| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Fisheries and Wildlife,3 Michigan State University, East Lansing, Michigan 48824
The Centre for Environment, Fisheries and Aquaculture Science,4 Weymouth, Dorset DT4 8UB, United Kingdom
ABSTRACT
The use of nuclear steroid receptors as ligand-activated transcription factors is a critical event in vertebrate evolution. It is believed that nuclear steroid receptors arose at or before the vertebrate radiation, except for an androgen receptor (Ar) that evolved only in the gnathostome line. We report an androgen-Ar complex in the male sea lamprey (Petromyzon marinus), an extant jawless vertebrate. The androgen with the highest affinity is not testosterone, but its direct precursor, androstenedione (Ad). To establish that the binding moiety in lamprey testis is a receptor—and not an "androgen-binding protein"—we have shown that it can be extracted from the nucleus as well as the cytosol, that the Ad-receptor complex binds to DNA, and that the receptor is approximately twice the size of an androgen-binding protein extracted from the Atlantic salmon testis. The capacity (and high affinity) of binding of the lamprey Ar is such that much of the Ad present in male lampreys becomes sequestered within the testis (as opposed to circulating in the plasma). Concentrations of Ad (but not of testosterone) in plasma and testis tissue are upregulated by injection of lamprey GnRH. Implantation of male lampreys with exogenous Ad significantly accelerates the development of the testis and growth of at least one secondary male characteristic. It appears that all classes of steroid hormones have contributed to the evolution of the regulatory complexity of steroid receptors found in modern vertebrates.
agnathan, androgen, androgen receptor, androstenedione, lamprey, receptor, steroid hormones, steroid hormone receptors, testis, testosterone
The nuclear receptor superfamily is an ancient lineage of transcription factors that are predominantly activated by small ligands [1–3]. As part of this superfamily, the steroid receptors have received much attention because they are considered an evolutionary advance, responsible for much of the regulatory complexity within the vertebrate lineage [4, 5]. Based largely on information garnered from the sea lamprey (Petromyzon marinus), an extant representative of the Agnathan lines that evolved over 500 million yr ago [6, 7], it has been proposed that the estrogen receptor was first to evolve [8–10] and that the progestogen and corticoid receptors arose next through gene duplications. In this scheme, the androgen receptor (Ar) arose only in the gnathostome line.
The supposition that lampreys lack an Ar has been supported by several lines of evidence, including the lack of binding activity of an androgen analogue (methyltrienolone) to testis extracts [11], the failure to amplify lamprey cDNA sequences homologous to known nuclear ars by PCR [8], and the absence of testosterone (17ß-hydroxyandrost-4-en-3-one, T) in lamprey plasma, even after the injection of high doses of GnRH [12]. These negative data have always been perplexing in view of the fact that juvenile and adult male lampreys exhibit testis development and secondary sexual characteristics similar to those of other vertebrates. We reasoned that some form of androgen signaling should be necessary for the developmental, physiological, and behavioral operations of male lamprey reproduction.
We further reasoned that the ligand for this putative receptor may not necessarily be T. In the present study, our objective was to establish the identity of the main androgen ligand in the lamprey, characterize its cognate receptor, and demonstrate that the ligand-receptor complex regulates lamprey reproductive physiology.
All experiments were approved by the Michigan State University Institutional Animal Care and Use Committee. Adult sea lampreys were captured during upstream migration by U.S. Fish and Wildlife Service and Canada Department of Fisheries and Oceans personnel and either transported to Michigan State University (East Lansing, MI; for incubation experiments) or to the U.S. Geological Survey Hammond Bay Biological Station (Millersburg, MI; for tissue collection, GnRH injection, and androstenedione [Ad] implant experiments). Tissues collected from fish held at Hammond Bay Biological Station were frozen in liquid nitrogen and held at –80°C until they were processed to separate cytosolic and nuclear fractions. Lampreys were classified as spermiating or ovulating if gentle pressure on the abdomen resulted in the release of gametes.
15
-Hydroxylated standards were synthesized by Ivan 
ern
[13, 14]. All radiolabeled steroids were obtained from Amersham (Piscataway, NJ). Synthetic steroids were obtained from Steraloids (Newport, RI) unless otherwise noted. Lamprey GnRH I [15] and lamprey GnRH III [16] were synthesized by Bachem Peptides (King of Prussia, PA). All other reagents were obtained from Sigma (St. Louis, MO) unless otherwise noted. All statistics were performed by StatView v5.0.1 (SAS, Cary, NC).
Use of Tissue Extract from Incubations as Label in Cytosol-Binding Assay
Finely diced testicular tissue (0.5 g) from a freshly killed prespermiating male lamprey (PSM) was incubated with 50 µCi (1,2,6,7-3H[N])-progesterone (P) in 10 ml of Leibovitz L15 media for 4.5 h at 12°C, similar to previous experiments [14]. The media were removed, and the tissues were extracted with ethanol and ethyl acetate. High-performance liquid chromatography (HPLC) was performed on the extract as described previously [13]. Fractions 11–70 were collected and counted for disintegrations per minute (d.p.m.).
To determine whether some HPLC fractions had corresponding receptors in presumed target tissues, cytosolic and nuclear fractions were obtained from prespermiating male testis and liver (see below). Aliquots of the raw tissue extract from the above incubation were evaporated under nitrogen, redissolved in Hepes buffer (25 mM Hepes, 10 mM NaCl, and 1 mM monothioglycerol, pH 7.4), and mixed with cytosolic or nuclear fractions from testis or liver. The tubes were incubated overnight at 4°C, placed on ice, and mixed with 0.5 ml of an ice-cold suspension of 1.25% charcoal and 0.125% dextran in Hepes buffer (dextran-coated charcoal solution [DCC]). After 5 min, the tubes were centrifuged at 1000 x g for 5 min, and the supernatants (containing the bound steroids) were collected. Of these, 100 µl was placed in a scintillation vial for direct counting, and the remainder was heated to 70°C (to denature the proteins), shaken with 2 ml of ethanol (to selectively dissolve the steroids), diluted 1:20 with distilled water, filtered (40-µm pore size; Millipore, Billerica, MA), and passed through a cartridge of activated octadecylsilane (Seppak C18; Waters, Millford, MA). Steroids were eluted with methanol, which was then dried down, mixed with 10 µg each of P and 15-hydroxyprogesterone (15-P) for standardization of elution times, fractionated by HPLC, and counted for d.p.m.
There was a peak in radioactivity that bound well to testis cytosol, but poorly to liver cytosol, at 51 min. This was the known elution time for Ad (androst-4-ene-3,17-dione) on HPLC. To establish the identity of this peak, some of it was run on thin layer chromatography (TLC) with standard Ad [13].
To confirm that Ad was not being metabolized to another steroid that could also bind to cytosolic receptors, radiolabeled Ad was mixed with testis cytosol and incubated at 5°C overnight. The cytosol was treated with DCC and extracted as described above. The TLC elution positions of the extract (2800 d.p.m.) and the Ad standard were analyzed.
Preparation of Cytosolic and Nuclear Fractions of Tissues
The method for purification of cytosolic and nuclear fractions of tissues was adapted from Patino and Thomas [17]. Briefly, frozen tissues were mixed 1:5 (weight:volume) in Hepes buffer and homogenized. The homogenate was centrifuged at 1000 x g for 15 min at 4°C. The supernatant from this step contained the cytosolic fraction, and the pellet contained the nuclear fraction. To purify the cytosolic fraction, the supernatant was removed to fresh tubes and centrifuged at 30 000 g for 1 h. The resultant supernatant was removed, and glycerol was added (10% volume); it was then frozen in aliquots at –80°C. To purify the nuclear fraction, the pellet from the original centrifugation step was washed three times in a 3x volume of Hepes buffer. After the third wash, the pellet was homogenized with a 3x volume of Hepes buffer with KCl (25 mM Hepes, 10 mM NaCl, 1 mM monothioglycerol, and 0.7 M KCl, pH 7.0) and incubated for 1 h on ice and then centrifuged at 30 000 x g for 1 h.
Characterization of Cytosolic-Binding Component
Saturation kinetics and Scatchard analysis. Radiolabeled Ad (0.31–5.00 nM) was added to 50 µl of testis cytosol diluted with 150 µl of Hepes buffer, either in the presence or absence of 0.5 µg of cold Ad (to determine nonspecific binding [BNS]). The tubes were incubated at 5°C overnight and then placed on ice. DCC was used to remove the free steroid, and the supernatant was counted. The concentration of binding sites (Bmax) and the dissociation constant (KD) were determined by hyperbolic regression by means of Sigmaplot v9.0 (SYSTAT, San Jose, CA). Protein concentrations were determined by the DC Protein Assay Kit for microplates (Bio-Rad, Hercules, CA) with BSA as a standard.
Steroid specificity.
Steroid specificity studies used one-point assays initially, followed by full competition curves for steroids that demonstrated the ability to displace radiolabeled Ad from the receptor. Radiolabeled Ad was mixed with testis cytosol diluted with Hepes buffer in the presence of different amounts of cold steroid (0.3 pg to 1.0 µg). Specificity was examined for T, 11-oxo-testosterone (17ß-hydroxyandrost-4-ene-3,11-dione;11OT), 15-hydroxytestosterone (15-T), 15ß-hydroxytestosterone (15ß-T), 5-dihydrotestosterone (17ß-hydroxy-5-androstan-3-one; 5-DHT), 11-hydroxyandrostenedione (11OH-Ad), 5-reduced Ad (5-Ad), 5ß-reduced Ad (5ß-Ad), mibolerone (Mb), 17-methyl-testosterone (MT), pregnenolone (P5), P, 17-hydroxyprogesterone (17-P), 15-P, cortisol (F), cortisone (B), corticosterone (E), 11-deoxycortisol (S), 11-deoxycorticosterone (DOC), 17ß-estradiol (E2), and estrone (E1). In addition, atypical 3-carbon-reduced forms of Ad and T (as synthesized in cell-free extracts by mollusks [18]) were examined, including 3ß,17ß-dihydroxy-androst-4-ene, 3ß,17ß-dihydroxy-androst-4-ene, 3ß-hydroxy-androst-4-en-17-one, and 3-hydroxy-androst-4-en-17-one.
Association and dissociation kinetics. The association rate was established by determining the specific binding of one-point assays with 5 nM radiolabeled Ad and 100-fold excess cold Ad at times ranging from 5 min to 8 h. To establish the dissociation rate, testis cytosol was incubated with 5 nM radiolabeled Ad overnight at 5°C. Aliquots of cytosol were removed and mixed with 500 nM cold Ad at times ranging from 5 min to 24 h prior to the reaction being stopped by the addition of DCC.
Size of the binding moiety. Cytosolic extract (5 ml) was loaded on to two desalting columns (PD-10; GE Healthcare Life Sciences, Piscataway, NJ) that were each eluted with 3.5 ml of distilled water. The eluates were mixed and lypholized to yield 75 mg of powder. Fifteen milligrams was redissolved in 500 µl of Hepes buffer and passed through a 0.22-µm filter. From the filtered solution, 150 µl was loaded onto a 1- x 30-cm Superdex HR200 size exclusion chromatography column (Amersham). The column was developed with Hepes buffer at room temperature at a rate of 0.5 ml/min, and fractions were collected in glass tubes at 1-min intervals. The fractions from just before the void until after the salt peak were then tested for binding activity with 30 000 d.p.m. tritiated Ad per tube. After a 3-h incubation at 4°C, the tubes were placed in crushed ice, and both free and bound tritiated Ad were then separated by the addition of DCC. Cytosolic extract of Atlantic salmon (Salmo salar) testis was treated exactly as described above, except that 15 mg of powder was initially redissolved in 350 µl of Hepes buffer.
A further 10 mg of lamprey cytosol powder was redissolved in 400 µl of ice-cold Hepes and mixed with 100 µl of a 5% (w/v) solution of digitonin in distilled water [19]. After 15 min, the mixture was passed through a 0.22-µm filter, and 150 µl was loaded onto the Superdex HR200 column. In contrast to the two experiments above, each fraction was split into two aliquots of 200 µl. To one was added 100 µl of buffer containing 250 ng of cold Ad, and to the other, 100 µl of buffer only was added. Tritiated Ad (30 000 d.p.m.) was added to all tubes, and separation of bound and free label was carried out after 3 h as described above. Both experiments were repeated with and without checking for nonspecific binding. A portion of goldfish (Carassius auratus) plasma (diluted 1:4) was tested once (data not shown). Several proteins of known mass were run under identical column conditions.
Characterization of Nuclear-Binding Component
Analyses were performed as described above, with the exception that 200 µl of testis nuclear extract was used in the binding experiments. Steroid specificity studies used one-point assays. Specificity was examined for T, 15-T, 5-DHT, 5-Ad, 5ß-Ad, P5, P, 17-P, 15-P, F, S, E2, and E1.
Cytosolic and nuclear fractions were purified from frozen tissues obtained from PSM and preovulating female sea lampreys. Tissues examined included brain, kidney, heart, gill, muscle, gut, gonad, and liver. Whole plasma was used in the binding experiments as well. Tissue fractions that exhibited significant specific binding to Ad in a one-point assay were used in Scatchard analyses to determine KD and Bmax. Tissues were also obtained from mature (spermiating male [SM] and ovulating female [OF]) adult sea lampreys, and these tissues were analyzed as well.
DNA-cellulose chromatography procedures were adapted from Pinter and Thomas [20]. TEDG buffer consisted of 50 mM Tris, 1.5 mM EDTA, 1 mM dithiothreitol, 10% (v/v) glycerol, and 0.2 mg/ml BSA, pH 7.4, at three concentrations of NaCl, 0.05 M (column buffer), 0.4 M (elution buffer), and 2.0 M (wash buffer). TEDM buffer consisted of 10 mM Tris, 1 mM EDTA, 5 mM dithiothreitol, 10 mM sodium molybdate, and 0.2 mg/ml BSA, pH 7.4, at the same three concentrations of NaCl. DNA-cellulose was obtained from Amersham, and 2 ml was packed in polypropylene mini-columns (Bio-Rad) in TEDG or TEDM column buffer. Testis cytosol was incubated with 5 nM tritiated Ad overnight at 5°C, and 250 µl was loaded on the column. The column was incubated at 5°C for 1 h. The column was then washed with 12 ml of column buffer to remove any free radiolabel or unbound receptors. Bound receptors were eluted with 6 ml of elution buffer, and the column was washed with 2 ml of wash buffer. One-milliliter fractions were collected.
Putative Function of Ad: In Vivo Experiments
To determine changes in plasma androgen concentrations in response to upstream hypothalamic-pituitary-gonadal axis stimulation, PSM lampreys were given two serial injections, 24 h apart, of either lamprey GnRH I or GnRH III (50 µg/kg of body mass) or 0.9% NaCl as a control (n = 12). Blood was sampled prior to the first injection and 6 h after the second injection. Total gonad weight was determined immediately after the second blood sampling, and a gonad sample was frozen in liquid nitrogen and stored at –80°C. RIA was performed as described by Bryan et al. [13, 14]; see Scott et al. [21] for details of Ad assay specificity; the antibody for the T assay was purchased from the Reproductive Endocrinology Laboratory at Colorado State University.
Steroids were extracted from testicular tissue by homogenizing frozen testes with ethanol and ethyl acetate. The extract was purified further by filtration and solid-phase extraction.
For individual samples, the extract was reconstituted in methanol and assayed by RIA as above. For chemical identification, extracted tissue samples from noninjected lampreys were reconstituted in HPLC buffers and fractionated, and the fractions corresponding to the elution position of Ad were evaporated under vacuum. The fractions were analyzed by liquid chromatography mass spectrometry (LCMS) at the Michigan State University Mass Spectrometry Facility. Extracts were also fractionated by HPLC and TLC, and the fractions were evaporated under vacuum. The fractions were then tested for their ability to displace radiolabeled Ad from the Ad-binding moiety. Differences between treatment groups in blood and tissue steroid concentrations were analyzed by an analysis of variance, followed by the Fisher least significant difference (LSD) test.
To test the effect of Ad on male reproductive physiology, migrating lampreys were collected in traps on the Cheboygan River (Michigan) and transported to the Hammond Bay Biological Station in May 2006. Sixty-four PSM fish were divided into eight tanks and acclimated at 16 ± 1°C for 5 days. Two tanks were used per treatment: low-dose (15 mg) pellet, high-dose (150 mg) pellet, placebo, and no implant. The 21-day release pellets formulated for fish were obtained from Innovative Research of America (Sarasota, FL). Each fish was checked for spermiation every 2–4 days by gently pressing on its abdomen. Blood samples were taken weekly to confirm steroid release. Measurements of the width of the dorsal rope, a secondary sex character observed only on mature male lampreys, were taken weekly with calipers (please see Supplemental Fig. 1, available online at www.biolreprod.org), and differences between treatment groups were tested by a two-way analysis of variance, with time and treatment as the independent variables, followed by the Fisher LSD test. Expressed fluids were checked under a microscope to determine the presence of sperm. Upon reaching spermiation, the animals were removed from the tanks. Differences between treatment groups in the time after implantation needed to reach maturation were tested with survival analysis by the Mantel-Cox test.
|
Specificity of the Binding Moiety for Ad
The incubation of PSM testis tissue with tritiated P resulted in several metabolites, which were then mixed with cytosol from the testis and liver (Supplemental Fig. 2, available online at www.biolreprod.org). Binding was observed to a metabolite at the known elution position of Ad on a reverse-phase HPLC column, and major binding was found to synthetic tritiated Ad. Saturation analysis revealed a specific, saturable binding moiety for Ad in testis cytosol (Fig. 1a). Binding analysis (Fig. 1b) showed that the binding moiety had a high affinity (mean ± SEM; KD = 0.69 ± 0.07 nM, n = 3) and a relatively high number of binding sites (Bmax = 755.37 ± 70.20 fmol/mg of protein, n = 3). Of the 24 steroids tested, Ad had the highest affinity to the cytosolic-binding moiety (Table 1 and Supplemental Fig. 3, available online at www.biolreprod.org). The association rate (T1/2) was 11.78 ± 4.10 min (n = 3), and the dissociation rate (T1/2) was 294.32 ± 29.77 min (n = 3; Supplemental Fig. 4, available online at www.biolreprod.org). Binding activity was also present in nuclear extracts from the testis. As with the cytosol, there was a high affinity (KD = 0.28 ± 0.04 nM, n = 3) and a high number of binding sites (Bmax = 429.92 ± 11.94 fmol/mg of protein, n = 3; Supplemental Fig. 5, available online at www.biolreprod.org). A similar pattern of steroid affinities was observed with the nuclear-binding moiety.
|
|
|
|
|
Evidence That the Binding Moiety Is a Receptor and Not a Binding Protein
The 3H-Ad-receptor complex was capable of specifically binding to DNA-cellulose and could be eluted with 0.40 M NaCl. The presence of sodium molybdate did not disrupt binding (Fig. 2) but instead slightly enhanced it.
Specific binding to 3H-Ad was also found in several other tissues of PSM and SM lampreys (Fig. 3). The number of binding sites in both the cytosolic and nuclear fractions increased with maturity in males (Table 2). There was no evidence of binding in the plasma. In females, binding was found only in cytosol from the kidney.
|
Based on its elution position on the gel filtration column (viz. the void), the binding moiety in lamprey testis cytosol had a very high mass (>1 mDa; Fig. 4). This was in complete contrast to the binding moiety for Ad that could be extracted from the testis of the Atlantic salmon, which eluted relatively late on the column. Given its high mass, we suspected that the lamprey binding moiety was membrane-bound. This was confirmed by treatment of the extract with digitonin, which converted the activity to a peak that was roughly twice the size of the androgen-binding activity in the salmon testis (approximately 440 kDa compared with approximately 200 kDa). We were unable to distinguish whether the binding moiety in the salmon testis was due to the presence of a specific androgen-binding protein (Abp) or to plasma-borne sex hormone-binding globulin (Shbg), as both proteins are derived from the same gene, although they have different sites of origin (testis vs. liver) and degree of glycosylation [22]. We did establish, however, that the binding activity in the salmon testis ran in exactly the same position as goldfish plasma Shbg (data not shown).
Ad Is Sequestered in the Testis in Association with the High-Affinity and High-Capacity Ar
Relatively low amounts of Ad were found in male plasma. At 6 h after GnRH or saline injection, plasma concentrations of Ad in PSM lampreys decreased to 0.59 ± 0.09 ng/ml in saline-injected males (P = 0.0162) and increased to only 1.55 ± 0.36 ng/ml (P = 0.0054) and 1.22 ± 0.16 ng/ml (P = 0.0090) in males treated with lamprey GnRH I and GnRH III, respectively (Fig. 5a; ANOVA, followed by the Fisher LSD). T plasma concentrations did not increase at all following GnRH injections (Fig. 5b, ANOVA, P > 0.05).
The discordance between the very high levels of putative receptors and the low amount of putative ligand in the plasma, even after stimulation with GnRH, led us to examine whether the Ad might be sequestered within the testis tissue. Ad levels in testicular tissue extracts of control animals were 0.08 ± 0.01 ng per gram of body weight (6.68 ± 2.13 ng/g of tissue; n = 8) and rose to 0.28 ± 0.02 ng Ad per gram of body weight (15.33 ± 1.94 ng/g of tissue) after injection of GnRH I (ANOVA followed by the Fisher LSD, P < 0.0001) and 0.16 ± 0.03 ng Ad per gram of body weight after injection of GnRH III (10.74 ± 2.13 ng/g of tissue; P = 0.0170; Fig. 5c). If the assumption that plasma volume is 5.5% body weight is used [23], then there is 2–3.5 times more Ad in the testis than in the entire blood volume. There was no significant change in the amount of T in the testis after GnRH injection (Fig. 5d; ANOVA, P > 0.05). LCMS of the HPLC fraction corresponding to the known elution position of Ad identified a molecule that had the same retention time and mass as the Ad standard, thus definitively identifying Ad in testis tissue.
To establish that Ad in testis tissue was the main ligand for the receptor, HPLC fractions of a testis extract were tested for their ability to displace tritiated Ad from the cytosolic-binding moiety. The major peak of binding activity corresponded to the elution position of Ad, although there was one other relatively minor (unidentified) peak (Fig. 6a). There was zero binding activity in the fraction corresponding to the elution position of T. A testis extract was also fractionated with TLC, and binding assays were performed on the fractions. The results further confirmed that Ad was the major ligand for the receptor in testis extract (Fig. 6b).
|
Based on the average Bmax and the protein concentration of the testis cytosol, we estimate that the testes of the PSMs have a capacity to bind 44.42 pmol of Ad per gram of testis tissue when all receptor sites are filled. The testes from the control lampreys from the above experiment contain an estimated 15.25 pmol of Ad per gram of testicular tissue, which is less than half the capacity. The testes from the lampreys injected with GnRH I and GnRH III contain 48.32 and 35.55 pmol Ad per gram of testicular tissue, respectively, which indicates saturation of the binding sites. In comparison, the plasma of the average male sea lamprey contains approximately 2.37 pmol/ml and increases to 5.42 or 4.27 pmol/ml after GnRH I or III injection.
Effects of Ad on Male Reproductive Physiology
PSM lampreys receiving high-dose pellets of Ad had an accelerated rate of maturation compared with lampreys receiving low-dose pellets, placebo pellets, or no pellet (survival analysis, Mantel-Cox test, P < 0.0001, Fig. 7). At 7 days after implantation, the rope widths of sea lampreys given high-dose pellets were significantly larger than the rope widths of sea lampreys in all other treatment groups (ANOVA followed by the Fisher LSD, P = 0.0029; Fig. 8).
|
|
While exploring the potential of a method for combining a common procedure that is used to identify steroids in fish (incubation of gonads with tritiated precursors) with a common procedure that is used to identify receptors (incubation of radioactive steroids with tissue cytosol), we have discovered a functional androgen-Ar complex in the sea lamprey. The inability of previous studies to detect an Ar in lampreys is understandable. These studies have used an unsuitable ligand to detect binding [11], looked for an ar transcript in the liver (a tissue in which the Ad-binding moiety is not found [8]), or measured classical androgens in plasma [12].
It has been established that the nuclear steroid receptor family of functional ligand-activated transcription factors has evolved only in the vertebrate line [8, 9]. However, the use of steroids as intercellular signals appears to have evolved independently several times through different pathways, including that of the membrane steroid receptors [24, 25], and there are unresolved data regarding the effects of classical steroids in mollusks [26]. We caution that, in the absence of information on the primary protein structure of this receptor, we are unable to state definitively that it belongs to the same gene family as other vertebrate nuclear steroid receptors.
We have taken into consideration that what we have discovered might be an Abp (homologous to the Shbg found in plasma) and not a receptor. Proteins that have the ability to bind androgens but that are not nuclear steroid receptors have been characterized in the testis cytosol of two other nonmammalian vertebrates—dogfish Squalus acanthias [27] and a urodele amphibian Necturus maculosus [28]. Although they share some of the same characteristics of the Ad-binding moiety in the lamprey (especially high capacity; see below), they differ in that in these other species, there was no binding in nuclear fractions nor to DNA. EDTA and monothioglycerol are known to disrupt the binding abilities of steroid-binding proteins [29], and yet the lamprey Ar binds ligand in their presence. In addition, the presence of high amounts of exogenous Ad in male sea lampreys had multiple effects on male reproductive physiology, including maturation and at least one secondary sex characteristic, which could only be mediated through some form of steroid-receptor complex. The lamprey Ad-binding moiety also had completely different chromatographic characteristics from the Ad-binding moiety in Atlantic salmon testis. In the latter activity, it was confirmed as either Abp or Shbg by comparing its elution position with Shbg in goldfish plasma, which has previously been shown to bind Ad [30]. The Ad-binding moiety in the lamprey testis appeared to be associated in some way with the membrane (most likely as neutrally buoyant micromicelles). Even when the activity was dissociated from the membrane by treatment with digitonin, it was still different from the teleost Abp or Shbg—being approximately twice as large. When comparing the characteristics of the Ar and Abp in sheep testis cytosol, Carreau et al. [31] found that the two proteins had masses of 192 and 90 kDa, respectively, on sucrose gradient centrifugation. These are roughly half the approximate masses (440 and 200 kDa, respectively) measured in the present study, suggesting that both proteins migrate as dimers under gel filtration conditions. However, this remains to be established. Although the chromatographic studies do not conclusively prove that the lamprey-binding moiety corresponds to a conventional nuclear Ar, they confirm that it does not have the characteristics of a typical Abp.
When the affinity, association rate, dissociation rate, and specificity of Ars and Abps from different species are examined, there are wide overlaps in their range of values and wide differences between species. The KD of the lamprey Ad-binding moiety in testis cytosol is 0.69 ± 0.07 nM, and Ar KDs range from 0.04 nM (rat pituitary [32]) to 5.2 nM (sheep epididymis [31]). The KDs of the Abps found in dogfish and Necturus testes also fall within this range (being 2.2–2.5 nM [27] and 1.0 nM [28], respectively). In the present study, the association rate T1/2 was less than 12 min, and the dissociation rate T1/2 was nearly 5 h, which also lie within the ranges previously recorded for Ars [33]. There are few published studies on the kinetics of Abps, but the T1/2 of the dissociation rate is usually only a few minutes [28, 31]. Like the binding moiety in the sea lamprey, both Ars and Abps often have affinities for steroids other than androgens (see below [27, 33–35]). Thus, we would suggest, it is impossible to distinguish between these two types of protein on the basis of any of the above-mentioned characteristics. In terms of binding capacity, however, it does seem to have been a "rule" that receptors have a low number of binding sites in comparison to Abps. The Bmax for the binding moiety in lamprey PSM testis was 755.37 ± 70.20 fmol/mg of protein (approximately 44.42 pmol/g of tissue), and published values for Bmax (as picomoles per gram of tissue) for Ars range from 0.094 in dogfish testis [35] to 68 in goldfish brain [30], with the majority of published values at the lower end of that range. The reported values of Bmax for Abps (listed as fentomole per milligram of protein) are 20–220 for dogfish [27] and 600 for Necturus [28]. In the present study, we have found a capacity of binding that is more similar to the Abps. Thus, it appears that this is a further characteristic that can no longer be used to distinguish a receptor from a steroid-binding protein.
It is the high capacity of the lamprey Ar, in combination with a relatively high affinity, fast association rate, and slow dissociation rate that appears to be responsible for most of the Ad in the lamprey being found in the testis tissue rather than in the blood plasma. The prevalence of Ar in males compared with females suggests that Ad plays a sex-specific role in males and demonstrates that Ar played an important role in regulating male reproduction at the beginning of vertebrate evolution. The presence of Ar in the heart and kidney of male lampreys and in the kidneys of female lampreys requires further investigation.
The results of the experiment with time-release implants of Ad do not conclusively establish a role for Ad. It is possible that the Ad is converted into another steroid, which then has effects on male reproductive physiology through a different receptor. It is also possible that Ad is acting nonspecifically via a receptor for another steroid. However, our experiments demonstrated that Ad is likely the main ligand for the Ar and that there are few other endogenous androgens present in lampreys. This therefore suggests that Ad is the true ligand for the binding moiety and that the physiological response observed in response to Ad implants was due to Ad itself. Previous investigations of androgen metabolism have shown that both Ad [36, 37] and T [13, 37, 38] are converted to 15-hydroxylated steroids. Our study has shown that Ad is not metabolized in cytosol, confirming that, like most other P450 enzymes, the 15-hydroxylase is located in the mitochondria. We have also shown that the binding moiety has little affinity for 15-T.
The steroidogenic capabilities of the testis appear similar to a previous study, which used tritiated P as a precursor [14]. However, the production of Ad was not observed at that time [14]. The likely explanation is that the previous study examined only the steroids found in the incubation media and not in the tissue extract. We have shown that Ad is sequestered by the testis tissue—and, indeed, it was in the tissue extract (and not in the medium) that we identified it in the present study. In a previous study [37], it was observed that a minor product of tritiated pregnenolone in lamprey testis was a steroid that coeluted with Ad on HPLC. However, this was also the elution position of dehydroepiandrosterone, and its identity was thus not fully established.
Perhaps the most perplexing observation of the present study was that the putative lamprey Ar has a 12% cross-reaction with 11-deoxycortisol—a steroid that was found in the plasma of lampreys over 25 yr ago [39]. The terminology for this steroid suggests that the lamprey Ar is actually the corticosteroid receptor. However, the receptor did not recognize the presence of any of the common corticosteroids or even 11-deoxycorticosterone. Also, if it were a corticosteroid receptor, it might have been expected to be as prevalent in females as in males and also present in a wider range of tissues. A lack of cross-reaction with P suggests that we have not inadvertently characterized the lamprey progesterone receptor. In this respect, however, the Ars of some other nonmammalian species have been found to have a much higher affinity for P (freshwater turtle, Trachemys scripta [33]; coho salmon, Onchorhynchus kisutch [34]; and shark [35]) than the lamprey Ar has for 11-deoxycortisol.
The only previous radioligand-binding study on lamprey steroid receptors [11] demonstrated a binding moiety (probable receptor) for E2. However, the finding that E2 does not bind to the Ad-binding moiety suggests that the Ad-binding moiety is not the estrogen receptor. The finding that the Ar does not bind 15-T suggests that either 15-T acts through a separate receptor or 15-T is not a functional hormone, even though it is produced in the testis and circulated in the plasma [13] and plasma concentrations increase following injections of GnRH [40, 41].
The presence of sodium molybdate generally disrupts a steroid receptor-ligand complex from binding to DNA by stabilizing receptors in their nontransformed, inactive state [42, 43]. Surprisingly, in the present study, the presence of sodium molybdate slightly enhanced binding. However, a previous study [44] found that sodium molybdate did not disrupt the binding of Ar-ligand complexes to DNA under certain circumstances. Additionally, previous studies of other steroid receptors have noted that the stabilizing effect of sodium molybdate is less effective if the receptor has already been transformed to its active state and that the transformation occurs when the ligand is bound [42, 43]. Because Ad is present in large amounts in the testis, it is possible that much of the Ar in the cytosol was in its activated state prior to the DNA-cellulose chromatography and was thus not prevented from binding by the presence of sodium molybdate. The binding of Ar1 from Atlantic croaker to DNA-cellulose was disrupted by sodium molybdate, but exposure to sodium molybdate did not change its sedimentation in a sucrose gradient, as is expected for nuclear steroid receptors [45], indicating that sodium molybdate interacts with some Ars in different ways.
We have shown that the earliest-evolving known form of an androgen signaling pathway uses Ad, and not T, as the androgenic hormone. In the steroid biosynthetic pathway, Ad is the direct precursor of T and an indirect precursor of other known major ligands of gnathostome Ars, including 5-DHT, as found in tetrapods, or 11-oxo-testosterone, as found in teleost fish [46]. Although Ad appears to have preceded T as the main androgen, it has not disappeared during the evolution of gnathostomes. In fact, it is an abundant steroid (and also a pheromone) in some teleost fish [47]. In humans, Ad is mainly synthesized by the adrenal gland and retains 1.5% of the affinity of 5-DHT for the human Ar [48]. It was at one time widely used as an alternative to T by athletes until its use was banned by the World Anti-Doping Agency and the Controlled Substance Act (USA, 2004). The evolutionary changes in the Ar resulting in the change in affinity from Ad to T and 5-DHT warrant further investigation.
ACKNOWLEDGMENTS
We thank Nicholas Johnson for his assistance with the GnRH injection and Ad implant experiments, Ashley Carney for her assistance with the RIAs, Roger Bergstedt and the staff at the U.S. Geological Survey Hammond Bay Biological Station for the use of materials and space, and the U.S. Fish and Wildlife Biological Stations at Marquette, Michigan, and the Canada Department of Fisheries and Oceans Sea Lamprey Control Centre at Sault Ste. Marie, Ontario, Canada, for providing lampreys. The photographs for Supplementary Figure 1 were provided by Nicholas Johnson (Michigan State University, East Lansing, MI).
FOOTNOTES
1Supported by NSF IOB 0450916 and by the Great Lakes Fishery Commission. 
Correspondence: 2Weiming Li, Department of Fisheries and Wildlife, 13 Natural Resources Building, Michigan State University, East Lansing, MI 48824. FAX: 517 432 1699; e-mail: liweim{at}msu.edu
Received: 22 February 2007.
First decision: 28 March 2007.
Accepted: 22 June 2007.
REFERENCES
ern I, Yun SS, Li W. 15-Hydroxytestosterone produced in vitro and in vivo in the sea lamprey, Petromyzon marinus Gen Comp Endocrinol 2003 132418–426[CrossRef][Medline]ern I, Li W. 15-Hydroxyprogesterone in male sea lampreys, Petromyzon marinus L Steroids 2004 69273–281, 20ß,21-trihydroxy-4-pregnen-3-one in ovaries of spotted seatrout (Cynoscion nebulosus) Gen Comp Endocrinol 1990 78204–217[CrossRef][Medline]-hydroxylated androgens by the testis and other tissues of the sea lamprey, Petromyzon marinus, in vitro Gen Comp Endocrinol 1982 4267–270- and 7-hydroxylase activity in gonadal tissue of the early-life stages of sea lampreys, Petromyzon marinus Comp Biochem Physiol Part B 2003 138119–127-Hydroxytestosterone induction by GnRH-I and GnRH-III in Atlantic and Great Lakes sea lamprey (Petromyzon marinus L.) Gen Comp Endocrinol 2004 136276–281[CrossRef][Medline]-hydroxylated sex steroids to gonadotropin-releasing hormones and pituitary extract in male sea lampreys (Petromyzon marinus) Gen Comp Endocrinol 2007 151108–115[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  M. S. Hossain, A. Larsson, N. Scherbak, P.-E. Olsson, and L. Orban Zebrafish Androgen Receptor: Isolation, Molecular, and Biochemical Characterization Biol Reprod, February 1, 2008; 78(2): 361 - 369. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
