Immunocytochemical identification of growth hormone, prolactin, and gonadotropin cells in the pituitary of male plaice (Pleuronectes platessa) during gonadal maturation

GENERAL

AND

COMPARATIVE

ENDOCRINOLOGY

85, 358-366 (1992)

lmmunocytochemical Identification of Growth Hormone, Prolactin, and Gonadotropin Cells in the Pituitary of Male Plaice (Pleuronectes platessa) during Gonadal Maturation D. M. POWER’ Ministry of Agriculture,

Fisheries and Food, Lowestoft, Suffolk, NR33 OHT, United Kingdom Accepted April 11, 1991

Somatotrophs, gonadotrophs, and prolactin (PRL) cells have been demonstrated in male plaice (Pleuronectes platessa) pituitary by immunocytochemistry. All cell types exhibited patterns of activity correlated with gonadal maturity. Immature and maturing male plaice (gonads very small or filling with sperm, stages I and III) had numerous densely staining somatotrophs and PRL cells but only a few weakly stained gonadotrophs. Mature plaice (fish in which sperm could be extruded under light pressure, stage VI) contained two populations of gonadotrophs. The PRL cells of these fish were densely stained and the somatotrophs more lightly stained. Spent plaice (thin, flabby gonads, stage VII) had little or no staining in PRL cells. Somatotrophs and gonadotrophs from these fish stained weakly with the antisera used in the study and, in some fish, vacuoles were observed in areas of the pituitary normally occupied by gonadotrophs. o NYZAcademic press.inc.

In teleost fish, hypophysial and gonadal hormones are intimately involved in the regulation of the timing of reproduction. Gonadotropin (GtH) produced by gonadotrophs in the pars distalis (PD) is most closely associated with reproduction, stimulating steroid production, uptake of vitellogenin, oocyte maturation and ovulation, and spermiation (Donaldson and Hunter, 1983; Goetz, 1983; Wallace and Selman, 1981). Most experimental data suggest that two other PD hormones, growth hormone (GH) and prolactin (PRL), are not involved in fish reproduction (Clarke and Bern, 1980). Their main functions so far identified in lish are the control of growth and osmoregulation, respectively. However, steroids produced by the gonads which have a profound effect on the gonadotrophs may affect GH ’ Address for correspondence: Universidade do Algarve, UCEH, Campus do Gambelas, 8000 Faro, Portugal.

and PRL cellular morphology (Olivereau and Olivereau, 1979). The plaice (Pleuronectes platessa), a temperate marine teleost, has a welldefined seasonal cycle of gonadal development and spawning (Simpson, 1959). The identity of steroids involved in reproduction is well documented (Canario and Scott, 1989, 1990). However, apart from a study by Barr and Hobson (1964), little information exists on pituitary cell morphology during maturation, when plasma gonadal steroid concentrations are elevated. In the present study, a combination of conventional histology and immunocytochemistry has been applied to the distribution of somatotrophs (GH-producing cells), gonadotrophs, and PRL cells in the plaice pituitary. The possible involvement of GH and PRL in gonadal maturation of wild male plaice was determined by correlating changes in cell number and distribution and intensity of staining (using immunocytochemistry) with gonadal stage. 358

00166480192 $1 SO Copyright All rights

0 1992 by Academic Press, Inc. of reproduction in any form reserved.

IDENTIFICATION

MATERIALS

OF

CELLS

AND METHODS

Tissue samples. Male plaice were collected from spawning grounds in the southern bight of the North Sea in January 1989. Since spawning occurs between December and March the fish collected were at various stages of maturity (Simpson, 1959). After capture, the fish were weighed, the stage of gonadal maturity was assessed (stage I, immature gonads; stage III, maturing gonads; stage VI, spermiating; stage VII, spent), and the brain with pituitary attached was removed and fixed in Bouin-Hollande sublimate for 24 hr at room temperature. Fixed tissue was transferred into 70% industrial methylated spirit, dehydrated, and embedded in low-melting point paraffin. Sagittal sections (4 pm) of plaice pituitary (n = 5 at each stage of maturity) were cut and mounted serially (three sections per slide) on glycerine/albumin-coated slides. Antibodies. Rabbit antisera directed against rainbow trout (Oncorhynchus mykiss) GH and chum salmon (Oncorhynchus keta) GH and PRL were raised as reported (Skarphedinsson et al., 1990) and salmon GtH antiserum was kindly donated by Dr. J. Sumpter. The antisera against trout GH, salmon PRL, and salmon GtH were used at dilutions of 1:2000 and 1:4000 in 0.02 M phosphate-buffered saline (PBS, pH 7.6). Immunocytochemical procedures. Sections were stained using the horseradish-peroxidase (HRPO) method (Stemberger, 1974). In brief, sections were dewaxed in xylene, rehydrated and incubated in PBS containing 3% bovine serum albumin for 20 min, and then incubated with primary antiserum for 12-18 hr at room temperature. The primary antiserum was rinsed off and the secondary antiserum, biotinylated antirabbit immunoglobulins (Dakopatts, Denmark), diluted 1:70 in PBS, was applied for 1 hr. Sections were then rinsed in PBS and incubated with streptavidinlinked HRPO (Dakopatts, Denmark). After a final wash, the HRPO complex was developed with a solution of diaminobenzidine tetrahydrochloride (0.2 mg . ml-‘) and 0.06% hydrogen peroxide. Sections were washed in tap water, dehydrated in alcohol, cleared in xylene, and mounted in DPX. In some experiments, antisera against GH, PRL, and GtH were preincubated with trout GH, salmon GH, salmon PRL (kindly donated by H. Kawauchi, School of Fisheries Science, Kitasato University, Japan), and partially purified salmon GtH, respectively, in concentrations of 100 or 1000 nmol . ml-i. Incubation with these proteins abolished staining. Other sections were incubated in the absence of primary or secondary antisera which also abolished staining. The gross morphology of the plaice pituitary used serial sections stained with Cleveland-Wolfe trichrome for tinctorial identification of the pituitary cell types.

IN

MALE

359

PLAICE

RESULTS Histology of the Pituitary

Gland

The general morphology of the male plaice pituitary (Fig. 1) conformed to that described by Barr and Hobson (1964), although some changes in cellular disposition occurred during the maturation cycle. Sagittal sections stained with the ClevelandWolfe trichrome method revealed a welldefined rostra1 pars distalis (RPD), which anteriorly contained faintly red cells interspersed occasionally with blue ones. The red cells were not arranged in follicles and appeared to be homologous with the prolactin cells described in other teleosts. A group of cells oval or round in shape and containing a large nucleus stained faintly dark brown to black and bordered this region. These cells were assigned as ACTH cells on the basis of staining and location. The proximal pars distalis (PPD), which lies adjacent to the RPD, contained chains of orange-staining cells (somatotrophs), interspersed with clear blue-staining cells. The clear blue cells were more numerous toward the periphery of the PPD and in some fish were also scattered throughout the pars intermedia. These cells were designated gonadotrophs.

RPD

FIG. 1. Diagrammatic sag&al section through the pituitary gland of plaice, indicating the main areas of staining with the antisera used in the study: GH, 8; PRL, I& GtH, *; RPD, rostral pars distalis; PPD, proximal pars distalis.

D. M. POWER

360 Immunocytochemistry Plaice Pituitary

of the

HRPO immunocytochemistry revealed intense, selective staining of specific cell types according to the antisera used and the maturational stage of the fish (Table 1). The stained cells generally corresponded to the cell types identified by conventional histology. Thus, the RPD contained cells stained by anti-salmon PRL sera, while anti-trout and anti-salmon GH sera stained cells in the PPD and anti-salmon GtH sera stained cells in the PPD which were not stained by antiGH sera. It is probable that some of the latter cells were thyrotrophs as the antiGtH sera were raised against the whole GtH molecule. With all the antisera used in the study, staining was confined to the cytoplasm of cells; nuclei were unstained. Staining of the PRL, GH, and GtH cells in male plaice sampled over a fortnight during the spawning period varied. In pituitaries from mature fish (stage VI), GtH cells were observed in both the main PPD (Fig. 2a) and between the RPD and the PPD. These cells were not interspersed with somatotrophs or prolactin cells but formed a discrete, homogeneous mass protruding into the RPD. The cells were relatively large and slightly elongated and had intensely staining cytoplasmic material (Fig. 2b). Immature and maturing plaice (gonadal stages I and III) demonstrated a higher general background stain and contained a few TABLE 1 THE CORRELATION OF GONADAL MATURITY WITH STAINING INTENSITY OF PITUITARY CELLS Gonadal stage I VI VII

antiPRL sera

anti-GH sera +++,t ++,r ++,r

t

+++,r ++,t -/+,*

r

antiGtH sera ++,r +++,t +,*

t

Note. t t , many cells; f , moderate number of cells; *, few cells; + + +, intense stain; + f, moderate stain; f, weak stain; - I + , little or no staining.

intensely staining GtH cells scattered throughout the PPD (Fig. 2~); there was no evidence of another population of gonadotrophs between the RPD and the PPD in these fish. The pituitary of spent fish (gonadal maturity stage VII) contained weakly staining GtH cells, confined to the PPD (Fig. 2d). In addition, there was a high general background stain and numerous degenerating gonadotrophs were observed. The PRL cells from plaice with gonadal maturity assessed as stages I, III, and VI showed evidence of nuclear hypertrophy and there was an accumulation of material in the cells, as evidenced by their dense staining with anti-salmon PRL (Figs. 3a3~). However, spent fish (stage VII) showed little or no staining of PRL cells with the antisera used in this study (Fig. 3d). Rather, there was a reduction in the number of PRL cells and an associated reduction in the overall size of the RPD. Somatotrophs appeared least affected by reproduction, although some differences were observed during the cycle. In immature and maturing fish (stages I and III), the PPD contained numerous cells with cytoplasm densely stained with antisera against trout and salmon GH (Figs. 4a and 4b). Pituitaries from stage VI plaice showed an apparent reduction in the size of the PPD occupied by somatotrophs, probably as a consequence of the development of GtH cells. Moreover, the somatotrophs of these fish did not appear to accumulate as much material as immature fish, based on the more lightly stained cytoplasm (Fig. 4~). Pituitaries from stage VII plaice also contained fewer somatotrophs with staining generally confined to the periphery of the cytoplasm (Fig. 4d). DISCUSSION

Immunocytochemistry has been used to localize GtH-, GH-, and PRL-secreting cells in the plaice pituitary and examine the

IDENTIFICATION

OF

CELLS

IN

MALE

PLAICE

361

FIG. 2. Sag&al sections through the pituitary gland of plaice at different stages of maturity, stained with anti-GtH serum. (a) Gonadotrophs in mature fish (stage VI), note the two populations of gonadotrophs. (b) Intensely staining material in the cytoplasm of gonadotrophs of mature fish. (c) Immature (stages I or III) plaice pituitary containing GtH cells scattered throughout the PPD. (d) Pituitary of spent plaice (stage VII), note the weakly stained and degenerating GtH cells. Arrows indicate typically stained cells. Scale bars: a and c = 4 urn; b and d = 1 urn.

changes that are associated with gonadal maturation. All the antisera used in the study (anti-trout and salmon GH, antisalmon GtH and PRL) cross-reacted with the corresponding plaice hormones, but at lower dilutions than those previously used with homologous species (Sharphedinsson et al., 1990) together with a higher general background stain. PRL is well known to be present in the pituitary of both freshwater and marine te-

leosts. In freshwater teleosts, PRL is important in osmoregulation (Ball and Ensor, 1967; Komourdjian and Idler, 1977). However, the role of PRL in marine teleosts is poorly understood. In fish, such as the flounder (Pleuronectes flesus) and the American plaice (Hippoglossoides plutessoides) which live in a varying osmotic environment, PRL probably plays a role in osmoregulation (Chadwick, 1970). In fish that live in a totally marine environment,

362

D. M. POWER

- ~--

.-

FIG. 3. Sagittal sections through the pituitary of plaice at different stages of maturity, stained with anti-prolactin serum. (a) Pituitary of immature (stages I or III) plaice showing densely stained PRL cells. (b) Intensely stained cytoplasm of PRL cells in the RPD of immature plaice. (c) Pituitary of plaice at gonadal stage VI. (d) Pituitary of plaice at gonadal stage VII demonstrating a total absence of staining with the antisera. Arrows indicate typically stained cells. Scale bars: a, c, and d = 4 pm; b = 1 pm.

where there is no demand for sodium retention, its role is unclear. In the present study, despite the constant osmotic conditions, PRL cell activity in male plaice at different stages varied. Although there is sparse evidence implicating PRL in teleost reproduction, the almost complete disappearance of PRL from the pituitary of stage

VI male plaice suggests that it may be associated with reproductive changes occurring at this time. Changes in PRL cell activity associated with reproduction have previously been noted in the silver eel (Olivereau and Olivereau, 1979) and the Indian catfish (Sundararaj and Keshavanath, 1976). In the latter case PRL was suggested

IDENTIFICATION

OF

CELLS

IN MALE

PLAICE

363

FIG. 4. Sagittal sections through the pituitary gland of plaice at different stages of maturity stained with trout anti-GH serum. (a) Pituitary of immature plaice (stages I or III), note the numerous densely stained somatotrophs. (b) High-power view of densely stained GH cells of immature fish. (c) Somatotrophs of mature plaice (stage VI), note the less numerous somatotrophs and the lighter cytoplasmic stain. (d) High-power view of somatotrophs of mature plaice (stage VI). Arrows indicate typically stained cells. Scale bars: a and c = 4 pm; b and d = 1 km.

to play a role in the development of the seminal vesicles (Sundararaj and Nayyar, 1969). Furthermore, specific binding sites for ovine PRL have been demonstrated in ovary and testes of tilapia (Sarotherodon mossambicus), although the role of these receptors is unclear (Edery et al., 1984). However, in male rats PRL enhances the action of luteinizing hormone (LH) on the testes possibly by stimulating LH receptor

numbers (Bartke and Dalterio, 1976; Johnson, 1974; Sharpe and McNeilly, 1979). Only one morphological type of gonadotroph could be identified in plaice by light microscopy. However, in stage VI fish two distinct populations of cells were identified on the basis of their spatial arrangement in the pituitary and appearance during this particular stage of gonadal ‘maturity. Two

364

D. M. POWER

morphologically distinct gonadotrophs have been identified ultrastructurally in sticklebacks (Gasterosteus aculeatus) (Slijkhs, 1978), loach (ikfisgurnus anguifficuudutus) (Ueda and Takahashi, 1980), winter flounder (Burton et al., 1981), goldfish (Curussius uurutus) (Olivereau, 1962), trout (S&no truttu furio) and saupe (Surpu (Boops) salpa) (Olivereau and Nagahama, 1983), and brown bullhead (Zctulurus nebulosus) (Farbridge et ui., 1985). Further studies will be required to determine if the two populations of cells identified in the present study represent two gonadotrophs which secrete different forms of gonadotropin as suggested in the salmon (Itoh et al., 1990; Suzuki et al., 1990). The correlation between the state of gonadal maturity and the activity of the gonadotrophs is in general agreement with Barr and Hobson (1964), who demonstrated that the gonadotrophic activity of extracted plaice pituitaries changed throughout the reproductive cycle. Thus, pituitary extracts from immature and spent fish had a low stimulatory action. Similarly, in the present study, the gonadotrophs from such fish were weakly stained. The reduction in cellular activity in spent fish was accompanied by the appearance of vacuoles in areas of the pituitary normally occupied by gonadotrophs. It is possible that these vacuoles are an expression of cellular exhaustion (Benjamin, 1981). Such a phenomenon has been observed in rainbow trout, where it has been proposed that such cavities arise as a consequence of degeneration of gonadotrophs which accompany sexual maturation (Robertson and Wexler, 1962). It is possible that the observed changes in GtH cell morphology during the sexual cycle may be caused by changes in circulating levels of gonadal steroids. Thus, exogenous gonadal steroids or gonadectomy respectively stimulate or inhibit gonadotrophs (Borg et al., 1986; Dufour et al., 1983; Goos et al., 1976; Groves and Batten, 1986;

Kaneko et al., 1986; Olivereau and Olivereau, 1979; Ueda and Takahashi, 1980). Seasonal changes in GH cell activity have previously been reported in fish (Swift and Pickford, 1965; Farbridge et al., 1985), although little direct information exists about GH cell activity during gonadal maturation. In the present study the small, immature, rapidly growing fish had the most active somatotrophs. Although somatotrophs from stages VI and VII fish were less numerous than those of stages I and III, whether the fewer somatotrophs reflect changes accompanying gonadal maturation or are part of the aging process is unclear. Finally, changes in GtH cell morphology are clearly related to reproduction while patterns of PRL and GH cell morphology during the sexual cycle renders it uncertain whether these two hormones are directly involved in reproduction. ACKNOWLEDGMENTS The author is grateful to Drs. Pat Ingleton and G. P. Arnold for reading the manuscript and providing helpful suggestions, and to Mrs. B. Turner who assisted in its preparation.

REFERENCES Ball, J. N., and Ensor, D. M. (1%7). Specific action of prolactin on plasma sodium levels in hypophysectomized Poecilin latipinna (Teleostei). Gen. Comp. Endocrinol. 8, 432-W. Barr, W. A., and Hobson, B. M. (1964). Endocrine control of the sexual cycle in the plaice, Pleuronectes platessa L. Gen. Comp. Endocrinol. 4, 608-613. Bartke, A., and Dalterio, S. (1976). Effects of prolactin on the sensitivity of the testis to LH. Biol. Reprod. 15, 90-93. Benjamin, M. (1981). Cysts and (large follicles) and colloid in pituitary glands. Gen. Comp. Endocrinol. 45, 425-445. Borg, B., Paulson, G., and Peute, J. (1986). Stimulatory effects of methyltestosterone on pituitary gonadotropic cells and testes Leydig cells of the three-spined stickleback, Gasterosteus aculeotus L. in winter. Gen. Comp. Endocrinol. 62, 54-61. Burton, M. P., Idler, D. R., and Bun Ng, T. (1981). The immunofluorescent location of teleost gonad-

IDENTIFICATION

OF

otropins and thyrotropins in flounder pituitary. Gen. Comp. Endocrinol. 43, 135-147. Canario, A. V. M., and Scott, A. P. (1989). Conjugates of ovarian steroids, including 17a,208dihydroxy-4-pregnen-3-one (maturation inducing steroid), accumulate in the urine of the marine teleost plaice (Pleuronectes platessa). J. Endocrinol. 123, Rl-R4. Canario, A. V. M., and Scott, A. P. (1990). Identification of, and development of radioimmunoassays for 17a,21-dihydroxy-4 pregnene-3,20-dione and 3a,l7a,21-trihydroxy-SB-pregnan-20-one in the ovaries of mature plaice (Pleuronectes plafess& Gen. Comp. Endocrinol. 78, 273-285. Chadwick, A. (1970). Pigeon crop-stimulating activity in the pituitary of the flounder (Pleuronectes fleSW). J. Endocrinol. 47, 46349. Clarke, W. C., and Bern, H. A. (1980). Comparative endocrinology of prolactin. In “Hormonal Proteins and Peptides” (C. H. Li, Ed.), Vol. 8, pp. 105-197. Academic Press, New York. Donaldson, E. M., and Hunter, G. A. (1983). Induced final maturation, ovulation, and spermiation in cultured fish. In “Fish Physiology” (W. S. Hoar, D. J. Randall, and E. M. Donaldson, Eds.), Vol. 9, Part B, pp. 117-170. Academic Press, New York.

Dufour, S., Deleme-Le Belle, N., and Fontaine, Y-A. (1983). Effects of steroid hormones on pituitary immunoreactive gonadotropin in European freshwater eel, Anguilla anguilla L. Gen. Comp. Endocrinol. 52, B&197. Edery, M., Young, G., Bern, H. A., and Steiny, S. (1984). Prolactin receptors in tilapia (Sarofherodon mossambicus) tissues: Binding studies using ‘Z51-labeled ovine prolactin. Gen. Comp. Endocrinol. 56, 19-23. Farbridge, K. J., Burke, M. Cl., and Leatherland, J. F. (1985). Seasonal changes in the structure of the adenohypophysis of the brown bullhead (Zctalurus nebulosus Lesueur). Cytobiosis 44, 4966. Goetz, F. W. (1983). Hormone control of oocyte final maturation and ovulation in fishes. Zn “Fish Physiology” (W. S. Hoar, D. J. Randall, and E. M. Donaldson, Eds.), Vol. 9, Part B, pp. 117-170. Academic Press, New York. Coos, H. J. Th., Selenrijk, R., and Peute, J. (1976). The gonadotropic cells in the pituitary of the black molly, Mollienisia lutipinna, and other teleosts identified by the immunofluorescence technique in normal and androgen-treated animals. Cell Tissue Res. 167, 211-219. Groves, D. J., and Batten, T. F. C. (1986). Direct control of the gonadotroph in a teleost, Poecilia latipinna: Gonadal steroids. Gen. Comp. Endocrinol. 61, 402416.

CELLS

IN

MALE

PLAICE

365

Itoh, H., Suzuki, K., and Kawauchi, H. (1990). The complete amino acid sequences of a subunits of chum salmon gonadotropins. Gen. Comp. Endocrinol. 78, 56-66. Johnson, D. C. (1974). Temporal augmentation of LH by prolactin in stimulation of androgen production by the testes of hypophysectomized male rats. Proc. Sot. Exp. Biol. Med. 145, 610-613. Kaneko, T., Aida, K., and Hanyu, I. (1986). Changes in ovarian activity and fine structure of pituitary gonadotrophs during spawning cycle of the Chichibu-goby Tridentiger obscurus. Bull. Jpn. Sot. Sci. Fish 52, 1923-1928. Komourdjian, M. P., and Idler, D. R. (1977). Hypophysectomy of rainbow trout, Salmo gairdneri, and its effect on plasmatic sodium regulation. Gen. Comp. Endocrinol. 32, 536-542. Olivereau, M. (1962). Cytologie de l’hypophyse du cyprin (Carassius auratus L.). C.R. Acad. Sci. Ser. 3 255, 2007-2009. Olivereau, M., and Olivereau, J. (1979). Estradiolpositive feedback on gonadotropic (GtH) cells in freshwater male silver eels. Gen. Comp. Endocrinol. 39, 247-261. Olivereau, M., and Nagahama, Y. (1983). Immunocytochemistry of gonadotropic cells in the pituitary of some teleost species. Gen. Comp. Endocrinol. 50,252-260. Robertson, 0. H., and Wexler, B. L. (1962). Histological changes in the pituitary gland of the rainbow trout (Salmo gairdneri) accompanying sexual maturation and spawning. J. Morphol. 110, 157169. Sharpe, R. M., and McNeilly, A. S. (1979). The effect of induced hyperprolactinaemia on leydig cell function and LH-induced loss of the LHreceptors in the rat testis. Mol. Cell. Endocrinol. 16, 19-27. Simpson, A. C. (1959). The spawning of the plaice (Pleuronectes platessa) in the North sea. Fish. Invest. London Ser. 2 22, (7), p. 111. Skarphedinsson, 0.. Power, D. M., and Ingleton P. M. (1990). Separation of rainbow trout (Salmo gairdnert), growth hormone by gel electrophoresis. Gen. Comp. Endocrinol. 80, 393-398. Slijkhuis, H. (1978). Ultrastructural evidence for two types of gonadotropic cells in the pituitary gland of the male three-spined stickleback (Gasterosfeus aculeatus). Gen. Comp. Endocrinol. 36, 639-641. Stemberger, L. A. (1974). In “Immunocytochemistry,” pp. 129-171. Prentice-Hall, Englewood Cliffs, NJ. Sundararaj, B. I., and Nayyar, S. K. (1%9). Effect of prolactin on the “seminal vesicles” and neural

366

D. M. POWER regulation

of prolactin

secretion in the cattish Gen. Comp. En-

Heteropneustes fossilis (Bloch). docrinol. Suppl. 2, 69-80.

Sundararaj, B. I., and Keshavanath, P. (1976). Effects of melatonin and prolactin treatment in the hypophysialovarian system in the catfish, Heteropneustes fossilis (Bloch). Gen. Comp. Endocrinol. 29, 84-96.

Suzuki, K., Kawauchi, H., and Nagahama, Y. (1990). Isolation and characterization of two distinct gonadotropins from chum salmon pituitary glands. Gen.

Cornp.

Endocrinol.

71, 292-301.

Swift, D. R., and Pickford, G. E. (1965). Seasonal variation in the hormone content of the pituitary gland of the perch, Perca fluviatilis L. Gen. Comp.

Endocrinol.

5, 354-36.5.

Ueda, H., and Takahashi, H. (1980). Responses of two different types of pituitary gonadotropins of the loach, Misgurnus anguillicaudatus, to gonadectomy and to exogenous sex steroids. Gen. Camp. Endocrinol.

40, 463472.

Wallace, R. A., and Selman, K. (1981). Cellular and dynamic aspects of oocyte growth in teleosts. Am. Zool. 21, 325-343.