Male Marmoset Monkeys Express an Adrenal Fetal Zone at Birth, But Not a Zona Reticularis in Adulthood
http://www.100md.com
内分泌学杂志 2005年第1期
Perinatal Research Laboratories (J.C.P., I.M.B.), Department of Obstetrics/Gynecology (J.C.P., D.H.A., I.M.B.), National Primate Research Center (D.H.A., W.S., A.J.A.), University of Wisconsin-Madison, Madison, Wisconsin 53715; Department of Biology (W.S.), University of California-Riverside, Riverside, California 92521; Population Health and Reproduction (A.D.N., A.J.C.), School of Veterinary Medicine, and Department of Medical Pharmacology and Toxicology (G.H., H.J.), University of California-Davis, Davis, California 95616; and Laboratory for Behavioural Neurobiology (C.R.P.), Swiss Federal Institute of Technology, Zurich CH-8603, Switzerland
Address all correspondence and requests for reprints to: Ian M. Bird, Ph.D, Perinatal Research Laboratories, 7E Meriter Hospital/Park, 202 South Park Street, Madison, Wisconsin 53715. E-mail: imbird@wisc.edu.
Abstract
Neonatal human males produce high levels of dehydroepiandrosterone (DHEA) and its sulfo-conjugated form (DS) that decline within a few months of birth, due to regression of the adrenal fetal zone (FZ). Adult male humans and rhesus monkeys produce C19 steroids in abundance from the adrenal zona reticularis (ZR). Male marmoset monkeys produce DS at birth, but unlike humans and rhesus monkeys, do not produce comparable amounts of DHEA and DS in adulthood. To determine whether male marmosets express a functional ZR in adulthood, we examined adult and neonatal male marmosets for the presence of a ZR and FZ, respectively. Exogenous ACTH failed to stimulate DHEA or DS in adults, and dexamethasone treatment failed to suppress DHEA and DS, although cortisol levels changed as expected. In steroidogenic tissues, the key proteins necessary to synthesize C19 steroids from pregnenolone are P450c17, 3?-hydroxysteroid dehydrogenase (3?-HSD), nicotinamide adenine dinucleotide phosphate (reduced) oxido-reductase cytochrome P450 (reductase), and cytochromeb5 (cytb5). Adult adrenal cross sections showed P450c17 and reductase protein expression throughout the cortex but showed no expected decrease in 3?-HSD and increase in cytb5 in the innermost region. Western analysis confirmed these data, demonstrating comparable P450c17 expression to rhesus monkeys, but not cytb5. HPLC analysis revealed similar 17-hydroxylase action on pregnenolone for adult marmoset and rhesus adrenal microsomes but greatly diminished 17,20-lyase activity in marmosets. Neonatal marmoset adrenals exhibited staining indicative of a putative FZ (with P450c17, reduced 3?-HSD and increased cytb5). We conclude that neonatal marmosets exhibit a C19 steroid-secreting FZ similar to humans, but adult males fail to acquire a functional ZR.
Introduction
THE COMMON MARMOSET (Callithrix jacchus) is a New World monkey that is widely used in biomedical research (1, 2, 3), including studies on stress (4) and the effects of behavioral subordination on adrenocortical function (5, 6, 7). Nonetheless, little is known of the structure or function of the fetal or adult marmoset adrenal and how it may differ from humans. Although adrenal morphology is relatively straightforward in other primate models, the histological zonation of the marmoset adrenal cortex has endured much controversy in the literature. Miraglia and Moreira (8) reported in 1969 that previous histological staining provided no conclusive evidence for the presence or absence of a zona reticularis (ZR) in either the male or the female marmoset adrenal: the putative ZR was indistinct and difficult to describe definitively. There was, however, clear morphological evidence for the presence of a zona glomerulosa (ZG) and a zona fasciculata (ZF). In their study, Miraglia and Moreira (8) interpreted morphological evidence in favor of a ZR, although it was difficult to visualize and staining techniques were nonimmunospecific. Levine et al. (9) also performed hematoxylin and eosin staining of marmoset adrenals and found evidence of a fetal zone (FZ) in neonates, but no expression of a ZR in adult males or females. The adults lacked morphologically distinct cell types between the mid-adrenal and the innermost adrenocortical region (the cells abutting the cortico-medullary junction). The finding of Levine et al. (9) was substantiated by an extensive measurement of circulating steroids in vivo. Both male and female marmosets had relatively high circulating levels of dehydroepiandrosterone (DHEA; males 1343 ± 266 ng/dl, females 3110 ± 1083 ng/dl; mean ± SEM), and the sulfoconjugate form DS at birth that declined within a few months, consistent with the regression of a functional FZ (9). In contrast to humans and other primates, such as rhesus monkeys, however, male marmosets do not produce large amounts of DHEA and DS upon reaching sexual maturity (9). Thus, the question remains: is the lack of DS production due to an endocrine suppression mechanism or to lack of a functional ZR in the adrenal cortex of adult male marmosets?
Since the aforementioned studies were performed, there has been considerable progress in our understanding of the enzymology of adrenal steroidogenesis and the factors that underlie normal ZR function. The adrenal cortex is known to be the source of mineralocorticoids, glucocorticoids, and, in both human and many nonhuman primates, C19 steroids. The ability to synthesize these three classes of steroids is achieved through discrete adrenal zonation (Ref.10 and Fig. 1). Mechanisms underlying zonation have been studied extensively in experimental models such as rat, mouse, and cow, and these have established that zonal function derives from the presence or, just as importantly, the absence of certain enzymes. Although these nonprimate models lack any significant ZR, studies in humans and nonhuman primates have demonstrated that the adult ZR is capable of producing C19 steroids, most specifically DHEA and DS, due to its high expression of 17-hydroxylase/17,20-lyase cytochrome P450 (P450c17) combined with low expression of competing 3?-hydroxysteroid dehydrogenase (3?-HSD) (Refs.11, 12, 13, 14, 15, 16 and Fig. 1). In addition, recent studies suggest that the ZR is also often rich in cytochromeb5 (cytb5) (14, 15, 16).
FIG. 1. Steroid biosynthesis in the adrenal cortex of higher mammals. Solid arrows, Predominant pathway in species expressing P450c17 in the adrenal gland. Dashed arrows, Steroid biosynthesis seen in species using the -4-lyase pathway.
Despite these advances in our understanding of ZR function, immunospecific staining for key steroidogenic enzymes in the adrenal, and correlative in vivo challenges of adrenocortical function, have not been performed on male marmosets. Our aim, therefore, was to localize the enzymes and cofactors influencing DHEA production to specific adrenal zones and to correlate those findings with results of functional manipulation of circulating levels of adrenal steroids. We found that although male marmosets exhibit a neonatal adrenal morphology and steroid production pattern that is developmentally similar to humans, they do not express a functional ZR upon maturation to adulthood. These unique properties make the male marmoset a highly unusual model that shares developmental similarities to humans, but in adulthood, is sufficiently different that it would be of considerable value in determining what factors may be necessary for normal human ZR development and function.
Materials and Methods
Materials
Unless otherwise stated, general reagents were obtained from Sigma-Aldrich (St. Louis, MO) or Fisher Scientific International (Pittsburgh, PA).
Animals
This research was conducted in accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals and the Animal Welfare Act and its subsequent amendments. All animal procedures were reviewed and approved by the Graduate School Animal Care and Use Committee of the University of Wisconsin-Madison. The National Primate Research Center at the University of Wisconsin-Madison (WPRC) is accredited by AAALAC as part of the UW-Madison Graduate School. Adult male marmosets, as detailed below, were housed indoors at the WPRC. They lived in aluminum and welded wire cages measuring 61 x 91 x 183 cm or 122 x 61 x 183 cm, or in a large room measuring 363 x 212 x 218 cm. Males were housed either singly, with an adult female pairmate or in a larger mixed-sex group, and had visual, olfactory, and auditory access to conspecifics in other cages. Lights were on from 0600–1800 h, and animals were fed at 1230–1330 h daily. Additional information on animal housing and husbandry has been published previously (6). Infant and adult rhesus monkey (Macaca mulatta) tissue was obtained, opportunistically, from healthy animals assigned as controls to other approved projects at the California National Primate Research Center.
Steroid responses in vivo
Adrenocortical responses to ACTH1–39 (ACTH1–39) were determined in four adult male marmosets, which were 38–62 months of age and weighed 365 ± 8 g (mean ± SEM) at the beginning of data collection. Four doses of human ACTH1–39 (0.1, 1, 10, and 100 μg/kg) and a saline control were administered to each male at intervals of at least 2 wk, with the order of treatments balanced across animals. Males were weighed and injected with dexamethasone (5 mg/kg, im; American Regent Laboratories, Shirley, NY) at 1700 h to suppress endogenous ACTH secretion. The following morning at 0900 h, each male was administered human ACTH1–39 or saline iv. Blood samples (0.2–0.6 ml) were collected immediately before, 60 min after, and 120 min after ACTH1–39 injection. In a separate experiment, adrenocortical responses to dexamethasone were determined in six adult male marmosets (25–54 months of age, 377 ± 11 g). Dexamethasone (5 mg/kg) was injected im at 1600 h, and blood samples (1.2 ml) were collected at 0900 h the day of and the day after dexamethasone treatment, within 3 min of initial disturbance to the animal. In both experiments, marmosets were briefly restrained in a marmoset restraint tube (17), whereas blood was collected from the femoral vein using a heparinized 1-cc syringe and 27-gauge needle. Samples were immediately placed on ice and centrifuged at 2000 rpm for 10 min, and the plasma fractions were aspirated and frozen at –20 C until assayed.
Hormone assays
All plasma hormone concentrations were determined by RIAs that had been validated for use with marmoset plasma at the WPRC Assay Services laboratories. The cortisol and testosterone assays were described previously (5, 18). Their intra- and interassay coefficients of variation (CVs) were 5.65% and 10.48%, respectively, for cortisol, and 2.2% and 5.5%, respectively, for testosterone. The remaining RIAs for corticosterone (antibody: Esoterix Endocrinology, Calabasas Hills, CA; tritiated trace: American Radiolabeled Chemicals, Inc., St. Louis, MO; reference preparation: Sigma-Aldrich), DHEA (antibody-coated tube RIA, TKDH2; DPC, Los Angeles, CA), DS (antibody-coated tube RIA, TKDS1; DPC) and aldosterone (antibody-coated-tube RIA, Active Aldosterone; Diagnostics Systems Laboratories, Inc., Webster, TX), produced the following intra- and interassay CVs: corticosterone, 2.2% and 9.0%; DHEA, 3.6% and 4.2%; DS, 1.7%, interassay CV not available for a single assay; and aldosterone, 2.3% and 5.4%. Corticosterone, testosterone, and DHEA RIAs were performed after celite chromatography.
The recovery of DS standards added to 50 μl of a marmoset plasma pool (n = 5) was 103.3 ± 2.2%. Serial dilution of a DS-spiked marmoset pool (50.00–0.39 μl; n = 8) gave a displacement curve parallel to that obtained with DS standards. The sensitivity of the assay was 5.0 μg/dl. The recovery of DHEA standards added to 100 μl of a marmoset plasma pool (n = 7) was 99.0 ± 2.4%. Serial dilution of a DHEA-spiked marmoset pool (100.00–1.56 μl; n = 7) gave a displacement curve parallel to that obtained with DHEA standards. The sensitivity of the assay was 0.5 ng/ml.
Immunohistochemistry (IHC)
Adrenals from adult male marmosets (aged 24–72 months, n = 6) were fixed in 10% formalin, processed routinely, and paraffin embedded. Adrenals from 1-d-old marmosets (n = 5 males, n = 1 female; 2 adrenals donated from the Laboratory of Behavioral Neurobiology, Swiss Federal Institute of Technology, and 4 from the WPRC) were fixed in 4% paraformaldehyde in 0.1 M sodium cacodylate buffer, processed routinely and embedded in paraffin wax. Tissues were then cut into 5-μm sections. All sections were deparaffinized and rehydrated in clearing solution and graded alcohols. Endogenous peroxidase activity was quenched by incubation in 30% hydrogen peroxide in 100% methanol for 30 min. Staining was detected using the Vectastain Elite ABC Kit (rabbit IgG) (Vector Laboratories, Inc., Burlingame, CA). Sections were subsequently stained using antiserum raised against human P450c17 (1:1000, chicken polyclonal, Dr. A. J. Conley, Department of Population Health and Reproduction, University of California-Davis), human placenta 3?-HSD (1:400, rabbit polyclonal, Dr. J. Ian Mason, Clinical Biochemistry, University of Edinburgh, Edinburgh, Scotland, UK), human cytb5 (1:3000, rabbit polyclonal, Dr. A. J. Conley), rat nicotinamide adenine dinucleotide phosphate (reduced) oxido-reductase cytochrome P450 (reductase; 1:3000, rabbit polyclonal, Dr. A. J. Conley), human side-chain-cleavage cytochrome P450 (P450scc; 1:1000, rabbit polyclonal, Dr. W. L. Miller, Department of Pediatrics, University of California-San Francisco), and human 21-hydroyxlase cytochrome P450 (P450c21; 1:3000, rabbit polyclonal, Dr. W. L. Miller). Normal nonimmune rabbit serum (1:1000, Vector Laboratories, Inc.) was used as a control. All sections were incubated with diluted normal goat serum (1:200) for 20 min at room temperature to block nonspecific binding. They were then incubated with primary antibody diluted in buffer at 4 C overnight. The sections were incubated in diluted secondary antibody (1:200, antirabbit, or 5 μg/ml, antichicken) and avidin-biotin-conjugated peroxidase for 30 min each at room temperature before exposure to 3-amino-9-ethylcarbazole (Vector Laboratories, Inc.) for 10 min also at room temperature. Nuclei were counterstained blue with Gill 2 hematoxylin (neat, 2 sec, rinsed 5 min in tap water) before sections were coverslipped with Paramount (Dako Corp., Carpinteria, CA). For photo-quality purposes, some sections were restained using 3,3'-diaminobenzidine (Vector Laboratories, Inc.) for 5 min at room temperature, counterstained as above, dehydrated, and coverslipped with permanent mounting media. Images were captured using bright-field illumination on a Zeiss Axioskop (Carl Zeiss, Oberkochen, Germany) fitted with a SPOT2 digital camera (through Zeiss), and digital-grayscale densitometry (Zeiss software, KS 300 version 3.0) was then used to quantify relative staining.
Microsomal preparation and Western immunoblotting
Whole frozen adrenals from six adult male marmosets (aged 60–132 months) and two infant rhesus monkeys (1 male and 1 female, aged 3 and 4 months, respectively) were homogenized directly in lysis buffer [0.1 M KPO4 (pH 7.4), 20% glycerol, 5 mM ?-mercaptoethanol, and 0.5 mM phenylmethylsulfonyl fluoride] at a ratio of 0.1 g/1 ml tissue to buffer. Cellular debris was removed by centrifugation at 1000 x g for 10 min. The supernatant was then spun down at 16,000 x g for 10 min under the same conditions. The premicrosomal pellet was stored at –80 C, and the supernatant was spun down at 100,000 x g for 60 min. The pellet was rehomogenized in dialysis buffer [lysis buffer with 1 mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate detergent] and stored at –80 C. Protein concentration of microsomal preparations was determined using a BCA Protein Assay Kit (Pierce Biotechnology, Inc., Rockford, IL). Ten (10) micrograms of total microsomal protein were size separated per lane on a 16% polyacrylamide gel (Bio-Rad Laboratories, Inc., Hercules, CA, 150 V constant, 1.5 h) and protein transferred to PVDF membrane (Bio-Rad Laboratories, Inc., 100 V, 1 h). Western blotting was then performed using the antihuman P450c17 and antihuman cytb5 antibodies (P450c17 1:10,000, chicken horseradish peroxidase-linked IgY secondary antibody 1:5000, cytb5 1:2000, donkey antirabbit horseradish peroxidase-linked IgG secondary antibody 1:10,000, detected by ECL (Amersham Biosciences, Piscataway, NJ) as previously described (19). Recombinant human cytb5 was used as a control.
P450c17 substrate preference and relative -4-lyase deficiency assay/analysis
Synthesis of DHEA by marmoset and rhesus P450c17 of adrenal origin and relative -4-lyase deficiency were examined by radio-HPLC analysis of substrate metabolism. Microsomal protein (20 μg) from the adrenals of six adult male marmosets (aged 60–132 months) and six juvenile and adult male rhesus monkeys (aged 3–96 months), prepared as above, were used in 0.5 ml assays. Pregnenolone (Steraloids, Newport, RI; [3H]-pregnenolone, PerkinElmer, Boston, MA) was used as the substrate to examine DHEA synthesis capabilities of marmoset and rhesus adrenal P450c17. Treatment with and without the 3?-HSD inhibitor trilostane (50 μM; Sanofi Research Division, Great Valley, PA) was performed for each subject. -4-Lyase deficiency was investigated using the same methods, but with 17-hydroxy-progesterone (Steraloids; 1, 2, 6, 7 [3H]-17-hydroxy-progesterone, American Radiolabeled Chemicals, St. Louis, MO) as the substrate; trilostane was not used in this study. For both studies, 10 μM substrate (1% hot, 99% cold) was combined with trilostane (when used) and evaporated to dryness under air. After evaporation, substrate (and trilostane) was resuspended in 460–467 μl assay buffer (50 mM KPO4 and 1 mM EDTA). Microsomal protein (20 μg) was added to the resuspension, and to maintain a constant supply of reducing equivalents (dihydronicotinamide adenine dinucleotide phosphate), a generating system consisting of 17 mM glucose-6-phosphate, 2 mM ?-nicotinamide adenine dinucleotide phosphate, 1 mM dihydronicotinamide adenine dinucleotide phosphate, and 1 U glucose-6-phosphate-dehydrogenase was also added. After incubation for 1 h in a 37 C water bath, 0.5 ml buffer was added. Extraction of steroids in the entire assay was performed with 10 ml methylene chloride. All samples were spun for 10 min at 2000 x g; the aqueous phase was then removed and the organic phase was poured into a clean tube and evaporated to dryness. All samples were resuspended in 1 ml HPLC-grade methanol and subjected to radio-HPLC. 17-Hydroxylase activity was determined from the sum of 17-hydroxylated metabolite (11-deoxycortisol, 17-hydroxy-pregnenolone, DHEA) radioactivity after conversion to amount of substrate metabolized.
Radio-HPLC
All solvents (including water) were of HPLC grade and all chemicals were of analytical grade. Radio-HPLC analysis of the samples was performed with an HP1100 liquid chromatographic system (Agilent Technologies, Menlo Park, CA) interfaced with a Radiomatic Model 500TR Flow Scintillation Analyzer (Packard Instrument Co., Meriden, CT). The HPLC system consisted of a solvent degasser, binary pump, autosampler, column heater, and a variable wavelength detector. Scintillation fluid (Ultima-flo, Packard Instrument Co.) flow rate was 3 ml/min. Samples (10 μl) were injected onto a 250 x 4.6 mm C18 HPLC column (BetaBasic-18, Keystone Scientific, Bellefonte, PA), then eluted with a mobile phase consisting of A (84% ammonium acetate buffer:8% MeOH:8% acetonitrile) and B (42% MeOH:42% acetonitrile:16% H2O). B was held at 20% for 5 min, increased from 20–95% by a linear ramp (1 ml/min) from 5–20 min, held at 95% from 20–40 min, reduced to 20% from 40–45 min, then reequilibrated at 20% from 45–50 min. Positive controls ([3H]pregnenolone and [3H]17-hydroxy-progesterone) were injected with each assay to verify the radiodetector response and retention time. When appropriate, methanol blanks were injected between samples to monitor carryover. Retention times and area-under-the-curve of the radioactive peaks were calculated using FLO-ONE software (Packard Instrument Co.).
Statistics
All values are expressed as mean ± SEM, unless otherwise stated. Dose-dependent effects of ACTH on cortisol were analyzed by one-way ANOVA followed by post hoc univariate F tests. Other in vivo steroid levels were compared by paired t tests. One-way ANOVA was performed on the densitometry of the adult male marmoset adrenal sections stained for P450c17 and 3?-HSD. Student’s t test was used to determine differences in whole adrenal microsomal substrate metabolism.
Results
Plasma steroids
Injection of ACTH1–39 in dexamethasone-suppressed adult male marmosets resulted in a dose-dependent rise in cortisol by 60 min and 120 min, as shown in Fig. 2A. The threshold dose for ACTH1–39 responsiveness was 1 μg/kg, and the maximal response was induced by 10 μg/kg at 60 min and by 100 μg/kg at 120 min. Pre-ACTH1–39 DS was present only at a level below the assay sensitivity threshold of 5 μg/dl (50 ng/ml), and no elevation into the detectable range occurred in response to 10 μg/kg of ACTH1–39 at 60 min post injection (Fig. 2B), the time of maximal cortisol response (Fig. 2A). DHEA was assayed in samples from marmosets treated with 100 μg/kg (Fig. 2B). DHEA concentrations approximated those at the sensitivity limit of the assay (0.5 ng/ml) before ACTH1–39 treatment and did not increase significantly in response to ACTH1–39. Plasma aldosterone and testosterone levels did not rise in response to ACTH1–39 (1.65-fold increase and 3.2-fold decrease, respectively; Fig. 2B), confirming the specificity of the ACTH1–39 response to the adrenals rather than the gonads. When baseline steroid levels were taken into account, however, aldosterone levels did exhibit a significant proportionate increase (P < 0.05), as would be expected using a supraphysiological dose of ACTH. Changes in circulating corticosterone levels mirrored those of cortisol, showing at least a 41-fold increase from below detectable levels (<0.35 ng/ml) in response to 10 μg/kg ACTH1–39 at 60 min. These latter findings were more consistent with a ZF rather than the ZG origin of corticosterone production and secretion because they were relatively similar to the degree of cortisol elevation, but not of aldosterone. The ratio of corticosterone:cortisol was 0.0173 indicating 1.73% leakage from the cortisol pathway.
FIG. 2. A, Plasma cortisol dose-response to ACTH1–39 at 0 (0900 h), 60 (1000 h), and 120 (1100 h) min post injection in four male marmosets, aged 38–62 months. Black circles represent 60-min values and open circles represent 120-min values. *, Significant difference between dose and saline control (P < 0.05); ^, significant difference between dose and previous dose (P < 0.05). B, Plasma steroid levels in response to 10 μg/kg or 100 μg/kg (DHEA only) ACTH1–39 at 0 (0900 h) and 60 (1000 h) min post injection in the same four animals as in panel A. Open bars, Pre-ACTH1–39 levels (after 5 mg/kg dexamethasome overnight treatment); black bars, ACTH1–39 treatment levels. Note log scale for steroid level. *, Significant difference from pre-ACTH1–39 level (P < 0.05); ^, significant difference in absolute response (P < 0.05). C, Plasma steroid levels in response to 5 mg/kg dexamethasone overnight treatment levels in response to 5 mg/kg dexamethasone overnight treatment (basal 0900 h, dexamethasone 0900 h) in six male marmosets, aged 25–54 months. Open bars represent basal, and black bars represent dexamethasone treatment levels. Note log scale for steroid level. *, Significant difference from basal level (P < 0.05).
The contrasting experiment using dexamethasone-only treatment to suppress pituitary ACTH revealed a corresponding reduction in plasma cortisol levels, as expected (Fig 2C). Consistent with the results of the ACTH1–39 challenge, DS and DHEA were again undetectable or near the limit of sensitivity, respectively. Plasma testosterone and aldosterone levels were not significantly affected by dexamethasone-only treatment, which is consistent with a gonadal origin of testosterone and with the fact that in the physiological range, ACTH is not a major regulator of aldosterone secretion. In addition, plasma corticosterone levels diminished along with cortisol, consistent with the ZF, and not the ZG, as the zone of origin for corticosterone biosynthesis.
IHC localization of steroidogenic enzymes in adult male marmosets
To compare adrenocortical enzyme expression patterns of marmosets to those of humans and rhesus, we stained fixed sections for the major steroidogenic enzymes that influence DHEA and DS synthesis. As expected, the adult male marmoset ZG was lacking in P450c17 expression, although staining was obvious throughout the rest of the adrenal cortex (Fig. 3A). Contrary to expression patterns in primates exhibiting a ZR, 3?-HSD stained throughout the adrenal cortex, including the cortico-medullary junction, and cytb5 staining was low or absent in the entire cortex, especially in the innermost region. Staining for cytochromes reductase, P450scc and P450c21 was observed throughout the cortex. Densitometry was performed on the sections to compare the P450c17:3?-HSD ratio (Fig. 3B). In the ZG, 3?-HSD expression was higher than P450c17 (P < 0.05). P450c17 was higher in the outer and inner ZF than the ZG, although expression decreased in the mid-ZF. 3?-HSD stained at similar levels throughout the entire ZF, up to the cortico-medullary junction, with no decrease in the innermost adrenocortical area.
FIG. 3. A, IHC staining for steroidogenic enzymes in six adult male marmosets, aged 24–72 months. Cap (adrenal capsule) ZFo (outer ZF), ZFm (mid-ZF), ZFi (inner ZF). Red or brown color indicates positive stain. B, Densitometry analysis to determine relative P450c17:3?-HSD ratios. Open bars represent P450c17, and black bars represent 3?-HSD staining. *, Significant difference between enzyme staining in a specific zone and P450c17 staining in the ZG (P < 0.05).
Western analysis of adrenal microsomes
The immunoblot of marmoset adrenal microsomes confirmed the specificity of the cytb5 antibody, and protein expression levels of cytb5 were consistent with the staining trends observed by IHC (Fig. 4). Indeed, marmosets expressed P450c17 protein at similar levels to rhesus controls, but expression of cytb5 protein was barely detectable compared with that observed in rhesus samples.
FIG. 4. Immunoblot of monkey protein probed for P450c17 and cytb5 from six marmoset males, aged 60–132 months, one rhesus male, aged 3 months and one rhesus female, aged 4 months. Recombinant human cytb5 (C), rhesus (Rh). Molecular markers are shown at left.
Neonatal adrenal IHC
In view of the poor evidence for a functional ZR in adult male marmosets but previous evidence for DHEA and DS production in neonates (9), we evaluated the neonatal adrenal by IHC. As seen in the adult marmosets and other species, P450c17 was undetectable in the developing ZG in the definitive zone (DZ; Fig 5). 3?-HSD stained throughout the DZ, but expression decreased dramatically in the area expected of a FZ. The drop in 3?-HSD expression also correlated with a marked increase in cytb5, and to some extent reductase, staining in the putative FZ.
FIG. 5. IHC staining for steroidogenic enzymes in five male and one female neonatal marmosets, aged 1 d. Red or brown color indicates positive stain. Cap, Adrenal capsule; "ZG", ZG of the DZ/neocortex; "ZF", ZF of the DZ/neocortex.
P450c17 substrate preference and relative -4-lyase deficiency
No information has been published to date concerning the substrate preference of marmoset P450c17, or the presence or absence of -4-lyase deficiency. Both marmoset and rhesus whole microsomal preparations showed no detectable production of progesterone, androstenedione or deoxycorticosterone, but showed substantial metabolism of pregnenolone to 17-hydroxy-pregnenolone and 11-deoxycortisol (Fig 6, A and B). The additional production of DHEA (32.49 ± 4.76% of total product) observed in male rhesus monkey preparations (Fig. 6B) was not observed in the adult male marmoset preparations (Fig. 6A). To determine more definitively whether the lack of DHEA production in adult male marmoset microsomes was a consequence of more rapid metabolism of 17-hydroxy-pregnenolone intermediate by 3?-HSD, experiments were replicated with the addition of trilostane, a 3?-HSD inhibitor. In the presence of trilostane (Fig. 6A), marmoset microsomal production of 11-deoxycortisol was reduced as expected, but this reduction was accompanied by an increase in 17-hydroxy-pregnenolone rather than an increase in DHEA.
FIG. 6. Radio-HPLC detected metabolites of pregnenolone substrate addition to whole adrenal microsomal preparations (marmosets-see Fig. 4; six rhesus males, aged 3–96 months). 17OHP5 (17-hydroxy-pregnenolone), P4 (progesterone), 17OHP4 (17-hydroxy-progesterone), A4 (androstenedione), 11-deoxyF (11-deoxycortisol), T (trilostane). Bars are expressed as group means ± SEM (A, marmoset) and (B, rhesus). Open bars represent the absence and black bars represent the presence of trilostane. C, DHEA/11-deoxycortisol ratios calculated from A and B. Ratios are shown without (–) or with (+) trilostane. Open bars represent marmoset and black bars represent rhesus ratios. *, Significant difference upon trilostane treatment (P < 0.05). D, Comparative 17-hydroxylase activities in marmoset and rhesus microsomes. Activities are shown without (–) or with (+) trilostane. Open bars represent marmoset and black bars represent rhesus activity. *, Significant difference upon trilostane treatment (P < 0.001).
Results determined from the net conversion of pregnenolone to 17-hydroxy-pregnenolone, 11-deoxycortisol, and DHEA showed 17-hydroxylase activities to be similar for rhesus and marmoset preparations (160.77 ± 6.45 nmol/mg·h vs. 144.02 ± 5.84 nmol/mg·h, respectively; P = 0.0673; Fig. 6D). In the presence of trilostane, apparent 17-hydroxylase activity was increased in both rhesus and marmoset microsomes (pregnenolone to 17-hydroxy-pregnenolone and DHEA, Fig. 6D).
A relative -4-lyase deficiency can be inferred in both marmosets and rhesus from the extensive conversion of pregnenolone to 11-deoxycortisol without any apparent androstenedione production (Fig. 6, A and B) but, in view of the importance of this metabolic step in determining the control of C19 steroid production, we sought to confirm it independently. Addition of 17-hydroxy-progesterone to adrenal microsomes did not yield any androstenedione production in either marmoset or rhesus preparations (Table 1).
TABLE 1. Products of whole adrenal microsomal conversion of 17OHP4 substrate as detected by radio-HPLC
Discussion
We have confirmed in this study that circulating levels of DHEA and DS in adult male common marmosets are much lower than in humans (9), suggesting either very poor lyase activity relative to hydroxylase in the adrenal cortex, or the lack of a significant ZR. In contrast to such differences between marmosets and humans, marmosets share with humans many important features of adrenal steroidogenic pathways. Only certain higher mammals express P450c17 in the adrenal, and such enzyme expression is essential for cortisol biosynthesis in the ZF. Of these cortisol-secreting species, only a subset—nonhuman primates, cows, sheep, and goats—show the same deficiency in -4–17,20-lyase activity observed in humans. This latter point is important in generating efficient cortisol biosynthesis without excessive C19 steroid production by the ZF (10). In addition, the relative degree of activation of cortisol release appears similar in marmosets to that in Old World primates and humans, as reflected by the use of marmosets as models for human adrenocortical function, stress, and psychopathology (3, 20, 21, 22). Marmosets are, therefore, one of a limited number of species that may be appropriate as a model for steroidogenesis in the human adrenal, and may be unique in that the effective absence of a ZR permits more straightforward interpretation of experimental manipulation of adrenal function at the endocrinological and molecular levels.
Before that can be achieved, however, it was first necessary to establish definitively whether the marmoset has a functional ZR in the adrenal cortex because the presence of such a zone has implications for nonovarian sources of estrogen and alterations in ovarian function (10). It was also desirable to know more about the substrate preferences of 17-hydroxylase and whether there was any associated -4-lyase deficiency because these two factors are key determinants of the species-specific control of both cortisol and C19 steroid biosynthesis (10).
We performed our studies in adult male marmosets to avoid the modulating effects of the ovarian cycle and female social status on adrenocortical activity (5, 6, 7, 22). We confirmed that ACTH1–39 increased cortisol in a dose-dependent manner in males similar to findings reported for female marmosets (22). Additionally, endogenously low levels of DHEA and DS did not respond to ACTH1–39, which corresponded with the prior findings of Levine et al. (9), and dexamethasone suppression yielded no detectable change in plasma DHEA or DS levels. Our in vivo data appear consistent with the absence of any C19 steroid-secreting ZR that is responsive to ACTH. The many-fold higher levels of cortisol than corticosterone suggests that 17-hydroxylase in adult male marmosets prefers pregnenolone as a substrate, and that conversion to 17-hydroxy-pregnenolone is the initial step of cortisol biosynthesis, i.e. as in humans, but not cows (10). The synthesis of abundant DHEA and DS requires preferential action of 17-hydroxylase on pregnenolone (Fig 1). Thus, the lack of any measurable increase in DHEA or DS in adult marmosets is not due to a lack of substrate preference by 17-hydroxylase for pregnenolone.
In human and rhesus adrenals, another hallmark of a functional ZR is coexpression of P450c17 with an elevated level of cytb5, and a P450c17: 3?-HSD ratio high enough to favor C19 steroid production (10). Indeed, the onset of C19 production at adrenarche has been associated with a marked reduction in 3?-HSD expression in the innermost adrenocortical area (11, 12, 13). Individual variation is seen, such that sometimes 3?-HSD is completely undetectable in the ZR, but, more commonly, a dramatically reduced level of 3?-HSD is seen in the ZR relative to the ZF. In this study, we demonstrated that marmosets, similar to humans and rhesus, do not express P450c17 in the ZG, but show P450c17 expression throughout the ZF up to the cortico-medullary junction. There was, however, no apparent or significant decrease in 3?-HSD staining in the innermost adrenocortical region, in contrast to that previously noted in humans and rhesus monkeys (11, 13, 14, 23). Additionally, although cytb5 has been localized to the C19 steroid-secreting ZR in both humans (15, 24, 25) and rhesus monkeys (14), and is thought to enhance 17,20-lyase activity through allosteric interactions with P450c17 (26), cytb5 was largely undetectable throughout the marmoset adrenal cortex, specifically in the innermost region. Thus, our findings for 3?-HSD and cytb5 in adult male marmoset adrenals are consistent with the absence of any functional ZR in these animals.
Our marmoset and rhesus monkey adrenal microsome substrate metabolism experiments provide functional evidence that adult male marmosets do not possess a functional C19 steroid-secreting ZR. Marmoset adrenal microsomes exhibited abundant 17-hydroxylase activity, equal in magnitude to that of rhesus monkeys but, unlike the latter, showed no evidence of significant DHEA synthesis, even in the presence of trilostane. Although the use of microsomes both enriches the enzyme and simplifies interpretation of results due to removal of mitochondrial enzymes, microsomal studies are not without their limitations and it is important to consider whether the results might be artifactual. In general terms, the lack of DHEA synthesis observed in a marmoset microsomal preparation could be attributed to an inability to metabolize the given substrate, 17-hydroxylase activity deficiency, 17,20-lyase activity deficiency, or reduced expression of P450c17 protein. Although DHEA was not a detectable product, marmoset adrenal microsomal preparations were capable of metabolizing pregnenolone to 17-hydroxy-pregnenolone and 11-deoxycortisol, as well as 17-hydroxy-progesterone to 11-dexoycortisol, suggesting reduced 17,20-lyase activity relative to rhesus.
Some validation of our microsomal methodology is given by the use of a positive control. DHEA was synthesized by the rhesus monkey adrenal microsomal preparations from pregnenolone, showing that DHEA synthesis can be achieved with the experimental design and substrate concentrations used in this study. Upon addition of trilostane in rhesus microsomes, the DHEA:11-deoxycortisol ratio increased significantly (P < 0.05, Fig. 6C). There was a significant increase in 17-hydroxy-pregnenolone in marmoset and rhesus preparations, confirming inhibition of conversion to -4 steroid products in both species. The lack of a corresponding increase in DHEA synthesis in marmosets by this method suggests that metabolism via 17,20-lyase is already saturated (and thereby inefficient) and that inhibition of 3?-HSD alone will not guarantee a significant increase in DHEA output. This shows clearly that the lack or inhibition of 3?-HSD is not the single determining factor for efficient 17,20-lyase activity in the adrenal.
The finding that 17,20-lyase activity was detectable in marmoset adrenal microsomes by a direct radiometric assay, albeit at 10% of that in rhesus monkey adrenal microsomal preparations (data not shown), is most likely due to the increased sensitivity of the radio-metric 17,20-lyase activity assay compared with the product analysis of radio-HPLC. It is equally possible that the failure to detect DHEA resulted in part from its further metabolism to other products such as DS that would not have been recovered during the organic extraction process. Regardless, it is clear that the 17,20-lyase activity of marmoset adrenal microsomes is substantially lower than that of rhesus monkey adrenal microsomes, even though P450c17 protein expression was similar.
Adrenal -4-lyase activity was examined using 17-hydroxy-progesterone as the substrate for the marmoset and rhesus monkey adrenal microsomes. In neither case did adrenal microsomal preparations metabolize 17-hydroxy-progesterone to detectable levels of androstenedione. Although it is possible to interpret this finding as support for a -4-lyase deficiency, such as seen in bovine and humans (27), it is also possible that the affinity of 21-hydroxylase for 17-hydroxy-pregnenolone is much greater than that of 17,20-lyase, thereby resulting in the complete conversion of 17-hydroxy-progesterone to 11-deoxycortisol (10). It is noteworthy that in our HPLC analysis of steroid metabolites, 21-hydroxylase activity was more apparent in microsomes from marmosets than in those from rhesus monkeys (Fig. 6, A and B). Future cloning and functional expression of marmoset P450c17 cDNA, such as described for rhesus monkey P450c17 (28) will be necessary to determine conclusively whether marmosets are truly -4-lyase deficient or whether it is competition with 21-hydroxylase that results in a physiologic -4-lyase deficiency. Knowledge of which mechanism is responsible for the observed -4-lyase deficiency is necessary before the marmoset can be used as a model for human adrenal and ovarian steroidogenesis studies.
The notable differences in ZR function of marmosets compared with humans and rhesus beg the question of whether there may also be differences in fetal adrenal development. Our examination of neonatal marmoset adrenals showed clear evidence for a FZ, again correlating with the findings of Levine et al. (9) using histochemical staining, and Suzuki et al. (15) and Parker et al. (29) using immunospecific and mRNA-specific methods. We conclude that marmosets resemble humans, more so than does the rhesus macaque (30), in possessing a C19 steroid-secreting region at birth, suggesting similar intrauterine development, and presumably parturition, processes. Therefore, we further conclude that the differences observed in adrenal zonation between humans and marmosets during adulthood develop after birth.
The fact that IHC indicated regions of high cytb5 expression in the innermost area of the neonate adrenal at a time when significant DS production is known to occur also makes it tempting to speculate that cytb5 may be a key factor in determining, or permitting, 17,20-lyase activity in the marmoset. It could be argued, however, that the adult male marmoset adrenal also shows staining of 3?-HSD up to the corticomedullary junction, and the neonatal adrenal shows a drop in 3?-HSD in the putative FZ, otherwise rich in cytb5, suggesting that it is the presence of 3?-HSD in adult marmoset adrenals that is preventing 17,20-lyase from using its substrate. Nonetheless, microsomal activity assays performed in the presence of trilostane failed to stimulate lyase activity. Thus, the lack of apparent lyase activity in adult males does not appear to be solely a function of competing 3?-HSD. Further studies will be necessary to determine whether cytb5 is indeed a direct determinant of lyase activity or whether increases in its expression are permissive or associated with an increase in lyase activity through another mechanism such as P450c17 phosphorylation (31).
In conclusion, we have shown that adult male marmosets lack significant circulating DHEA and DS because they lack a functional zona reticularis. Despite this difference from humans and rhesus monkeys, the development of the marmoset in utero involves the development of a FZ, capable of C19 steroid biosynthesis, that regresses before adulthood (9). Preliminary analyses of steroid responses in vivo and of microsome preparations in vitro suggest that marmoset 17-hydroxylase prefers pregnenolone as a substrate and shows comparatively poor -4-lyase activity. The existence of a primate model with similar fetal adrenal development to humans, but lacking a clearly defined or functional ZR, may prove useful for a number of studies. Examples include studies of ZF function or of ovarian steroidogenesis without the added complication of ZR-derived steroids, and examination of factors (increases in cytb5 or decreases/inhibition of 3?-HSD) that may be able to induce a ZR phenotype in the ZF. Such studies may in turn illuminate as-yet poorly understood phenomena, including adrenarche and adrenopause, and poorly understood endocrinopathies that relate to control of 17-hydroxylase activity, factors controlling androgen production and factors associated with a lack of adrenal androgen production such as mutations associated with isolated 17,20-lyase deficiency, which may then give insight into possible clinical treatment.
Acknowledgments
We thank Brynn Hogan and Brian Horman for technical assistance, Dan Wittwer and Fritz Wegner in WPRC Assay Services for performing hormone assays, and Amy Usborne, DVM in Pathology Services at WPRC for providing marmoset adrenal tissue. We also thank Samantha Mapes, Jo Corbin, and Dr. Francisco Moran in Population Health and Reproduction at University of California-Davis for assistance in microsomal preparation and assays, and Dr. Alice Tarantal and staff at the California National Primate Research Center for help in collecting rhesus adrenal tissue.
References
Abbott DH, Barnett DK, Colman RJ, Yamamoto ME, Schultz-Darken NJ 2003 Aspects of common marmoset basic biology and life history important for biomedical research. Comp Med 53:339–350
Mansfield K 2003 Marmoset models commonly used in biomedical research. Comp Med 53:383–392
Dettling AC, Feldon J, Pryce CR 2002 Early deprivation and behavioral and physiological responses to social separation/novelty in the marmoset. Pharmacol Biochem Behav 73:259–269
Gould E, Tanapat P, McEwen BS, Flugge G, Fuchs E 1998 Proliferation of granule cell precursors in the dentate gyrus of adult monkeys is diminished by stress. Proc Natl Acad Sci USA 95:3168–3171
Saltzman W, Schultz-Darken NJ, Scheffler G, Wegner FH, Abbott DH 1994 Social and reproductive influences on plasma cortisol in female marmoset monkeys. Physiol Behav 56:801–810
Saltzman W, Schultz-Darken NJ, Wegner FH, Wittwer DJ, Abbott DH 1998 Suppression of cortisol levels in subordinate female marmosets: reproductive and social contributions. Horm Behav 33:58–74
Saltzman W, Prudom SL, Schultz-Darken NJ, Wittwer DJ, Abbott DH 2004 Social suppression of cortisol in female marmoset monkeys: role of circulating ACTH levels and glucocorticoid negative feedback. Psychoneuroendocrinology 29:141–161
Miraglia T, Moreira E 1969 The adrenal cortex of the marmoset (Callithrix jacchus). Acta Anat (Basel) 72:520–532
Levine J, Wolfe LG, Schiebinger RJ, Loriaux DL, Cutler Jr GB 1982 Rapid regression of fetal adrenal zone and absence of adrenal reticular zone in the marmoset. Endocrinology 111:1797–1802
Conley AJ, Bird IM 1997 The role of cytochrome P450 17 -hydroxylase and 3 ?-hydroxysteroid dehydrogenase in the integration of gonadal and adrenal steroidogenesis via the 5 and 4 pathways of steroidogenesis in mammals. Biol Reprod 56:789–799
Gell JS, Atkins B, Margraf L, Mason JI, Sasano H, Rainey WE, Carr BR 1996 Adrenarche is associated with decreased 3?-hydroxysteroid dehydrogenase expression in the adrenal reticularis. Endocr Res 22:723–728
Endoh A, Kristiansen SB, Casson PR, Buster JE, Hornsby PJ 1996 The zona reticularis is the site of biosynthesis of dehydroepiandroseterone and dehydroepiandrosterone sulfate in the adult human adrenal cortex resulting from its low expression of 3?-hydroxysteroid dehydrogenase. J Clin Endocrinol Metab 81:3558–3565
Gell JS, Carr BR, Sasano H, Atkins B, Margraf L, Mason JI, Rainey WE 1998 Adrenarche results from development of a 3?-hydroxysteroid dehydrogenase-deficient adrenal reticularis. J Clin Endocrinol Metab 83:3695–3701
Mapes S, Corbin CJ, Tarantal A, Conley A 1999 The primate adrenal zona reticularis is defined by expression of cytochrome b5, 17-hydroxylase/17,20-lyase cytochrome P450 (P450c17) and NADPH-cytochrome P450 reductase (reductase) but not 3?-hydroxysteroid dehydrogenase/5–4 isomerase (3?-HSD). J Clin Endocrinol Metab 84:3382–3385
Suzuki T, Sasano H, Takeyama J, Kaneko C, Freije WA, Carr BR, Rainey WE 2000 Developmental changes in steroidogenic enzymes in human postnatal adrenal cortex: immunohistochemical studies. Clin Endocrinol (Oxf) 53:739–747
Rainey WE, Carr BR, Sasano H, Suzuki T, Mason JI 2002 Dissecting human adrenal androgen production. Trends Endocrinol Metab 13:234–239
Hearn JP 1977 Restraining device for small monkeys. Lab Anim 11:261–262
Baker JV, Abbott DH, Saltzman W 1999 Social determinants of reproductive failure in male common marmosets housed with their natal family. Anim Behav 58:501–513
Moran FM, Ford JJ, Corbin CJ, Mapes SM, Njar VC, Brodie AM, Conley AJ 2002 Regulation of microsomal P450, redox partner proteins, and steroidogenesis in the developing testes of the neonatal pig. Endocrinology 143:3361–3369
Cilia J, Piper DC 1997 Marmoset conspecific confrontation: an ethologically-based model of anxiety. Pharmacol Biochem Behav 58:85–91
Johnson EO, Kamilaris TC, Carter CS, Calogero AE, Gold PW, Chrousos GP 1996 The biobehavioral consequences of psychogenic stress in a small, social primate (Callithrix jacchus jacchus). Biol Psychiatry 40:317–337
Saltzman W, Prudom SL, Schultz-Darken NJ, Abbott DH 2000 Reduced adrenocortical responsiveness to adrenocorticotropic hormone (ACTH) in socially subordinate female marmoset monkeys. Psychoneuroendocrinology 25:463–477
Conley AJ, Pattison JC, Bird IM 2004 Variations in adrenal androgen production among (nonhuman) primates. Semin Reprod Med 22:311–326
Yanase T, Sasano H, Yubisui T, Sakai Y, Takayanagi R, Nawata H 1998 Immunohistochemical study of cytochrome b5 in human adrenal gland and in adrenocortical adenomas from patients with Cushing’s syndrome. Endocr J 45:89–95
Narasaka T, Suzuki T, Moriya T, Sasano H 2001 Temporal and spatial distribution of corticosteroidogenic enzymes immunoreactivity in developing human adrenal. Mol Cell Endocrinol 174:111–120
Auchus RJ, Lee TC, Miller WL 1998 Cytochrome b5 augments the 17,20-lyase activity of human P450c17 without direct electron transfer. J Biol Chem 273:3158–3165
Fevold HR, Lorence MC, McCarthy JL, Trant JM, Kagimoto M, Waterman MR, Mason JI 1989 Rat P450 (17) from testis: characterization of a full-length cDNA encoding a unique steroid hydroxylase capable of catalyzing both 4- and 5-steroid 17,20-lyase reactions. Mol Endocrinol 3:968–975
Arlt W, Martens JW, Song M, Wang JT, Auchus RJ, Miller WL 2002 Molecular evolution of adrenarche: structural and functional analysis of p450c17 from four primate species. Endocrinology 143:4665–4672
Parker Jr CR, Faye-Peterson O, Stankovic AK, Mason JI, Grizzle WE 1995 Immunohistochemical evaluation of the cellular localization and ontogeny of 3?-hydroxysteroid dehydrogenase/ 5–4 isomerase in the human fetal adrenal gland. Endocr Res 21:69–80
Mapes S, Corbin CJ, Tarantal A, Parker CR, Conley A 2002 Cytochrome b5 (b5) in the developing rhesus adrenal. Endocr Res 28:623 (Abstract)
Pandey AV, Mellon SH, Miller WL 2003 Protein phosphatase 2A and phosphoprotein SET regulate androgen production by P450c17. J Biol Chem 278:2837–2844(J. Christina Pattison, Da)
Address all correspondence and requests for reprints to: Ian M. Bird, Ph.D, Perinatal Research Laboratories, 7E Meriter Hospital/Park, 202 South Park Street, Madison, Wisconsin 53715. E-mail: imbird@wisc.edu.
Abstract
Neonatal human males produce high levels of dehydroepiandrosterone (DHEA) and its sulfo-conjugated form (DS) that decline within a few months of birth, due to regression of the adrenal fetal zone (FZ). Adult male humans and rhesus monkeys produce C19 steroids in abundance from the adrenal zona reticularis (ZR). Male marmoset monkeys produce DS at birth, but unlike humans and rhesus monkeys, do not produce comparable amounts of DHEA and DS in adulthood. To determine whether male marmosets express a functional ZR in adulthood, we examined adult and neonatal male marmosets for the presence of a ZR and FZ, respectively. Exogenous ACTH failed to stimulate DHEA or DS in adults, and dexamethasone treatment failed to suppress DHEA and DS, although cortisol levels changed as expected. In steroidogenic tissues, the key proteins necessary to synthesize C19 steroids from pregnenolone are P450c17, 3?-hydroxysteroid dehydrogenase (3?-HSD), nicotinamide adenine dinucleotide phosphate (reduced) oxido-reductase cytochrome P450 (reductase), and cytochromeb5 (cytb5). Adult adrenal cross sections showed P450c17 and reductase protein expression throughout the cortex but showed no expected decrease in 3?-HSD and increase in cytb5 in the innermost region. Western analysis confirmed these data, demonstrating comparable P450c17 expression to rhesus monkeys, but not cytb5. HPLC analysis revealed similar 17-hydroxylase action on pregnenolone for adult marmoset and rhesus adrenal microsomes but greatly diminished 17,20-lyase activity in marmosets. Neonatal marmoset adrenals exhibited staining indicative of a putative FZ (with P450c17, reduced 3?-HSD and increased cytb5). We conclude that neonatal marmosets exhibit a C19 steroid-secreting FZ similar to humans, but adult males fail to acquire a functional ZR.
Introduction
THE COMMON MARMOSET (Callithrix jacchus) is a New World monkey that is widely used in biomedical research (1, 2, 3), including studies on stress (4) and the effects of behavioral subordination on adrenocortical function (5, 6, 7). Nonetheless, little is known of the structure or function of the fetal or adult marmoset adrenal and how it may differ from humans. Although adrenal morphology is relatively straightforward in other primate models, the histological zonation of the marmoset adrenal cortex has endured much controversy in the literature. Miraglia and Moreira (8) reported in 1969 that previous histological staining provided no conclusive evidence for the presence or absence of a zona reticularis (ZR) in either the male or the female marmoset adrenal: the putative ZR was indistinct and difficult to describe definitively. There was, however, clear morphological evidence for the presence of a zona glomerulosa (ZG) and a zona fasciculata (ZF). In their study, Miraglia and Moreira (8) interpreted morphological evidence in favor of a ZR, although it was difficult to visualize and staining techniques were nonimmunospecific. Levine et al. (9) also performed hematoxylin and eosin staining of marmoset adrenals and found evidence of a fetal zone (FZ) in neonates, but no expression of a ZR in adult males or females. The adults lacked morphologically distinct cell types between the mid-adrenal and the innermost adrenocortical region (the cells abutting the cortico-medullary junction). The finding of Levine et al. (9) was substantiated by an extensive measurement of circulating steroids in vivo. Both male and female marmosets had relatively high circulating levels of dehydroepiandrosterone (DHEA; males 1343 ± 266 ng/dl, females 3110 ± 1083 ng/dl; mean ± SEM), and the sulfoconjugate form DS at birth that declined within a few months, consistent with the regression of a functional FZ (9). In contrast to humans and other primates, such as rhesus monkeys, however, male marmosets do not produce large amounts of DHEA and DS upon reaching sexual maturity (9). Thus, the question remains: is the lack of DS production due to an endocrine suppression mechanism or to lack of a functional ZR in the adrenal cortex of adult male marmosets?
Since the aforementioned studies were performed, there has been considerable progress in our understanding of the enzymology of adrenal steroidogenesis and the factors that underlie normal ZR function. The adrenal cortex is known to be the source of mineralocorticoids, glucocorticoids, and, in both human and many nonhuman primates, C19 steroids. The ability to synthesize these three classes of steroids is achieved through discrete adrenal zonation (Ref.10 and Fig. 1). Mechanisms underlying zonation have been studied extensively in experimental models such as rat, mouse, and cow, and these have established that zonal function derives from the presence or, just as importantly, the absence of certain enzymes. Although these nonprimate models lack any significant ZR, studies in humans and nonhuman primates have demonstrated that the adult ZR is capable of producing C19 steroids, most specifically DHEA and DS, due to its high expression of 17-hydroxylase/17,20-lyase cytochrome P450 (P450c17) combined with low expression of competing 3?-hydroxysteroid dehydrogenase (3?-HSD) (Refs.11, 12, 13, 14, 15, 16 and Fig. 1). In addition, recent studies suggest that the ZR is also often rich in cytochromeb5 (cytb5) (14, 15, 16).
FIG. 1. Steroid biosynthesis in the adrenal cortex of higher mammals. Solid arrows, Predominant pathway in species expressing P450c17 in the adrenal gland. Dashed arrows, Steroid biosynthesis seen in species using the -4-lyase pathway.
Despite these advances in our understanding of ZR function, immunospecific staining for key steroidogenic enzymes in the adrenal, and correlative in vivo challenges of adrenocortical function, have not been performed on male marmosets. Our aim, therefore, was to localize the enzymes and cofactors influencing DHEA production to specific adrenal zones and to correlate those findings with results of functional manipulation of circulating levels of adrenal steroids. We found that although male marmosets exhibit a neonatal adrenal morphology and steroid production pattern that is developmentally similar to humans, they do not express a functional ZR upon maturation to adulthood. These unique properties make the male marmoset a highly unusual model that shares developmental similarities to humans, but in adulthood, is sufficiently different that it would be of considerable value in determining what factors may be necessary for normal human ZR development and function.
Materials and Methods
Materials
Unless otherwise stated, general reagents were obtained from Sigma-Aldrich (St. Louis, MO) or Fisher Scientific International (Pittsburgh, PA).
Animals
This research was conducted in accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals and the Animal Welfare Act and its subsequent amendments. All animal procedures were reviewed and approved by the Graduate School Animal Care and Use Committee of the University of Wisconsin-Madison. The National Primate Research Center at the University of Wisconsin-Madison (WPRC) is accredited by AAALAC as part of the UW-Madison Graduate School. Adult male marmosets, as detailed below, were housed indoors at the WPRC. They lived in aluminum and welded wire cages measuring 61 x 91 x 183 cm or 122 x 61 x 183 cm, or in a large room measuring 363 x 212 x 218 cm. Males were housed either singly, with an adult female pairmate or in a larger mixed-sex group, and had visual, olfactory, and auditory access to conspecifics in other cages. Lights were on from 0600–1800 h, and animals were fed at 1230–1330 h daily. Additional information on animal housing and husbandry has been published previously (6). Infant and adult rhesus monkey (Macaca mulatta) tissue was obtained, opportunistically, from healthy animals assigned as controls to other approved projects at the California National Primate Research Center.
Steroid responses in vivo
Adrenocortical responses to ACTH1–39 (ACTH1–39) were determined in four adult male marmosets, which were 38–62 months of age and weighed 365 ± 8 g (mean ± SEM) at the beginning of data collection. Four doses of human ACTH1–39 (0.1, 1, 10, and 100 μg/kg) and a saline control were administered to each male at intervals of at least 2 wk, with the order of treatments balanced across animals. Males were weighed and injected with dexamethasone (5 mg/kg, im; American Regent Laboratories, Shirley, NY) at 1700 h to suppress endogenous ACTH secretion. The following morning at 0900 h, each male was administered human ACTH1–39 or saline iv. Blood samples (0.2–0.6 ml) were collected immediately before, 60 min after, and 120 min after ACTH1–39 injection. In a separate experiment, adrenocortical responses to dexamethasone were determined in six adult male marmosets (25–54 months of age, 377 ± 11 g). Dexamethasone (5 mg/kg) was injected im at 1600 h, and blood samples (1.2 ml) were collected at 0900 h the day of and the day after dexamethasone treatment, within 3 min of initial disturbance to the animal. In both experiments, marmosets were briefly restrained in a marmoset restraint tube (17), whereas blood was collected from the femoral vein using a heparinized 1-cc syringe and 27-gauge needle. Samples were immediately placed on ice and centrifuged at 2000 rpm for 10 min, and the plasma fractions were aspirated and frozen at –20 C until assayed.
Hormone assays
All plasma hormone concentrations were determined by RIAs that had been validated for use with marmoset plasma at the WPRC Assay Services laboratories. The cortisol and testosterone assays were described previously (5, 18). Their intra- and interassay coefficients of variation (CVs) were 5.65% and 10.48%, respectively, for cortisol, and 2.2% and 5.5%, respectively, for testosterone. The remaining RIAs for corticosterone (antibody: Esoterix Endocrinology, Calabasas Hills, CA; tritiated trace: American Radiolabeled Chemicals, Inc., St. Louis, MO; reference preparation: Sigma-Aldrich), DHEA (antibody-coated tube RIA, TKDH2; DPC, Los Angeles, CA), DS (antibody-coated tube RIA, TKDS1; DPC) and aldosterone (antibody-coated-tube RIA, Active Aldosterone; Diagnostics Systems Laboratories, Inc., Webster, TX), produced the following intra- and interassay CVs: corticosterone, 2.2% and 9.0%; DHEA, 3.6% and 4.2%; DS, 1.7%, interassay CV not available for a single assay; and aldosterone, 2.3% and 5.4%. Corticosterone, testosterone, and DHEA RIAs were performed after celite chromatography.
The recovery of DS standards added to 50 μl of a marmoset plasma pool (n = 5) was 103.3 ± 2.2%. Serial dilution of a DS-spiked marmoset pool (50.00–0.39 μl; n = 8) gave a displacement curve parallel to that obtained with DS standards. The sensitivity of the assay was 5.0 μg/dl. The recovery of DHEA standards added to 100 μl of a marmoset plasma pool (n = 7) was 99.0 ± 2.4%. Serial dilution of a DHEA-spiked marmoset pool (100.00–1.56 μl; n = 7) gave a displacement curve parallel to that obtained with DHEA standards. The sensitivity of the assay was 0.5 ng/ml.
Immunohistochemistry (IHC)
Adrenals from adult male marmosets (aged 24–72 months, n = 6) were fixed in 10% formalin, processed routinely, and paraffin embedded. Adrenals from 1-d-old marmosets (n = 5 males, n = 1 female; 2 adrenals donated from the Laboratory of Behavioral Neurobiology, Swiss Federal Institute of Technology, and 4 from the WPRC) were fixed in 4% paraformaldehyde in 0.1 M sodium cacodylate buffer, processed routinely and embedded in paraffin wax. Tissues were then cut into 5-μm sections. All sections were deparaffinized and rehydrated in clearing solution and graded alcohols. Endogenous peroxidase activity was quenched by incubation in 30% hydrogen peroxide in 100% methanol for 30 min. Staining was detected using the Vectastain Elite ABC Kit (rabbit IgG) (Vector Laboratories, Inc., Burlingame, CA). Sections were subsequently stained using antiserum raised against human P450c17 (1:1000, chicken polyclonal, Dr. A. J. Conley, Department of Population Health and Reproduction, University of California-Davis), human placenta 3?-HSD (1:400, rabbit polyclonal, Dr. J. Ian Mason, Clinical Biochemistry, University of Edinburgh, Edinburgh, Scotland, UK), human cytb5 (1:3000, rabbit polyclonal, Dr. A. J. Conley), rat nicotinamide adenine dinucleotide phosphate (reduced) oxido-reductase cytochrome P450 (reductase; 1:3000, rabbit polyclonal, Dr. A. J. Conley), human side-chain-cleavage cytochrome P450 (P450scc; 1:1000, rabbit polyclonal, Dr. W. L. Miller, Department of Pediatrics, University of California-San Francisco), and human 21-hydroyxlase cytochrome P450 (P450c21; 1:3000, rabbit polyclonal, Dr. W. L. Miller). Normal nonimmune rabbit serum (1:1000, Vector Laboratories, Inc.) was used as a control. All sections were incubated with diluted normal goat serum (1:200) for 20 min at room temperature to block nonspecific binding. They were then incubated with primary antibody diluted in buffer at 4 C overnight. The sections were incubated in diluted secondary antibody (1:200, antirabbit, or 5 μg/ml, antichicken) and avidin-biotin-conjugated peroxidase for 30 min each at room temperature before exposure to 3-amino-9-ethylcarbazole (Vector Laboratories, Inc.) for 10 min also at room temperature. Nuclei were counterstained blue with Gill 2 hematoxylin (neat, 2 sec, rinsed 5 min in tap water) before sections were coverslipped with Paramount (Dako Corp., Carpinteria, CA). For photo-quality purposes, some sections were restained using 3,3'-diaminobenzidine (Vector Laboratories, Inc.) for 5 min at room temperature, counterstained as above, dehydrated, and coverslipped with permanent mounting media. Images were captured using bright-field illumination on a Zeiss Axioskop (Carl Zeiss, Oberkochen, Germany) fitted with a SPOT2 digital camera (through Zeiss), and digital-grayscale densitometry (Zeiss software, KS 300 version 3.0) was then used to quantify relative staining.
Microsomal preparation and Western immunoblotting
Whole frozen adrenals from six adult male marmosets (aged 60–132 months) and two infant rhesus monkeys (1 male and 1 female, aged 3 and 4 months, respectively) were homogenized directly in lysis buffer [0.1 M KPO4 (pH 7.4), 20% glycerol, 5 mM ?-mercaptoethanol, and 0.5 mM phenylmethylsulfonyl fluoride] at a ratio of 0.1 g/1 ml tissue to buffer. Cellular debris was removed by centrifugation at 1000 x g for 10 min. The supernatant was then spun down at 16,000 x g for 10 min under the same conditions. The premicrosomal pellet was stored at –80 C, and the supernatant was spun down at 100,000 x g for 60 min. The pellet was rehomogenized in dialysis buffer [lysis buffer with 1 mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate detergent] and stored at –80 C. Protein concentration of microsomal preparations was determined using a BCA Protein Assay Kit (Pierce Biotechnology, Inc., Rockford, IL). Ten (10) micrograms of total microsomal protein were size separated per lane on a 16% polyacrylamide gel (Bio-Rad Laboratories, Inc., Hercules, CA, 150 V constant, 1.5 h) and protein transferred to PVDF membrane (Bio-Rad Laboratories, Inc., 100 V, 1 h). Western blotting was then performed using the antihuman P450c17 and antihuman cytb5 antibodies (P450c17 1:10,000, chicken horseradish peroxidase-linked IgY secondary antibody 1:5000, cytb5 1:2000, donkey antirabbit horseradish peroxidase-linked IgG secondary antibody 1:10,000, detected by ECL (Amersham Biosciences, Piscataway, NJ) as previously described (19). Recombinant human cytb5 was used as a control.
P450c17 substrate preference and relative -4-lyase deficiency assay/analysis
Synthesis of DHEA by marmoset and rhesus P450c17 of adrenal origin and relative -4-lyase deficiency were examined by radio-HPLC analysis of substrate metabolism. Microsomal protein (20 μg) from the adrenals of six adult male marmosets (aged 60–132 months) and six juvenile and adult male rhesus monkeys (aged 3–96 months), prepared as above, were used in 0.5 ml assays. Pregnenolone (Steraloids, Newport, RI; [3H]-pregnenolone, PerkinElmer, Boston, MA) was used as the substrate to examine DHEA synthesis capabilities of marmoset and rhesus adrenal P450c17. Treatment with and without the 3?-HSD inhibitor trilostane (50 μM; Sanofi Research Division, Great Valley, PA) was performed for each subject. -4-Lyase deficiency was investigated using the same methods, but with 17-hydroxy-progesterone (Steraloids; 1, 2, 6, 7 [3H]-17-hydroxy-progesterone, American Radiolabeled Chemicals, St. Louis, MO) as the substrate; trilostane was not used in this study. For both studies, 10 μM substrate (1% hot, 99% cold) was combined with trilostane (when used) and evaporated to dryness under air. After evaporation, substrate (and trilostane) was resuspended in 460–467 μl assay buffer (50 mM KPO4 and 1 mM EDTA). Microsomal protein (20 μg) was added to the resuspension, and to maintain a constant supply of reducing equivalents (dihydronicotinamide adenine dinucleotide phosphate), a generating system consisting of 17 mM glucose-6-phosphate, 2 mM ?-nicotinamide adenine dinucleotide phosphate, 1 mM dihydronicotinamide adenine dinucleotide phosphate, and 1 U glucose-6-phosphate-dehydrogenase was also added. After incubation for 1 h in a 37 C water bath, 0.5 ml buffer was added. Extraction of steroids in the entire assay was performed with 10 ml methylene chloride. All samples were spun for 10 min at 2000 x g; the aqueous phase was then removed and the organic phase was poured into a clean tube and evaporated to dryness. All samples were resuspended in 1 ml HPLC-grade methanol and subjected to radio-HPLC. 17-Hydroxylase activity was determined from the sum of 17-hydroxylated metabolite (11-deoxycortisol, 17-hydroxy-pregnenolone, DHEA) radioactivity after conversion to amount of substrate metabolized.
Radio-HPLC
All solvents (including water) were of HPLC grade and all chemicals were of analytical grade. Radio-HPLC analysis of the samples was performed with an HP1100 liquid chromatographic system (Agilent Technologies, Menlo Park, CA) interfaced with a Radiomatic Model 500TR Flow Scintillation Analyzer (Packard Instrument Co., Meriden, CT). The HPLC system consisted of a solvent degasser, binary pump, autosampler, column heater, and a variable wavelength detector. Scintillation fluid (Ultima-flo, Packard Instrument Co.) flow rate was 3 ml/min. Samples (10 μl) were injected onto a 250 x 4.6 mm C18 HPLC column (BetaBasic-18, Keystone Scientific, Bellefonte, PA), then eluted with a mobile phase consisting of A (84% ammonium acetate buffer:8% MeOH:8% acetonitrile) and B (42% MeOH:42% acetonitrile:16% H2O). B was held at 20% for 5 min, increased from 20–95% by a linear ramp (1 ml/min) from 5–20 min, held at 95% from 20–40 min, reduced to 20% from 40–45 min, then reequilibrated at 20% from 45–50 min. Positive controls ([3H]pregnenolone and [3H]17-hydroxy-progesterone) were injected with each assay to verify the radiodetector response and retention time. When appropriate, methanol blanks were injected between samples to monitor carryover. Retention times and area-under-the-curve of the radioactive peaks were calculated using FLO-ONE software (Packard Instrument Co.).
Statistics
All values are expressed as mean ± SEM, unless otherwise stated. Dose-dependent effects of ACTH on cortisol were analyzed by one-way ANOVA followed by post hoc univariate F tests. Other in vivo steroid levels were compared by paired t tests. One-way ANOVA was performed on the densitometry of the adult male marmoset adrenal sections stained for P450c17 and 3?-HSD. Student’s t test was used to determine differences in whole adrenal microsomal substrate metabolism.
Results
Plasma steroids
Injection of ACTH1–39 in dexamethasone-suppressed adult male marmosets resulted in a dose-dependent rise in cortisol by 60 min and 120 min, as shown in Fig. 2A. The threshold dose for ACTH1–39 responsiveness was 1 μg/kg, and the maximal response was induced by 10 μg/kg at 60 min and by 100 μg/kg at 120 min. Pre-ACTH1–39 DS was present only at a level below the assay sensitivity threshold of 5 μg/dl (50 ng/ml), and no elevation into the detectable range occurred in response to 10 μg/kg of ACTH1–39 at 60 min post injection (Fig. 2B), the time of maximal cortisol response (Fig. 2A). DHEA was assayed in samples from marmosets treated with 100 μg/kg (Fig. 2B). DHEA concentrations approximated those at the sensitivity limit of the assay (0.5 ng/ml) before ACTH1–39 treatment and did not increase significantly in response to ACTH1–39. Plasma aldosterone and testosterone levels did not rise in response to ACTH1–39 (1.65-fold increase and 3.2-fold decrease, respectively; Fig. 2B), confirming the specificity of the ACTH1–39 response to the adrenals rather than the gonads. When baseline steroid levels were taken into account, however, aldosterone levels did exhibit a significant proportionate increase (P < 0.05), as would be expected using a supraphysiological dose of ACTH. Changes in circulating corticosterone levels mirrored those of cortisol, showing at least a 41-fold increase from below detectable levels (<0.35 ng/ml) in response to 10 μg/kg ACTH1–39 at 60 min. These latter findings were more consistent with a ZF rather than the ZG origin of corticosterone production and secretion because they were relatively similar to the degree of cortisol elevation, but not of aldosterone. The ratio of corticosterone:cortisol was 0.0173 indicating 1.73% leakage from the cortisol pathway.
FIG. 2. A, Plasma cortisol dose-response to ACTH1–39 at 0 (0900 h), 60 (1000 h), and 120 (1100 h) min post injection in four male marmosets, aged 38–62 months. Black circles represent 60-min values and open circles represent 120-min values. *, Significant difference between dose and saline control (P < 0.05); ^, significant difference between dose and previous dose (P < 0.05). B, Plasma steroid levels in response to 10 μg/kg or 100 μg/kg (DHEA only) ACTH1–39 at 0 (0900 h) and 60 (1000 h) min post injection in the same four animals as in panel A. Open bars, Pre-ACTH1–39 levels (after 5 mg/kg dexamethasome overnight treatment); black bars, ACTH1–39 treatment levels. Note log scale for steroid level. *, Significant difference from pre-ACTH1–39 level (P < 0.05); ^, significant difference in absolute response (P < 0.05). C, Plasma steroid levels in response to 5 mg/kg dexamethasone overnight treatment levels in response to 5 mg/kg dexamethasone overnight treatment (basal 0900 h, dexamethasone 0900 h) in six male marmosets, aged 25–54 months. Open bars represent basal, and black bars represent dexamethasone treatment levels. Note log scale for steroid level. *, Significant difference from basal level (P < 0.05).
The contrasting experiment using dexamethasone-only treatment to suppress pituitary ACTH revealed a corresponding reduction in plasma cortisol levels, as expected (Fig 2C). Consistent with the results of the ACTH1–39 challenge, DS and DHEA were again undetectable or near the limit of sensitivity, respectively. Plasma testosterone and aldosterone levels were not significantly affected by dexamethasone-only treatment, which is consistent with a gonadal origin of testosterone and with the fact that in the physiological range, ACTH is not a major regulator of aldosterone secretion. In addition, plasma corticosterone levels diminished along with cortisol, consistent with the ZF, and not the ZG, as the zone of origin for corticosterone biosynthesis.
IHC localization of steroidogenic enzymes in adult male marmosets
To compare adrenocortical enzyme expression patterns of marmosets to those of humans and rhesus, we stained fixed sections for the major steroidogenic enzymes that influence DHEA and DS synthesis. As expected, the adult male marmoset ZG was lacking in P450c17 expression, although staining was obvious throughout the rest of the adrenal cortex (Fig. 3A). Contrary to expression patterns in primates exhibiting a ZR, 3?-HSD stained throughout the adrenal cortex, including the cortico-medullary junction, and cytb5 staining was low or absent in the entire cortex, especially in the innermost region. Staining for cytochromes reductase, P450scc and P450c21 was observed throughout the cortex. Densitometry was performed on the sections to compare the P450c17:3?-HSD ratio (Fig. 3B). In the ZG, 3?-HSD expression was higher than P450c17 (P < 0.05). P450c17 was higher in the outer and inner ZF than the ZG, although expression decreased in the mid-ZF. 3?-HSD stained at similar levels throughout the entire ZF, up to the cortico-medullary junction, with no decrease in the innermost adrenocortical area.
FIG. 3. A, IHC staining for steroidogenic enzymes in six adult male marmosets, aged 24–72 months. Cap (adrenal capsule) ZFo (outer ZF), ZFm (mid-ZF), ZFi (inner ZF). Red or brown color indicates positive stain. B, Densitometry analysis to determine relative P450c17:3?-HSD ratios. Open bars represent P450c17, and black bars represent 3?-HSD staining. *, Significant difference between enzyme staining in a specific zone and P450c17 staining in the ZG (P < 0.05).
Western analysis of adrenal microsomes
The immunoblot of marmoset adrenal microsomes confirmed the specificity of the cytb5 antibody, and protein expression levels of cytb5 were consistent with the staining trends observed by IHC (Fig. 4). Indeed, marmosets expressed P450c17 protein at similar levels to rhesus controls, but expression of cytb5 protein was barely detectable compared with that observed in rhesus samples.
FIG. 4. Immunoblot of monkey protein probed for P450c17 and cytb5 from six marmoset males, aged 60–132 months, one rhesus male, aged 3 months and one rhesus female, aged 4 months. Recombinant human cytb5 (C), rhesus (Rh). Molecular markers are shown at left.
Neonatal adrenal IHC
In view of the poor evidence for a functional ZR in adult male marmosets but previous evidence for DHEA and DS production in neonates (9), we evaluated the neonatal adrenal by IHC. As seen in the adult marmosets and other species, P450c17 was undetectable in the developing ZG in the definitive zone (DZ; Fig 5). 3?-HSD stained throughout the DZ, but expression decreased dramatically in the area expected of a FZ. The drop in 3?-HSD expression also correlated with a marked increase in cytb5, and to some extent reductase, staining in the putative FZ.
FIG. 5. IHC staining for steroidogenic enzymes in five male and one female neonatal marmosets, aged 1 d. Red or brown color indicates positive stain. Cap, Adrenal capsule; "ZG", ZG of the DZ/neocortex; "ZF", ZF of the DZ/neocortex.
P450c17 substrate preference and relative -4-lyase deficiency
No information has been published to date concerning the substrate preference of marmoset P450c17, or the presence or absence of -4-lyase deficiency. Both marmoset and rhesus whole microsomal preparations showed no detectable production of progesterone, androstenedione or deoxycorticosterone, but showed substantial metabolism of pregnenolone to 17-hydroxy-pregnenolone and 11-deoxycortisol (Fig 6, A and B). The additional production of DHEA (32.49 ± 4.76% of total product) observed in male rhesus monkey preparations (Fig. 6B) was not observed in the adult male marmoset preparations (Fig. 6A). To determine more definitively whether the lack of DHEA production in adult male marmoset microsomes was a consequence of more rapid metabolism of 17-hydroxy-pregnenolone intermediate by 3?-HSD, experiments were replicated with the addition of trilostane, a 3?-HSD inhibitor. In the presence of trilostane (Fig. 6A), marmoset microsomal production of 11-deoxycortisol was reduced as expected, but this reduction was accompanied by an increase in 17-hydroxy-pregnenolone rather than an increase in DHEA.
FIG. 6. Radio-HPLC detected metabolites of pregnenolone substrate addition to whole adrenal microsomal preparations (marmosets-see Fig. 4; six rhesus males, aged 3–96 months). 17OHP5 (17-hydroxy-pregnenolone), P4 (progesterone), 17OHP4 (17-hydroxy-progesterone), A4 (androstenedione), 11-deoxyF (11-deoxycortisol), T (trilostane). Bars are expressed as group means ± SEM (A, marmoset) and (B, rhesus). Open bars represent the absence and black bars represent the presence of trilostane. C, DHEA/11-deoxycortisol ratios calculated from A and B. Ratios are shown without (–) or with (+) trilostane. Open bars represent marmoset and black bars represent rhesus ratios. *, Significant difference upon trilostane treatment (P < 0.05). D, Comparative 17-hydroxylase activities in marmoset and rhesus microsomes. Activities are shown without (–) or with (+) trilostane. Open bars represent marmoset and black bars represent rhesus activity. *, Significant difference upon trilostane treatment (P < 0.001).
Results determined from the net conversion of pregnenolone to 17-hydroxy-pregnenolone, 11-deoxycortisol, and DHEA showed 17-hydroxylase activities to be similar for rhesus and marmoset preparations (160.77 ± 6.45 nmol/mg·h vs. 144.02 ± 5.84 nmol/mg·h, respectively; P = 0.0673; Fig. 6D). In the presence of trilostane, apparent 17-hydroxylase activity was increased in both rhesus and marmoset microsomes (pregnenolone to 17-hydroxy-pregnenolone and DHEA, Fig. 6D).
A relative -4-lyase deficiency can be inferred in both marmosets and rhesus from the extensive conversion of pregnenolone to 11-deoxycortisol without any apparent androstenedione production (Fig. 6, A and B) but, in view of the importance of this metabolic step in determining the control of C19 steroid production, we sought to confirm it independently. Addition of 17-hydroxy-progesterone to adrenal microsomes did not yield any androstenedione production in either marmoset or rhesus preparations (Table 1).
TABLE 1. Products of whole adrenal microsomal conversion of 17OHP4 substrate as detected by radio-HPLC
Discussion
We have confirmed in this study that circulating levels of DHEA and DS in adult male common marmosets are much lower than in humans (9), suggesting either very poor lyase activity relative to hydroxylase in the adrenal cortex, or the lack of a significant ZR. In contrast to such differences between marmosets and humans, marmosets share with humans many important features of adrenal steroidogenic pathways. Only certain higher mammals express P450c17 in the adrenal, and such enzyme expression is essential for cortisol biosynthesis in the ZF. Of these cortisol-secreting species, only a subset—nonhuman primates, cows, sheep, and goats—show the same deficiency in -4–17,20-lyase activity observed in humans. This latter point is important in generating efficient cortisol biosynthesis without excessive C19 steroid production by the ZF (10). In addition, the relative degree of activation of cortisol release appears similar in marmosets to that in Old World primates and humans, as reflected by the use of marmosets as models for human adrenocortical function, stress, and psychopathology (3, 20, 21, 22). Marmosets are, therefore, one of a limited number of species that may be appropriate as a model for steroidogenesis in the human adrenal, and may be unique in that the effective absence of a ZR permits more straightforward interpretation of experimental manipulation of adrenal function at the endocrinological and molecular levels.
Before that can be achieved, however, it was first necessary to establish definitively whether the marmoset has a functional ZR in the adrenal cortex because the presence of such a zone has implications for nonovarian sources of estrogen and alterations in ovarian function (10). It was also desirable to know more about the substrate preferences of 17-hydroxylase and whether there was any associated -4-lyase deficiency because these two factors are key determinants of the species-specific control of both cortisol and C19 steroid biosynthesis (10).
We performed our studies in adult male marmosets to avoid the modulating effects of the ovarian cycle and female social status on adrenocortical activity (5, 6, 7, 22). We confirmed that ACTH1–39 increased cortisol in a dose-dependent manner in males similar to findings reported for female marmosets (22). Additionally, endogenously low levels of DHEA and DS did not respond to ACTH1–39, which corresponded with the prior findings of Levine et al. (9), and dexamethasone suppression yielded no detectable change in plasma DHEA or DS levels. Our in vivo data appear consistent with the absence of any C19 steroid-secreting ZR that is responsive to ACTH. The many-fold higher levels of cortisol than corticosterone suggests that 17-hydroxylase in adult male marmosets prefers pregnenolone as a substrate, and that conversion to 17-hydroxy-pregnenolone is the initial step of cortisol biosynthesis, i.e. as in humans, but not cows (10). The synthesis of abundant DHEA and DS requires preferential action of 17-hydroxylase on pregnenolone (Fig 1). Thus, the lack of any measurable increase in DHEA or DS in adult marmosets is not due to a lack of substrate preference by 17-hydroxylase for pregnenolone.
In human and rhesus adrenals, another hallmark of a functional ZR is coexpression of P450c17 with an elevated level of cytb5, and a P450c17: 3?-HSD ratio high enough to favor C19 steroid production (10). Indeed, the onset of C19 production at adrenarche has been associated with a marked reduction in 3?-HSD expression in the innermost adrenocortical area (11, 12, 13). Individual variation is seen, such that sometimes 3?-HSD is completely undetectable in the ZR, but, more commonly, a dramatically reduced level of 3?-HSD is seen in the ZR relative to the ZF. In this study, we demonstrated that marmosets, similar to humans and rhesus, do not express P450c17 in the ZG, but show P450c17 expression throughout the ZF up to the cortico-medullary junction. There was, however, no apparent or significant decrease in 3?-HSD staining in the innermost adrenocortical region, in contrast to that previously noted in humans and rhesus monkeys (11, 13, 14, 23). Additionally, although cytb5 has been localized to the C19 steroid-secreting ZR in both humans (15, 24, 25) and rhesus monkeys (14), and is thought to enhance 17,20-lyase activity through allosteric interactions with P450c17 (26), cytb5 was largely undetectable throughout the marmoset adrenal cortex, specifically in the innermost region. Thus, our findings for 3?-HSD and cytb5 in adult male marmoset adrenals are consistent with the absence of any functional ZR in these animals.
Our marmoset and rhesus monkey adrenal microsome substrate metabolism experiments provide functional evidence that adult male marmosets do not possess a functional C19 steroid-secreting ZR. Marmoset adrenal microsomes exhibited abundant 17-hydroxylase activity, equal in magnitude to that of rhesus monkeys but, unlike the latter, showed no evidence of significant DHEA synthesis, even in the presence of trilostane. Although the use of microsomes both enriches the enzyme and simplifies interpretation of results due to removal of mitochondrial enzymes, microsomal studies are not without their limitations and it is important to consider whether the results might be artifactual. In general terms, the lack of DHEA synthesis observed in a marmoset microsomal preparation could be attributed to an inability to metabolize the given substrate, 17-hydroxylase activity deficiency, 17,20-lyase activity deficiency, or reduced expression of P450c17 protein. Although DHEA was not a detectable product, marmoset adrenal microsomal preparations were capable of metabolizing pregnenolone to 17-hydroxy-pregnenolone and 11-deoxycortisol, as well as 17-hydroxy-progesterone to 11-dexoycortisol, suggesting reduced 17,20-lyase activity relative to rhesus.
Some validation of our microsomal methodology is given by the use of a positive control. DHEA was synthesized by the rhesus monkey adrenal microsomal preparations from pregnenolone, showing that DHEA synthesis can be achieved with the experimental design and substrate concentrations used in this study. Upon addition of trilostane in rhesus microsomes, the DHEA:11-deoxycortisol ratio increased significantly (P < 0.05, Fig. 6C). There was a significant increase in 17-hydroxy-pregnenolone in marmoset and rhesus preparations, confirming inhibition of conversion to -4 steroid products in both species. The lack of a corresponding increase in DHEA synthesis in marmosets by this method suggests that metabolism via 17,20-lyase is already saturated (and thereby inefficient) and that inhibition of 3?-HSD alone will not guarantee a significant increase in DHEA output. This shows clearly that the lack or inhibition of 3?-HSD is not the single determining factor for efficient 17,20-lyase activity in the adrenal.
The finding that 17,20-lyase activity was detectable in marmoset adrenal microsomes by a direct radiometric assay, albeit at 10% of that in rhesus monkey adrenal microsomal preparations (data not shown), is most likely due to the increased sensitivity of the radio-metric 17,20-lyase activity assay compared with the product analysis of radio-HPLC. It is equally possible that the failure to detect DHEA resulted in part from its further metabolism to other products such as DS that would not have been recovered during the organic extraction process. Regardless, it is clear that the 17,20-lyase activity of marmoset adrenal microsomes is substantially lower than that of rhesus monkey adrenal microsomes, even though P450c17 protein expression was similar.
Adrenal -4-lyase activity was examined using 17-hydroxy-progesterone as the substrate for the marmoset and rhesus monkey adrenal microsomes. In neither case did adrenal microsomal preparations metabolize 17-hydroxy-progesterone to detectable levels of androstenedione. Although it is possible to interpret this finding as support for a -4-lyase deficiency, such as seen in bovine and humans (27), it is also possible that the affinity of 21-hydroxylase for 17-hydroxy-pregnenolone is much greater than that of 17,20-lyase, thereby resulting in the complete conversion of 17-hydroxy-progesterone to 11-deoxycortisol (10). It is noteworthy that in our HPLC analysis of steroid metabolites, 21-hydroxylase activity was more apparent in microsomes from marmosets than in those from rhesus monkeys (Fig. 6, A and B). Future cloning and functional expression of marmoset P450c17 cDNA, such as described for rhesus monkey P450c17 (28) will be necessary to determine conclusively whether marmosets are truly -4-lyase deficient or whether it is competition with 21-hydroxylase that results in a physiologic -4-lyase deficiency. Knowledge of which mechanism is responsible for the observed -4-lyase deficiency is necessary before the marmoset can be used as a model for human adrenal and ovarian steroidogenesis studies.
The notable differences in ZR function of marmosets compared with humans and rhesus beg the question of whether there may also be differences in fetal adrenal development. Our examination of neonatal marmoset adrenals showed clear evidence for a FZ, again correlating with the findings of Levine et al. (9) using histochemical staining, and Suzuki et al. (15) and Parker et al. (29) using immunospecific and mRNA-specific methods. We conclude that marmosets resemble humans, more so than does the rhesus macaque (30), in possessing a C19 steroid-secreting region at birth, suggesting similar intrauterine development, and presumably parturition, processes. Therefore, we further conclude that the differences observed in adrenal zonation between humans and marmosets during adulthood develop after birth.
The fact that IHC indicated regions of high cytb5 expression in the innermost area of the neonate adrenal at a time when significant DS production is known to occur also makes it tempting to speculate that cytb5 may be a key factor in determining, or permitting, 17,20-lyase activity in the marmoset. It could be argued, however, that the adult male marmoset adrenal also shows staining of 3?-HSD up to the corticomedullary junction, and the neonatal adrenal shows a drop in 3?-HSD in the putative FZ, otherwise rich in cytb5, suggesting that it is the presence of 3?-HSD in adult marmoset adrenals that is preventing 17,20-lyase from using its substrate. Nonetheless, microsomal activity assays performed in the presence of trilostane failed to stimulate lyase activity. Thus, the lack of apparent lyase activity in adult males does not appear to be solely a function of competing 3?-HSD. Further studies will be necessary to determine whether cytb5 is indeed a direct determinant of lyase activity or whether increases in its expression are permissive or associated with an increase in lyase activity through another mechanism such as P450c17 phosphorylation (31).
In conclusion, we have shown that adult male marmosets lack significant circulating DHEA and DS because they lack a functional zona reticularis. Despite this difference from humans and rhesus monkeys, the development of the marmoset in utero involves the development of a FZ, capable of C19 steroid biosynthesis, that regresses before adulthood (9). Preliminary analyses of steroid responses in vivo and of microsome preparations in vitro suggest that marmoset 17-hydroxylase prefers pregnenolone as a substrate and shows comparatively poor -4-lyase activity. The existence of a primate model with similar fetal adrenal development to humans, but lacking a clearly defined or functional ZR, may prove useful for a number of studies. Examples include studies of ZF function or of ovarian steroidogenesis without the added complication of ZR-derived steroids, and examination of factors (increases in cytb5 or decreases/inhibition of 3?-HSD) that may be able to induce a ZR phenotype in the ZF. Such studies may in turn illuminate as-yet poorly understood phenomena, including adrenarche and adrenopause, and poorly understood endocrinopathies that relate to control of 17-hydroxylase activity, factors controlling androgen production and factors associated with a lack of adrenal androgen production such as mutations associated with isolated 17,20-lyase deficiency, which may then give insight into possible clinical treatment.
Acknowledgments
We thank Brynn Hogan and Brian Horman for technical assistance, Dan Wittwer and Fritz Wegner in WPRC Assay Services for performing hormone assays, and Amy Usborne, DVM in Pathology Services at WPRC for providing marmoset adrenal tissue. We also thank Samantha Mapes, Jo Corbin, and Dr. Francisco Moran in Population Health and Reproduction at University of California-Davis for assistance in microsomal preparation and assays, and Dr. Alice Tarantal and staff at the California National Primate Research Center for help in collecting rhesus adrenal tissue.
References
Abbott DH, Barnett DK, Colman RJ, Yamamoto ME, Schultz-Darken NJ 2003 Aspects of common marmoset basic biology and life history important for biomedical research. Comp Med 53:339–350
Mansfield K 2003 Marmoset models commonly used in biomedical research. Comp Med 53:383–392
Dettling AC, Feldon J, Pryce CR 2002 Early deprivation and behavioral and physiological responses to social separation/novelty in the marmoset. Pharmacol Biochem Behav 73:259–269
Gould E, Tanapat P, McEwen BS, Flugge G, Fuchs E 1998 Proliferation of granule cell precursors in the dentate gyrus of adult monkeys is diminished by stress. Proc Natl Acad Sci USA 95:3168–3171
Saltzman W, Schultz-Darken NJ, Scheffler G, Wegner FH, Abbott DH 1994 Social and reproductive influences on plasma cortisol in female marmoset monkeys. Physiol Behav 56:801–810
Saltzman W, Schultz-Darken NJ, Wegner FH, Wittwer DJ, Abbott DH 1998 Suppression of cortisol levels in subordinate female marmosets: reproductive and social contributions. Horm Behav 33:58–74
Saltzman W, Prudom SL, Schultz-Darken NJ, Wittwer DJ, Abbott DH 2004 Social suppression of cortisol in female marmoset monkeys: role of circulating ACTH levels and glucocorticoid negative feedback. Psychoneuroendocrinology 29:141–161
Miraglia T, Moreira E 1969 The adrenal cortex of the marmoset (Callithrix jacchus). Acta Anat (Basel) 72:520–532
Levine J, Wolfe LG, Schiebinger RJ, Loriaux DL, Cutler Jr GB 1982 Rapid regression of fetal adrenal zone and absence of adrenal reticular zone in the marmoset. Endocrinology 111:1797–1802
Conley AJ, Bird IM 1997 The role of cytochrome P450 17 -hydroxylase and 3 ?-hydroxysteroid dehydrogenase in the integration of gonadal and adrenal steroidogenesis via the 5 and 4 pathways of steroidogenesis in mammals. Biol Reprod 56:789–799
Gell JS, Atkins B, Margraf L, Mason JI, Sasano H, Rainey WE, Carr BR 1996 Adrenarche is associated with decreased 3?-hydroxysteroid dehydrogenase expression in the adrenal reticularis. Endocr Res 22:723–728
Endoh A, Kristiansen SB, Casson PR, Buster JE, Hornsby PJ 1996 The zona reticularis is the site of biosynthesis of dehydroepiandroseterone and dehydroepiandrosterone sulfate in the adult human adrenal cortex resulting from its low expression of 3?-hydroxysteroid dehydrogenase. J Clin Endocrinol Metab 81:3558–3565
Gell JS, Carr BR, Sasano H, Atkins B, Margraf L, Mason JI, Rainey WE 1998 Adrenarche results from development of a 3?-hydroxysteroid dehydrogenase-deficient adrenal reticularis. J Clin Endocrinol Metab 83:3695–3701
Mapes S, Corbin CJ, Tarantal A, Conley A 1999 The primate adrenal zona reticularis is defined by expression of cytochrome b5, 17-hydroxylase/17,20-lyase cytochrome P450 (P450c17) and NADPH-cytochrome P450 reductase (reductase) but not 3?-hydroxysteroid dehydrogenase/5–4 isomerase (3?-HSD). J Clin Endocrinol Metab 84:3382–3385
Suzuki T, Sasano H, Takeyama J, Kaneko C, Freije WA, Carr BR, Rainey WE 2000 Developmental changes in steroidogenic enzymes in human postnatal adrenal cortex: immunohistochemical studies. Clin Endocrinol (Oxf) 53:739–747
Rainey WE, Carr BR, Sasano H, Suzuki T, Mason JI 2002 Dissecting human adrenal androgen production. Trends Endocrinol Metab 13:234–239
Hearn JP 1977 Restraining device for small monkeys. Lab Anim 11:261–262
Baker JV, Abbott DH, Saltzman W 1999 Social determinants of reproductive failure in male common marmosets housed with their natal family. Anim Behav 58:501–513
Moran FM, Ford JJ, Corbin CJ, Mapes SM, Njar VC, Brodie AM, Conley AJ 2002 Regulation of microsomal P450, redox partner proteins, and steroidogenesis in the developing testes of the neonatal pig. Endocrinology 143:3361–3369
Cilia J, Piper DC 1997 Marmoset conspecific confrontation: an ethologically-based model of anxiety. Pharmacol Biochem Behav 58:85–91
Johnson EO, Kamilaris TC, Carter CS, Calogero AE, Gold PW, Chrousos GP 1996 The biobehavioral consequences of psychogenic stress in a small, social primate (Callithrix jacchus jacchus). Biol Psychiatry 40:317–337
Saltzman W, Prudom SL, Schultz-Darken NJ, Abbott DH 2000 Reduced adrenocortical responsiveness to adrenocorticotropic hormone (ACTH) in socially subordinate female marmoset monkeys. Psychoneuroendocrinology 25:463–477
Conley AJ, Pattison JC, Bird IM 2004 Variations in adrenal androgen production among (nonhuman) primates. Semin Reprod Med 22:311–326
Yanase T, Sasano H, Yubisui T, Sakai Y, Takayanagi R, Nawata H 1998 Immunohistochemical study of cytochrome b5 in human adrenal gland and in adrenocortical adenomas from patients with Cushing’s syndrome. Endocr J 45:89–95
Narasaka T, Suzuki T, Moriya T, Sasano H 2001 Temporal and spatial distribution of corticosteroidogenic enzymes immunoreactivity in developing human adrenal. Mol Cell Endocrinol 174:111–120
Auchus RJ, Lee TC, Miller WL 1998 Cytochrome b5 augments the 17,20-lyase activity of human P450c17 without direct electron transfer. J Biol Chem 273:3158–3165
Fevold HR, Lorence MC, McCarthy JL, Trant JM, Kagimoto M, Waterman MR, Mason JI 1989 Rat P450 (17) from testis: characterization of a full-length cDNA encoding a unique steroid hydroxylase capable of catalyzing both 4- and 5-steroid 17,20-lyase reactions. Mol Endocrinol 3:968–975
Arlt W, Martens JW, Song M, Wang JT, Auchus RJ, Miller WL 2002 Molecular evolution of adrenarche: structural and functional analysis of p450c17 from four primate species. Endocrinology 143:4665–4672
Parker Jr CR, Faye-Peterson O, Stankovic AK, Mason JI, Grizzle WE 1995 Immunohistochemical evaluation of the cellular localization and ontogeny of 3?-hydroxysteroid dehydrogenase/ 5–4 isomerase in the human fetal adrenal gland. Endocr Res 21:69–80
Mapes S, Corbin CJ, Tarantal A, Parker CR, Conley A 2002 Cytochrome b5 (b5) in the developing rhesus adrenal. Endocr Res 28:623 (Abstract)
Pandey AV, Mellon SH, Miller WL 2003 Protein phosphatase 2A and phosphoprotein SET regulate androgen production by P450c17. J Biol Chem 278:2837–2844(J. Christina Pattison, Da)