Abstract
Two distinct nuclear androgen receptors (ARs) were identified in brain and ovarian tissues of kelp bass, Paralabrax clathratus, termed kbAR1 and kbAR2, which correspond to the two nuclear ARs we have previously characterized in Atlantic croaker, Micropogonias undulatus, termed acAR1 and acAR2. Scatchard analysis of nuclear fractions of whole brain tissue demonstrated that kbAR1 had a single class of high-affinity binding sites for testosterone (T; Kd of 1.8 nM and Bmax of 1.0 pmol/g tissue), whereas cytosolic fractions of kbAR2 ovarian tissue had a single class of high-affinity binding sites for dihydrotestosterone (DHT; Kd of 0.1 nM and Bmax of 0.5 pmol/g tissue). Competition studies showed that both kbAR1 and kbAR2 were specific for androgens. However, kbAR1 bound only T with high affinity, whereas kbAR2 bound DHT, mibolerone, 17α-methyl-testosterone, T, and 11-ketotestosterone with high affinity. In addition, we examined the binding affinities of dichlorodiphenyltrichloroethane and its derivatives, several hydroxylated polychlorinated biphenyl (PCB) congeners, PCB mixtures, and the fungicide vinclozolin and its two metabolites M1 and M2 for the two ARs in Atlantic croaker ovarian, testicular, and brain tissues and in kelp bass ovarian and brain tissues. Only 4,4′-PCB-3-OH and 2′,5′-PCB-3-OH demonstrated greater than 50% displacement of [3H]testosterone from either acAR1 or kbAR1. In contrast, with the exception of vinclozolin, all of the xenobiotics examined demonstrated binding to acAR2 in testicular and ovarian tissues. The binding affinities were highest in the testicular tissue with M2, 2,2′5′-PCB-4-OH, and o,p′-DDD all binding with EC50s less than 10 μM. The binding affinities of xenobiotics to kbAR2 in ovarian tissue were similar to their binding affinities for ovarian acAR2. The finding that AR1 and AR2 possess different binding affinities for natural androgens and synthetic steroids, as well as for xenobiotics, suggests that the activities of androgens and of certain xenobiotics will depend upon the type of AR present within the target tissue.
Introduction
Recently, we characterized two distinct nuclear androgen receptors (ARs) in the teleost fish, Atlantic croaker (Micropogonias undulatus) [1]; we have designated the AR found in the brain AR1, and the second AR, found in multiple tissues, AR2. These two ARs possess different tissue distributions and distinct steroid-binding specificities and physical characteristics, which suggests that they may mediate different actions of androgens. However, the precise cellular and tissue distributions and physiological functions of AR1 and AR2 are currently not known. The biochemical characteristics of AR2 are quite similar to those of mammalian ARs [2], whereas the binding characteristics of AR1 are heretofore unique to teleosts. The identification of two different ARs in Atlantic croaker is not unexpected, since multiplicity within the superfamily of nuclear steroid receptors is widespread. Multiple subtypes of the nuclear estrogen receptor (ER) have been identified in mammals [3], as well as multiple isoforms of the AR in amphibians and mammals [4, 5], and the progesterone receptor (PR) [6] and glucocorticoid [7] and mineralocorticoid receptors [8] in mammals. These studies suggest, therefore, that multiple isoforms and subtypes are necessary components of the pleiotropic responses of nuclear steroid receptors.
ARs have been biochemically characterized in only a few species of fish, and cDNA sequences have recently been published for an AR in Japanese eel (Anguilla japonica) testes [9], as well as for two isoforms, ARα and ARβ, in rainbow trout (Oncorhynchus mykiss) testes [10]. The fish ARs, first described in brown trout (Salmo trutta) skin [11] and subsequently in goldfish (Carassius auratus) brain tissue [12] and rainbow trout (Oncorhynchus mykiss) lymphocytes [13], have high affinity for testosterone (T) and low affinity for dihydrotestosterone (DHT) and the important fish androgen, 11-ketotestosterone (11-KT), and correspond closely to AR1 in Atlantic croaker. The AR partially characterized in coho salmon (Oncorhynchus kisutch) ovaries demonstrates higher affinity for DHT and 11-KT than for T [14], and has characteristics similar to AR2 in Atlantic croaker. The transactivation studies with the eel AR cDNA demonstrate that, in vitro, DHT, 11-KT, and T all induce high transcriptional activity [9], indicating that this receptor may have binding characteristics similar to those of AR2. Whether the biochemical and binding-specificity characteristics of ARα and ARβ [10] correspond to AR1 and AR2 is not yet known. Together, these data suggest that multiple ARs may be present throughout teleosts; however, their coexistence has been demonstrated in only two species, Atlantic croaker [1] and rainbow trout [10].
Extensive evidence for xenobiotic-induced endocrine disruption has been obtained with a wide variety of compounds in laboratory studies and in natural populations of fish and wildlife [15–17]. These studies have focused primarily upon the ER as the mechanism through which the endocrine system may be compromised; however, it is now apparent that chemicals can disrupt reproductive endocrine function through a variety of mechanisms at several sites along the hypothalamus-pituitary-gonadal axis [18–20]. For example, xenobiotic chemicals have been shown to interact with other steroid receptors, both membrane receptors [20] and nuclear receptors such as the PR [21] and the AR [22–25]. Although there is mounting evidence of impaired or altered sexual differentiation and reproductive function in natural fish populations inhabiting contaminated environments [26], there have been few attempts to investigate the effects of xenobiotics on androgen function or binding [27] in teleosts, and there are no reports of xenobiotic binding to multiple teleostean ARs.
Several pesticides and industrial chemicals, as well as phytochemicals, have been shown to bind to ARs in mammals. The organochlorine pesticide dichlorodiphenyltrichloroethane (DDT) and its derivatives bind to the AR in rats [22, 24], through which they are capable of antagonizing the actions of DHT in vitro. In vivo studies in rats demonstrated that p,p′-DDE can act as a potent antiandrogen, causing reproductive impairment in the offspring of exposed females [23, 28]. In addition, several other pesticides including the isomers δ and γ of hexachlorocyclohexane, pentachlorophenol, dieldrin, and atrazine can bind to the AR [24]. Additionally, two metabolites of the fungicide vinclozolin, M1 and M2, which cause in vivo reproductive impairment consistent with antiandrogenic action, demonstrate affinity for the AR [23]. Several phytochemicals have also recently been identified as potential antiandrogens [29]. It is concluded from these studies that the AR is potentially an important site of action for endocrine-disrupting chemicals.
A comprehensive understanding of endocrine disruption mediated by xenobiotic binding to nuclear steroid receptors must include a consideration of multiple receptor subtypes, their differential distributions, and possible differences in their susceptibility to xenobiotic interference. Multiple estrogen and androgen receptors have been shown to possess differential steroid-binding affinities and tissue distributions [1, 30], suggesting variable tissue responses and tissue-specific sensitivity to xenobiotics. Moreover, the two ER subtypes and ERs in different tissues have different binding affinities for certain phytoestrogens and xenobiotics [31]. Thus, some receptors and target tissues may be more susceptible than others to endocrine disruption by a particular xenobiotic. One further complicating factor is that despite similar binding affinities, different receptor subtypes or isoforms may invoke different functional responses [6, 32]. Therefore, it is necessary to broaden the scope of endocrine toxicology studies to include consideration of all forms of hormone receptors and their affinity for xenobiotics.
In the present study, we confirm that the two nuclear ARs, acAR1 and acAR2, which we previously identified in Atlantic croaker brain and ovarian tissue, respectively [1], are also present in the brain and ovarian tissue of kelp bass, Paralabrax clathratus (termed kbAR1 and kbAR2). In addition, we examine the differential binding affinities of DDT and its derivatives, several hydroxylated polychlorinated biphenyl (OH-PCB) congeners, PCB mixtures, and the fungicide vinclozolin and its two metabolites M1 and M2 for the two ARs in both Atlantic croaker and kelp bass tissues.
Materials and Methods
Chemicals
[1,2,6,7-3H] Testosterone (92.0 Ci/mmol) and [1,2,4,5,6,7-3H] 5α-dihydrotestosterone (127.0 Ci/mmol) were purchased from New England Nuclear (Boston, MA) and stored at −20°C. The unlabeled steroids were purchased from either Steraloids, Inc. (Wilton, NH) or from Sigma Chemical Company (St. Louis, MO). Mibolerone was a gift from the Upjohn Laboratories (Kalamazoo, MI). O,p′-DDT, p,p′-DDT, o,p′-DDE, o,p′-DDD, and p,p′-DDD were obtained from Chem Services (Westchester, PA); p,p′-DDE was obtained from Supelco (Bellefonte, PA); Aroclor 1242 was provided by Applied Marine Science (Livermore, CA); and 4,4′-dichloro-3-biphenylol (4,4′-PCB-3-OH), 2′,5′-dichloro-3-biphenylol (2′,5′-PCB-3-OH), 2,2′,5′-trichloro-4-biphenylol (2,2′,5′-PCB-4-OH), 2′,3′,4′,5′-tetrachloro-4-biphenylol (2′,3′,4′,5′-PCB-4-OH), Aroclor 1254, and Aroclor 1260 were purchased from Ultra Scientific (Kingston, RI). All steroids and xenobiotics were stored in 95% ethanol at −20°C. Chemicals and salts used for making the buffers were purchased from Sigma and from Fisher Scientific (Pittsburgh, PA). The scintillation cocktail was a mixture of 4 L toluene, 16 g PPO (7,5-diphenyl-oxazole), and 0.4 g POPOP (1,4-bis[5-phenyl–2-oxazolyl]-benzene) and 400 ml methanol.
Animals and Tissue Sampling
Adult male and female Atlantic croaker, between 14 and 28 cm in length, were collected during the reproductive season in the fall by either gill net or trawl from the bays near Port Aransas, Texas. Fish were maintained in circular, recirculating tanks under constant photoperiod and temperature regimes and fed a commercial fish food diet daily. Fish were acclimated in the laboratory for at least 1 mo prior to any tissue collection. Fish were rapidly decapitated, and tissues were removed, placed on ice, and used immediately or frozen on dry ice and stored at −80°C for up to 1 yr. Forty mature, recrudesced male and female adult kelp bass were collected in June during their reproductive season by hook and line in the Southern California Bight within one-half mile of Dana Point, California. Immediately after capture, the fish were rapidly decapitated and the tissues removed and frozen on dry ice. The tissue samples were shipped on dry ice to the University of Texas Marine Science Institute, where they were stored at −80°C and analyzed within 6 mo.
Buffers
Assay buffers for AR1 measurement were H-1, the homogenization buffer (50 mM Tris-HCl, 1 mM EDTA, 12 mM monothioglycerol, 30% glycerol [v:v], pH 7.5 at 4°C); W, the washing buffer (10 mM Tris-HCl, 2 mM MgCl2, 2 mM monothioglycerol, 250 mM sucrose, 10% glycerol [v:v], pH 7.5 at 4°C); and E-1, the extraction buffer (H-1+0.7 M KCl). The assay buffer for AR2 measurement was H-2, the homogenization buffer (50 mM Tris-HCl, 10 mM sodium molybdate, 1 mM EDTA, 12 mM monothioglycerol, 10% glycerol [v:v], pH 7.4 at 4°C).
Preparation of Tissue for AR1 and AR2 Assays
Whole brain tissue, comprising the olfactory bulbs and everything posterior to the medulla oblongata including the pituitary, was prepared as described previously [1]. Briefly, brain tissue homogenates in H-1 buffer (1:10 w:v) were centrifuged at 2500 × g for 15 min, and the resulting supernatant (S1) was spun at 160 000 × g for 1 h. The cytosolic fraction (S1) was charcoal stripped to remove endogenous steroids. The initial 2500 × g pellet was washed three times in ice-cold W buffer. After the wash, the pellet was resuspended in E-1 buffer (1:10 initial tissue w:v) and incubated for 1 h with vortexing at 15-min intervals. The suspension was then centrifuged at 160 000 × g for 1 h and the supernatant provided the nuclear fraction. The tissue preparations were either used immediately or frozen at −80°C, where binding activity remained constant for at least 1 wk.
Ovarian tissue homogenates were prepared in H-2 buffer (1:10 w:v) as described previously [1]. All subsequent steps were identical to those for preparing brain cytosolic fractions. Testicular tissue homogenates were prepared similarly to those of the ovary, with the exception that the tissue was diluted in H-2 buffer at a higher concentration (1:5 w:v).
Saturation Kinetics and Scatchard Analysis
Except where noted, all tissue preparations were incubated at 4°C between 14 and 20 h for measurement of AR1 and between 8 and 14 h for AR2 measurement. The dissociation constant, Kd, and receptor concentration, Bmax, of AR1 and AR2 were estimated by Scatchard [33] plots using the program Deltagraph (DeltaPoint Inc., Monterey, CA). To measure total binding, 50 μl of [3H]T or [3H]DHT (0.1–6 nM), dried under N2 and redissolved in buffer, was added to duplicate tubes containing 250-μl aliquots of tissue. In parallel sets of tubes, 100-fold excess of either unlabeled T or DHT was added to determine nonspecific binding.
Competition Studies with Steroids and Xenobiotics
The unlabeled steroid competitors, dissolved in 95% ethanol, were pipetted into test tubes and dried under N2. The 3H ligand was added to the tissue extracts immediately prior to adding aliquots of the mixture to test tubes containing the steroid competitors. Various concentrations of unlabeled steroids were incubated with 4 nM [3H]T for 12–26 h for AR1 and with 1 nM [3H]DHT for 8–12 h for AR2. Competitive binding of unlabeled xenobiotics was determined by pipetting 1) 50 μl of 3H ligand in buffer, 2) xenobiotic competitor dissolved in 95% ethanol, and 3) 250 μl of tissue sample into the test tubes, resulting in a final concentration of ethanol of 1% for acAR1, acAR2, and kbAR1 and 2% for kbAR2. These were the highest concentrations of ethanol that did not disrupt receptor binding (data not shown). Samples were vortexed and incubated for 12–16 h for AR1 and 8–12 h for AR2. Competition curves of unlabeled T were run in each of the AR1 assays and of unlabeled DHT for the AR2 assays as standards, and an appropriate concentration of ethanol was added for xenobiotic competition assays. Various concentrations of unlabeled steroids and xenobiotics were incubated with 4 nM [3H]T for 12–26 h for AR1 and 1 nM [3H]DHT for 8–12 h for AR2. The relative binding affinities (RBAs) of steroids for AR1 and AR2 were expressed as a percentage of the maximum specific binding of T and DHT, respectively. The pH of buffers H-1 and E-1 were lowered to 7.4 prior to preparing the tissue for the competitive binding assays of vinclozolin and the two metabolites M1 and M2 because of the pH sensitivity of vinclozolin.
Separation of Bound from Free Steroid
Free steroid was separated from bound by the dextran-coated charcoal (DCC) method. An equal volume of DCC (50 mM Tris-HCl, 1 mM EDTA, 10% glycerol [v:v], 1% Norit-A charcoal [w:v], 0.1% dextran T-70 [w:v], pH 7.5 at 4°C) was added to the samples, and the mixture was incubated for 5 min prior to centrifugation at 3000 × g for 5 min at 4°C. For the separation of bound from free steroid in testicular tissue preparations, the 1% DCC was incubated for 10 min prior to centrifugation. The supernatants were decanted into 7-ml scintillation vials; 5 ml of the standard scintillation cocktail was added, and the radiation was measured. The radioactivity within each sample was determined by counting for 5 min in a Beckman LS 6000SC scintillation counter (Beckman Instruments, Fullerton, CA).
Results
Binding Characteristics of Kelp Bass ARs
Saturation analysis of kelp bass ARs
Saturation analysis of [3H]T binding to kbAR1 in the nuclear fractions of pooled, gonadally recrudesced, male and female whole brain homogenates demonstrated that there was high affinity and saturable binding (Fig. 1). A Scatchard plot (Fig. 1, inset) of the saturation data showed a single class of high-affinity ARs with a dissociation constant, Kd, of 1.8 nM and a receptor concentration, Bmax, of 1.0 pmol/g brain tissue. The cytosolic fraction demonstrated similar results with a Kd of 2.2 nM and a Bmax of 1.2 pmol/g brain tissue (results not shown).
![A saturation curve of [3H]T binding to kbAR1 in a nuclear fraction of pooled whole brain homogenates from male and female kelp bass. Specific binding (SB) was determined by subtracting nonspecific binding (NSB), 1 μM T plus [3H]T, from total binding (TB). Inset: Scatchard analysis of the SB](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/biolreprod/61/4/10.1095_biolreprod61.4.1152/1/m_bire-61-04-25-f01.jpeg?Expires=1709964080&Signature=Yvwgl1LAUpi5ONyAm4Kt0dMn5XrfjIxfi1J1r4nGJCAtInd8uP8u2dmkhHGvWK3emdTF8OYE-v-jvAM1gnayKnchP-FdObDZiEtaxhIuD9K4zTAg6TnihRaN3q5FSUcfuVxJzA~c5ywItUBXyTeAKIrmJoQXY4ivrje0rmH-C0ZgfrVl3s~fGIp8kffsDOaEbGiKZPJnfbmCPyx6luWuFspTT7NEXgpOIqKRRcEybuPGa7rrnxyAOUo52trlPj1wwAKgVydf3Ddf6jXFVPiKAR3kiGLwN4qyFqOaD1afjNZDQpCyMXZJIXJrtk0Sg2WOL7HgYh7SkDXkuib-0oLyDw__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
A saturation curve of [3H]T binding to kbAR1 in a nuclear fraction of pooled whole brain homogenates from male and female kelp bass. Specific binding (SB) was determined by subtracting nonspecific binding (NSB), 1 μM T plus [3H]T, from total binding (TB). Inset: Scatchard analysis of the SB
Saturation analysis of [3H]DHT binding to kbAR2 in the cytosolic fractions of gonadally recrudesced ovarian homogenates demonstrated that there was high affinity and saturable binding (Fig. 2). Scatchard plots (Fig. 2, inset) of the saturation data showed a single class of high-affinity ARs with a Kd of 0.1 nM and a Bmax of 0.5 pmol/g tissue. Binding of [3H]DHT in the nuclear fraction of ovarian tissue was not determined.
![A saturation curve of [3H]DHT binding to kbAR2 in a cytosolic fraction of ovarian homogenates from kelp bass. SB was determined by subtracting NSB, 100 nM DHT plus [3H]DHT, from TB. Inset: Scatchard analysis of the SB](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/biolreprod/61/4/10.1095_biolreprod61.4.1152/1/m_bire-61-04-25-f02.jpeg?Expires=1709964080&Signature=ep~MuY8zyJvK2OcN~~f86apKxw7QCwZp40gaI2xmV2e0fCFQ9pwk1WuEPGT5G2yaEmLD4wLqmDzhs8M7vir9vg1zpN9uMDA7ShgEkTf8Z~5FiRgJCJaskOH-aaIcKDbOhXtBFjbeiWNx89wd8iN04QW2yVVaxtEjfU3sBLNF-ST1aEmFlQXuxw6PB7UuaO1LDnnFNBMxxP073AU~jWAN79gSk71L8dkkryKANaVnJAKTLu5Pxb631GkGmL0cT59sjKGmqev06upF7Ojdy7yGLXhq7KVwXbh2EggTXQUOZ8LjeGxD6tQp0LhqQfYfvYaOH6j7gm-Sc-tHUqYJzLwGng__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
A saturation curve of [3H]DHT binding to kbAR2 in a cytosolic fraction of ovarian homogenates from kelp bass. SB was determined by subtracting NSB, 100 nM DHT plus [3H]DHT, from TB. Inset: Scatchard analysis of the SB
Steroid specificity of kelp bass ARs
The steroid competition curves were parallel, indicating competitive binding between the unlabeled steroid competitors and either 4 nM [3H]T for kbAR1 or 1 nM [3H]DHT for kbAR2 (Fig. 3, A and B). This allowed for the 50% displacement of the radiolabeled ligand (EC50) to be determined and the RBAs to be calculated (Table 1). The steroid competition studies demonstrated that kbAR1 was specific for androgens and showed high affinity for T, whereas all of the other steroids tested had much lower binding affinities (RBA < 10%). DHT and the synthetic androgen 17α-methyl-17β-hydroxy-4-androsten-3-one (MT) bound with one and two orders of magnitude lower affinity, respectively, whereas at 1 μM the synthetic androgen mibolerone caused only 40% displacement of T, which was equivalent to the binding of estradiol-17β (E2) and higher than that of the important fish androgen 11-KT (Fig. 3A, Table 1). The kbAR1 demonstrated low affinity for the progestogen 17,20β-P but no affinity for either androstenedione or the progestogen 20β-S at concentrations up to 1 μM (Fig. 3A, Table 1).
The concentrations of steriods causing 50% displacement (EC50) of [3H]T from nuclear AR1 and [3H]DHT from cystolicAR2 binding in kelp bass brain and ovarian tissues, respectively, and their relative binding affinities (RBA).
| Steriod | EC50 (nM)a,b | RBA (%)c | ||
|---|---|---|---|---|
| kbAR1 | kbAR2 | kbAR1 | kbAR2 | |
| DHT (5α-dihydrotestosterone) | 150 | 0.3 | 2 | 100 |
| T (testosterone) | 3.0 | 2.0 | 100 | 15 |
| 11-KT (17β-hydroxy-4-androstene- | L | 3.0 | 0 | 17 |
| 3,11-dione) | ||||
| Androstenedione | nd | 90 | 0 | 0.33 |
| Mibolerone (7α,17α-dimethyl-17β- | L | 0.3 | 0 | 100 |
| hydroxy-4-estren-3-one) | ||||
| MT (17α-methyl-17β-hydroxy-4- | 450 | 0.3 | 0.7 | 100 |
| androsten-3-one) | ||||
| E2 (17β-estradiol) | L | 300 | 0 | 0.1 |
| 20β-S (17,20β,21-trihydroxy-4- | nd | 500 | 0 | 0.06 |
| pregnen-3-one) | ||||
| 17,20β-P (17,20β-dihydroxy-4- | L | 150 | 0 | 0.2 |
| pregnen-3-one) | ||||
| P4 (progesterone) | — | 30 | — | 1 |
| Cortisol | — | nd | — | 0 |
| 21-P (deoxycorticosterone) | — | 50 | — | 0.6 |
| Steriod | EC50 (nM)a,b | RBA (%)c | ||
|---|---|---|---|---|
| kbAR1 | kbAR2 | kbAR1 | kbAR2 | |
| DHT (5α-dihydrotestosterone) | 150 | 0.3 | 2 | 100 |
| T (testosterone) | 3.0 | 2.0 | 100 | 15 |
| 11-KT (17β-hydroxy-4-androstene- | L | 3.0 | 0 | 17 |
| 3,11-dione) | ||||
| Androstenedione | nd | 90 | 0 | 0.33 |
| Mibolerone (7α,17α-dimethyl-17β- | L | 0.3 | 0 | 100 |
| hydroxy-4-estren-3-one) | ||||
| MT (17α-methyl-17β-hydroxy-4- | 450 | 0.3 | 0.7 | 100 |
| androsten-3-one) | ||||
| E2 (17β-estradiol) | L | 300 | 0 | 0.1 |
| 20β-S (17,20β,21-trihydroxy-4- | nd | 500 | 0 | 0.06 |
| pregnen-3-one) | ||||
| 17,20β-P (17,20β-dihydroxy-4- | L | 150 | 0 | 0.2 |
| pregnen-3-one) | ||||
| P4 (progesterone) | — | 30 | — | 1 |
| Cortisol | — | nd | — | 0 |
| 21-P (deoxycorticosterone) | — | 50 | — | 0.6 |
EC50 of each compound is calculated from competition curves (see Figs. 3 and 4) and is equal to the concentration of competitor necessary to displace 50% of either 4 nM [3H]T from AR1 or 1 nM [3H]DHT from AR2.
nd, no displacement of [3H]ligand at the maximum concentration of competitor; L, less than 50% displacement at highest concentration tested (1 μM); —, compound was not tested.
RBA is calculated from the ratio of the EC50s of the various competitors to the EC50s of either T or DHT; the RBA′s ofT and DHT were set at 100 for AR1 and AR2 binding, respectively.
The concentrations of steriods causing 50% displacement (EC50) of [3H]T from nuclear AR1 and [3H]DHT from cystolicAR2 binding in kelp bass brain and ovarian tissues, respectively, and their relative binding affinities (RBA).
| Steriod | EC50 (nM)a,b | RBA (%)c | ||
|---|---|---|---|---|
| kbAR1 | kbAR2 | kbAR1 | kbAR2 | |
| DHT (5α-dihydrotestosterone) | 150 | 0.3 | 2 | 100 |
| T (testosterone) | 3.0 | 2.0 | 100 | 15 |
| 11-KT (17β-hydroxy-4-androstene- | L | 3.0 | 0 | 17 |
| 3,11-dione) | ||||
| Androstenedione | nd | 90 | 0 | 0.33 |
| Mibolerone (7α,17α-dimethyl-17β- | L | 0.3 | 0 | 100 |
| hydroxy-4-estren-3-one) | ||||
| MT (17α-methyl-17β-hydroxy-4- | 450 | 0.3 | 0.7 | 100 |
| androsten-3-one) | ||||
| E2 (17β-estradiol) | L | 300 | 0 | 0.1 |
| 20β-S (17,20β,21-trihydroxy-4- | nd | 500 | 0 | 0.06 |
| pregnen-3-one) | ||||
| 17,20β-P (17,20β-dihydroxy-4- | L | 150 | 0 | 0.2 |
| pregnen-3-one) | ||||
| P4 (progesterone) | — | 30 | — | 1 |
| Cortisol | — | nd | — | 0 |
| 21-P (deoxycorticosterone) | — | 50 | — | 0.6 |
| Steriod | EC50 (nM)a,b | RBA (%)c | ||
|---|---|---|---|---|
| kbAR1 | kbAR2 | kbAR1 | kbAR2 | |
| DHT (5α-dihydrotestosterone) | 150 | 0.3 | 2 | 100 |
| T (testosterone) | 3.0 | 2.0 | 100 | 15 |
| 11-KT (17β-hydroxy-4-androstene- | L | 3.0 | 0 | 17 |
| 3,11-dione) | ||||
| Androstenedione | nd | 90 | 0 | 0.33 |
| Mibolerone (7α,17α-dimethyl-17β- | L | 0.3 | 0 | 100 |
| hydroxy-4-estren-3-one) | ||||
| MT (17α-methyl-17β-hydroxy-4- | 450 | 0.3 | 0.7 | 100 |
| androsten-3-one) | ||||
| E2 (17β-estradiol) | L | 300 | 0 | 0.1 |
| 20β-S (17,20β,21-trihydroxy-4- | nd | 500 | 0 | 0.06 |
| pregnen-3-one) | ||||
| 17,20β-P (17,20β-dihydroxy-4- | L | 150 | 0 | 0.2 |
| pregnen-3-one) | ||||
| P4 (progesterone) | — | 30 | — | 1 |
| Cortisol | — | nd | — | 0 |
| 21-P (deoxycorticosterone) | — | 50 | — | 0.6 |
EC50 of each compound is calculated from competition curves (see Figs. 3 and 4) and is equal to the concentration of competitor necessary to displace 50% of either 4 nM [3H]T from AR1 or 1 nM [3H]DHT from AR2.
nd, no displacement of [3H]ligand at the maximum concentration of competitor; L, less than 50% displacement at highest concentration tested (1 μM); —, compound was not tested.
RBA is calculated from the ratio of the EC50s of the various competitors to the EC50s of either T or DHT; the RBA′s ofT and DHT were set at 100 for AR1 and AR2 binding, respectively.
Competition curves of the binding of various natural and synthetic steroids to A) kbAR1 in nuclear extracts of pooled male and female brain tissue and B) to kbAR2 in ovarian cytosolic extracts of kelp bass. See Table 1 for key to steroid abbreviations
In contrast, kbAR2 showed higher affinity for a much broader range of androgens. DHT and the two synthetic androgens, mibolerone and MT, bound equally well to kbAR2 with RBAs of 100%, whereas T and 11-KT bound with slightly less affinity (Fig. 3B and Table 1). Androstenedione, with an RBA of 0.3%, was the only androgen tested that did not bind with high affinity. Non-androgenic steroids also bound to kbAR2, progesterone (P4) binding with moderate affinity followed by deoxycorticosterone (21-P), 17,20β-P, E2, and 20β-S (Table 1). Cortisol was the only steroid tested that did not bind to kbAR2 at concentrations up to 1 μM (Table 1).
Xenobiotic Binding to Kelp Bass and AtlanticCroaker ARs
Binding of DDT derivatives
The derivatives of DDT—o,p′- and p,p′-DDT, o,p′- and p,p′-DDE, and o,p′- and p,p′-DDD—demonstrated no affinity for acAR1 in nuclear fractions of whole brain homogenates with concentrations up to 500 μM (n = 3) (Table 2). Likewise, the DDT derivatives showed no affinity for kbAR1 in nuclear fractions of whole brain homogenates with concentrations up to 500 μM for p,p′-DDT and o,p′- and p,p′-DDE and 100 μM for o,p′-DDT and o,p′- and p,p′-DDD (n = 3) (Table 2).
Concentrations of xenobiotics causing 50% displacement (EC50) of [3H]T from nuclear AR1 and [3H]DHT from cytosolic AR2 in both Atlantic croaker and kelp bass brain and ovarian tissues, respectively.a
| Xenobiotic | Atlantic croaker EC50 (μm) | Kelp bass EC50(μM) | ||
|---|---|---|---|---|
| acARI | acAR2 | kbAR1 | kbAR2 | |
| o,p′-DDT | nd | 200 | nd | nd |
| p,p′-DDT | nd | L | nd | 350 |
| o,p′-DDE | nd | L | nd | 500 |
| p,p′-DDE | nd | 200 | nd | L |
| o,p′ -DDD | nd | 190 | nd | nd |
| p,p′-DDD | nd | L | nd | nd |
| 4,4′-PCB-3-OH | 1000 | 480 | — | — |
| 2′,5′-PCB-3-OH | L | 210 | 700 | 350 |
| 2,2′,5′-PCB-4-OH | L | 45 | nd | 100 |
| 2′,3′,4′,5′-PCB-4-OH | nd | 140 | — | — |
| Aroclor 1242 | — | — | nd | 420 |
| Aroclor 1254 | nd | 480 | nd | 260 |
| Aroclor 1260 | nd | L | — | — |
| Vinclozolin | nd | nd | — | — |
| M1 | nd | 100 | — | — |
| M2 | nd | 3.5 | — | — |
| Xenobiotic | Atlantic croaker EC50 (μm) | Kelp bass EC50(μM) | ||
|---|---|---|---|---|
| acARI | acAR2 | kbAR1 | kbAR2 | |
| o,p′-DDT | nd | 200 | nd | nd |
| p,p′-DDT | nd | L | nd | 350 |
| o,p′-DDE | nd | L | nd | 500 |
| p,p′-DDE | nd | 200 | nd | L |
| o,p′ -DDD | nd | 190 | nd | nd |
| p,p′-DDD | nd | L | nd | nd |
| 4,4′-PCB-3-OH | 1000 | 480 | — | — |
| 2′,5′-PCB-3-OH | L | 210 | 700 | 350 |
| 2,2′,5′-PCB-4-OH | L | 45 | nd | 100 |
| 2′,3′,4′,5′-PCB-4-OH | nd | 140 | — | — |
| Aroclor 1242 | — | — | nd | 420 |
| Aroclor 1254 | nd | 480 | nd | 260 |
| Aroclor 1260 | nd | L | — | — |
| Vinclozolin | nd | nd | — | — |
| M1 | nd | 100 | — | — |
| M2 | nd | 3.5 | — | — |
nd, No displacement of [3H]ligand at the maximum concentration of competitor (see text for concentration); L, less than 50% displacement; —, compound was not tested.
Concentrations of xenobiotics causing 50% displacement (EC50) of [3H]T from nuclear AR1 and [3H]DHT from cytosolic AR2 in both Atlantic croaker and kelp bass brain and ovarian tissues, respectively.a
| Xenobiotic | Atlantic croaker EC50 (μm) | Kelp bass EC50(μM) | ||
|---|---|---|---|---|
| acARI | acAR2 | kbAR1 | kbAR2 | |
| o,p′-DDT | nd | 200 | nd | nd |
| p,p′-DDT | nd | L | nd | 350 |
| o,p′-DDE | nd | L | nd | 500 |
| p,p′-DDE | nd | 200 | nd | L |
| o,p′ -DDD | nd | 190 | nd | nd |
| p,p′-DDD | nd | L | nd | nd |
| 4,4′-PCB-3-OH | 1000 | 480 | — | — |
| 2′,5′-PCB-3-OH | L | 210 | 700 | 350 |
| 2,2′,5′-PCB-4-OH | L | 45 | nd | 100 |
| 2′,3′,4′,5′-PCB-4-OH | nd | 140 | — | — |
| Aroclor 1242 | — | — | nd | 420 |
| Aroclor 1254 | nd | 480 | nd | 260 |
| Aroclor 1260 | nd | L | — | — |
| Vinclozolin | nd | nd | — | — |
| M1 | nd | 100 | — | — |
| M2 | nd | 3.5 | — | — |
| Xenobiotic | Atlantic croaker EC50 (μm) | Kelp bass EC50(μM) | ||
|---|---|---|---|---|
| acARI | acAR2 | kbAR1 | kbAR2 | |
| o,p′-DDT | nd | 200 | nd | nd |
| p,p′-DDT | nd | L | nd | 350 |
| o,p′-DDE | nd | L | nd | 500 |
| p,p′-DDE | nd | 200 | nd | L |
| o,p′ -DDD | nd | 190 | nd | nd |
| p,p′-DDD | nd | L | nd | nd |
| 4,4′-PCB-3-OH | 1000 | 480 | — | — |
| 2′,5′-PCB-3-OH | L | 210 | 700 | 350 |
| 2,2′,5′-PCB-4-OH | L | 45 | nd | 100 |
| 2′,3′,4′,5′-PCB-4-OH | nd | 140 | — | — |
| Aroclor 1242 | — | — | nd | 420 |
| Aroclor 1254 | nd | 480 | nd | 260 |
| Aroclor 1260 | nd | L | — | — |
| Vinclozolin | nd | nd | — | — |
| M1 | nd | 100 | — | — |
| M2 | nd | 3.5 | — | — |
nd, No displacement of [3H]ligand at the maximum concentration of competitor (see text for concentration); L, less than 50% displacement; —, compound was not tested.
Each of the DDT derivatives demonstrated parallel competition curves suggesting that the binding was competitive to acAR2 in cytosolic fractions of both testicular and ovarian tissues, with greater than 40% displacement of 1 nM [3H]DHT with concentrations up to 500 μM (Fig. 4). O,p′-DDD had the highest affinity for acAR2 in both the testicular and ovarian tissues with EC50s of 10.5 and 190 μM, respectively (Tables 2 and 3). In ovarian tissue, p,p′-DDE and o,p-DDT also inhibited greater than 50% of the [3H]DHT binding to acAR2; and with the exception of p,p′-DDE, the o,p′-isomers of DDT and DDD were more effective competitors than the p,p′-isomers (Table 2) in ovarian tissue, whereas in testicular tissue, all of the o,p′-isomers demonstrated higher affinity for acAR2 than the p,p′-isomers. All six isomers inhibited greater than 50% of the [3H]DHT binding to acAR2 in testicular tissue, with an order of magnitude higher affinity than that found in ovarian tissue (Table 3). We found that in ovarian tissue, but not the testicular tissue, p,p′-DDE from Chem Services had no affinity for acAR2, whereas the Supelco lot had similar affinities to acAR2 in the two tissues (data not shown). The six DDT derivatives demonstrated lower binding affinities for kbAR2 than for acAR2. At concentrations of 1 mM competitor, p,p′-DDT, with an EC50 of 350 μM, had the highest affinity for kbAR2, followed by o,p′-DDE (Table 2). Although an RBA could not be calculated for p,p′-DDE, there was nearly 50% displacement of 1 nM [3H]DHT with 1 mM competitor. The other DDT derivatives were tested at concentrations less than 500 μM and did not show any displacement of [3H]DHT from kbAR2.
Concentrations of xenobiotics causing 50% displacement (EC50) of [3H]DHT from cytosolic AR2 in Atlantic croaker testicular tissue.
| Xenobiotic | EC50a (μM) |
|---|---|
| o,p′-DDT | 12 |
| p,p′-DDT | 35 |
| o,p′-DDE | 16 |
| p,p′-DDE | 40 |
| o,p′-DDD | 10.5 |
| p,p′-DDD | 35 |
| 4,4′-PCB-3-OH | 20 |
| 2′,5′-PCB-3-OH | 20 |
| 2,2′,5′-PCB-4-OH | 10 |
| 2′,3′,4′,5′-PCB-4-OH | L |
| Aroclor 1254 | 30 |
| Aroclor 1260 | 500 |
| Vinclozolin | nd |
| M1 | 65 |
| M2 | 3.5 |
| Xenobiotic | EC50a (μM) |
|---|---|
| o,p′-DDT | 12 |
| p,p′-DDT | 35 |
| o,p′-DDE | 16 |
| p,p′-DDE | 40 |
| o,p′-DDD | 10.5 |
| p,p′-DDD | 35 |
| 4,4′-PCB-3-OH | 20 |
| 2′,5′-PCB-3-OH | 20 |
| 2,2′,5′-PCB-4-OH | 10 |
| 2′,3′,4′,5′-PCB-4-OH | L |
| Aroclor 1254 | 30 |
| Aroclor 1260 | 500 |
| Vinclozolin | nd |
| M1 | 65 |
| M2 | 3.5 |
nd, No displacement of [3H]ligand at the maximum concentration of competitor (see text for concentration); L, less than 50% displacement.
Concentrations of xenobiotics causing 50% displacement (EC50) of [3H]DHT from cytosolic AR2 in Atlantic croaker testicular tissue.
| Xenobiotic | EC50a (μM) |
|---|---|
| o,p′-DDT | 12 |
| p,p′-DDT | 35 |
| o,p′-DDE | 16 |
| p,p′-DDE | 40 |
| o,p′-DDD | 10.5 |
| p,p′-DDD | 35 |
| 4,4′-PCB-3-OH | 20 |
| 2′,5′-PCB-3-OH | 20 |
| 2,2′,5′-PCB-4-OH | 10 |
| 2′,3′,4′,5′-PCB-4-OH | L |
| Aroclor 1254 | 30 |
| Aroclor 1260 | 500 |
| Vinclozolin | nd |
| M1 | 65 |
| M2 | 3.5 |
| Xenobiotic | EC50a (μM) |
|---|---|
| o,p′-DDT | 12 |
| p,p′-DDT | 35 |
| o,p′-DDE | 16 |
| p,p′-DDE | 40 |
| o,p′-DDD | 10.5 |
| p,p′-DDD | 35 |
| 4,4′-PCB-3-OH | 20 |
| 2′,5′-PCB-3-OH | 20 |
| 2,2′,5′-PCB-4-OH | 10 |
| 2′,3′,4′,5′-PCB-4-OH | L |
| Aroclor 1254 | 30 |
| Aroclor 1260 | 500 |
| Vinclozolin | nd |
| M1 | 65 |
| M2 | 3.5 |
nd, No displacement of [3H]ligand at the maximum concentration of competitor (see text for concentration); L, less than 50% displacement.
Competition curves of the binding of various DDT derivatives to acAR2 in cytosolic extracts of A) ovarian and B) testicular tissue. Each data point is the average of 3–5 assays with SEM < 9%
Binding of PCBs
At concentrations of 1 mM, 4,4′-PCB-3-OH caused 50% displacement, and 2,2′,5′-PCB-4-OH and 2′,5′-PCB-3-OH caused 35% and 25% displacement, respectively, whereas 2′,3′,4′,5′-PCB-4-OH caused no displacement at concentrations of 1 mM of 4 nM [3H]T from acAR1 in nuclear fractions of whole brain homogenates (Fig. 5, Table 2). The 2′,3′,4′,5′-PCB-4-OH congener demonstrated over 300% binding at 1 mM in two of three assays (data not shown), suggesting that accurate measurements of binding for this compound could not be made under these assay conditions. Neither of the PCB mixtures tested, Aroclor 1254 or 1260, demonstrated binding to the acAR1 (Table 2). Of the two OH-PCB congeners tested, only 2′,5′-PCB-3-OH at concentrations of 1 mM demonstrated competitive binding to kbAR1 (Table 2). No displacement of [3H]T from kbAR1 was observed with the other hydroxylated congener 2,2′,5′-PCB-4-OH at 1 mM and the PCB mixtures Aroclor 1242 and 1254, at concentrations of 34 and 296 μM, respectively.
Competition curves of the binding of four OH-PCBs to acAR1 in pooled male and female brain nuclear extracts. Each data point is the average of 3–4 assays with SEM < 10%, except for 2′,3′,4′,5′-PCB-4-OH, which has an n = 1
At concentrations of up to 1 mM, each of the four PCB congeners tested—4,4′-PCB-3-OH, 2′,5′-PCB-3-OH, 2,2′,5′-PCB-4-OH, and 2′,3′,4′,5′-PCB-4-OH—demonstrated competitive displacement of [3H]DHT from acAR2 in cytosolic ovarian and testicular tissue preparations (Fig. 6). 2,2′,5′-PCB-4-OH had the highest affinity for acAR2 in both testicular and ovarian tissues, with EC50s of 10 and 45 μM, respectively. In ovarian tissues, the affinities of the other three congeners to acAR2 were one third to one tenth lower than that of 2,2′,5′-PCB-4-OH. In testicular tissue, both 4,4′-PCB-3-OH and 2′,5′-PCB-3-OH bound with half of the affinity of 2,2′,5′-PCB-4-OH, whereas at 50 μM, 2′,3′,4′,5′-PCB-4-OH demonstrated 35% inhibition of [3H]DHT binding to acAR2. This differed from the findings with the ovarian tissue, in which 2′,3′,4′,5′-PCB-4-OH bound to acAR2 with higher affinity than either 4,4′-PCB-3-OH or 2′,5′-PCB-3-OH. However, similar to the binding of 2′,3′,4′,5′-PCB-4-OH to acAR1 in brain tissue, at 1 mM competitor, 300% binding was observed (data not shown), which suggests that accurate measurements of binding for this compound could not be made using these assay conditions. The PCB mixture Aroclor 1254 demonstrated binding to acAR2 in both testicular and ovarian tissue with EC50s of 30 and 480 μM, respectively (Fig. 6, Tables 2 and 3). The more chlorinated PCB mixture Aroclor 1260, at 1 mM, showed only 25% displacement of [3H]DHT from acAR2 in ovarian tissue (Fig. 6), whereas in testicular tissue it demonstrated greater than 50% displacement at 1 mM with an EC50 of 500 μM. The results obtained in the kbAR2 competition assays were similar to the results of ovarian acAR2 binding, in which the 2,2′,5′-PCB-4-OH and the 2′,5′-PCB-3-OH congeners demonstrated competitive binding to kbAR2 at concentrations of up to 1 mM with EC50s of 100 μM and 350 μM, respectively (Table 2). The two PCB mixtures, Aroclor 1242 and 1254, also competitively displaced [3H]DHT from kbAR2 with EC50s of 420 μM and 260 μM, respectively (Table 2).
Competition curves of the binding of four OH-PCBs and two PCB mixtures to acAR2 in cytosolic extracts of A) ovarian and B) testicular tissue. Each data point is the average of 3 assays with SEM < 10%, except for 2′5′-PCB-3-OH at 1 mM, which had an SEM = 25% in testicular tissue
Binding of vinclozolin and its metabolites
The fungicide vinclozolin did not bind to either acAR1 in brain homogenates or acAR2 in ovarian and testicular homogenates (Fig. 7, Tables 2 and 3). The two primary vinclozolin metabolites, M1 and M2, bound to acAR2 in both testicular and ovarian tissues. M1 bound with higher affinity to testicular acAR2 than to ovarian acAR2; EC50s were 65 and 100 μM, respectively. On the other hand, the affinities of M2 for acAR2 in the two tissues were identical, with EC50s of 3.5 μM. Neither metabolite demonstrated any affinity for acAR1 at concentrations up to 500 μM. The binding of vinclozolin and its metabolites to either kbAR1 or kbAR2 was not examined.
Competition curves of the binding of vinclozolin and its metabolites to acAR2 in cytosolic extracts of A) ovarian and B) testicular tissue. Each data point is the average of 3 assays with SEM < 8%
Protein concentrations of ovarian and testicular tissue preparations
Protein concentrations within Atlantic croaker testicular homogenates were 0.9 ± 0.05 ng/ml compared to 2.9 ± 0.2 ng/ml in Atlantic croaker ovarian homogenates. The binding affinities of o,p′-DDD, 2,2′,5′-PCB-4-OH, and M2 to AR2 in diluted ovarian homogenates (1.1 ng/ml protein) were not different from those in undiluted ovarian homogenates (data not shown). BSA was used as the standard for the determination of protein concentration via the method of Bradford [34].
Discussion
The results of this study demonstrate that kelp bass possess multiple ARs that are equivalent to the ARs we have previously characterized in brain and ovarian tissues of the Atlantic croaker [1]. The kelp bass ARs were not fully characterized but were assayed under conditions previously established for measurement of Atlantic croaker ARs. A nuclear receptor was identified in brain tissue of kelp bass that had [3H]T-binding characteristics (Kd and Bmax) in both the nuclear and cytosolic components similar to those for [3H]T binding to AR1 in Atlantic croaker. A distinguishing characteristic of acAR1 is its rank order of steroid specificities: T >> DHT > MT >> 11-KT > mibolerone > androstenedione > E2 > 20β-S. A similar steroid-specificity profile was seen with kbAR1, in which T >> DHT > MT >> mibolerone = E2 > 11-KT. Although kbAR1 had higher affinity for E2 than did acAR1, the affinities for the six androgens tested were nearly identical, and both acAR1 and kbAR1 bound only T with high affinity. We also identified the receptor equivalent to acAR2 in the cytosolic fraction of ovarian tissue of kelp bass, which specifically bound [3H]DHT with a Kd and Bmax similar to those of acAR2. Nuclear binding of kbAR2 was not measured, but the more complete characterization of acAR2 demonstrated that this receptor is a nuclear steroid receptor [1]. The steroid-specificity profile of kbAR2 (DHT = mibolerone = MT > T = 11-KT >> androstenedione > E2 > 20β-S) was slightly different from that of acAR2 (DHT > T > mibolerone = MT > 11-KT > androstenedione > E2 > 20β-S). However, acAR2 is similar to kbAR2 in that both receptors demonstrate a broad specificity for androgens and bind both natural and synthetic androgens with high affinity.
The presence of multiple ARs has now been confirmed in representatives of two different families of advanced fishes belonging to the order Perciformes, Atlantic croaker (Sciaenidae) [1] and kelp bass (Serranidae). Based upon similarities in the steroid-specificity profiles and binding characteristics, AR1-like receptors have also been detected in goldfish (Cyprinidae) [12] and in the salmonids, brown trout [11] and rainbow trout [13]. T binds with the highest affinity to the AR1-like receptors in all five teleost species, followed by DHT and then, in the species in which it was tested, MT. Thereafter, the steroid-specificity profiles of the lower-affinity steroids vary. In addition, within the salmonid family, AR2-like receptors have been found in coho salmon [14]. In general, the AR2 receptors investigated to date bind a broader range of androgens, and their specificity profiles show greater species differences, than the AR1 receptors. The occurrence of both AR1-like and AR2-like receptors in Salmonidae, a family with ancestral characteristics, as well as in Sciaenidae and Serranidae, two families with more derived characteristics, suggests that multiple ARs are widespread among teleost fish.
High levels of androgens are usually detected in the blood of both male and female fish during gonadal recrudescence. Both T and 11-KT are present in high concentrations in most male teleosts, whereas only T is present in the circulation in most females [35]. The EC50s for T binding to kbAR1 and kbAR2 are similar, which suggests that the androgenic action of T can be mediated by binding to either AR1 or AR2. In contrast, 11-KT binds with high affinity only to kbAR2, suggesting that the actions of this steroid are likely mediated through binding to AR2. However, our current knowledge of the binding specificities of AR1 and AR2 is limited, leaving open the possibility that other metabolites of T may be important within specific target tissues. In addition, the AR2-like receptors examined to date have variable specificity profiles for T, 11-KT, and DHT. While all three steroids bind with high affinity, the rank order of their specificity is different, suggesting that their relative physiological importance may differ among teleost species. Currently, there is little evidence linking physiological functions of the major teleost androgens with an AR-mediated mechanism of action. In order to understand the relative physiological importance of androgens within target tissues, the specific tissue and cellular distributions of the two teleostean ARs need to be directly examined, since their distributions govern which steroid can act at which target site. Likewise, with the presence of multiple ARs in fish it is necessary to understand the tissue and cellular distributions of the various androgens.
While few studies demonstrate exclusive actions of either T or 11-KT on physiological processes in fish, there is evidence that 11-KT is a potent androgen inducing certain male-specific traits, e.g., gonadal differentiation [36], secondary sexual characteristics [37–39] and suppression of the stress response [40]. In most cases, the results of these studies have provided only circumstantial evidence for any major physiological role of 11-KT, and the in vivo actions of other androgens have not been ruled out. However, physiological studies with the Japanese eel, Anguilla japonica, suggest that spermatogenesis appears to be mediated solely by 11-KT [41]. An AR has recently been cloned from eel testes that demonstrates high in vitro transcriptional activity in response to 11-KT [9], but DHT, T, and several synthetic androgens also demonstrate high transcriptional activity [9]. There is less evidence correlating direct, AR-mediated actions of T with physiological processes, even though there is ample evidence that T plays an important role in both males and females at the hypothalamus and pituitary in the feedback control of gonadotropin (GTH) secretion [42]. In some cases it is evident that T is acting via aromatase as E2 [43, 44], whereas in male African catfish, Clarias gariepinius, the T metabolite, 11-KT, controls GTH secretion [45]. To date, the only evidence that demonstrates the potential for differential regulation by T and 11-KT, and a potential AR-mediated action of T, is in the control of GTH surges in goldfish. In female goldfish (Carassius auratus), 11-KT induced male-specific spawning behavior and GTH surges [46], whereas T induced a surge of GTH similar to that found in ovulating females [47]. Thus, T and 11-KT may have specific, differential actions controlling GTH release in goldfish that could be mediated by AR1 and AR2, respectively.
Results of the present study show that a major contaminant of aquatic and terrestrial environments, DDT and its derivatives, can bind to ARs in two species of fish, and therefore could potentially disrupt androgen function in both male and female teleosts. In mammals, it has recently been shown that p,p′-DDE can bind to the rat AR in vitro and act as an antiandrogen in vivo [23, 48]. Moreover, the other derivatives of DDT—DDE and DDD—also bind to the rat AR (rAR) and have antiandrogenic activities in vitro [23, 25]. The EC50s of o,p′- and p,p′-DDT, and p,p′-DDD for AR in rat prostate tissue [23], were 2- to 8-fold higher than the EC50s of these same compounds for acAR2 in testicular tissue, while the EC50 of p,p′-DDE was 8-fold lower in testicular acAR2 than in prostate rAR. On the other hand, the binding of the DDT compounds to either ovarian acAR2 or kbAR2 was an order of magnitude lower than that found with either testicular acAR2 or rAR. This may indicate that certain tissues are more susceptible to endocrine disruption by these chemicals; however, on the basis of the present study it is not possible to differentiate the mechanism contributing to these tissue differences. A similar difference was found for the testicular and hepatic ERs in the Atlantic croaker, where most of the steroids and xenobiotics that were examined bound with higher affinity for the testicular ER than for the hepatic ER [49]. The lack of binding of the DDT congeners to either acAR1 or kbAR1 suggests that any agonism or antagonism of the AR caused by DDT derivatives will occur via AR2, rather than AR1. The persistence of p,p′-DDE within the environment and within organisms [23] makes the competitive binding of p,p′-DDE to AR2 the most environmentally relevant among the DDT congeners. In kelp bass collected off of the Southern California Bight, p,p′-DDE accounted for 97% of the total DDT burden in livers, at mean levels of around 3.3 ppm [50]. The EC50 of p,p′-DDE binding to acAR2 was 12 ppm, suggesting that tissue levels of DDE were slightly lower than that necessary to cause 50% displacement of the endogenous ligand, but still sufficient to cause 35–40% displacement. However, within fish, the relationship between the in vivo activity of a xenobiotic and its in vitro receptor binding is not clear.
It has been previously demonstrated that the hydroxylated metabolites of various PCB congeners are capable of binding to the mammalian ER acting as either estrogens or antiestrogens [51]; however, to our knowledge this is the first report demonstrating competitive binding to ARs in vertebrates by both OH-PCB congeners and mixtures of PCBs. The hydroxylated PCB congeners 4,4′-PCB-3-OH, 2,2′,5′-PCB-4-OH, and 2′,5′-PCB-3-OH bound to acAR1 or kbAR1. Since these same congeners, in particular 2,2′,5′-PCB-4-OH, are also capable of binding to acAR2 and kbAR2, they could possibly disrupt the endocrine system by binding to both AR1 and AR2. However, the toxicological importance of OH-PCBs to the fish endocrine system is not known. Hydroxylated metabolites of various PCB congeners have been identified in birds and mammals, including humans [51]; however, it has been demonstrated that fish do not hydroxylate PCBs with the same efficiency as tetrapods [52]. In addition, the PCB metabolites retained in fish, as well as in other wildlife and humans, are the more heavily chlorinated PCBs [53]. In both Atlantic croaker and kelp bass, the tri- and tetra-chloro OH-PCBs bound with higher affinity to ovarian AR2 than did the dichloro OH-PCBs. However, the tetra-chloro OH-PCB, 2′,3′,4′,5′-PCB-4-OH, had the lowest affinity for testicular acAR2. In contrast, the less chlorinated OH-PCBs bound with higher affinity to acAR1 and kbAR1. The PCB mixtures are potentially more toxicologically relevant than the hydroxylated PCBs, and the major environmental contaminant Aroclor 1254 bound to testicular acAR2 with an EC50 equivalent to 9 ppm, which suggests that this PCB mixture has potential to act via the AR and influence androgenic function in a vertebrate species.
The vinclozolin metabolites M1 and, in particular, M2 bound to acAR2 in both testicular and ovarian tissue with relatively high affinity, suggesting that these metabolites have the potential to act as potent endocrine disrupters of the AR in fish. The affinities of M1 and M2 for acAR2 are nearly identical to their affinities for the AR in rats. In rats, these two metabolites act as antagonists of the AR in vitro [54], and vinclozolin has been shown to act as an antiandrogen, presumably via M1 and M2, in vivo [48]. M1 and M2 are the primary metabolites of vinclozolin found within plants, in the soil, and within rodent tissue [23]. However, to our knowledge their presence has not been reported within fish tissues.
In conclusion, we have identified two nuclear ARs, AR1 and AR2, in the kelp bass that correspond to the two nuclear ARs in Atlantic croaker previously characterized by us [1]. These two ARs possess different binding affinities for natural androgens and synthetic steroids as well as for xenobiotics. The presence of multiple ARs suggests an additional level of differential regulation of target genes and of xenobiotic interference that has not been previously recognized. Therefore, to understand the physiological roles of these receptors in mediating androgen actions, it is necessary to determine their specific cellular and tissue distributions as well as their differential regulation throughout the reproductive cycle. This will provide the background necessary to obtain a comprehensive understanding, not only of the normal physiology of androgen action, but also of the sites and mechanisms of xenobiotic interference with the endocrine system mediated by binding to ARs.
Acknowledgments
The authors wish to thank Upjohn Laboratories for their donation of mibolerone and L.E. Gray for the donation of vinclozolin and the metabolites M1 and M2. In addition, we thank W.R. and S. Lawson for their assistance with fish care and the members of the P. Thomas lab for assistance in collecting fish.
References
Author notes
This research was supported by PHS grant ESO 4214, Texas A&M University Sea Grant College Program grant RIMBT-3 and a NOAA grant to P.T. and the E.J. Lund Fellowship, University of Texas Marine Science Institute to T.S.
