INFORMATION TO USERS

This manuscript ,has been reproduced from the microfilm master. UMI

films the text directly from the original or copy submitted. Thus, some

thesis and dissertation copies are in typewriter face, while others may

be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of thecopy submitted. Broken or indistinct print, colored or poor qualityillustrations and photographs, print bleedtbrough, substandard margins,

and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete

mannsenpt and there are missing pages, these will be noted. Also, ifunauthorized copyright material had to be removed, a note will indicate

the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by

sectioning the original, beginning at the upper left-hand comer and

contimiing from left to right in equal sections with small overlaps. Each

original is also photographed in one exposure and is included in

reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced

xerographically in this copy. Higher quality 6" x 9" black: and white

photographic prints are available for any photographs or illustrations

appearing in this copy for an additional charge. Contact UMI directly

to order.

UMIA Bell & Howell Information Company

300 North Zeeb Road. Ann A~bor. MI48106·1346 USA313/761-4700 800:521-0600

REGULATION OF PROLACTIN AND CHANGES IN PROLACTIN AND GROWTH

HORMONE IN OSMOREGULATION, METABOLISM, AND REPRODUCTION IN

THE TILAPIA, OREOCHROMIS MOSSAMBICUS.

A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THEUNIVERSITY OF HAWAII IN PARTIAL FULFILLMENT OF THE

REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN

ZOOLOGY

MAY 1995

By

Gregory Martin Weber

Dissertation Committee:

E. Gordon Grau, ChairmanShannon Atkinson

Christopher L. BrownGillian D. Bryant-Greenwood

Fred I. Kamemoto

UMI Number: 9532638

OMI Microform 9532638Copyright 1995, by OMI Company. All rights reserved.

This microform edition is protected against unauthorizedcopying under Title 17, United States Code.

UMI300 North Zeeb RoadAnn Arbor, MI 48103

I dedicate this dissertation to my wife, Sue, in

appreciation of her constant support, and to my daughter,

Katelyn, who likes to see her name in books.

(MMS)

iii

ACKNOWLEDGEMENTS

I wish to extend my thanks to my thesis committee,

and especially to my advisor, Prof. E. Gordon Grau for his

support that allowed me to pursue a diversity of projects,

including some which are not encompassed in this disserta­

tion. I would like to acknowledge the contributions of

Buel Rodgers with whom I ran the first experiments of this

work. This dissertation also includes studies conducted

with Drs. Hal Richman and Russell Borski. I would like to

thank Drs. Tatsuya Sakamoto and Takashi Yada for teaching

me the RIA procedures and Prof. Tetsuya Hirano for provid­

ing the antibodies. Among the many other people who have

helped me along the way and may have also contributed to

this work are: Benny Ron, Craig Morrey, Brian Shepherd,

Steve Shimoda, Prof. Milton Stetson, Prof. Howard Bern,

Prof. Yoshitaka Nagahama, Dr. Richard Nishioka, Prof.

Nancy Sherwood, Dr. Matthew Grober, and Philippa Melamed.

iv

ABSTRACT

These studies addressed the regulation of the prolac­

tins (PRL) and growth hormone (GH) in the tilapia, Oreo­

chromis mossambicus. Regulation of the PRL cell was

investigated in the context of its potential roles in

osmoregulation, metabolism, and reproduction. The in

vitro release of the tPRLs (tPRL1 77 and tPRL1 SS) was

measured as well as changes in serum concentrations and

pituitary content of the PRLs and GH.

These studies provide evidence that it is the osmotic

gradient across the cell membrane that leads to an in­

crease in osmotic pressure, and not osmolality per se,

which accounts for the osmoreceptivity of the tilapia PRL

cell. Prolactin release from pituitary tissues (rostral

pars distalis; RPD) was inhibited when RPD were incubated

in medium made hyperosmotic by the addition of NaCI or the

membrane-impermeant molecule, mannitol; but not by the

addition of the permeant molecules, urea or ethanol.

Injections of a gonadotropin-releasing hormone (GnRH)

analog elevated serum concentrations of the tPRLs. Fur­

thermore, native forms of GnRH stimulated PRL release in

vitro with the following order of potency: chicken-GnRH-II

> salmon-GnRH > seabream-GnRH. Prolactin release was

accompanied by an increase in intracellular free ca2 + ,

suggesting Ca2 + operates as a second messenger in mediat-

v

ing the effects of GnRH on PRL release. Estradiol-178 and

testosterone potentiated the effects of GnRH on PRL re­

lease in vi tro.

Serum concentrations of the tPRLs were elevated in

response to fasting, followed by increases in serum con­

centrations and pituitary content of GH. Serum concentra­

tions and pituitary content of GH and serum concentrations

of PRL were highest late in the brooding phase of the

reproductive cycle. Reduced food intake during brooding

may contribute to changes in PRL and GH serum concentra­

tions and pituitary content. Nevertheless, patterns of

changes in serum and pituitary levels of the tPRLs and GH

observed during the reproductive cycle had characteristics

that were both similar to and distinct from patterns

observed during fasting, suggesting the hormones may have

actions in both metabolism and reproduction. In addition,

salinity altered the patterns of changes in serum and

pituitary levels of the tPRLs, but not GH, observed during

the reproductive cycle.

vi

TABLE OF CONTENTS

DEDICATION iiiACKNOWLEDGMENTS i vABSTRACT vLIST OF TABLES viiiLIST OF FIGURES ixLIST OF ABBREVIATIONS xiiiCHAPTER I: Introduction , 1CHAPTER II: Alteration in osmotic pressure and not

osmolality are coupled to the sustained release ofthe osmoregulatory hormone, prolactin, in theeuryhaline teleost, tilapia(Oreochromis mossambicus) 34

INTRODUCTION 34MATERIALS AND METHODS 40RESULTS 45DISCUSSION 57

CHAPTER III: Gonadotropin-releasing hormone functionsas a prolactin-releasing factor in the tilapia,Oreochromis mossambicus 62

INTRODUCTION 62MATERIALS AND METHODS 65RESULTS 72DISCUSSION 97

CHAPTER IV: Changes in serum concentrations andpituitary content of the prolactins and growthhormone with fasting in the tilapia,Oreochromis mossambicus '" 108

INTRODUCTION 108MATERIALS AND METHODS '" 112RESULTS 117DISCUSSION 136

CHAPTER V: Changes in serum concentrations andpituitary content of the prolactins and growthhormone during the reproductive cycle of femaletilapia, Oreochromis mossambicus, adapted tofresh water and seawater 145

INTRODUCTION 145MATERIALS AND METHODS 150RESULTS 156DISCUSSION 175

CHAPTER IV: Conclusions 196REFERENCES 201

vii

LIST OF TABLES

Table Page1. The effects of fasting up to 21 days on condition

factor, body weight and standard length in maleand female tilapia with a high condition factor ... 123

2. The effects of fasting up to 31 days on conditionfactor, body weight and standard length in maletilapia with a low condition factor 132

viii

LIST OF FIGURES

Figure Page1. The effects of increasing medium osmolality by

the addition of membrane permeant (urea) andimpermeant (mannitol) molecules on PRL releasefrom the tilapia RPD during 18-20 hr staticincubation 48

2. The effects of increasing medium osmolality bythe addition of the membrane permeant moleculeurea, on PRL release from the tilapia RPD during18-20 hr static incubation '" 50

3. The effects of increasing medium osmolality bythe addition of the membrane permeant moleculeethanol, on PRL release from the tilapia RPDduring 18-20 hr static incubation 53

4. The effects of increasing medium osmolality bythe addition of membrane permeant (urea) andimpermeant (mannitol) molecules on PRL releasefrom the tilapia RPD during perifusion incubation .. 56

5. The effects of intraperitoneal injections ofmGnRHa on serum PRL concentrations in male andfemale tilapia..................................... 74

6. The effects of mGnRHa on PRL release from thetilapia RPD during 18-20 hr static incubation inisosmotic medium 77

7. The effects of mGnRHa on PRL release from thetilapia RPD during 18-20 hr static incubation inhyposmotic medium, isosmotic medium andhyperosmotic medium................................ 79

8. The effects of cGnRH-I, cGnRH-II and sGnRH on PRLrelease from the tilapia RPD during 18-20 hrstatic incubation '" 82

ix

9. The effects of cGnRH-II, sGnRH and sbGnRH on PRLrelease from the tilapia RPD during 18-20 hrstatic incubation 85

10. The effects of sbGnRH on PRL release from thetilapia RPD during perifusion incubation 87

11. The effects of cGnRH-II and changes in mediumosmolality on [Ca++]i in dispersed PRL cellsduring perifusion incubation 89

12. The effects of decreasing concentrations ofcGnRH-II and changes in medium osmolality on[Ca++]i in dispersed PRL cells and PRL releasefrom tilapia RPD during perifusion incubation 91

13. The effects of increasing concentrations ofcGnRH-II and changes in medium osmolality on[Ca++]i in dispersed PRL cells and PRL releasefrom tilapia RPD during perifusion incubation 93

14. The effects of E2, T, sGnRH and combinationsof the steroid hormones and sGnRH on PRL releasefrom the tilapia RPD during 3 hr static incubation. 96

15. The effects of 2-phenoxyethanol on serumconcentrations of the tPRLs and GH in tpe tilapia .. 115

16. Changes in serum concentrations of the tPRLs andGH during fasting in male and female tilapia witha high condition factor 119

17. Changes in pituitary content of the tPRLs and GHduring fasting in male and female tilapia witha high condition factor , 122

18. Changes in gonadosomatic index during fasting inmale and female tilapia with a high conditionfactor , 126

x

19. Changes in serum concentrations of the tPRLs andGH during fasting in male tilapia with a lowcondition factor 128

20. Changes in pituitary content of the tPRLs and GHduring fasting in maletilapia with a lowcondition factor 131

21. Changes in hepatosomatic index during fasting inmale tilapia with a low condition factor 135

22. Changes in follicle mass during the brooding phaseof the reproductive cycle in the female tilapia ... 158

23. Changes in condition factor and hepatosomaticindex during the brooding phase of thereproductive cycle in FW-adapted female tilapiamaintained in an outdoor tank 161

24. Comparison of condition factors and hepatosomaticindices for FW- and 8W-adapted female tilapiawhich were either brooding post-yolksac larvae ornot brooding and at the end of vitellogenesis 163

25. Changes in serum concentrations of the tPRLs andGH during the brooding phase of the reproductivecycle in FW-adapted female tilapia maintained in anoutdoor tank 166

26. Changes in pituitary content of the tPRLs and GHduring the brooding phase of the reproductive cyclein FW-adapted female tilapia maintained in anoutdoor tank 168

27. Changes in serum concentrations of the tPRLs and GHduring the reproductive cycle in FW-adapted femaletilapia maintained in indoor tanks 171

xi

28. Changes in serum concentrations of the tPRLs andGH during the brooding phase of the reproductivecycle in 8W-adapted female tilapia maintained inindoor tanks...................................... 174

29. Changes in pituitary content of the tPRLs and GHduring the brooding phase of the reproductive cyclein 8W-adapted female tilapia maintained in indoortanks ' 177

xii

LIST OF ABBREVIATIONS

BW " Body weightcAMP '" 3'S'-cyclic adenosine

monophosphatecGnRH-I '" Chicken-gonadotropin-releasing

hormonecGnRH-II Chicken-gonadotropin-releasing

hormoneCF Condition factorcfGnRH Catfish-gonadotropin-releasing

hormoneE2 Estradiol-178FFA " Free fatty acidsFW Fresh waterGH Growth hormoneGHRH Growth hormone-releasing hormoneGnRH '" Gonadotropin-releasing hormoneGSI Gonadosomatic indexGtH GonadotropinHSI Hepatosomatic indexmGnRHa Mammal-gonadotropin-releasing

hormone analogmOsmolal MilliosmolalmOsm MilliosmolalPRL ProlactinRPD Rostral pars distalissbGnRH Seabream-gonadotropin-releasing

hormonesGnRH Salmon-gonadotropin-releasing

hormoneSL 'Standard lengthSW SeawaterT Testosterone

xiii

Chapter I

INTRODUCTION

Prolactin (PRL) and growth hormone (GH) are members

of the same polypeptide family and are thought to be

derived from a common ancestral gene (Niall et al., 1971)

Not surprisingly, PRL and GH have many overlapping actions

and are regulated by many of the same factors, although

often in opposite directions. The spectrum of actions of

each of these hormones throughout the vertebrates includes

effects on reproduction, growth and development, metabo­

lism, and osmoregulation (Clarke and Bern, 1980; Loretz

and Bern, 1982; Sakamoto et al., 1993; Nicoll, 1982).

Often, the hormones regulate these different functions

simultaneously. The cells secreting the hormones must,

therefore, receive regulatory information from a variety

of sources to be used in determining the secretion rate of

the hormones. In order to understand how PRL and GH

accomplish the regulation of diverse functions, changes in

hormones that are related to these activities must be

characterized and the factors regulating PRL and GH must

be identified. This has been the thrust of my research

and is the subject of this thesis.

1

Two tiiapia proiactins

The rostral pars distalis (RPD) of the adenohypophy­

sis of the tilapia releases two distinct PRL molecules

which are encoded by different genes and are not derived

through the differential processing of the same transla­

tion product (Specker et ai., 1985b; Yamaguchi et ai.,

1988). The larger PRL (tPRL1 88) contains 188 amino acid

residues and has a molecular weight of 20.8 kDa, whereas

the smaller PRL (tPRL1 77) contains 177 amino acid residues

and has a molecular weight of 19.6 kDa (Yamaguchi et ai.,

1988) .

There is limited evidence that the two tPRLs may have

distinct as well as overlapping functions. The two PRL

molecules show similar effects in the tilapia sodium

retaining assay, a well characterized bioassay that

measures the osmoregulatory activity of PRL (Specker et

ai., 1985b). Nevertheless, Borski and colleagues (1992)

have shown that the ratio of the total amount of tPRL1 8 8

to tPRL1 77 is greater in the pituitary of freshwater

(FW)-adapted tilapia, Oreochromis mossambicus, compared

with seawater (SW)-adapted tilapia. Specker and col­

leagues (1985a) found that tPRL1 88 and not tPRL177 has

growth-promoting activity in the tilapia. On the other

hand, Shepherd and colleagues (1994) found that injections

of tPRL1 77 and not tPRL1 88 are effective at restoring

2

sulfate and thymidine incorporation in vi'tro by branchial

cartilage of hypophysectomized tilapia. The assay is a

well characterized assay for skeletal growth. Shepherd

and colleagues also found that tPRL1 77, but not tPRL1 88,

ovine-PRL or salmon-PRL, can displace GH from high­

affinity, low-capacity binding sites in the tilapia liver.

A functional distinction between the two PRLs in regards

to reproduction or metabolism within the tilapia has not

been elucidated. Rubin and Specker (1992) found similar

activities of the two PRLs in affecting steroid release

from testicular tissues of courting and non-courting male

tilapia.

Tilapia pituitary pland morphology

The anatomical arrangement of the tilapia pituitary

gland imparts certain advantages as a model to study PRL

and GH cell physiology. The distinctive features of the

tilapia pituitary are described below in the next section,

followed with a description of how these features have

been exploited to study PRL and GH'cell function. The

morphology of the tilapia pituitary gland has been de­

scribed in detail (Dharmamba and Nishioka, 1968; Bern et

al., 1975; Nishioka et al., 1988). Two features of the

pituitary gland common to most teleosts including the

tilapia, are strikingly different from those of mammals.

3

They are: 1) hormone-producing cell types are segregated

into discrete areas of the pituitary gland, and 2) hypo­

thalamic fibers innervate the adenohypophysis (Dharmamba

and Nishioka, 1968; Bern et al., 1975; Ball, 1981). In

the tilapia, the proximal pars distalis contains GH cells

in addition to gonadotropin (GtH) and thyrotropin cells.

The RPD contains PRL and adrenocorticotropin cells. The

PRL cells are located in the anterior region of the RPD

and adrenocorticotropin cells are located close to the

neurohypophysis. Hypothalamic fibers innervate the proxi­

mal pars distalis and terminate in close proximity to the

hormone-secreting cells. In contrast, hypothalamic fibers

do not leave the neurohypophysis in the RPD, but terminate

on an adjacent basement membrane. Separating the PRL

cells and the hypothalamic fibers are the basement

membrane, the adrenocorticotropin cells, and a layer of

stellate cell processes (see review, Nishioka et al.,

1988). Nishioka and colleagues (1988) pose the possibili­

ty that the stellate cells are involved in the movement of

neurohormonal factors from the neurohypophysial nerve

endings to the PRL cells.

Tilapia PRL and GH cells as models for studying PRL and GH

cell function

Interactions among hormones in heterogeneous cell

4

populations complicate the study of the regulation of

individual cell types. Segregation of cell types within

the teleost pituitary facilitates separation of certain

cell types for study. A tissue containing a nearly homo­

geneous population of PRL cells can be obtained for in

vitro study by a simple dissection of the anterior RPD.

The RPD tissue obtained in this way is 95-99% PRL cells.

An additional attribute of the PRL cell which lends itself

to study is its sensitivity to small physiological changes

in osmolality. This sensitivity allows for the control of

baseline release to facilitate the investigation of poten­

tially important regulators of PRL secretion (cf. Grau et

al., 1982). Cells of the proximal pars distalis can also

be separated from cells of the pars intermedia and RPD by

simple dissection. Separation of GH, GtH and thyrotropin

cells from each other is more difficult. Currently, most

investigators utilize mammalian clonal cell lines derived

from PRL tumors to study the regulation of PRL cells

(Tashjian et al., 1970; Lamberts and MacLeod, 1990; Sato

et al., 1990). A concern with the use of tumor cell lines

is that one can not be sure that they are studying normal

cell function as opposed to tumor cell function.

Direct innervation of the tilapia adenohypophysis by

hypothalamic neurosecretory fibers enables the tracing of

fibers from the hypothalamus to the region of the

5

pituitary where the hormone is released. Since the hor­

mone-secreting cells of the pituitary are segregated,

neurosecretory fibers can be followed to determine at

which cells the neurosecretory factors are released, and

therefore, may regulate. Furthermore, neurosecretory

factors reaching the pituitary can be easily identified

due to the high quantity of the factors in the'nerve

terminals in the pituitary. I have taken advantage of

this feature and determined with colleagues, that sea­

bream-gonadotropin-releasing hormone is the most abundant

form of gonadotropin-releasing hormone (GnRH) in the

tilapia pituitary (Weber et al., 1994).

Hormonal regulation of teleost PRL and GH cell activity

The teleost PRL cell responds to many molecules in a

way that is similar to the manner in which the molecules

have been shown to be effective in mammals. Basal PRL

secretion in teleosts and mammals is under inhibitory

control from the hypothalamus. When connections with the

hypothalamus are severed at the pituitary stalk, as in

auto-transplants, the pituitary secretes high levels of

PRL. In mammals, this inhibition appears to be controlled

predominantly by dopamine (cf. Lamberts and MacLeod, 1990)

while in the tilapia, somatostatin appears to be the

predominant suppressor molecule (Grau et al., 1982, 1985).

6

Thyrotropin-releasing hormone stimulates PRL release in

mammals and tilapia, while vasoactive intestinal peptide

is a stimulator in mammals and an inhibitor in tilapia

(Barry and Grau, 1986; Kelley et al., 1988; cf. Lamberts

and MacLeod, 1990). A non-hypothalamic hormone, cortisol,

has been shown to be a potent inhibitor of PRL release in

the tilapia (Wigham et al., 1977; Borski et al., 1991).

More related to reproduction, it has been shown that

the tilapia PRL cells are stimulated to release PRL in

response to estradiol-17B (E2) and testosterone (T),

(Wigham et al., 1977; Barry and Grau, 1986; Borski et al.,

1991). Other steroids including progestins have been

found to have no effect on PRL release (Borski et al.,

1991). Furthermore, stimulation of PRL cells by thyrotro­

pin-releasing hormone was only observable following pre­

treatment with E2 (Barry and Grau, 1986). Stimulation of

PRL release from PRL cells by E2 and T suggests an

increase in PRL activity during phases of the reproductive

cycle when circulating E2 and T levels are elevated.

Control of GH cell activity in teleosts is not as

well characterized as control of PRL cell activity (see

review, Nishioka et ai, 1988). In mammals, GH secretion

is primarily under the control of hypothalamic regulators.

Growth hormone-releasing hormone (GHRH) stimulates GH

secretion and somatostatin inhibits GH secretion. It is

7

clear that the inhibitory role of somatostatin on GH cell

activity has been conserved in teleosts including the

tilapia (Fryer et ai., 1979). Studies examining the

effects of GHRH on GH release have been mixed. Neverthe­

less, GHRH was found to be a potent stimulator of GH

release in the only study to use native GHRH (Vaughan et

ai., 1992). Vaughan and colleagues (1992) isolated and

synthesized GHRH from the common carp and found it able to

elevate circulating GH levels and stimulate release of GH

from pituitary cells. In this same study a native GnRH,

salmon GnRH (sGnRH), was found to be equipotent with carp

GHRH at the one concentration compared (100 nm). Marchant

et ai. (1989) have shown that GnRH is a potent stimulator

of GH release in the goldfish. Injections of a human GHRH

fragment, fragment 1-29, elevated plasma GH levels in a

tilapia hybrid. Furthermore, GnRH fragment 1-29 and carp

GHRH stimulated GH release from pituitary fragments with

similar potency (Melamed et ai., 1995). In the same

study, injections of a sGnRH superactive analog was more

effective than the human GHRH fragment in elevating GH

levels. In addition, sGnRH was more potent than either

GHRH form in stimulating GH release in vitro.

Cortisol is a potent stimulator of GH release, yet

direct effects of sex steroids on GH secretion in the

tilapia have not been demonstrated (Nishioka et ai., 1985;

8

Helms et al., 1987). There is evidence, however, that

androgens have modulatory effects on GH-releasing factors,

similar to the effects of E2 on the induction of PRL

release by thyrotropin-releasing hormone described earlier

(Barry and Grau, 1986). Melamed (1993) has shown that

GHRH and GnRH are only effective in vivo and in vitro in a

tilapia hybrid when the fish are reproductively mature.

Furthermore, the releasing factors can be effective in

vivo with reproductively immature animals if the fish are

first injected with T or the synthetic androgen, 17a­

methyltestosterone. Growth hormone-releasing hormone and

GnRH are also effective in vitro with tissues from

reproductively immature tilapia, if the tissues are

co-incubated with these steroids (Melamed, 1993). In con­

trast to sex steroids, the glucocorticoid cortisol has a

direct and potent stimulatory effect on GH secretion

(Nishioka et al., 1985; Helms et al., 1987). Finally,

insulin-like growth factor-I inhibits GH release in tele­

osts as in mammals (Perez-Sanchez et al., 1992).

Possible role of GnRH in PRL regulation

One area of regulation of reproduction in which the

teleost system may differ from the mammalian system is

with regard to GnRH. First, while most eutherian mammals

have only one form of GnRH, mammalian-GnRH, teleosts, like

9

most vertebrates examined to date, have multiple forms

(see review, Sherwood et ai., 1994). By convention, the

GnRH molecules are named for the animals from which they

were first identified. There is evidence for specificity

of function for the different forms of the peptide.

Chicken-GnRH II (cGnRH-II) was found to be more potent in

releasing GtH from peri fused fragments and dispersed cells

of the goldfish pituitary than sGnRH, while sGnRH was

found to be the more potent of the two in releasing growth

hormone (GH) from the pituitary fragments, although equip­

otent with dispersed cells (Chang et ai., 1989; Peter et

ai., 1990).

The possibility that one or more of the multiple

forms of GnRH present in the tilapia may stimulate PRL

release was investigated as part of the studies described

in this dissertation. Marchant and colleagues (1989) have

shown that GnRH stimulates GH release in the goldfish, and

Melamed and colleagues (1995) have shown the same for a

tilapia hybrid. However, Cook and colleagues (1991) were

not able to detect GnRH binding to PRL cells of the gold-

fish despite binding to GtH and GH cells.

The involvement of GnRH in PRL regulation is not well

understood in mammals or in fishes. In mammals, GnRH has

been shown to stimulate PRL release in vivo when basal PRL

release is within normal ranges, but inhibit PRL release

10

when basal PRI release is unusually high as in hyperpro­

lactinemia (Debeljuk et al., 1985j Kugu et al., 1988j

Sridaran et al., 1988). Injections of GnRH have been

shown to increase PRL cell activity in the Atlantic salm­

on, based on an immunocytological and electron microscopi­

cal study (Ekengren et al., 1978). The authors suggest

that this effect was mediated via GnRH stimulation of GtH

release which in turn, stimulated E2 release, a potent

stimulator of PRL cell activity.

One example of PRL stimulation by GnRH in a mammal

may not fit the teleost. Denef and Andries (1983) have

shown that in the fourteen-day-old rat, GnRH stimulates

gonadotroph cells to release a paracrine factor that

stimulates PRL release. This factor is not one of the

GtHs. They demonstrated that PRL release is stimulated

when a population of dispersed pituitary cells, including

both PRL and GtH cells, are treated with GnRH. However,

if GtH cells are removed from the culture, GnRH has no

effect. Furthermore, when PRL cells are incubated in

media from GnRH stimulated GtH cells, PRL release is

increased. Recent studies suggest that angiotensin II may

be this paracrine factor (Becu-Villalobes et al., 1994).

The anatomical arrangement of the tilapia pituitary makes

this same paracrine interaction less likely. The PRL

11

cells of the tilapia are segregated into the RPD while the

GtH cells are in the proximal pars distalis.

The role of Ca2 + in tilapia PRL release

Indirect evidence suggests that PRL release in

response to an osmotic signal is Ca2+ dependent in the

tilapia (see reviews, Grau and Helms, 1989; Grau et al.,

1994). Recently, Borski (1993) has used the Ca 2+­

sensitive fluorescent dye, fura-2, to show that intracel­

lular free Ca 2+ concentrations ([Ca2+]i) increase within

20 sec after exposure to reduced osmolality medium,

further supporting a role for Ca2+ in PRL release. Now

that homologous radioimmunoassays are available for tila­

pia PRLs, the time-course for PRL release in conjunction

with changes in [Ca2+]i, in response to osmotic stimuli,

should be investigated. Jobin and Chang (1992) have also

shown that native GnRH stimulates increases in [Ca 2+]i in

GtH and GH cells of goldfish. As part of my studies, I

examined simultaneous changes in [Ca2+]i and the release

of the two tPRLs in response to reductions in medium

osmolality and GnRH.

Roles and regulation of PRL and GH in osmoregulation

The tilapia, o. mossambicus, is a euryhaline teleost

fish that reproduces and thrives in habitats ranging in

12

salinity from fresh water (FW) to hypersaline seawater

(SW). Prolactin and GH are regulators of physiological

processes in the tilapia, which confer the ability to

adapt to various salinities.

Prolactin plays a central role in FW adaptation and

GH is increasingly believed to play a role in SW adapta­

tion in tilapia and other euryhaline teleost fishes.

Prolactin maintains extracellular salt and water balance

when the fish are in a low salinity environment by acting

on virtually all osmoregulatory tissues including the

gills, integument, urinary bladder, intestine and kidney

to reduce water permeability and increase sodium retention

(Clarke and Bern, 1980; Hirano, 1986). Growth hormone has

been shown to act in SW adaptation although its mode of

action is not as well known as for PRL. Studies suggest

that GH affects ion transport at the gills and may be

working in conjunction with cortisol and insulin-like­

growth factors, or through a stimulation of cortisol and

insulin-like-growth factors (see review, Sakamoto et al.,

1993, Sakamoto et al., 1994).

Consistent with the roles of PRL and GH in osmoregu­

lation, PRL cell activity is enhanced when tilapia are

maintained in FW and is reduced in SW fish, while the

reverse is true for GH cell activity (Dharmamba and

Nishioka, 1968; Clarke et al., 1973; Nagahama et al.,

13

-- - _.-_.-- ---~-----------

1975; Nishioka et al., 1993; Borski et al., 1994). Small

changes in medium osmolality, well within the physiologi­

cal range of the tilapia, alter PRL and GH release from

pituitary tissues (Nagahama et al., 1975; Grau et al.,

1982; Helms et ali 1987). Grau and colleagues (1982) have

shown that PRL release from PRL cells is increased in

response to reductions in medium osmolality within 10-20

mini time-course studies have not been conducted for GH

cells. Nagahama and colleagues (1975) have shown that the

PRL cells respond to changes in medium osmolality and not

sodium or chloride ions per se, by demonstrating that the

response of PRL cells to changes in medium osmolality are

the same whether the changes in medium osmolality are due

to differences in NaCl concentration or are made by

changes in the membrane-impermeant molecule, mannitol. It

has not been determined whether changes in medium osmolal­

ity or the osmotic gradient across the cell membrane

leading to a change in osmotic pressure, evoke the osmotic

response. This question was addressed as part of this

dissertation.

Circulating levels of GH and PRL change during adap­

tation of o. mossambicus to different salinities.

Circulating and pituitary levels of PRL were observed to

decrease when tilapia were transferred or acclimated from

FW to SW and to increase when transferred or acclimated

14

from SW to FW (Nicoll et ai., 1981; Borski et ai., 1992;

Ayson et ai., 1993; Yada et ai., 1994).

Borski and colleagues (1994) found that the pituitary

content of GH was almost twice as high in male tilapia

raised in SW for 7 months compared with male tilapia

raised in FW fQr 7 months. In addition, pituitary GH

levels were reduced in male fish transferred from SW to FW

and increased in fish transferred from FW to SW after 49

days. Ayson and colleagues (1993) did not see an increase

in pituitary GH in O. mossambicus (sex not reported) 3-4

weeks after transfer from FW to SW. Yada et ai. (1994)

observed an increase in plasma GH concentrations shortly

after transfer of male o. mossambicus from FW to 70~-SW

but no change in levels in females. Growth hormone con­

centrations in the plasma were reduced shortly after

transfer from SW to FW in both sexes, compared to animals

transferred from SW to SW, at the same time points.

Circulating levels of GH were no longer significantly

different from controls, in either direction of transfer

and either sex, by 7 days after transfer. Sakamoto and

colleagues (1994) also observed an increase in plasma GH

concentrations following transfer of O. mossambicus from

FW to SW, whereas plasma GH was not altered following

transfer from SW to FW. Furthermore, the increase in

plasma GH was observed in both males and females and at

15

the end of the study, day 14 after transfer, plasma GH was

elevated in males but not in females (Sakamoto, personal

communication) .

The patterns of circulating GH concentrations

observed in o. mossambicus following transfer from FW to

SW or SW to FW are similar to those observed in salmonids

undergoing comparable salinity changes (see review,

Sakamoto et al., 1993; Yada et al., 1994; Sakamoto et al.,

1994). Circulating levels of GH increase transiently or

do not change following transfer from FW to SW or SW to FW

in salmonids (see review, Sakamoto et al., 1993). Trans­

fer of rainbow trout and coho salmon to SW was accompanied

by an increase in metabolic clearance rate and the

calculated secretion rate of GH whereas clearance kinetics

did not change in coho salmon transferred from SW to FW

(Sakamoto et al., 1990, 1991). Whether there are changes

in metabolic clearance rate or secretion rate of GH in

tilapia following adaptation to FW or SW has not been

determined.

Roles and regulation of PRL and GH in metabolism

Prolactin and GH have effects on protein, carbohy­

drate, and lipid metabolism in mammals (see reviews,

Nicoll, 1974; Fain, 1980; Davidson, 1987). These metabol­

ic actions appear to be conserved for GH in teleosts,

16

while the actions of PRL in metabolism have received

little attention.

Prolactin can be lipolytic or lipogenic in teleosts,

depending on the influences of temperature and photoperi­

od, circadian rhythms, and development (Lee and Meier,

1967; de Vlaming and Pardo, 1974; de Vlaming et ai., 1975;

Horseman and Meier, 1979; Sheridan, 1986). There has been

only one study examining circulating PRL concentrations

with changes in metabolic state. In this study, circulat­

ing PRL levels were not altered in Kokanee salmon

(Oncorhynchus nerka), fasted for 30 days (McKeown et ai.,

1975). Nevertheless, Rodgers and colleagues (1992) found

both pituitary content of PRL and basal release from RPD

in vitro are decreased after 2 weeks of fasting in O.

mossambicus. In this same study, PRL release in vitro was

negatively correlated with the concentration of essential

amino acids added to the incubation medium and was

unaffected by D-glucose concentration. Although PRL

appears to have metabolic actions in teleosts, the roles

of PRL in regulating metabolism are still unclear. I

examined changes in serum and pituitary levels of PRL in

response to fasting as part" of this thesis.

The role of GH in metabolism has received more atten­

tion in teleosts than has the role of PRL, but is not

without controversy. Growth hormone treatment has been

17

shown to increase free fatty acids (FFA) in the circula­

tion of rainbow trout and goldfish and to mobilize lipid

reserves from the liver of rainbow trout and coho salmon

(Minick and Chavin, 1970; Leatherland and Nuti, 1981;

Sheridan, 1986). In addition, GH injections increased

muscle FFAs but not plasma FFAs or glucose, and increased

liver glycogen in Kokanee salmon (McKeown et al., 1975).

Different responses to GH treatment were observed with the

eel, Anguilla japonica, and tilapia, o. mossambicus.

Injections of bovine-GH in hypophysectomized eels

increased serum amino acid concentration and did not alter

plasma lipid concentration (Inui et al., 1985). Injec­

tions of bovine-GH stimulate the release of glycogen and

not lipid from liver reserves of the tilapia, and increase

serum concentrations of amino acids and glucose but not

protein, lipid or cholesterol (Leung et al., 1991).

Consistent with this elevation in circulating amino acids

and glucose with fasting, Rodgers and colleagues (1992)

found an increase in GH release from pituitary tissues in

response to reductions in essential amino acids and

glucose in the incubation medium, in the same species.

Together the results of these studies suggest that GH

cells act in the regulation of, and are directly respon­

sive to changes in the blood concentrations of amino acids

and glucose in the tilapia.

18

Increases in circulating and pituitary levels of GH

have been observed in fasted teleosts. Increases in

plasma GH levels have been observed within 3 weeks of

fasting in rainbow trout by Wagner and McKeown (1986), and

within 1 week by Sumpter et al. (1991). No change in

plasma GH was observed after 30 days of fasting in Kokanee

salmon (McKeown et al., 1975). Farbridge and Leatherland

(1992b) observed a bi-modal response to fasting in rainbow

trout. Plasma GH concentrations were lower in fasted fish

than in fed fish at 2 weeks of fasting, followed by an

elevation in circulating GH concentrations in fasted fish

at 4 weeks. Plasma GH concentrations were elevated at 2

weeks of fasting in a repeat of the study. Melamed (1993)

observed a rise in plasma GH within 16 days of fasting in

a tilapia hybrid. Also in tilapia, o. mossambicus, Rogers

and colleagues (1992) found that pituitary content of GH

was elevated with 2 weeks of fasting.

Farbridge et al. (1992) found that feeding rainbow

trout only once every five days resulted in reduced plasma

GH and suggested that GH concentrations are depressed by

low feeding rates. This is not consistent with the study

on the tilapia hybrid, by Melamed (1993). The tilapia at

the start of the experiment by Melamed, were described as

"sub-optimally fed". When the fish were switched to an

increased feeding rate, plasma GH levels fell within 8

19

days. The nutritional state of the animal prior to fast­

ing and changes in nutritional state during fasting is

likely to affect the hormonal response to subsequent

fasting. For this reason, I examined changes in serum and

pituitary levels of GH and PRL, in response to fasting, in

tilapia which had been well-fed and tilapia which had been

on a restricted diet. I also measured changes in body

parameters including body weight, length, gonad weight,

and liver weight in response to fasting as a means of

assessing changes in the nutritional state of the animals.

The reproductive cycle of the female tilapia

Prolactin and GH have actions in reproduction in

teleosts. Before reviewing these actions, I will first

discuss what is known about the reproductive cycle of the

female tilapia. Tilapias are important food fishes world­

wide and therefore their reproductive habits and physiolo­

gy have received much attention. The tilapia o. mossambi­

cus, can reproduce in FW and in SW. After the female lays

her eggs and they are fertilized by the male, she picks up

the eggs and broods them in her buccal cavity. Brooding

continues for about 3 weeks. The female tilapia o. mos­

sambicus, can reproduce every 20-25 days if they do not

brood. If the female does brood, the inter-spawn period

is extended to approximately 40 days. Smith and Haley

20

(1987, 1988) have described sex steroid and ovarian

morphology profiles of the reproductive cycle for brooding

and non-brooding female O. mossambicus. Plasma steroid

hormone levels and ovarian tissue responsiveness to GtH at

different phases of the breeding cycle have also been

examined for another tilapia species, O. aureus, by

Bogomolnaya and colleagues (1984).

Smith and Haley (1987) found that postovulatory

follicles remain viable over the length of the brooding

period in fish that brood, but regress quickly in those

females which do not. They have provided evidence that

these structures produce steroid hormones, most strongly

during the first 7 days after spawning. In addition,

blood levels of E2 and T are elevated during the latter

phase of the brooding cycle, even though oocytes consti­

tuting the next clutch to be spawned are arrested in early

vitellogenesis during this time. Both E2 and T are also

elevated during brooding in O. aureus (Bogomolnaya and

colleagues, 1984). Smith and Haley (1987) found that

ovarian growth is not arrested in non-brooding female

tilapia that are fasted, however, the oocytes show signs

of atresia at latter stages. Smith and Haley (1987, 1988)

suggest that E2 and testosterone, and postovulatory folli­

cles, may be involved in parental care behavior, as well

as the arresting of oocyte growth during brooding and the

21

protection of oocytes from atresia in the tilapia. Based

on what is known about the actions of PRL and GH in

reproduction and metabolism, PRL and GH should also be

considered in these roles. Furthermore, PRL and GH may be

involved in the elevation of the steroid hormone levels

and in prolonging the life of the postovulatory follicles.

In addition to the reproductive roles PRL and GH may

have during the reproductive cycle of the tilapia, the

hormones may be called upon to regulate osmotic homeosta­

sis and metabolism. This would depend on both the salini­

ty in which the animals reproduce and on metabolic changes

that may occur in response to reduced feeding during the

brooding phase of the reproductive cycle. For this reason

I characterized changes in serum and pituitary levels of

the tPRLs and GH during the reproductive cycle of FW- and

SW-adapted female tilapia as well as fed and fasted

tilapia as part of my studies.

Prolactin and teleost perentzeI care behaviors

The roles of PRL in vertebrate reproduction are

diverse; often they are associated with the nurturing of

young. These include nestbuilding and protective behav­

iors as in birds, rabbits and rats, and mitogenic effects

associated with maternal tissue derived feeding, such as

22

mammary development and lactation and pigeon cropsac

development (cf. Nicoll and Bern, 1971).

Similar actions have been attributed to PRL in

teleosts; these include parental care behaviors such as

fanning of eggs and nests by sticklebacks, bluegills and

cichlids. Also in cichlids, "calling movements" to fry, a

behavior exhibited by o. mossambicus, and reduced feeding

behavior suggested to prevent the cannibalism of young

during brooding, have been attributed to PRL. Prolactin

has been shown to induce mucus secretion in the discus

fish, a source of nutrition for developing fry (Noble et

al., 1938; Blum and Fielder, 1965; Slijkhuis et al., 1984;

DeRuiter et al., 1986; Kindler et al., 1991). The tilapia

is a cichlid and displays parental care behaviors similar

to those attributed to PRL in other cichlids. Thus, PRL

may be involved in the regulation of parental care behav­

ior in the tilapia.

Prolactin cell activity was shown to be increased in

male sticklebacks displaying fanning behavior, based on

ultrastructure morphometry and r3H] - l y s i ne incorporation

rate of PRL cells (Slijkhuis et al., 1984). However, the

same group (Wendelaar Bonga et al., 1984) found no differ­

ences in PRL cell activity between brooding and non-brood­

ing female tilapia using these same techniques. While

23

this conclusion may be correct, further investigation is

warranted.

Roles of PRL in teleost reproduction

Evidence suggests that PRL has roles in reproduction

in teleosts in addition to regulating behavior. As I have

discussed earlier, the sex steroids E2 and T stimulate PRL

cell activity directly and sensitize PRL cells to thyro­

tropin-releasing hormone stimulation. Studies examining

changes in circulating or pituitary PRL levels with

reproductive cycles are limited but do provide support for

PRL having roles in reproduction. Prolactin has been

implicated in the regulation of vitellogenesis (covered in

detail in the next section) and steroidogenesis. Finally,

specific binding of ovine-PRL has been observed in mem­

brane preparations of ovary and testis of tilapia (Edery

et al., 1984).

Hirano et al. (1986) and Prunet et al. (1990) both

cite preliminary data suggesting plasma PRL concentrations

change during the reproductive cycle in female salmonids

and suggest an inverse correlation with plasma progestins.

The rise in PRL in rainbow trout described by Prunet and

colleagues (1990) did not occur until 4 weeks after

ovulation. Changes in prolactin bioactivity in sera and

pituitary of the FW catfish, Clarias batrachus, assessed

24

using a pigeon crop sac assay, paralleled changes in GtH

concentrations (Singh and Singh, 1981). In contrast, no

changes in ultrastructure of PRL cells were observed

during the reproductive cycle of the sailfin molly,

Poecilia latipinna (Young and Ball, 1983).

To date, there have been only three studies on the

effects of teleost PRL on steroidogenesis. Singh and

colleagues (1988) used purified salmon PRL to examine

gonadal steroidogenesis in hypophysectomized Fundulus

heteroclitus, and found salmon PRL significantly increased

plasma concentrations of T in males but had no effect on

steroid levels in females and no effect on steroid release

in vitro by either testes or ovaries despite preventing

the decline in gonadal weight usually associated with

hypophysectomy. Both tPRLs, tPRL1 88 and tPRL1 7 7, were

tested for their effects on E2 production by vitellogenic

oocytes of the guppy, Poecilia reticulata, (Tan et al.,

1988). Prolactin1 88 stimulated E2 production from all

stages of vitellogenic oocytes of the guppy with the

strongest response elicited from oocytes at the beginning

of vitellogenesis. Prolactin177 was found to have no

consistent effect. Neither of these studies used homolo­

gous PRL, and therefore, the specificity of the responses

to the hormones is open to question.

Rubin and Specker (1992) conducted the only study to

25

examine the effects of homologous PRL on steroidogenesis.

They examined the effects of the two tPRLs on steroid

production by testicular tissues of courting and non­

courting male tilapia. Rubin and Specker (1992) found

similar effects with both tPRLs. The tPRls stimulated T

production in testicular tissue from courting males, but

not in testicular tissue from non-courting males. The

tPRLs also increased ovine-luteinizing hormone-stimulated

T production in courting males but were inhibitory in

non-courting males.

Blum and Weber (1968), showed that injections of

ovine-PRL increased steroid-3g-ol-dehydrogenase activity

in the cichlid, Aquideus pulcher. Young and colleagues

(1983), reported that ovine-PRL increases E2 and 17a,20B­

dihydroxy-4-pregnen-3-one production by ovarian follicles.

Collectively, changes in circulating PRL levels during the

reproductive cycle, effects of exogenous PRL on steroido­

genesis and vitellogenesis, and detection of ovarian and

hepatic PRL receptors strongly indicate a role for PRL in

control of fish reproduction.

Roles of GH in teleost reproduction

Several lines of evidence suggest GH may have roles

in fish reproduction. Circulating GH levels change during

the reproductive cycle of fishes; GH preparations affect

26

steroidogenesis and gonadal development; and GH receptors

are present in the ovary and in the liver, the site of

vitellogenin production.

The recent availability of homologous radioimmunoas­

says led to the discovery that GH is present at moderate

levels in the circulation during vitellogenesis (oocyte

growth) or spermatogenesis in several fish species, and

then increases abruptly during or just prior to the spawn­

ing period (Stacey et ai., 1984; Marchant and Peter 1986;

Bjornsson et ai., 1991; Swanson 1991). In pituitaries of

vitellogenic striped bass (genus Morone) , immunoreactive

GH cells were strongly labeled with a heterologous

antiserum to fish GH (Huang and Specker, 1994). The

density of GH cells and their intensity of staining then

decreased in spawning fish, suggesting changes in GH

secretion are linked to final maturation. Changes in

serum GH and GtH levels are closely correlated during the

ovulatory GtH surge in goldfish (Yu et ai., 1991). Sumpt­

er and colleagues, (1991b) on the other hand, found no

significant elevations in GH with reproductive cycle in

female rainbow trout until after ovulation. They suggest

the rise in GH after ovulation and the rise in GH observed

in maturing males, were due to starvation effects and were

not associated with reproduction per se.

27

In salmonids, mammalian GH preparations have long

been known to enhance ovarian growth, increase circulating

levels of sex steroids and promote in vitro ovarian ste­

roidogenesis (Higgs et al., 1976; Fostier et al.,

1983; Young et al., 1983). Similar effects of bovine-GH

were recently observed in spotted seatrout (Singh and

Thomas, 1993). In combined treatments, stimulation of

steroidogenesis by bovine-GH and human chorionic GtH

were additive, confirming that bovine-GH did not merely

potentiate GtH action. Furthermore, bovine-GH increased

follicular aromatase activity, an action dependent upon

the synthesis of new RNA and regulatory protein(s). The

gonadotropic and steroidogenic actions of mammalian GH

have been confirmed using purified and recombinant fish

hormones. Injections of recombinant GH stimulate gonadal

growth in hypophysectomized killifish and elevate circu­

lating sex steroids in immature rainbow trout (Singh et

al., 1988; Danzmann et al., 1990). The recombinant GH

stimulated steroidogenesis by follicles isolated from

hypophysectomized killifish or trout, even in the absence

of GtH preparations (Singh et al., 1988). Purified carp

or chum salmon GH potentiated carp GtH-II stimulation of

in vitro steroidogenesis by both vitellogenic and preovu­

latory goldfish follicles (Van Der Kraak et al., 1990).

However, GH was ineffective when used alone.

28

Most oocyte growth in fishes can be accounted for by

the uptake of a yolk precursor protein (vitellogenin) that

is synthesized and secreted by the liver of maturing

females (vitellogenesis) under the influence of circulat­

ing estrogens, primarily E2 (Specker and Sullivan, 1994).

In frogs and turtles, GH stimulates hepatic synthesis of

vitellogenin in vitro (Carnevali et al., 1992; Ho et al.,

1985). The effect is also seen in vivo in hypophysecto­

mized turtles (Ho et ai., 1982), implying direct action of

GH on the liver independent of circulating GtH. Recently,

Kwon and Mugiya (1994) demonstrated that GH or PRL as well

as E2 are essential for vitellogenin synthesis in the eel.

Although E2 injections could induce vitellogenin synthesis

in intact and sham operated immature eels, they were

ineffective in hypophysectomized animals, confirming that

pituitary hormones play a role in initiating vitellogenin

synthesis. Growth hormone or PRL added to the incubation

medium of cultured eel hepatocytes greatly potentiated

weak responsiveness to E2 alone. Growth hormone and

insulin are known to regulate uptake of vitellogenin

(growth) by cultured oocytes of various vertebrates,

including fishes (reviewed by Specker and Sullivan, 1994)

Finally, specific membrane receptors for GH have

recently been detected in salmon ovary and testes (Le Gac

et ai., 1991; Mourot et al., 1992), confirming that the

29

teleost gonad is a GH target. Hepatic receptors for GH

were described earlier for a variety of teleosts (Gray and

Kelley, 1991; Hirano, 1991; Sakamoto and Hirano, 1991; Yao

et al., 1991). Collectively, the maturational changes in

circulating GH levels, effects of exogenous GH on steroi­

dogenesis and vitellogenesis or oocyte growth, and detec­

tion of ovarian and hepatic GH receptors strongly indicate

a role for GH in control of fish reproduction.

Research ojectives

The spectrum of actions of PRL and GH in teleosts

include effects on osmoregulation, metabolism and repro­

duction. Often, the hormones are called upon to regulate

these different functions simultaneously. The cells

secreting the hormones must therefore receive regulatory

information from a variety of sources to be used in

determining the secretion rate of the hormones. In order

to understand how PRL and GH may be involved in regulating

diverse functions, hormone changes with these functions

must be characterized and the factors regulating PRL and

GH must be identified.

Changes in serum and pituitary levels of the tPRLs

and GH with changes in environmental salinity have been

well characterized in the tilapia O. mossambicus, but

changes with metabolic or reproductive state have received

30

little attention. There is still much to'learn about the

regulation of the tPRLs and GH in osmoregulation,

metabolism, and reproduction. The overall objectives of

my thesis research were to characterize changes in serum

concentrations and pituitary content of the tPRLs and GH

with changes in metabolic and reproductive states and to

characterize further the regulation of these hormones.

Prolactin and GH cells of the tilapia can detect

changes in the osmotic concentration of extracellular

fluids and respond by altering the release of their

hormones (Helms et ai., 1987; cf. Grau et al., 1994). It

is not known whether osmoreceptive cells, such as the PRL

and GH cells, detect changes in the osmolality of fluids

or changes in the osmotic gradient across the cell mem­

brane which lead to an increase in osmotic pressure on the

cell membrane. The first objective of my thesis research

was to make this determination.

Gonadotropin-releasing hormone has been shown to

stimulate not only GtH release in teleosts, including the

tilapia, but also GH release (Marchant et al., 1989;

Melamed et al., 1995). Since PRL and GH are closely

related molecules, derived from a common ancestral gene

and regulated by many of the same factors, my second

objective was to determine whether GnRH has effects on PRL

release. As part of this study, it was also my objective

31

to determine whether calcium operates as a second messen­

ger in mediating the effects of GnRH on PRL release.

Little is known about the roles of PRL and GH in

the regulation of metabolism in teleosts. Treatment of

teleosts including the tilapia, with heterologous PRLs and

GHs indicates that these hormones have effects on the

mobilization of metabolites. There have been few studies

inquiring into the existence of correlations between

changes in circulating and pituitary levels of GH with

alterations in metabolic state in teleosts. Furthermore,

there has been only 1 study describing changes in circu­

lating PRL concentrations and 1 describing changes in

pituitary PRL content with changes in metabolic state

(McKweon et al., 1975; Rodgers et al., 1992). The third

objective of my research was to characterize changes in

serum concentrations and pituitary content of the tPRLs

and GH in relation to alterations in metabolic state

induced by fasting.

Prolactin and GH have reproductive actions in tele­

osts. There have been few studies which examine changes

in these hormones with the reproductive cycle of teleosts.

The fourth objective of my thesis work was to characterize

changes in serum concentrations and pituitary content of

the tPRLs and GH during the reproductive cycle of the

female tilapia. As part of this study, my objective was

32

to determine whether environmental salinity alters the

patterns of serum and pituitary levels of the tPRLs and GH

observed during the reproductive cycle.

33

Chapter II

ALTERATIONS IN OSMOTIC PRESSURE AND NOT OSMOLALITY ARE

COUPLED TO THE SUSTAINED RELEASE OF THE OSMOREGULATORY

HORMONE, PROLACTIN, IN THE EURYHALINE TELEOST, TILAPIA

(OREOCHROMIS MOSSAMBICUS)

INTRODUCTION

Prolactin is the FW-adapting hormone of many euryha­

line teleosts including the tilapia, O. mossambicus.

Prolactin acts on all osmoregulatory tissues to stimulate

ion transport and to reduce ion and water permeability

(see reviews by Clarke and Bern, 1980; Hirano, 1986; Brown

and Brown, 1987; Grau and Helms, 1990). Consistent with

this osmoregulatory action, prolactin cell activity, and

blood and pituitary PRL levels are higher in FW-adapted

tilapia than in SW-adapted tilapia, (Dharmamba and Nishio­

ka, 1968; Nicoll et al., 1981; Borski et al., 1992; Ayson

et al., 1993). Clearly defined osmoregulatory actions

have yet to be defined in mammals.

In addition to the osmoregulatory actions of PRL in

the tilapia, the tilapia PRL cell appears to be an osmore­

ceptor. Support for this notion comes from 3 lines of

evidence. First, studies have demonstrated changes in

34

sustained PRL release, PRL synthesis and PRL mRNA levels

in direct response to small changes in medium osmolality

that are well within the range of plasma osmolalities

observed in vivo (Nagahama et al., 1975; Wigham et al.,

1977; Grau et al., 1981; Kelly et al., 1988; Grau et al.,

1994; Yoshikawa et al., 1995). Furthermore, the effect of

changes in osmolality on PRL release appears to be mediat­

ed by second messenger systems. Prolactin release in

response to changes in medium osmolality is Ca2 + and cAMP

dependent and is accompanied by an increase in intracellu­

lar Ca 2 + (Grau et al., 1981, 1982; Richman et al., 1990,

1991; Borski, 1992; Grau et al., 1994). The increased

release of PRL in response to decreases in osmolality is

specific to PRL cells. The secretory response of PRL

cells to changes in osmolality is not shared by the close­

ly related GH cells of the tilapia. The release of GH

from GH cells of the tilapia is either unchanged or inhib­

ited by decreases in medium osmolality (Helms et al.,

1987; c.f. Grau and Helms, 1989).

The studies demonstrating that the PRL cells of the

tilapia are osmoreceptive cells were conducted using

pituitary tissues of the tilapia which are a nearly homo­

geneous population of PRL cells (95-95%; Nishioka et al.,

1988). These tissues are from the anterior most region of

the rostral pars distalis (RPD). Thus, the PRL cells were

35

responding directly to changes in medium osmolality. The

ability to isolate a nearly homogeneous population of PRL

cells obviates the impediments to investigating most

osmoreceptive cells.

The vasopressin cells of the mammalian hypothalamus

are an example of the complexity of most osmoreceptive

cells. The cell bodies of these neurosecretory cells are

located mainly in the supraoptic nucleus with a smaller

number in the paraventricular nucleus. In both sites, the

vasopressin cells are found among a variety of cell types

and synapse with many others. The axons of vasopressin

neurons project a considerable distance from these loca­

tions to capillaries in the posterior pituitary, the site

of vasopressin secretion. This kind of complexity in

morphology and arrangement has virtually blocked attempts

to clarify the mechanisms by which the osmotic signals

alter vasopressin secretion. In fact, it is still unknown

whether vasopressin neurons are directly osmosensitive, or

are governed by adjoining osmoreceptive cells, possibly

around the anteroventral border of the third ventricle

(see Guyton, 1991).

The combination of PRL's actions in regulating osmot­

ic homeostasis together with the PRL cell1s ability to

respond directly to changes in osmolality demonstrates

that the PRL cell both monitors and regulates the osmolal-

36

ity of extracellular fluids. Little is known about the

cellular mechanisms which monitor the osmolality of extra­

cellular fluids and lead to changes in the release of

osmoregulatory hormones. Basic to this understanding is

the identification of the osmotic signal recognized by the

osmoreceptive cell. Nagahama and colleagues (1975) using

the tilapia PRL cell as a model, showed that osmoreceptive

cells respond to changes in osmolality and not sodium.

They demonstrated that changes in medium osmolality by the

addition of NaCI or the membrane-impermeant molecule

mannitol, result in similar reductions in PRL releases

from PRL cells. Since this study, there has been little

progress towards characterizing the osmotic signal recog­

nized by osmoreceptive cells. Studies on cells which are

not involved in the osmoregulation of extracellular fluids

point to the possibility that osmotic pressure and not

osmolality may be the osmotic signal recognized by osmore­

ceptive cells.

Studies have shown that many cells, including mamma­

lian PRL cells, release hormones in response to changes in

medium osmotic pressure and not medium osmolality per se

(Blackard et al., 1975; Sato et al., 1991; Wang et al.,

1991). These cells include adenohypophysial cells and B­

cells of the pancreatic islets of the rat. Thyrotropin is

released by thyrotropin cells and insulin is released from

37

pancreatic cells in response to reductions in medium

osmolality. These studies have shown that when the NaCI

concentration of isosmotic media is reduced and replaced

with membrane-impermeant molecules, release is not al­

tered, however, when the same amount of NaCI is replaced

with membrane-permeant molecules release is increased.

Unlike membrane-impermeant molecules, membrane-permeant

molecules reach an equilibrium across the cell membrane

and do not contribute appreciably to the osmotic pressure

on the cell membrane. The osmolality inside and outside

the cell are increased equally by membrane-permeant

molecules.

Unlike the tilapia PRL cells, the release of hormones

by the rat PRL, thyrotropin and B-pancreatic cells is not

sustained with continued reduced osmolality (Blackard et

al., 1975; Sato et al., 1991; Wang et al., 1991). There

is a burst of release followed by a return to baseline

levels of release within 10 min of exposure to reduced

osmolality. This temporally limited response to changes

in osmolality is inconsistent with osmolality or osmotic

pressure being a signal which regulates the release of

hormones to serve osmoregulatory functions. Unlike PRL in

the tilapia, the hormones released by these cells have not

been shown to posses clearly defined osmoregulatory ac­

tions in mammals. Thus, it is not likely that a sustained

38

response to osmotic stimuli would be required. Blackard

and colleagues (1975) noted that a reduction in serum

osmolality of the magnitude required to increase insulin

release (20 mOsmolal) does not occur physiologically in

the rat and therefore the significance of this response is

questionable. Furthermore, Sato and colleagues (1991),

also working with the rat, made isotonic medium containing

the membrane-permeant molecule ethanol by removing NaCI to

reduce the osmolality of the medium 80 mOsmolal and then

replacing it with ethanol. The reduction in medium osmo­

lality before the addition of ethanol in the studies by

Sato and colleagues (1991) was an even greater change in

osmolality than in the studies by Blackard and colleagues

(1975) .

The studies just described raise the question of

whether the sustained release of PRL from PRL cells of the

tilapia, in response to reductions in medium osmolality,

results from differences in osmotic pressure and not

osmolality per se. Do cells respond to the overall osmo­

lality of the medium or do they respond to the osmotic

pressure of the medium which results from the osmotic

gradient between the cell cytoplasm and the medium? To

this end, I examined the response of PRL cells to various

medium osmolalities achieved by varying the concentrations

of membrane-permeant or membrane-impermeant molecules.

39

Tilapia RPD were incubated individually in wells for 18-20

hr to examine whether the effect of reduced medium osmotic

pressure on PRL release is sustained. Prolactin tissues

were also incubated in peri fusion to characterize the

short-term response of PRL cells to reduced medium osmotic

pressure and osmolality. Prolactin release in response to

a reduction in osmotic pressure must be sustained for

osmotic pressure to be the physiologically important

signal recognized by osmosensitive cells in monitoring

extracellular osmotic conditions. Sustained increases in

PRL cell activity and hormone effects are necessary for

PRL to maintain osmotic homeostasis in the tilapia.

MATERIALS AND METHODS

Animals

Tilapia, o. mossambicus were collected from brackish

water streams and maintained in FW in outdoor tanks at a

temperature of 22-25°C for at least 2 months before use.

The fish were fed to satiation twice daily with Purina

trout chow (Purina Mills, Inc., St. Louis, MO).

Static incubations

Mature male tilapia 30-60 g were decapitated and

40

their pituitaries removed. Each pituitary was dissected

and the RPD was placed into a well of a Falcon 96-well

culture plate with 100 ~l of Kreb's bicarbonate-Ringer

solution containing glucose (500 mg/l) , L-glutamine (290

mg/l) , and Eagles's minimal essential medium (50X MEM, 20

ml/l: GIBCO; Grand Island, NY) (Wigham et ai., 1977). The

osmolality of the incubation medium was adjusted by

varying -the concentration of NaCl (or treatment molecules)

and measured using a Wescor vapor osmometer (Wescor;

Logan, UT). Treatment media were hyposmotic medium, (300

mOsmolal), or hyposmotic medium made hyperosmotic (355

mOsmolal) by the addition of either NaCl, the permeant

molecules; urea or ethanol, or the impermeant molecule;

D-mannitol. The medium was gassed for 10 min with 95%

O2/5% CO2 (pH = 7.3). The tissues were incubated at 28 ±

1°C under a humidified atmosphere of 95% O2/5% CO2 and

placed on a gyratory platform (80 rpm) for 18-20 hr. All

chemicals were purchased from Sigma, St. Louis, MO unless

otherwise indicated.

Perifusion incubations

Mature male tilapia 200-300 g were used for the

perifusion studies. The RPD were dissected and placed 8

per chamber in parallel chambers with 3 chambers per

treatment. The perifusion apparatus consisted of a peris-

41

taltic pump (Technicon proportional pump model III,

Dublin, Ireland) connected to a fraction collector (ISCO

model 328, Lincoln, NE). The chambers containing the

tissues consisted of a 3.5 cm length of glass tubing (cut

from a 100-200 ~l microdispenser tube; Drummond, Broomall,

PAl sealed at both ends with 120 ~l mesh Nitex screen.

The chambers were placed after the pump. The dead volume

between the introduction of the test media and the

fraction collector was 1.1 mI. Tissues were preincubated

in control medium (355 mOsmolal) overnight and continued

until the introduction of treatment media. Unlike the

static culture medium, the medium used in the perifusion

studies contained 50 I.U./ml penicillin and 0.05 mg/ml

streptomycin. Fractions (100 ~l) were collected every 2

minutes into 1.5-ml polypropylene microcentrifuge tubes

containing 50 ~l of 2% bovine serum albumin in phosphate

buffered saline.

Electrophoresis

Prolactin release during 18-20 hr incubations was

quantified using a combination of gel electrophoresis and

densitometry (Specker et al., 1985a; modified by Kelley et

al, 1988). At the termination of the incubations, media

and tissue were placed in sodium dodecyl sulfate (SDS)-2­

mercaptoethanol buffer, ultrasonically disrupted (Heat

42

Systems W-385 sonicator, Heat Systems-Ultrasonics, Inc.;

Farmingdale, NY) and boiled for 3 min. Samples were then

stored at -80 aC or measured immediately.

The tPRLs were separated using SDS-polyacrylamide gel

electrophoresis (PAGE). A vertical slab gel electrophore­

sis apparatus (Bio-Rad; Richmond, CA) was used. The

samples were stacked in a 4% 37.5:1 acrylamide:bis-acryla­

mide gel and separated in a 15% 37.5:1 acrylamide:bis­

acrylamide gel (12 cm long, 0.15 cm thick). Samples were

subject to electrophoresis at 30 rnA constant current/gel

for 4-5 hr using a voltage- and current-regulated power

supply (ISCOi Lincoln, NE). The gels were stained with

Coomassie blue R-250 dissolved in a 10% methanol, 5%

acetic acid solution. The gels were destained in a 10%

methanol, 7% acetic acid solution until clearly

discernible bands were observed and then stored in a 7%

acetic acid solution. Bands of both tPRLs were quantified

by using a densitometer and proprietary software (Hoefer

Scientific, San Francisco, CA). For all static

incubations, the release of PRL into the incubation medium

was normalized as a percentage of the total PRL in the

tissue and the medium.

Radioimmunoassays

Prolactins in medium collected during peri fusion

43

incubations were measured using the homologous radioimmu­

noassays developed by Ayson et al., (1993) as modified by

Yada et al. (1994). These radioimmunoassays were

also used in studies described in following chapters for

measuring the tPRLs and GH in serum and pituitaries. r

will describe all three assays at this time for the sake

of clarity. Hormones used as labeled-ligand were labeled

with 125r using chloromine-T. The assays were performed

using a double antibody method under disequilibrium condi­

tions. Samples were run in duplicate. The tPRL1 7 7

antiserum (P177-9-4) does not cross-react with either

tPRL1 88 or GH. The tPRL1 8 8 antiserum (P188-1-3) does not

cross-react with tPRL1 7 7 and only slightly (3.1%) with GH.

The GH antiserum (G-4-4) shows slight cross-reactivity

with tPRL1 7 7 (0.4%) and tPRL1 88 (1.6%). Sample values

were corrected for cross-reactivity using the formula:

Corrected value = measured value - %cross-reaction X value

of cross-reacting hormone/laO.

Repeated measurements within the same assay and in

subsequent assays of a pool of FW o. mossambicus gave

intra- and interassay coefficients of variation of 5.9 and

8.8% for tPRL1 77, 6.3 and 9.5% for tPRL1 88, and 7.8 and

10.4% for GH. The sensitivity of the assays, defined as

the amount of measured hormone that can be distinguished

from zero dose and calculated as twice the standard devia-

44

tion at zero dose, was 0.18-0.40 ng/ml for tPRL1 77,

0.19-0.45 ng/ml for tPRL1 88, and 0.14-0.30 ng/ml for GH

when 50 ~l of serum was used. Non-specific binding did

not exceed 5% of total counts.

Statistical analysis

Differences among groups were determined using analy­

sis of variance and the least significant difference test

for a priori pairwise comparisons (Steel and Torrie,

1980). Experiments with only two groups were analyzed

using the unpaired Student's t-test. Data are expressed as

means ± SE.

RESULTS

Static Incubations

Effects of increasing medium osmolality by the addition of

membrane-permeant (urea) and impermeant (mannitol) mole­

cules on PRL release from RPD of the tilapia

The purpose of this study was to determine whether

changes in medium osmolality or changes in the osmotic

gradient across the cell membrane evoke sustained changes

in PRL release. The release of both tPRLs was greater for

tissues incubated for 18-20 hr in medium with reduced

osmolality (300 mOsmolal) compared with release from

45

----------- -----

tissues incubated in medium with increased osmolality (355

mOsmolal), when the increase in osmolality is due to the

addition of NaCl or mannitol (Fig. 1; P < 0.001). In­

creased medium osmolality did not affect PRL release when

the difference in osmolality was derived from the addition

of the membrane-permeant molecule, urea. Mannitol and

NaCl were equally effective at reducing PRL release when

added to medium to increase the osmolality of the medium.

Effects of increasing medium osmolality by the addition of

the membrane-permeant molecule urea, on PRL release from

RPD of the tilapia

The effect of urea on PRL release was investigated.

The membrane-permeant molecule urea had no effect on PRL

release when added to hyposmotic medium (300 mOsmolal),

raising the osmolality 20 mOsmolals and making the medium

isosmotic (320 mOsmolal), or when added to isosmotic

medium, raising the osmolality 35 mOsmolals and making the

medium hyperosmotic (355 mOsmolal) to tilapia plasma (Fig.

2) .

Effects of increasing medium osmolality by the addition of

the membrane-permeant molecule ethanol, on PRL release

from RPD of the tilapia

The purpose of this study was to examine the effects

46

Figure 1. Effects of increasing medium osmolality by theaddition of membrane-permeant (urea) and impermeant(mannitol) molecules on PRL release from RPD of tilapia.Treatment media were made by adding NaCl, mannitol or ureato hyposmotic medium (300 mOsmolal) to increase theosmolality of the media to 355 mOsmolal, numbers indicatebasal osmolality of media plus additional mOsmolalscontributed by mannitol or urea (mean ± BE; N = 12 pertreatment; *** P < 0.001, ns = not significant atP < 0.05).

47

o tPRl188

[] tPRL177 ns

80

***70 I

***

0

NaCI NaCI NaCI NaCI300 355 300 300

+ +Mannitol Urea

55 55

10

60

(I)tn 50ca(I)-(I)'- 40..J~D-

30'#.

20

48

Figure 2. Effects of increasing medium osmolality by theaddition of the membrane permeant molecule urea, on PRLrelease from RPD of tilapia. Treatment media were made byadding NaCI or urea to hyposmotic medium (300 mOsmolal) toincrease the osmolality of the media to the designatedosmolalities, numbers indicate basal osmolality of mediaplus additional mOsmolals contributed by urea (mean ± BE;N = 12 per treatment; ns = not significant at P < 0.05).

49

70 D tPRL188

EJ tPRL177ns ,,'

60 I

50

CDtncaCD 40-e~

0::: 30C.

"eft.20

10

NaCI300

NaCI300

+Urea20

50

NaCI320

nsI

"....

NaCI320

+Urea35

NaCI355

of a second membrane-permeant molecule on PRL release from

RPD. The addition of the membrane-permeant molecule,

ethanol, to medium of varying osmolalities had no effect

on PRL release from RPD (Fig. 3). Prolactin release from

tissues incubated in 300 mOsmolal medium without ethanol

was not significantly different from PRL release from

tissues incubated in the same medium increased to 330

mOsmolal by the addition of ethanol. Prolactin release

from tissues incubated in 330 mOsmolal medium without

ethanol was not significantly different from PRL release

from tissues incubated in the same medium increased to 355

mOsmolal by the addition of ethanol. Prolactin release

from tissues incubated in 355 mOsmolal medium without

ethanol was not significantly different from PRL release

from tissues incubated in the same medium increased to 405

mOsmolal by the addition of ethanol.

Peri fusion Incubations

Time-course for PRL release in response to a change in

medium osmolality from hyperosmotic (355 mOsmolal) to

hyposmotic (300 mOsmolal) , or to hyposmotic medium made

hyperosmotic by the addition of the permeant molecule

urea, or the impermeant molecule mannitol

This study was aimed at investigating the time-course

of possible changes in the release of the tPRLs from PRL

51

Figure 3. Effects of increasing medium osmolality by theaddition of the membrane permeant molecule ethanol on PRLrelease from RPD of tilapia. Media was made by adding NaCIor ethanol to hyposmotic medium (300 mOsmolal) to increasethe osmolality of the media to the designated osmolali­ties, numbers indicate basal osmolality of media plusadditional mOsmolals contributed by ethanol (mean ± BE;

N = 12 per treatment; ns = not significant at P < 0.05).

52

70 D tPRL188nsI

[] tPRL17760

...

53

cells when medium was altered from hyperosmotic medium to

either hyposmotic medium (300 mOsmolal), hyposmotic medium

made hyperosmotic with the addition of the membrane­

permeant molecule, urea (355 mOsmolal), or hyposmotic

medium made hyperosmotic by the addition of the membrane

impermeant molecule, mannitol. The insert in Figure 4

shows the changes in osmolality of the medium in response

to the introduction of hyposmotic medium (300 mOsmolal)

into the perifusion system, following hyperosmotic medium

(355 mOsmolal). The hyposmotic medium was introduced

at the beginning of fraction 1 and the drop in medium

osmolality in the collection tubes occurs primarily in

fraction 4, or 6-8 minutes later.

Prolactin release was not altered when the medium was

switched from hyperosmotic medium to hyperosmotic medium

containing mannitol (Fig. 4). Prolactin release increased

when the medium was switched from hyperosmotic medium to

hyposmotic medium, or to hyperosmotic medium containing

urea. Prolactin release increased more rapidly when RPD

were exposed to hyposmotic medium than when RPD were

exposed to hyperosmotic medium containing urea. Prolac­

tin177 release was greater in RPD exposed to hyposmotic

medium than in RPD exposed to hyperosmotic medium contain­

ing urea, within 2 min after the onset of exposure to the

new medium (fraction 4), but was no longer different at 4

54

Figure 4. Effects of increased medium osmolality by theaddition of membrane permeant (urea) and impermeant(mannitol) molecules on tPRL1 77 release (top graph) andtPRL1 88 (bottom graph) from RPD of tilapia. Media wasmade by adding NaCl, mannitol or urea to hyposmotic medium(300 mOsmolal) to increase the osmolality of the media to355 mOsmolal (hyperosmotic media). Tissues were incubatedin hyperosmotic medium raised to 355 mOsmolal with NaClbefore the media was switch to the treatment media whichwere 300 mOsmolal or hyposmotic medium made hyperosmoticwith mannitol or urea. Fractions are 2 min fractions withnumbering starting with the first sample collected at thetime the treatment medium was added to the peri fusionsystem. The insert shows the rate of osmolality change inthe collected medium when hyposmotic medium is introducedinto the perifusion system (mean ± SE; N = 12 per treat) .

55

~360iii'0 340E

16 Mannitol~ 320-- ---- § 3000

II) 14 :a)( -e- Urea ~ 280- 0 1 2 3 4 5 6 7 8- 12E -*- 300-0') 10c--.. 8cec

60e.......... 4~

-I2c::

a....0

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

16--0II) 14X::-12E-0') 10c-.....c 8ec

60eco 4co~

-I0::: 2a....

00 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Fraction (2 min)

56

min (fraction 5) after the onset of exposure to the new

medium. Prolactin1 88 release was greater in RPD exposed

to hyposmotic medium than in RPD exposed to hyperosmotic

medium containing urea, within 2 and 4 min after the RPD

were first exposed to the new medium (fractions 4 and 5) ,

but was no longer different at 6 min (fraction 6) after

the onset of exposure to the new medium. Peak PRL release

was observed earlier in RPD exposed to hyposmotic medium

than in RPD exposed to hyperosmotic medium containing

urea. Prolactin1 77 release peaked by 4 min after the

onset of exposure to hyposmotic medium and about 10-12 min

after the onset of exposure to hyperosmotic medium with

urea. Prolactin1 88 release also peaked by 4 min after the

onset of exposure to hyposmotic medium and about 8-10 min

after the onset of exposure to hyperosmotic medium con­

taining urea. At 22 min after the initiation of exposure

(fraction 19) to treatment media, the release of both

tPRLs was similar for RPD exposed to hyposmotic medium and

RPD exposed to hyperosmotic medium containing urea.

DISCUSSION

Prolactin release was high in RPD incubated for 18-20

hr in hyposmotic medium (300 mOsmolal) or medium made

hyperosmotic (355 mOsmolal) with the membrane-permeant

57

molecules; urea or ethanol, but was low in RPD incubated

in medium made hyperosmotic with NaCI or the impermeant

molecule, mannitol. These data suggest that it is the

osmotic gradient that leads to an increase in osmotic

pressure, and not the medium osmolality per se, which

accounts for the osmoreceptivity of the tilapia PRL cell.

Furthermore, the greater release from tissues incubated in

the medium with urea compared with mannitol and NaCI was

not due to a stimulatory effect of the urea. This was

demonstrated by showing that the addition of urea to media

of various osmolalities, including osmolalities with

reduced baseline release of the tPRLs, did not increase

PRL release (Fig. 2).

Results similar to these obtained with urea were

obtained with an additional permeant molecule, ethanol.

Ethanol had no effect on PRL release, further supporting

the notion that PRL cells of the tilapia respond to

changes in osmotic pressure (Fig. 3). Contrary to our

findings, Sato et al., (1991) reported isotonic ethanol

solutions stimulate PRL and thyrotropin release from rat

adenohypophysial cells along with inducing cell swelling.

In the study by Sato and colleagues, medium at 300 mOsmo­

lal was diluted 80 mOsmolal with distilled water and the

osmolality restored to 300 mOsmolal with ethanol. The

induced release as well as the observed swelling could

58

have been induced by the reduced osmolality of the diluted

medium with the permeant ethanol molecules having no

effect.

Prolactin177 and tPRL18 8 release from RPD incubated

in hyposmotic medium and in hyperosmotic medium containing

urea were the same at 22 min after the onset of exposure

to these media. Furthermore, release of both tPRLs in RPD

exposed to both media, were greater than the release of

both PRLs in RPD exposed to hyperosmotic medium containing

mannitol. These data support the results of the my static

incubation experiments and the conclusion that PRL release

in response to changes in osmotic pressure is sustained

beyond the 10 min response observed in cells which secrete

hormones that are not known to regulate the osmolarity of

extracellular fluids, specifically: mammalian PRL, thyro­

tropin and pancreatic B-cells (Blackard et al., 1975; Sato

et al., 1991; Wang et al., 1991).

The rise in release of the tPRLs upon exposure of RPD

to hyposmotic medium made hyperosmotic with urea was

slower than that observed after the introduction of hypos­

motic medium, ~lthough PRL release in both treatments

converged by 22 min after the initiation of exposure (Fiq.

4). Thus, urea affected only the early phase of the

response and not the later phase of the response.

The difference in response of tissues incubated in

59

hyposmotic medium compared with tissues incubated in

hyposmoticmedium made hyperosmotic with urea is consist­

ent with urea having a diffusion rate across the cell

membrane of the PRL cells that is slower than that of

water. Until the urea molecules diffuse across the

membrane, there is an osmotic gradient across the cell

membrane. The osmotic pressure on the membrane due to the

medium made hyperosmotic with NaCI or mannitol, and the

medium made hyperosmotic with urea would be similar imme­

diately after exposure and thus, PRL release would be

expected to be the same. As the urea molecules diffuse

across the membrane, the osmotic gradient across the cell

membrane decreases as does the osmotic pressure. As this

decrease in osmotic pressure occurs, PRL release in­

creases. The time required for the urea to diffuse across

the membrane is the cause of the delay in PRL release

observed between the RPD exposed to hyposmotic medium and

RPD exposed to hyposmotic medium with added urea. Once

the urea comes into equilibrium across the membrane, only

the impermeant molecules in the initial hyposmotic medium

affect the osmotic pressure across the cell membrane.

This is the same in both the hyposmotic medium and the

hyperosmotic medium containing urea. At this point,

release of the PRLs would be expected to be equal in RPD

incubated in the two media, which is what is observed at

60

22 min after the initiation of exposure to the two media.

In summary, alterations in osmotic pressure and not

osmolality are coupled to the sustained release of the

osmoregulatory hormone, prolactin, in the euryhaline

teleost, O. mossambicus.

61

Chapter III

GONADOTROPIN-RELEASING HORMONE FUNCTIONS AS A

PROLACTIN-RELEASING FACTOR IN THE TILAPIA,

OREOCHROMIS MOSSAMBICUS

INTRODUCTION

Teleosts, like most vertebrates other than placental

mammals, have multiple forms of GnRH in the brain which

differ in their spatial distribution throughout the brain.

Three endogenous forms of GnRH have been identified in the

brain of tilapia: salmon-GnRH (sGnRH), chicken-GnRH-II

(cGnRH-II) and seabream-GnRH (sbGnRH) (Weber et al.,

1994). The GnRH molecules are named for the first animal

from which they were identified. The presence of multiple

forms of GnRH together with multifarious distributions of

these forms has led to speculation that separate functions

may be subserved by distinct GnRH peptides. These func­

tions may extend beyond the regulation of GtHs since GnRH

has also been shown to function as a regulator of GH in

several teleosts, including tilapia (Marchant et al.,

1989; Melamed et al., 1995).

Growth hormone and PRL are part of the same polypep­

tide family and are thought to be derived from a common

62

ancestral gene (Niall et al., 1971). Not surprisingly,

PRL and GH are regulated by many of the same factors:

somatostatin, cortisol, and changes in osmotic pressure

(cf. Nishioka et al., 1988). Despite these commonalities

between GH and PRL, and the knowledge that GnRH regulates

GH release, the effects of GnRH on PRL release in teleosts

have received little attention.

There is little direct evidence to support or oppose

a role for GnRH as a regulator of PRL in teleosts. In the

Atlantic salmon, injections of GnRH have been shown to

increase PRL cell activity, based on light and electron

microscopical evidence (Ekengren et al., 1978). There was

an increase in the synthetic and exocytotic activity of

the cells. It was thought, however, that this effect was

mediated via GtH and E2, with GnRH inducing rises in

circulating concentrations of GtH and then E2 which in

turn stimulated the increase in PRL cell activity. Gona­

dotropin-releasing hormone did not bind to PRL cells of

the goldfish, suggesting that GnRH may not be a direct

regulator of PRL cell function in this species (Cook et

al., 1991).

In mammals, GnRH has been found to stimulate PRL

release indirectly via paracrine secretions of GtH cells

(Denef and Andries, 1983; Lamberts et al., 1989). There

is strong evidence that angiotensin II from gonadotrophs

63

is a paracrine mediator of this action (cf. Saavedra,

1992; Becu-Villalobos et al., 1994). In mammals,

prolactin cells are found in very close association with

GtH cells (Sato, 1980). In contrast, the PRL cells in the

tilapia are segregated into the anterior most region of

the RPD as an almost homogeneous mass, while the GtH cells

are located in the proximal pars distalis (Bern et al.,

1975). This anatomical arrangement of the tilapia pitui­

tary gland makes a paracrine-mediated effect through GtH

cells less likely.

Jobin and Chang (1992) have recently shown that

native GnRHs stimulate increases in [Ca2+]i in GtH and GH

cells of the goldfish. Borski (1993) has also recently

shown that [Ca2+]i is increased in PRL cells of the

tilapia with exposure to reduced medium osmolality. The

effects of GnRH on changes in [Ca2+]i in PRL cells of a

teleost have not been investigated.

The purpose of the present study was to: 1) determine

whether GnRH injections can affect circulating levels of

PRL; 2) determine whether GnRH can alter the release of

PRL from tilapia RPD and to compare potencies of the

native GnRH forms in releasing PRL; 3) characterize the

effect of medium osmolality on the ability of GnRH to

stimulate PRL release; 4) determine whether GnRH alters

[Ca2+]i and compare temporal changes in [Ca2+]i with

64

temporal changes in PRL release; and 5) determine whether

the steroid hormones, E2 or T, alter the affect of GnRH on

PRL cells; both steroids stimulate PRL release.

MATERIALS AND METHODS

Animals

Tilapia, o. mossambicus, were collected from brackish

water streams and maintained in FW in outdoor tanks at a

temperature of 22-25°C for at least 2 months before use.

The tanks were supplied with a continuous flow of FW and

constant aeration. The fish were fed to satiation twice

daily with Purina trout chow. All fish were mature adults

and ranged in size from 70-120 g body weight.

Injection study

Mature male and female tilapia were used in the

study. Fish were moved to indoor oval fiberglass tanks

(60 1), 8 individuals per tank, and allowed to acclimate

under a fixed photoperiod of 14L:10D to the tanks for at

least 2 weeks before injections. The tanks were supplied

with a continuous flow of FW and constant aeration. The

fish were not fed on the day of the experiment. Fish were

anesthetized with 2-phenoxyethanol at a concentration of 1

mlll for 1 min before being weighed and given intraperito-

65

neal injections of an analog of mammalian~GnRH (mGnRHai

des-Glyl0, [d-Ala6JmGnRHi Sigma, St. Louis, MO) dissolved

in 0.9% saline (0.1 ~g mGnRHa/g body weight) or saline

for injected controls. Non-injected controls were also

anesthetized. Injected animals received an injection

volume of 1 ~l/g body weight. Fish to be sampled at 11 hr

post-injection were injected 9 hr before the other treat­

ments so that all treatments were sampled between 3-6 PM.

Fish were also anesthetized for blood collection.

Static incubations

Procedures for static incubations are described in

detail in chapter II. Mature male tilapia were used in

the study. The sGnRH, cGnRH-II and chicken GnRH-I (cGnRH­

I) used in the study comparing these 3 forms, was pur­

chased from Peninsula Laboratories, San Carlos, CA. The

sGnRH, cGnRH-II and sbGnRH used in the study comparing

these 3 forms were a gift from the Salk Institute. The

GnRHs were dissolved in hyperosmotic medium to 1 roM and

the steroid hormones (Sigma) were dissolved in absolute

ethanol to 1 roM before dilution in medium. In studies

which included steroid hormones which were dissolved in

100% ethanol, all treatments received equal volumes of

ethanol.

66

Perifusion incubations

Procedures for perifusion incubations are described

in detail in chapter II. Eight RPD from mature males were

placed in each of 6 parallel chambers. The tissues were

preincubated in control medium (355 mOsmolal) overnight.

Seabream-GnRH (Salk Institute) was introduced as 5 min

pulses in graded doses of 0, 0.1, 1, 10, 100, and

1000 nm with 2 hr between pulses. Ten min fractions were

collected. Half of the chambers received increasing

concentrations of GnRH and the other half decreasing

concentrations.

Cell dispersion for fCa 2+Ji measurement

Cells from RPD were dispersed following the method

described by Borski (1993). Prolactin cells of the RPD

were dispersed by first placing a single tissue in 0.5 ml

of a 0.25% solution of porcine trypsin 1-300 (US Biochem,

Cleveland, OH) in phosphate buffered saline (PBS; pH =

7.5; 355 mOsmolal) for 45 min at 28 ± ICC and then gently

passing the tissues through a 1 ml pipette 5-10 times.

The cells were centrifuged (250 X g, 5 min, 25 ± ICC), the

supernatant was decanted, and the cells were resuspended

and triturated in 1 ml of trypsin free phosphate buffered

saline. The cells were rinsed in this way 3 additional

times and then resuspended in culture medium adjusted to

67

355 mOsmolal. The cells were plated onto a glass cover­

slip coated with 0.1 mg/ml of poly-L-lysine. Cells were

preincubated in 355 mOsmolal medium for at least 12 hr

prior to the determination of [Ca2+]i.

Monitoring fCa 2+Ji by dual microspectrofluorometry with

fura-2

Validations and procedures used in this study were

described in detail by Borski (1993). The procedures are

described again, briefly. Prolactin cells, plated

on poly-L-Iysine coated cover slips and incubated in

hyperosmotic medium, were loaded with 10 ~M of fura-2/AM

(Molecular Probes, Eugene, OR), for 90 min at 28 ± 1°C.

The fura-2/AM was solubilized in anhydrous dimethyl

sulfoxide (DMSOj Aldrich Chemical, Milwaukee, WI) to a

concentration of 10 mM prior to its final dilution to 10

~M « 0.10% DMSO v/v). The fura-2/AM is cleaved to its

impermeable form, fura-2 by endogenous esterases within

the PRL cells. The cover slips were then mounted in a

chamber that allows the cell to be peri fused and the

chamber was in turn mounted on a microscope stage. The

PRL cells were perfused for at least 30 min in hyperosmot­

ic medium to allow the fura-2/AM to further deesterify

(Grynkiewcz et al., 1985j Lewis et al., 1988).

Single cell measurements of the fura-2 ratio were

68

made with a dual excitation spectrofluorometer (ARCM-MIC­

N, Spex Industries, Edison, NJ) interfaced with a

Diaphot-TMD inverted microscope (CF 40 X oil immersion

fluorite objective, Nikon). The microscope was equipped

with fluorescence optics, a 50 watt halogen illuminator,

epifluorescence illumination, and a quartz nose-piece (for

UV). Excitation light alternated between 340 and 380 nM

(narrow bandpass filters, SPEX) by a computer-controlled

chopper mirror. Fluorescent emission intensity was

transduced every 2 sec by a photomultiplier tube focused

on a single PRL cell after it had passed through a 500 nM

emission filter. A pinhole (1 mm) placed in the epiillu­

mination path restricted the UV illumination to only the

cell of interest.

All data are expressed as the relative intensity of

the ratio of fura-2 fluorescence excited at 340 nm (fura-2

bound to Ca 2+) to that excited by 380 nm (free fura-2)

from which background (autofluorescence) was subtracted.

Shifts in this ratio (340/380) result directly from

changes in [ca2+]i which are independent of

dye concentration, cell thickness, and absolute optical

efficiency of the instrument (Tsien et ai., 1985;

Grynkiewicz et ai., 1985; cf. Poenie et al., 1986).

69

Incubations for measuring changes in [Ca2+]i in

conjunction with PRL release

Experimental media (hyperosmotic medium with or

without cGnRH-II, and hyposmotic medium) were maintained

at 28 ± 1°C in hanging 60 ml plastic syringes connected to

an eight-point manifold perifusate selector (Hamilton Co.,

Reno, NV) via one-way stop cocks and polyethylene tubing.

The manifold output is connected to the input port of the

perifusion chamber by another piece of polyethylene tub­

ing. The rate of perifusion through the chamber was

maintained at 300 ~l/min by keeping the height of the

syringes and volume of all solutions in the syringes

constant throughout the experiment. Chicken-GnRH-II was

used in the study because it showed the widest range of

stimulation in our static incubation studies. The

cGnRH-II was dissolved in hyperosmotic medium (0.1 pM

cGnRH-II, 355 mOsmolal). Hyposmotic medium (300 mOsmolal)

was introduced to the system at the end of each run.

The amount of the tPRLs released from the dispersed

cells was insufficient to measure tPRL release into the

media. For this reason, in 2 runs of the experiment, 4

RPD were placed in a chamber taken from the perifusion

system described earlier and placed inline after the

chamber with the dispersed cell preparation used for

[ca2+]i measurements. The experimental medium flowed

70

through the chamber with the cells for [ca2+]i

measurements and then through polyethylene tubing to the

chamber with the 4 RPD, and finally to a fraction

collector. One min fractions were collected and tPRL in

the media were measured as described for the perifusion

system earlier.

Quantification of PRLs

Prolactin release was quantified using a combination

of gel electrophoresis and densitometry for 18-20 hr

cultures and radioimmunoassay for 3 hr cultures and serum.

The electrophoresis and radioimmunoassay procedures are

described in detail in chapter II. For perifusion incuba­

tions, the data are expressed as percent of baseline

release which is the sum of the concentrations of hormone

in each of the three fractions following and including the

introduction of the GnRH (b1,b2,b3), divided by the sum of

the hormone in the three fractions preceding the introduc­

tion of the GnRH (a1,a2,a3), expressed as percentage

((b1+b2+b3)/(a1+a2+a3) X 100).

Statistical analysis

Differences among groups were determined using analy­

sis of variance and the least significant difference test

for a priori pairwise comparisons (LSD) analysis. Due to

71

the much greater response to GnRH than to the steroid

hormones in the experiment comparing these responses, the

mean sum of squares used in the LSD analysis was derived

using only values from groups being compared (Snedecor and

Cochran, 1980). Experiments with only two groups were

analyzed using the unpaired Student's t-test. Data are

expressed as means ± SE.

RESULTS

The effects of mGnRHa injections on serum concentrations

of PRL

Intraperitoneal injections of 0.1 ~g mGnRHa/g body

weight elevated serum tPRL1 88 and tPRL1 77 concentrations

of mature male and female tilapia at 1 hr after injection,

over concentrations observed in vehicle-injected and non­

injected controls (Fig. 5; P < 0.01). Serum hormone

concentrations were not significantly different from

controls at 11 hr after injection. Injections did not

alter GH serum concentrations.

The effects of mGnRHa on the in vitro release of PRL from

tilapia RPD incubated in isosmotic medium

To determine the response of PRL cells to GnRH, RPD

were incubated for 18-20 hr in isosmotic medium

72

Figure 5. Effects of intraperitoneal injections of0.1 ~g/ g body weight of mGnRHa on serum PRL levels inmale (M) and female (F) O. mossambicus at 1 and 11 hrspost-injection (non-injected, black bars; vehicleinjected, open bars; mGnRHa injected, diagonal bars)(mean ± SE, N = 10-12 per treatment from 2 replicateexperiments, ** P < 0.01 and *** P < 0.001, comparisonsare to injected controls) .

73

1 hr

40tPRL188

- *- *E 35 *- *C')

..:. 30 *c0 25+:if!

20 11 hr-c Non-(1)o 15 injectedc0o 10E:::s 5~

CDen0

F M F M F M

tPRL177

M

11 hr

FM

1 hr

***

FMF

Non­injected

_18-.€ 16C')

":'14

5 12+:ie 10-; 8o5 6(,) 4E2 2(1)

en 0

74

(320 mOsmolal) containing graded concentrations of mGnRHa

ranging from 0.01 nM to 1 ~M. Tissues were incubated in

isosmotic medium (320 mOsmolal) which elicits moderate

baseline PRL release so that either a stimulatory or

inhibitory effect of GnRH on PRL release could be ob­

served. Incubation with mGnRHa stimulated the release of

tPRL177 and tPRL188 with the greatest response elicited at

a concentration of 10 nM mGnRHa (Fig. 6; P < 0.01).

Release is variable with isosmotic medium. For this

reason, data from each of two :replicate experiments were

normalized as a percent of controls before being combined.

The effects of medium osmolality on mGnRHa stimulated

release of PRL from tilapia RPD

To determine whether GnRH could overcome the effects

of medium osmolality on PRL release, RPD were incubated

for 18-20 hr in hyposmotic medium (300 mOsmolal), isosmot­

ic medium (320 mOsmolal), or hyperosmotic medium (355

mOsmolal), with graded concentrations of mGnRHa. The

release of both PRLs showed a typically inverse correla­

tion with medium osmolality in the absence of GnRH (Fig.

7). Release of both PRLs was stimulated by mGnRHa at a

concentration of 100 nM when incubated in isosmotic and

hyperosmotic medium (P < 0.01 and 0.001 respectively)

There was no significant difference among treatments

75

Figure 6. Prolactin release from RPD of o. mossambicusduring 18-20 hr incubations in isosmotic medium(320 mOsmolal), over a range of mGnRHa concentrations.

Due to variability of baseline release in isosmoticmedium, the data are expressed as a percent change fromthe control values for each replicate of the experiment(mean ± SE, N = 10-12 per treatment from 2 replicateexperiments, ** P < 0.01).

76

200

160(I)tilCU(I)-e..I 1200::C.e.-(I)

en 80cns

..c:o?ft.

40

D tPRL188

D tPRL177***

*

o 0.01 0.1 1 10 100 1000

log mGnRHa (nM)

77

Figure 7. Prolactin release from RPD of O.mossambicusduring 18-20 hr incubations with mGnRHa, in hyposmoticmedium (300 mOsmolal), isosmotic medium (320 mOsmolal) andhyperosmotic medium (355 mOsmolal; mean ± SE, N : 10-12per treatment from 2 replicate experiments, ** P < 0.01and *** P < 0.001) .

78

***355320300

-

- + **rt- f+-

~-

r--

+- +-

II I I I I I I I I I I

60

o

~ 50ns(1)- 40e~30"t"'"

....Ia:: 20D....,';f!. 10

o 10 100 o 10 100 o 10 100

60300 320 355

***~ 50eu(1)

Q) 40a..

~30~

....IIX: 20a..+of

~ 10o

**

o 10 100 o 10 100 o 10 100

mGnRHa (nM)

79

~~ ... _- -~-_._--------

exhibiting the greatest stimulation for each osmolality,

possibly because stimulation was near maximal under each

condition. Better separation of treatment effects was

observable with hyperosmotic medium due to a lower base­

line release. For this reason, subsequent studies were

conducted in hyperosmotic medium.

The relative potency of various GnRH molecules on the

release of PRL from tilapia RPD

The potency of the GnRH forms were compared using

18-20 hr static incubations of RPD in hyperosmotic medium.

These incubations differ from the previous incubations in

that the tissues were preincubated in hyperosmotic control

medium overnight and the medium was changed and tissues

rinsed 3 times before the treatment medium was added.

This long preincubation reduced baseline release and

provided a better separation of responses. The response

of tPRLs were similar to each other in both studies. In

the first set of experiments (Fig. 8), cGnRH-I, a form of

GnRH which was not found in the tilapia, was the least

potent form of GnRH tested. Chicken-GnRH-I did not elic­

iting a significant increase in the release of either PRL

until a concentration of 1000 nM. Both native forms of

GnRH, sGnRH and cGnRH-II were more potent than cGnRH-I. A

significant response with sGnRH was observed with 1 nM

80

Figure 8. Prolactin release from RPD of O. mossambicusduring 18-20 hr incubations in hyperosmotic medium(355 mOsmolal), over a range of cGnRH-I, cGnRH-II andsGnRH concentrations (mean ± SE, N = 10-12 per treatmentfrom 2 replicate experiments, * P < 0.05 and*** P < 0.001).

81

82

sGnRH. Chicken GnRH-II was found to be the most potent of

the three forms tested. A significant increase in the

release of both tPRLs was observed with the lowest concen­

tration tested, 0.01 nm cGnRH-II. All three forms evoked

the maximum response observed. When three forms of GnRH

that are native to the tilapia were cOTI~ared (Fig. 9),

cGnRH-II was again found to be the most potent form of the

native GnRHs, followed by sGnRH. Seabream GnRH was the

least potent for stimulating the release of both tPRLs.

Effects of sbGnRH on PRL release from tilapia RPD in

perifusion incubation

Single short exposures (5 min pulse) of RPD to sbGnRH

stimulated the release of both PRLs in a dose-related

manner at concentrations similar to that observed with the

18-20 hr static incubations (Fig. 10). A stimulation was

first observed at a concentration of 10 nM sbGnRH

(P < 0.05).

The effects of cGnRH-II on [ca2+J i and PRL release

Chicken GnRH-II and hyposmotic medium evoked in­

creases in [ca2+]i and PRL release (Figs. 11, 12, 13).

Chicken GnRH-II was effective at a concentration as low as

0.1 pM. An increase in [ca2+]i was observed within 1 min

of exposure to cGnRH-II or hyposmotic medium. A reduction

83

Figure 9. Prolactin release from RPD of o. mossambicusduring 18-20 hr incubations in hyperosmotic medium(355 mOsmolal), over a range of cGnRH-II, sGnRH, andsbGnRH concentrations (mean ± SE, N = 6 per treatment,* P < 0.05, ** P < 0.01 and *** P < 0.001).

84

85

Figure 10. Prolactin release from RPD of o. mossambicusduring perifusion incubation in hyperosmotic medium(355 mOsmolal), in response to a range of sbGnRHconcentrations delivered in 5 min pulses. Fractions werecollected at 10 min intervals and the data are presentedas the percent of total hormone released into the mediumin the 3 fractions following the introduction of thesbGnRH, compared to the total hormone in the 3 fractionspreceding the introduction of the sbGnRH, at eachconcentration (mean ± SE, N = 10 RPD per column and 6columns per GnRH form, * P < 0.05 and ** P < 0.001).

86

220-e- tPRL188

200 --II- tPRL177

-- **eu 180 ***tneu..c **'eft. 160-(I)tneuCI) 140-G)~

..Ia:: 120a..

100

80 -l.--,---.,.----r----r---,-------r-

o 0.1 1 10 100 1000

Log sbGnRH (nM)

87

Figure 11. Changes in [Ca++]i from dispersed PRL cells ofO. mossambicus during perifusion incubation inhyperosmotic medium (355 mOsmolal) I in response to a rangeof cGnRH-II concentrations and a switch to hyposmoticmedium (300 mOsmolal). The [ca++]i is expressed as therelative intensity of the ratio of fura-2 fluorescenceexcited at 340 nm (fura-2 bound to Ca++) to that excitedby 380 nm (free fura-2) .

88

89

\Do

Figure 12. Prolactin release from 4 RPD and changes in [Ca++]ifrom dispersed PRL cells of O.mossambicus during peri fusionincubation in hyperosmotic medium (355 mOsmolal), in response toa range of decreasing cGnRH-II concentrations and a switch tohyposmotic medium (300 mOsmolal). Fractions were collected at 1min and the PRL data are presented as ng/ml of medium. The[Ca++]i is expressed as the relative intensity of the ratio offura-2 fluorescence excited at 340 nm (fura-2 bound to Ca++) to~hat excited by 380 nm (free fura-2).

:::- 180 ---tPRL188.E 160Cl -tPRL177..s 140 Control5 120 GnRH GnRH 300

~ 100 GnRH 0.01 nM 0.0001 nM Control mQsm... 355 mQsm.... 1.0 nM 355 rnOsm 355c 80CD 355 mOsmuc 600e 40E 20::s...(I) 0en

0 10 20 30 40 50 60 70\D Fractions (1 min)I-'

5.5~GnRH

_ 5 0.01 nM

g 4.5 ~ 355 mQsm

~ 40~M 3.5

~ 3~~GnRH i GnRH "~r" I Control~ 2.5 i 1.0nm 0.0001 nM

~ 2 355mQsm 355 mQsm300mQsm

1.5 ..... - ... - ..... - - - .. - ............

0 10 20 30 40 50 60 70Minutes

\DN

Figure 13. Prolactin release from 4 RPD and changes in [Ca++]ifrom dispersed PRL cells of o. mossambicus during peri fusionincubation in hyperosmotic medium (355 mOsmolal) I in response toa range of increasing cGnRH-II concentrations alternating withhyperosmotic medium without cGnRH-II and a switch to hyposmoticmedium (300 mOsmolal). Fractions were collected at 1 min and thePRL data are presented as ng/ml of medium. The [Ca++]iis expressed as the relative intensity of the ratio of fura-2fluorescence excited at 340 nm (fura-2 bound to .Ca++) to thatexcited by 380 nm (free fura-2) .

_ 600

15001 --..-- tPRL188n -- tPRL177S 400+:ica GnRH Controlb 300 GnRHr:: GnRH Control 1.0nm 300(1) 0.01 nM Controlg 200 0.0001 nM 355 355 355 3550 355 \ rnOJm mOsm rnOsm rnOsmuE 100 rnOsm ¢~ ---~- ~a.. - ----(1) 0CIJ

0 10 20 30 40 50 60 70~

w Fractions (1 min)4.

GnRHGnRH-g 3.5 0.0001 nM 1.0nmM 355 355-0 3 rnOsm rnOsm ~~M-;2.5 Control+:i 300a:I 2c:= mOsm rnOsm

1.5 ........ - ... -

0 10 20 30 40 50 60 70Minutes

in [ca2+]i was observed within 1 min after exposure to

control medium (hyperosmotic medium without cGnRH-II, Fig.

13). Prolactin release in response to both GnRH and

hyposmotic medium appeared to be biphasic. An initial,

large response was followed by lower levels that were

still above baseline, even though [ca2+]i remained high

(Figs. 12 and 13). A dose response to cGnRH-II was not

observed for [ca2+]i. Repeated exposures to cGnRH-II with

10 min breaks between exposures evoked increases in

[ca2+]i that were similar in magnitude to the first re­

sponse but PRL release did not appear to be restimulated

(Fig 13). Exposure to hyposmotic medium was able to evoke

an increase in PRL release following a failure in repeated

GnRH exposures even though both elicited similar responses

in [Ca2+] i (Fig. 13).

The effects of T and E2 on sGnRH induced PRL release

Treatment with sGnRH (100 nm), T (10 nM), and E2

(10 nM) stimulated the release of both PRLs from RPD

during 3 hr incubations in hyperosmotic medium, following

an overnight preincubation in control medium (Fig. 14; all

P < 0.05). Combined treatments resulted in greater re­

lease of both PRLs than either sGnRH or steroid treatments

alone (P < 0.05 for comparisons for each individual treat-

94

Figure 14. Prolactin release from RPD of O. mossambicusduring 3 hr incubations in hyperosmotic medium(355 mOsmolal), with 10 nM E2, 10 nM T, 100 nM sGnRH or acombination of the sGnRH and the steroid hormones(mean ± SE, N = 10-12 per treatment from 2 replicateexperiments, * P < 0.05, ** P < 0.01 and *** P < 0.001).

95

*

sGnRH sGnRH sG nR H&E2 &T

***

***

**

Tc

6

7 D tPRl188

[] tPRL177

1

96

ment compared with treatments in which they were used in

combination with one of the other hormones) .

DISCUSSION

Effects of GnRH injections on PRL release

Injections of the tilapia with the GnRH analog,

mGnRHa, elevated serum concentrations of the tPRLs. These

data provide evidence that GnRH either stimulates the PRL

cells to release PRL directly, or stimulates less direct

pathways that lead to PRL release from PRL cells. The

injections almost certainly lead to an increase in GtH

release which in turn leads to an increase in the release

of steroids from the gonads. I present data that confirm

earlier findings that T and E2 stimulate PRL release in

the tilapia (Barry and Grau, 1986; Borski et al., 1991).

Taken together, the rise in circulating concentrations of

PRL in the tilapia in response to GnRH injections, could

have been mediated via GtH and then steroid hormones.

Ekengren and colleagues (1978) suggested this pathway to

explain how GnRH injections increased PRL cell activity in

the Atlantic salmon.

The rise in circulating PRL levels within 1 hr of a

single injection of mGnRHa (0.1 ~g/g body weight) occurred

within a time frame consistent with a GtH and steroid

97

-_.._-------- . -_._--_._-----

mediated response. Intraperitoneal injections of

catfish-GnRH (0.125 ~g/g body weight) elevated plasma

concentrations of GtH-II and 11-ketotestosterone in

African catfish, (Clarias gariepinus), within 1 hr post­

injection, and cGnRH-II injections (dosage as low as 0.002

~g/g body weight) elevated plasma GtH-II concentrations

within 0.5 hr (Schulz et al., 1993). Each of these time

points represent the earliest time points examined. Both

forms of GnRH are found in the pituitary of the African

catfish with catfish-GnRH being 37-fold more abundant than

cGnRH-II.

The rapid elevation in steroid concentrations in

response to GnRH injections makes in vivo studies of the

effects of GnRH on PRL release difficult to interpret in

intact sexually mature fish. A stimulation of PRL levels

in gonadectomized fish in response to GnRH injections may

be a better indication that GnRH directly stimulates PRL

release, however, the GnRH response may be mediated via

other factors. In the rat, GnRH has been shown to stimu­

late PRL release via paracrine action involving GtH cells

(Denef and Andries, 1983). For these reasons, in vitro

examinations of the effects of GnRH on PRL cells was

emphasized in my study.

The rise in serum PRL concentrations induced by GnRH

injections demonstrates that increases in circulating GnRH

98

can affect PRL levels, regardless of the mechanism. Thus,

the interpretation of in vivo treatments with GnRHs,

intended to elevate GtH or GH, should consider the effects

of changes in PRL levels on the observed responses to the

treatments.

In vitro evidence that GnRH functions as a PRL-releasing

factor

All three forms of GnRH that are native to the tila­

pia stimulated the release of tPRL1 88 and tPRL1 7 7 from RPD

in vitro. Concentrations of GnRH found to stimulate PRL

release were well within the range for receptor-mediated

action. Furthermore, the range of concentrations of GnRH

capable of stimulating PRL release was similar to those

shown to stimulate GtH and GH release from pituitary

fragments of teleosts including tilapia (Levavi-Sivan and

Yaron, 1989; Peter et al, 1990; Melamed et al., 1995).

These data, and the fact that the RPD is an almost homoge­

neous mass of PRL cells, suggest that GnRH is a direct

regulator of PRL cell activity in the tilapia, o. mossam­

bicus.

Consistent with the notion that GnRH's effects on PRL

release are direct and receptor mediated, treatment with

cGnRH-II evoked an increase in [ca2+]i within 1 min of

exposure and an elevation in PRL release within 2 min.

99

These data suggest that calcium is operating as a second

messenger in mediating the effects of GnRH on PRL release.

The pathway of GnRH action on PRL cells of the tila­

pia clearly differs from that of the rat. In the rat,

GnRH has been shown to stimulate PRL release via paracrine

action involving GtH cells (Denef and Andries, 1983). The

RPD of the tilapia does not contain either GtH cells or GH

cells. Growth hormone cell have also been to respond to

GnRH in teleosts, including the tilapia (Marchant et ai.,

1989; Melamed et ai., 1995).

The pathway of GnRH action on PRL cells of the

tilapia may also differ from that of the goldfish. Cook

and colleagues found negligible binding of a sGnRH analog

to prolactin cells of the goldfish compared with binding

to GtH and GH cells. Either the PRL cells of the goldfish

do not respond to GnRH or the receptors are only expressed

during certain periods, for example, discrete stages of

development or reproduction. Prolactin is a FW-adapting

hormone in teleosts and may always be expressed in euryha­

line teleosts such as the tilapia (Grau et ai., 1994).

Alternatively, the direct regulation of PRL cells by GnRH

may be a late development in the evolution of teleosts.

Direct regulation of PRL cells by GnRH may exist in highly

derived teleosts as the tilapia but not in lower teleosts

as the goldfish or in mammals. An investigation of GnRH

100

receptors on PRL cells of the tilapia needs to be conduct­

ed as well as studies on the effects of GnRH on PRL

release in other fishes, including the goldfish.

Potencies of the 3 native forms of GnRH

The tilapia and seabream have the same 3 native forms

of GnRH (Powell et al., 1994; Weber et al., 1994).

Interestingly, their order of potency are the same for in

vivo stimulation of GtH-II release in seabream (Zohar et

al., 1995) and in vitro stimulation of PRL release in

tilapia. Seabream GnRH is by far the most abundant form

in the pituitary of the tilapia (N. Sherwood, personal

communication) and seabream (Powell, et al., 1994).

Seabream-GnRH is also the least potent in stimulating PRL

or GtH release, and cGnRH-II is the most potent in both

circumstances. The cause for differences in potency in

stimulating PRL release is at the pituitary level. Dif­

ferences in peripheral degradation by specific peptidases

at the liver or kidney, as listed among the possible

factors contributing to the same order of potency in

seabream, can not be a factor in my in vitro studies.

Reasons for the differing potencies of the various

GnRH molecules include differences in receptor affinities,

differences in intracellular degradation rates of the GnRH

molecules within the pituitary, and activation of differ-

101

ent signal transduction cascades. A single type of GnRH

receptor has been found in the pituitary of all fish

examined to date with the exception of the goldfish. This

suggests that the different GnRH forms are competing for

the same receptor. Fish found to have a single type of

GnRH receptor in the pituitary include the African catfish

(Leeuw et ai., 1988), winter flounder (Crim et ai., 1988),

stickleback (Andersson et ai., 1989), and seabream (Pagel­

son and Zohar, 1992). More than one type of receptor has

been identified in the goldfish (Habibi et ai., 1987).

The native GnRH peptide with the greater GtH-II­

releasing activity has been found to have greater receptor

affinity in African catfish (cGnRH-II > catfish-GnRH;

Schulz et ai. 1993) and goldfish (cGnRH-II > sGnRH; Peter

et ai. 1990; Habibi, 1991). Nevertheless, in the same

study in the goldfish, sGnRH was more active in releasing

GH.

Gonadotropin-releasing hormone is degraded within the

pituitary cells of teleosts (Goren et ai., 1990; Zohar et

ai., 1990). The pituitary cells posses cytosolic pepti­

dases that degrade GnRH after it has been internalized.

Differences in biological activities of various GnRH

molecules have been shown to be attributable to their

differing susceptibility to degradation by these

peptidases (Goren et ai., 1990; Zohar et ai., 1990).

102

Chang and colleagues (1993) have demonstrated that

different native GnRH forms activate different signal

transduction pathways in GtH cells of the goldfish. They

suggest that binding of the two GnRH molecules to the same

receptor leads to distinct receptor-conformation changes

and different G-protein linkages. These different path­

ways may result in different potencies for the different

GnRH molecules.

The order of potency appears to parallel the

evolution of the GnRH peptides. The cGnRH-II peptide is

present in cartilaginous fishes, sGnRH first appears in

early teleosts and sbGnRH in late teleosts (cf. Sherwood

et al. 1994). The presence of a single GnRH receptor in

most teleosts suggests the peptides may have evolved

independently of the receptor or receptors and therefore

the older peptides may have a greater affinity to the

receptor or receptors. On the other hand, if the pitui­

tary cells which respond to GnRH have evolved specific

peptidases to break down the biologically relevant GnRH

form, then the biologically relevant GnRH peptide would

have the shorter half life and lower biological activity

(Goren et al., 1990; Zohar et al., 1990). Either

possibility is consistent with sbGnRH being the biologi­

cally relevant regulator of GtHs, GH and PRLs in the

pituitary. Furthermore, since sbGnRH is delivered by

103

neurosecretory fibers directly to the cells of the

pituitary, sbGnRH can be delivered in high concentrations,

and thus, a greater potency may not be necessary for the

hormone to have its effect.

Roles for the different forms of GnRH

The dominance of sbGnRH in the pituitaries suggests

that sbGnRH may be the GnRH form which is the regulator of

GtHs, GH and PRLs. Nevertheless, examinations of pitui­

tary and circulating levels of native GnRHs under many

varying physiological and developmental conditions are

needed before regulation of pituitary cells by other GnRH

forms can be dismissed. Schulz and colleagues (1993)

found catfish-GnRH to be in greater abundance than cGnRH­

II in the pituitaries of the African catfish, Clarias

gariepinus, with cGnRH-II being the more potent stimulator

of GtH-II release in vivo and in vitro. The researchers

present the possibility that catfish-GnRH may be the

regulator of moderate levels of circulating GtH and the

more potent cGnRH-II may be the regulator of GtH surges,

such as those associated with spawning.

Injections and implants of native forms of GnRH and

analogs have been shown to stimulate GtH and GH release.

Thus, pituitary tissues can respond to GnRH in the blood.

Whether circulating GnRH has biological significance in

104

most teleosts has yet to be determined. Gonadotropin­

releasing hormone and a GnRH binding protein have been

identified in the circulation of the goldfish. Further­

more, the concentrations of GnRH in the circulation were

within a range found to stimulate GtH and GH release in

vitro in the same species (Huang and Peter, 1988; Peter et

al., 1990). In preliminary studies GnRH was not detected

in the serum of tilapia (N. Sherwood, personal communica­

tion). Nonetheless, my in vitro studies suggest that

cGnRH-II, which is endogenous in the tilapia (Weber et

al., 1994), may be able to stimulate PRL release at

concentrations which are below the detectable limits of

the radioimmunoassay used to measure the GnRH.

Effects of T and E2 on GnRH stimulated PRL release

The steroid hormones, E2 and T, are able to

enhance the response of PRL cells to GnRH, increasing

percent release over 3-fold. The steroids may regulate

the sensitivity of the PRL cells to GnRH stimulation

during the reproductive cycle. Alternatively, GnRH

molecules may regulate the sensitivity of PRL cells to T

and E2 stimulation. The potentiation of the response of

PRL cells by GnRH and thyrotropin-releasing hormone by sex

steroids supports the suggestion by Barry and Grau (1986)

105

- -------------------'----

that there may be a shift in the control of PRL secretion

with changes in the reproductive state of the tilapia.

Hypothalamic stimulators of PRL cells in the tilapia

Prolactin secretion in teleosts is thought to be

predominantly under inhibitory control, however, evidence

had suggested that a hypothalamic prolactin-releasing

factor was involved in the control of PRL secretion in

fish (see reviews, Clarke and Bern, 1980; Ball, 1981).

Barry and Grau (1986) have shown that thyrotropin­

releasing hormone stimulates PRL release in the tilapia

but only following E2 preincubation. Until the present

study, thyrotropin-releasing hormone was the only hypotha­

lamic factor shown to have the capability to stimulate PRL

release in the tilapia.

Summary

Gonadotropin-releasing hormone appears to function as

a PRL-releasing factor in the tilapia, o. mossambicus, and

this effect appears to be direct. All endogenous GnRH

molecules stimulate the release of the two tPRLs, tPRL1 77

and tPRL1 88, from RPD with the following order of potency:

cGnRH-II > sGnRH > sbGnRH. Furthermore, the data suggest

that calcium is operating as a second messenger in mediat-

106

ing the effects of GnRH on PRL release and the sex ster­

oids, °E2 and T, potentiate the response of PRL cells to

GnRH.

107

Chapter IV

CHANGES IN SERUM AND PITUITARY LEVELS OF PROLACTIN AND

GROWTH HORMONE WITH FASTING IN THE TILAPIA,

OREOCHROMIS MOSSAMBICUS

INTRODUCTION

Prolactin and GH have effects on protein, carbohy­

drate, and lipid metabolism in mammals (see reviews,

Nicoll, 1974; Fain, 1980; Davidson, 1987). The +oles of

PRL and GH in metabolism in teleosts are not well charac­

terized. Prolactin has been shown to be either lipolytic

or lipogenic depending on many factors including tempera­

ture, photoperiod, circadian rhythms, development and

circulating levels of other hormones (review see, Baker

and Wigham, 1979; Sheridan, 1986). Interestingly, plasma

PRL concentrations are not altered by 30 days of fasting

in Kokanee salmon, (Oncorhynchus nerka; McKeown et al.,

1975). This is the only study to measure circulating PRL

levels during fasting in a teleost. Nevertheless, Rodgers

and colleagues (1992) found both pituitary concentrations

of PRL and basal PRL release from RPD in vitro to be

decreased after 2 weeks of fasting in o. mossambicus. In

this same study, PRL release in vitro was negatively

108

correlated with essential amino acids concentration but

was unaffected by D-glucose concentration in the medium.

Although actions of PRL in metabolism appear to be con­

served, evidence for a change in secretion rate with

fasting is limited.

The roles of GH in metabolism and fasting in teleosts

have received more attention than those of PRL, but are

not without controversy. Growth hormone treatment has

been shown to increase circulating levels of free fatty

acids (FFAs) in the rainbow trout and goldfish and to

mobilize lipid reserves from the liver of rainbow trout

and coho salmon (Minick and Chavin, 1970; Leatherland and

Nuti, 1981; Sheridan, 1986). In addition, GH injections

increase liver glycogen and muscle FFAs in Kokanee salmon,

but not plasma FFAs or glucose (McKeown et al., 1975).

Growth hormone injections do not appear to stimulate

lipolysis in the eel, Anguilla japonica, and tilapia, o.

mossambicus, but instead appear to stimulate the mobiliza­

tion of amino acids and glucose. Injections of bovine-GH

in hypophysectomized eels increase serum amino acid con­

centrations and do not alter plasma lipid concentrations

(Inui et al., 1985). Injections of bovine-GH stimulate

the release of glycogen and not lipid from liver reserves

and increases serum concentrations of amino acids and

glucose but not protein, lipid or cholesterol in the

109

tilapia (Leung et ai., 1991). Consistent with this

elevation in circulating amino acids and glucose, Rodgers

and colleagues (1992) found an increase in GH release from

the proximal pars distalis of tilapia in response to

reductions in essential amino acids and glucose in the

medium. Together these studies suggest that GH cells in

the tilapia act in the regulation of, and are directly

responsive to, changes in the blood concentrations of

amino acids and glucose. Thus, GH would be expected to

increase with fasting in the tilapia, as in mammals.

Increases in circulating and pituitary levels of GH

have been observed in fasted teleosts. Increases in

plasma GH concentrations in rainbow trout have been ob­

served within 3 weeks of fasting in studies by Wagner and

McKeown (1986), and within 1 week in studies by Sumpter et

ai. (1991a). No change in plasma GH concentrations was

observed after 30 days of fasting in Kokanee salmon

(McKeown et ai., 1975). Farbridge and Leatherland (1992)

observed a bi-modal response to fasting in rainbow trout,

first exhibiting a decline in plasma GH concentrations at

2 weeks compared with fed controls, followed by an eleva­

tion at 4 weeks. Plasma GH concentrations in the fed

controls may not have been different from initial values

and plasma GH levels were elevated at 2 weeks of fasting

in a repeat of the study. Melamed (1993) observed an

110

increase in plasma GH concentrations within 16 days of

fasting in "sub-optimally" fed tilapia hybrids (0. niloti­

cus X O. aureus). Also in tilapia, o. mossambicus,

pituitary content of GH was elevated with 2 weeks of

fasting (Rodgers et al., 1992).

Farbridge et al. (1992) found that feeding rainbow

trout only once every five days resulted in reduced plasma

GH and suggested that GH concentrations are depressed by

low feeding rates. This is not consistent with the study

on the tilapia hybrid, by Melamed (1993). The tilapia at

the start of the experiment by Melamed, were described as

"sub-optimally fed". When the fish were switched to an

increased feeding rate, plasma GH concentrations fell

within 8 days. Clearly, the nutritional state of the

animal prior to fasting and changes in nutritional state

during fasting affects the hormonal response to subsequent

fasting.

In order to learn more about the roles of PRL and GH

in tilapia metabolism, I have characterized changes in

serum concentrations and pituitary content of the tPRLs

and GH with 21 and 31 days of fasting in well-fed tilapia

with a high condition factor (CF; fat tilapia) and tilapia

with a low CF (thin tilapia). The fasting should lead to

an alteration in metabolic state requiring the mobiliza­

tion of energy stores. If the tPRLs or GH are involved in

111

the mobilization of these energy stores, then their serum

concentrations or content in the pituitary may be altered.

Changes in body weight (BW) , standard length (SL) , gonad

weight, and liver weight were recorded to gauge the

effects of fasting on the metabolic state of the animals.

MATERIALS AND METHODS

Animals

Tilapia, o. mossambicus, were collected from brackish

water streams and maintained in FW in outdoor tanks (22­

25°C), for at least 3 months before use. The fish were

fed to satiation twice daily with Purina trout chow. Fish

were not fed on the day of sampling.

Fish were transferred to an indoor facility with a

fixed photoperiod of 14L:10D. Fish were placed in oval

fiberglass tanks (60 1). The tanks were outfitted with a

continuous water flow (-23°C) and constant aeration. The

fish ranged in size from about 60-100 g. Fish were al­

lowed to acclimate to the tanks for at least 2 weeks

before the start of the experiments.

Fish were anesthetized with 2-phenoxyethanol (1 mIll)

for blood collection. Blood was drawn from the hemal arch

within 5 min of introduction to the anesthesia, using a

28-gauge needle attached to a 1-ml disposable syringe.

112

Blood was stored on ice and the serum was separated from

the formed elements by centrifugation and then stored at

-Boec until it was analyzed for PRLs and GH by radioimmu­

noassay. The radioimmunoassays are detailed in chapter

II. Anesthetization with 2-phenoxyethanol did not affect

serum PRL and GH levels (Fig. 15).

Additional measurements and collections of tissues

varied with experiments. Body weight, SL, gonad weight

and liver weight were recorded and pituitaries were col­

lected and placed in 20 ~l of distilled water in 1.5-ml

microcentrifuge tubes and stored at -Boec until they were

analyzed for PRLs and GH by radioimmunoassay. Pituitaries

were sonicated in phosphate buffered saline and diluted in

phosphate buffer with 1% bovine serum albumin (Sigma).

Protein content was determined in the undiluted samples

using a Bio-Rad (Richman CA) protein assay kit. Data were

expressed as micrograms of hormone per pituitary (~g/pit.)

as opposed to micrograms of hormone per milligram of

protein for 2 reasons. First, the fish in each study were

similar in size and therefore should have similar size

pituitaries. Secondly, pituitary proteins; PRLs, GH, and

GtHs would be expected to change dramatically under the

conditions of the studies, possibly affecting pituitary

protein content. For example, PRL content will appear to

decrease if there is a large increase in pituitary GH

113

Figure 15. Effects of 2-phenoxyethanol (1 mIll) onprolactin and growth hormone serum concentrations in thetilapia, Oreochromis mossambicus (mean ± SE,N 10-15 per treatment, ns = not significant atP < 0.05).

114

15

14

13

12

:=- 11E......C) 10c:---c: 9o.-...E 8...; 7CJ

§ 6Co)

E 5:s...Q) 4en

3

2

1

ns D control

EJ 2-phenoxyethanol

tPRL188 tPRL177 GH

115

content and PRL is normalized to pituitary protein.

Gonadosomatic index (GS1; ((gonad weight (g)/ (g)) X 100)

hepatosomatic index (HS1; ((liver weight (g)/BW (g)) X

100) and condition factor (CF; ((BW (g)/(SL (cm))3) X 100)

were calculated. When sampling required killing the

animals, the fish were quickly decapitated following

anesthetization.

Statistical analysis

Differences among groups were determined using analy­

sis of variance and the least significant difference test

for a priori pairwise comparisons analysis (Steel and

Torrie, 1980). Comparisons between two means were per­

formed using the unpaired Student's t-test. Data are

expressed as means ± SE.

Study 1

Two fasting studies were conducted. The first study

examined the effects of 10 and 21 days of fasting on

female and 21 days of fasting on male tilapia with a high

condition factor (-3.6). The study was replicated in two

trials. Body weights, SL and gonad weight measurements

were recorded for the second trial of the study. In both

trials of the study fish were placed 5 to a tank and there

were two tanks per treatment (10 treatments). Fasting

116

started on a staggered basis so that all fish were sampled

on the same day between 1300 and 1600. Fish were fed to

satiation twice daily with Purina trout chow when not

fasting.

Study 2

The second study examined the effects of 10, 21, and

31 days of fasting on male tilapia with a low condition

factor (-3.2). Study 2 was conducted following the same

protocol as Study 1 except 6 fish were placed in each tank

and were fed once daily when not fasted as part of the

study. The fish used in the second study were fed a

limited ration once daily for 2 months prior to the start

of the study. Body weights, SL, gonad weight and liver

weight measurements were recorded for the study.

RESULTS

The effects of fasting on serum concentrations and

pituitary content of the PRLs and GH in male and female

tilapia with a high condition factor

Serum levels of tPRL1 7 7 increased from 1.8 ± 0.3

ng/ml in fed female tilapia, to 6.2 ± 1.3 ng/ml with 10

days of fasting (P < 0.001) and 5.3 ± 0.5 ng/ml with 21

days of fasting (Fig. 16; P < 0.01). A similar trend

117

Figure 16. Changes in serum concentrations of prolactinand growth hormone during fasting in male (M) and female(F) tilapia Oreochromis mossambicus with a high conditionfactor (mean ± SE, N = 10-12 per treatment; * P < 0.05,** P < 0.01, *** P < 0.001 ).

118

tPRL188

*

..-.-E.......C) 10c---co;:; 8e....,s::(1)(J 6co(J

E 4::s'-(I,)

tn

o 1021 0 21

F Mo 10 21 0 21

F M

Days Fasted

o 10 21 0 21

F M

-

-

- tPRl177

**- ***

- t*. GH

-

-~ rf &r.f. rfl

li rJI I I I I I I I I I I I

14

o

12

2

119

was not significant for tPRL1 88 (ANOVA P = 0.059). Serum

tPRL1 77 and tPRL1 8 8 concentrations were elevated in male

tilapia fasted for 21 days. Prolactin177 concentrations

increased from 1.7 ± 0.2 to 6.7 ± 1.3 ng/ml with 21 days

of fasting (P < 0.01) and tPRL1 88 serum concentrations

increased from 7.5 ± 0.8 to 12.6 ± 2.1 ng/ml (P < 0.05).

Serum GH concentrations were not affected by 21 days of

fasting in either sex.

Pituitary content of tPRL177 was elevated in female

tilapia fasted 21 days compared with females fasted 10

days or fed (Fig. 17; P < 0.01). Fasting did not affect

pituitary tPRL1 77 content in male tilapia or tPRL1 88

content in male or female tilapia. Pituitary content of

GH increased from 3.9 ± 0.6 ~g/pit. in fed female tilapia

to 6.1 ± 0.6 ~g/pit. in female tilapia fasted for 21 days,

and increased from 3.9 ± 0.2 ~g/pit. in fed male tilapia

to 5.4 ± 0.4 ~g/pit. in male tilapia fasted for 21 days

(P < 0.001 and P < 0.01 respectively) .

The effects of fasting on condition factor, body weight,

standard length and GSI in male and female tilapia with a

high condition factor

At the start of the study, CFs were similar for male

and female tilapia, ranging from 3.58 to 3.77 (Table 1).

120

Figure 17. Changes in pituitary content of prolactin andgrowth hormone during fasting in male (M) and female (F)tilapia Oreochromis mossambicus with a high conditionfactor (mean ± SE, N = 10-12 per treatment; ** P < 0.01,*** P < 0.001).

121

tPRL188

***

F M

o 1021 0 10o 1021 0 10

F M

Days Fasted

F M

o 1021 0 10

-

-

-

rfGH

-tPRL177

d **-

~a* ~ai~

- r!l

rIlrf-

I I II I I 1 -II I -Io

14

12

.-.......'0. 10--..en~-..... 8c

CI)......C0(J 6~C\1.....-::s 4.....-e,

2

122

Table 1. Effects of fasting up to 21 days on condition factor,body weight (g) and standard length (cm) in male and femaletilapia with a high condition factor (CF > 3.6; mean ~ SE).

FemalesDays BW-B BW-E %CH t- SL-B SL-E %CH t- CF-B CF-E %CH t-T

f-'NW

fed

10

21

ANOVA

Malesfed

21

t-T

81. 66.1

79.62.5

74.95.4ns

73.23.5

79.94.4ns

80.75.6

75.12.5

69.04.7ns

78.63.0

74.24.1ns

-1. 2 ns 12.920.35

-5.6 ns 12.820.18

- 7 . 9 ns 12. 600.31

ns

7.3 ns 12.590.22

-7.0 ns 13.030.26

ns

13.170.35

12.880.17

12.600.32

ns

13.070.20

13.060.25

ns

1. 90 ns

0.47 ns

-0.07 ns

3.76 ns

0.20 ns

3.720.053.770.063.690.06

ns

3.650.043.580.08

ns

3.490.053.510.063.410.06

ns

3.520.063.310.07

*

-6.3 **

-7.0 **

-7.4 **

-3.6 ns

-7.7 *

(B, beginning; E, end; BW, body weight; SL, standard length; %CH, % change;t-T, t-Test comparisons; * P < 0.05, ** P < 0.01; ns, not significant)

Condition factors were reduced by 21 days of fasting in

males compared with values for the same animals at the

start of the study (P < 0.05), and with fed males at the

end of the study (P < 0.05). The CFs for females in all

treatments were lower at the end of the study compared

with initial values (all P < 0.01). Condition factors

were not different among treatments with females, at the

end of fasting. There were no significant changes in BW

or SL with treatment. Nevertheless, there was a signifi­

cant drop in GSI for females fasted for 21 days (Fig. 18;

P < 0.01). The follicles of fasted fish were soft, sug­

gesting that atresia was beginning. The GSI for males did

not change with fasting.

The effects of fasting up to 31 days on serum

concentrations and pituitary content of the PRLs and GH in

male tilapia with a low condition factor

I did not see an expected increase in serum GH with

21 days of fasting with fish that had a high CF. For this

reason, the study was repeated with animals with a lower

CF, and fasting was extended to 31 days.

Serum concentrations of tPRL1 77 and tPRL1 88, but not

GH, were elevated in tilapia fasted for 10 days (Fig. 19;

both P < 0.05). Serum tPRL177 increased from 0.6 ± 0.1

ng/ml to 9.0 ± 2.8ng/ml and tPRL1 88 increased from

124

Figure 18. Changes in gonadosomatic index (8S1) duringfasting in male (M) and female (F) tilapia Oreochromismossambicus with a high condition factor (mean ± SE,N = 10 per treatment; ** P < 0.01).

125

2110

Females

o21o

-

-r-- f-

-r-- :-

-

**r- I--

-

-

-

Males

- rr- ~

o

4

2

3

1

1.5

0.5

4.5

3.5

2.5-ene

Days Fasted

126

Figure 19. Changes serum concentrations of prolactin andgrowth hormone during fasting in male tilapia Oreochromismossambicus with a low condition factor (mean ± SE,N = 9-10 per treatment; * P < 0.05, ** P < 0.01,*** P < 0.001).

127

**

o 10 21 31o 10 21 31o 10 21 31

- tPRL177 tPRL188

-** **

- *

*r-

- * -I-

*-r-

-

- GH

- +*-

r!--

III r-lrlrlI I I r -, J I I I I I I I I

20

18

-..-E 16........C)

..:.. 14e0.-~ 12COL.~

s::CI> 10CJe0 8CJ

E::s 6L.CI>en

4

2

0

Days Fasted

128

5.7 ± 0.3 ng/ml to 12.2 ± 1.3 ng/ml, comparing fish that

had been fed to fish that had been fasted. Serum PRL

concentrations were elevated in fish fasted for 21 and 31

days, compared with fed fish. Growth hormone concentra­

tions were less than 1 ng/ml in fish fasted up to 21 days

and then increased to 3.5 ± 0.7 ng/ml in fish fasted for

31 days (P < 0.001).

Pituitary tPRL1 8 8 levels were not different among fed

animals and animals fasted for 10 and 21 days. Pituitary

content of tPRL1 8 8 decreased with day 31 of fasting (Fig.

20; P < 0.01). Pituitary content of tPRL1 77 was not

significantly affected by fasting. Pituitary GH content

increased with fasting. Growth hormone rose from 7.9 ±

0.8 ~g/pit. in fed animals, to 13.0 ± 1.9 ~g/pit. in

animals fasted for 21 days (P < 0.05), and 14.3 ± 2.1

~g/pit. in tilapia fasted for 31 days (P < 0.01).

The effects of fasting up to 31 days on condition factor,

body weight, standard length, GSI and RSI in male tilapia

with a low condition factor

Body weight and SL were measured at the start of the

experiment, at the beginning of the fasting period, and at

the end of the fasting period (Table 2). Changes in BW,

SL and CF were compared from the start of the fasting

period to the end of the fasting period, as well as from

129

Figure 20. Changes pituitary content of prolactin andgrowth hor~one during fasting in male tilapia Oreochromismossambicus with a low condition factor (mean ± SE,N = 9-10 per treatment; * P < 0.05, ** P < 0.01).

130

25 -tPRL188

.

. 20 - GH...-c.---.CD

**=:t-...., tPRL177 *s:: 15 - **(I)....,c:0CJ

~

-t tC'O 10...-~ t rr iI-...-'l.

5 -

I I I Io 10 21 31I I I I I I I I I Io 10 21 31 0 10 21 31

Days Fasted

131

Table 2. Effects of fasting up tQ 31 days on condition factor, body weight (g)and stnadard length (cm) in male tilapia with a low condition factor

(CF > 3.2; Mean ± BE)

Start of fasting and end of fastingDays BW-B BW-E %CH t-T SL-B BL-E %CH t-T CF-B CF-E %CH t-TFed 8S.2 89.2 14.21 14.21 3.07 3.07

5.7 5.7 0.31 0.31 0.05 0.0510 89.5 86.6 -3.3 ns 14.14 14.11 -0.23 ns 3.10 3.02 -2.7 ns

6.2 6.6 0.35 0.41 0.03 0.0321 80.4 72.0 -10.5 ns 13.70 13.58 -0.87 ns 3.11 2.86 -8.1 **

3.6 3.2 0.23 0.23 0.05 0.0431 68.4 60.6 -11.3 ns 12.78 12.73 -0.39 ns 3.21 2.88 -10.3 ***

I-'w 5.50 4.70 0.37 0.36 0.05 0.04N

ANOVA * *** * ** n ***

Start of study and end of fastingFed 74.6 89.2 19.6 ns 13.31 14.21 6.75 ns 3.10 3.07 -1.1 ns

5.5 5.7 0.35 0.31 0.04 0.0510 72.7 86.6 19.0 ns 13.28 14.11 6.20 ns 3.07 3.02 -1.5 ns

4.4 6.6 0.32 0.41 0.04 0.0321 74.4 72.0 -3.2 ns 13.44 13.58 1. 03 ns 3.05 2.86 -6.1 **

3.5 3.2 0.23 0.23 0.04 0.0431 68.4 60.6 -11.3 ns 12.78 12.73 -0.39 ns 3.21 2.88 -10.3 ***

5.50 4.70 0.37 0.36 0.05 0.04ANOVA ns *** ns ** ns ***

(B, beginning; E, end; BW, body weight; SL, standard length; %CH, % change; t-T,t-Test comparisons; * p < 0.05, ** P < 0.01, *** P < 0.001;ns = not significant)

the start of the experiment to the end of the fasting

period, and the start of the experiment to the start of

the fasting period. The CFs did not differ among treat­

ment groups from the start of the experiment to the start

of fasting. This consistency in CF suggests that the

nutritional status of the animals remained unchanged

throughout the period before fasting was initiated.

Condition factor was reduced from 3.11 ± 0.05 to 2.86 ±

0.04 (P < 0.01) with 21 days of fasting, and from 3.21 ±

0.05 to 2.88 ± 0.04 (P < 0.001) with 31 days of fasting.

The difference (P < 0.05) in BW and SL among treat­

ment groups at the start of the fasting period is due to

the slightly smaller size of the fish in the group to be

fasted for 31 days at the start of the study, combined

with the growth of the fish during the interval between

the start of the experiment and the beginning of the

fasting period in the other groups. Body weight, SL and

CF decreased as fasting increased (P < 0.001, P < 0.01, P

< 0.001 respectively). There was a significant increase

in BW in the group fasted for 10 days that occurred be­

tween the start of the experiment and the start of the

fasting period (P < 0.05).

Hepatosomatic index was reduced at 10 days of fasting

(Fig. 21; P < 0.001). The HSI of fed fish was 1.72 ± 0.07

compared with 0.97 ± 0.05 for fasted fish. There was no

133

Figure 21. Changes in hepatosomatic index (HSI) duringfasting in male tilapia Oreochromis mossambicus with a lowcondition factor (mean ± SE, N = 10 per treatment;*** p < 0.001 ).

134

*

312110o

-

-- -

-

-

-

***

rf-***-

+ **

rl--

-

-

-

2

o

1.8

1.4

1.6

1.2

0.4

0.8

0.2

0.6

-UJ 1J:

Days Fasted

135

further reduction in HSI as fasting continued. There were

no differences in GSI among treatments at the end of the

study.

DISCUSSION

Serum concentrations and pituitary content of GH and

serum concentrations of tPRL1 7 7 and tPRL1 8 8 were elevated

in response to fasting in o. mossambicus. These data

suggest that GH and the tPRLs have roles in the regulation

of intermediary metabolism in the tilapia; o. mossambicus.

Furthermore, serum concentrations of the tPRLs increased

earlier in response to fasting than GH, and thus, the

tPRLs and GH may have different roles in regulating metab­

olism.

An increase in circulating GH appeared later in

response to fasting in o. mossambicus than has been ob­

served in most other teleosts. Serum GH concentrations

were elevated at 31 days of fasting in male tilapia with a

low CF. Serum concentrations of GH were unchanged for the

first 21 days of fasting in female o. mossambicus, with a

high CF, and in males with a high or a low CF. Increases

in circulating GH concentrations were observed within 1-3

weeks in rainbow trout (Oncorhynchus mykiss; Wagner and

McKeown, 1986; Sumpter et al., 1991) and within 16 days in

136

a tilapia hybrid (0. niloticus X o. aureus; Melamed,

1993). On the other hand, McKeown et ale (1975) did not

see an elevation in plasma GH concentrations in Kokanee

salmon fasted for 30 days.

The difference in the timing of the increase in

circulating GH levels among the studies may be due to a

variety of factors. Among these factors are differences

in basal metabolic rates of the different animals. These

differences may be due simply to species or species dif­

ferences combined with differences in developmental stage

or environment. For example, in my study there was only a

-31% difference in BW change between male tilapia that had

been fed and male tilapia that had been fasted for 31

days. The rainbow trout in the study by Wagner and

McKeown (1986) that had been fed were -250% greater in BW

than the fish that had been fasted for 21 days. The

greater weight loss observed in the trout compared with

the tilapia is likely due to differences in metabolism

related to the tilapia being a herbivore and the trout

being a carnivore. A second factor Gontributing to the

timing of an increase in circulating GH levels in response

to fasting is the metabolic state of the animals at the

start of the fasting period. The tilapia hybrid used in

the fasting study by Melamed (1993) were described as

IIsub-optimallyll fed. There is no further description of

137

the condition of the animals. In that study, the plasma

GH concentrations started at over 200 ng/ml and increased

to over 500 ng/ml following 16 days of fasting, or dropped

to 15-20 ng/ml following 16 days of "optimal feeding". In

my study, fish with low and high CFs had serum GH

concentrations below 2 ng/ml before fasting. The elevated

GH levels prior to fasting in the tilapia hybrids may

suggest that their energy reserves were more depleted than

any of the fish in the present study. Thus, the hybrid

tilapia may have required an increase in GH sooner after a

further reduction in feeding, to mobilize additional

energy stores or protect essential resources from catabo­

lism. Alternatively, the difference in response between

the 2 tilapias may be due to species differences in the

hormonal regulation of metabolism. The hybrid is selected

for fast growth and has higher baseline GH levels. Re­

gardless of the differences in the timing of the increases

in circulating GH levels, circulating concentrations of GH

were elevated in the tilapia in response to fasting as in

most species of teleosts examined.

The pituitary content of GH was elevated at 21 days

of fasting in all groups in both studies; male and female

tilapia that had a high CF (-3.6) prior to fasting, and

male tilapia that had a low CF (-3.2) prior to fasting.

The pituitary content of GH was not elevated at 10 days of

138

fasting in any of these groups. This is consistent with a

previous report by Rodgers and colleagues (1992) that

pituitary GH content was elevated with 2 weeks of fasting

in this same species. Increased pituitary GH content was

observed prior to an increase in serum GH concentration in

male tilapia fasted for 31 days. These data suggest that

pituitary GH may increase before release from the pitui­

tary is increased.

An earlier change in pituitary GH content with

fasting, than serum concentration with fasting, suggests

that changes in pituitary GH content may be the better

indicator of changes in GH cell activity with changes in

metabolic state accompanying fasting. Consistent with

this notion, pituitary GH content, but not serum concen­

tration, was elevated at 21 days of fasting, and at 21

days of fasting there were signs of changes in metabolic

state in the tilapia which had been fasted. In both

studies CF was reduced in male tilapia fasted for 21 days

compared with fed fish and compared with values from the

same fish prior to fasting. In addition, HSI was reduced

by 10 days of fasting in male tilapia suggesting that

liver energy stores were mobilized. The GSls of female

tilapia fasted for 21 days were reduced compared to those

of fed fish, suggesting ovarian growth was compromised.

Serum tPRL177 concentrations were increased with 10

139

days of fasting in male and female tilapia and in tilapia

with a low and a high CF. This is the first report of an

increase in circulating PRL concentrations in response to

fasting in a teleost. Prolactin177 appears more sensitive

to fasting than tPRLI 8 8. Unlike serum tPRL1 77

concentrations, serum tPRL1 8 8 concentrations were not

elevated in female tilapia with a high CF, which had been

fasted for 10 or 21 days. In the same study, serum

tPRL1 7 7 concentrations were increased 3 fold and tPRL1 8 8

levels were increased by about 70% in male tilapia at 21

days of fasting compared with fed tilapia. Serum tPRL1 7 7

and tPRL1 8 8 concentrations were elevated with 10 days of

fasting in male tilapia with a low CF.

Rodgers and colleagues (1992) found that RPD removed

from tilapia (0. mossambicus) fasted for 2 weeks had

reduced spontaneous release of PRL. The pituitary content

of PRL together with the hormone released into the medium

during 20 hr incubations, was also lower for the fasted o.

mossambicus (sex not identified). The increase in serum

PRL concentrations I observed in response to fasting may

at first appear to be inconsistent with the reduction in

spontaneous PRL release from RPD with fasting, reported by

Rodgers and colleagues (1992). It is possible that the

reduction in PRL release from RPD of fasted fish, observed

by Rodgers and colleagues, was the result of a depletion

140

of stored PRL in the RPD, prior to removal of the tissues.

This interpretation is consistent with the increase in

circulating tPRLs I observed with fasting and the reduced

PRL content in the pituitaries of the fasted tilapia in

the studies by Rodgers and colleagues.

Pituitary content of tPRL1 77 was elevated at 21 days

of fasting in female tilapia with a high CF whereas pitui­

tary content of tPRL177 in males and tPRL1 88 in males and

females were not altered by fasting. On the other hand,

pituitary content of tPRL1 88 was reduced in male tilapia

with a low CF at 31 days of fasting while there was no

change in pituitary tPRL1 77 content. The effect of

fasting on pituitary PRL content appears to be sex specif­

ic, however, more data are needed before this notion can

be confirmed.

The differences in serum concentration and pituitary

content of PRL and GH among fed and fasted tilapia ob­

served in this study are assumed to be in response to a

need to mobilize or protect energy reserves in response to

a switch from anabolic to catabolic activities, or at

least an increase in catabolic activities with fasting.

Several lines of evidence support that there were changes

in metabolic state with the imposed fasting. In the

second study with tilapia with a low CF, there was a

decline in CF at 21 and 31 days of fasting. Also in the

141

second study, there was a dramatic decrease in HSI by day

10 of fasting suggesting a mobilization and possible

depletion of available energy reserves in the liver. In

the first study, there was a decline in CF for males but

no decline in CF related to fasting for females. The

females may not have been in the same nutritional state at

the end of the study even though there were no differences

in CF between fed and fasted female tilapia. Although not

significant, there appeared to be a trend towards a de­

crease in BW and SL with increased fasting time. There

may have been reduced linear growth with fasting keeping

pace with any decline in body weight, thus, a decline in

CF with fasting was not observed. There may not have been

sufficient resources for maintaining linear growth in fish

which started with a low CF in Study 2 and therefore

changes in nutritional state were more easily detected in

changes in CF with fasting. A decline in GSI together

with an apparent softening of the follicles suggests that

the females in Study 1 were suffering from a nutritional

deficit, enough to impact ovarian development.

Although there were differences in changes in CF in

response to fasting between the two studies, suggesting

that there were differences between the two studies as to

the changes in metabolic states, there were no discernible

differences in hormone response. In both studies, one

142

with low CF (-3.2) and one with high CF (-3.6), pituitary

content but not serum concentrations of GH increased with

21 days of fasting. In addition, serum concentrations of

tPRL1 7 7 were increased by 10 days of fasting in both

studies and serum concentrations of tPRL1 8 8 were elevated

in males with 21 days of fasting in both studies. The

consistency of the responses may be due to the limited

sampling. Serum PRL concentrations may have increased

earlier within the first 10 days of fasting, and pituitary

GH content may have increased earlier between days 10 and

21 of fasting in the tilapia with a lower CF than in the

tilapia with a higher CF. On the other hand, the

metabolic or nutritional state of the tilapia at the start

of fasting may be of less importance than a change in

nutritional state.

The differences in response to fasting by PRL and GH

suggest the hormones may have disparate actions. Serum

tPRL1 7 7 increased much earlier than serum GH concentra­

tions. Prolactin may be mobilizing energy stores from the

liver which undergoes most of its weight loss within 10

days. Growth hormone may then rise to mobilize other

energy reserves or to inhibit the mobilizing or cataboliz­

ing of needed resources. Further study is needed to

understand the roles of PRL and GH in metabolic processes

associated with fasting in the tilapia. Among these

143

studies are kinetic studies to see if there are changes in

the metabolic clearance rates of the tPRLs or GH associat­

ed with fasting and studies on changes in body composition

and blood metabolites in response to fasting and homolo­

gous hormone treatment. Even then, these studies must be

conducted under different physiological conditions and at

different developmental stages.

In summary, serum concentrations of tPRL1 77

and tPRL1 88 increased soon after the initiation of

fasting while changes in pituitary content of GH and serum

concentrations of GH were not observed until late in

fasting. Prolactin1 77 appeared to be more sensitive to

changes in nutritional state than tPRL1 88. Changes in

pituitary content of GH were observed before changes in

serum concentrations of the hormone while the opposite was

true with PRL. In addition, the PRL response to fasting

may be sex specific. Prolactin1 77 content in the

pituitary increased with fasting in female tilapia with no

change in male tilapia. A decline in pituitary tPRL1 88

content was observed in males with a low CF at 31 days of

fasting but there were not females in the experiment for

comparison. These studies provide evidence that the tPRLs

and GH are active in the regulation of metabolism in the

tilapia, O. mossambicus and that the tPRLs and GH may have

disparate functions.

144

Chapter V

CHANGES IN SERUM CONCENTRATIONS AND PITUITARY CONTENT OF

PROLACTIN AND GROWTH HORMONE DURING THE REPRODUCTIVE CYCLE

OF FEMALE TILAPIA, OREOCHROMIS MOSSAMBICUS ADAPTED TO

FRESH WATER AND SEAWATER

INTRODUCTION

Prolactin and GH have actions in osmoregulation,

metabolism, and reproduction (Nicoll, 1974; Clarke and

Bern, 1980; Loretz and Bern, 1982; Hirano, 1986). These

processes are intertwined within the reproductive habits

of the tilapia, O. mossambicus. Oreochromis mossambicus

can reproduce in both FW and SW. Furthermore, feeding is

reduced in the female tilapia when it broods offspring in

its buccal cavity. Brooding lasts for approximately 3

weeks. The subject of this study was the roles of the

tPRLs and GH in the regulation of these disparate though

overlapping processes during reproduction in the female

tilapia.

Prolactin and GH may regulate reproductive functions

in the female tilapia. Prolactin has been implicated in

parental care in teleosts including cichlids. Behaviors

shown to be elicited by PRL are expressed in female tila-

145

pia. These behaviors include "calling movements" to fry

and reduced feeding behavior which may prevent the canni­

balism of young (Noble et al., 1938; Neil, 1964; Blum and

Fielder, 1965; Blum, 1966). Prolactin has also been shown

to induce mucus secretion as a source of nutrition for

developing fry (see review, Nicoll, 1974).

Fish PRLs and GHs have been shown to stimulate ovar­

ian growth and steroidogenesis in teleosts (Singh et al.,

1988; Tan et al., 1988; Danzmann et al., 1990; Van Der

Kraak et al., 1990). In the tilapia, tPRLs have been

shown to alter steroid release from testicular tissues of

courting and non-courting males in vitro (Rubin and

Specker, 1992). The tPRLs stimulate T production from

testis fragments of courting males but have no effect on

testicular tissues from non-courting males. The tPRLs

augment the response of testicular tissues from courting

males to ovine-luteinizing hormone but inhibit this

response in testis from non-courting males. In addition,

tPRL1 8 8 has been shown to stimulate E2 production by

vitellogenic oocytes of the guppy, Poecilia reticulata

(Tan et al., 1988). Further supporting a direct action of

PRL and GH in regulating steroidogenesis in teleosts is

the identification of PRL and GH binding sites in the

ovary. Prolactin binding sites have been identified in

the ovary of the tilapia and GH receptors have been iden-

146

tified in the ovaries of salmonids (Edery et al., 1984;

Mourot et al., 1992).

Reproductive hormones have been shown to stimulate or

modulate the release of PRL and GH in tilapia. The ster­

oid hormones, E2 and T have been shown to stimulate PRL

cells (Nagahama et al., 1975; Wigham et al., 1977; Barry

and Grau, 1986; Borski et al., 1991) and modulate the

response of PRL and GH cells to hypothalamic factors

including thyrotropin-releasing hormone, GnRH and GHRH

(Barry and Grau, 1986; Melamed et al., 1995).

Prolactin and GH regulate osmoregulatory functions in

the tilapia and may be active in these roles depending on

the salinity in which the tilapia is reproducing (cf.

Nishioka et al., 1988; cf. Grau et al., 1994). Pituitary

content and blood concentrations of the tPRLs are greater

in FW- than in SW-adapted tilapia (Nicoll et al., 1981;

Borski et al., 1992; Ayson et al.; 1993, Yada et al.,

1994). Furthermore, Borski and colleagues (1992) have

shown that the two tPRLs are differentially regulated by

environmental salinity (Borski et al., 1992). Pituitary

content of GH is greater in SW- than in FW-adapted tilapia

(Borski et al., 1994). Yada et al. (1994) observed an

increase in circulating GH shortly after transfer of male

o. mossambicus from FW to 70%-SW but no change in females.

In the same study, GH levels in the plasma were reduced

147

shortly after transfer from SW to FW in both sexes,

compared to animals at the same time point which had been

transferred from SW to SW. Interestingly, circulating

levels of GH were no longer different from controls, in

either direction of transfer and either sex, by 7 days

after transfer. Sakamoto and colleagues (1994) also

observed an increase in plasma GH following transfer of o.

mossambicus from FW to SW whereas plasma GH was not

altered following transfer from SW to FW. Furthermore,

the increase in plasma GH was observed in both males and

females and at the end of the study, day 14 after trans­

fer, plasma GH was elevated in males but not in females

(Sakamoto, personal communication).

Finally, the tPRLs and GH may regulate metabolic

activities in the brooding female tilapia which experi­

ences a reduction in food intake. I have shown in chapter

IV that serum and pituitary levels of the tPRLs and GH

increase in response to fasting. Changes in hormone

levels observed during the reproductive cycle of the

female tilapia may therefore derive from changes in nutri­

tional state. Sumpter and colleagues (1991a,b) have

determined that changes in circulating GH concentrations

during the reproductive cycle of male and female rainbow

trout are a consequence of changes in metabolic state and

148

not due to reproduction per se. A similar situation may

exist with the tilapia.

The foregoing discussion suggests that reproductive,

metabolic and osmoregulatory requirements of the female

tilapia may contribute to PRL and GH cell activity during

the reproductive cycle. In chapter IV, I have described

changes in serum concentrations and pituitary content of

the tPRLs and GH that occur with changes in metabolic

state induced by fasting. In the studies described in

the present chapter I characterized changes in serum

concentrations and pituitary content of the tPRLs and GH

that occur over the course of the reproductive cycle of

the female tilapia. I also examined the effects of the

salinity on the patterns of changes in the tPRLs and GH in

the serum and pituitary, observed during the reproductive

cycle. Changes in BW, SL, CF, GSI and HSI that occur over

the course of the reproductive cycle were also examined in

an attempt to evaluate the contribution of changes in

metabolic state to hormone changes. The aim was to gain

insight into how the tilapia balances the regulation of

different functions by the same hormones, by comparing

hormone patterns for female tilapia reproducing in FW and

SW, and hormone patterns observed during fasting.

Comparing changes in hormones during the reproductive

cycle of fish maintained under various conditions may help

149

to clarify which aspects of the hormone patterns are

intrinsic to the reproductive cycle. Three studies were

conducted. Study 1 was conducted in an outdoor tank with

FW animals. This study was conducted to examine changes

in serum concentrations and pituitary content of the tPRLs

and GH in tilapia that are reproducing in an environment

that is conducive to their natural behavior. Studies 2

and 3 were conducted in indoor tanks. The fish and envi­

ronmental conditions for studies 2 and 3 were similar

except that in Study 2, the fish were adapted to FW and in

Study 3 the fish were adapted to SW. Study 2 was also

designed to expand and refine sampling to include earlier

stages of the reproductive cycle. The purpose of Study 3

was to provide for a comparison of serum and pituitary

patterns of PRL and GH between FW- and SW-adapted tilapia.

MATERIALS AND METHODS

Animals

Tilapia, O. mossambicus were collected from brackish

water streams and maintained in FW in outdoor tanks at a

temperature of 22-25°C for at least 2 months before use.

The fish were fed to satiation twice daily with Purina

trout chow. Fish were handled and sampled using proce­

dures described in chapter IV. Prolactins and GH in serum

150

and pituitaries were measured using the homologous ra­

dioimmunoassays developed by Ayson et al., (1993) and

modified by Yada et al. (1994) as described in chapter

III. Pituitaries were collected and prepared for measure­

ment by radioimmunoassay as described in chapter IV.

The reproductive stage of the animals was classified

using two criteria. Follicle mass of the largest clutch

of follicles in the ovary was used to classify non­

brooding females, and stage of larval development of the

young was used to classify brooding females. Follicle

mass was determined by collecting and weighing a group of

20 follicles from the most advanced clutch of follicles

within the ovary. Follicle growth was found to be linear

over time starting from spawning in o. mossambicus

(Weber et al., 1992). Non-brooding females were grouped

according to follicle mass, to the nearest mg. Follicle

mass and brooding stage were recorded to investigate

changes in ovarian growth during the brooding phase of the

reproductive cycle. The days from spawning for each stage

were estimated from the stage of development of the eggs

and larvae, based on previous studies that examined the

rate of larval development in o. mossambicus (Weber et

al., 1989; Okimoto et al., 1993).

Initially, brooding females were classified into 10

stages based on the developmental stage of the young.

151

These stages are: early-egg, eyes not pigmented; mid-egg,

eyes pigmented but no spinal movement; late-egg, spinal

movement; early-yolksac, up until 50% of the larvae can

swim in the water column; mid-yolksac, up until pigmenta­

tion covers the sides of the larvae and yolk can only be

seen through the bottom of the fish; late-yolksac, up

until yolk can not be found in the larvae; females

brooding post-yolksac larvae. In most of the present

studies, these stages have been collapsed into just 3

brooding stages: females brooding eggs, females brooding

yolksac larvae, and females brooding post-yolksac larvae.

Females designated as simply non-brooding females were

females with follicles greater than 4 mg. Four mg

follicles are not found in females which are brooding.

Follicle mass does not usually exceed 3 mg in brooding

tilapia (Fig. 22). Follicles weigh about 6 mg at

ovulation.

Statistical analysis

Differences among groups were determined using analy­

sis of variance and the least significant difference test

for a priori pairwise comparisons analysis (Steel and

Torrie, 1980). Comparisons between two means were per­

formed using the unpaired Student's t-test. Data are

expressed as means ± SEe

152

Study 1

Fish were transferred to an outdoor vinyl-lined

broodstock tank (18 ft diameter) with a slow, continuous

flow of FW. The fish were not sampled for at least 2

months after transfer to allow the females time to begin

reproducing. Females averaged about 150-200 g. Water

temperature was dependent on ambient temperatures and

ranged from -22-26°C. An algal bloom was maintained in

the tank as a dietary supplement by controlling water

exchange rate. The tank also accumulated detritus upon

which the tilapia can feed. Detritus-like matter was seen

in the gut of brooding and non-brooding tilapia. The

females in this tank were observed to stay near the center

of the tank with the males establishing nest cites around

the perimeter.

Females were sampled from the outdoor tank on 3

occasions, once each in July, November and April, for a

total of 126 fish. Fish were collected by hand net and

placed immediately in anesthesia. Blood samples were

collected from all fish, and pituitaries were collected in

April and November as described in chapter IV. Also in

November; SL and liver weight was recorded for all females

for the determination of HSI and CF, and follicle mass was

recorded for brooding females for the examination of

153

ovarian growth during the brooding phase. Only females

with follicles larger than 4 mg were identified as non­

brooding females to insure that they were not brooding

females which had released their clutch before capture.

During the April collection, the time of capture was

recorded. Prolactin and GH levels were not correlated

with time from the start of collection, suggesting that

the collection method did not affect hormone levels (data

not shown).

Study 2

Fish were transferred to an indoor facility with a

fixed photoperiod of 14L:10D. Fish were distributed among

3 fiberglass tanks (500 1) fitted with a plexiglass win­

dow. In each tank, 20 females were housed with 2 males.

Females weighed 60-100 g. The tanks were outfitted with a

continuous flow of FW, constant aeration, and submersible

heaters that maintained the water temperature at 26.5 ±

0.5°C. The indoor tanks had limited detritus buildup.

The fish were not sampled for at least 8 weeks following

their introduction to the tanks. At the time of sampling,

females that were brooding were identified through the

windows of the tanks and removed first. This prevented

the females from releasing the clutches of young they were

brooding before I had a chance to identify the females as

154

brooding females. Females which had not been brooding but

ware in stages earlier than late vitellogenesis can not be

distinguished with certainty from brooding females with

similar size follicles that had been brooding but released

their clutch prior to capture. Therefore, sampling of

females during the early stages of ovarian growth required

that all females be identified as brooding females or non­

brooding females prior to collection.

Only non-brooding females whose follicle mass exceed­

ed 1.5 mg were included in the study. This is because

females with a follicle mass of less than 1.5 mg may not

have been reproductively mature or may have terminated

brooding prematurely. A mass of 1.5 mg was chosen as the

lower limit because FW females brooding post-yolksac

larvae were found to have follicles with a mass of just

under 1.5 mg (Fig. 22). Samples were collected on 5

occasions from July through December, from a total of 177

females. Body weight, SL, liver weight and gonad weight

were recorded during the final sample collection for

determination of HSI, CF and CF with gonad weight

subtracted from BW.

Study 3

Study 3 was the same as Study 2 except the fish were

adapted to SW. Fish were introduced into the tanks as in

155

Study 2 and the salinity of the water was slowly increased

to full-strength SW over a 1 week period. The tanks were

then supplied with a continuous flow of SW. The fish

were not sampled for at least 8 weeks following this

acclimation to SW. The non-brooding females were all

females which had a follicle mass greater than 4 mg.

Follicle mass was determined for all brooding SW fish for

the examination of ovarian growth during the brooding

phase. Samples were collected on 2 occasions in October

and 1 in November from a total of 53 females. Body

weight, SL and liver weight were recorded during the final

sample collection for the determination of RSI and CF.

RESULTS

Follicle growth during the brooding phase of the

reproductive cycle

Follicle growth during the brooding phase of the

reproductive cycle was examined in fish collected in Study

1 (FW, outdoor tank) and Study 3 (SW, indoor tank) .

Follicles in fish from Study 1 increased in size from 0.22

± 0.07 mg in females brooding early stage eggs to 1.31 ±

0.17 mg in females brooding post-yolksac larvae (Fig. 22).

Follicles in fis~ from Study 3 increased in size from 0.30

± 0.01 mg in females brooding early stage eggs to

156

Figure 22. Changes in follicle mass during the broodingperiod of the tilapia Oreochromis mossambicus. For eachstudy, points marked by different letters are significant­ly different (P < 0.05). Points without letters are N=land were not included in the analysis (mean ± SE; study 1N: EE=3, ME=5, LE=l, EY=6, MY=l, LY=6, PY=6, 28 fishtotal; study 3 N: EE=3, ME=4, LE=2 EY=3, MY=3, LY=5, PY=9,29 fish total). EE, early-egg; ME, mid-egg; LE, late-egg;EY, early-yolksac larvae; MY, mid-yolksac larvae; LY,late-yolksac larvae; PY, post-yolksac larvae broodingfemales.

157

3-. Study 1

b

2.5• Study 3 b

.-.tn

2E-tnU'J

beu 1.5:ECI)-U.--- 10

LL

0.5

aO-+--r---1r---T~--"'--'---r--1r---T~--,--.--.....--,.....-,.--r-~-'--r-I

o 2 4 6 8 10 12 14 16 18 20( EE ME LE EY MY LY PY)

Approximate day from spawning(Stage)

158

2.34 ± 0.31 mg in females brooding post-yolksac larvae.

Follicles in females brooding late-yolksac larvae were

similar in size to follicles in females brooding post­

yolksac larvae in both studies.

Changes in condition factor and hepatosomatic index with

brooding

The CF of FW females maintained in the outdoor tank

(study 1) did not significantly change with brooding

stage. The HSI was reduced in females carrying

early-yolksac larvae, about day 5-6 after spawning,

compared with non-brooding females (Fig. 23; P < 0.05).

Detritus-like matter was seen in the gut of brooding and

non-brooding tilapia in Study 1 but not in any of the

brooding fish in studies 2 or 3. Condition factor and HSI

were reduced in FW (Study 2) and SW (Study 3) females

brooding post-yolksac larvae compared with non-brooding

females (Fig. 24; P < 0.05). The drop in CF in FW fish

(Study 2) was significant even when gonad weights were

subtracted from body weights before CF was calculated

(P < 0.05).

159

Figure 23. Changes in condition factor (CF) andhepatosomatic index (HSI) during the brooding period ofthe tilapia Oreochromis mossambicus. For each subject,points marked by different letters are significantlydifferent (P < 0.05). Points without letter are N=l andwere not included in the analysis (mean ± SE; study 1 N:NB=8, EE=3, ME=S, LE=l, EY=6, MY=l, LY=6, PY=6, 28 fishtotal; study 3 N: NB=7, EE=3, ME=4, LE=2 EY=3, MY=3, LY=S,PY=9, 36 fish total). NB, non-brooding; EE, early-egg; ME,mid-egg; LE, late-egg; EY, early-yolksac larvae; MY, mid­yolksac larvae; LY, late-yolksac larvae; PY, post-yolksaclarvae brooding females.

160

4

-e- HSI

1.5 CF

3.5 +

b

a

a

f2

3

-en:J:...... 2.5LLo

o 2 4 6 8 10 12 14 16 18 20

( NB EE ME LE EY MY LY py)

Approximate day from spawning

(Stage)

161

Figure 24. Changes in condition factor (CF), CFcalculated without gonad weights included in body weightvalues (CF-GW), and hepatosomatic index (HSI) in non­brooding (NB) female tilapia Oreochromis mossambicus andfemale tilapia brooding post-yolksac-larvae, adapted tofresh water (FW) and seawater (SW) (mean ± SE, FW: N = 7-9per time point; SW: N = 4; * P < 0.05, ** P < 0.01,*** P < 0.001).

162

4 CF-GW

3.5

3

2.5-en~ 2LL(J

1.5

1

0.5

FW

CF

FW SW

163

D NB

D py

HSI

FW SW

Serum concentrations and pituitary content of PRL and GH

during the brooding cycle in FW-adapted female tilapia,

maintained in outdoor tanks (Study 1)

This study was conducted to examine changes in serum

concentrations and pituitary content of the tPRLs and GH

in tilapia that are reproducing in an environment that is

conducive to their natural behavior. Serum tPRL1 7 7

concentrations but not tPRL1 8 8 concentrations were

elevated in female tilapia at the end of the brooding

phase over levels in non-brooding females (Fig. 25).

Serum concentrations of tPRL1 77 were elevated in

females brooding post-yolksac larvae over those observed

in non-brooding females or in females brooding either eggs

or yolksac-larvae (P < 0.05). Serum concentrations of

tPRL1 8 8 remained at about 3-4 ng/ml throughout the brood­

ing cycle. Serum GH concentrations were also elevated in

female tilapia at the end of the brooding phase compared

with concentrations in non-brooding females. Serum GH

concentrations were below 1 ng/ml in non-brooding females

and in egg-brooding females before increasing to 2.3 ± 0.3

ng/ml in females brooding yolksac larvae (P < 0.01) and

4.9 ± 0.7 ng/ml in females brooding post-yolksac larvae

(P < 0.001).

Pituitary content of tPRL1 77 changed over the brood­

ing cycle whereas tPRL1 88 did not (Fig. 26). Pituitary

164

Figure 25. Changes in serum concentrations of prolactinand growth hormone during the brooding period of thetilapia Oreochromis mossambicus in fresh water.(mean ± SE, N = 30-47 per treatment for serum levels,

N = 24-35 for pituitary levels; 126 fish total;* p < 0.05, ** P < 0.01, *** P < 0.001 ). N, non-broodingfemales; E, females brooding eggs; Y, females broodingyolksac larvae; PY, females brooding post-yolksac larvae.

165

6 -

***tPRL177 tPRL188 GH

5 -.-.-E

*--ene- 4-e0.-..CO.....e 3 -CI)CJ **C0CJ ...~

E 2 -:::sa..Q)

tJ)

1 - rt rt

N E V PY N E Y PY

Stage

166

N E Y PY

Figure 26. Changes in pituitary content of prolactin andgrowth hormone during the brooding period of the tilapiaOreochromis mossambicus in fresh water. (mean ± SE,N = 30-47 per treatment for serum levels, N = 24-35 forpituitary levels; 126 fish total; * P < 0.05, ** P < 0.01,*** P < 0.001 ). N, non-brooding females; E, femalesbrooding eggs; Y, females brooding yolksac larvae; PY,females brooding post-yolksac larvae.

167

20-

18-tPRL177 tPRL188 GH ***

16-

.12-

_ 14-....-c.--C):0:-

8-

** *6- ~ If4- t rf

.....s::: 10­(1).....s::::ou~ca......-:::s......-n.

2-

N E Y PYN E Y PYo....L..l-....L.-L--I...-.a..-...L-....L.-L_--'--L-I.--I...-.L..-..I-...L.-L_---L..-L-I.--J-...L-.L......J.-L...

N E Y PY

Stage

168

content of tPRL1 77 increased from 4.3 ± 0.3 ~g/pit. in

non-brooding females to 6.0 ± 0.5 ~g/pit. in females

brooding eggs (P < 0.01), and 5.8 ± 0.3 ~g/pit. in females

brooding yolksac-larvae (P < 0.05). Pituitary tPRL1 77

content in females brooding post-yolksac larvae was not

different from those at any other stage. Pituitary con­

tent of tPRL188 remained at about 8-10 ~g/pit. throughout

the brooding cycle. Pituitary content of GH increased

from 10.5 ± 1.3 ~g/pit. in non-brooding females to 17.6 ±

1.1 ~g/pit. in females brooding post-yolksac larvae

(P < 0.001).

Serum PRL and GH concentrations during the reproductive

cycle of the female tilapia adapted to FW and maintained

in indoor tanks (Study 2).

The purpose of this study was to expand and refine

sampling to include earlier stages of the reproductive

cycle. Serum tPRL1 77 concentrations were lowest at the

end of ovarian growth, when the follicles average 5-6 mg

(4.6 ng/ml). Serum tPRL17 7 then increased to 7.8 ± 0.6

ng/ml in egg-brooding females (P < 0.01) and remained

elevated throughout the remainder of the brooding phase

(Fig. 27). The concentrations were not different among

the non-brooding stages or between the early ovarian

growth stages and the brooding stages. A significant

169

Figure 27. Changes in serum prolactin and growth hormoneconcentrations during the reproductive cycle of the femaletilapia Oreochromis mossambicus (mean ± SE,N = 12-48 per time point, 177 fish total, for each hor­mone, points marked by different letters are significantlydifferent at P < 0.05). Stages are classified by folliclemass (mg) and developmental stage of brooded young. E,females brooding eggs; Y, females brooding yolksac larvae;PY, females brooding post-yolksac larvae.

170

15 -.- GH

~ tPRL188

.-. 12--II- tPRL177

-E--C)c:.....e0 9.-...,f!.... bc(I)uc 60oE~ c~

(I)

UJ 3a

a a a

02 3 4 5 6 E y py

Stage

171

effect of reproductive stage with tPRL1 8 8 concentration

was not detected by analysis of variance (P = 0.14),

however, the pattern for tPRL1 88 appeared similar to that

of tPRL1 77. Growth hormone concentrations were low

throughout the ovarian growth phase in non-brooding fe­

males «1.5 ng/ml), and increased with the continuation of

brooding to reach 2.5 ± 0.3 ng/ml in females brooding

yolksac larvae (P < 0.001 compared with egg-brooding

females) and 3.3 ± 0.5 in females brooding post-yolksac

larvae (P < 0.05 compared with yolksac larvae brooding

females) .

Serum concentrations and pituitary content of PRL and GH

during the brooding cycle in SW-adapted female tilapia,

maintained in indoor tanks (Study 3)

The purpose of this study was to provide for a com­

parison of serum and pituitary patterns of PRL and GH

between FW- and SW-adapted tilapia. Only those non­

brooding females with follicles larger than 4 mg were

included in the study, for the reasons discussed above.

The acclimation of tilapia to SW for a minimum of 8

weeks reduced circulating tPRL1 77 to undetectable levels.

Serum tPRL1 88 concentration was reduced to -4 ng/ml, from

-9 ng/ml measured in FW tilapia under similar conditions

(from Study 2; Fig. 28). Serum tPRL1 8 8 in non-brooding

172

Figure 28. Changes in serum concntrations of prolactinand growth hormone during the brooding phase of the repro­ductive cycle in SW-adapted female tilapia, Oreochromismossambicus, maintained in indoor tanks; (mean ± SE,N = 6-20 per treatment, 53 fish total; * P < 0.05,*** P < 0.001 compared with non-brooding females unlessotherwise indicated). N, non-brooding females; E, femalesbrooding eggs; Y, females brooding yolksac larvae; PY,females brooding post-yolksac larvae.

173

tPRL188

*

**

N E Y PYN E Y PY

-GH '*

-r-

-

-*

..-1-

-

t +

5

o

4--E--C)s:::-e0 3.-....CO=-....s:::CI)us:::

20oE::J...(1)

en1

Stage

174

females was not significantly different from females

at any stage of the brooding phase. Serum tPRL1 8 8

concentrations were increased in females brooding

post-yolksac larvae over levels in females brooding eggs.

Serum GH concentrations increased as brooding continued.

Serum GH increased from 1 ng/ml in non-brooding and in egg

brooding females to 1.5 ± 0.2 ng/ml in females brooding

yolksac larvae (P < 0.05) and 3.7 ± 0.9 ng/ml in females

brooding post-yolksac larvae (P < 0.001, compared with

non-brooding females). Serum GH concentrations and

changes in concentrations were similar in all studies.

In SW, pituitary content of tPRL1 77 and tPRL1 88 did

not differ among females at various stages of the brooding

cycle (Fig. 29). Growth hormone content was elevated in

females brooding post-yolksac larvae compared with all

other stages (P < 0.05).

DISCUSSION

Patterns of changes in the tPRLs in serum and

pituitaries of the female tilapia, O. mossambicus, during

the reproductive cycle were affected by the salinity to

which the fish were adapted. Patterns of changes in GH,

on the other hand, were not affected by salinity. Fur­

thermore, the patterns for the tPRLs and GH observed

175

Figure 29. Changes in pituitary content of prolactin andgrowth hormone during the brooding phase of thereproductive cycle in SW-adapted female tilapia, Oreochro­mis mossambicus, maintained in indoor tanks; (mean ± SE,N = 6-20 per treatment, 53 fish total; * P < 0.05 comparedwith non-brooding females). N, non-brooding females; E,females brooding eggs; Y, females brooding yolksac larvae;PY, females brooding post-yolksac larvae.

176

14-

tPRL177 tPRL188 GH *12-

-

4-

6-

::s...-

~:! 10-.-::s...-Q.--.C)== 8-­..c:CI)..c:ou~CO...-a.

2-

N E Y PY N E Y PY N E Y PY

Stage

177

during the reproductive cycle had characteristics that

were both similar to and distinct from patterns observed

during fasting as described in chapter IV. These findings

suggest that changes in metabolic state may influence tPRL

and GH cell activity during the reproductive cycle al­

though they can not account for all the changes in serum

concentration and pituitary content of the hormones.

Overall, serum PRL concentrations were highest in

females brooding post-yolksac larvae in each of the three

studies. Nonetheless, changes in serum and pituitary

levels with the reproductive cycle appeared to be small

and there were differences among studies as to when PRL

levels increased during the brooding phase. The observed

patterns of serum concentration and pituitary content of

the tPRLs do not support the notion that changes in serum

concentrations of the tPRLs are driving parental care

behavior or ovarian function in the female tilapia. On

the other hand, the patterns of serum concentration and

pituitary content of GH during the reproductive cycle were

consistent in all three studies. Serum GH concentrations

increased during the brooding phase of the reproductive

cycle followed by an increase in pituitary content of the

hormone. These changes in serum and pituitary GH levels

appear to be a characteristic component of the reproduc­

tive cycle of the female tilapia and suggest that GH has

178

functions intrinsic to reproduction. In the following

segments, I will first discuss changes in the physiology

of the female tilapia during the reproductive cycle and

then relate these changes to the patterns of serum and

pituitary tPRLs and GH levels observed during the repro-

ductive cycle.

Ovarian growth during the brooding phase of the

reproductive cycle

Ovarian follicles grew in size throughout most of the

brooding phase in the 2 studies in which it was examined,

Studies 1 and 3. Follicles reached their maximum weight

when the females were brooding late-yolksac larvae, about

day 12 days post-spawning, and then stopped growing

throughout the remainder of the brooding phase. Follicles

increased in weight from less than 0.4 mg the day after

spawning up to 1.3 and 2.3 mg by the end of the brooding

phase in Study 1 and Study 3 fish respectively. The

difference in growth between Study 1 and Study 3 may be

due to a combination of factors including salinity, water

temperature, and size of the fish. The females in Study 1

were adapted to FW, maintained in at outdoor tank at

ambient temperatures (-22-26), and were about 150-200 g.

The females in Study 3 were adapted to SW, maintained

indoors in tanks heated to 26.5 ± 0.5°C, and were

179

60-100 g. What is of interest is that the follicles

stopped growing at about the same stage of the reproduc­

tive cycle in both studies and not at the same size.

Smith and Haley (1987, 1988) gave a similar account

of ovarian growth during the brooding phase of this

species following a histological examination of ovarian

development. They reported that ovarian growth was

arrested between days 15-25 after spawning. In addition,

they found that ovarian growth did not cease in females

which did not brood, even when fasted for 20 days. The

oocytes of some of the fasted females showed softening, an

early sign of atresia. I also observed a reduction in GSI

and softening of follicles in fasted female tilapia

(chapter IV). Smith and Haley (1988) suggested that the

interruption of oocyte growth is not due simply to a

reduction in food intake. This notion is supported by the

consistency in the timing of the interruption in oocyte

growth in both the studies by Smith and Haley (1987, 1988)

and in my studies; and the finding that oocyte growth is

interrupted at a particular stage, regardless of follicle

size. The size of the follicles and the stage of the

brooded young were not recorded in the studies by Smith

and Haley (1987, 1988).

Smith and Haley (1988) showed that E2 and T were

elevated during brooding and suggested that these steroids

180

may be involved in either brooding behavior or in

protecting the oocytes from atresia. Furthermore, Smith

and Haley (1987) found that postovulatory follicles

are steroidogenic and persist longer in brooding than non­

brooding females. They suggested that these structures

may also be involved in either brooding behavior,

protecting the oocytes from atresia or inhibiting oocyte

growth in the latter part of the brooding phase. I have

found that PRL and GH are elevated during the brooding

phase, and thus, may participate in these actions. More­

over, PRL and GH may have roles in regulating circulating

concentrations of levels of E2 and T, and in prolonging

the life of the postovulatory follicles. Alternatively,

PRL and GH may have metabolic roles during the brooding

phase, mobilizing and directing energy reserves.

Changes in condition factor and hepatosomatic index with

brooding in FW-and SW-adapted tilapia

Condition factor and RSI declined in females as

brooding continued. These data suggest that nutritional

status declined during the brooding phase of the reproduc­

tive cycle. The decline in CF between non-brooding

females and females brooding post-yolksac larvae was

significant for animals sampled from the indoor tanks

(Studies 2 and 3) but not for females from the outdoor

181

tank (Study 1). One possible reason CF was not reduced in

females brooding post-yolksac larvae in the outdoor tank

is that the females had resumed feeding. The outdoor tank

had detritus on the bottom and the fish had detritus-like

matter in their guts. Furthermore, there was more space

in the outdoor tank, possibly allowing the brooding fe­

males to release their clutches and feed. I think the

females in Study 1 did undergo changes in condition during

brooding that I was unable to detect. Although CF did not

decline significantly in females in Study 1, the mean CF

in females brooding yolksac larvae from the outdoor tank

(Study 1) was 12% lower than in non-brooding females.

This is the same difference in mean CF for non-brooding

females and females brooding post-yolksac larvae in

studies 2 and 3, where those differences were significant.

The 12% decline in CF observed during the brooding phase

is equal to or greater than reductions in CF that occurred

over the course of 21 days of fasting in female tilapia

(7.5%) and 31 days of fasting in male tilapia (10.5%;

chapter IV) .

In the present study, RBI was reduced from 2.3 in

non-brooding females to below 1.5 in females brooding

post-yolksac larvae, in all studies. A significant

reduction was observed in females brooding early-yolksac

larvae, about 6 days after spawning, (Study 1). The early

182

decline in HSI with brooding is comparable to a decline in

HSI observed with 10 days of fasting (chapter IV). The

drop in HSI during brooding was most likely due to nutri­

tional demands rather than a decline associated with a

decrease in vitellogenin synthesis activity. In support

of this notion, HSI was reduced from 2.4 in non-brooding

females to 1.7 in females brooding early-yolksac larvae,

even though oocyte growth continues beyond this stage

(Study 1). In addition, studies suggest that tilapia

produce vitellogenin which is secreted into mucus as a

source of nutrition for developing post-yolksac larvae

(Kishida and Specker, 1994).

Effect of salinity on serum and pituitary patterns of tPRL

during the brooding cycle

Salinity affected baseline serum concentrations and

the patterns of the tPRLs in serum and pituitaries during

the brooding cycle of the tilapia. Circulating titers

of the tPRLs were greatly reduced in SW fish (Study 3)

compared with FW fish (Study 2) held under similar condi­

tions. Prolactin177 was usually not detectable in the

serum of the SW-adapted tilapia. Reduced serum concentra­

tions of the tPRLs in the SW tilapia compared with levels

in FW tilapia is consistent with previous findings in o.

mossambicus (Nicoll et al., 1981; Borski et al., 1992;

183

Ayson et al., 1993; Yada et al., 1994). Differences in

levels and patterns of the tPRLs in serum and pituitaries

with salinity suggest that the tPRLs perform osmoregulato­

ry actions during the reproductive cycle of female tilapia

adapted to FW. Nevertheless, large differences in serum

tPRL concentrations in FW- compared with SW-adapted

tilapia do not imply that the hormones do not have actions

other than in osmoregulation. Low levels of the tPRLs may

be adequate for the actions of the tPRLs in reproduction

or metabolism.

Changes in serum tPRL1 8 8 were also different in the

FW- and SW-adapted tilapia during the reproductive cycle.

In FW, serum concentrations of tPRL1 88 did not change over

the reproductive cycle. In SW, by contrast, serum tPRL1 88

concentrations were elevated in females that were brooding

post-yolksac larvae over levels in females brooding eggs.

No changes in pituitary content of tPRL1 77 or tPRL1 88 were

observed in the SW-adapted tilapia. No changes in pitui­

tary content of tPRL1 88 were observed in the FW fish from

the outdoor tank (Study 1), however, pituitary tPRL1 77

content was elevated in females brooding eggs and females

brooding yolksac larvae, over levels in non-brooding

females.

184

Effect of salinity on patterns of GH in serum and

pituitaries during the brooding cycle

Salinity did not influence patterns of changes in

serum or pituitary GH during the brooding cycle of the

tilapia. Increases in circulating GH concentrations,

in response to transfer from FW to SW, have yet to be

observed in the tilapia, o. mossambicus, to last longer

than 14 days. Yada et al. (1994) observed a transient

increase in plasma GH concentrations in male tilapia after

transfer from FW to 70%-SW but no change in females.

Sakamoto and colleagues (1994) also observed an increase

in plasma GH following transfer of o. mossambicus from

FW to SW, whereas plasma GH was not altered following

transfer from SW to FW. Furthermore, the increase in

plasma GH following transfer to SW was observed in both

males and females, and at the end of the study, day 14

after transfer, plasma GH was elevated in males but not in

females (Sakamoto, personal communication). Transient

increases in blood GH occur after SW transfer in coho

salmon and rainbow trout. These rises are accompanied by

increases in metabolic clearance rate and secretion rates

of GH (Sakamoto et al., 1990, 1991). Whether there are

salinity-specific differences in metabolic clearance and

in the secretion of GH in the tilapia remains to be deter-

185

mined. Borski and colleagues (1994) found that pituitary

GH content was elevated in o. mossambicus reared in SW for

7 months and in tilapia transferred from FW to SW for 49

days. Nonetheless, in the present studies, changes in

serum and pituitary levels of GH during the brooding cycle

were similar in FW- and SW-adapted tilapia.

A final consideration is that female tilapia may not

respond to changes in salinity as do males. Only males

were used in the studies by Borski and colleague (1994).

Furthermore, Yada et al. (1994) did not observe changes in

plasma GH concentrations in female tilapia following

transfer from FW to SW despite seeing increases with

males.

Changes in serum concentrations and pituitary content of

PRL during the reproductive cycle of the female tilapia

Overall, serum PRL concentrations were highest in

females brooding post-yolksac larvae in each of the three

studies. In the FW-adapted fish, serum concentrations of

tPRL1 77 increased in brooding females over levels in

those in the later stages of vitellogenesis. Nonetheless,

the patterns of change in tPRL serum and pituitary levels

varied with rearing condition. In fish reared in the

outdoor tank, serum tPRL1 7 7 concentrations were not sig­

nificantly elevated until late in the brooding phase, that

186

is when females were brooding post-yolksac larvae. Serum

concentrations of tPRL1 77 were elevated early in the

brooding phase when tilapia were held in indoor tanks.

Serum concentrations of tPRL188 did not change sig­

nificantly in tilapia adapted to FW. In SW, by contrast,

serum concentrations of tPRL1 88 were elevated when females

were brooding post-yolksac larvae above those measured in

females brooding eggs. Prolactin1 77 was not detectable

in most of the serum samples taken from the SW-adapted

tilapia. For this reasn, it could not be determined

whether serum tPRL1 77 concentrations might vary in a

physiologically important manner.

The pattern of serum tPRL concentrations in non­

brooding tilapia during the oocyte growth phase may be a

continuation of the pattern observed in females brooding

post-yolksac larvae (Study 2). Female O. mossambicus

can spawn at 22 to 40 day intervals with oocyte growth

immediately following the termination of brooding (Neil,

1966; Smith and Haley, 1987). Serum tPRL concentrations

may increase with brooding and then, following the termi­

nation of brooding, slowly decrease as ovarian growth

progresses. Changes in serum tPRL1 88 in the FW fish

reared in the indoor tanks appear to parallel changes in

tPRL1 77, Thus, the same factors which control tPRL1 77

levels may also govern tPRL1 8 8.

187

Few conclusions can be drawn from the pituitary

patterns of the tPRLs observed during the reproductive

cycle. Pituitary content of tPRL1 77 was elevated in FW

females brooding eggs or brooding yolksac larvae compared

with levels in non-brooding females, late in vitellogene­

sis. On the other hand, there were no changes in tPRL1 7 7

content in the pituitaries from SW tilapia. There were no

changes in pituitary content of tPRL1 88 in FW or SW ani­

mals during the brooding cycle.

Although serum PRL concentration was highest in

females brooding post-yolksac larvae in all three studies,

these changes were relatively small compared with changes

observed with fasting (chapter IV) and transfer between FW

and SW (Yada et al., 1994). There was also an inconsist­

ency regarding when serum PRL concentrations begin to

increase in the brooding phase. This suggests that

changes in serum PRL concentrations do not drive parental

care or ovarian function in the female tilapia. This view

is supported by the conclusions of Wendelaar Bonga and

colleagues (1984) who found that PRL cell activity did not

vary between non-brooding and brooding female tilapia,

whether adapted to FW or SW. Their conclusions were based

on ultrastructure morphometry and incorporation rates of

[3 H]-lysine into PRL cells.

This conclusion does not preclude PRL from having

188

roles in reproduction in the tilapia, including actions in

parental care or ovarian function. It is not necessary

that there be large changes in PRL for PRL to have

affects. Changes may be occurring in other aspects of the

physiology of the animals associated with reproduction

that makes target tissues either less receptive or more

receptive to PRL stimulation. This may include changes in

PRL receptors in target tissues or changes in hormones

that either potentiate or antagonize the actions of PRL.

Changes in serum concentrations and pituitary content of

GH during the reproductive cycle of the female tilapia

The patterns of change in serum and pituitary GH were

consistent in all 3 studies. Serum GH concentrations

increased beginning when females were brooding yolksac­

larvae and continued to increase further in females

brooding post-yolksac larvae. Pituitary GH content was

elevated in females brooding post-yolksac larvae both in

SW and in FW. This suggests that an increase (> 3 fold)

in serum GH during the brooding stages of the reproductive

cycle is a characteristic component of the reproductive

cycle of the female tilapia.

189

Factors affecting PRL and GH cell activity during the

reproductive cycle of the female tilapia

Prolactin and GH cell activity is likely to be influ­

enced by many factors during reproduction. Among these

are salinity, steroid hormone concentrations, and nutri­

tion.

Salinity appeared to influence baseline PRL serum and

pituitary levels as well as the patterns in which serum

and pituitary PRL changed over the reproductive cycle.

Salinity did not influence the patterns of changes in

serum and pituitary GH over the reproductive cycle. Both

E2 and T are elevated in brooding tilapia (Smith and

Haley, 1988). The steroid hormones, E2 and T, have been

shown to stimulate PRL release directly and potentiate the

stimulatory effects of GnRH and thyrotropin-releasing

hormone on PRL release in the tilapia (Barry and Grau,

1986; Borski et al., 1991; chapter III). These same

steroid hormones have also been shown to potentiate the

stimulatory actions of GnRH and GHRH on GH release in a

tilapia hybrid (Melamed 1993) .

Changes in serum tPRL concentrations that occur over

the course of the reproductive cycle in female tilapia

vary with rearing condition. This may suggest that PRL

serves functions other than, or in addition to, regulating

aspects of reproduction, that are nonetheless tied to the

190

reproductive cycle. This may include metabolic actions.

Changes in CF and HSI observed during the reproductive

cycle were similar or greater than changes observed with

21 days of fasting in studies described in chapter IV. In

those studies, serum concentrations of the tPRLs were

increased with 10 days of fasting in male and female

tilapia. Thus, changes in metabolic state during the

reproductive cycle may be contributing to the changes in

serum tPRL concentrations observed during the course of

the reproductive cycle.

Sumpter et al. (1991a,b) found increases in

circulating GH concentrations were inversely correlated

with changes in CF during fasting and CF during the

reproductive cycle of male and female rainbow trout. They

concluded that changes in GH observed during the reproduc­

tive cycle were due to a loss of condition and not a

consequence of reproduction per se. Reductions in CF and

HSI during the brooding phase of the reproductive cycle in

tilapia suggest that a similar situation may exist in this

species.

The 12% decline in CF observed during the brooding

phase is equal to or greater than reductions in CF that

occurred over the course of 21 days of fasting in female

tilapia (7.5%) and 31 days of fasting in male tilapia

(10.5%; chapter IV). Pituitary content of GH were elevat-

191

ed in the female tilapia fasted for 21 days compared with

females which were fed, and serum concentrations of GH and

pituitary contents of GH were elevated in the male tilapia

fasted for 31 days. Overall, the decreases in HSI and CF

during the brooding phase of the reproductive cycle are

consistent with the idea that serum and pituitary levels

of GH are elevated in response to changes in metabolic

state occurring during the brooding phase of the reproduc­

tive cycle.

Although tPRL and GH serum concentrations and

pituitary content were elevated during brooding, the

patterns were different than with fasting. Serum tPRL17 7

concentrations were elevated prior to increases in tPRL1 77

in pituitaries of female tilapia during fasting. By con­

trast, pituitary tPRL1 7 7 levels were elevated prior to an

elevation in serum levels during the brooding cycle.

Serum concentrations of tPRL1 88 appeared to change in a

manner which is similar to tPRL17 7 during fasting and the

reproductive cycle. In both instances, these changes were

not statistically significant (chapter IV) .

Fasting caused elevations in serum and pituitary

levels of GH in o. mossambicus (chapter IV) . Increases in

pituitary content of GH were first observed at 21 days

of fasting in female tilapia and in females brooding

post-yolksac larvae (about 20 days post-spawning). Thus,

192

changes in pituitary content of GH occurred over a similar

time-course in both fasting and brooding female tilapia.

Nevertheless, serum levels of GH increased before pitui­

tary levels in females during the brooding cycle whereas

the reverse was found during fasting. These differences

in patterns between fasting and brooding animals suggest

that the changes in serum concentration and pituitary

content of the tPRLs and GH observed during the brooding

phase of the reproductive cycle are not due simply to a

decrease in food intake.

Possible actions of PRL and GH during the reproductive

cycle of the female tilapia

The present study suggests that the tPRLs have ac­

tions in osmoregulation during the reproductive cycle in

FW adapted-tilapia. The present study also suggests that

the tPRLs and GH may have actions associated with the

brooding phase of the reproductive cycle. It is unclear,

however, whether these actions are more closely aligned to

the regulation of metabolism or reproduction.

Prolactin and GH may have metabolic roles closely

associated with reproduction, ushering resources to or

away from ovarian growth and vitellogenin production. The

tPRLs and GH may release metabolites for vitellogenin

production or protect energy stores and protein from

193

utilization during this period when food intake is low.

The tPRLs and GH may have actions in parental behav­

ior, steroidogenesis, or maintenance of postovulatory

follicles. Possible actions in parental care include

"calling movements" to fry, reduced feeding behavior to

prevent feeding on young, and mucus production. Wedelaar

Bonga and colleagues (1984) dismissed a role for PRL in

parental care in the tilapia due to the absence of a

change in PRL cell activity. r have shown that PRL levels

are elevated in FW tilapia at the end of the brooding

period. Furthermore, as r have argued previously, large

changes in PRL release may not be necessary for PRL to

have actions. Changes in target tissue receptors and in

interacting hormones which may modulate PRL1s actions, may

change during the reproductive cycle.

r have mentioned previously that, Smith and Haley

(1988) have shown that E2 and T levels are elevated

during the brooding phase and steroidogenic postovulatory

follicles persist longer in brooding females than in

non-brooding females. The tPRLs or GH may be involved in

regulating these events. Fish PRLs including tPRLs, and

GHs have been shown to stimulate ovarian growth and ste­

roidogenesis in teleosts (Singh et al., 1988; Tan et al.,

1988; Danzmann et al., 1990; Van Der Kraak et al., 1990;

Rubin and Specker, 1992).

194

Clearly, more work is required before the actions of

PRL and GH during the reproductive cycle of the female

tilapia are clear. Studies should include: 1) metabolic

clearance studies for the tPRLs, GH and steroid hormones;

2) studies on the efficacy of native hormones to elicit or

maintain parental behavior and mucus production; 3) in

vitro studies on the effects of the tPRLs and GH on ste­

roidogenesis from ovarian tissues, including postovulatory

follicles; 4) studies on the efficacy of the tPRL and GH

to maintain postovulatory follicle function in non-brood­

ing tilapia; and 5) studies to determine whether tPRLs, GH

or steroid hormones can protect oocytes from atresia

during fasting.

195

Chapter VI

CONCLUSIONS

Consistent with PRL's role in FW osmoregulation, PRL

cells of the tilapia release PRL in response to small

changes in medium osmolality that are well within the

range of plasma osmolalities observed in vivo (Nagahama et

al., 1975; Wigham et al., 1977; Grau et al., 1981, Kelly

et al., 1988). I have shown that alterations in osmotic

pressure and not osmolality are coupled to the sustained

release of the osmoregulatory hormones, tPRL1 77 and

tPRL1 88. This is the first study to describe a direct

response to osmotic pressure by a cell with a clearly

demonstrated osmotic output. Prolactin release was high

in RPD incubated for 18-20 hr in hyposmotic medium (300

mOsmolal) or medium made hyperosmotic (355 mOsmolal) with

the mernbrane-permeant molecules, urea or ethanol, but was

low in RPD incubated in medium made hyperosmotic with NaCl

or the mernbrane-impermeant molecule, mannitol. These data

suggest that it is the osmotic gradient across the cell

membrane that leads to an increase in osmotic pressure,

and not the medium osmolality per se, which accounts for

the osmoreceptivity of the tilapia PRL cell. In perifu­

sion studies, tPRL1 77 and PRL1 88 release was the same from

RPD incubated in hyposmotic medium and in hyperosmotic

196

medium containing urea, at 22 min after the onset of expo­

sure to these media. Furthermore, release was greater in

both media compared with tPRL release from RPD incubated

in hyposmotic media made hyperosmotic with the membrane­

impermeant molecule, mannitol. These data provide addi­

tional evidence that the increase in tPRL release in

response to reductions in osmotic pressure is sustained.

I have shown that injections of mGnRHa elevate serum

concentrations of the tPRLs and incubation of RPD in

medium containing GnRH molecules that are native to the

tilapia, increases PRL release into the medium. These

data provide evidence that GnRH functions as a PRL­

releasing factor in the tilapia, O. mossambicus. The

forms of GnRH endogenous to the tilapia stimulated the

release of the tPRLs from RPD with the following order of

potency: cGnRH-II > sGnRH > sbGnRH. This action appears

to be directly on the PRL cells. Support for this notion

comes from several lines of evidence. First, RPD are a

nearly homogeneous population of PRL cells (95-99% pure) ,

GtH and GH cells which have been shown to respond to GnRH

are not present in the RPD. Second, the range of GnRH

concentrations used in the studies were compatible with

receptor mediated action. Thirdly, cGnRH-II evoked an in­

crease in [ca2+]i within 1 min of exposure and an eleva­

tion in PRL release within 2 min. These data suggest that

197

calcium operates as a second messenger in mediating the

effects of GnRH on PRL release.

Prolactin secretion in teleosts is thought to be

predominantly under inhibitory control, however, evidence

had suggested that a hypothalamic prolactin-releasing

factor was involved in the control of PRL secretion in

fish (reviewed by Clarke and Bern, 1980; Ball, 1981)

Barry and Grau (1986) have shown that thyrotropin­

releasing hormone stimulates PRL release in the tilapia

but only following E2 preincubation. Until the present

study, thyrotropin-releasing hormone was the only hypotha­

lamic factor shown to have the capability to stimulate PRL

release in the tilapia. Furthermore, the steroid hormones

E2 and T are able to potentiate the response of PRL cells

to GnRH, increasing percent release over 3 fold. The

steroids may regulate the sensitivity of the PRL cells to

GnRH stimulation durIng the reproductive cycle. Alterna­

tively, GnRH molecules may regulate the sensitivity of PRL

cells to T and E2 stimulation. The potentiation of the

response of PRL cells to GnRH and thyrotropin-releasing

hormone by sex steroids supports the suggestion by Barry

and Grau (1986) that there may be a shift in the control

of PRL secretion with changes in the reproductive state of

the tilapia.

Serum concentration and pituitary content of GH and

198

serum concentrations of tPRL1 77 and tPRL1 88 were elevated

in response to fasting in o. mossambicus. These data

suggest that GH and the tPRLs have roles in the regulation

of intermediary metabolism in the tilapia, o. mossambicus.

This is the first report of changes in circulating PRL

levels in response to fasting in a teleost. Furthermore,

serum concentrations of the tPRLs increased earlier in

response to fasting than GH, and thus, the tPRLs and GH

may have different roles in regulating metabolism.

Patterns of changes in serum and pituitary tPRL

levels observed during the reproductive cycle of the

female tilapia, o. mossambicus, were affected by the

salinity to which the fish were adapted, whereas GH pat­

terns were not affected by salinity. The patterns for the

tPRLs and GH observed during the reproductive cycle had

characteristics that were both similar to and distinct

from patterns observed during fasting. In addition,

changes in CF and HSI observed during the brooding phase

of the reproductive cycle were similar in magnitude to

changes observed with fasting which were accompanied by

changes in serum tPRL and GH concentrations and GH pitui­

tary content. These findings suggest that changes in

metabolic state may influence tPRL and GH cell activity

during the reproductive cycle although they can not ac-

199

count for all the changes in serum concentrations and

pituitary content of the hormones.

Overall, serum PRL concentrations were highest in

females brooding post-yolksac larvae. Nonetheless,

changes in serum and pituitary levels with the reproduc­

tive cycle appeared small and there were discrepancies

among studies as to when serum PRL concentrations in­

creased during the brooding phase. The observed patterns

of serum concentrations and pituitary content of the tPRLs

do not support the notion that changes in serum concentra­

tions of the tPRLs are driving parental care behaviors or

ovarian function in the female tilapia. On the other

hand, the pattern of serum and pituitary GH levels during

the reproductive cycle was consistent among three studies.

Serum GH concentrations increased during the brooding

phase of the reproductive cycle, followed by an increase

in pituitary content of the hormone. These changes in

serum and pituitary GH levels appear to be a characteris­

tic component of hormone changes during the reproductive

cycle of the female tilapia and suggest that GH has func­

tions intrinsic to reproduction.

200

REFERENCES

Andersson, E., Borg, B. and de Leeuw, R. 1989.Characterization of gonadotropin-releasing hormone bindingsites in the pituitary of the three-spined stickleback,Gasterosteus aculeatus. Gen. Compo Endocrinol. 76:41-45.

Ayson, F.G., Kaneko, T., Tagawa, M., Hasegawa, S., Grau,E.G., Nishioka, R.S., King, D.S., Bern, H.A. and Hirano,T. 1993. Effects of acclimation to hypertonic environmenton plasma and pituitary levels of two prolactins andgrowth hormone in two species of tilapia, Oreochromismossambicus and Oreochromis niloticus. Gen. CompoEndocrinol. 89:138-148.

Baker, B.I. and Wigham, T. 1979. Endocrine aspects ofmetabolic coordination in teleosts. Symp. Zool. Soc. Lond.44:89-103.

Ball, J.N. 1981. Hypothalamic control of pars distalis infishes, amphibians, and reptiles. Gen. Compo Endocrinol.,44:135-170.

Barry, T.P. and Grau, E.G. 1986. Estradiol-17g and thyro­tropin-releasing hormone stimulate prolactin release fromthe pituitary gland of a teleost fish in vitro. Gen. CompoEndocrinol. 62:306-314.

Becu-Villalobes, D., Lacau-Mengido, I.M., Thyssen, S.M.,Diaz-Torga, G.B. and Libertun, C. 1994. Effects of LHRHand ANG II on prolactin stimulation are mediated byhypophysial AT1 receptor subtype. Am. J. Physiol.266:E274-E278.

Bern, H.A., Nishioka, R.S. and Nagahama, Y. 1975. Therelationship between nerve fibers and adenohypophysialcell types in the cichlid teleost, Tilapia mossambica. In:Researches Biologiques Contemporaines. L. Arvy, ed.Imprimerie Vagner, Nancy, pp. 170-194.

201

Bjornsson, B.Th., Stefanson, S.O., Taranger, G.L., Hansen,T., Walther, B.TH. and Haux, C. 1991. Photoperiodiccontrol of plasma growth hormone levels and sexualmaturation of adult Atlantic salmon. In: Proceedings ofthe 4th International Symposium on Reproductive Physiologyof Fish. A. P. Scott, J. P. Sumpter, D. E. Kime, and M. S.Rolfe, eds., University of East Anglia, U.K., July 7-12,1991. Fish Symp 91, pubs., Sheffield, U.K .. pp. 161.

Blackard, W.G., Kikuchi, M., Rabinovitch, A. and Reynold,A.E. 1975. An effect of hyposmolarity on insulin releasein vitro. Am. J. Physiol. 228:703-713.

Blum, V. 1966. Zur hormonalen steuerung der brutpflegeeiniger cichliden. Zool. Jb. Physiol. Bd. 72:264-290.

Blum, V. and Fielder, K. 1965. Hormonal control ofreproductive behavior in some cichlid fish. Gen. CompoEndocrinol. 5:186-196.

Blum, V. and Weber, K.M. 1968. The influence of prolactinon the activity of steroid-3B-ol dehydrogenase in theovary of the cichlid fish, Aequidens pulcher. Experientia24:1259-1260.

Bogomolnaya, A., Rothbard, S. and Yaron, Z. 1984. Sexsteroids in tilapia at various phases of the breedingcycle: Preliminary in vivo and in vitro studies. Gen.Compo Endocrinol. 53:452.

Borski, R.J. 1993. The Regulation of Prolactin Releasefrom the Pituitary of the Tilapia, Oreochromismossambicus, by Cortisol and Environmental Salinity. Ph.Dthesis, University of Hawaii.

Borski, R.J., Helms, L.M.H., Richman, N.H.III and Grau,E.G. 1991. Cortisol rapidly reduces prolactin release andcAMP and 45 Ca2+ accumulation in the cichlid fish pituitaryin vitro. Proc. Natl. Acad. Sci. USA 82:2758-2762.

Borski, R.J., Hansen, M.U., Nishioka, R.S. and Grau, E.G.1992. Differential processing of the two prolactins of thetilapia, (Oreochromis mossambicus) in relation toenvironmental salinity. J. Exp. Zool. 264:46-54.

202

Borski, R.J., Yoshikawa, J.S.M., Madsen, S.S., Nishioka,R.S., Zabetian, C., Bern, H.A. and Grau, E.G. 1994.Effects of environmental salinity on pituitary growthhormone content and cell activity in the euryhalinetilapia, Oreochromis mossambicus. Gen. Compo Endocrinol.95:483-494.

Brown, P.S. and Brown S.C. 1987. Osmoregulatory actions ofprolactin and other adenohypophysial hormones. In:Vertebrate Endocrinology: Fundamentals and BiomedicalImplications, Vol. 2: Regulation of Water andElectrolytes. P. K. T. Pang, M. P. Schreibman, and W. H.Sawyer eds., Academic Press, San Diego, pp. 45-84.

Carnevali, 0., Mosconi, G., Yamamoto, K., Kobayashi, T.,Kikuyama, S. and Polzonetti-Magni, A.M. 1992. Hormonalcontrol of in Vitro vitellogenin synthesis in Ranaesculenta liver: Effects of mammalian and amphibian growthhormone. Gen. Compo Endocrinol. 88:406-414.

Chang, J.P, Cook, H., Freedman, G.L., Wiggs, A.J., Somoza,G.M., De Leeuw, R. and Peter, R.E. 1990. Use of apituitary cell dispersion method and primary culturesystem for the studies of gonadotropin-releasing hormoneaction in the goldfish, Carassius auratus. Gen. CompoEndocrinol. 77:256-273.

Chang, J.P., Jobin, R.M. and Wong, A.O.L. 1993.Intracellular mechanisms mediating gonadotropin and growthhormone release in the goldfish, Carassius auratus. FishPhysiol. Biochem. 11:25-33.

Clarke, W.C. 1973. Disc-electrophoretic identification ofprolactin in the cichlid teleost Tilapia and Cichlasomaand densitiometric measurements of its concentration inTilapia pituitaries during salinity transfer experiments.Can. J. Zool. 51:687-695.

Clarke, W.C. and Bern, H.A. 1980. Comparative endocrinolo­gy of prolactin. In: Hormonal Proteins and Peptides. Vol.8, C. H. Li, ed., Academic Press, New York, pp. 105-197.

203

Cook, H. Berkenbosch, J.W., Fernhout, M.J., Yu, K-L.,Peter, R.E., Chang, J.P. and Rivier, J.E. 1991.Demonstration of gonadotropin releasing-hormone receptorson gonadotrophs and somatotrophs of the goldfish: anelectron microscope study. Reg. Pep. 36:369-378.

Crim, L.W., St. Arnaud, R., Lavoie, M., Labrie, F. 1988. Astudy of LH-RH receptors in the pituitary gland of thewinter flounder (Pseudopleuronectes americanus Walbaum) .Gen. Compo Endocrinol. 69:372-377.

Danzman, R.G., Van Der Kraak, G.J., Chen, T.T. and D.A.Powers. 1990. Metabolic effects of bovine growth hormoneand genetically engineered rainbow trout growth hormone inrainbow trout (Oncorhynchus mykiss) reared at a hightemperature. Can. J. Fish. Aquat. Sci. 47:1292-1301.

Davidson, M. 1987. Effect of growth hormone oncarbohydrate and lipid metabolism. Endocr. Rev. 8:115-131.

Debeljuk, L., Aleman, I.T., and Schally, A.V. 1985.Antiserum to LH-RH blocks haloperidol-inducedhyperprolactinemia in female rats. Neuroendocrinol.40:185-187.

Denef, C. and Andries, M. 1983. Evidence for paracrineinteraction between gonadotrophs and lactotrophs inpituitary cell aggregates. Endocrinology 112:813-822.

DeRuiter, A.J.H., Wendelaar Bonga, S.E.W., Slijkuis, H.and Baggerman, B. 1986. The effect of prolactin on fanning-behavior in the male three-spined stickleback,Gasterosteus aculeatus L. Gen. Compo Endocrinol. 64:273-283.

De Vlaming, V.L. and Pardo, R.J. 1974. In vitro effects ofprolactin on liver lipid metabolism in the teleost fish,Notemiqonus chrysoleucas. Amer. Zool. 14:1270A.

De Vlaming, V.L., Sage, M. and Tiegs, R. 1975. A diurnalrhythm of pituitary prolactin activity with diurnaleffects of mammalian and teleostean prolactin on totalbody lipid deposition and liver metabolism in teleostfishes. J. Fish. BioI. 7:717-726.

204

Dharmamba, M. and Nishioka, R.S. 1968. Response of"prolactin-secreting" cells of Tilapia mossambica toenvironmental salinity. Gen. Compo Endocrinol. 6:295-302.

Edery, M., Young, G., Bern, H. and Steiny, S. 1984.Prolactin receptors in tilapia (Sarotherodon mossambicus)tissues: binding studies using 125I-labeled ovineprolactin. Gen. Compo Endocrinol. 56:19-23.

Ekengren, B., Peute, J. and Fridberg, G. 1978. Gonadotrop­ic cells in the Atlantic Salmon, Salmo salor: Anexperimental immunocytological, electron microscopicalstudy. Cell Tiss. Res. 191:187-203.

Fain, J.N. 1980. Hormonal regulation of lipid metabolismfrom adipose tissue. In: Biochemical Actions of Hormones.Vol. III, G. Litwack, ed. Acedemic Press, New York, pp.120-204.

Farbridge, K.J., Flett, P.A. and Leatherland, J.F. 1992.Temporal effects of restricted diet and compensatoryincreased dietary intake on thyroid function, plasmagrowth hormone levels and tissue lipid reserves of rainbowtrout Oncorhynchus mykiss. Aquaculture 104:157-174.

Farbridge, K.J. and Leatherland, J.F. 1992. Temporalchanges in plasma thyroid hormone, growth hormone and freefatty acid concentrations, and hepatic 5"-monodeiodinaseactivity, lipid and protein content during chronic fastingand re-feeding in rainbow trout (Oncorhynchus mykiss) .Fish Physiol. Biochem. 10:245-257.

Fostier, A., Jalabert, B., Billard, R., Brenton, B. andZohar, Y. 1983. The gonadal steroids., In: FishPhysiology, Vol. lOB, (W. S. Hoar, D. J. Randall and E. M.Donaldson, eds.) Academic Press Inc., New York.pp. 277 - 3 72 .

Fryer, J.N., Nishioka, R.S. and Bern, H.A. 1979.Somatostatin inhibition of teleost growth hormonesecretion. Gen. Compo Endocrinol. 39:244-246.

205

Goren, A., Zohar, Y., Fridkin, M., Elhanati, E. and Koch,Y. 1990. Degradation of gonadotropin releasing hormone inthe gilthead seabream, Sparus aurata: I. Cleavage ofnative salmon GnRH, mammalian LHRH and their analogs inthe pituitary. Gen. Compo Endocrinol. 79:291-305.

Grau, E.G. and Helms, L.M.H. 1989. The tilapia prolactincell: A model for stimulus-secretion coupling. FishPhysiol. Biochem. 7:11-19.

Grau, E.G. and Helms, L.M.H. 1990. The tilapia prolactincell-twenty-five years of investigation. In: Progress inComparative Endocrinology. A. Epple, C. G. Scanes, and M.H. Stetson eds., Wiley-Liss, Inc. pp. 534-540.

Grau, E.G., Nishioka, R.S. and Bern H.A. 1981. Effects ofosmotic pressure and calcium ion on prolactin release invitro from rostral pars distalis of the tilapiaSarotherodon mossambicus. Gen. Compo Endocrinol.45:405-408.

Grau, E.G., Nishioka, R.S. and Bern H.A. 1982. Effects ofsomatostatin and urotensin lIon tilapia pituitaryprolactin release and interactions between somatostatin,osmotic pressure, Ca2 + , and adenosine 3 1,5'-monophosphate

in prolactin release in vitro. Endocrinol. 110:910-915.

Grau, E.G., Nishioka, R.S. Young, G. and Bern, H.A. 1985.Somatostatin-like immunoreactivity in the brain of andpituitary of three teleost fish species. Gen. CompoEndocrinol. 65:12-18.

Grau, E.G., Richman, N.H. III and Borski, R.J. 1994.Osmoreception and a simple endocrine reflex of theprolactin cell of the tilapia oreochromis mossambicus. In:Prospectives in Comparative Endocrinology. K. G. Davey, R.E. Peter and S. S. Tobe, eds., National Research councilof Canada, Ottawa, Canada, pp. 251-256.

Gray, E.S. and Kelley, K.M. 1991. Growth regulation in thegobiid teleost, Gillichthys mirabilis: roles of growthhormone, hepatic growth hormone receptors and insulin-likegrowth factor-I. J. Endocrinol. 131:57-66.

206

Grynckiewicz, G., Poenie, M., and Tsien, R.Y. (1985). Anew generation of Ca++ indicators with greatly improvedfluorescence properties. J. BioI. Chern. 260:3440-3450.

Guyton, A.C. 1991. Textbook of Medical Physiology. EighthEdition. W.B. Saunders Co., Philadelphia, 1014pp.

Habibi, H.R. 1991. Homologous desensitization ofgonadotropin-releasing hormone (GnRH) receptors ingoldfish pituitary: effects of native GnRH peptides and asynthetic GnRH antagonist. BioI. Reprod. 44:275-283.

Habibi, H.R., Peter, R.E., Sokolowska, M., Rivier and J.E.Vale, W.W. 1987. Characterization of gonadotropin­releasing hormone (GnRH) binding to pituitary receptors ingoldfish (Carassius aurata). BioI. Reprod. 36:844-853.

Helms, L.H.M., Grau, E.G., Shimoda, S.K., Nishioka, R.S.and Bern, H.A. 1987. Studies on the regulation of growthhormone release from the proximal pars distalis of maletilapia, Oreochromis mossambicus, in vitro. Gen. CompoEndocrinol. 65:48-55.

Higgs, D.A. Donaldson, E.M., Dye, H.M. and McBride, J.R.1976. Influence of bovine growth hormone and L-thyroxineon growth, muscle composition, and histological structureof the gonads, thyroid, pancreas, and pituitary of cohosalmon (Oncorhynchus kisutch). J. Fish. Res. Board Canada33:1585-1603.

Hirano, T. 1986. The spectrum of prolactin action inteleosts. In: Comparative Endocrinology: Developments andDirections. C. L. Ralph, ed., Alan R. Liss, New York,pp. 53-74.

Hirano, T. 1991. Hepatic receptors for homologous growthhormone in the eel. Gen. Compo Endocrinol. 81:383-390.

Ho, S.-M., Taylor, S. and Callard, I.P. 1982. Effects ofhypophysectomy and growth hormone on estrogen-inducedvitellogenesis in the freshwater turtle, Chrysemys picta.Gen. Compo Endocrinol. 48:254-260.

207

Ho, S.-M., Wangh, L.J. and Collard, I.P. 1985. Sexualdifferences in the in vitro induction of vitellogenesis inthe turtle: Role of the pituitary and growth hormone.Compo Biochem. Physiol. 81B:467-472.

Horseman, N.D. and Meier, A.H. 1979. Circadianphase-dependent prolactin mechanisms in hepaticlipogenesis of a teleost fish. J. Endocrinol. 82:367-372.

Huang, L. and Specker, J.L. 1994. Growth hormone- andprolactin-producing cells in the pituitary gland ofstriped bass (Morone saxatilis): immunocytochemicalcharacterization at different stages of maturity. Gen.Compo Endocrinol. 94:225-236.

Huang, Y.P. and Peter, R.E. 1988. Evidence for gonadotro­pin-releasing hormone binding protein in goldfish(Carassius auratus) serum. Gen. Compo Endocrinol.69:308-316.

Inui Y., Miwa, S. and Ishioka, H. 1985. Effects ofmammalian growth hormone on amino acid nitrogenmobilization in the eel. Gen. Compo Endocrinol.59:287-294.

Jobin, R.M. and Chang, J.P. 1992. Actions of two nativeGnRHs and PKC activators on goldfish pituitary cells.Studies on intracellular levels and gonadotropin release.Cell Calcium 13:531-540.

Kelley K.M., Nishioka, R.S.and Bern, H.A. 1988. Noveleffect of vasoactive intestinal polypeptide and peptidehistidine isoleucine: inhibition of in vitro secretion ofprolactin in the tilapia, Oreochromis mossambicus. Gen.Compo Endocrinol. 72:97-106.

Kindler, P.M., Bahr, J.M. Gross. M.R. and Philipp, D.P.1991. Hormonal regulation of parental care behavior innesting male bluegills: Do the effects of bromocriptinesuggest a role for prolactin? Physiol. Zool. 64:310-322.

208

Kishida, M., and Specker,surface mucus of tilapia,possibility for uptake byExp. Zoo1 . (In press) .

J.L. 1994. Vitellogenin in the(Oreochromis mossambicus) :the free-swimming embryos. J.

Kugu, K., Taketani, Y. andprolactin levels caused byhormone. Endocrinol. Japon

Mizuno, M. 1988. Changes inluteinizing hormone releasing35:545-548.

Kwon, H.C. and Mugiya, Y. 1994. Involvement of growthhormone and prolactin in the induction of vitellogeninsynthesis in primary hepatocyte culture in the eel,Anguilla japonica. Gen. Compo Endocrinol. 93:51-60.

Lamberts, S.W.J. and MacLeod, R.M. 1990. Regulation ofprolactin secretion at the level of the lactotroph.Physiol. Rev. 70:279-318.

Lamberts, W.S., Uitterlinden, P., Reubi, J-C.and de Jong,F.H. 1989. Effects of gonadotropin-releasing hormone andits agonist on prolactin secretion from normal andtumorous pituitary cells. Neuroendocrinology 49:157-163.

Leatherland, J.F. and Nuti, R.N. 1981. Effects of bovinegrowth hormone on plasma FFA concentrations and liver,muscle and carcass lipid content in rainbow trout, Salmogairdneri Richardson. J. Fish BioI. 19:487-498.

Lee R. and Meier, A.H. 1967. Diurnal variations offattening response to prolactin in the golden topminnow,Fundulus heteroclitus. J. Exp. Zool. 71:275-276.

Leeuw de R., Conn, P.M., vanlt Veer, C., Goos, H.J.Th. andvan Oordt, P.G.W.J. 1988. Characterization of the receptorfor gonadotropin-releasing hormone in the pituitary of theAfrican catfish, Clarias gariepinus. Fish Physiol.Biochem. 5:99-107.

209

LeGac, F., Ollitrault, M., Loir, M. and LeBail, P.Y. 1991.Binding and action of salmon growth hormone (sGH) inmature trout testes. In: Proceedings of the 4thInternational Symposium on Reproductive Physiology ofFish. A. P. Scott, J. P. Sumpter, D. E. Kime and M. S.Rolfe, eds., University of East Anglia, U.K., July 7-12,1991. Fish Symp 91, Pubs., Sheffield, U.K., pp. 117-111.

Leung, T.C., Ng, T.B. and Woo, N.Y.S. 1991. Metaboliceffects of bovine growth hormone in the tilapiaOreochromis mossambicus. Compo Biochem. Physiol.99A:633-636.

Levavi-Sivan, B. and Yaron, Z. 1989. Gonadotropinsecretion from peri fused tilapia pituitary in relation togonadotropin-releasing hormone, extracellular calcium, andactivation of protein kinase C. Gen. Compo Endocrinol.75:187-194.

Lewis, D.L., Goodman, M.B., John, P.A.S., and Barker, J.L.(1988). Calcium current and fura-2 signals influorescence-activated cell sorted lactotrophs andsomatotrophs of rat anterior pituitary. Endocrinolgy.123:611-621.

Loretz, C.A. and Bern, H.A. 1982. Prolactin andosmoregulation in vertebrates. Neuroendocrinology35:292-304.

Marchant, T.A., Chang, J.P., Nahornial, C.S. and Peter,R.E. 1989. Evidence that gonadotropin-releasing hormonealso functions as a growth hormone-releasing factor in thegoldfish. Endocrinology 124:2509-2518.

Marchant, M.A. and Peter, R.E. 1986. Seasonal variationsin body growth rates and circulating levels of growthhormone in the goldfish, Carassius auratus. J. Exp. Zool.237:231-239.

McKeown, B.A., Leatherland, J.F. and Johns, T.M. 1975. Theeffect of growth hormone and prolactin on the mobilizationof free fatty acids and glucose in the Kokanee salmon,Oncorhynchus nerka. Compo Biochem. Physiol. B50:425-430.

210

Melamed, P., Eliahu, N., Levavi-Sivan, B., Ofir, M.,Farchi-Pisanty, 0., Rentier-Delrue, F., Smal, J. Yaron, Z.and Naor, Z. 1995. Hypothalamic and thyroidal regulationof growth hormone in tilapia. Gen. Compo Endocrinol.97:13-30.

Melamed, P.M. 1993. Hypothalamic Factors Regulating theSecretion of Growth Hormone in Tilapia. Masters thesis,Tel-Aviv University.

Minick, M.C. and Chavin, W. 1970. Effect of pituitaryhormones upon serum free fatty acids in goldfish(Carassius auratus L.). Amer. Zool. 9:1083A.

Mourot, B., Fostier, A., Yao, K., LeGac, F. and Le Bail,P.Y. 1992. Specific binding of salmonid growth hormone inrainbow trout (Oncorhynchus mykiss) ovary. In: Proceedingsof the 2nd International Symposium on Fish Endocrinology,Saint Malo, France, June 1-4, 1992, p. 73.

Nagahama, Y., Nishioka, R.S., Bern, H.A. and Gunther, R.L.1975. Control of prolactin secretion in teleosts, withspecial reference to Gillichthys mirabilis and Tilapiamossambica. Gen. Compo Endocrinol. 25:166-188.

Neil, E.H. 1964. An analysis of color changes and socialbehavior of Tilapia mossambica. Univ. Calif. Publ. Zool.75:1-58.

Neil, E.H. 1966. Observations on the behavior of Tilapiamossambica (Pisces, Cichlidae) in Hawaiian ponds. Copeia1:50-56.

Niall, H.D., Hogan, M.L., Sauer, R., Rosenblum, I.Y. andGreenwood, F.C. 1971. Sequences of pituitary and placentallactogenic and growth hormones: Evolution from aprimordial peptide by gene reduplication. Proc. Nat. Acad.Sci. USA 68:866-871.

Nicoll, C.S. 1974. Physiological actions of prolactin. In:Handbook of Physiology, sec. 7, Endocrinology, vol. 4, pt.2, The Pituitary Gland and Its Neuroendocrine Control., E.Knobil and W. H. Sawyer eds., Williams & Wilkins,Baltimore, pp. 253-292.

211

Nicoll, C.S. 1982. Prolactin and growth hormone:Specialists on one hand and mutual mimics on the other.Prespect. BioI. Med. 25:369-381.

Nicoll, C.S. and Bern, H.A. 1971. On the actions ofprolactin among vertebrates: is there a commondenominator? In: Ciba Foundation Symposium on LactogenicHormones. G. E. W. Wolstenholme and J. Knight, eds.,Churchill Livingston, New York, pp. 299-324.

Nicoll, C.S., Wilson, S.W., Nishioka, R.S. and Bern, H.A.1981. Blood and pituitary prolactin levels in tilapia(Sarotherodon mossambicus; Teleostei) from differentsalinities as measured by a homologous radioimmunoassay.Gen. Compo Endocrinol. 44:365-373.

Nishioka, R.S., de Jesus, E.G.T. and Hyodo, S. 1993.Localization of mRNAs for a pair of prolactins and growthhormone in the tilapia pituitary using in situhybridization with oligonucleotide probes. Gen. CompoEndocrinol. 89:72-81.

Nishioka, R.S., Grau, E.G. and Bern, H.A. 1985. In vitrorelease of growth hormone from the pituitary gland oftilapia (Oreochromis mossambicus). Gen. Compo Endocrinol.60:90-94.

Nishioka, R.S., Kelley, K.M. and Bern, H.A. 1988. Controlof prolactin and growth hormone secretion in teleostfishes. Zool. Sci. 5:267-280.

Noble, G.K., Kumpf, K.F. and Billings, V.N. 1938. Theinduction of brooding behavior in the jewel fish.Endocrinology 23:353-359.

Okimoto, D.K., Weber, G.M. and Grau, E.G. 1993. Theeffects of thyroxine and propylthiouracil treatment onchanges in body form associated with a possible develop­mental thyroxine surge during post-hatching development oftilapia, Oreochromis mossambicus. Zool. Sci. 10:803-811.

212

Pagelson, G. and Zohar, Y. 1992. Characterization ofgonadotropin-releasing hormone (GnRH) receptors in thegilthead seabream (Sparus aurata). BioI. Reprod.47:1004-1008.

Perez-Sanchez, J., Weil, C. and Le Bail, P.-Y. 1992.Effects of human insulin-like growth factor-Ion releaseof growth hormone by rainbow trout (Oncorhynchus mykiss)pituitary celts. J. Exp. Zoo1. 262:287-290.

Peter, R.E., Habibi, H.R., Chang, J.P., Nahorniak, C.S.Yu, K.L., Huang,Y.P. and Marchant, T.A. 1990. Actions ofgonadotropin-releasing hormone (GnRH) in the goldfish. In:Progress in Comparative Endocrinology., A. Epple, C.Scanes and M. H. Stetson, eds., Wiley-Liss, Inc., NewYork, pp. 393-399.

Poenie, M., Alderton, J., Tsien, R.Y., and Steinhardt,R.A. 1986. Changes of free calcium levels with stages ofthe cell division cycle. Nature. 315:147-149.

Powell, J.F.F., Zohar, Y., Elizur, A., Park, M., Fischer,W.H., Craig, A.G., Rivier, J.E., Lovejoy, D.A. andSherwood, N.M. 1994. Three forms of gonadotropin­releasing hormone characterized from brains of onespecies, Proc. Natl. Acad. Sci. USA 91:12081-12085.

Prunet, P. Avella, M., Fostier, A., Bjornsson, B.T.,Boeuf, G. and Haux, C. 1990. In: Progress in ComparativeEndocrinology, A. Epple, C. G. Scanes, and M. H. Stetsoneds., Wiley-Liss, Inc., New York, pp. 547-552.

Richman, N.H., Helms, L.M.H., Ford, C.A., Beneshin, C.,Pang, P.K.T., Cooke, I.M. and Grau, E.G. 1990. Effects ofdepolarizing concentrations of K+ and reduced osmoticpressure on calcium accumulation by the rostral parsdistalis of tilapia, Oreochromis mossambicus. Gen. CompoEndocrinol. 77.292-297.

Richman, N.H., Ford, C.A., Helms, L.M.H., Cooke, I.M.Pang, P.K.T. and Grau, E.G. 1991. The loss of 45 Ca++associated with prolactin release from the tilapia,Oreochromis mossambicus rostral pars distalis. Gen. CompoEndocrinol. 83:56-67.

213

Rodgers, B.D., Helms, L.M.H. and Grau, E.G. 1992. Effectsof fasting, medium glucose, and amino acid concentrationson prolactin and growth hormone release, in vitro, fromthe pituitary of the tilapia, Oreochromis mossambicus.Gen. Compo Endocrinol. 86:344-351.

Rubin, D.A. and Specker, J.L. 1992. In vitro effects ofhomologous prolactins on testosterone production by testesof tilapia (Oreochromis mossambicus). Gen. CompoEndocrinol. 87:189-196.

Sakamoto, T. and Hirano, T. 1991. Growth hormone receptorsin the liver and osmoregulatory organs of rainbow trout:Characterization and dynamics during adaptation toseawater. J. Endocrinol. 130:425-433.

Sakamoto, T., Iwata, M. and Hirano, T. 1991. Kineticstudies of growth hormone and prolactin during adaptationof coho salmon, Oncorhynchus kisutch, to differentsalinities. Gen. Compo Endocrinol. 82:184-191.

Sakamoto, T., McCormick, S.D. and Hirano, T. 1993.Osmoregulatory actions of growth hormone and its mode ofaction in salmonids: A review. Fish Physiol. Biochem.11:155-164.

Sakamoto, T., Ogasawara, T. and Hirano, T. 1990. Growthhormone kinetics during adaptation to a hyperosmoticenvironment in rainbow trout. J. Compo Physiol. B160:1-6.

Sakamoto, T., Shepherd, B.S., Nishioka, R.S., Madsen,S.S., Siharath, J., Richman, N.H.III, Chen, T.T., Bern,H.A. and Grau, E.G. 1994. Osmoregulatory actions of growthhormone in an advanced teleost. Amer. Zool. 35:43A.

Saavedra, J.M. 1992. Brain and pituitary angiotensin.Endocr. Rev. 13:329-380.

Sato, N., Wang, X., Greer, M.A., Greer, S.E., McAdams, S.and Oshima, T. 1990. Medium hyposmolarity stimulatesprolactin secretion in GH4C1 cells by inducing an increasein cytosolic free calcium. Endocrinol. 127:957-964.

214

Sato, N., Wang, X. and Greer, M.A. 1991. Hormone secretionstimulated by ethanol-induced cell swelling in normal ratadenohypophysial cells. Am. J. Physiol. 260:E946-E950.

Sato, S. 1980. Post-natal development, sexual differenceand sexual cycle variation of prolactin cells in rats;special references to the topographic affinity of agonadotroph. Endocrinol. Jpn. 27:573-583.

Schulz, R.W., Bosma, P.T., Zandbergen, M.A., Van DerSanden, M.C.A., Van Dijk, W., Peute, J., Bogerd, J. andGoos, H.J.Th. 1993. Two gonadotropin-releasing hormones inthe African catfish, Clarias gariepinus: Localization,pituitary receptor binding, and gonadotropin releaseactivity. Endocrinology 133:1569-1577.

Shepherd, B.S., Sakamoto, T., Nishioka, R.S., Richman,N.H., Mori, I., Hirano, T., Bern, H.A. and Grau, E.G.1994. Growth related activities of the growth hormone andtwo prolactins in the hypophysectomized, euryhalinetilapia (Oreochromis mossambicus). In: Applications ofEndocrinology to Pacific Rim Aquaculture, University ofCalifornia, Davis Bodega Marine Laboratory,September 7-11, 1994.

Sheridan, M.A. 1986. Effects of thyroxine, cortisol,growth hormone, and prolactin on lipid metabolism of cohosalmon, Oncorhynchus kisutch, during smoltification. Gen.Compo Endocrinol. 64:220-238.

Sherwood, N.M., Parker, D.B., McRory, J.E. and Lescheid,D.W. 1994. Molecular evolution of growth hormone-releasinghormone and gonadotropin-releasing hormone. In: FishPhysiology. Vol. XIII, w. S. Hoar and D. J. Randall eds.,Academic Press, New York, pp. 3-66.

Singh, H., Griffith, R.W., Takashi, A., Kawauchi, H.,Thomas, P. and Stegeman, J.J. 1988. Regulation of gonadalsteroidogenesis in Fundulus heteroclitus by recombinantsalmon growth hormone and purified salmon prolactin. Gen.Compo Endocrinol. 72:144-153.

215

Singh, S.P. and Singh, T.P. 1981. Effect of sex steroidson pituitary and serum prolactin level in ovariectomizedcatfish, Clarias batrachus. Annales d'Endocrinologie42:57-62.

Singh, H. and Thomas, P. 1993. Mechanisms of stimulatoryaction of growth hormone on ovarian steroidogenesis inspotted seatrout, Cynoscion nebulosus. Gen. CompoEndocrinol. 89:341-353.

Slijkuis, H., De Ruiter, A.J.H., Baggerman, B. andWendelaar Bonga, S.E. 1984. Parental fanning behavior andprolactin cell activity in the male three-spined stickle­back Gasterosteus aculeatus L. Gen. Compo Endocrinol.54:297-307.

Smith, C.J. and Haley, S.R. 1987. Evidence of steroidogen­esis in postovulatory follicles of the tilapia,Oreochromis mossambicus. Cell Tiss. Res. 247:675-687.

Smith, C.J. and Haley, S.R. 1988. Steroid profiles of thefemale tilapia, Oreochromis mossambicus, and correlationwith oocyte growth and mouthbrooding behavior. Gen. CompoEndocrinol. 69:88-98.

Snedecor, G.W. and Cochran, W.G. 1980. StatisticalMethods. Iowa State Univ. Press, Ames, 507p.

Specker, J.L., Ingleton, P.M. and Bern, H.A. 1985a.Comparative physiology of the prolactin cell. In:Prolactin Secretion: A multidisciplinary Approach. F. Menaand C. M. Valverde-R. eds., Academic Press, New York.pp. 17-30.

Specker, J.L., King, D.S., Nishioka, R.S., Shirahata, K.,Yamaguchi, K. and Bern, H.A. 1985b. Isolation and partialcharacterization of a pair of prolactins released in vitroby the pituitary of a cichlid fish, Oreochromismossambicus. Proc. Nat. Acad. Sci. USA. 82:7490-7494.

216

Specker, J.L. and Sullivan, C.V. 1994. Vitellogenesis infishes: status and perspectives. In: Perspectives inComparative Endocrinology., K. G. Davey, R. E. Peter andS. S. Tobe, eds., National Research Council of Canada,Ottawa, Canada, pp. 304-315.

Sridaran, R., Ghose, M. and Mahesh V.B. 1988. Inhibitoryeffects of a gonadotropin-releasing hormone agonist on theluteal synthesis of progesterone, estradiol receptors, andprolactin surges during early pregnancy. Endocrinology123:1740-1746.

Stacey, N.E., MacKenzie, D.S., Marchant, T.A., Kyle, A.L.and Peter, R.E. 1984. Endocrine changes during naturalspawning in the white sucker, Catostomus commersoni. I.gonadotropin, growth hormone, and thyroid hormones. Gen.Compo Endocrinol. 56:333-348.

Steel, R.G.D., and Torrie, J.H. 1980. Principles andProcedures of Statistics: A Biometrical Approach. McGraw­Hill, New York.

Sumpter, J.P., LeBail, P.Y., Pickering, A.D., Pottinger,T.G. and Carragher, J.F. 1991a. The effect of starvationon growth and plasma growth hormone concentrations ofrainbow trout, Oncorhynchus mykiss. Gen. CompoEndocrinol. 83:94-102.

Sumpter, J.P., Lincoln, R.F., Bye, V.J., Carragher, J.F.and LeBail, P.Y. 1991b. Plasma growth hormone levelsduring sexual maturation in diploid and triploid rainbowtrout (Oncorhynchus mykiss). Gen. Compo Endocrinol.83:103-110.

Swanson, P. 1991. Physiology of salmon gonadotropins I andII. In: Proceedings of the 3rd MAFF Homeostasis Workshop:Endocrine Control of Reproduction and Early Development ofFish, National Research Institute of Aquaculture, Nansei,Watarai, Mie, JAPAN, November 19-20, 1991, pp. 26-30.

Tan, C.H., Wong, L.Y., Pank, M.K. and Lam, T.J. 1988.Tilapia prolactin stimulates estradiol-17B synthesis invitro in vitellogenic oocytes of the guppy Poeciliareticulata. J. Exp. Zool. 248:361-364.

217

Tashjian, A.H., Bancroft, F.C. and Levine, L. 1970.Production of both prolactin and growth hormone by clonalstrains of rat pituitary tumor cells: Differential effectsof hydrocortisone and tissue extracts. J. Cell. BioI.47:61-70.

Tsien, R.Y., Rink, T.J., and Poenie, M. 1985. Measurementof cytosolic free Ca 2 + in individual small cells usingfluorescence microscopy with dual excitation wavelengths.Cell Calcium. 6:145-157.

Van Der Kraak, G., Rosenblum, P.M. and Peter, R.E. 1990.Growth hormone-dependent potentiation of gonadotropin­stimulated steroid production by ovarian follicles of thegoldfish. Gen. Compo Endocrinol. 79:233-239

Vaughan, J.M., Rivier, J., Spiess, J. Peng, C., Chang,J.P., Peter, R.E. and Vale, W. 1992. Isolation andcharacterization of hypothalamic growth hormone releasingfactor from common carp, Cyprinus carpio. Endocrinology56:539-549.

Wagner, G.F. and McKeown, B.A. 1986. Development of asalmon growth hormone radioimmunoassay. Gen. CompoEndocrinol. 62:452-458.

Wang, X., Sato, N., Greer, M.A., McAdams, S. and Greer,S.E. 1991. Dual effect of osmotic cell swelling onprolactin secretion by acutely dispersed adenohypophysealcells. Life Sci. 48:617-622.

Weber, G. M., Borski, R. J., Powell, J. F. F., Sherwood,N. M. and Grau, E. G., 1994. In vivo and in vitrq effectsof gonadotropin-releasing hormone on prolactin in thetilapia, Oreochromis mossambicus. Amer. Zool. 34:121A.

Weber, G.M., Okimoto, D.K., Richman, N.H.III, and Grau,E.G. 1992. Patterns of thyroxine and triiodothyronine inserum and follicle-bound oocytes of the tilapia,Oreochromis mossambicus, during oogenesis. Gen. CompoEndocrinol. 85:392-404.

218

Weber, G.M., Tom, C. and Grau, E.G. 1989. Effects ofmaternal injections of 3,3',5-triiodo-L-thyronine (T3) onlarval development of the tilapia, Oreochromismossambicus. In: Hormones and the Environment., D. K. O.Chan, ed., University of Hong Kong, U.K., December 18-20,1989. Published by Society for the study of endocrinology,metabolism and reproduction, c/o Dept. of Medicine, Univ.of Hong Kong, Hong Kong, U.K., pp. 103-104.

Wendelaar Bonga, S.E., De Ruiter, A.J.H., Slijkhuis, H.and Dirks, L. 1984. Is parental care in fish prolactindependent? Gen. Compo Endocrinol. 53:460A.

Wigham, T., Nishioka, R.S. and Bern, H.A. 1977. Factorsaffecting in vitro activity of prolactin cells in theeuryhaline teleost Sarotherodon mossambicus (Tilapiamossambica). Gen. Compo Endocrinol. 32:120-131.

Yada, T., Hirano, T. and Grau, E.G. 1994. Changes inplasma levels of the two prolactins and growth hormoneduring adaptation to different salinities in theeuryhaline tilapia, Oreochromis mossambicus. Gen. CompoEndocrinol. 93:214-223.

Yamaguchi, K., Specker, J.L., King, D.S., Yokoo, Y.,Nishioka, R.S., Hirano, T. and Bern, H.A. 1988. Completeamino acid sequence of a pair of fish (tilapia)prolactins, tPRL1 7 7 and tPRI'188' J. BioI. Chern.263:9113-9121.

Yao, K., Niu, P.-D., Le Gac, F. and Le Bail, P.-Y. 1991.Presence of GH specific binding sites in rainbow trout(Oncorhynchus mykiss) tissues: characterization of thehepatic receptor. Gen. Compo Endocrinol. 81:72-82.

Yoshikawa, J.S.M., Borski, R.J. and Grau, E.G. 1995. Theeffects of acclimation salinity and in vitro mediumosmotic pressure on the incorporation of 3H-leucine intothe two prolactins of the tilapia, Oreochromismossambicus. J. Exp. Zool. (In press).

219

Young, G. and Ball, J.N. 1983. Ultrastructural changes inthe adenohypophysis during the ovarian cycle of theviviparous teleost Poecilia latipinna. Gen. CompoEndocrinol. 52:86-101.

Young, G., Veda, H. and Nagahama, Y. 1983. Estradiol-17Band 17a,20B-dihydroxy-4-pregnen-3-one production byisolated ovarian follicles of amago salmon (Oncorhynchusrhodurus) in response to mammalian pituitary and placentalhormones and salmon gonadotropin. Gen. Compo Endocrinol.52:329-335.

Yu, K.L., Peng, C. and Peter, R.E. 1991. Changes in brainlevels of gonadotropin-releasing hormone and serum levelsof gonadotropin and growth hormone in goldfish duringspawning. Can. J. Zool. 69:182-188.

Zohar, Y., Elizur, A., Sherwood., N.M., Powell, J.F.F.,Rivier, J.E., and Zmora, N. 1995. Gonadotropin-releasingpotencies of the three native forms of gonadotropin-re­leasing hormones present in the brain of the gilthead seabream, Sparus aurata. Gen. Compo Endocrinol. (In Press)

Zohar, Y., Goren, A., Fridkin, M., Elhanati, E. and Koch,Y. 1990. Degradation of gonadotropin releasing hormone inthe gilthead seabream, Sparus aurata: II. Cleavage ofnative salmon GnRH, mammalian LHRH and their analogs inthe pituitary, kidney, and liver. Gen. Compo Endocrinol.79:306-319.

220