sensitive period for sex determination in T. scripta.
0 1992 Academic Press. Inc
Although the physiology of sexual differentiation in mammals and birds has been well described, particularly with regard to the characterization of steroidogenesis and the morphogenetic effects of steroids (reviewed in Adkins-Regan, 1981; George and Wilson, 1988), research on sexual differen-
tiation in reptiles has been sparse. Previous work focused on localization of steroidogenie enzymes and the effects of exogenous steroids on sexual di@erentiation (reviewed in Raynaud and Pieau, 1985; Crews et al., 1987). While these studies have provided indirect evidence for the role of steroids during early development of reptiles, dl~rect evidence that steroids are present, and thus r Department of Cell and Structural Etiology, Stopford Building, University of Manchester, Manchester knowledge of sex-specific or ontogenetic patterns of steroidogenesis, is lacking. Ml3 9PT, UK. IO 0016~6480/92 $4.00 Copyright 8 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
ADRENAL
STEROIDOGENESIS
The hormonal theory of sexual differentiation holds that the genotype acts directly to produce testes or ovaries, and then the remainder of sexual development proceeds independently of direct genetic effects (Jost, 1953, 1970). The gonads direct the sexual differentiation of somatic cells by secreting hormonal signals (e.g., sex steroids and Miillerian inhibiting substance) at critical times in embryonic development. This theory has been generally supported in the mammal and bird species studied. However, contrary to the hormonal theory, studies on American and Australian marsupials (Renfree and Short, 1988; Shaw et al., 1990) have demonstrated that events once thought to be under gonadal (i.e., steroidal) control in mammalian sexual development are actually under direct genetic control. Many somatic sexual dimorphisms precede the first indication of testicular development (and presumably steroid secretion) in these species, thus suggesting that the hormonal theory of sexual differentiation is less pervasive than previously believed. A conspicuous problem in applying the hormonal theory to reptiles is the phenomenon of temperature-dependent sex determination (TSD), which is widespread in three of the four reptilian orders (Bull, 1980). In TSD, sex is established not by genotype but by egg incubation temperature. Clearly, the initial differentiation of testes or ovaries in reptiles with TSD cannot follow the hormonal theory. However, the question of whether the remainder of sexual differentiation in reptiles proceeds according to the hormonal theory remains unanswered. One of the implicit goals of the present study was to establish a framework for testing this in reptiles. A particular advantage of using a species with TSD to study sexual differentiation is that it allows u priori knowledge of the direction of sexual development. This facilitates the comparison of steroidogenic activity in the presumptive sexes prior to any visible signs of sexual differentiation. In
AND
TSD
11
general, this type of study is difficult or impossible in animals with genotypic ( = chromosomal) sex determination because they must manifest some diagnostic feature of sexual differentiation before the sexes can be separated for experimental comparisons. For the present study, we chose a reptile with TSD, the red-eared slider turtle (Trachemys scripta), which invariably produces 100% males at low egg incubation temperatures and 100% females at comparatively high egg incubation temperatures (Crews et al., 1991; Wibbels et al., 1991). The primary goal of the present study was to describe in vitro steroidogenesis during early development, particularly in relation to the temperature-sensitive period for sex determination. Highly specific and sensitive radioimmunoassays (RIAs) were used to assess the in vitro secretion of steroids. Pilot incubations of embryonic T. scripta adrenal-kidney tissues (AKs) and adrenal-kidney-gonad complexes (AKGs) with [7-3H]pregnenolone followed by TLC and HPLC separations of the radiolabeled products allowed us to tentatively identify (cf. Sandor and Idler, 1972) progesterone and corticosterone as major peaks (unpublished observations). Although androstenedione, dehydroepiandrosterone, and testosterone were not identified from these incubations, immunoreactivity to each of these steroids has been reported in young turtles (Emys orbicularis, Pieau et al., 1982). Furthermore, in adult T. scripta testes, testosterone has been positively identified by Bourne and Licht (1985), and androstenedione and dehydroepiandrosterone have been tentatively identified by Garstka et al. (1991). Estradiol has not previously been identified in T. scripta; however, it has been identified in other reptiles (e.g., Chrysemys picta, Callard et al., 1976; Naja nuja, Lance and Lofts, 1977). Estradiol is known to reverse gonadal sex in T. scripta (Crews et al., 1991) as well as in other reptiles (Pieau, 1974; Bull et al., 1988; Crews et al., 1989), thus suggesting its possible in-
12
WHITE
AND
volvement in normal sexual differentiation. In addition to the assessment of steroid secretion using RIA, incubations of gonads with [7-3H]pregnenolone followed by TLC separation of the radiolabeled products were also performed. We incubated those tissues most likely to be steroidogenically active during sexual differentiation: gonads, AKs, and AKGs. This approach allowed us to evaluate the relative in vitro steroidogenic activity of these tissues. By measuring tissue secretion at specific developmental stages before, during, and after the temperaturesensitive period for sex determination, we were able to describe the ontogeny of steroidogenesis. Possible sex differences in steroid secretion were examined by using embryos from both male-producing and female-producing temperatures. MATERIALS
AND METHODS
Treutment of animals. T. scripta eggs were acquired on the day of laying from a commercial supplier (R. Kliebert Farms) in Hammond, Louisiana. The eggs were transported to Port Aransas, Texas, and kept at 27 2 3” for approximately 10 days. Inviable eggs were discarded and the remaining eggs were placed in covered plastic trays containing moistened vermiculite (1 g vermiculite:0.75 g water). Containers of eggs were placed at either 26.0 ? 1.0” or 31 .O 2 1.0” prior to the temperature-sensitive period for sex determination. Embryos were staged throughout development by periodic examination of external morphology according to criteria previously described for the snapping turtle, Chelydra serpentina (Yntema, 1968). In T. scripta, the temperature-sensitive period begins at about stage 16 (embryo weight -0.17 g) and continues through stage 19 (embryo weight -0.55 g) (Wibbels et al., 1991) which comprises the middle one-third of the total egg incubation period. Hatching occurs at stage 26 (weight -6.5 g). Incubation at 26” invariably produces only males and 31” yields only females (Crews et al., 1991; Wibbels et al., 1991). Reagents. High-purity, glass-distilled ethyl acetate and hexane were obtained from Burdick and Jackson Laboratories (Muskegan, MI). [7-3H]Pregnenolone (13 Ci/mmol) was purchased from New England Nuclear (Boston, MA). [1,2,6,7-3H]Progesterone (85 Ci/mmol), [i,2,6,7-3H]androstenedione (97 Ci/mmol), [I ,2,6,73H]dehydroepiandrosterone (65 Ci/mmol), and [1,23H]corticosterone (53 Ci/mmol) were purchased from
THOMAS
Amersham (Arlington Heights, IL). ]1,2,6,7‘H]Testosterone (90 Ciimmol) and [2,4,6,7‘Hlestradiol (100 Ciimmol) were purchased from Research Products International (Mount Prospect, IL). Radioisotopes were purified by celite-propylene glycol column chromatography before use. Progesterone, androstenedione, and dehydroepiandrosterone, and corticosterone antisera were obtained from Endocrine Sciences (Calabasas Hills, CA). Testosterone and estradiol antisera were generated in rabbits in our laboratory to testosterone-3- and estradiol-3-o-carboxymethyloxime thyroglobulin conjugates, respectively. All other materials were obtained from general laboratory vendors. Tissue incubations for RIA. To describe the ontogeny of steroidogenesis in AK&, embryos from stages 15, 17, 19, and 21 were decapitated and AKGs were removed with the aid of a stereomicroscope. The AKGs were rinsed briefly in incubation medium (Dulbecco’s modified Eagle’s medium F-12 HAM, Hepes buffered (#D8900; Sigma St. Louis, MO), pH 7.6, supplemented with 60 mg penicillin and 100 mg streptomycin/liter), and then placed in 16 x 100 mm borosilicate glass tubes containing 1.0 ml incubation medium. AKGs from 10 animals (stage 15) or 6 animals (stages 17, 19, and 21) were pooled for each sample (n = 3 samples for stage 15; n = 4 samples for stages 17, 19, and 21). Total protein concentration was cl.6 mg (see below) per milliliter of medium for all incubations. The number of samples and tissue pieces per sample were chosen so that incubations comprising one group of replicates could be started ~1 hr after dissection. Tissues were incubated in an oxygen-enriched atmosphere at 28.5 +- 0.5”. a temperature precisely intermediate to the two egg incubation temperatures, to allow direct comparison of the steroid secretion rates by tissues from the two temperature groups. An assumption of this experimental protocol is that a 2.5 shift from egg incubation temperature to tissue incubation temperature would cause at most a minor shift in steroidogenic enzyme kinetics. At the end of the 6-hr incubation period, samples were frozen and stored at - 20” until assayed for steroid content. For comparing steroidogenesis by gonads versus AKs, the above procedures were used with the following modifications. AKGs were separated into gonadal and AK components and incubated separately for 1, 3, and 18 hr to give information on the time course of in vitro steroidogenesis. Tissue and media were analyzed for steroid content, as described below. Stage 21 female tissues were used for these incubations because preliminary work (data not shown) indicated that they were the most steroidogenic of those examined. Tissue extractions for RIA. Tissues were thawed and transferred to another glass tube containing 0.5 ml deionized water and 25 ~1 recovery tracer consisting of -500 dpm [3H]testosterone in borate-bovine serum al-
ADRENAL
STEROIDOGENESIS
bumin (BSA) buffer (0.4 g boric acid, 0.2 g BSA, and 0.1 g sodium azide!liter deionized water, pH 8.0). Samples were sonicated for 30 set and then extracted once with 5 ml hexane: ethyl acetate (1: 1, v/v). Extraction efficiency was determined for each sample and ranged from 81 to 90%. Because previous tests showed no significant difference between recovery of testosterone and progesterone tracers (data not shown), a single tracer was used to monitor the recovery of both steroids. After solvent extraction, protein content was determined by the method of Bradford (1976) using BSA as standard. The protein content of individual gonads, AKs, and AKGs from stage 21 females was 4.3 * 0.4, 118.6 +- 3.9, and 123.0 5 7.1 pg, respectively (mean * SE, n Z= 9 pools of 12 pieces of tissue for each value given). RIAs. Both progesterone and testosterone concentrations were determined from the same tissue extract. Extracts were dried at 45” under N,, reconstituted with 300 pl borate-BSA buffer, and vortexed gently for 5 min. The sample tubes were covered and incubated at 45” for 2 hr to increase reconstitution efficiency. Reconstituted samples were divided into five parts: two dilution aliquots each for measurement of testosterone (100 and 50 ~1) and progesterone (50 and 25 pl), and one aliquot for recovery determination (50 ~1). Aliquots for assay were incubated at 4” overnight with the appropriate antiserum (50 ~1) and tracer (50 pl, -8000 dpm). To increase sensitivity of the assays, antiserum concentrations were set to give 30-35% total binding. Total volumes were adjusted to 150 and 200 pl with borate-BSA buffer for the progesterone and testosterone assays, respectively. Bound steroid was separated from unbound steroid by adding 500 ul Dextran-coated charcoal (0.02 g Dextran T-70 and 0.2 g No&-A charcoaliliter borate-BSA buffer) to each tube, followed by centrifugation at 3000g. The supernatants were decanted into polyethylene scintillation vials, 5 ml of scintillation cocktail (4.0 g 2,5diphenyloxazole and 0.1 g 1,4-bis-[5-phenyl-2oxazolyl] benzene/liter toluene) was added, and the vials were shaken for 10 min to extract the steroids from the aqueous supernatant into the cocktail. Radioactivity was measured with a liquid scintillation spectrometer. Media were assayed as described above except for the following changes. Samples were not extracted before assay, but were assayed directly using duplicate aliquots of 100 pl of medium for testosterone and estradiol and 50 pl of medium for progesterone, corticosterone, androstenedione, and dehydroepiandrosterone. Total volumes were adjusted with ligand-free medium to 225 ~1 for testosterone and estradiol assays and 200 ul for progesterone, corticosterone, androstenedione, and dehydroepiandrosterone assays. Intraassay variation was determined on samples of ligand-free medium spiked with sufficient pure ligand
AND
TSD
13
(10-80 pg, depending on the assay) to give 50-75% total binding. The coefficient of variation (Chard, 1987) for each assay was
