GENERAL
ANDCOMPARATIVE
ENDOCRINOLOGY
s&252-258(1985)
Functional
Specificity of Gonadotropin Fundulus heteroclitusl
CHRISTOPHER
L. BROWN,* E. GORDON GRAU,~ ANDMILTON
Physiology
Section,
School
of Life
and Health
Sciences,
University
and Thyrotropin
of Delaware,
in
H. STETSON~ Newark,
Delaware
19716
Accepted July 25, 1984 Gonadotropic hormones (GTHs) and thyrotropic hormones (TSHs) reportedly bear close evolutionary and structural relationships, and the thyroid appears to be active in reproduction in some fish species. We tested the sensitivity of the thyroid of Fund&s heteroclitus to glycoprotein hormones from mammalian and piscine sources. Six mammalian glycoprotein hormones, including four gonadotropins and two thyrotropins, produced dose-dependent elevations in serum thyroxin. A release of endogenous gonadotropins was elicited by injecting GnRH. This resulted in gonadal stimulation, with no alteration in circulating thyroxin levels and the rate of radioiodine uptake. We also treated fish with partially purified salmon gonadotropin (SG-GlOO). The gonadotropic actions of this extract were confirmed by steroid elevations, and again T4 and lz51 uptake remained at resting levels. The lack of response of the thyroid gland to fish gonadotropins suggests that TSH receptors in Fundulus heteroclitus can differentiate between endogenous thyrotropin and gonadotropin(s), even though most heterologous glycoprotein hormones are thyrotropic. Q I985 Academic PXSS, IN.
A method of extracting glycoprotein hormones from bovine pituitary glands was described by Fontaine (1963). Two such extracts were found to have thyrotropic activity. One, believed to be bTSH, was stimulatory to both mammalian and teleost thyroids; the second was only active in the teleost bioassays. Fontaine initially called the latter extract “bovine heterothyrotropic factor” (HTF). The observation that HTF was capable of activating both the thyroid and gonads of hypophysectomized Fundulus heteroclitus (Pickford and Grant, 1968) was consistent with Fontaine’s (1969a) conclusion that the HTF fraction was composed of bovine gonadotropins. An assortment of mammalian gonadotro’ Supported by NSF Research Grants PCM78-06684 and PCM81-11384. Portions of this work have been published in abstract form (Grau and Stetson, 1977c; Brown et al., 1982). 2 Current address: Department of Zoology, University of California, Berkeley, Calif. 3 Current address: Department of Zoology, University of Hawaii, Honolulu, Hawaii. 4 To whom reprint requests should be sent. 252 0016-6480185 $1.50 Copyright 0 1985 by Academic Press. Inc. All rights of reproduction in any form reserved.
pins and thyrotropins have since been shown to stimulate thyroxin release in F. heteroclitus (Grau, 1978). In fact, purified gonadotropins representing all four classes of tetrapods have intrinsic thyrotropic potency when administered to the teleost Gillichythys mirabilis (MacKenzie, 1982). A parallel situation is seen in birds and reptiles that have a heterothyrotropic response to frog LH (MacKenzie et al., 1978; MacKenzie, 1980). Oddly, the degree of response to foreign glycoprotein hormones seldom reflects phylogenetic relationships (Licht et al., 1977). The responsiveness of the teleost thyroid to various gonadotropins raises interesting evolutionary questions. It has been postulated that LH, FSH, and TSH arose from a single ancestral glycoprotein by way of a series of three gene duplications (Fontaine and Burzawa-Gerard, 1977). It is not yet known at what stage of vertebrate evolution these hormones emerged structurally and began to assume divergent functions. Structural similarities in the g subunits seem largely responsible for recognition of
GTH-TSH
SPECIFICITY
heterologous gonadotropins by gonadotropin receptors (Fontaine-Bertrand et al., 1981), and may be important in the heterothyrotropic actions of these hormones. Sage (1973) surmised that TSH may have been derived from a gonadotropin, and that the thyroid system may have originally had a role in reproduction. Perhaps the established cases of thyroid involvement in fish reproduction (e.g., Detlaff and Davydova, 1979; Chakraborti et aE., 1983) represent vestiges of primitive thyroid function. Endogenous gonadotropins may have retained some thyrotropic actions in fish. This could account for the responsiveness of the teleostean thyroid to all gonadotropins tested, as opposed to the mammalian thyroid gland, which seems to respond only to mammalian TSH (Fontaine, 1969b). Most studies on possible thyrotropic actions of fish gonadotropins have employed heterologous GTHs, and the results have been mixed. Fontaine (1969b) treated trout with a carp GTH extract and noted no change in 13110dine uptake. Donaldson and McBride (1974) injected partially purified chinook salmon gonadotropin (SG-GlOO) into sockeye salmon and reported histological signs of thyroid stimulation, due either to TSH contamination or an intrinsic property of the gonadotropin. Questions about the purity of glycoprotein hormone extracts were also raised by Ng et al. (1982), who found that in general it was possible to purify independently gonadotropic and thyrotropic fractions from carp, salmon, plaice, and flounder pituitaries, as tested in a flounder bioassay. There was some overlap of function in a few extracts, though, such as the flounder “62K DEIII CMII” fraction, which caused increases in both T3 levels and vitellogenesis. In this ,paper we examine the gonadal and/or thyroid response of Fundulus hetevoclitus to glycoprotein hormones of mammalian origin, a teleo’st gonadotropin preparation and endogenous gonadotropin.
253
IN FISH
MATERIALS
AND METHODS
Procurement and handling offish. Adult Fund&s heteroclitus of both sexes (5- 10 g body weight) were trapped in baited minnow traps in the tidal waters of Canary Creek, Lewes, Delaware. Fish were brought back to the laboratory and were held in a 1.5 k liter pool containing 20%~ synthetic seawater (Forty Fathoms, Marine Enterprises, Timonium, Md.) at 20 +- 1”. Tetramin was fed daily, and a photoperiod of LD 14:lO was maintained. After 1 week of laboratory acclimation, fish (N = 6- 10) were transferred to IO-liter plastic aquaria for experimentation. Experiment 1: heterothyrotropic agents. Six different glycoprotein hormones of mammalian origin were administered to fish (collected in midsummer) to test their thyrotropie potency. These included bovine TSH (NIH-TSH-BS), ovine TSH (NLH-TSH-S8), ovine LH (NIH-LH-S16), ovine FSH (NIH-FSH-SS), human chorionic gonadotropin (Sigma GC-2), and pregnant mare serum gonadotropin (Sigma G-4877). Each was injected intraperitoneally in 0.06 ml of a 0.6% NaCl solution. After four daily injections (@ 1600 hr), fish were sacrificed on the morning of the fifth day and serum was separated by centrifugation and stored frozen. Serum thyroxin was measured by RIA (validated by Grau and Stetson, 1977a). A stimulatory range of dosages was determined for each hormone after determining the minimal dose capabie of elevating T, levels over those of saline-injected controls. Since the units of potency are based on bioassay activity and are arbitrarily established for each hormone, we could not compare hormone dosages directly. We therefore expressed the degree of thyroidal response ([TJ) to increasing dosages of these compounds, as compared to the minimally stimulatory dose found for each (Fig. 1). The quantities of hormones used were as follows: oTSH, 100 kg-400 p+g: bTSH, 200 pg-1.2 mg; oFSH, 5 Kg-180 pg; oLH, 1 pg- 1.8 mg: PMSG, 0.1 pg- 100 pg. The minimal dose of HCG used was 25 IU (weight units were not provided with this hormone). Thirty-seven to 58 fish were used to generate each response curve in Fig. I. The slopes of the dose-response curves (rendered linear by using a log scale) were compared by Iinear regression and correlation. Experiment 2: elevation of endogenous GTH. Fish were collected for this experiment in the first week of March, 1981, shortly before the onset of the breeding season. Groups of fish were injected ip with ,synthetic GnRH (Calbiochem, San Diego, Calif.) in total doses of 1 and 4 p,g, using a 4-day injection par&d&m like that used in Experiment 1. One group of hypophysectomized fish was treated with the larger dose; of GnRH (4 kg/4 days). Controls were injected with 0.6% saline. At the time of sacrifice and serum collection, the number of male fish exhibiting nuptial coloration and
254
BROWN,
GRAU, AND STETSON
I 2 0.6 8 0.5 t
H
/ FIG. 1. Semi-logarithmic plots of the (log) dose-response curves for several glycoprotein hormones. Hormones are indicated as follows: oTSH, solid circles; oFSH, solid squares; oLH, triangles; HCG, open squares; bTSH, asterisks; PMSG, open circles. Each point represents the mean response of LO-12 fish. female fish with ovulated eggs were noted for each group. Serum T4 was measured by RIA. Radioiodine uptake was measured (as per Grau and Stetson, 1977b) in GnRH-treated fish (4 pg14 days) and saline-injected control fish. Experiment 3: effects of SG-GlOO. Partially purified salmon gonadotropin (SG-GlOO) was injected ip at doses of 1 and 10 pg in 0.06 ml 0.6% NaCliday for 4 days. Other groups were given four injections of oTSH (NIH-TSH-S9, 0.2 IU total) or the saline vehicle. The effects of these treatments on thyroid function were assessed by measuring serum T, (RIA) and radioiodine uptake. The significance of the radioiodine uptake data was determined by analysis of variance of GnRH, SGGlOO, and saline-treated controls. As this experiment was conducted in the fall, after the spawning season, it was necessary to use indicators of gonadal response other than the signs of induced spawning seen in Experiment 2. Radioimmunoassays for estradiol (validated by Bradford, 1983) and testosterone (validated by Donham, unpublished) were used. Testosterone was assayed by RIA after serum extraction with methylene chloride. A commercial antiserum preparation (Wien Antibodies, Inc., Succasunna, NJ.) was used and the bound and free testosterone were separated by dextran-coated charcoal. Accuracy was determined by addition of testosterone, ranging from 25 to 300 pg, to charcoal-stripped serum. The equation of the regression line, fitted to 21 points, was y = 0.94.x -+ 9.0, the correlation coefficient (r) was 0.94. The efficiency of extraction, as measured by recovery of added tritiated testosterone to each sample was 84.5 2 0.72% (n = 70). The coefficient of variation of 17 replicates of a Fundulus heteroclitus serum pool (sample size: 10 ~1) measured in a single
H G,nRH
WI
c
G12yGEy
FIG. 2. The effect of synthetic GnRH on serum thyroxin. H, Hypophysectomized fish; C, saline-injected controls. N = 6- 10 per group. The dosages of GnRH indicated are the total doses administered over 4 days. Vertical lines represent standard error of the mean. Fifty to sixty percent of the pituitary-intact fish given GnRH (either dose) showed ovulation (females) or nuptial coloration (males). assay was 12.8%. All samples were measured in one assay and so were not affected by interassay variation. Serum testosterone and estradiol measurements were compared between experimental groups by analysis of variance and Duncan’s multiple range test.
RESULTS Experiment 1. In Figure 1, the natural logarithms of the doses of the glycoproteins used are plotted against the thyroidal response (serum T4) they elicited. Within the range of greatest sensitivity, the dosemean response for all treatments was essentially linear when plotted in this fashion. This linearity is reflected in the correlation coefficients (r2) generated for each treatment. They were 0.97 for bTSH, 0.99 for oTSH, 0.99 for oLH, 0.99 for oFSH, 0.86 for HCG, and 0.90 for PMSG. The slope of the regression line generated for oTSH is significantly greater than that of any other treatment (P < 0.01). The remaining treatments (bTSH, oLH, oFSH, PMSG, and HCG) produced similar dose-responses. These data indicate that the thyroid of this species responds more readily to oTSH than to the other hormones tested, but that it is responsive to all available glycopro-
GTH-TSH
C
SPECIFICITY
TSH
FIG. 3. Effects af two different doses of partially purified salmon gonadotropin (SG) on serum estradiol, as compared to saline-injected controls (C), and ovine TSH injected fish. Shaded bars represent levels in male fish and open bars represent levels in females. Vertical lines depict SEM. N = 6-10 per group.
teins. No such responses have been seen in Fund&s given saline injections (0.06 ml of 0.6% NaCl; for examples see Grau and Stetson, 1977a, 1979). Experiment 2. One and 4 Fg of GnRH promoted egg deposition or nuptial coloration in a majority (50-60%) of specimens. Saline-injected controls and GnRH-treated hypophysectomized fish showed no indications of nuptial coloration or ovulation. Serum thyroxin levels for all treatment groups are shown in Fig. 2. Hypophysectomy produced the expected decline in serum T,, which was unaffected by GnRH. Likewise, GnRII was without a measurable effect on serum T, in intact fish, as compared to the saline controls. The T, levels in controls are typical of resting T, levels commonly observed in this species (for example, see Grau and Stetson, 1977a, b). Radioiodine uptake was not affected by GnRH treatment. Experiment 3. The gonadotropic potency of SG-GlOO at a dosage of 4 kg/fish/4 days was confirmed by significant elevations (P < 0.01) in serum gonadal steroid levels in both sexes. Interestingly, the resting concentrations of both estradiol (Fig. 3) and testosterone (Fig. 4) were equal in the two sexes’. Highly significant increases (P < 0.01) in serum’ estradiol were induced by SG-Cl00 treatment at this dosage, and a
255
IN FISH
significant (P < 0.01) sex dependence is seen. A lo-fold increase in the SG-Cl00 dose did not cause a further increase in the estrogen titre. Ovine TSH had no significant effect on serum estradiol levels in either sex, but caused significant elevations of serum testosterone in both (P < 0.05). Treatment with 4 pg SC-G100 produced testosterone elevations comparable to those obtained with oTSH, and any apparent sex differences were insignificant. A 10x increase in SG-G100 dosage further elevated testosterone levels, but some sexrelated differences became apparent, although at this dosage only six samples (four female, two male) had ample serum remaining for testosterone assay after,the other assays. SG-GlOO had no significant effect on radioiodine uptake or thyroxin release (Fig. 5). DISCUSSfON
The thyroid gland of F. heteroclitus responded readily to every glycoprotein of mammalian origin with which it was challenged. The system appeared to be far more sensitive to oTSH than to any of the other glycoproteins. The dose-dependent -thyrotropic actions of four mammalian gonadotropins in this teleost lit the pattern first reported by Fontaine (1969a). Experiments with heterologous hormones, however, give unpredictable indications of the actions of the homologous hormones (discussed at length by Licht et al., 1977). Since this species showed such classic heterothyrotropic responses, we considered it suitable for experimentation directed at related issues left unresolved by earlier investigations. Foremost among these was the possibility that endogenous gonadotropins in fish might have some inherent thyrotropic activity. Both doses of G&H tested proved capable of promoting sufficient gonadotropin release to induce ovulation in females and nuptial coloration in males in a majority (50-60%) of specimens. This experiment (2) was conducted with fish that were already approaching a reproductively active
256
BROWN,
GRAU, AND STETSON
FIG. 4. Effects of partially purified salmon gonadotropin (SG) on serum testosterone levels in male (shaded bars) and female fish (open bars). Saline controls (C) and oTSH-treated groups are also shown. Vertical lines depict SEM. N = 6-10 per group.
state; short-term treatment with GnRH at other times of the year is less effective. Synthetic GnRH has been shown to cause a rapid release of gonadotropin from trout pituitaries both in viva (Crim and Cluett, 1974) and in vitro (Crim and Evans, 1980); the latter effect is markedly enhanced in steroid-primed fish. The failure of any of our saline-injected controls to display nuptial coloration or ovulation confirms that the spawning of the GnRH-injected groups was induced artificially. That this gonadal stimulation involved gonadotropin release is reinforced by the lack of response of hypophysectomized fish to GnRH. Serum thyroxin and ‘2SIodine uptake remained at basal levels in GnRH-stimulated fish, indicating that the elevated levels of endogenous gonadotropin had no major effect on the thyroid system. This is consistent with our observation that fish collected at various times in the natural (semilunar) spawning cycle maintain uniformly basal T, and T, levels and similar radioiodine uptake capacity irrespective of reproductive status (unpublished). These results suggest that there is a high degree of discrimination between endogenous TSH and GTH by the TSH receptors in this species. Independence of function at the hypothalamic level is also inferred, at least as far as GnRH is concerned. Peter (1970) observed that lesioning the nucleus lateralis tuberis (NLT)
c
oTSH
S G-G100 4Ki WJg
FIG. 5. Effects of partially purified salmon gonadotropin in two doses (SG) on serum thyroxin, as compared to saline-injected controls (C) and oTSH-treated fish. Vertical lines depict SEM. N = 6-10 per group.
pars posterior influenced both the thyroid and reproductive systems in the goldfish. This hypothalamic nucleus is reported to be particularly rich in immunoreactive GnRH in Xiphophorus (Schreibman et al., 1979). There is strong evidence that the primary control of TSH release in teleosts is inhibitory (Peter, 1970; Grau and Stetson, 1977a), but there is still disagreement on the chemical nature of the compound(s) involved. The unchanged indicators of thyroid function in our GnRH-treated fish provide evidence against a major role of this peptide in the regulation of thyrotropin release (though native GnRH may differ biologically from this synthetic compound). Partially purified salmon gonadotropin (SG-GlOO) clearly had stimulatory effects on the gonads of both sexes of Fundulus. Treatment of females with forty micrograms of this extract produced increases in both circulating estradiol and testosterone. Males treated similarly also showed increases in both steroids. The observed elevations in testosterone in the females and estradiol in the males are not without precedent; naturally spawning plaice (Pleuronectes) display both phenomena (Wingfield and Grimm, 1977). The gonadotropic effect of oTSH (statis-
GTH-TSH
SPECIFICITY
tically significant serum testosterone elevation, Fig. 4) may have been due to contamination with ovine gonadotropins. According to NIH specifications, the LH and FSH content of oTSH injections delivered to fish amounts to under 0.25 pg per injection, which is below the dosage necessary for demonstration of heterothyrotropic effects (see Grau and Stetson, 1979). This is approximately 25% of the lowest dose of SG-GlOO (on a weight:weight basis) used here to demonstrate gonadotropic effects (Figs. 3, 4). Such a dose of gonadotropins may be capable of causing a release of gonadal steroids, or it is possible that the TSH had some intrinsic gonadotropic activity, which could be either physiological or pharmacological in nature. It is also conceivable that mammalian glycoprotein hormones could somehow alter steroid hormone turnover. More complete interpretation of these data awaits further experimentation. The lack of thyrotropic effects of SGGLOO in Fundulus differs from reports of other species. In female freshwater perch (Am&as) SG-GlOO stimulated thyroid activity, but only if the ovaries were intact (Chakraborti et ad., 1983). Thyroid cell height and peroxidase activity decreased following ovariectomy and were partially restored with estrogen replacement. These authors proposed that gonadotropin induces ovarian steroid release, which in turn activates the thyroid; this could explain their empirical observation that peak thyroid activity is concurrent with spawning. In contrast, the ovaries were not necessary for the demonstration of a thyrotropic effect of SG-GlOO in Oncarhyncus nevka (Donaldson and McBride, 1974). It was concluded that either some TSH contamination or some inherent thyrotropic activity of,GTH were possible, with the bulk of evidence behind the first theory. The absence of direct thyrotropic effects of SG-GlOO in Anabas and Fund&s does not necessarily contradict this explanation. 0. nerka may be more sensitive to a very small amount
257
IN FISH
of TSH in SG-GlOO as a consequence of its phylogenetic relationship to 0. tsawyfscka (from which the pituitary extract is made). In general, our results indicate a separation of function between the thyroid and gonadal systems in Fundulus. Despite a typical heterothyrotropic response to mammalian gonadotropins, the thyroid of this species is insensitive to endogenous GTH and a partially purified salmonid GTH. If indeed the thyroid system originated as an accessory to reproduction, traces of the initial role are not readily demonstrated in Fundulus. Perhaps natural selection has favored developmental and climate-related thyroid actions in this species, while some other, more primitive fishes may still reiy on the thyroid in a reproductive capacity. ACKNOWLEDGMENTS SG-GlOO was a gift of Dr. E. Donaldson. Mammalian hormones were supplied by the National Pituitary Agency. Steroid radioimmunoassays were performed by C S , Bradford and Dr. R. S. Donham. B. Hamilton helped in the collection of fish.
REFERENCES Bradford, C. S. (1983). Changing steroid concentrations during the simulunar spawning cycle of the Killifish Fundulus heteroclitus. Master’s Thesis, School of Life and Health Sciences, University of Delaware. Brown, C. L., Grau, E. G., and Stetson, M. H. (1982). Endogenous gonadotropin does not have heterothyrotropic activity in Funduhs iaeteroclitus. Amer. Zoo/. 21, 1025. Chakraborti, P., Rakshit, D. K., and Bhattacharya, S. (1983). Influence ,of season, gonadotropins, and gonadal hormones on the thyroid activity of freshwater perch, Anabns testudinetis (Bloch). Ccmad. J. Zool. 61, 359-364. Crim, L. W., and Cluett, D. M. (1974). Elevation of plasma gonadotropin concentration in response to mammalian gonadotropin releasing hormone (GRH) treatment of the male brown trout as determined by radioimmunoassay. Endo. Res. Comm. 1. 101-110. Crim, L. W., and Evans, D. M. (1980). LH-RH-Stimulated gonadotropin release from the rainbow trout pituitary gland: An in vitro assay for detection of teleost gonadotropin releasing factor(s). Gen.
Camp.Endocrinoi.
40, 283-290.
Dettlaff, T. A., and Davydova, S. I. (1979). Differen-
258
BROWN,
GRAU, AND STETSON
tial sensitivity of cells of follicular epithelium and oocytes in the stellate sturgeon to nnfavorabie conditions, and correlating influence of triiodothyronine. Gen. Camp. Endocrinol. 39, 236-243. Donaldson, W. M., and McBride, J. R. (1974). Effect of ACTH and salmon gonadotropin on interrenal and thyroid activity of gonadotomized adult sockeye salmon (Oncorhynchus nerka). J. Fish. Res. Board Canad. 31, 1211-1214. Fontaine, Y.-A. (1963). Sur des facteurs a activite hettrothyreotrope presents dans les hypophyses de Mammiferes. Definition et purification. Gen. Camp. Endocrinol. 3, 701. Fontaine, Y.-A. (1969a). Studies on the heterothyrotropic activity of preparations of mammalian gonadotropins on teleost fish. Gen. Comp. Endocrinol. Suppl. 2, 417-424. Fontaine, Y.-A. (3969b). La speflcitt zoologique des proteines hypophysaires capables de stimuler la thyroide. Acta Endocr. Suppl. 136, l-154. Fontaine, Y.-A., and Burzawa-Gerard, E. (1977). Esquisse de l’evolution des hormones gonadotropes et thyreotropes des vertebres. Gen. Comp. Endocrinol. 32, 341-347. Fontaine-Bertrand, E., Marchelidon, J., Salmon, C., and Fontaine, Y.-A. (1981). Les sous-unites des lutropines: une activite gonadotrope intrinseque de (Y et le role de p dans le determinisme de la specificit zoologique d’action. C. R. Hebd. Se’anc. Acad. Sci. Paris. Ser. III 292, 507-510. Grau, E. G., and Stetson, M. H. (1977a). Pituitary autotransplants in Fundulus heteroclitus: Effect on thyroid function. Gen. Comp. Endocrinol. 32, 427-43 1. Grau, E. G., and Stetson, M. H. (1977b). The effects of prolactin and TSH on thyroid function in Fundulus heteroclitus. Gen. Comp. Endocrinol. 33, 329-335. Grau, E. G., and Stetson, M. H. (1977c). Thyroidal responses to exogenous mammalian hormones in Fundulas heteroclitus. Amer. Zool. 17, 8.57. Grau, E. G., and Stetson, M. H. (1979). Growth hormone is thyrotropic in Fundulus heterociitus. Gen. Comp. Endocrinol. 39, l-8.
Grau, E. G. (1978). Control mechanisms in the teleost hypothalamo-hypophysial-thyroidal axis. Ph.D. Dissertation, School of Life and Health Sciences, University of Delaware. Licht, P., Papkoff, H., Farmer, S. W., Muller, C. H., Tsui, H. W., and Crews, D. (1977). Evolution of gonadotropin structure and function. Recent Prog. Horm. Res. 33, 169-248. MacKenzie, D. S. (1980). Phylogenetic patterns in the specificity of thyroid response to pituitary glycoprotein hormones. Amer. Zool. 20, 858. MacKenzie, D. S. (1982). Stimulation of the thyroid gland of a teleost fish, Gillichthys mirabilis, by tetrapod pituitary glycoprotein hormones. Comp. Biochem. Physiol. 72A, 477-482. MacKenzie, D. S., Licht, l?, and Papkoff, H. (1978). Thyrotropin from amphibian (Rana catesbeiana) pituitaries and evidence for heterothyrotropic activity of bullfrog luteinizing hormone in reptiles. Gen. Comp. Endocrinol. 36, 566-574. Ng, T. B., Idler, D. R., and Eales, J. G. (1982). Pituitary hormones that stimulate the thyroidal system in teleost fishes. Gen. Comp. Endocrinol. 48, 372-389. Peter, R. E. (1970). Hypothalamic control of thyroid gland activity and gonadal activity in the goldfish, Carassius auratus. Gen. Camp. Endocrinol. 14, 334-356. Pickford, G. E., and Grant, F. B. (1968). The response of hypophysectomized male killifish (Fund&s heteroclitus) to thyrotropin preparations and to the bovine heterothyrotropic factor. Gen. Comp. Endocrinol. 10, l-7. Sage, M. (1973). The evolution of thyroidal function in fishes. Amer. Zool. 13, 899-905. Schreibman, M. P., Halpern, L. R., Goos, H. J. Th., and Margolis-Kazan, H. (1979). Identification of luteinizing hormone-releasing hormone (LH-RH) in the brain and pituitary gland of a fish by immunocytochemistry. J. Exp. Zool. 210, 153-160. Wingfield, J. C., and Grimm, A. S. (1977). Seasonal changes in plasma cortisol, testosterone, and oestradiol-17P in the plaice, Pleuronectes platessa L. Gen. Comp. Endocrinol. 31, l-11.