- Email: [email protected]
Seasonal variation in specific plasma- and target-tissue binding of androgens, relative to plasma steroid levels, in the brown trout, Salmo trutta L
GENERAL
AND
COMPARATIVE
ENDOCRINOLOGY
70, 334344 (1988)
Seasonal Variation in Specific Plasma- and Target-Tissue Binding of Androgens, Relative to Plasma Steroid Levels, in the Brown Trout, Salmo trutta L. T. G. POTTINGER The Freshwater
Biological
Association,
The Ferry United
House, Far Kingdom
Sawrey,
Ambleside.
Cumbria
LA22
OLP,
Accepted December 12, 1987 The circulating levels of the plasma androgens, testosterone and 1I-ketotestosterone, and their specific binding to skin cytosol, skin nuclear extract, and plasma were determined in mature male and immature male and female brown trout during a single spawning cycle. 1I-Ketotestosterone was not bound by any of the fractions examined whereas testosterone was bound with high affinity to plasma (/Q, = 32.6 nM), skin cytosol (k, = 16.9 ni’M), and skin nuclear extract (k, = 2.6 r&f). The binding capacity of each fraction varied independently with time. In mature male fish an increase in specific binding of testosterone to nuclear extract, from 77 to 269 fmol mg- ’ protein, occurred between September and November, coincident with peak androgen levels. Following the spawning period and the decline in androgen levels, nuclear-binding capacity in mature fish dropped to a level similar to that of immature fish by June. Nuclear binding in immature fish remained in the range 25-75 fmol mgg ’ protein throughout. Plasma-binding capacity of both mature and immature fish declined during the spawning period, from 190 to 125 nM in mature fish and from 360 to 125 nM in immature fish. Plasma-binding capacity in both mature and immature fish increased following spawning to reach levels of 340 nM (mature) and 250 nM (immature). Little change was observed in cytosol-binding capacity of either mature or immature fish. The results suggest that androgen-induced structural changes in the integument are predominantly testosterone stimulated, are initiated by an increase in the concentration of a specific testosterone-binding protein within the nucleus, may be potentiated by a drop in plasma testosterone-binding capacity, and that a cytosol-binding protein of intermediate affinity for testosterone may maintain a high intracellular concentration of steroid. 0 1988 Academic Press, Inc.
The control and coordination of reproductive processes in teleost fish, in common with higher vertebrates, is mediated by a range of steroid hormones of gonadal origin, under direct control of one or more pituitary gonadotropins (Fostier et al., 1983; Idler and Ng, 1983). Studies on the cyclical nature of the reproductive process in fish have provided much information on patterns of hormone secretion (Kime and Manning, 1982; Scott and Sumpter, 1983; Scott et al., 1980a, b). However, hormonally influenced physiological processes may be subject to control in a more subtle manner than simply by fluctuations in circulat-
ing hormone levels. Target tissues responding to a given hormone are characterized by the presence of specific receptor proteins (Baulieu, 1979) and it is known that receptor concentration can vary with time within a single tissue, processes such as castration, hypophysectomy, and treatment with exogenous hormones may all influence receptor numbers (Dubois and Almon, 1984; Sanborn et al., 1984; Smith and Shuster, 1984; Tezon and Blaquier, 1983). Furthermore, a specific steroid-binding system exists in the plasma of many species and may be of some significance in altering free hormone levels and therefore the concentra334
0016~6480188 $1.50 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
SEASONAL
VARIATION
IN
ANDROGEN
tion of hormone available to target tissues (Moll and Rosenfield, 1986). The aim of the present study was to examine three components of a teleost hormone-target-tissue complex; total plasma steroid level, specific plasma-binding capacity, and target-tissue receptor concentration, to determine whether interactions occur and assess their significance. The system chosen for study was the skin of mature male brown trout, Saho trutta, the thickening and demucification of which is an androgen-dependent secondary sexual character (Pottinger and Pickering, 1985a, b). A putative androgen receptor has been identified within brown trout skin (Pottinger, 1987) and brown trout plasma is known to bind androgens in a specific, saturable manner (Pottinger, 1986, 1987).
BINDING
IN TROUT
335
by the addition of 5 vol of 0.7 M KC1 in homogenization buffer, followed by freezing at -70°C for 1 hr. Thawed extract was centrifuged at 4” for 1 hr at 50,OOOgand the clear supernatant (a crude nuclear extract) was frozen in aliquots at - 70” until required for assay. All binding assays were carried out-within 3 days of the collection of material. Quantification of steroid binding. The binding of [3H]testosterone to skin cytosol and crude nuclear extract was quantified by saturation analysis. Nuclear extract containing 2.0-5.0 mg protein ml-’ was incubated in duplicate 200+1 aliquots with 100 p,l homogenization buffer containing 0.6-20.0 nM [I ,2,6,7-3H]testosterone (91.7 Ci mmol- ‘, Amersham) both with and without a loo-fold excess of inert testosterone. For plasma and cytosol, which possess considerably higher binding capacity, a different protocol was adopted. Aliquots of 200 ~1 of cytosol or 200 p.1of plasma (diluted 1:lO with buffer) were incubated in duplicate with 100 ~1 buffer containing increasing amounts of inert testosterone (5-160 nM) and a constant amount of [3H]testosterone (1.25 nM) both with and without a IOO-fold excess of inert testosterone. For both methods samples were incubated overnight MATERIALS AND METHODS (18 hr) at 4”. Fish. One hundred 2 + brown trout (Hungerford Unbound steroid was removed by the addition to strain) were divided equally into two 1500-liter outeach tube of 200$ charcoal suspension (1.25% chardoor fiberglass tanks in July 1986, supplied with a concoal, 0.125% dextran in buffer) followed by incubation stant flow of lake water (35 liters min- ‘). One hundred on ice for 5 min and centrifugation at 4” for 5 min at 1 + fish of the same strain were similarly divided into IOOOg. Aliquots of supernatant were removed, added a further two tanks. The fish were fed once daily with to 5.0 ml scintillant (Unisolve 1, Koch-Light), and commercial feed at a rate of 1% body weight day- ‘. counted under standard 3H conditions. Specific bindSampling and tissue preparation. The fish were ing was calculated from nonspecific and total binding sampled at monthly intervals between August 1986 and the equilibrium dissociation constant (k,)- and and June 1987. At each sampling time eight mature 2 + maximum number of binding sites (N,,,) were calcumales and eight immature 1 + fish were removed. Fish lated according to Scatchard (1949). The binding of were netted from each pond, four at a time and anes- [3H] 1I-ketotestosterone was assessed in an identical thetized in 2-phenoxyethanol (0.5 ml liter-‘). Blood fashion to that of [3H]testosterone. Labeled llwas collected into a heparinized syringe from the cau- ketotestosterone was prepared from [3H]cortisol as da1 vessels; fish were killed by a blow to the head and described previously (Pottinger, 1987). skin was removed from both flanks in 2-cm squares Protein determination. The protein concentration of and placed immediately into buffer (50 mM Tris-HCl, the various fractions was measured by the method of pH 7.4) on ice. A total of -60 cm* per fish was norOhnishi and Barr (1978). mally removed. Tissue was rinsed twice in buffer to Radioimmunoassays. Plasma testosterone and llremove excess mucus, trimmed to remove adherent ketotestosterone levels were measured by radioimmumuscle, and minced finely with scissors. Minced tissue noassay according to the methods of Pottinger and was homogenized in 2 vol of buffer (50 mM Tris-HCl, Pickering (1985a) although antisera of higher specific0.1 mM EDTA, 50 mM dithiothreitol, 10 mM sodium ity were employed allowing omission of the TLC stage molybdate, pH 7.4) on ice using an IKA Ultra Turrax (see Pickering et al., 1987 for details). (TWO) for IO-set bursts with cooling between bursts. Statistical analyses. The concentration of binding Homogenate was centrifuged at 4” for 10 min at 1OOOg sites in plasma, cytosol, and nuclear extract were septo prepare a crude nuclear pellet. Supernatant was trans- arately analyzed by multifactorial analysis of variance ferred to clean tubes and centrifuged at 4” for 1 hr at (ANOVA, Genstat) with maturity (mature/immature) 50,OOOg. The clear supernatant (cytosol) was aspirated and time as factors. From a plot of the residuals and stored at - 70” until required for assay. The pellets against fitted values appropriate transformations (V or containing nuclear material were washed three times log) were selected, where necessary, to improve hoin 10 vol of homogenization buffer and then extracted mogeneity of variance. The levels of significance given
336
T. G. POTTINGER
in the paper are derived from these analyses but for ease of presentation, data are given as arithmetic means + SEM.
RESULTS Hormone levels. During the period August 1986 to June 1987, plasma llketotestosterone levels rose from 5.0 ng ml-’ to reach a peak of 53 .O ng ml- ’ in November, declining to < 2.0 ng ml-’ by March (Fig. la). Levels of testosterone rose from 4.0 ng ml-’ in August to peak in October at 15.0 ng ml- ’ before declining to < l.Ongml-’ in February (Fig. lb). Levels of both hormones in immature fish remained below 1.0 ng ml- ’ throughout the period studied and have not, therefore, been plotted. Plasma steroid-binding capacity. Total
plasma androgen-binding capacity was calculated from the specific binding (mg-’ protein) and the protein concentration in the plasma at each sample. The results are presented in Fig. 2. Both mature and immature fish showed significant changes in total binding capacity with time (P < 0.001) and a significant maturity X time interaction was also noted (P < 0.001). This was resolved as significantly higher binding capacity in immature fish during September and October (P < 0.001) and significantly lower during May (P < 0.05). Overall, plasma-binding capacity in immature fish declined significantly from 360 + 26 nM (n = 8 throughout) in September to a low of 128 * 21 ti in January (P < 0.001) before rising to 258 t 23 ti in March (P < 0.001) and declining to 139 t 17 in May (P <
SEASONAL
VARIATION
IN
ANDROGEN
BINDING
IN
337
TROUT
Month
2. Total testosterone-binding capacity (nM) of brown trout plasma from mature (-O-) immature (- -O- -) fish during the period September to June (Z + SEM, N = 8). FIG.
0.001). A similar, though less marked, pattern was observed in mature fish. Binding capacity declined from 193 t 13 nIL4 in September to 130 * 9 nM in November (P < 0.05) before rising to a peak of 341 -t 33 in March (P < 0.001) and then dropping to 243 + 38 nM in May (P < 0.05). There was no significant difference in affinity of mature
and
or immature plasma for testosterone, overall k,, = 32.6 + 1.0 nM (N = 144). Cytosol-binding capacity. Cytosolbinding capacity varied significantly with time in both mature and immature fish (P < 0.001, Fig. 3). Overall, binding capacity in immature fish declined from 782 5 41 fmol mg-’ protein in September to 292 * 53
1000
c” .z
800
252 g ‘5 00 &h IiO ‘cn
SE
a-gg
600
400
0-
In‘:0
200
0 0
Month
3. Testosterone-binding capacity (IV,,,) of skin cytosol from mature (-04 (- -O- -) brown trout between September and June (F + SEM, n = 8). FIG.
and immature
338
T. G. POTTINGER
fmol mg-’ protein in May (P < 0.01). In mature fish, binding capacity increased from 602 + 57 fmol mg - ’ protein in November to a peak of 944 + 103 fmol mg-’ in March (P -=cO.Ol), thereafter declining to less than 600 fmol mg-’ protein by May. During the period December to May, binding capacity in mature fish was significantly higher than that in immature fish (P < 0.05 - 0.001). There was no significant difference in affinity for testosterone between mature and immature cytosol. Overall, k, = 16.9 + 0.5 n&! (n = 144). Nuclear extract binding capacity. A significant effect of maturity (P < O.Ol), time (P < O.OOl), and a significant maturity x time interaction (P < 0.01) were observed in the nuclear extract binding data (Fig. 4). No measurable change occurred in nuclear binding capacity of immature fish, levels remained in the range 25-75 fmol mg-’ protein throughout the period under study. In mature fish, however, binding capacity increased from 77 _+ 12 fmol mg- ’ protein in September to 269 + 31 fmol mg-’ protein in November (P < 0.001) before declining gradually to a level of 69 It 11 fmol mg - ’ 300
protein in May. Levels of binding in mature fish were significantly higher than in immature fish (P < 0.01-0.001) at all samples except June. There was no difference in K. of nuclear binding between mature and immature fish, kr, = 2.6 * 0.1 ti 9 (n = 144). Binding of 13H]1 1 -ketotestosterone. At no point during the investigation was measurable and reproducible specific binding of [3H] 11-ketotestosterone observed, in any of the fractions studied. DISCUSSION
The hormone profile measured in maturing male brown trout during this experiment resembles closely that previously observed in brown trout (Kime and Manning, 1982; Pottinger and Pickering, 1985a; 1987). Although not directly monitored in the present study, the period during which increased dermal and epidermal thickness and reduced superficial goblet cell concentration are observed occurs between September and April (Pickering, 1977; Pottinger and Pickering, 1985a). Previously, it has been suggested that maturity-related
*
200
100 .
01
,
I
I
I
I
I
I
I
I
I
I
1
A
S
0
N
D
J
F
M
A
M
J
J
Month
4. Testosterone-binding capacity (N,,,) of skin nuclear extract from mature (-O-) immature (--O- -) brown trout between September and June (F + SEM, n = 8). FIG.
and
SEASONAL
VARIATION
IN
ANDROGEN
changes in skin structure were invoked by elevated 1 1-ketotestosterone (1 l-KT) levels, because of both the timing of the changes (Pottinger and Pickering, 1985a) and the ability of exogenous 1 l-KT to induce similar changes in the skin of immature salmonids (Idler et al., 1961; McBride and van Overbeeke, 1971; Pottinger and Pickering, 1985b). However, specific highaffinity binding of 1l-KT, indicative of the presence of an intracellular receptor protein, was not detected in the skin of mature male brown trout (Pottinger, 1987). This observation has been confirmed by the present study. [3H]testosterone (T) was, however, bound by all three fractions examined, plasma, skin cytosol, and skin nuclear extract. The binding capacity of each fraction was found to vary, independently, with time. An increase in measurable binding of 3H-T to nuclear extract occurred in mature fish between September and November. During the same period there was a drop in plasma-binding capacity but no changes were observed in cytosolic binding capacity. Plasma-binding capacity rose from a low point in November to peak in March. In immature fish, no major change was apparent in nuclear or cytosol-binding capacity, they remained low throughout. However, plasma T-binding capacity changed, in immature fish, in a manner similar to that observed in mature fish. There is little information available, for any vertebrate group, on changes during the reproductive period in any of the parameters examined in the present study, other than hormone levels. The blood of vertebrates contains proteins which bind steroids with high affinity and limited capacity (Martin, 1980). A plasma protein specifically binding sex steroids, a presumptive sex hormone-binding globulin (SHBG), has been identified in most higher vertebrates and many lower, including the cod Gadus morhua (Freeman and Idler, 1971), Atlantic salmon Salmo salar (Freeman and Idler, 1971; Lazier et al., 1985).
BINDING
IN TROUT
339
rainbow trout Salmo gairdneri (Fostier and Breton, 1975), and brown trout (Pottinger, 1987). Although none of these studies provide information on temporal changes in binding capacity, in some mammals higher concentrations of SHBG are found in immature than mature animals (Lobl, 1981), for instance there is an inverse relationship between SHBG and plasma testosterone levels in the marmoset monkey (Hodges et al., 1983). In reptiles and amphibians so far studied the relationship between plasma steroid levels and binding capacity is less clear. In female, though not male, alligators SHBG declines during the breeding season (Ho et al., 1987). In male newts, Taricha granulosa, no change in SHBG-binding capacity occurs during the reproductive period (Moore et al., 1983). During the present investigation, plasma T-binding capacity was lowest during the period OctoberJanuary, coincident with peak androgen levels and spawning activity, and highest in March when androgen levels had returned to basal levels. This conforms to a broad pattern of an inverse relationship between binding capacity and androgen levels. The range of plasma-binding capacity observed during the present investigation (- 100-400 nM) compares with a range of binding capacity reported for other vertebrates extending from 23-45 nM in the amphibian Taricha granulosa (Moore et al., 1983) to 2-8 $V in female rainbow trout (Fostier and Breton, 1975). This value is markedly higher than that observed in male brown trout during the present investigation. However, it has been noted that absolute plasma-binding capacity is directly related to maximum plasma steroid levels in a given species (Salhanick and Callard, 1979). In male brown trout T levels rarely exceed 20 ng ml- ‘, in female rainbow trout levels up to 400 ng ml - ’ have been reported (Scott and Sumpter, 1983). In functional terms it has been suggested that plasmabinding proteins protect circulating steroids from inactivation (by conjugation and clear-
340
T.
G.
POTTINGER
ante) and may provide a delivery system to target tissues (Martin, 1980). An alternative view suggests that binding proteins inhibit the biological effectiveness of steroids (Moll and Rosenfield, 1986 and refs therein), an interpretation supported by Fostier and Breton (1975) working on rainbow trout. In the present investigation although significant decline in T-binding capacity, to a level of 130 nM, occurs during the spawning period, this capacity still exceeds the maximum plasma levels of total T (15 ng ml-’ - 50 nM) occurring during this period. The binding protein has little affinity for the other major androgen, ll-KT (Fostier and Breton, 1975; Pottinger, 1987), therefore relatively little ll-KT would be bound, even in the presence of maximal SHBG capacity. Androgens have been implicated in the control of SHBG levels in mammals (see Mooradian et al., 1987) but plasma androgen-binding capacity of immature brown trout fluctuates in an almost identical manner to that of mature fish and throughout the period studied plasma androgens in immature fish remained at a level below 1.O ng ml. The plasma androgen-binding system in brown trout would therefore appear to be controlled by factors outside the pituitary-gonadal axis. Binding of 3H-T to the nuclear fraction of brown trout skin increased nearly four-fold during the period September to November. Previous work has established this to be a period of maximal increase in dermal and epidermal thickness (Pottinger and Pickering, 1985a). Whether the increase in binding sites represents a de nova synthesis of receptor or an increase in receptor occupancy leading to elevated nuclear localization is not clear. It has been shown that intact steroid-receptor complexes are more likely to remain within the nucleus during homogenization than unoccupied receptors (Welshons et al., 1984). However, the number of nuclear-binding sites did not increase until September-October, T levels had already begun to rise by August which suggests oc-
cupancy alone may not explain the increase in measurable binding sites. No attempt was made in the present study to distinguish between occupied and unoccupied nuclear-binding sites. The complete exchange of endogenous (unlabeled) ligand for exogenous tracer is reported to require between 18 hr in the mouse kidney (Isomaa et al., 1982) to 48 hr in the mouse submandibular gland (Kyakumoto, 1986). It is therefore possible that exchange was not completed during the incubation period of 18 hr employed in the present study. This may have resulted in an underestimation of the number of binding sites during the period of maximum plasma T levels. A further factor to be considered is whether the extraction regime employed in this study was sufficiently rigorous to remove the majority of binding sites from the nucleus. Although nonextractable binding sites have been reported to occur in androgen target organs (Barrack et al., 1983) these may represent as few as 5% of the total androgen-receptor content of a tissue (Amet et al., 1986). There was a significant increase in cytosolic binding sites between November and March (P < 0.01) in mature male fish coinciding with the decline in nuclear binding. However, there was no decline in the number of cytosol-binding sites coinciding with the increase in nuclear-binding sites, the affinity of cytosol sites for T was consistently lower than that of nuclear extract (k, = 16.9 I-&! c.f. k, = 2.7 nM), and Scatchard plots gave no indication that cytosol contained two discrete populations of binding protein. It seems unlikely that the increase in nuclear binding is simply a translocation of binder from the cytosol compartment. Although peak 11-KT levels coincide with peak nuclear-binding capacity, the binder has little measurable affinity for 1 lKT (Pottinger, 1987). Regulation of a receptor for one hormone by another is known to occur in various mammalian systems, most notably in some estrogen sensitive tissues which also contain progesterone receptors.
SEASONAL
VARIATION
IN
ANDROGEN
An increase in progesterone receptor numbers can be induced by estrogen administration (Muldoon, 1980). It might, therefore, be suggested that in the present study close correlation of the increase in nuclear testosterone binding with plasma 1 l-KT levels reflects a causal relationship. However, if such were the case, specific 1 l-KT binding ought to exist at some stage during the reproductive cycle. More widely observed is a direct link between plasma hormone levels and receptor concentration in target tissue. Castration results in a drop in androgen-binding sites and T administration replenishes binding sites in rat epididymis (Tezon and Blaquier, 1983), androgen treatment increases binding sites in cultured human epididymis (Vazquez et al., 1986), and has a similar effect in rat ventral prostrate (Blondeau et al., 1982). An analogous system has been observed in other vertebrates. Androgen-binding sites in the larynx of male frogs, Xenopus Zaevis, decline following castration and increase after administration of T (Segil et al., 1987). Androgen binding in the uropygial gland of castrated quail is increased by the administration of T (Amet et al., 1986). This effect was also observed in the thumb pad of castrated and androgen-treated Rana escuZenta (Delrio et al., 1980). The only information available on fish refers to the hepatic estrogen receptor of SaZmo salar, the concentration of which increases in response to estrogen administration (Lazier et al., 1985). In the present investigation, peak plasma T levels precede peak Tbinding capacity by one month. Administration of T and ll-KT to immature fish would clarify which steroid actively increases the number of T-binding sites in the skin. The binding of T to the cytosolic fraction of skin is curious by comparison to other systems examined. Scatchard analysis suggests only one population of binding sites is present with a kt, (16.9 nA4) intermediate between that of plasma (k, = 32.6 nM) and
BINDING
IN
TROUT
341
nuclear extract (k, = 2.6 nM). This binding appears not to be a plasma contaminantdespite large fluctuations in plasma-binding capacity, cytosol-binding levels remain relatively constant. Similar lower affinity cytosol-binding proteins have been reported in mammals (Dekloet and McEwen, 1976; Koch et al., 1976) and turtle (Salhanick and Callard, 1979). This differs markedly from other systems where cytosol and nuclear binding are functionally indistinguishable (e.g., Lazier et al., 1985; Amet et al., 1986; Kleis-San Francisco and Callard, 1986). It is possible that lower affinity cytosolic binding proteins are involved in maintaining local concentrations of steroids and promoting delivery to the nuclear compartment (Salhanick and Callard, 1979). In conclusion, the period during which brown trout skin thickens and demucifies coincides with elevated plasma androgen levels, an increase in salt-extractable nuclear androgen binding in the skin and a decline in plasma androgen-binding capacity. Although both T and ll-KT, when administered to immature fish, stimulate epidermal and dermal thickening (Pottinger and Pickering, 1985a) throughout the spawning period, specific binding exist within the skin for T only. This suggests therefore that, as a target tissue, skin is equipped to respond to T rather than llKT. The large reduction in mucus-secreting goblet cells within the epidermis, which is a feature of maturation (Pottinger and Pickering, 1985a) and can be induced in immature fish by 1 l-KT but not T (Pottinger and Pickering, 1985b), may represent a mechanism ultimately under the control of 1 l-KT but one in which the steroid acts at a site outside the integument. For example, prolactin is known to stimulate goblet cell differentiation in teleost epidermis (Ogawa, 1970; Mattheij et al., 1972; Bonga, 1978; Schwerdtfeger, 1979), further work may indicate whether 1l-KT has a suppressive effect on prolactin secretion. The paradox re-
342
T. G. POTTINGER
mains, however, that 1 l-KT is also potent in stimulating epidermal and dermal thickening. Conversion of 1 l-KT within the skin is a possibility yet to be investigated. Relatively few steroids have been tested for potency in displacing bound [3H]T (Pottinger, 1987), an as yet unidentified metabolite of 1 l-KT may prove to be active at the same binding sites as T. In the absence of comparable data from other species it is difficult to present a functional interpretation of skin cytosol and plasma binding. The drop in plasma binding capacity during the period of elevated androgen levels, if its effect is to potentiate androgen action by increasing the proportion of unbound steroid circulating, suggests the plasma steroid-binding protein may be an important factor modulating hormone activity. The cytosol binder, apparently distinct from both nuclear and plasma binders and with an affinity for T intermediate between the two, may be involved in maintaining a high intracellular concentration of steroid. ACKNOWLEDGMENTS The author thanks Dr. A. D. Pickering for helpful discussion and critical comments on the manuscript, Janet Pollard and Janice Fletcher for maintaining the experimental fish, and Trevor Furnass for preparing the illustrations. This work was financed by the Natural Environment Research Council.
Blondeau, J.-P., Baulieu, E.-E., and Robel, P. (1982). Androgen-dependent regulation of androgen nuclear receptor in the rat ventral prostrate. Endocrinology
Endocrinol.
Amet, Y., Abalain, J.-H., di Stefano, S., Daniel, J.-Y., Tea, K., Floch, H. H., and Robel, P. (1986). Androgen regulation of the androgen receptor of the quail uropygial gland: Application of a [3H] mibolerone exchange assay. J. Endocrinol. 109, 299306. Barrack, E. R., Bujnovszky, P., and Walsh, P. C. (1983). Subcellular distribution of androgen receptors in human normal, benign hyperplastic and malignant prostatic tissues. Characterization of nuclear salt resistant receptors. Cancer Res. 43, 1107-1113. Baulieu, E.-E. (1979). Aspects of steroid hormonetarget cell interactions. Adv. Exp. Med. Biol. 117, 377-399.
34, 265-275.
De Kloet, E. R., and McEwen, B. S. (1976). A putative glucocorticoid receptor and a transcortin-like macromolecule in pituitary cytosol. Biochim. Biophys.
Acta
421, 115-123.
Delrio, G., Citarella, F. and D’Istria, M. (1980). Androgen receptor in the thumb pad of Rana esculenta: Dynamic aspects. J. Endocrinol. 85, 279282. DuBois, D. C., and Almon, R. R. (1984). Perineal muscles: Possible androgen regulation of glucocorticoid receptor sites in the rat. J. Endocrinol. 102, 225-229.
Fostier, A., and Breton, B. (1975). Binding of steroids by plasma of a teleost: The rainbow trout, Salmo gairdneri.
J. Steroid
Biochem.
6, 345-35
1.
Fostier, A., Jalabert, B., Billard, R., Breton, B., and Zohar, Y. (1983). The gonadal steroids. In “Fish Physiology” (W. S. Hoar, D. J. Randall, and E. M. Donaldson, Eds.), Vol 9A, pp. 277-372. Academic Press, New York. Freeman, H. C., and Idler, D. R. (1971). Binding affinities of blood proteins for sex hormones and corticosteroids in fish. Steroids 17, 233-250. Ho, S.-M., Lance, V., and Megaloudis, M. (1987). Plasma sex-steroid binding protein in a seasonally breeding reptile, Alligator mississippiensis. Gen. Comp. Endocrinol. 65, 121-132. Hodges, J. K., Eastman, S. A. K., and Jenkins, N. (1983). Sex steroids and their relationship to binding proteins in the serum of the marmoset monkey (Callithrix
REFERENCES
110, 19261932.
Bonga, S. E. W. (1978). The effect of changes in external sodium, calcium, and magnesium concentrations on prolactin cells, skin, and plasma electrolytes of Gasterosteus aculeatus. Gen. Camp.
jacchus).
J. Endocrinol.
96, 443-450.
Idler, D. R., Bitners, I. I., and Schmidt, P. J. (1961). 1 I-ketotestosterone: An androgen for sockeye salmon. Canad. J. Biochem. Physiol. 39, 17371742. Idler, D. R., and Ng, T. B. (1983). Teleost gonadotropins: Isolation, biochemistry and function. In “Fish Physiology” (W. S. Hoar, D. J. Randall, and E. M. Donaldson, Eds.) Vol9A, pp. 187-221. Academic Press, New York. Isomaa, V., Pajunen, A. E. I., Bardin, C. W., and Janne, 0. A. (1982). Nuclear androgen receptors in the mouse kidney: Validation of a new assay. Endocrinology
111, 833-843.
Kime, D. E., and Manning, N. J. (1982). Seasonal patterns of free and conjugated androgens in the brown trout Salmo trutta. Gen. Comp. Endocrinol. 48, 222-23 1.
SEASONAL
VARIATION
IN
ANDROGEN
Kleis-San Francisco, S. K., and Callard, I. P. (1986). Progesterone receptors in the oviduct of a viviparous snake (Nerodiu): Correlations with ovarian function and plasma steroid levels. Gen. Camp. Endocrinol.
63, 220-229.
Koch, B., Lutz, B., Briaud, B., and Mialhe, C. (1976). Heterogeneity of pituitary glucocorticoid binding evidence for a transcortin-like compound. Biochim.
Biophys.
Acta
444, 497-507.
Kyakumoto, S., Kurokawa, R., Ohara-Nemoto, Y., and Ota, M. (1986). Sex difference in the cytosolic and nuclear distribution of androgen receptor in mouse submandibular gland. J. Endocrinol. 108, 267-273. Lazier, C. B., Lonergan, K., and Mommsen, T. P. (1985). Hepatic estrogen receptors and plasma estrogen-binding activity in the Atlantic salmon. Gen.
Comp.
Endocrinol57,
234-245.
Lobl, T. J. (1981). Androgen transport proteins: Physical properties, hormonal regulation and possible mechanism of TeBG and ABP action. Arch. Androl. 7, 133-151. Martin, B. (1980). Steroid-protein interactions in nonmammalian vertebrates: Distribution, origin, regulation, and physiological significance of plasma steroid binding proteins. In “Steroids and Their Mechanism of Action in Nonmammalian Vertebrates” (G. Delrio and J. Brachet, Eds.) pp. 63-73. Raven Press, New York. Mattheij, J. A. M., Stroband, H. W. J., Kingma, F. J., and van Oordt, P. G. W. J. (1972). Prolactin and osmoregulation in the cichlid fish, Cichlasoma biocellatum, and the effect of this hormone on the thyroid, gills, and skin. Gen. Comp. Endocrinol
18, 607.
McBride, J. R., and van Overbeeke, A. P. (1971). Effects of androgens, estrogens and cortisol on the skin, stomach, liver, pancreas and kidney in gonadectomized adult sockeye salmon (Oncorhynthus
nerka).
J. Fish Res. Board
Canad.
28, 485-
490. Moll, G. W., and Rosenfield, R. L. (1986). Estradiol inhibition of pituitary luteinizing hormone release is antagonized by serum proteins. J. Steroid Biothem.
25, 309-314.
Mooradian, A. D., Morley, J. E., and Korenman, S. G. (1987). Biological actions of androgens, Endoer. Rev. 8, l-28. Moore, F. L., Spielvogel, S. P., Zoeller, R. T., and Wingfield, J. (1983). Testosterone-binding protein in a seasonally breeding amphibian. Gen. Comp. Endocrinol. 49, 15-2 1. Muldoon, T. G. (1980). Regulation of steroid hormone activity. Endocr. Rev. 1, 339-364. Ogawa, M. (1970). Effects of prolactin on the epiderma1 mucous cells of the goldfish Carassius auratus L. Canad. J. Zool. 48, 501-503.
BINDING
IN TROUT
343
Ohnishi, S. T., and Barr, J. K. (1978). A simplified method of quantitating protein using the biuret and phenol reagents. Anal. Biochem. 86, 193-200. Pickering, A. D. (1977). Seasonal changes in the epidermis of the brown trout, Salmo trutta L. J. Fish Biol.
10, 561-566.
Pickering, A. D., Pottinger, T. G., Carragher, J., and Sumpter, J. P. (1987). The effects of acute and chronic stress on the levels of reproductive hormones in the plasma of mature male brown trout, Salmo trutta L. Gen. Comp. Endocrinol. 68, 249259. Pottinger, T. G. (1986). Estrogen-binding sites in the liver of sexually mature male and female brown trout, Salmo trutta L. Gen. Comp. Endocrinol. 61, 120-126. Pottinger, T. G. (1987). Androgen binding in the skin of mature male brown trout, Salmo trutta L. Gen. Comp.
Endocrinol.
66, 224-232.
Pottinger, T. G., and Pickering, A. D. (1985a). Changes in skin structure associated with elevated androgen levels in maturing male brown trout, Salmo trutta L. J. Fish Biol. 26, 745-753. Pottinger, T. G., and Pickering, A. D. (1985b). The effects of 1I-ketotestosterone and testosterone on the skin structure of brown trout, Salmo trutta L. Gen.
Comp.
Endocrinol.
59, 335-342.
Pottinger, T. G., and Pickering, A. D. (1987). Androgen levels and erythrocytosis in maturing brown trout, Salmo trutta L. Fish Physiol. Biochem. 3, 121-126. Salhanick, A. R., and Callard, I. P. (1979). Sex steroid binding proteins in non-mammalian vertebrates. Adv.
Exp.
Med.
Biol.
117, 441-459.
Sanborn, B. M., Wagle, J. R., and Steinberger, A. (1984). Control of androgen cytosol receptor concentrations in sertoli cells: Effect of androgens. Endocrinology
114, 2388-2393.
Scatchard, G. (1949). The attractions of proteins for small molecules and ions. Ann. N.Y. Acad. Sci. 51, 660-672.
Schwerdtfeger, W. K. (1979). Morphometrical studies of the ultrastructure of the epidermis of the guppy, Poecilia reticulata Peters, following adaptation to seawater and treatment with prolactin. Gen.
Comp.
Endocrinol.
38, 476-483.
Scott, A. P., Bye, V. J., and Baynes, S. M. (1980a). Seasonal variations in sex steroids of female rainbow trout (Salmo gairdneri Richardson). J. Fish Biol.
17, 587-592.
Scott, A. P., Bye, V. J., Baynes, S. M., and Springate, J. C. (1980b). Seasonal variations in plasma concentrations of 1 1-ketotestosterone and testosterone in male rainbow trout (Salmo gairdneri Richardson). J. Fish Biol. 17, 495-505. Scott, A. P., and Sumpter, J. P. (1983). A comparison of the female reproductive cycles of autumnspawning and winter-spawning strains of rainbow
344
T. G. POTTINGER
trout (Salmo gairdneri Richardson). Gen. Comp. Endocrinol. 52, 79-85. Segil, N., Silverman, L., and Kelley, D. B. (1987). Androgen-binding levels in a sexually dimorphic muscle of Xenopus laevis. Gen. Comp. Endocrinol. 66, 95-101. Smith, K, and Shuster, S. (1984). Effect of adrenalectomy and steroid treatment on rat skin cytosol glucocorticoid receptor. J. Endocrinol. 102, 161165.
Tezon, 3. G., and Blaquier, J. A. (1983). Androgens control androgen-binding sites in rat epididymis. Endocrinology 113, 1025-1030. Vazquez, M. H., de Larminar, M. A., and Blaquier, J. A. (1986). Effect of androgen on androgen receptors in cultured human epididymis. J. Endocrinol. 111, 343-348. Welshons, W. V., Lieberman, M. E., and Gorski, J. (1984). Nuclear location of unoccupied oestrogen receptors. Nature (London) 307, 747-749.
AND
COMPARATIVE
ENDOCRINOLOGY
70, 334344 (1988)
Seasonal Variation in Specific Plasma- and Target-Tissue Binding of Androgens, Relative to Plasma Steroid Levels, in the Brown Trout, Salmo trutta L. T. G. POTTINGER The Freshwater
Biological
Association,
The Ferry United
House, Far Kingdom
Sawrey,
Ambleside.
Cumbria
LA22
OLP,
Accepted December 12, 1987 The circulating levels of the plasma androgens, testosterone and 1I-ketotestosterone, and their specific binding to skin cytosol, skin nuclear extract, and plasma were determined in mature male and immature male and female brown trout during a single spawning cycle. 1I-Ketotestosterone was not bound by any of the fractions examined whereas testosterone was bound with high affinity to plasma (/Q, = 32.6 nM), skin cytosol (k, = 16.9 ni’M), and skin nuclear extract (k, = 2.6 r&f). The binding capacity of each fraction varied independently with time. In mature male fish an increase in specific binding of testosterone to nuclear extract, from 77 to 269 fmol mg- ’ protein, occurred between September and November, coincident with peak androgen levels. Following the spawning period and the decline in androgen levels, nuclear-binding capacity in mature fish dropped to a level similar to that of immature fish by June. Nuclear binding in immature fish remained in the range 25-75 fmol mgg ’ protein throughout. Plasma-binding capacity of both mature and immature fish declined during the spawning period, from 190 to 125 nM in mature fish and from 360 to 125 nM in immature fish. Plasma-binding capacity in both mature and immature fish increased following spawning to reach levels of 340 nM (mature) and 250 nM (immature). Little change was observed in cytosol-binding capacity of either mature or immature fish. The results suggest that androgen-induced structural changes in the integument are predominantly testosterone stimulated, are initiated by an increase in the concentration of a specific testosterone-binding protein within the nucleus, may be potentiated by a drop in plasma testosterone-binding capacity, and that a cytosol-binding protein of intermediate affinity for testosterone may maintain a high intracellular concentration of steroid. 0 1988 Academic Press, Inc.
The control and coordination of reproductive processes in teleost fish, in common with higher vertebrates, is mediated by a range of steroid hormones of gonadal origin, under direct control of one or more pituitary gonadotropins (Fostier et al., 1983; Idler and Ng, 1983). Studies on the cyclical nature of the reproductive process in fish have provided much information on patterns of hormone secretion (Kime and Manning, 1982; Scott and Sumpter, 1983; Scott et al., 1980a, b). However, hormonally influenced physiological processes may be subject to control in a more subtle manner than simply by fluctuations in circulat-
ing hormone levels. Target tissues responding to a given hormone are characterized by the presence of specific receptor proteins (Baulieu, 1979) and it is known that receptor concentration can vary with time within a single tissue, processes such as castration, hypophysectomy, and treatment with exogenous hormones may all influence receptor numbers (Dubois and Almon, 1984; Sanborn et al., 1984; Smith and Shuster, 1984; Tezon and Blaquier, 1983). Furthermore, a specific steroid-binding system exists in the plasma of many species and may be of some significance in altering free hormone levels and therefore the concentra334
0016~6480188 $1.50 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
SEASONAL
VARIATION
IN
ANDROGEN
tion of hormone available to target tissues (Moll and Rosenfield, 1986). The aim of the present study was to examine three components of a teleost hormone-target-tissue complex; total plasma steroid level, specific plasma-binding capacity, and target-tissue receptor concentration, to determine whether interactions occur and assess their significance. The system chosen for study was the skin of mature male brown trout, Saho trutta, the thickening and demucification of which is an androgen-dependent secondary sexual character (Pottinger and Pickering, 1985a, b). A putative androgen receptor has been identified within brown trout skin (Pottinger, 1987) and brown trout plasma is known to bind androgens in a specific, saturable manner (Pottinger, 1986, 1987).
BINDING
IN TROUT
335
by the addition of 5 vol of 0.7 M KC1 in homogenization buffer, followed by freezing at -70°C for 1 hr. Thawed extract was centrifuged at 4” for 1 hr at 50,OOOgand the clear supernatant (a crude nuclear extract) was frozen in aliquots at - 70” until required for assay. All binding assays were carried out-within 3 days of the collection of material. Quantification of steroid binding. The binding of [3H]testosterone to skin cytosol and crude nuclear extract was quantified by saturation analysis. Nuclear extract containing 2.0-5.0 mg protein ml-’ was incubated in duplicate 200+1 aliquots with 100 p,l homogenization buffer containing 0.6-20.0 nM [I ,2,6,7-3H]testosterone (91.7 Ci mmol- ‘, Amersham) both with and without a loo-fold excess of inert testosterone. For plasma and cytosol, which possess considerably higher binding capacity, a different protocol was adopted. Aliquots of 200 ~1 of cytosol or 200 p.1of plasma (diluted 1:lO with buffer) were incubated in duplicate with 100 ~1 buffer containing increasing amounts of inert testosterone (5-160 nM) and a constant amount of [3H]testosterone (1.25 nM) both with and without a IOO-fold excess of inert testosterone. For both methods samples were incubated overnight MATERIALS AND METHODS (18 hr) at 4”. Fish. One hundred 2 + brown trout (Hungerford Unbound steroid was removed by the addition to strain) were divided equally into two 1500-liter outeach tube of 200$ charcoal suspension (1.25% chardoor fiberglass tanks in July 1986, supplied with a concoal, 0.125% dextran in buffer) followed by incubation stant flow of lake water (35 liters min- ‘). One hundred on ice for 5 min and centrifugation at 4” for 5 min at 1 + fish of the same strain were similarly divided into IOOOg. Aliquots of supernatant were removed, added a further two tanks. The fish were fed once daily with to 5.0 ml scintillant (Unisolve 1, Koch-Light), and commercial feed at a rate of 1% body weight day- ‘. counted under standard 3H conditions. Specific bindSampling and tissue preparation. The fish were ing was calculated from nonspecific and total binding sampled at monthly intervals between August 1986 and the equilibrium dissociation constant (k,)- and and June 1987. At each sampling time eight mature 2 + maximum number of binding sites (N,,,) were calcumales and eight immature 1 + fish were removed. Fish lated according to Scatchard (1949). The binding of were netted from each pond, four at a time and anes- [3H] 1I-ketotestosterone was assessed in an identical thetized in 2-phenoxyethanol (0.5 ml liter-‘). Blood fashion to that of [3H]testosterone. Labeled llwas collected into a heparinized syringe from the cau- ketotestosterone was prepared from [3H]cortisol as da1 vessels; fish were killed by a blow to the head and described previously (Pottinger, 1987). skin was removed from both flanks in 2-cm squares Protein determination. The protein concentration of and placed immediately into buffer (50 mM Tris-HCl, the various fractions was measured by the method of pH 7.4) on ice. A total of -60 cm* per fish was norOhnishi and Barr (1978). mally removed. Tissue was rinsed twice in buffer to Radioimmunoassays. Plasma testosterone and llremove excess mucus, trimmed to remove adherent ketotestosterone levels were measured by radioimmumuscle, and minced finely with scissors. Minced tissue noassay according to the methods of Pottinger and was homogenized in 2 vol of buffer (50 mM Tris-HCl, Pickering (1985a) although antisera of higher specific0.1 mM EDTA, 50 mM dithiothreitol, 10 mM sodium ity were employed allowing omission of the TLC stage molybdate, pH 7.4) on ice using an IKA Ultra Turrax (see Pickering et al., 1987 for details). (TWO) for IO-set bursts with cooling between bursts. Statistical analyses. The concentration of binding Homogenate was centrifuged at 4” for 10 min at 1OOOg sites in plasma, cytosol, and nuclear extract were septo prepare a crude nuclear pellet. Supernatant was trans- arately analyzed by multifactorial analysis of variance ferred to clean tubes and centrifuged at 4” for 1 hr at (ANOVA, Genstat) with maturity (mature/immature) 50,OOOg. The clear supernatant (cytosol) was aspirated and time as factors. From a plot of the residuals and stored at - 70” until required for assay. The pellets against fitted values appropriate transformations (V or containing nuclear material were washed three times log) were selected, where necessary, to improve hoin 10 vol of homogenization buffer and then extracted mogeneity of variance. The levels of significance given
336
T. G. POTTINGER
in the paper are derived from these analyses but for ease of presentation, data are given as arithmetic means + SEM.
RESULTS Hormone levels. During the period August 1986 to June 1987, plasma llketotestosterone levels rose from 5.0 ng ml-’ to reach a peak of 53 .O ng ml- ’ in November, declining to < 2.0 ng ml-’ by March (Fig. la). Levels of testosterone rose from 4.0 ng ml-’ in August to peak in October at 15.0 ng ml- ’ before declining to < l.Ongml-’ in February (Fig. lb). Levels of both hormones in immature fish remained below 1.0 ng ml- ’ throughout the period studied and have not, therefore, been plotted. Plasma steroid-binding capacity. Total
plasma androgen-binding capacity was calculated from the specific binding (mg-’ protein) and the protein concentration in the plasma at each sample. The results are presented in Fig. 2. Both mature and immature fish showed significant changes in total binding capacity with time (P < 0.001) and a significant maturity X time interaction was also noted (P < 0.001). This was resolved as significantly higher binding capacity in immature fish during September and October (P < 0.001) and significantly lower during May (P < 0.05). Overall, plasma-binding capacity in immature fish declined significantly from 360 + 26 nM (n = 8 throughout) in September to a low of 128 * 21 ti in January (P < 0.001) before rising to 258 t 23 ti in March (P < 0.001) and declining to 139 t 17 in May (P <
SEASONAL
VARIATION
IN
ANDROGEN
BINDING
IN
337
TROUT
Month
2. Total testosterone-binding capacity (nM) of brown trout plasma from mature (-O-) immature (- -O- -) fish during the period September to June (Z + SEM, N = 8). FIG.
0.001). A similar, though less marked, pattern was observed in mature fish. Binding capacity declined from 193 t 13 nIL4 in September to 130 * 9 nM in November (P < 0.05) before rising to a peak of 341 -t 33 in March (P < 0.001) and then dropping to 243 + 38 nM in May (P < 0.05). There was no significant difference in affinity of mature
and
or immature plasma for testosterone, overall k,, = 32.6 + 1.0 nM (N = 144). Cytosol-binding capacity. Cytosolbinding capacity varied significantly with time in both mature and immature fish (P < 0.001, Fig. 3). Overall, binding capacity in immature fish declined from 782 5 41 fmol mg-’ protein in September to 292 * 53
1000
c” .z
800
252 g ‘5 00 &h IiO ‘cn
SE
a-gg
600
400
0-
In‘:0
200
0 0
Month
3. Testosterone-binding capacity (IV,,,) of skin cytosol from mature (-04 (- -O- -) brown trout between September and June (F + SEM, n = 8). FIG.
and immature
338
T. G. POTTINGER
fmol mg-’ protein in May (P < 0.01). In mature fish, binding capacity increased from 602 + 57 fmol mg - ’ protein in November to a peak of 944 + 103 fmol mg-’ in March (P -=cO.Ol), thereafter declining to less than 600 fmol mg-’ protein by May. During the period December to May, binding capacity in mature fish was significantly higher than that in immature fish (P < 0.05 - 0.001). There was no significant difference in affinity for testosterone between mature and immature cytosol. Overall, k, = 16.9 + 0.5 n&! (n = 144). Nuclear extract binding capacity. A significant effect of maturity (P < O.Ol), time (P < O.OOl), and a significant maturity x time interaction (P < 0.01) were observed in the nuclear extract binding data (Fig. 4). No measurable change occurred in nuclear binding capacity of immature fish, levels remained in the range 25-75 fmol mg-’ protein throughout the period under study. In mature fish, however, binding capacity increased from 77 _+ 12 fmol mg- ’ protein in September to 269 + 31 fmol mg-’ protein in November (P < 0.001) before declining gradually to a level of 69 It 11 fmol mg - ’ 300
protein in May. Levels of binding in mature fish were significantly higher than in immature fish (P < 0.01-0.001) at all samples except June. There was no difference in K. of nuclear binding between mature and immature fish, kr, = 2.6 * 0.1 ti 9 (n = 144). Binding of 13H]1 1 -ketotestosterone. At no point during the investigation was measurable and reproducible specific binding of [3H] 11-ketotestosterone observed, in any of the fractions studied. DISCUSSION
The hormone profile measured in maturing male brown trout during this experiment resembles closely that previously observed in brown trout (Kime and Manning, 1982; Pottinger and Pickering, 1985a; 1987). Although not directly monitored in the present study, the period during which increased dermal and epidermal thickness and reduced superficial goblet cell concentration are observed occurs between September and April (Pickering, 1977; Pottinger and Pickering, 1985a). Previously, it has been suggested that maturity-related
*
200
100 .
01
,
I
I
I
I
I
I
I
I
I
I
1
A
S
0
N
D
J
F
M
A
M
J
J
Month
4. Testosterone-binding capacity (N,,,) of skin nuclear extract from mature (-O-) immature (--O- -) brown trout between September and June (F + SEM, n = 8). FIG.
and
SEASONAL
VARIATION
IN
ANDROGEN
changes in skin structure were invoked by elevated 1 1-ketotestosterone (1 l-KT) levels, because of both the timing of the changes (Pottinger and Pickering, 1985a) and the ability of exogenous 1 l-KT to induce similar changes in the skin of immature salmonids (Idler et al., 1961; McBride and van Overbeeke, 1971; Pottinger and Pickering, 1985b). However, specific highaffinity binding of 1l-KT, indicative of the presence of an intracellular receptor protein, was not detected in the skin of mature male brown trout (Pottinger, 1987). This observation has been confirmed by the present study. [3H]testosterone (T) was, however, bound by all three fractions examined, plasma, skin cytosol, and skin nuclear extract. The binding capacity of each fraction was found to vary, independently, with time. An increase in measurable binding of 3H-T to nuclear extract occurred in mature fish between September and November. During the same period there was a drop in plasma-binding capacity but no changes were observed in cytosolic binding capacity. Plasma-binding capacity rose from a low point in November to peak in March. In immature fish, no major change was apparent in nuclear or cytosol-binding capacity, they remained low throughout. However, plasma T-binding capacity changed, in immature fish, in a manner similar to that observed in mature fish. There is little information available, for any vertebrate group, on changes during the reproductive period in any of the parameters examined in the present study, other than hormone levels. The blood of vertebrates contains proteins which bind steroids with high affinity and limited capacity (Martin, 1980). A plasma protein specifically binding sex steroids, a presumptive sex hormone-binding globulin (SHBG), has been identified in most higher vertebrates and many lower, including the cod Gadus morhua (Freeman and Idler, 1971), Atlantic salmon Salmo salar (Freeman and Idler, 1971; Lazier et al., 1985).
BINDING
IN TROUT
339
rainbow trout Salmo gairdneri (Fostier and Breton, 1975), and brown trout (Pottinger, 1987). Although none of these studies provide information on temporal changes in binding capacity, in some mammals higher concentrations of SHBG are found in immature than mature animals (Lobl, 1981), for instance there is an inverse relationship between SHBG and plasma testosterone levels in the marmoset monkey (Hodges et al., 1983). In reptiles and amphibians so far studied the relationship between plasma steroid levels and binding capacity is less clear. In female, though not male, alligators SHBG declines during the breeding season (Ho et al., 1987). In male newts, Taricha granulosa, no change in SHBG-binding capacity occurs during the reproductive period (Moore et al., 1983). During the present investigation, plasma T-binding capacity was lowest during the period OctoberJanuary, coincident with peak androgen levels and spawning activity, and highest in March when androgen levels had returned to basal levels. This conforms to a broad pattern of an inverse relationship between binding capacity and androgen levels. The range of plasma-binding capacity observed during the present investigation (- 100-400 nM) compares with a range of binding capacity reported for other vertebrates extending from 23-45 nM in the amphibian Taricha granulosa (Moore et al., 1983) to 2-8 $V in female rainbow trout (Fostier and Breton, 1975). This value is markedly higher than that observed in male brown trout during the present investigation. However, it has been noted that absolute plasma-binding capacity is directly related to maximum plasma steroid levels in a given species (Salhanick and Callard, 1979). In male brown trout T levels rarely exceed 20 ng ml- ‘, in female rainbow trout levels up to 400 ng ml - ’ have been reported (Scott and Sumpter, 1983). In functional terms it has been suggested that plasmabinding proteins protect circulating steroids from inactivation (by conjugation and clear-
340
T.
G.
POTTINGER
ante) and may provide a delivery system to target tissues (Martin, 1980). An alternative view suggests that binding proteins inhibit the biological effectiveness of steroids (Moll and Rosenfield, 1986 and refs therein), an interpretation supported by Fostier and Breton (1975) working on rainbow trout. In the present investigation although significant decline in T-binding capacity, to a level of 130 nM, occurs during the spawning period, this capacity still exceeds the maximum plasma levels of total T (15 ng ml-’ - 50 nM) occurring during this period. The binding protein has little affinity for the other major androgen, ll-KT (Fostier and Breton, 1975; Pottinger, 1987), therefore relatively little ll-KT would be bound, even in the presence of maximal SHBG capacity. Androgens have been implicated in the control of SHBG levels in mammals (see Mooradian et al., 1987) but plasma androgen-binding capacity of immature brown trout fluctuates in an almost identical manner to that of mature fish and throughout the period studied plasma androgens in immature fish remained at a level below 1.O ng ml. The plasma androgen-binding system in brown trout would therefore appear to be controlled by factors outside the pituitary-gonadal axis. Binding of 3H-T to the nuclear fraction of brown trout skin increased nearly four-fold during the period September to November. Previous work has established this to be a period of maximal increase in dermal and epidermal thickness (Pottinger and Pickering, 1985a). Whether the increase in binding sites represents a de nova synthesis of receptor or an increase in receptor occupancy leading to elevated nuclear localization is not clear. It has been shown that intact steroid-receptor complexes are more likely to remain within the nucleus during homogenization than unoccupied receptors (Welshons et al., 1984). However, the number of nuclear-binding sites did not increase until September-October, T levels had already begun to rise by August which suggests oc-
cupancy alone may not explain the increase in measurable binding sites. No attempt was made in the present study to distinguish between occupied and unoccupied nuclear-binding sites. The complete exchange of endogenous (unlabeled) ligand for exogenous tracer is reported to require between 18 hr in the mouse kidney (Isomaa et al., 1982) to 48 hr in the mouse submandibular gland (Kyakumoto, 1986). It is therefore possible that exchange was not completed during the incubation period of 18 hr employed in the present study. This may have resulted in an underestimation of the number of binding sites during the period of maximum plasma T levels. A further factor to be considered is whether the extraction regime employed in this study was sufficiently rigorous to remove the majority of binding sites from the nucleus. Although nonextractable binding sites have been reported to occur in androgen target organs (Barrack et al., 1983) these may represent as few as 5% of the total androgen-receptor content of a tissue (Amet et al., 1986). There was a significant increase in cytosolic binding sites between November and March (P < 0.01) in mature male fish coinciding with the decline in nuclear binding. However, there was no decline in the number of cytosol-binding sites coinciding with the increase in nuclear-binding sites, the affinity of cytosol sites for T was consistently lower than that of nuclear extract (k, = 16.9 I-&! c.f. k, = 2.7 nM), and Scatchard plots gave no indication that cytosol contained two discrete populations of binding protein. It seems unlikely that the increase in nuclear binding is simply a translocation of binder from the cytosol compartment. Although peak 11-KT levels coincide with peak nuclear-binding capacity, the binder has little measurable affinity for 1 lKT (Pottinger, 1987). Regulation of a receptor for one hormone by another is known to occur in various mammalian systems, most notably in some estrogen sensitive tissues which also contain progesterone receptors.
SEASONAL
VARIATION
IN
ANDROGEN
An increase in progesterone receptor numbers can be induced by estrogen administration (Muldoon, 1980). It might, therefore, be suggested that in the present study close correlation of the increase in nuclear testosterone binding with plasma 1 l-KT levels reflects a causal relationship. However, if such were the case, specific 1 l-KT binding ought to exist at some stage during the reproductive cycle. More widely observed is a direct link between plasma hormone levels and receptor concentration in target tissue. Castration results in a drop in androgen-binding sites and T administration replenishes binding sites in rat epididymis (Tezon and Blaquier, 1983), androgen treatment increases binding sites in cultured human epididymis (Vazquez et al., 1986), and has a similar effect in rat ventral prostrate (Blondeau et al., 1982). An analogous system has been observed in other vertebrates. Androgen-binding sites in the larynx of male frogs, Xenopus Zaevis, decline following castration and increase after administration of T (Segil et al., 1987). Androgen binding in the uropygial gland of castrated quail is increased by the administration of T (Amet et al., 1986). This effect was also observed in the thumb pad of castrated and androgen-treated Rana escuZenta (Delrio et al., 1980). The only information available on fish refers to the hepatic estrogen receptor of SaZmo salar, the concentration of which increases in response to estrogen administration (Lazier et al., 1985). In the present investigation, peak plasma T levels precede peak Tbinding capacity by one month. Administration of T and ll-KT to immature fish would clarify which steroid actively increases the number of T-binding sites in the skin. The binding of T to the cytosolic fraction of skin is curious by comparison to other systems examined. Scatchard analysis suggests only one population of binding sites is present with a kt, (16.9 nA4) intermediate between that of plasma (k, = 32.6 nM) and
BINDING
IN
TROUT
341
nuclear extract (k, = 2.6 nM). This binding appears not to be a plasma contaminantdespite large fluctuations in plasma-binding capacity, cytosol-binding levels remain relatively constant. Similar lower affinity cytosol-binding proteins have been reported in mammals (Dekloet and McEwen, 1976; Koch et al., 1976) and turtle (Salhanick and Callard, 1979). This differs markedly from other systems where cytosol and nuclear binding are functionally indistinguishable (e.g., Lazier et al., 1985; Amet et al., 1986; Kleis-San Francisco and Callard, 1986). It is possible that lower affinity cytosolic binding proteins are involved in maintaining local concentrations of steroids and promoting delivery to the nuclear compartment (Salhanick and Callard, 1979). In conclusion, the period during which brown trout skin thickens and demucifies coincides with elevated plasma androgen levels, an increase in salt-extractable nuclear androgen binding in the skin and a decline in plasma androgen-binding capacity. Although both T and ll-KT, when administered to immature fish, stimulate epidermal and dermal thickening (Pottinger and Pickering, 1985a) throughout the spawning period, specific binding exist within the skin for T only. This suggests therefore that, as a target tissue, skin is equipped to respond to T rather than llKT. The large reduction in mucus-secreting goblet cells within the epidermis, which is a feature of maturation (Pottinger and Pickering, 1985a) and can be induced in immature fish by 1 l-KT but not T (Pottinger and Pickering, 1985b), may represent a mechanism ultimately under the control of 1 l-KT but one in which the steroid acts at a site outside the integument. For example, prolactin is known to stimulate goblet cell differentiation in teleost epidermis (Ogawa, 1970; Mattheij et al., 1972; Bonga, 1978; Schwerdtfeger, 1979), further work may indicate whether 1l-KT has a suppressive effect on prolactin secretion. The paradox re-
342
T. G. POTTINGER
mains, however, that 1 l-KT is also potent in stimulating epidermal and dermal thickening. Conversion of 1 l-KT within the skin is a possibility yet to be investigated. Relatively few steroids have been tested for potency in displacing bound [3H]T (Pottinger, 1987), an as yet unidentified metabolite of 1 l-KT may prove to be active at the same binding sites as T. In the absence of comparable data from other species it is difficult to present a functional interpretation of skin cytosol and plasma binding. The drop in plasma binding capacity during the period of elevated androgen levels, if its effect is to potentiate androgen action by increasing the proportion of unbound steroid circulating, suggests the plasma steroid-binding protein may be an important factor modulating hormone activity. The cytosol binder, apparently distinct from both nuclear and plasma binders and with an affinity for T intermediate between the two, may be involved in maintaining a high intracellular concentration of steroid. ACKNOWLEDGMENTS The author thanks Dr. A. D. Pickering for helpful discussion and critical comments on the manuscript, Janet Pollard and Janice Fletcher for maintaining the experimental fish, and Trevor Furnass for preparing the illustrations. This work was financed by the Natural Environment Research Council.
Blondeau, J.-P., Baulieu, E.-E., and Robel, P. (1982). Androgen-dependent regulation of androgen nuclear receptor in the rat ventral prostrate. Endocrinology
Endocrinol.
Amet, Y., Abalain, J.-H., di Stefano, S., Daniel, J.-Y., Tea, K., Floch, H. H., and Robel, P. (1986). Androgen regulation of the androgen receptor of the quail uropygial gland: Application of a [3H] mibolerone exchange assay. J. Endocrinol. 109, 299306. Barrack, E. R., Bujnovszky, P., and Walsh, P. C. (1983). Subcellular distribution of androgen receptors in human normal, benign hyperplastic and malignant prostatic tissues. Characterization of nuclear salt resistant receptors. Cancer Res. 43, 1107-1113. Baulieu, E.-E. (1979). Aspects of steroid hormonetarget cell interactions. Adv. Exp. Med. Biol. 117, 377-399.
34, 265-275.
De Kloet, E. R., and McEwen, B. S. (1976). A putative glucocorticoid receptor and a transcortin-like macromolecule in pituitary cytosol. Biochim. Biophys.
Acta
421, 115-123.
Delrio, G., Citarella, F. and D’Istria, M. (1980). Androgen receptor in the thumb pad of Rana esculenta: Dynamic aspects. J. Endocrinol. 85, 279282. DuBois, D. C., and Almon, R. R. (1984). Perineal muscles: Possible androgen regulation of glucocorticoid receptor sites in the rat. J. Endocrinol. 102, 225-229.
Fostier, A., and Breton, B. (1975). Binding of steroids by plasma of a teleost: The rainbow trout, Salmo gairdneri.
J. Steroid
Biochem.
6, 345-35
1.
Fostier, A., Jalabert, B., Billard, R., Breton, B., and Zohar, Y. (1983). The gonadal steroids. In “Fish Physiology” (W. S. Hoar, D. J. Randall, and E. M. Donaldson, Eds.), Vol 9A, pp. 277-372. Academic Press, New York. Freeman, H. C., and Idler, D. R. (1971). Binding affinities of blood proteins for sex hormones and corticosteroids in fish. Steroids 17, 233-250. Ho, S.-M., Lance, V., and Megaloudis, M. (1987). Plasma sex-steroid binding protein in a seasonally breeding reptile, Alligator mississippiensis. Gen. Comp. Endocrinol. 65, 121-132. Hodges, J. K., Eastman, S. A. K., and Jenkins, N. (1983). Sex steroids and their relationship to binding proteins in the serum of the marmoset monkey (Callithrix
REFERENCES
110, 19261932.
Bonga, S. E. W. (1978). The effect of changes in external sodium, calcium, and magnesium concentrations on prolactin cells, skin, and plasma electrolytes of Gasterosteus aculeatus. Gen. Camp.
jacchus).
J. Endocrinol.
96, 443-450.
Idler, D. R., Bitners, I. I., and Schmidt, P. J. (1961). 1 I-ketotestosterone: An androgen for sockeye salmon. Canad. J. Biochem. Physiol. 39, 17371742. Idler, D. R., and Ng, T. B. (1983). Teleost gonadotropins: Isolation, biochemistry and function. In “Fish Physiology” (W. S. Hoar, D. J. Randall, and E. M. Donaldson, Eds.) Vol9A, pp. 187-221. Academic Press, New York. Isomaa, V., Pajunen, A. E. I., Bardin, C. W., and Janne, 0. A. (1982). Nuclear androgen receptors in the mouse kidney: Validation of a new assay. Endocrinology
111, 833-843.
Kime, D. E., and Manning, N. J. (1982). Seasonal patterns of free and conjugated androgens in the brown trout Salmo trutta. Gen. Comp. Endocrinol. 48, 222-23 1.
SEASONAL
VARIATION
IN
ANDROGEN
Kleis-San Francisco, S. K., and Callard, I. P. (1986). Progesterone receptors in the oviduct of a viviparous snake (Nerodiu): Correlations with ovarian function and plasma steroid levels. Gen. Camp. Endocrinol.
63, 220-229.
Koch, B., Lutz, B., Briaud, B., and Mialhe, C. (1976). Heterogeneity of pituitary glucocorticoid binding evidence for a transcortin-like compound. Biochim.
Biophys.
Acta
444, 497-507.
Kyakumoto, S., Kurokawa, R., Ohara-Nemoto, Y., and Ota, M. (1986). Sex difference in the cytosolic and nuclear distribution of androgen receptor in mouse submandibular gland. J. Endocrinol. 108, 267-273. Lazier, C. B., Lonergan, K., and Mommsen, T. P. (1985). Hepatic estrogen receptors and plasma estrogen-binding activity in the Atlantic salmon. Gen.
Comp.
Endocrinol57,
234-245.
Lobl, T. J. (1981). Androgen transport proteins: Physical properties, hormonal regulation and possible mechanism of TeBG and ABP action. Arch. Androl. 7, 133-151. Martin, B. (1980). Steroid-protein interactions in nonmammalian vertebrates: Distribution, origin, regulation, and physiological significance of plasma steroid binding proteins. In “Steroids and Their Mechanism of Action in Nonmammalian Vertebrates” (G. Delrio and J. Brachet, Eds.) pp. 63-73. Raven Press, New York. Mattheij, J. A. M., Stroband, H. W. J., Kingma, F. J., and van Oordt, P. G. W. J. (1972). Prolactin and osmoregulation in the cichlid fish, Cichlasoma biocellatum, and the effect of this hormone on the thyroid, gills, and skin. Gen. Comp. Endocrinol
18, 607.
McBride, J. R., and van Overbeeke, A. P. (1971). Effects of androgens, estrogens and cortisol on the skin, stomach, liver, pancreas and kidney in gonadectomized adult sockeye salmon (Oncorhynthus
nerka).
J. Fish Res. Board
Canad.
28, 485-
490. Moll, G. W., and Rosenfield, R. L. (1986). Estradiol inhibition of pituitary luteinizing hormone release is antagonized by serum proteins. J. Steroid Biothem.
25, 309-314.
Mooradian, A. D., Morley, J. E., and Korenman, S. G. (1987). Biological actions of androgens, Endoer. Rev. 8, l-28. Moore, F. L., Spielvogel, S. P., Zoeller, R. T., and Wingfield, J. (1983). Testosterone-binding protein in a seasonally breeding amphibian. Gen. Comp. Endocrinol. 49, 15-2 1. Muldoon, T. G. (1980). Regulation of steroid hormone activity. Endocr. Rev. 1, 339-364. Ogawa, M. (1970). Effects of prolactin on the epiderma1 mucous cells of the goldfish Carassius auratus L. Canad. J. Zool. 48, 501-503.
BINDING
IN TROUT
343
Ohnishi, S. T., and Barr, J. K. (1978). A simplified method of quantitating protein using the biuret and phenol reagents. Anal. Biochem. 86, 193-200. Pickering, A. D. (1977). Seasonal changes in the epidermis of the brown trout, Salmo trutta L. J. Fish Biol.
10, 561-566.
Pickering, A. D., Pottinger, T. G., Carragher, J., and Sumpter, J. P. (1987). The effects of acute and chronic stress on the levels of reproductive hormones in the plasma of mature male brown trout, Salmo trutta L. Gen. Comp. Endocrinol. 68, 249259. Pottinger, T. G. (1986). Estrogen-binding sites in the liver of sexually mature male and female brown trout, Salmo trutta L. Gen. Comp. Endocrinol. 61, 120-126. Pottinger, T. G. (1987). Androgen binding in the skin of mature male brown trout, Salmo trutta L. Gen. Comp.
Endocrinol.
66, 224-232.
Pottinger, T. G., and Pickering, A. D. (1985a). Changes in skin structure associated with elevated androgen levels in maturing male brown trout, Salmo trutta L. J. Fish Biol. 26, 745-753. Pottinger, T. G., and Pickering, A. D. (1985b). The effects of 1I-ketotestosterone and testosterone on the skin structure of brown trout, Salmo trutta L. Gen.
Comp.
Endocrinol.
59, 335-342.
Pottinger, T. G., and Pickering, A. D. (1987). Androgen levels and erythrocytosis in maturing brown trout, Salmo trutta L. Fish Physiol. Biochem. 3, 121-126. Salhanick, A. R., and Callard, I. P. (1979). Sex steroid binding proteins in non-mammalian vertebrates. Adv.
Exp.
Med.
Biol.
117, 441-459.
Sanborn, B. M., Wagle, J. R., and Steinberger, A. (1984). Control of androgen cytosol receptor concentrations in sertoli cells: Effect of androgens. Endocrinology
114, 2388-2393.
Scatchard, G. (1949). The attractions of proteins for small molecules and ions. Ann. N.Y. Acad. Sci. 51, 660-672.
Schwerdtfeger, W. K. (1979). Morphometrical studies of the ultrastructure of the epidermis of the guppy, Poecilia reticulata Peters, following adaptation to seawater and treatment with prolactin. Gen.
Comp.
Endocrinol.
38, 476-483.
Scott, A. P., Bye, V. J., and Baynes, S. M. (1980a). Seasonal variations in sex steroids of female rainbow trout (Salmo gairdneri Richardson). J. Fish Biol.
17, 587-592.
Scott, A. P., Bye, V. J., Baynes, S. M., and Springate, J. C. (1980b). Seasonal variations in plasma concentrations of 1 1-ketotestosterone and testosterone in male rainbow trout (Salmo gairdneri Richardson). J. Fish Biol. 17, 495-505. Scott, A. P., and Sumpter, J. P. (1983). A comparison of the female reproductive cycles of autumnspawning and winter-spawning strains of rainbow
344
T. G. POTTINGER
trout (Salmo gairdneri Richardson). Gen. Comp. Endocrinol. 52, 79-85. Segil, N., Silverman, L., and Kelley, D. B. (1987). Androgen-binding levels in a sexually dimorphic muscle of Xenopus laevis. Gen. Comp. Endocrinol. 66, 95-101. Smith, K, and Shuster, S. (1984). Effect of adrenalectomy and steroid treatment on rat skin cytosol glucocorticoid receptor. J. Endocrinol. 102, 161165.
Tezon, 3. G., and Blaquier, J. A. (1983). Androgens control androgen-binding sites in rat epididymis. Endocrinology 113, 1025-1030. Vazquez, M. H., de Larminar, M. A., and Blaquier, J. A. (1986). Effect of androgen on androgen receptors in cultured human epididymis. J. Endocrinol. 111, 343-348. Welshons, W. V., Lieberman, M. E., and Gorski, J. (1984). Nuclear location of unoccupied oestrogen receptors. Nature (London) 307, 747-749.