AN ABSTRACT OF THE THESIS OF
Martin Stone Fitzpatrick
Philosophy in
for the degree of Doctor of
Fisheries Science presented on May 29, 1990
Title:
The Endocrine Regulation of Final Oocyte Maturation and Sex
Differentiation in Salmonids.
Abstract approved:
Redacted for Privacy
Carl B. Schreck
Sexual maturation and sex differentiation comprise facets of a
common theme: reproduction. The endocrine system regulates many
of the critical physiological processes necessary for reproduction and
offers a framework within which technologies can be developed for
controlling sexual maturation and sex differentiation. The studies
described in this thesis were undertaken to improve the
understanding of the endocrine control of these critical stages of
development in salmonids.
Final ovarian maturation in salmon is accompanied by dynamic
changes in plasma hormone levels. Ovulation can be accelerated
through the use of hormones such as gonadotropin releasing hormone
or its analogs (GnRHa). The effectiveness of GnRHa often depends on
the timing of treatment. To determine if plasma concentrations of
steroids can be used to predict the sensitivity of adult female coho
salmon (Oncorhynchus kisutch) to GnRHa, circulating levels of
testosterone, 17a,2013-dihydroxyprogesterone (DHP), and estradiol
were measured before and after injection with GnRHa to accelerate
ovulation. We found that high levels of testosterone were predictive of
early response of coho salmon to GnRHa treatment.
The correlation between testosterone and ovulatory response to
GnRHa suggested a possible functional relation. However,
implantation or injection of testosterone, 17a-methyltestosterone
(MT), or the antiandrogen, cyproterone acetate (CA), before or with
GnRHa treatment did not affect the ovulatory response of coho or
chinook salmon (0. tshawytscha) to GnRHa. Chinook salmon treated
with MT alone had accelerated ovulation in comparison to controls.
If steroids are involved in sex differentiation, steroids must be
produced early in development. In vitro production of steroids in both
coho salmon and rainbow trout (0. mykiss) was assessed from hatch
through sex differentiation. Cortisol, androstenedione, testosterone,
and estradiol were produced just after hatching by tissue explants that
contained anterior kidneys and gonads of coho salmon. To circumvent
the problem of not knowing the sex of individuals until after sex
differentiation, single-sex populations of rainbow trout were produced
by gynogenesis or androgenesis. Tissue explants produced more
androstenedione than testosterone or estradiol. More androgens were
produced by testes and more estradiol was produced by ovaries within
6 to 10 weeks of hatching. Dietary treatment with estradiol or MT
inhibited gonadal steroid secretion.
Electrophoresis of gonadal homogenates from salmonids
revealed several sex-specific bands. In particular, a prominent band of
about 50,000 daltons was apparent in ovaries but not testes.
Production of sex-specific proteins may be affected by dietary steroid
treatment.
Endocrine Regulation of Final Oocyte Maturation
and Sex Differentiation in Salmonids
by
Martin Stone Fitzpatrick
A THESIS
submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Doctor of Philosophy
Completed May 29, 1990
Commencement June 1991
APPROVED:
Redacted for Privacy
Professor of Fisheries in charge of major
Redacted for Privacy
Head of Department of Fisheries and Wildlife
Redacted for Privacy
Dean of G
ate SchooCI
Date thesis is presented
May 29, 1990
Acknowledgements
Many people provided the guidance and inspiration that made
this study possible. My major professor, Dr. Carl Schreck, furnished
me with the wherewithal to initiate and complete my research
through his support and advice in every aspect of my dissertation.
wish to thank him for having confidence in me. Funding for this
I
research came from the Oregon Sea Grant Program, the U.S.
Department of Agriculture, and Sigma Xi, the Scientific Research
Society.
Dr. J. Michael Redding was a key collaborator on many of the
studies and I spent many fruitful hours with Mike discussing science,
society, and departmental gossip. Sam Bradford, Grant Feist, Cameron
Sharpe, Steve Stone, and other members of the Oregon Cooperative
Fishery Research Unit provided their excellent technical assistance
and repartee during many of the studies. Dr. Gary Thorgaard
(Washington State University) gave us monosex populations of trout
which made two of the five studies possible. I had numerous
conversations with Dr. Reynaldo Pathos on many topics, and to
Reynaldo I owe much of the inspiration for the in vitro work.
My major professor is coauthor on Chapters II through VI. In
Chapter II, Dr. Redding and Mr. Frank Ratti are coauthors for their
help in collecting the data. In Chapter III, Dr. Redding collaborated
on the experimental design and data collection; Dr. Penny Swanson
analyzed the plasma samples for GtH I and
II.
Redding and Swanson as coauthors of Chapter
I recognize Drs.
III.
In Chapter IV, Dr.
Redding performed the cortisol analysis and is, therefore, a coauthor.
In Chapter VI, Dr. Redding helped collect the tissue samples; Drs.
Miranda and Buhier trained me to perfoi iii SDS-PAGE analysis,
allowed me to use their facilities, and provided me with advice
throughout the study. All are coauthors on Chapter VI.
Finally, I would like to thank my family for their love and
encouragement throughout my studies. My beloved Wife, Hillary Egna,
gave me much needed spiritual support as well as excellent technical
assistance during all these years. She has helped me through some
very difficult times, as well as shared in times of triumph.
TABLE OF CONTENTS
I.
INTRODUCTION
Background
Goals and Objectives
Organization of the Thesis
II.
IN VITRO STEROID SECRETION DURING EARLY
DEVELOPMENT OF COHO SALMON
Abstract
Introduction
Materials and Methods
Results
Discussion
Acknowledgements
V.
10
11
16
28
33
EFFECTS OF ANDROGENS ON GONADOTROPINRELEASING HORMONE ANALOG INDUCED OVULATION
IN PACIFIC SALMON
34
Abstract
Introduction
Materials and Methods
Results
Discussion
Acknowledgements
IV.
7
PLASMA TESTOSTERONE CONCENTRATION PREDICTS
THE OVULATORY RESPONSE OF COHO SALMON
(ONCORHYNCHUS KISUTCH) TO GONADOTROPINRELEASING HORMONE ANALOG
8
Abstract
9
Introduction
Materials and Methods
Results
Discussion
Acknowledgements
III.
1
6
IN VITRO STEROID SECRETION DURING EARLY
DEVELOPMENT OF RAINBOW TROUT: ONSET OF
PITUITARY CONTROL AND EFFECTS OF DIETARY
STEROID TREATMENT
Abstract
Introduction
Materials and Methods
Results
Discussion
Acknowledgements
35
36
38
44
86
92
94
95
96
97
99
109
111
112
113
114
116
119
138
146
VI.
SEX-SPECIFIC GONADAL PROTEINS IN SALMONIDS
DURING DEVELOPMENT
Abstract
Introduction
Materials and Methods
Results
Discussion
Acknowledgements
VII.
CONCLUSION
Final Maturation
Sex Differentiation
147
148
149
151
154
164
168
169
169
173
VIII. BIBLIOGRAPHY
177
APPENDIX
189
LIST OF FIGURES
Figure
Page
1.
Induced ovulation of coho salmon treated on 16
and 23 October 1984 and on 22 October 1985
with GnRHa or saline.
17
2.
Mean plasma concentrations of estradiol -17J3,
17a,2013-dihydroxy-4-pregnen-3-one (DHP), and
testosterone in female coho salmon on the day of
21
treatment with either gonadotropin-releasing
hormone analog (GnRHa) or saline.
3.
Mean plasma concentrations of estradio1-175,
17a,20f3-dihydroxy-4-pregnen-3-one (DHP), and
testosterone in female coho salmon on the day of
treatment with either gonadotropin-releasing
hormone analog (GnRHa) or saline.
4.
Mean plasma concentrations of estradiol -17J3,
25
17a,20J3-dihydroxy-4-pregnen-3-one (DHP), and
testosterone in female coho salmon before (-7 and
0 days) and after (+2 days) treatment with
gonadotropin-releasing hormone analog (GnRHa).
5.
Experimental design for the 1986 and 1987 coho 40
salmon studies.
6.
Experimental design for the 1987 and 1988
chinook salmon studies.
43
7.
Mean days to ovulation (from Day 0) for coho
salmon during the 1986 study.
45
8.
Induced ovulation in coho salmon in 1986.
47
9.
Mean days to ovulation from Day 0 for coho
salmon during the 1987 study.
50
10.
Induced ovulation in coho salmon in 1987.
52
11.
Mean days to ovulation from Day 0 for chinook
salmon during the 1987 study.
54
12.
Induced ovulation in chinook salmon in 1987.
56
23
13.
Mean days to ovulation from Day 0 for chinook
salmon during the 1988 study.
58
14.
Induced ovulation in chinook salmon in 1988.
60
15.
Mean plasma concentrations of testosterone in
female coho salmon in 1986.
62
16.
Mean plasma concentrations of testosterone in
female coho salmon in 1987.
64
17.
Mean plasma concentrations of testosterone in
female chinook salmon in 1987.
66
18.
Mean plasma concentrations of testosterone in
female chinook salmon on 28 Jul 1988.
68
19.
Mean plasma concentrations of testosterone in
early and late ovulating female coho salmon in
71
Mean plasma concentrations of testosterone in
early and late ovulating female coho salmon in
73
Mean plasma concentrations of testosterone in
early and late ovulating female chinook salmon in
75
22.
Mean plasma concentrations of testosterone in
early and late ovulating female chinook salmon in
1988.
77
23.
Mean plasma concentrations of gonadotropin
(GtH) I in female coho salmon from Group B
in 1986.
80
24.
Plasma concentrations of gonadotropin (GtH) II
in female coho salmon from Group B in 1986.
82
25.
Mean levels of cortisol in media from incubations
of coho salmon gonads and anterior kidneys
treated with partly purified salmon gonadotropin
(SG) or porcine adrenocorticotropin (pACTH).
100
26.
Mean levels of androstenedione in media from
incubations of coho salmon gonads and anterior
kidneys treated with partly purified salmon
102
1986.
20.
1987.
21.
1987.
gonadotropin (SG) or porcine adrenocorticotropin
(pACTH).
27.
Mean levels of testosterone in media from
104
incubations of coho salmon gonads and anterior
kidneys treated with partly purified salmon
gonadotropin (SG) or porcine adrenocorticotropin
(pACTH).
28.
Mean levels of estradiol in media from
106
incubations of coho salmon gonads and anterior
kidneys treated with partly purified salmon
gonadotropin (SG) or porcine adrenocorticotropin
(pACTH).
29.
Mean levels of androstenedione in media from
incubations of rainbow trout tissue explants
(n= 6 to 9 for each treatment).
120
30.
Mean levels of testosterone in media from
incubations of rainbow trout tissue explants
(n= 6 to 9 for each treatment).
123
31.
Mean levels of estradiol in media from
incubations of rainbow trout tissue explants
(n= 6 to 9 for each treatment).
125
32.
Mean levels of androstenedione in media from
incubations of rainbow trout gonads (0 = control
media; G = media containing 100 ng of SG).
127
33.
Mean levels of androstenedione in media from
incubations of rainbow trout anterior kidneys
(0 = control media; G = media containing
130
34.
Mean levels of testosterone in media from
incubations of rainbow trout gonads (0 = control
media; G = media containing 100 ng of SG).
132
35.
Mean levels of estradiol in media from
incubations of rainbow trout gonads (0 = control
media; G = media containing 100 ng of SG).
134
36.
Histology of gonads at various times after
136
fertilization from rainbow trout fed a control diet.
37.
Histology of gonads at various times after
fertilization from rainbow trout fed diets
100 ng of SG).
139
(between 63 and 139 days post-fertilization)
containing 3 4/kg of estradiol
(E2)
of 17a-methyltestosterone (MT).
or 5 4/kg
38.
SDS-PAGE of tissue homogenates from coho
salmon.
155
39.
Native gel electrophoresis of tissue homogenates
from coho salmon.
157
40.
SDS-PAGE of tissue homogenates from rainbow
trout.
160
SDS-PAGE of subcellular fractions from rainbow
162
42.
HPLC chromatogram of methyltestosterone (MT)
in plasma from juvenile coho salmon at 1, 4, and
10 days after treatment with MT microcapsules.
192
43.
Plasma levels of methyltestosterone (MT) in
juvenile coho salmon treated (on Day 0) with
MT microcapsules made with either tripalmitin
or lecithin.
194
44.
HPLC chromatogram of cyproterone acetate (CA)
in plasma from adult female coho salmon at
1 week after treatment with CA microcapsules.
196
41.
trout (17-month old control diet fish).
LIST OF TABLES
Table
Page
I.
Experimental design for 1984 and 1985 female
coho salmon GnRHa studies.
13
II.
Plasma levels of steroids (ng/mL) in female coho
salmon injected with saline on both 22 and 24
October 1985.
20
III.
1986 Coho Study. Viabilities of eggs 11 days after
fertilization.
84
IV.
1986 Coho Study. Gender of offspring from
treated females.
85
ENDOCRINE REGULATION OF FINAL OOCYTE MATURATION
AND SEX DIFFERENTIATION IN SALMONIDS
I: INTRODUCTION
Background
The cultural and economic importance of salmonids to the
Pacific Northwest makes research on the biology of these fishes
especially pertinent. The study of reproduction offers particular
promise for the development of technologies useful in the
preservation and culture of salmonids. The endocrine system
coordinates many of the critical physiological processes that allow for
successful reproduction.
Final ovarian maturation in salmonid fishes, which is regulated
by the endocrine system (Goetz 1983), is accompanied by dynamic
changes in the plasma levels of steroids (Scott et al.. 1983) and
gonadotropin (Van Der Kraak et at. 1983). Specifically, circulating
levels of both gonadotropin and 17a,206-dihydroxy-4-pregnen-3-one
(17a-hydroxy-20J3-dihydroprogesterone or DHP) rise and the level of
estradio1-17J3 decreases (Van Der Kraak et at. 1983, 1985).
Gonadotropin, secreted by the pituitary, induces the release of DHP in
the ovary, and DHP, in turn, stimulates the final maturation of oocytes
(Jalabert 1976). In salmonids, two distinct gonadotropins (GtH I and
GtH II) have been recently isolated that have similar steroidogenic
capacities (Suzuki et at. 1988a), but appear to differ in synthesis and
secretion during development (Suzuki et al. 1988b; Swanson et al.
2
1989). GtH II is generally regarded as the "maturational"
gonadotropin and GtH I has predominance during the juvenile stage.
A gonadotropin-releasing hormone (GnRH) has been isolated and
characterized from the hypothalamus of chum salmon, Oncorhynchus
keta (Sherwood et al. 1983). Synthetic analogs of GnRH (GnRHa) have
been used to successfully accelerate final maturation in coho salmon,
0. kisutch (Van Der Kraak et a/. 1983, 1985; Fitzpatrick et a/. 1984),
Atlantic salmon, Salmo salar (Crim et at. 1983a; Crim and Glebe 1984),
and rainbow trout, 0. mykiss (Crim et at. 1983b). Treatment of female
coho salmon with GnRHa elevated plasma gonadotropin (Van Der
Kraak et al. 1983) and DHP (Van Der Kraak et at. 1984). In female
coho salmon and steelhead trout, 0. mykiss, plasma estradiol was
depressed and plasma androgens were elevated after treatment with
serial injections of salmon gonadotropin and GnRHa (Sower et al.
1984). These studies demonstrated that specific changes occur in the
endocrine system after treatment with maturation inducing hormones
such as GnRHa. Emphasis has usually been placed on subsets of fish
within the experimental populations that ovulate soon after treatment.
Less well studied are the fish that do not respond quickly to
treatment. A large percentage of female Atlantic salmon did not
ovulate when given GnRHa 45 days before the mean spawning date of
control fish. Furthermore, these non-responsive fish did not show the
characteristic elevation of plasma gonadotropin seen in responsive fish
(Crim and Glebe 1984). Similarly, in coho salmon, plasma steroid
profiles in females that did not ovulate within 2 wk after GnRHa
injection differed from those in females that ovulated within 2 wk
(Van Der Kraak et al. 1985). Thus, maturational status, as reflected by
3
the endocrine system, may determine the sensitivity of the female to
GnRHa.
Although typically considered a male hormone, testosterone
reaches very high levels in the plasma of female coho salmon during
final maturation (Fitzpatrick et al. 1986)--in fact, the levels are higher
in females than in males. It is not known what, if any, function the
high levels of testosterone have in final ovarian maturation.
Testosterone enhances the effectiveness of gonadotropin on the in
vitro maturation of oocytes in amago salmon, 0. rhodurus (Young et al.
1982; Nagahama et at. 1983). In vivo pretreatment of juvenile rainbow
trout, 0. mykiss, with testosterone resulted in pituitary accumulation
of gonadotropin which was released in vitro after stimulation with
GnRHa (Crim and Evans 1980). Methyltestosterone given in
conjunction with GnRHa to milkfish, Chanos chanos, significantly
stimulated maturation compared to controls or fish given GnRHa alone
(Lee et al. 1986a,b). The high levels of testosterone found in females
near final maturation have been postulated to be responsible for
stimulating gonadotropin synthesis (Fitzpatrick et al. 1987; Kobayashi
et a/. 1989a,b), and Zohar (1989) suggested that a failure in GtH
release underlies the lack or delay of final oocyte maturation observed
in some fish.
The presence or absence of the Y chromosome may determine
the sex in some vertebrates, but the sexual phenotype (which includes
the gonads and the accessory sex organs) results from a developmental
cascade of biochemical events that involves more than just the Y
chromosome. In mammals, the development of the sex of an
4
individual requires a translation of genetic sex into gonadal sex. In
some species of reptiles and fishes that rely on temperature-
dependent sex determination, genetic sex may have little meaning
(Crews and Bull 1987); nevertheless, the actual differentiation of the
gonad likely requires the activation of genes. Steroids can direct
gonadal development to the phenotype opposite that of the genotype
in many non-mammalian vertebrates, which led Yamamoto (1969) to
postulate that steroids may be the natural inducers of gonadal sex
differentiation in fishes. However, steroids cannot completely invert
the sex of most mammalian gonads (Jost 1965; Wilson et al. 1980) and
recent work has focused on single gene differences between males
and females (Page et a/. 1987). In the many teleosts in which steroidtreatment does cause sex-inversion (see Hunter and Donaldson 1983),
it is unclear if the effect has pharmacological or truly physiological
causes.
Rainbow trout (Oncorhynchus mykiss) embryos accumulated
radioactive pregnenolone, progesterone, dehydroepiandrostenedione,
and estradiol, although only pregnenolone and progesterone were
metabolized significantly (Anti la 1984). Gonadal homogenates from
rainbow trout synthesized progestins by 50 days after fertilization (21
days after hatching), androgens by 100 days, and estrogens by 200
days (van den Hurk et al. 1982). The latter studies used incubations of
homogenized tissues in the presence of radioactive precursors--a
technique which attends to the question of synthetic potential but
does not address the question of natural production. Any sex
differences between males and females in the capacity to synthesize
steroids may be of degree rather than kind -in other words, both
5
males and females may be able to make any particular steroid, but
profound differences may exist in the amount produced. Furthermore,
gonadal steroids may not be the prime factors influencing sex
differentiation. Van den Hurk and van Oordt (1985) suggest that the
interrenal may direct masculine sex differentiation through the
production of androstenedione derivatives.
The biochemical events that comprise sex differentiation remain
poorly understood. The nature of the inducing substance(s) has and
continues to be a rich area of inquiry. Witschi (1934) hypothesized
the existence of hormone-like inducing substances responsible for
setting off the chain of events leading to female and male gonadal
development. Although some investigators postulated that sex
steroids were the endogenous sex inducers (Burns 1961; Yamamoto
1969), Witschi (1967) argued that the substances were proteins. The
discovery and further study of H-Y antigen--a protein seemingly
invariably associated with the heterogametic sex of many vertebrates-
offered hope that the inducer substance had been found (Wachtel and
Koo 1981). However, sexual differences in H-Y antigen were lacking
in several species of teleosts (Muller and Wolf 1979) and the antigen
was absent in a laboratory strain of testes-bearing rats (McLaren et al.
1984). Recently, the H-Y gene has been mapped to a location outside
that of the purported sex determining region of the Y chromosome
(Simpson et al. 1987). Such evidence has made the H-Y antigen
theory problematic--indeed, it seems to preclude H-Y antigen from
being the primary inducer.
6
The identification of the testes-determining region of the Y
chromosome (ZFY--for "Zinc Finger protein on Y) in several
mammalian species has again raised hopes that an inducer substance
may shortly be found (Page et at. 1987). A complex phenomenon such
as differentiation likely involves the differential production of proteins
throughout the developmental cascade. Therefore, the study of
steroidal regulation of protein production may provide important
insights into the process of sex differentiation.
Goals and Objectives
The goal of my thesis is to expand the understanding of the
endocrine regulation of final sexual maturation and sex differentiation
in salmonid fishes. The objectives of the individual studies were:
1) to determine if plasma concentrations of steroids can be used
to predict the ovulatory response of coho salmon to GnRHa;
2) to determine if androgens or antiandrogens affect the
induction of ovulation by GnRHa;
3) to determine if sexual dimorphism exists in steroid
production during early development of coho salmon;
4) to determine if sexual dimorphism exists in steroid
production during early development of rainbow trout and if
exogenous steroid treatment affects steroid production; and
5) to determine the existence and intracellular location of sexspecific gonadal proteins in salmonids and the effects of
exogenous steroid treatment on sex-specific protein
production.
7
Organization of the Thesis
The thesis is organized into seven chapters with an appendix.
Chapters II through VI were prepared in manuscript form with each
chapter consisting of an abstract, introduction, materials and methods
section, results section, and discussion. Chapter VII is an overall
summary and conclusion of the work.
8
IL PLASMA TESTOSTERONE CONCENTRATION PREDICTS THE
OVULATORY RESPONSE OF COHO SALMON (ONCORHYNCHUS
KISUTCH) TO GONADOTROPIN-RELEASING HORMONE ANALOG1
Martin S. Fitzpatrick
J. Michael Redding2
Frank D. Ratti3
and Carl B. Schreck
Oregon Cooperative Fishery Research Unit4
Oregon State University, Corvallis, OR 97331, USA
10regon State University, Agricultural Experiment Station Technical
Paper No. 8006.
2present address: Department of Biology, Tennessee Technological
University, Cookeville, TN.
3Oregon Aqua Foods, Inc., 8870 Marco la Road, Springfield, OR,
97477, USA
4Cooperators are Oregon State University, Oregon Department of Fish
and Wildlife, and U.S. Fish and Wildlife Service.
9
Abstract
Plasma concentrations of testosterone, 17a,2013-dihydroxy-4pregnen-3-one (DHP), and estradio1-1713 were determined in hatchery
populations of coho salmon (Oncorhynchus kisutch) before and after
the fish were injected with gonadotropin-releasing hormone analog
(GnRHa) to accelerate ovulation. Treatment with GnRHa significantly
accelerated maturation over that observed in saline-injected controls;
however, a number of females that received GnRHa did not ovulate for
as long as 8 weeks after treatment. Before treatment, the mean
plasma levels of DHP did not differ between early- and late-ovulating
females; estradiol differed sometimes, but not always; and testosterone
was always higher in the early ovulating fish. Thus, plasma
concentrations of testosterone appear to be useful for predicting the
sensitivity of coho salmon to GnRHa treatment for induced ovulation.
10
Introduction
Final ovarian maturation in salmonid fishes, which is regulated
by the endocrine system (Goetz 1983), is accompanied by dynamic
changes in the plasma levels of steroids (Scott et a/. 1983) and
gonadotropin (Van Der Kraak et at. 1983). Specifically, circulating
levels of both gonadotropin and 17a,205-dihydroxy-4-pregnen-3-one
(17a,20)3-dihydroxyprogesterone or DHP) rise and the level of
estradiol -17J3 decreases (Van Der Kraak et al. 1983, 1985).
Gonadotropin, secreted by the pituitary, induces the release of DHP in
the ovary, and DHP, in turn, stimulates the final maturation of oocytes
(Jalabert 1976).
A gonadotropin-releasing hormone (GnRH) has been isolated and
characterized from the hypothalamus of chum salmon, Oncorhynchus
keta (Sherwood et at. 1983). Synthetic analogs of GnRH (GnRHa) have
been used successfully to accelerate final maturation in coho salmon,
0. kisutch (Van Der Kraak et at. 1983, 1985; Fitzpatrick et al. 1984),
Atlantic salmon, Salmo salar (Crim et al. 1983a; Crim and Glebe 1984),
and rainbow trout, 0. mykiss (Crim et at. 1983b). Treatment of female
coho salmon with GnRHa elevated plasma gonadotropin (Van Der
Kraak et al. 1983) and DHP (Van Der Kraak et al. 1984). In female
coho salmon and steelhead trout, 0. mykiss, plasma estradiol was
depressed and plasma androgens were elevated after treatment with
serial injections of salmon gonadotropin and GnRHa (Sower et al.
1984). These studies demonstrated that specific changes occur in the
endocrine system after treatment with maturation inducing hormones
11
such as GnRHa. Emphasis has usually been placed on subsets of fish
within the experimental populations that ovulate soon after treatment.
Less well studied are the fish that do not respond quickly to
treatment. A large percentage of female Atlantic salmon did not
ovulate when given GnRHa 45 d before the mean spawning date of
control fish. Furthermore, these non-responsive fish did not show the
characteristic elevation of plasma gonadotropin seen in responsive fish
(Crim and Glebe 1984). Similarly, in coho salmon, plasma steroid
profiles in females that did not ovulate within 2 wk after GnRHa
injection differed from those in females that ovulated within 2 wk
(Van Der Kraak et at. 1985). Thus, maturational status, as reflected by
the endocrine system, may determine the sensitivity of the female to
GnRHa.
Plasma concentrations of estradiol, DHP, and testosterone are
highly correlated to the developmental stages of the ovaries of coho
salmon during final maturation (Fitzpatrick et at. 1986). The objective
of the present study, conducted at the freshwater hatchery of Oregon
Aqua Foods, Inc., Springfield, Oregon, during fall 1984 and fall 1985,
was to determine if the plasma concentrations of certain steroids
could be used to predict the ovulatory response of coho salmon to
GnRHa.
Materials and Methods
Experimental Animals. On 16 October 1984 (T0), 200 females
were sorted from the hatchery brood stock. Of these, 100 were from
a population designated "mature," which had been at the hatchery
12
longer than any other group of fish (having returned earlier from the
ocean). The coloration of all members of this group indicated
advanced sexual maturity. The other 100 fish were from a group
designated "less-mature," which had been at the hatchery for the
shortest time. Coloration of members of this group differed from that
of fish of the "mature" group. On T0, 50 "mature" and 50 "less-mature"
fish were tagged, sampled, and treated. On 23 October 1984 (T0'), the
remaining untreated females (50 each) from the original "mature" and
"less-mature" groups were tagged, sampled, and treated.
Eleven females from the "mature" group died during anesthesia
on 23 October and to make up for this loss, twenty females from the
original raceway that contained the "mature" group of fish were then
injected on 24 October. As controls for the injection process, 26
uninjected females from the original "mature" group raceway were
placed with the experimental animals.
On 15 October 1985 (T_7), 66 females were isolated from the
hatchery brood stock; 44 were selected from a group of fish that had
returned earlier than any other group and were therefore expected to
mature before those of any other group, and the remaining 22 females
came from another raceway containing animals that were not expected
to mature as soon as the first group.
Sampling and Treatment. In 1984, we sampled blood from each
female on the day of first injection with either GnRHa (des-Glylo[DAla6]1.H-RH-ethylamide; Syndel Biochemicals) or saline (To or To') (see
Table I). The females received a second injection of either GnRHa or
saline 3 days later. In 1985, we sampled blood from each female 1 wk
13
Table I. Experimental design for 1984 and 1985 female coho salmon GnRHa
studies. Key: b = blood sampled; i = GnRHa or saline injected; c = ovulation
checked.
1984
To
n
50 mature
50 less-mature
50 mature
50 less-mature
16 Oct.
T'o
19 Oct.
23 Oct.
26 Oct.
30 Oct.
b, i
b, i
11 Dec.
c
c
c
b, i
b, i
1985
T_7
To
T2
n
15 Oct.
22 Oct.
24 Oct.
66
b
b, i
b, i
31 Oct.
3 Dec.
14
before first injection with GnRHa or saline (T_7), on the day of
injection (To), and on the day of second injection, 2 d later (T2).
We anesthetized females with ethyl p-aminobenzoate
(benzocaine), tagged them individually, and collected a blood sample
from the caudal vein with a Vacutainer (Becton-Dickinson) coated with
lithium heparin. On treatment days, the fish were injected
intraperitoneally either with 5 lig of GnRHa--about 2.014/kg body
weight--or with a saline solution, as described by Fitzpatrick et al.
(1984). All fish were checked manually for ovulation every 3 to 4 d in
1984 and 4 to 6 d in 1985 for the first month after injection and every
week thereafter. Females were considered to have ovulated if eggs
were extruded when gentle pressure was applied to the abdomen.
Ovulated fish were removed from the experimental population. In
1984, the experiment lasted until all the females had ovulated (48 and
49 d after To and T0', respectively). In 1985, females that ovulated on
day 45 or thereafter (designated 45+) were grouped together for
analysis.
In 1984, a total of 18 females died before ovulation, but they
were still used to calculate the cumulative percentages of fish that had
ovulated in each treatment group. Nine other fish divided among four
different treatment groups, which were mistakenly killed before
complete ovulation, were not considered in the analysis because it was
impossible to determine the date of ovulation. Mortality was equal in
all treatment groups.
In 1985, a total of four fish either died during the course of the
experiment or lost their tags and were not considered in the analysis
except when we determined the cumulative percentage of ovulation in
15
the treatment group. Two fish killed before complete ovulation were
not considered in the analysis. When a single fish constituted the total
number of females ovulating on a particular date, it was not considered
in the statistical analysis of plasma steroid content. Because we were
interested in the value of circulating steroids in predicting the
ovulatory response after injection of GnRHa, we combined data on
steroid levels from all GnRHa-treated animals for statistical analyses.
Three females treated with GnRHa in 1985 had high levels of DHP on
the day of first injection--suggesting that they were very close to
ovulation before treatment (all ovulated by day 9); these fish were not
used in the statistical analysis of plasma steroid content.
Assays. The blood samples were held on ice until centrifugation,
after which the plasma was drawn off and stored frozen (-20°C) until
analysis. Plasma samples were analyzed for DHP, estradiol -17j3, and
testosterone by radioimmunoassay using the procedure of Fitzpatrick
et al. (1986) with the exception that a single aliquot of 50 1.1.14 of plasma
was extracted with diethyl ether, the extract was reconstituted in 1
mL of phosphate-buffered saline, and then portions of reconstituted
extract were assayed for each steroid.
Statistics. Data for time elapsed from the first injection to
ovulation between GnRHa- and saline-treated groups were analyzed by
a Mann-Whitney U test. The plasma steroid data for individuals
through the 3 sampling dates in 1985 were analyzed initially by
Friedman's analysis of variance for dependent samples. Data showing
significant variation were then subjected to Wilcoxon's sign test to
determine significant differences between the sampling dates. Groups
were established based on the number of days elapsed to ovulation;
16
then steroid data were compared between groups for any particular
sampling date and treatment by Kruskal-Wallis analysis of variance for
independent samples. Groups demonstrating significant variation
were then analyzed with a Mann-Whitney U test. Definitive statistical
significance was set at p<0.05.
Results
Inducement of Ovulation
Fish of the "mature" group that received GnRHa on 16 October
1984 (T0) ovulated in significantly fewer days than did saline-injected
controls, but this effect was not apparent in the "less-mature" groups
(Fig. 1). Fish treated with GnRHa on 23 October 1984 (To') from both
the "mature" and "less-mature" groups ovulated sooner than their
respective saline-injected controls; likewise, "mature" fish injected
with saline on To' ovulated sooner than uninjected "mature" controls.
In 1985, females that received GnRHa ovulated in significantly
fewer days than did saline-injected control fish (Fig. 1)--21.1 + 2.4 vs.
36.1 + 3.3 (mean ± SE).
Analysis of Steroid Concentrations
Plasma concentrations of hormones on To and To' in 1984 were
analyzed as a function of the number of days elapsed to ovulation after
treatment. Data were combined for fish that received the same
treatment on a given date regardless of maturity status. The 1985 data
for the plasma steroid levels were analyzed in two ways--in both of
which the fish were grouped by the number of days elapsed to
ovulation. The first analysis compared values temporally in fish from
17
Figure 1. Induced ovulation of coho salmon treated on 16 and 23
October 1984 and on 22 October 1985 with GnRHa or saline. The
numbers of fish that ovulated after being injected with GnRHa or
saline, or were not injected, are expressed as cumulative percentages
of their respective groups.
18
100
16 October 1984
-------°
80
60
40
0 Mature -GnRHa
20
0 Less mature -GnRHa
Mature-Saline
--0-- Less-mature-Saline
0
0
100
10
20
30
40
50
60
23 October 1984
--a
80
60
40
a Mature-GnRHa
20
--A-- Mature-Uninjected
Mature-Saline
0 Less-mature-GnRHa
4-- Less-mature-Saline
0
0
100
10
20
30
40
50
60
22 October 1985
80
60
40
20
0 GnRHa
0-- Saline
0
10
20
30
40
Days after First Injection
Figure 1.
50
60
19
any particular group (ovulating 9, 14, or 45+ d after treatment)
between the three sampling dates (T_7, To, and T2); the second
compared the values between groups on a given sampling date. For
example, first, the levels of DHP in all females that ovulated 9 d after
GnRHa injection were compared between the various sampling dates
(T_7, To, and T2); second, DHP in females that ovulated after 9 d were
compared to that of females that ovulated after 14 and 45+ d on a
given sampling date (T_7, To, and T2).
The numbers of control females that ovulated on either 9, 14, or
30 d after saline injection were too small to allow a rigorous statistical
comparison of plasma hormone levels over the sampling period;
nonetheless, trends were apparent (Table II) as described below.
17a, 2016-Dihydroxy-4-pregnen-3-one
In 1984, the levels of DHP on To or To' did not vary as a function
of days elapsed to ovulation after treatment with either GnRHa or
saline (Figs. 2 and 3).
In 1985, no significant difference in the concentrations of DHP
on T_7 was detected between groups injected with. saline, but on To
DHP was higher in fish that ovulated after 14 d than in those that
ovulated after 45+ d. On T2, DHP tended (p<0.10) to be higher in
saline-injected fish that ovulated after 14 d than after 30 or 45+ d
(Table II).
In GnRHa-treated females, plasma DHP was lower on T_7 than on
To or T2 in all groups, and lower on To than on T2 (Fig. 4). On both T_7
and To, there was no significant difference in plasma DHP between
groups, but on T2 DHP was higher in fish that ovulated after 9 or 14 d
than after 45+ d (Fig. 4).
Table II. Plasma levels of steroids (ng/mL) in female coho salmon injected with saline on both 22 and 24
October 1985. Days to ovulation indicates the time elapsed from 22 October to ovulation.
Days to
ovulation
17a,20B-Dihydroxy4-pregnen-3-one
n
9
1
14
3
30
45+
2
12
T.7
4
3 ± 0.4
2 ± 0.5
2 ± 0.1
To
350
7 ± 0.6
3 ± 0.3
3 + 0.2
T2
685
13 ±6.0
4±0.5
3 ±0.2
Testosterone
Estradio1-176
T_7
To
T2
59
4
3
T_7
486
To
T2
736
448
54±9 28±12 19±12 398±76 442±76 545±131
77±26 50±14 38± 18 261 ±12 329±40 340±52
70-3
60 + 4
39 +3 242 + 16 238 + 14
199 + 23
21
Figure 2. Mean plasma concentrations of estradio1-173, 17a,2013dihydroxy-4-pregnen-3-one (DHP), and testosterone in female coho
salmon on the day of treatment with either gonadotropin-releasing
hormone analog (GnRHa) or saline. Each bar and vertical line
represents the mean and standard error for the indicated number of
females in groups differentiated by the number of days elapsed to
ovulation from 16 October 1984 (T0). Means of groups identified by
days to ovulation which share the same horizontal line are not
significantly different (note: the minimum sample size matrices
required by the Mann-Whitney test are 2 x 8, 3 x 5, or 4 x 4).
22
GnRHa
Saline
80 -
80 -
60 -
40 -
20 -
0
0
10 -
10
8E
rn
6
0
4-
4-
_=_
Z.
20
400 -
400 10
17 14 21 34 48 41
17 28 21 24 48 34 41
300 -
300 -
200 -
200 -
100
100
13
10
14
17
21
34
41
48
17
21
24
Days elapsed to ovulation from To
Figure 2.
28
34
6
3
41
48
23
Figure 3. Mean plasma concentrations of estradiol -17j3, 17a,205dihydroxy-4-pregnen-3-one (DHP), and testosterone in female coho
salmon on the day of treatment with either gonadotropin-releasing
hormone analog (GnRHa) or saline. Each bar and vertical line
represents the mean and standard error for the indicated number of
females in groups differentiated by the number of days elapsed to
ovulation from 23 October 1984 (T0'). Means of groups identified by
days to ovulation which share the same horizontal line are not
significantly different (note: the minimum sample size matrices
required by the Mann-Whitney test are 2 x 8, 3 x 5, or 4 x 4).
24
GnRHa
0
.
150
80
-
60
-
40
-
20
-
o
20
-
10
-
Saline
14 10 17 49 21 27 34 41
110
E
13)
c
70 -
0-
_T__
5.
0
-
,
0
500
-
400
-
300
300
-
200
200
-
100
-
500
7
400
100
.1_
S.
10
14 17 21 34 49
C-
-r
14
17 27 49 34 41
10 21
_T__
_T_
_s_
16
0
2
5
2
10
14
17
14
17
21
34
49
Days elapsed to ovulation from To
Figure 3.
6
7
9
9
4
27
34
41
49
0
10
21
25
Figure 4. Mean plasma concentrations of estradiol -17J3, 17a,2053dihydroxy-4-pregnen-3-one (DHP), and testosterone in female coho
salmon before (-7 and 0 days) and after (+2 days) treatment with
gonadotropin-releasing hormone analog (GnRHa). Each value
represents the mean and standard error from females (n = 11-13) in
groups differentiated by the number of days elapsed to ovulation from
22 October 1985 (0 day), i.e., 9, 14, and 45+ days after first treatment.
Means marked by an asterisk are significantly different from others on
that particular day.
26
200
Days elapsed to
ovulation
9
El
E
-6)
c
14
El 45+
100
tL
0
1...1172.2.Throonwwww
0
-7 days
treatment day
Time relative to treatment
Figure 4.
+2 days
27
Estradio1-1713
In 1984, the levels of estradiol on To did not vary as a function of
days elapsed to ovulation after treatment with either GnRHa or saline
(Fig. 2). Plasma levels of estradiol in fish treated with GnRHa on To'
did not differ between days elapsed to ovulation; however, in saline-
injected fish, plasma estradiol on To' was lower in fish that ovulated on
day 10 than on days 34 or 41, and in fish that ovulated on day 14 than
on days 21, 27, 34, or 41 (Fig. 3).
In 1985, there were no significant differences in plasma
estradiol on T_7, To, or T2 between groups that ovulated 14, 30, or 45+
d after saline injection (Table II). In females that ovulated by either 9
or 14 d after the first injection of GnRHa, estradiol levels were higher
on T_7 than on To or T2, and higher on To than on T2 (Fig. 4). No
significant temporal changes were observed in estradiol in fish that
ovulated 45+d after administration of GnRHa.
Plasma estradiol on T_7 was not significantly different between
fish that ovulated after 9, 14, or 45+ d but on To it was lower in fish
that ovulated after 9 d than in those that ovulated after 14 or 45+ d
(Fig. 4). On T2, estradiol was lower in fish ovulating after 9 or 14 d
than after 45+ d.
Testosterone
On To in 1984, testosterone was higher in saline-treated females
that ovulated 17 d after injection than in fish that ovulated 24, 34, or
41 d after treatment, and in fish that ovulated on day 28 than in fish
that ovulated on days 34, 41, or 48. Testosterone was higher in
GnRHa-treated females that ovulated on day 10 than on days 34, 41, or
28
48, and was higher in females that ovulated on day 17 than on day 41
(Fig. 2).
Testosterone on To' was higher in saline-injected fish that
ovulated on day 10 than on days 34 or 41; higher in fish that ovulated
on day 14 than on days 34, 41, or 49; and higher in fish that ovulated
on day 21 than on days 34 or 41. Plasma testosterone was higher in
GnRHa-treated females that ovulated on day 7 than on days 21 or 34,
and in fish that ovulated on day 10 than on days 34 or 49 (Fig. 3).
In 1985, testosterone was similar in all saline-injected groups
on T_7, but on To it was higher in fish that ovulated after 14 d than
after 45+ d. On T2, testosterone levels tended (p<0.10) to be higher
in fish ovulating after 14 d than after 30 or 45+ d (Table II).
In all groups treated with GnRHa, plasma testosterone was
significantly higher on T2 than on T_7 or To (Fig. 4). Plasma
testosterone was similar in all groups on T_7, although it tended to be
higher (p<0.12) in fish that ovulated after 9 or 14 d than after 45+ d.
On To and T2 testosterone was higher in fish ovulating after 9 or 14 d
than after 45+ d (Fig. 4).
Discussion
Elevated concentrations of testosterone in plasma were strongly
correlated with early ovulation in both control and GnRHa-injected
female coho salmon in both 1984 and 1985. Estradiol and DHP
concentrations were not consistently related to ovulatory timing,
although temporal changes in their concentrations during the
sampling period in 1985 seemed to reflect ovulatory responsiveness.
29
Information about the hormonal status of female salmon could be
useful in predicting ovulatory timing in hatchery populations.
Coloration and historical dates of maturation are now the criteria
typically used in hatcheries to predict when females will mature.
Inasmuch as changes in the environment may lead to year-to-year
variations in the timing of spawning runs and in coloration, these
selection criteria may not be adequate. Final maturation and ovulation
are under endocrine control. Oocyte maturation has been correlated
to increased plasma DHP and testosterone and to decreased estradiol
(Scott et at. 1983; Van Der Kraak et al. 1984; Fitzpatrick et al. 1986).
Information on hormonal status may be especially important in
situations where ovulation is to be induced by GnRH and its analogs, a
technique that is often but not always successful (Fig.1; Van Der Kraak
et al. 1983, 1985; Crim and Glebe 1984; Fitzpatrick et at. 1984).
This variable success suggests that there is a limit to the
effectiveness of GnRHa in accelerating ovulation and this limit to
effectiveness may be related to the length of time that treatment
occurs before natural ovulation. Crim and Glebe (1984) demonstrated
that in Atlantic salmon treated with a pelleted GnRH analog, only 30%
of females ovulated when implanted 45 days before the spawning time
of control fish, but 94% ovulated when the implantation was made 28
days before that spawning time. In the present study in 1984, 29% of
the fish injected on 16 October and 52% of those injected on 23
October ovulated within 2 wk. Thus, delaying GnRHa treatment 1 wk
increased the percentage of females that responded within 2 wk of
treatment.
30
The value of plasma estradiol concentration as a predictive index
was questionable: in 1984 the concentration varied as a function of
ovulatory response in only one of four treatment regimes (salineinjected, 23 October; Fig. 3); and in 1985, neither at 1 wk before nor
on the day of treatment did estradiol differ significantly between fish
that ovulated after 14 d and those that ovulated after 45+ d.
Therefore, only in females ovulating 9 d after injection of GnRHa did
information about estradiol levels on the day of treatment provide a
clue as to how quickly the fish would ovulate after treatment. In this
study, low estradiol suggested that the animals would ovulate soon
after treatment, but higher estradiol levels meant only that ovulation
would occur at some time between 14 and 45 d or more after
treatment. Plasma estradiol declined between 15 and 22 October
1985 in saline- and GnRHa-injected fish--significantly in fish that
ovulated within 14 days after treatment with GnRHa. Thus, although
temporal changes in estradiol may have some predictive value, the
predictive utility is lessened due to the requirement of multiple
sampling.
DHP is of little use in predicting how the animals will respond to
treatment. In salmonids, administration of gonadotropin in vivo raises
plasma concentrations of DHP (Scott et at. 1982; Wright and Hunt
1982; Van Der Kraak et al. 1985), and in vitro incubation of ovaries
with gonadotropin increases follicular production of DHP (Young et al.
1982, 1983; Jalabert and Fostier 1984). It has been demonstrated
that various analogs of GnRH increase the plasma levels of
gonadotropin in coho salmon (Van Der Kraak et al. 1983, 1985),
Atlantic salmon (Crim et at. 1983a; Crim and Glebe 1984), and
31
rainbow trout (Crim et al. 1983b). In the present study, plasma
concentrations of DHP did not differ between early- and late-ovulating
groups before treatment; however, by 2 d after the initial injection in
1985, levels of DHP were higher in fish that would ovulate by 9 or 14 d
than in those that required 45 d or longer. This finding is consistent
with those of Van Der Kraak et al. (1985) in coho salmon. Increases in
DHP levels between 15 and 24 October 1985 in all groups prevented
clear discrimination based on temporal changes.
In both 1984 and 1985, data on plasma testosterone
concentration on the day of treatment with saline or GnRHa allowed
clear discrimination between early- or late-ovulating fish. This
relation tended (p<0.12) to hold for samples taken 1 wk before
treatment with GnRHa in 1985. Testosterone levels were also higher
in early- than in late-ovulating fish 2 d after treatment. It is
impossible to specify a particular plasma concentration of testosterone
that will guarantee successful response to GnRHa. However, of the 49
GnRHa-treated females with plasma testosterone levels on the
treatment date above 200 ng/mL (1984 and 1985 data combined),
81.6% ovulated within 2 wk of treatment. Conversely, of the 46
GnRHa-treated females with testosterone levels less than 150 ng/mL,
only 19.6% ovulated within 2 wk. Thus, high testosterone levels
indicate the likelihood of a successful response (i.e. ovulation within 2
wk), whereas lower levels indicate minimal response. Caution should
be exercised when extending these results to other situations where
genetic and environmental factors differ from those in the present
study.
32
The function of testosterone in adult female salmon is unclear,
although the predictive value of testosterone for responsiveness to
GnRHa suggests a functional relation. Androgens, including
testosterone, enhance the synthesis and storage of gonadotropin in
the pituitaries of juvenile salmonids (Crim and Evans 1979; Gielen et
a/. 1982; van den Hurk et al. 1984). Crim and Evans (1980)
demonstrated that in vivo pretreatment of immature rainbow trout
with testosterone resulted in gonadotropin accumulation in the
pituitary gland which was readily released in vitro upon stimulation
with GnRH. Testosterone has also been shown to enhance the effect
of gonadotropin on the in vitro maturation of oocytes in amago salmon,
0. rhodurus (Nagahama et al. 1980; Young et al. 1982).
In view of the present results, it is plausible that the increased
levels of testosterone in early-ovulating fish resulted in increased
synthesis of gonadotropin by the pituitary. Upon stimulation by
GnRHa, gonadotropin was probably released, causing increased
production of DHP by the ovarian follicles. The decreasing output of
estradiol may have permitted this action, as it has been shown that
high levels of estradiol decrease the responsiveness of ovarian follicles
of rainbow trout to gonadotropin (Jalabert and Fostier 1984).
Testosterone is aromatized to estradiol -17J3 in the granulosa layer of
the ovarian follicles of amago salmon (Kagawa et al. 1982). It is not
known what effect testosterone has on aromatization activity during
this stage of maturation. Further investigation is required to
determine if the increase in plasma testosterone levels is a cause or an
effect of the decrease in circulating estradiol.
33
The results of our study demonstrate the effectiveness of GnRHa
in accelerating maturation in coho salmon. These findings are
consistent with those in previous studies (Van Der Kraak et al. 1983,
1985; Fitzpatrick et at. 1984). Although GnRHa represents an
effective tool in the management of hatchery brood stock populations,
it does not accelerate maturation in all treated females in a given
hatchery population. Therefore, to improve the efficiency of GnRHa
treatment, the fish should be screened before treatment to select the
animals most likely to respond or, alternatively, treatment should be
postponed until most of the animals can be expected to respond
favorably. Measurement of the circulating levels of testosterone
appears to be a promising method for the screening of female coho
salmon.
Acknowledgments
We thank the staff of Oregon Aqua Foods, Inc. for their technical
assistance. Sam Bradford, Hillary Egna, and Alec Mau le provided
excellent technical help during the sampling. This study was
supported by the Oregon State University College Sea Grant Program
and by a Grant-in-Aid of Research from Sigma Xi, The Scientific
Research Society, to one of us (MSF).
34
III. EFFECTS OF ANDROGENS ON GONADOTROPIN-RELEASING
HORMONE ANALOG INDUCED OVULATION
IN PACIFIC SALMON
Martin S. Fitzpatrick'
J. Michael Redding''
Penny Swanson2
Carl B. Schreckl
'Oregon Cooperative Fishery Research Unit
Oregon State University, Corvallis OR 97331
2School of Fisheries
University of Washington, Seattle WA 98195
*Current address: Department of Biology, Tennessee Technological
University, Cookeville, TN.
35
Abstract
Plasma testosterone levels in female Pacific salmon increase
during the period of final maturation; however, the function of
elevated testosterone levels remain unknown. Female coho
(Oncorhynchus kisutch) and chinook (0. tshawytscha) salmon were
implanted or injected with testosterone, 17a-methyltestosterone
(MT), or the androgen inhibitor, cyproterone acetate (CA), before or
concurrently to the administration of gonadotropin-releasing hormone
analog (GnRHa). Treatment of coho and chinook salmon with either
androgen did not increase the effectiveness of GnRHa for inducing
ovulation. Chinook salmon showed indications of accelerated ovulatory
response to MT alone relative to controls. Treatment coho salmon
with CA significantly decreased circulating testosterone, but resulted
in no significant delay in time elapsed to ovulation.
36
Introduction
The efficiency of fish culture operations can be enhanced
through careful control of the reproductive cycle. For an increasing
number of fish species under culture, hormone treatments have
become an important tool for controlling sexual maturation. Initially,
gonadotropins were shown to be successful inducers of maturation and
ovulation (Clemens and Sneed 1962; FAO 1975; and Lam 1982).
However, the difficulty and cost in obtaining gonadotropins from
pituitaries led to the use of hypothalamic hormones, especially the
gonadotropin-releasing hormone (also called luteinizing hormonereleasing hormone) and its superactive synthetic analogs (Donaldson
et at. 1981; Sower et at. 1982; Crim et a/. 1983a,b; Donaldson and
Hunter 1983; Van Der Kraak et at. 1983; Crim and Glebe 1984; Sower
et al. 1984; Fitzpatrick et a/. 1984; Van Der Kraak et al. 1985).
Endocrine control of final ovarian maturation in fishes underlies
the success of these hormone treatments (Goetz 1983). Changes in
the plasma levels of steroids and gonadotropin accompany final
maturation in salmonids (Scott et at. 1983; Van Der Kraak et at. 1983;
Fitzpatrick et al. 1986). Gonadotropin, secreted by the pituitary,
induces the release of 17a,2013-dihydroxy-4-pregnen-3-one (DHP)
from the ovary, and DHP, in turn, stimulates the final maturation of the
oocytes (Jalabert 1976). Gonadotropin-releasing hormone (GnRH)
controls the release of gonadotropin from the pituitary (Schally et al.
1973). Synthetic analogs of GnRH (GnRHa) have been shown to
elevate plasma gonadotropin (Van Der Kraak et at. 1983), plasma DHP
(Van Der Kraak et a/. 1984) and testosterone (Fitzpatrick et al. 1987)
37
in coho salmon, Oncorhynchus kisutch. In salmonids, two distinct
gonadotropins (GtH I and GtH II) have been recently isolated that have
similar steroidogenic capacities (Suzuki et at. 1988a), but appear to
differ in synthesis and secretion during development (Suzuki et at.
1988b; Swanson et al. 1989). GtH II is generally regarded as the
"maturational" gonadotropin and GtH I has predominance during the
juvenile stage.
It is not known what, if any, function the high levels of
testosterone have in final ovarian maturation. Testosterone enhances
the effectiveness of gonadotropin on the in vitro maturation of oocytes
in amago salmon, 0. rhodurus (Young et al. 1982; Nagahama et a/.
1983). In vivo pretreatment of juvenile rainbow trout, 0. mykiss, with
testosterone resulted in pituitary accumulation of gonadotropin which
was released in vitro after stimulation with GnRHa (Crim and Evans
1980). Methyltestosterone given in conjunction with GnRHa to
milkfish, Chanos chanos, significantly stimulated maturation compared
to controls or fish given GnRHa alone (Lee et al. 1986a,b). The high
levels of testosterone found in females near final maturation have been
postulated to be responsible for stimulating gonadotropin synthesis
(Fitzpatrick et al. 1987; Kobayashi et al. 1989a,b).
Significant numbers of female salmonids do not ovulate soon
after GnRHa treatment (Crim and Glebe 1984; Fitzpatrick et al. 1987),
especially when treatment occurs too long before the time of natural
ovulation. Zohar (1989) suggested that a failure in GtH release
underlies the lack of--or delay in--final oocyte maturation. We have
shown previously that high plasma testosterone levels are predictive of
successful ovulatory response to GnRHa in coho salmon (Fitzpatrick et
38
a/. 1987).
Perhaps a lack of sufficient pituitary GtH is the reason for
the occasional failure of GnRHa. In the following study, we tested the
hypothesis that augmentation of the GnRHa treatment with androgens
to enhance GtH synthesis would lead to increased induction of
ovulation in coho and chinook salmon, 0. tschawytscha. Our
experimental approach entailed the use of several distinct delivery
systems, different timing of treatments, doses of hormones, and the
use of the antiandrogen, cyproterone acetate, which has been reported
to be both antiandrogenic and antigonadotropic in mammals
(Neumann et al. 1977).
Materials and Methods
Animals. The experiments were performed with coho salmon
from the Oregon Aqua Foods, Inc., Springfield, Oregon hatchery and
with chinook salmon from the Little White Salmon National Fish
Hatchery in southwestern Washington.
Chemicals. GnRHa (des-Gly10[D-Ala61-LHRH-ethylamide),
testosterone (T), 17a-methyltestosterone (MT), tripalmitin, polyvinyl
alcohol were obtained from Sigma Chemical Company, St. Louis, Mo.
Cyproterone acetate (CA) was a gift from Schering Corporation,
Bloomfield, N.J. Menhaden oil was a gift from Zapata Haynie
Corporation, Reedville, Va. Kaomel was a gift from Durkee Foods, Inc.,
Lafayette, CA.
Treatment Protocol. The first experiment was designed to test
the effectiveness of T implants given 1 or 2 weeks before the first of
two GnRHa injections. We also compared the effect of treatment
39
timing by temporally separating implantation and injection. In 1986,
approximately 250 female coho salmon were isolated from the rest of
the hatchery brood stock and assigned to one of three groups. On 22
Sep (Group A) and 7 Oct (Groups B and C), the each fish was
anesthetized, tagged with a metal jaw tag, and a blood sample was
taken. On the day of implantation, each fish was given an implant
(volume = 1 mL) of either T (1.25 mg) dissolved in cocoa butter or
cocoa butter alone ("sham"). On the days of injection, each fish was
given an injection (volume = 0.5 mL) of either 5 lag of GnRHa in saline
or saline alone. Group A was given implants on 22 Sep and injections
on 6 and 8 Oct. Group B was given implants on 7 Oct and injections
on 21 and 23 Oct. Group C was given implants on 14 Oct and
injections on 21 and 23 Oct. Blood samples were taken weekly
through the first injection date (beginning with the implantation date
for Groups A and B; and 1 week before implantation for Group C) and a
final blood sample was taken on the second injection date. After the
last injection, females were checked by palpitation twice a week for
the first month and weekly thereafter to determine if ovulation had
occurred. The experimental design is summarized in Figure 5. Blood
samples were centrifuged immediately and the plasma was stored at
200C until analysis by radioimmunoassay for T, GtH I, and GtH II.
To assess the effect of treatment on viability of the offspring, a
sample of eggs from each female was incubated for 11 days at 12 °C at
which time the eggs were fixed in Stockards solution and
subsequently checked for embryonic development. To determine the
effect of testosterone treatment on sex differentiation of the offspring,
1986 Coho Study
-14
-7
0
2
check for
ovulation
implant
implant
testosterone or
sham
testosterone or
sham
Group C: 7 Oct
Group A: 22 Sep
Group B: 7 Oct
nject
GnRHa
or
saline
1987 Coho Study
-7
0
check for
ovulation
inject
MT, CA, or
sham
microcapsules
inject
GnRHa
or
saline
Figure 5. Experimental Design for the 1986 and 1987 mho salmon studies.
41
the remaining eggs were allowed to hatch and the offspring were
pooled according to the treatment received by the mother. The fish
were grown for 6 months in circular tanks at Oregon State University's
Smith Farm Hatchery at which time they were killed and examined to
determine sex.
In addition to the cocoa butter implants used in 1986, we tested
a microencapsulation technique (developed with Dr. Chris Langdon,
Oregon State University) that provided for better hormone delivery
and easier administration to the fish. The microcapsules consisted of
menhaden oil, kaomel, tripalmitin, polyvinyl alcohol, and water.
Steroids (MT or CA) were dissolved in the menhaden oil (8% w/v),
then kaomel (32% w/v) and tripalmitin (48% w/v) were added and
mixed well. To this mixture, 2% (w/v) polyvinyl alcohol (250% v/v)
was added and immediately sonicated. The mixture was diluted in one
half by ice cold water. The initial test of the microcapsules took place
using juvenile coho salmon that were each treated with 20 mg/kg body
weight of MT microcapsules (see Appendix I). Blood samples were
taken at 1, 4, and 10 days after injection, the plasma samples were
pooled for each date, and then analyzed for the presence of MT by
High Performance Liquid Chromatography (using the procedure of
Patirio et al. 1987). In a preliminary study, we found that this
technique can be used to detect MT in plasma 10 days after
microcapsule implantation (see Appendix I).
In 1987, we designed experiments to test the effects of
pretreatment with microencapsulated MT or the antiandrogen CA on
GnRHa-induced ovulation (Fig. 5). A single group of coho salmon
females was isolated on 29 Oct. The fish given an implant (1.5 mL) of
42
either sham microcapsules, 20 mg MT microcapsules, or 20 mg CA
microcapsules. On 5 Nov, the fish were injected (0.5 mL) with either
saline or 6.7 lig of GnRHa. A single injection of GnRHa was used in the
hope that such a suboptimal treatment would allow the effects of MT
or CA to be better manifested. Blood samples were taken on 29 Oct
and 5 Nov.
We designed the following experiment to test the effect of MT
pretreatment on GnRHa-induced maturation in chinook salmon (Fig.
6). On 15 Jul 1987, a single group of chinook salmon females was
isolated from the brood stock, each fish given an implant of sham
microcapsules, 50 mg MT microcapsules, or 5 mg MT microcapsules.
On 27 and 30 Jul, the fish were injected with either saline or 20 1.tg of
GnRHa (the dose was higher than that for coho salmon because the
chinook salmon were about four times larger). Blood samples were
taken on 15, 27, and 30 Jul, at the times of implant and injection.
We tested the action of both pretreatment and concurrent
treatment with MT on GnRHa-induced maturation of chinook salmon
in 1988 (Fig. 6). On 18 Jul, a single group of chinook salmon females
was isolated and implanted with sham microcapsules or 50 mg MT
microcapsules. On 28 Jul, each fish had a blood sample taken and was
given an injection of either saline, 5 mg of MT in saline, 20 lag of
GnRHa, or 5 mg MT + 20 fig of GnRHa. Fish were checked weekly to
determine if ovulation had occurred.
Radioimmunoassay. All plasma samples were analyzed for T
levels as described in Fitzpatrick et al. (1986, 1987). The assay has
been validated for coho salmon plasma (Fitzpatrick et al. 1986); the
assay was validated for chinook salmon plasma by the determination of
1987 Chinook Study
-13
0
3
check for
ovulation
in ect
inject
GnRHa
or
MT or sham
microcapsules
saline
1988 Chinok Study
-10
0
check for
ovulation
inject
MT or sham
microcapsules
inject
GnRHa,
MT+GnRHa,
MT, or
saline
Figure 6. Experimental Design for the 1987 and 1988 chinook studies
44
parallelism between serial dilutions of pooled chinook salmon plasma
and the standards. The average extraction efficiency for both coho and
chinook salmon plasma was greater than 90%. The cross-reactivity of
the antibody with MT and CA was < 0.1%.
Plasma levels of gonadotropin (GtH) I and GtH II were
determined in coho salmon plasma from Group B of the 1986
experiment. The methods used for radioimmunoassays of coho GtH I
and II (Swanson et a/. 1989) were similar to those described in detail
by Suzuki et al. (1988a) for chum salmon, 0. keta, GtH I and II.
Statistics. The time elapsed to ovulation was compared between
treatment groups using the Mann-Whitney U test after initial KruskalWallis analysis of variance. Plasma levels of hormones were compared
between treatment groups on any given sampling date using MannWhitney U test after initial Kruskal-Wallis analysis of variance. Plasma
levels of hormones were compared between sampling dates within
treatment groups using Wilcoxon sign-rank test for dependent
samples after initial Friedman's analysis of variance. In all analyses,
the level of significance was set at P < 0.05.
Results
Days to Ovulation
For the 1986 Coho Study, no significant differences existed
between the treatments with regard to the number of days elapsed to
ovulation in Group A (Figs. 7-8). In Groups B and C, females treated
with sham implants followed by GnRHa injection (S:G) and females
treated with testosterone implants followed by GnRHa (T:G) ovulated
45
Figure 7. Mean days to ovulation (from Day 0) for coho salmon during
the 1986 study. Values given are means (± S.E.) of fish that ovulated
after being treated with a testosterone (Test.) or sham implant
followed by an injection 1 (Group C) or 2 (Groups A and B) weeks later
with GnRHa or saline (on Day 0). Means marked by an "*" indicate
significant differences from unmarked means within that Group.
46
Sham:Saline
Group A
Figure 7.
Group B
Group C
Ea
Sham:GnRHa
0
Test.: Saline
12
Test.:GnRHa
47
Figure 8. Induced ovulation in coho salmon in 1986. The numbers of
fish that ovulated after being treated with a testosterone or sham
implant 1 (Group C) or 2 (Groups A and B) weeks before an injection
with GnRHa or saline (on Day 0) are expressed as cumulative
percentages of their respective groups.
48
Group A
100
80
a Sham:Saline
Sham:GnRHa
0 Test.:Saline
Test.:GnRHa
0
0
100
10
20
30
40
50
60
Group B
80
ar
60
a)
a Sham:Saline
CL 0
Sham:GnRHa
CD
40
7-0-3
E
8
z
0 Test.:Saline
Test.:GnRHa
20
0
0
100 _
80
10
20
30
40
Group C
_
a)
0
if)
0-
c
C)
60
0
o
Sham:GnRHa
Test.:Saline
Test.:GnRHa
40
cis 5
50
20
0
0
10
20
Days Elapsed to Ovulation
Figure 8.
--a Sham:Saline
30
49
significantly earlier than females in either sham-implanted: saline
injected (S:S) or testosterone-implanted: saline-injected (T:S).
However, no differences between T:G and S:G groups nor between T:S
and S:S groups were observed. In the 1987 Coho Study, no significant
differences existed between treatment groups with regard to days
elapsed to ovulation (Figs. 9-10). In the 1987 Chinook Study, the high
dosage MT implant followed 2 weeks later by GnRHa injections
(MT:G), MT implant followed by saline injections (MT:S), and S:G
females ovulated significantly sooner than did S:S females (Figs. 1112). In the 1988 Chinook Study, no significant differences existed
between treatments with regard to time to ovulation (Figs. 13-14).
Effects of implants on plasma testosterone
In the 1986 Coho Study, testosterone implants did not result in
elevated circulating testosterone, as determined one week after
implantation in Group A; however, in Groups B and C, testosterone-
implanted fish had higher testosterone levels one week after
implantation (Fig. 15). In the 1987 Coho Study, MT and CA
microcapsule-implanted fish had significantly decreased testosterone
levels one week after implant (Fig. 16). In the 1987 Chinook Study,
the effects of MT microcapsules on plasma testosterone were
equivocal as MT:S fish showed no difference in testosterone 2 weeks
after implantation, but both MT:G groups demonstrated increased
testosterone levels (Fig. 17). In the 1988 Chinook Study, testosterone
concentrations were not different between females with MT
microcapsules and females with sham microcapsules (blood samples
taken 10 days after implantation; Fig. 18).
50
Figure 9. Mean days to ovulation from Day 0 for coho salmon during
the 1987 study. Values given are means (± S.E.) of fish that ovulated
after being treated with methyltestosterone (MT), cyproterone acetate
(CA), or sham microcapsules 1 week before an injection (on Day 0)
with GnRHa or saline.
51
23
El
1987 Coho
Figure 9.
Sham:Saline
Sham:GnRHa
MT: Saline
MT: GnRHa
CA: GnRHa
52
Figure 10. Induced ovulation in coho salmon in 1987. The numbers of
fish that ovulated after being treated with methyltestosterone (MT),
cyproterone acetate (CA), or sham microcapsules 1 week before an
injection (on Day 0) with GnRHa or saline are expressed as cumulative
percentages of their respective groups.
53
1987 Coho Salmon
100
80
60
Sham:Saline
40
Sham:GnRHa
MT:Saline
20
MT:GnRHa
CA:GnRHa
0
10
20
Days Elapsed to Ovulation
Figure 10.
30
54
Figure 11. Mean days to ovulation from Day 0 for chinook salmon
during the 1987 study. Values given are means (± S.E.) of fish that
ovulated after being treated with methyltestosterone (MT) or sham
microcapsules 13 days before the first (on Day 0) of two injections
with GnRHa or saline (on Day 0). Means marked by an "*" indicate
significant differences from unmarked means.
55
ea
ri
a
1987 Chinook
Figure 11.
Sham:Saline
Sham:GnRHa
MT:Saline
MT:GnRHa
low MT:GnRHa
56
Figure 12. Induced ovulation in chinook salmon in 1987. The
numbers of fish that ovulated after being treated with
methyltestosterone (MT) or sham microcapsules 13 days before the
first (on Day 0) of two injections with GnRHa or saline are expressed
as cumulative percentages of their respective groups.
57
1987 Chinook Salmon
100
80
Sham:Saline
Sham:GnRHa
MT:Saline
20
MT:GnRHa
low MT:GnRHa
0
10
20
Days Elapsed to Ovulation
Figure 12.
30
58
Figure 13. Mean days to ovulation from Day 0 for chinook salmon
during the 1988 study. Values given are means (± S.E.) of fish that
ovulated after being treated with methyltestosterone (MT) or sham
microcapsules 10 days before an injection (on Day 0) with GnRHa,
GnRHa + MT, MT, or saline (on Day 0).
59
20
_
Sham:Saline
Sham:GnRHa
Sham:MT+GnRHa
10_
leg
MT:MT
MT:GnRHa
ea
0
1988 Chinook
Figure 13.
MT:MT+GnRHa
60
Figure 14. Induced ovulation in chinook salmon in 1988. The
numbers of fish that ovulated after being treated with
methyltestosterone (MT) or sham microcapsules 10 days before an
injection (on Day 0) with GnRHa, GnRHa + MT, MT, or saline are
expressed as cumulative percentages of their respective groups.
61
1988 Chinook Salmon
100
80
#6
G)
2
cu c
60
13- 0
0 Sham:Saline
0 Sham:GnRHa
0 :47..
> to
ci
40
-5O
E
of
&-- Sham:MT+GnRHa
MT:MT
20
---11--
MT:GnRHa
--i----
MT:MT+GnRHa
1
0
10
20
Days Elapsed to Ovulation
Figure 14.
30
62
Figure 15. Mean plasma concentrations of testosterone in female coho
salmon in 1986. Females were treated with testosterone or sham
implant (22 Sep for Group A; 7 Oct for Group B; 14 Oct for Group C)
and injected with GnRHa or saline (6 and 8 Oct for Group A; 21 and
23 Oct for Groups B and C). Each value represents the mean (± S.E.).
Means marked by "*" are significasntly different from unmarked
means within that treatment Group and by "**" are significantly
different from all other means within that treatment Group.
63
300 -
200
Group A
_
IZ
100
III
_
Ea
22 Sep
29 Sep
6 Oct
8 Oct
0
300
**
**
Group B
200 -
*
rE
7 Oct
14 Oct
21 Oct
23 Oct
El
7 Oct
14 Oct
21 Oct
23 Oct
173
100
IIIII
_
0
**
**
*
Figure 15.
64
Figure 16. Mean plasma concentrations of testosterone in female coho
salmon in 1987. Fish were treated with methyltestosterone (MT),
cyproterone acetate (CA), or sham microcapsules on 29 Oct and with
GnRHa or saline on 5 Nov. Each value represents the mean (± S.E.).
Means marked by "*" are significantly different from unmarked means
within that treatment.
CA:GnRHa
MT:GnRHa
MT:Saline
Sham:GnRHa
Sham:Saline
00
co
N.)
C)
r.
o
0
<
Z
01
*
0
0
iv
Testosterone (ng/ml)
(A)
00
66
Figure 17. Mean plasma concentrations of testosterone in female
chinook salmon in 1987. Fish were treated with methyltestosterone
(MT) or sham microcapsules on 15 Oct and with GnRHa or saline on
27 and 30 Jul. Each value represents the mean (± S.E.). Means
marked by "*" are significantly different from unmarked means within
that treatment.
MT:GnRHa
MT:Saline
Sham:GnRHa
Sham:Saline
Sa*.gt4"
O
C
C.-
9
--s.
CTI
C_.
'"%e
*
v`4,0
Waete.A.
C CC-
\
sage
Testosterone (ng /mI)
N)
68
Figure 18. Mean plasma concentrations of testosterone in female
chinook salmon on 28 Jul 1988. Fish were treated with
methyltestosterone (MT) or sham microcapsules on 18 Jul and with
GnRHa, GnRHa + MT, MT, or saline on 28 Jul. Each value represents
the mean (_± S.E.).
69
1988 Chinook salmon
200
-
150
_
gg
100 Ea
50
El
El
_
0
28 Jul
Figure 18.
Sham:Saline
Sham:GnRHa
Sham:MT+GnRHa
MT:MT
MT:GnRHa
MT:MT+GnRHa
70
Plasma testosterone in early and late ovulators
The females in each treatment were divided into two groups
designated "early ovulators" as indicated by ovulation taking place
within 14 days of injection with either GnRHa or saline and "late
ovulators" as indicated by ovulation taking place in more than 14 days
(this designation was based on the Oregon Aqua Foods management
goal of inducing ovulation within 2 weeks of injectionanything later
than 2 weeks decreases the benefits of the treatment because of
increased mortality, increased labor costs, and inability to produce
eyed eggs in a timely manner for filling orders).
In the 1986 Coho Study, early ovulators tended to have higher
testosterone levels than late ovulators in Groups B and C (Fig. 19).
Only two fish in Group A ovulated within two weeks -both had
testosterone levels higher than the means for the late ovulators. In
the 1987 Coho Study (Fig. 20), early ovulators had higher testosterone
than late ovulators only in MT:G and MT:S groups (this difference
occurred in both pre- and post-implant plasma concentrations). In
the 1987 Chinook Study (Fig. 21), early and late ovulators showed no
consistent differences in plasma testosterone. In the 1988 Chinook
Study (Fig. 22), early ovulators had higher testosterone than late
ovulators in MT:G and MT:MT+G groups, but all other groups had no
significant differences between early and late ovulators.
Effects of GnRHa injection on Testosterone
In the 1986 Coho Study and 1987 Chinook Study, blood samples
were taken after GnRHa injection. In coho salmon, plasma
testosterone levels were no higher 2 days after the first GnRHa
injection than on the day of injection in Group A (Fig. 15). In Groups B
71
Figure 19. Mean plasma concentrations of testosterone in early and
late ovulating female coho salmon in 1986. Each value represents the
mean (± S.E.) of fish that ovulated either within 14 days (early
ovulators) or in more than 14 days (late ovulators) of the first injection
with either GnRHa or saline (on 6 Oct for Group A; 21 Oct for Groups
B and C). Arrows indicate the date of testosterone implant. An "*"
indicates significant differences between 5_14 and >14 groups on the
specified date (p < 0.05).
72
Group A
500
Sham:
Saline
400
_
300
_
Test:
Saline
Sham:
GnRHa
Test:
GnRHa
200_
100_
efa-4-4
1111111 I
0
111
0Q. t r)
a) a)
N CV
CV 0) CD cO
N CV
(I) U)
a) 0 0
(f)
N 0) CO c0 CV CO CD CO
al
CV CV
I
a)
C:2- Q. r) tj Ct. Q. .,'rt -
I
111
Q. Ct_
00 0
NN NCI
CD c0
Group B
Days Elapsed
to Ovulation
0--411-- >14
I
I
I
U
I
I
I
I
o
I
Li U U
00000
O 0 0 0 0 0 0 0 0 0 0 0 0U 0 0 0
c.)
(.2
:1' I c
Group C
c.)
N
r- Tt N
N
1 Cr) N. cr
N CV
500
400
300
200
100
0
O'
N-
Figure 19.
,c2*
S ' 8 8 6 ' 8 6 ' 8 8 8 8 O' 8 8 6'
sr
cv
r-
N
cv
r-
N
.cA N
73
Figure 20. Mean plasma concentrations of testosterone in early and
late ovulating female coho salmon in 1987. Each value represents the
mean (± S.E.) of fish that ovulated either within 14 days (early
ovulators) or in more than 14 days (late ovulators) of injection (on 5
Nov) with either GnRHa or saline. Implantation of sham,
methyltestosterone (MT), or cyproterone acetate (CA) microcapsules
took place on 29 Oct. An "*" indicates significant differences between
.14 and >14 groups on the specified date (p < 0.05).
5 Nov
29 Oct
5 Nov
29 Oct
5 Nov
29 Oct
4111
*
4=.
V
IA
Or-1
ir, p
0= 0>
oo
2a)
i
Do
0= --i
K
ca
Do il)
5 Nov
F2: fa
u) u)
O
O
0= (I)
=
O
O
29 Oct
O
O
5 Nov
110
O
0
P P
29 Oct
O
Testosterone (ng /mI)
O
O
(Y1
75
Figure 21. Mean plasma concentrations of testosterone in early and
late ovulating female chinook salmon in 1987. Each value represents
the mean (± S.E.) of fish that ovulated either within 14 days (early
ovulators) or in more than 14 days (late ovulators) of the first injection
(on 27 July) with either GnRHa or saline. Implantation of sham or
methyltestosterone (MT) microcapsules took place on 15 Jul. An "*"
indicates significant differences between
and >14 groups on the
specified date (p < 0.05).
27 Jul30 Jul15 Jul27 Jul30 Jul15 Jul27 Jul30 Jul-
is A 11.-
15 Jul27 Jul30 Jul
15 Jul27 Jul30 Jul-
0
V
li
0
0
_...
IA
1)
0
0
i
K.)
Testosterone (ng/ml)
CD
Cu
GO M
0
0
CO
77
Figure 22. Mean plasma concentrations of testosterone in early and
late ovulating female chinook salmon in 1988. Each value represents
the mean (± S.E.) of fish that ovulated either within 14 days (early
ovulators) or in more than 14 days (late ovulators) of the first injection
(on 28 July) with either GnRHa or saline. Implantation of sham or
methyltestosterone (MT) microcapsules took place on 18 Jul. An "*"
indicates significant differences between <14 and >14 groups on the
specified date (p < 0.05).
28 Jul
28 Jul
i*
V
10i
IA
*
1-0-1
*
28 Jul
28 Jul
I-0-1
NIA
a
a
28 Jul -
28 Jul-
o
*
1
a
0
-I
71
--I
3
m-
s')
+ ri)
CD
(/)
+
Testosterone (ng/ml)
79
and C (which received GnRHa 15 days after Group A), testosterone
concentrations were elevated 2 days after GnRHa injection in both
sham and testosterone pretreated fish (Fig. 15). In chinook salmon,
plasma testosterone was elevated :3 days after GnRHa injection in 2 of
the 3 GnRHa-treated groups (Fig. 17).
Effects of implants and injections on GtH I and GtH II
In the 1986 Coho Study, plasma levels of GtH I and GtH II were
determined on the days of injection with either GnRHa or saline in
Group B (i.e., 14 and 16 days after implantation). Testosterone
implantation had no significant effect on gonadotropins as determined
by comparing the plasma levels of GtH I or II on 21 Oct between the
various treatments. Testosterone implantation did not modulate the
response of females to the first GnRHa injection as measured by a
comparison of the GtH I and II levels on 23 Oct between the
treatment groups. Between 21 and 23 Oct, GtH I decreased
significantly in all treatments (Fig. 23) and GtH II increased
significantly in both groups treated with GnRHa (Fig. 24). The levels
of GtH I and II were not consistently different between early and late
ovulators on either date.
Effects of implants on egg viability and gender of offspring
No significant effect was observed of testosterone implants on
the viability of the eggs or on the sex ratio of the offspring of treated
females (Tables III and IV). Eggs from females treated with GnRHa
had significantly lower viabilities than those from saline-treated fish
for both testosterone- and sham-implanted coho in Group C.
80
Figure 23. Mean plasma concentrations of gonadotropin (GtH) I in
female coho salmon from Group B in 1986. Females had been treated
with a testosterone or sham implant on 7 Oct followed by an injection
of GnRHa or saline on both 21 and 23 Oct. Therefore, each value
represents the mean (± S.E.) for females after implantation but before
the first injection (21 Oct) and after the first injection (23 Oct).
Means marked by "*" are significantly different from unmarked means
within that treatment group.
P
No
1-1
CD
O'Cl
:11)
Test.:GnRHa
TAq.:RnlinP
Sham:GnRHa
Sham:Saline
a
I
000
0
Ca --
N.) 1\3
EI
__.
a
GtH I (ng/ml)
I
aN3
oo
82
Figure 24. Mean plasma concentrations of gonadotropin (GtH) II in
female coho salmon from Group B in 1986. Females had been treated
with a testosterone or sham implant on 7 Oct followed by an injection
of GnRHa or saline on both 21 and 23 Oct. Therefore, each value
represents the mean (± S.E.) for females after implantation but before
the first injection (21 Oct) and after the first injection (23 Oct).
Means marked by "*" are significantly different from unmarked means
within that treatment group.
Test.:GnRHa
Test.:Saline
Sham:GnRHa
Sham:Saline
0
0
N3
0
G3
4 =.
00
00
G.) -+
N3 N.)
1:1
0
1.11.1.11q
0
__.
GtH II (ng /mI)
*
84
Table III.
1986 Coho Study. Viabilities of eggs 11 days after
fertilization. Values given for egg viabilities are mean
percentages ± S.E. from the indicated number of females.
Significant differences between treatments are indicated
by different letters.
Treatment
Implant
Date
n
Egg
Viabilities
5
8
96.7 ± 0.9
76.6 ± 8.7
Sham:Saline
Sham:GnRHa
Testosterone:Saline
Testosterone:GnRHa
22 Sep
Sham:Saline
Sham:GnRHa
Testosterone:Saline
Testosterone:GnRHa
7 Oct
t,
7
8
9
9
70.4 ± 10.5
66.2 + 10.6
74.8 ± 14.3
43.9 ± 12.4
Sham: Saline
Sham: GnRHa
14 Oct
9
7
8
6
22.9 + 12.8b
Testosterone:Saline
Testosterone:GnRHa
6
6
83.9 ± 12.2
87.4 + 4.7
72.2 ± 9.8a
82.5 + 5.8a
26.1 ± 14.2b
85
Table N. 1986 Coho Study. Gender of offspring from treated
females. Values given are percentages of males of the
indicated numbers of individuals.
Treatment
n
Sham:Saline
Sham:GnRHa
Testosterone:Saline
Testosterone:GnRHa
50
99
62
100
Percentage
of Males
50.0
48.5
56.5
52.0
86
Discussion
Our results indicate that pretreatment with androgens, or
concurrent treatment with androgens and GnRHa, did not increase
the effectiveness of GnRHa for inducing ovulation in Pacific salmon.
Neither T implantation, MT microcapsules, nor MT injection
augmented GnRHa action. However, MT microcapsules alone resulted
in accelerated maturation compared to controls in one instance with
chinook salmon. These results contrast with the demonstrated
success of androgen treatment in accelerating maturation in GnRHatreated milkfish (Lee et a/. 1986a) and to the inhibition of GnRHainduced maturation by MT in striped mullet (Kelley et al. 1987). The
dosages used in the present study were similar to those used in
studies of milkfish (Lee et a/. 1986a,b) and juvenile salmonids (Crim et
al. 1981; Swanson and Dickhoff 1988).
Testosterone pretreatment did not result in increased plasma
GtH II either before or following GnRHa injection. The females that
received T implants later in the spawning season had significantly
elevated T one week after implantation, but after two weeks when the
first GnRHa treatment occurred, the levels of T were similar between
T- and sham-implanted fish. Stimulation of pituitary GtH II synthesis
may have been obscured by the normal increase in GtH II synthesis
that occurs at this time in salmonids (Kawauchi et a/. 1989). Shamimplanted females showed the same significant increase in plasma
GtH II that T implanted females demonstrated 2 days after GnRHa
administration. In juvenile salmonids, when pituitary GtH II synthesis
is stimulated by in vivo treatment with T, in vitro GnRHa treatment of
87
pituitaries stimulates the release of GtH II (Swanson and Dickhoff
1988). We were unable to demonstrate this effect of T on GtH II in
the plasma of adult salmon. In the maturing salmon used in the
present study, the T implant may not have been of sufficient dosage or
given at the correct time to allow detection of plasma concentrations
above normal levels. A number of ovariectomized goldfish that were
implanted with T demonstrated a GtH surge three months later
similar to that observed during normal spontaneous ovulation
(Kobayashi et al. 1989a,b). Similarly, successful treatment of milkfish
with MT + GnRHa occurred when the androgen was given several
months in advance of the expected spawning date (Lee et a/. 1986a).
The increase in testosterone that was observed 2 days after GnRHa
injection in Groups B and C in 1986 was probably stimulated via GtH
II; therefore, the lack of T increase in females treated with GnRHa in
Group A may indicate pituitary insensitivity due to insufficient
synthesis or release of GtH II or perhaps inability to overcome
inhibitory controls. Van Der Kraak et a/. (1986) found that treatment
of coho salmon with a dopamine antagonist significantly enhanced the
effectiveness of GnRHa in inducing ovulation, suggesting that
dopamine acts as a inhibitor of GtH release. It is not known what
effect T has on dopamine regulation of GnRH in teleosts.
Elevated T was correlated to quicker ovulatory response in many
of the experimental groups; however, the results were not as dramatic
as those observed in an earlier study (Fitzpatrick et al. 1987). The
lack of consistent results may have been due to the designation in this
study of "late ovulators" as fish that spawned after 14 days, whereas in
the earlier study, "late ovulators" spawned in 45 or more days. In the
88
present study, only in coho salmon from the earliest treatment group
in 1986 did ovulation occur in individuals 45 days after treatment, so
the distribution of time elapsed to ovulation for most groups was
narrower. In the 1987 coho study, GnRHa did not significantly
accelerate maturation compared to controls probably because the
controls were already close to ovulation. In support of this suggestion,
the mean days to ovulation were lower than in previous studies with
this stock of fish (Fitzpatrick et al. 1984, 1987; Table I) and plasma
levels of T in all groups were relatively high on 29 Oct at the initiation
of the experiment.
The function of the elevated T during final maturation remains
enigmatic. Androgen treatment led to significant acceleration of
ovulation in only one instance (MT:saline-treated chinook salmon) and
never in conjunction with treatment with GnRHa. One difficulty with
the present study could be that ovulation was used as the experimental
endpoint. Ovulation is under the control of gonadotropin through nonsteroidal means and final oocyte maturation, which involves the
resumption of meiosis, may be temporally separate from ovulation
(Goetz 1983). Therefore, any effect of testosterone on the resumption
of meiosis may have been masked by natural delays in ovulation.
Treatment of coho salmon with CA, an antiandrogen, did not
significantly delay ovulation despite significantly decreasing circulating
testosterone. However, this conclusion may be premature as viewing
the data by cumulative percentage of fish that ovulated versus time
(Fig. 10) indicates an apparent delay in ovulatory response in the CA-
treated group compared to the other GnRHa-treated groups. The CA:G
group at seven days post-GnRHa injection had about the same
89
percentage ovulation as did the S:S group, but reached about the same
percentage as the S:G and MT:G groups at 11 days post-GnRHa
injection. The levels of T in coho on the day of CA-treatment were
already elevated compared to the levels observed in early samples
from coho in 1986 and to those reported previously (Fitzpatrick et al.
1986; 1987). Therefore, the optimal point for inhibiting the
physiological actions of testosterone may have passed. In addition,
acute effects of CA on GnRHa-induced ovulation were not investigated
and may thus require further study if the elevated testosterone after
GnRHa injection is important to the ovulatory response. In the only
groups in which the effects of an acute treatment with androgens were
assessed, two groups of chinook salmon given a combined injection of
MT and GnRHa had higher cumulative percentages of females that
ovulated than those given GnRHa alone at all the times that the fish
were checked. However, these results did not translate into
significant differences in mean days elapsed to ovulation (Fig. 14).
This discordance in the cumulative percentage calculation stems from
the inclusion of animals that had suffered mortality or tag loss, or
simply did not ovulate during the study, whereas the statistical analysis
of days to ovulation includes only those animals that ovulated.
In addition to actions on pituitary synthesis of GtH in juveniles
(Crim and Evans 1980; Swanson and Dickhoff 1988), testosterone has
been linked to other processes that may have bearing on final
maturation. Testosterone has been reported to modulate the synthesis
of pituitary GnRH receptors in rats (Frager et al. 1981; Clayton and
Catt 1981; Giguere et al. 1981), although these reports generally point
to a down-regulation of receptors and such down-regulation would
90
seemingly be in contrast to the hypothesized positive role of
testosterone in female teleosts. Recent work has suggested that
endogenous opiates may modulate GnRH and dopamine release in
male goldfish (Rosenblum and Peter 1989) and testosterone may
determine hypothalamic sensitivity to opiate antagonists in rats
(Nikolarakis et al. 1986). In female salmonids, T may be acting beyond
the hypothalamo-pituitary-gonadal axis. Finally, testosterone may be
involved in the behavior of female salmon which under natural
conditions would be undergoing spawning migrations and nest-
building. In talapoin monkeys, testosterone levels were lower in
subordinate females than in either dominant or intermediate-ranking
females suggesting that social hierarchy may regulate androgen levels
(Batty et at. 1986). Li ley et a/. (1986) speculated that high levels of
testosterone may play a role in the aggressive behavior observed in
female rainbow trout defending nests during the time of spawning.
The current results support the concept of a "window of
sensitivity" to GnRHa-induced maturation (Fitzpatrick et al. 1987). In
1986, coho salmon were generally insensitive to GnRHa when injected
on 6 Oct but demonstrated marked acceleration when injected on 21
Oct. Kobayashi et al. (1988) found a similar window of sensitivity in
kisu (Sallago japonica), a fish that undergoes final maturation during a
24-hour period. In a germinal vesicle breakdown (GVBD) bioassay,
kisu underwent DHP-induced GVBD only at a certain time of day.
Sensitivity to GtH preceded that for DHP leading Kobayashi et al.
(1988) to speculate that GtH develops the oocyte's sensitivity to DHP
through stimulation of DHP receptors. High levels of T observed
during this time may also play a role in the modulation of DHP
91
receptors. The sensitivity of female milkfish to GnRHa + MT
treatment was correlated strongly to the diameter of the oocytes, with
successful spawning occurring only when egg size was between 700
and 750 pm (Tamaru and Lee 1987). These authors also found that
serum T increased slowly when egg diameters grew from 300 to 500
p.m and then dramatically increased when the size reached 700
Thus, the sensitivity to maturation-inducing agents may depend on the
availiabilty of receptors which, in turn, may depend on developmental
status of the oocyte (which in the case of the milkfish is reflected by
size and plasma testosterone).
An apparent limit to GnRHa-effectiveness harkens the need for
new approaches to the induced-maturation problem. The window of
sensitivity to GnRHa may not be expandable and instead efforts should
be made to accelerate the physiological events preceding GnRHasensitivity. Nutrition seems a likely avenue to this end in species that
continue to feed up to or near the time of maturation. However,
Pacific salmon, which do not feed once they enter fresh water, may be
problematic. Other methods such as decreased photoperiod (Bromage
et a/. 1984), which have been already proven effective in accelerating
the reproductive cycle, may be useful for pushing female salmonids to
the threshold of GnRHa-sensitivity. The use of androgens to enhance
the effectiveness of GnRHa for induction of ovulation in Pacific salmon
does not appear to be fruitful in the manner tested in this project.
However, the published success of androgen supplementation of
GnRHa in milkfish (Lee et al. 1986a,b) and the trend toward earlier
ovulation observed in the acutely MT-treated chinook salmon in 1988
92
(Fig. 2) warrants further investigation of the use of steroids in
conjunction with established maturation-accelerating regimens.
The development of the microcapsule delivery system holds
promise as an effective means for chronically elevating steroids. The
great advantage of microcapsules is their fluid characteristic. Unlike
implantation of pellets that require a large puncture into the fish or
cocoa butter that is very difficult to administer and must be kept
heated, microcapsules can be injected as easily as saline. The effective
elevation time for the microcapsules used in the current study appears
to be around 1 week as determined by the observed decrease in
plasma testosterone in MT-treated coho salmon in 1987 and the lack
of any observed effects on testosterone 10 days after MT-implant in
chinook salmon.
In summary, pretreatment with androgens or simultaneous
treatment with androgens and GnRHa did not augment GnRHa
effectiveness for induced maturation in coho or chinook salmon
females under the described treatment regimes. It may be that longer
term exposure or different dosage of androgens would increase the
efficacy of GnRHa as has been demonstrated in other species. The
development of the microcapsule delivery system should facilitate
further testing of androgens and antiandrogens.
Acknowledgements
We would like to thank the staffs of Oregon Aqua Foods and Little
White Salmon National Fish Hatchery for providing animals and
holding facilities. Sam Bradford, Cameron Sharpe, Grant Feist, Steve
93
Stone, and Alec Mau le provided excellent technical assistance. The
funds for this study were partly provided by the U.S. Department of
Agriculture, under grant No. 86-CRSR-2-2876.
94
IV. IN VITRO STEROID SECRETION DURING
EARLY DEVELOPMENT OF COHO SALMON
Martin S. Fitzpatrick
J. Michael Redding'
Carl B. Schreck
Oregon Cooperative Fishery Research UniT2, Oregon State University,
Corvallis, Oregon 97331
'present address: Department of Biology, Tennessee Technological
University, Cookeville, TN.
2Cooperators are the Fish and Wildlife Service, Oregon Department of
Fisheries and Wildlife, and Oregon State University.
95
Abstract
A static in vitro system was used to assess production of steroids
in coho salmon (Oncorhynchus kisutch) during early development.
Tissue pieces that included the gonads and interrenals were incubated
in the presence or absence or partly purified salmon gonadotropin
(SG) or porcine ACTH. Cortisol, androstenedione, testosterone, and
estradiol were produced soon after hatching. Within one month of the
onset of feeding, the tissues showed sensitivity to SG stimulation. A
sexually distinct pattern of in vitro testosterone and estradiol
production emerged near the time that the gonads showed obvious
signs of sexual dimorphism with more androstenedione and
testosterone produced by the testes and more estradiol produced by
ovaries.
96
Introduction
Research on sex differentiation in teleosts reveals a spectrum of
potential mechanisms, including induction by sex steroids (Yamamoto
1969; Hunter and Donaldson 1983). Steroids and/or steroidogenesis
appear very early in development. Rainbow trout (Oncorhynchus
mykiss) embryos are capable of accumulating radioactive
pregnenolone, progesterone, dehydroepiandrostenedione, and
estradiol (Anti la 1984), although in this study only pregnenolone and
progesterone were metabolized significantly. Gonadal homogenates
from rainbow trout synthesized progestins by 50 days after fertilization
(21 days after hatching), androgens by 100 days, and estrogens by 200
days (van den Hurk et al. 1982). Both of these studies on steroid
synthetic capacity of developing rainbow trout employed incubations of
radioactive steroids followed by thin layer and/or column
chromatography. Results from this technique are suggestive of
steroidogenic potential; however, little can be said about physiological
amounts of steroids produced by the tissue. In addition, the technique
requires bathing the tissue in a relatively enormous amount of steroid
and this conceivably could affect enzyme reaction kinetics and
equilibria.
If steroids are acting as inducers of sex differentiation in
teleosts, one possible mechanism is that sex differences exist in the
ability of steroidogenic tissue to synthesize steroids. Van den Hurk et
al. (1982) found no sex differences in synthetic capacity of gonads
until well after the period of sex differentiation. However, the
97
conclusions drawn from this study may be limited to which steroids
are capable of being synthesized- -not how much. Any sex differences
between males and females in the capacity to synthesize steroids are
likely to be of degree rather than kind -in other words, both males
and females will probably be able to make any particular steroid, but
profound differences may exist in the amount produced. Furthermore,
gonadal steroids may not be the prime factors influencing sex
differentiation. Van den Hurk and van Oordt (1985) suggest that the
interrenal may direct masculine sex differentiation through the
production of androstenedione derivatives.
The objective of this study was to determine if sex dimorphism
exists in steroid production by gonads and interrenals during early
development that may indicate a role of steroids in the process of sex
differentiation. Steroid production in coho salmon (Oncorhynchus
kisutch) in an in vitro incubation system was quantified by
radioimmunoassays; thus differences in amount of steroids produced
could be detected.
Materials and Methods
Animals. Coho salmon eggs (from Oregon Aqua Foods, Inc.,
Springfield, OR) were fertilized and raised until the eyed egg stage at
Smith Farm hatchery (Oregon State University) at which time they
were transported to the laboratory where they were held in a circular
tank containing water at approximately 10 °C. Experiments took place
twice before the onset of feeding (56 and 70 days post-fertilization
98
[DPF)) and three times after the onset of feeding (86, 103, and 126
DPF). In those instances in which the animals had been feeding, fish
were starved for the 24 hr before sampling. At the time of each
experiment, animals were removed from the tank and killed by a blow
to the head. For each sample, the viscera (and yolk sac when
necessary), head, and tail were removed and an explant of tissue
consisting of the anterior kidney (which contains the interrenal cells)
and gonads was placed in 1 mL of ice cold medium, which was the
same as that described in Patin° et al. (1986) except that the NaC1
content was increased to 130 mM.
Incubation. The explants were incubated for 18 hr in 1 mL of
incubation medium on a Junior orbit shaker (100 rpm) in a constant
temperature room (140C) after two 1-hr washes (the explants were
removed to tubes containing fresh media after each wash). For the
incubation, each explant was placed either in media alone, media
containing 2.5 pg of partially purified salmon gonadotropin (SG,
courtesy of Dr. E.M. Donaldson, West Vancouver Laboratory, West
Vancouver, British Columbia), or media containing 10 mU of porcine
adrenocorticotropic hormone (pACTH; Sigma Chemical Company, St.
Louis, Mo.). The doses of SG and pACTH used were those which would
give a stimulatory release of estradiol from ovaries (unpublished
observations) and cortisol from anterior kidneys (Patric) et at. 1986),
respectively, in juvenile coho salmon. At the end of the incubation
period, the media samples were frozen until analysis for steroids by
radioimmunoassay.
99
Assays. For each steroid measured, a 3001.11, aliquot of sample
media was extracted twice in diethyl ether and reconstituted in
phosphate buffered saline containing gelatin. The extraction
efficiencies for all steroids were above 90%. Aliquots of the
reconstituted extract were then removed to culture tubes and assayed
by radioimmunoassay for estradiol, testosterone, and androstenedione
using the method of Fitzpatrick et at. (1986) and for cortisol using the
method of Redding et at. (1984).
Statistics. The steroid levels differentiated by the various
treatment groups (and where possible, by sex) were analyzed by
Kruskal-Wallis analysis of variance within each date. Groups
demonstrating significant differences were then analyzed by a MannWhitney U test. Definitive statistical significance was set at p < 0.05.
Results
Explants taken on 56, 70, 86, 103, and 136 days postfertilization (DPF) all produced measurable amounts of steroids (Figs.
25-28). Within a week of hatch (56 DPF), more cortisol and
androstenedione were produced than testosterone and estradiol.
There was no significant stimulation of steroid production by either
SG or pACTH. Just before first feeding (70 DPF), gonads and anterior
kidneys treated with SG produced significantly more testosterone
than controls. A trend towards stimulation of estradiol levels in SG
treated explants appeared although it was not statistically significant
(p < 0.08). There were no differences between controls and the levels
100
Figure 25. Mean levels of cortisol in media from incubations of coho
salmon gonads and anterior kidneys treated with partly purified
salmon gonadotropin (SG) or porcine adrenocorticotropin (pACTH).
Values indicated are means (± S.E.)
d
1500
1000 500
0
I
ri rn No
56
70
Yd
no data
n
86
103
Days post-fertilization
Figure 25.
Control
SG
III pACTH
136
102
Figure 26. Mean levels of androstenedione in media from incubations
of coho salmon gonads and anterior kidneys treated with partly
purified salmon gonadotropin (SG) or porcine adrenocorticotropin
(pACTH). Values indicated are means (± S.E.). Means marked by "*"
indicate significant stimulation by SG over controls of the same sex (if
genders could be distinguished on that date). Means marked by "a"
indicate significant differences between sexes under similar
incubation conditions on that date.
350
300
250
*a
6
200
Control
150
a9
9
100
pACTH
50
9
0
56
70
86
103
Days post-fertilization
Figure 26.
El SG
136
104
Figure 27. Mean levels of testosterone in media from incubations of
coho salmon gonads and anterior kidneys treated with partly purified
salmon gonadotropin (SG) or porcine adrenocorticotropin (pACTH).
Values indicated are means (± S.E.). Means marked by "*" indicate
significant stimulation by SG over controls of the same sex (if genders
could be distinguished on that date). Means marked by "a" indicate
significant differences between sexes under similar incubation
conditions on that date.
`a
*
Control
El SG
pACTH
Days post-fertilization
Figure 27.
106
Figure 28. Mean levels of estradiol in media from incubations of coho
salmon gonads and anterior kidneys treated with partly purified
salmon gonadotropin (SG) or porcine adrenocorticotropin (pACTH).
Values indicated are means (± S.E.). Means marked by "*" indicate
significant stimulation by SG over controls of the same sex (if genders
could be distinguished on that date). Means marked by "a" indicate
significant differences between sexes under similar incubation
conditions on that date.
195
155
115
75
Control
SG
35
pACTH
*a
a
56
70
86
103
Days post-fertilization
Figure 28.
136
108
of steroids produced in pACTH treated tissues. Just after first feeding
(day 86), no significant differences were observed in steroids
produced although a trend towards stimulation of estradiol was seen in
both SG and pACTH groups (p < 0.09). The levels of testosterone and
estradiol produced in all incubations were lower than those observed
at 70 DPF.
By 103 DPF, ovaries and testes could be distinguished from one
another under the dissecting microscope. Testosterone levels were
higher in control males than in control females and were higher in
SG-treated males than in SG-treated females. There was no
stimulation of testosterone by SG in either males or females. Estradiol
levels were almost significantly (p < 0.09) higher in control females
than in control males and were definitively higher in SG-treated
females than in SG-treated males. SG-treated females also showed
higher levels of estradiol than control females. When sex was not
taken into account, SG stimulated production of cortisol over that
observed in controls, but pACTH did not.
By 136 DPF, estradiol, testosterone, and androstenedione all
varied significantly (cortisol was not measured). Estradiol was
significantly higher in control females than in control males and
higher in SG-treated females than in SG-treated males. SG stimulated
production of estradiol over control levels in females. Testosterone
and androstenedione were higher in control males than in control
females and higher in SG-treated males than in SG-treated females.
SG stimulated the production of testosterone in males and
androstenedione in males and females relative to controls.
109
Discussion
An in vitro system has been developed for the combined
incubation of gonads and anterior kidneys (which contain the
interrenal cells) in coho salmon. Anti la (1984) and van den Hurk et al.
(1982) found only limited synthesis of steroids in developing rainbow
trout using incubations of radioactive precursor steroids followed by
chromatography. Before the onset of feeding, the steroids that were
produced appeared to be restricted to progestins. In coho salmon,
this does not appear to be the case as androstenedione, testosterone,
and estradiol were all detected by radioimmunoassay before first
feeding. It is not clear if the differences between studies is due to
species differences or differences in methods employed for detection.
Experiments with incubations of rainbow trout tissues indicate that at
least androstenedione is produced before first feeding (Chapter V).
There also appears to be an increase in the resting output of all
steroids measured between hatching and first feeding, although just
after first feeding the production appears to decrease. The pattern
and timing of steroid production is very similar to that reported by
Feist et al. (1990) for the whole-body steroid contents of coho salmon.
The decrease following the onset of feeding could be due, in part, to
negative feedback of steroids on the pituitary or hypothalamus as the
pelletized salmon diet contains considerable amounts of steroids (Feist
and Schreck 1990).
Van den Hurk (1982) found immunocytochemical evidence for
pituitary gonadotropin in rainbow trout 50 days after fertilization (17
110
days after hatching). The gonadotropin cells increased in number and
size by 100 days after fertilization. This study did not attempt to
answer the question of gonadal sensitivity to gonadotropin during early
development. Our results indicate that after hatching but before first
feeding, coho salmon gonadal and anterior kidney incubations showed
sensitivity to SG in the production of testosterone and perhaps
estradiol. After first feeding, sensitivity to SG decreased with only
estradiol showing a trend towards stimulation. By the time that the
sex of the gonads could be distinguished through the dissecting
microscope (103 DPF), female explants were sensitive to SG
stimulation for the production of estradiol and male explants showed a
trend towards sensitivity for the production of testosterone. By 136
DPF, SG stimulated estradiol and androstenedione in the female
explants, and testosterone and androstenedione in the male explants.
The lack of consistent stimulation of cortisol production by pACTH
(especially in the later experiments) may be due to excessive
trimming of the tissues. In the larger fish, the explants were cut to
minimum sizes to facilitate incubation in 1 mL of medium and,
therefore, interrenal tissue may have been inadvertently lost.
These results strongly suggest that the hypothalamo-pituitarygonadal and -interrenal axes may be developed very early in
development. We have demonstrated the presence of gonadal and
anterior kidney sensitivity to a partially purified gonadotropin very
early in development. It remains to be seen if steroids are the primary
inducers of sex differentiation.
111
Acknowledgements
The funds for this study were partly provided by the U.S.
Department of Agriculture, under grant No. 86-CRSR-2-2876.
112
V. IN VITRO STEROID SECRETION DURING EARLY DEVELOPMENT
OF RAINBOW TROUT:
ONSET OF PITUITARY CONTROL AND
EFFECTS OF DIETARY STEROID TREATMENT
Martin S. Fitzpatrick
Carl B. Schreck
Oregon Cooperative Fishery Research Unit', Oregon State University,
Corvallis, Oregon 97331
1Cooperators are the Fish and Wildlife Service, Oregon Department of
Fisheries and Wildlife, and Oregon State University.
113
Abstract
Sex differentiation in many teleost species can be controlled by
treatment with steroids. To investigate the development of
steroidogenesis during both natural and controlled sex differentiation,
the production of androstenedione, testosterone, and estradiol were
determined in tissues from populations of all-female and all-male
rainbow trout (Oncorhynchus mykiss). At various times from hatching
through gonadal sex differentiation, explants of steroidogenic tissues
were incubated in vitro alone or in the presence of partly purified
salmon gonadotropin (SG) and the resulting media were assayed for
steroids. Androstenedione was produced at higher levels than any
other steoid measured, was stimulated by SG, and was produced
primarily by the interrenal at the earliest stages. Both resting and SG-
stimulated production of testosterone and estradiol were sexually
distictive in differentiated gonads. Dietary treatment of rainbow trout
with either estradiol or 17a-methyltestosterone (MT) inhibited in
vitro gonadal steroid production--an effect that persisted in MT-fed
fish even after withdrawal of dietary steroids.
114
Introduction
Sex determination can be defined as the events (usually viewed
as genetic) that fix the sex of the germ cell and gonad (Austin et al.
1981); whereas sex differentiation can be viewed as the development
of a phenotype that is indicative of a particular sex (McCarrey and
Abbott 1979). The presence or absence of the Y chromosome may
determine the sex in some vertebrates, but the sexual phenotype
(which includes the gonads and the accessory sex organs) results from
a developmental cascade of biochemical events that involves more than
just the Y chromosome. In mammals, the development of the sex of
an individual requires a translation of genetic sex into gonadal sex. In
some species of reptiles and fishes that rely on temperature-
dependent sex determination, genetic sex may have little meaning
(Crews and Bull 1987); nevertheless, the actual differentiation of the
gonad probably requires the activation of genes. The control of
gonadal sex differentiation has attracted and continues to attract the
interest of many researchers. Steroids can direct gonadal
development to the phenotype opposite that of the genotype in many
non-mammalian vertebrates, which led Yamamoto (1969) to postulate
that steroids may be the natural inducers of gonadal sex differentiation
in fishes. However, steroids cannot completely invert the sex of most
mammalian gonads (Jost 1965; Wilson et al. 1980) and recent work
has focused on single gene differences between males and females
(Page et al. 1987). In the many teleosts in which steroid-treatment
does cause sex-inversion (see Hunter and Donaldson 1983), it is
unclear if the effect has pharmacological or truly physiological causes.
115
Rainbow trout (Oncorhynchus mykiss) are characterized by
female homogamety and male heterogamety similar to most
mammalian species (Yamazaki 1983). Gonadal sex differentiation in
rainbow trout reportedly occurs near the time of yolk sac absorption
and the onset of first feeding (van den Hurk and Slof 1981). Recently,
Feist et al. (1990) found that whole-body levels of steroids are
bimodally distributed in coho salmon (0. kisutch) around the time of
first feeding, suggesting that sexual differences in steroids may exist at
the time of sexual differentiation. In rainbow trout, the capacity of
gonads to produce steroids from precursors has been observed very
early in development (Anti la 1984); however, the ability of gonads to
synthesize androgens and estrogens from precursors reportedly does
not occur until after gonadal differentiation (van den Hurk et al. 1982).
The latter studies used incubations of homogenized tissues in the
presence of radioactive precursors--a technique that attends to the
question of synthetic potential but does not address the question of
natural production. Furthermore, the use of bisexual populations in
these studies limited the ability to discriminate between males and
females until after gonadal differentiation.
The following study describes the in vitro production of steroids
in resting and gonadotropin-stimulated tissues from monosex
populations of rainbow trout and describes the effects of dietary
steroid treatment (for sex inversion) on steroid production. The
objectives of the study were to determine: when, during early
development, the gonads and interrenal begin to secrete androgens
and estrogens; when the gonads become sensitive to gonadotropic
control; and whether dietary steroid treatment affects gonadal
116
steroidogenesis. The gender of individual rainbow trout can be
discriminated through karyotyping (Thorgaard 1977). However, to
avoid the difficulty of karyotyping each sample and to ensure equal
sample sizes for each experiment, we created monosex populations of
rainbow trout; thus, the sexes for entire experimental groups were
known prior to gonadal sex differentiation.
Materials and Methods
Animals. Eggs from Shasta strain rainbow trout (obtained
courtesy of Dr. Jerry Hendricks, Food Science Laboratory, Oregon
State University) were collected in early December 1987 and shipped
to Dr. Gary Thorgaard, Washington State University, for production of
monosex populations. All-female populations were produced by
gynogenesis using the method of Thorgaard and Allen (1987), which
entailed fertilizing eggs with Atlantic salmon (Salmo salar) milt
inactivated with ultraviolet light followed 10 min after fertilization by a
10 min heat shock (290C), which blocked the second meiotic division
and restored diploidy to the embryos. All-male populations were made
by fertilizing eggs with milt from an androgenetically-produced male
that had sired all-male offspring the previous year. Androgenesis was
induced using the method of Scheerer et al. (1986), which is similar
to that for gynogenesis except that the eggs instead of the sperm were
inactivated with ultraviolet light. The eyed eggs were returned to
Oregon State University where they were maintained in flow-through
troughs (water temperature varied from 8 to 110C during the
experiments). After hatching, the all-male and all-female populations
117
were each further separated into three equal-sized groups. At the
onset of feeding, each group received an Oregon Moist Pellet-based
diet. Experimental diets were made by spraying estradiol (E2; 5 ppm)
or 17a-methyltestosterone (MT; 3 ppm) dissolved in EtOH (the
control diet was made by spraying EtOH alone) onto pellets and then
allowing the EtOH to evaporate. The fish were fed up to six times per
day during early stages of development and at least twice a day
thereafter for 11 weeks at which time all groups were placed on a
non-EtOH-treated control diet. At six months after hatching, 50 fish
from each experimental group were killed by overdose with anesthetic
(tricaine methansulfonate, MS-222) and the gonads were examined to
determine sex ratios.
In vitro experiments. At five times during development--55, 78,
95, 117, and 147 days post-fertilization (DPF)--fish were killed by
overdose of anesthetic and held on ice during dissection. Dissection
on 55 DPF consisted of removal of the yolk sac, viscera, head and tail
from each fish leaving an explant of tissue that consisted of the
anterior kidney (which contained interrenal tissue) and gonads which
was placed in 1 mL of incubation media (Patin et al. 1986) in a 24
well plate. The presence of gonads could not be verified under the
dissection scope, but in separate experiments using coho salmon,
dissections of this sort revealed head kidney and gonads in histological
sections (unpublished observations). On 78, 95, 117, and 147 DPF,
gonadal and anterior kidney explants from each fish were placed in
separate wells containing media. After the last dissections, the plates
were placed on a Junior Orbit shaker (100 rpms) in a constant
temperature room (150C) for two 1.5-hr intervals. After the first
118
interval, the tissues were removed to wells containing fresh media.
After the second interval, the tissues were removed to wells
containing either fresh media or fresh media with 100 ng of partly
purified salmon gonadotropin (SG; courtesy of Dr. E.M. Donaldson,
West Vancouver Laboratory, Department of Fisheries and Oceans, West
Vancouver, B.C.). Media used during the incubation period contained
1 mM NAD and NADPH. Tissues were incubated for 18 hr at the end
of which media aliquots were removed and stored frozen (-200C) until
analysis for steroids.
Radioimmunoassays. The levels of androstenedione (An),
testosterone (T), and estradiol (E2) were determined in 300 [IL
aliquots using methods similar to those described in Schreck et al.
(1989) for An and Fitzpatrick et al. (1986) for T and E2. Briefly, the
aliquots were extracted twice in excess diethyl ether and the
evaporated extract was reconstituted in 200 !IL of phosphate-buffered
saline containing gelatin (PBSG). Standards of authentic steroids (in
EtOH) from 1.25 to 250 pg were evaporated to dryness in media
blanks (that had been put through the extraction process) and then
reconstituted in PBSG. Samples and standards were incubated with
100 pi, of antisera for 30 mM at room temperature. All tubes were
incubated with 100 'IL of 3H-steroid overnight at 40C. The tubes were
placed on ice and 1 mL (An and T) or 0.5 mL (E2) of dextran-coated
charcoal was added to each. After 10 mM, the tubes were centrifuged
(2200 x g) for 20 min and 0.5 mL of each tube was decanted into a
scintillation vial and counted. The antisera for the An assay
(Endocrine Sciences, Tarzana, CA) was diluted 1: 600 in PBSG, for the
T assay (Endocrine Sciences, Tarzana, CA) was diluted 1:1000 in
119
PBSG, and for the E2 assay (Dr. G. Nishwender, Colorado State
University) was diluted 1:60,000. 3H-steroids were diluted in PBSG to
obtain approximately 10,000 (An) or 15,000 (T and E2) dpm/100
The average extraction efficiency was 96.9% for An, 82.0% for T, and
87.6% for E2. The concentrations of steroid from serial dilutions of
sample media were parallel to the standard curve for each steroid.
Histology. On 55, 78, 117, and 147 DPF, 3 to 5 fish from each
experimental group were killed by overdose with MS-222,
decapitated, and then fixed in buffered formalin. The samples were
imbedded in paraffin and 5 pm thick sagital sections were made
through each sample. The sections were stained with hematoxylineosin.
Results
In vitro steroid production in untreated fish
Androstenedione. On 55 DPF, the production of An in the
combined gonad-anteriod kidney incubations was stimulated by SG in
females, but not in males (Fig. 29). On 78 DPF, male gonads produced
significantly more An than female gonads, but neither group
demonstrated stimulation by SG. On 95 and 117 DPF, male gonads
were stimulated by SG to produce An, but female gonads were not. No
differences between sexes in resting output were observed on these
dates. On 147 DPF, resting end SG-exposed testes produced more An
than similarly treated ovaries; SG-exposed testes produced more An
than resting testes. Interrenal production of An
120
Figure 29. Mean levels of androstenedione in media from incubations
of rainbow trout tissue explants (n= 6 to 9 for each treatment). The
fish hatched 38 day post-fertilization (DPF) and began to feed between
59 and 63 DPF. Explants were incubated in either control media or
media containing partly purified salmon gonadotropin (SG). "9" and
"d" indicate animals destined to be females and males, respectively.
"*" indicate significant (p < 0.05) stimulation of steroid production by
SG; "a" indicates significant (p < 0.05) difference between sexes under
similar incubation conditions (i.e. control or SG).
Gonads
500
Combined Gonads/Anterior kidneys
400
300
70
60
0
'a :5 50
a)
200
100
t) 40
Thaw -jE
E
0
30
20
500
io
Anterior
kidneys
9
400
0
55
Control
MI SG (100 ng/mL)
300
200
100
no
data
no
data
117
147
0
78
Days post-fertilization
Figure 29.
95
122
was stimulated by SG on 78 and 95 DPF, and there were no
differences between sexes.
Testosterone. In combined gonad and anterior kidney
incubations, T was non-detectable on 55 DPF (Fig. 30). On 78 DPF, T
was first detected in gonadal incubations from both sexes, but no sex
differences in resting T production appeared until 147 DPF at which
time unstimulated testicular output was higher than unstimulated
ovarian output. Secretion of T was stimulated by SG from testes on 95,
117, and 147 DPF, but never from ovaries. Female anterior kidneys
demonstrated SG-stimulated T production at both 78 and 95 DPF,
whereas male anterior kidneys did not. However, no differences were
observed when either resting or SG-stimulated T output was
compared between sexes.
Estradiot. On 55 and 78 DPF, E2 production by explants was
non-detectable (Fig. 31). Ovarian output of E2 was barely detectable
on 95 DPF but testicular production remained non-detectable. On 117
and 147 DPF, both resting and SG-stimulated E2 levels were higher
from ovaries than from testes, and ovaries demonstrated stimulation
by SG, whereas testes did not. Anterior kidney production of E2 in
both sexes was extremely low or non-detectable at all times measured.
In vitro steroid production in steroid-treated fish
Androstenedione. The resting production of An from testes was
significantly lower in MT-fed fish on 117 and 147 DPF and in E2-fed
fish on 147 DPF compared with control-fed fish (Fig. 32). Levels
tended to be lower on all other dates; however, the differences were
not statistically significant. There was no effect of dietary steroids on
123
Figure 30. Mean levels of testosterone in media from incubations of
rainbow trout tissue explants (n= 6 to 9 for each treatment). The fish
hatched 38 day post-fertilization (DPF) and began to feed between 59
and 63 DPF. Explants were incubated in either control media or
media containing partly purified salmon gonadotropin (SG). "9" and
"d" indicate animals destined to be females and males, respectively.
"*" indicate significant (p < 0.05) stimulation of steroid production by
SG; "a" indicates significant (p < 0.05) difference between sexes under
similar incubation conditions (i.e. control or SG).
as
Gonads
**
200
Combined Gonads/Anterior kidneys
150
a
100
**
a
50
cc
9d
NAV
0
2. I)
of
17;
Anterior
kidneys
9
'-7,*
30
CD
9
55
IE1
Zi
Control
SG (100 ng/mL)
20
10
0
Figure 30.
**
78
Days post-fertilization
95
117
147
125
Figure 31. Mean levels of estradiol in media from incubations of
rainbow trout tissue explants (n= 6 to 9 for each treatment). The fish
hatched 38 day post-fertilization (DPF) and began to feed between 59
and 63 DPF. Explants were incubated in either control media or
media containing partly purified salmon gonadotropin (SG). "9" and
" indicate animals destined to be females and males, respectively.
"*" indicate significant (p < 0.05) stimulation of steroid production by
SG; "a" indicates significant (p < 0.05) difference between sexes under
similar incubation conditions (i.e. control or SG).
a
a
Gonads
200
Combined Gonads/Anterior kidneys
150
100
10
a
50
0
30
0
55
Anterior
kidneys
20
Control
SG (100 ng/mL)
10
e
0
78
Days post-fertilization
Figure 31.
9
d
95
117
147
127
Figure 32. Mean levels of androstenedione in media from incubations
of rainbow trout gonads (0 = control media; G = media containing 100
ng of SG). Fish were fed diets containing steroids for 76 days from the
onset of feeding (63 DPF) through 139 DPF (n = 6 for each treatment).
Concentrations of steroids in diets were 5 and 3 µg /kg for estradio117J3 and 17a-methyltestosterone, respectively. An "*" indicates
steroid levels significantly (p < 0.05) different from levels in control
diet fish; "a" indicates significant (p < 0.05) stimulation of steroid
levels by SG within the same dietary treatment.
G
1100
Males
Females
900
m
c
o
700
.21
c ? 400
cD
_
400
Diet
a
0 Control
G
300
El Estradiol
III Methyltestosterone
tr; E
9.)300
c
ct
200
a
-c3
G
200
.---
100
100
0
0
78
95
117
147
78
95
Days post-fertilization
Figure 32.
117
147
129
the resting output of An from ovaries on any of the dates. Production
of An from testes incubated with SG was lower in MT- and E2-fed fish
compared with control-fed fish at all times tested; however, ovarian
An output in response to SG was not affected by dietary steroid
treatment. Interrenal production of An in both resting and SGstimulated incubations was not affected by dietary steroid treatment
(Fig. 33).
Testosterone. In comparison to control-fed fish, the resting
production of T from testes was significantly lower in MT-fed fish at
95, 117, and 147 DPF and in E2-fed fish at 147 DPF (Fig. 34). No
significant differences were detected in resting T output from ovaries.
Production of T from SG-exposed testes was lower in MT- and E2-fed
fish compared with control-fed fish at 95, 117, and 147 DPF. The
output of T from SG-exposed ovaries did not differ between treatment
groups.
Estradiol. The resting production of E2 from ovaries was
significantly lower in MT-fed fish at 95, 117, and 147 DPF and in E2-
fed fish at 117 DPF compared with control-fed fish (Fig 35). No effect
of dietary steroid treatment was observed on testicular E2 production.
Levels of E2 from SG-exposed ovaries were significantly lower in MTfed fish at 95, 117, and 147 DPF and in E2-fed fish at 95 and 117 DPF
compared with control-fed fish. Dietary steroid treatment did not
affect E2 production from SG-exposed testes.
Histology
On 55 DPF, the gonads of fish from all-female and all-male
populations were undifferentiated long thin structures with few
primordial germ cells (Fig. 36a and 36b). On 78 DPF, the female
130
Figure 33. Mean levels of androstenedione in media from incubations
of rainbow trout anterior kidneys (0 = control media: G = media
containing 100 ng of SG). Fish were fed diets containing steroids for
76 days from the onset of feeding (63 DPF) through 139 DPF (n = 6 for
each treatment). Concentrations of steroids in diets were 5 and 3
4/kg for estradiol -17j3 and 17a-methyltestosterone, respectively. An
"a" indicates significant (p < 0.05) stimulation of steroid levels by SG
within the same dietary treatment.
as
500
Females
500
Males
a
Diet
o
a)
C
'17:31'
a)
C
400
Estradiol
300
111
300
Mothyltestosterone
E
_
17; E 200
c
400
El Control
100
aaa
aaa
aaa
200
G
G
.....=
G
100
0 i,
......,
,/,
no data
no data
0
,
78
95
117
147
0
rell
78
0
r
95
Days post-fertilization
Figure 33.
no data
no data
117
147
132
Figure 34. Mean levels of testosterone in media from incubations of
rainbow trout gonads (0 = control media; G = media containing 100
ng of SG). Fish were fed diets containing steroids for 76 days from the
onset of feeding (63 DPF) through 139 DPF (n = 6 for each treatment).
Concentrations of steroids in diets were 5 and 3 14/kg for estradiol175 and 17a-methyltestosterone, respectively. An "*" indicates
steroid levels significantly (p < 0.05) different from levels in control
diet fish; "a" indicates significant (p < 0.05) stimulation of steroid
levels by SG within the same dietary treatment.
a
G
Males
200
,......,
200
Diet
0 Control
--,-,
O:)... 13
150
___
150
Estradiol
II Mothyltestosterone
0 a)
ti;
rn
100
100
G
0E
ti -a
a 50
H
1
0
a
a
G
0 ,--11 11
11 11
78
95
50
le
1
117
147
0
78
Days post-fertilization
Figure 34.
95
117
147
134
Figure 35. Mean levels of estradiol in media from incubations of
rainbow trout gonads (0 = control media; G = media containing 100
ng of SG). Fish were fed diets containing steroids for 76 days from the
onset of feeding (63 DPF) through 139 DPF (n = 6 for each treatment).
Concentrations of steroids in diets were 5 and 3 .1,g/kg for estradiol17J3 and 17a-methyltestosterone, respectively. An "*" indicates
steroid levels significantly (p < 0.05) different from levels in control
diet fish; "a" indicates significant (p < 0.05) stimulation of steroid
levels by SG within the same dietary treatment.
as
a
Males
200
G
Females
Diet
200
Control
150
Estradiol
150
111 Mothyltestosterone
100
4(7)
E
W
-EL"
100
G
50
G
G
G
II
0
78
95
117
;I;
147
50
QJ G
72.
.-
2
0
.
78
Days post-fertilization
Figure 35.
2/
G
95
117
i/11.
147
i
136
Figure 36. Histology of gonads at various times after fertilization from
rainbow trout fed a control diet. "9" and "d" indicate animals
destined to be females and males, respectively. All photographs were
taken at 400x power except for i and j which were taken at 100x
power.
9
Figure 36.
138
gonads contained mostly stromal cells and the few germ cells had
light cytoplasms with darker nuclei (Fig. 36c). Most of the male
gonadal sections looked similar to those from females; however, other
sections contained cells with mottled cytoplasms (Fig. 36d). On 117
DPF, female gonads contained many clear germ cells and had
distinctly lamellar structures (Fig. 36e) whereas the male gonads had
no lamellar structures and contained many germ cells arranged in
cysts (Fig. 360. On 147 DPF, the female gonads contained many large
perinucleolar oocytes and had extremely well-developed lamellar
structures (Fig. 36g and 1); the male gonads looked similar to those
from 117 DPF but contained many more germ cells (Fig. 36h and j).
Females treated with E2 had gonads similar to those of controlfed females on all dates (Fig 37a-c); gonads of E2-treated males had
occasional large oocytes on 147 DPF but contained mostly germ cells
similar to those in control-fed males (Fig. 37g-i). Gonads from MT-fed
females and males showed progressive loss of germ cells and
increased areas of connective tissues (Fig. 37d-f; 37g-i). In females,
the loss of germ cells was nearly complete whereas in males, germ
cells were still apparent on 147 DPF.
Discussion
We have demonstrated that both interrenal and gonadal tissues
are capable of producing steroids very early in development of rainbow
trout. Van den Hurk et al. (1982) reported that only progestins could
be produced from pregnenolone in the undifferentiated gonads of
rainbow trout, and that androgens and estrogens were synthesized
139
Figure 37. Histology of gonads at various times after fertilization from
rainbow trout fed diets (between 63 and 139 days post-fertilization)
containing 3 µg /kg of estradiol (E2) or 5 14/kg of 17amethyltestosterone (MT). "?" and "e" indicate animals destined to be
females and males, respectively. All photographs were taken at 400x
power.
78
9
di
Figure 37.
117
147
141
from precursors only well after sex differentiation. In the current
study, undifferentiated gonads similar to those used by van den Hurk
et al. (1982) were incubated together with head kidneys. The amount
of An produced in separate incubations of gonads and head kidneys on
78 DPF and results of others (van den Hurk et at. 1982) suggest that
the An produced on 55 DPF came predominantly from interrenal
tissue and not from the undifferentiated gonads. The first indications
of gonadal sex differentiation have been observed near the time of
initial feeding in rainbow trout (Takashima et at. 1980; van den Hurk
and Slof 1981; Le Brun et at. 1982) and coho salmon (Piferrer and
Donaldson 1989). In the current study, androstenedione and
testosterone were produced at detectable levels in incubations of
gonads taken shortly after the onset of feeding, but estradiol was not.
Furthermore, gonadal An production was higher from males than from
females. Histological examination of the gonads at this time revealed
only slight differences between males and females; however, the small
number of samples examined may have limited the detection of
differentiation in a majority of each population. Differences analogous
to those reported in the literature for trout and salmon were apparent
at 117 DPF and sex differences in An, T, and E2 production were
marked by this stage of development.
The testicular secretion of An was greater than that of T at all
times. The significance of the early production of An remains unclear,
but the characterization of An as an atypical testicular androgen (van
den Hurk and van Oordt 1985) may have been too hasty. An is
certainly produced in large amounts relative to the other steroids that
we measured in rainbow trout and coho salmon (see Chapter IV). Van
142
den Hurk et al. (1982) reported that before differentiation the gonads
could not bioconvert An, but that after differentiation the testes could
convert it to 1113-hydroxyandrostenedione. Treatment with 11_5hydroxyandrostenedione has been reported to masculinize rainbow
trout, but treatment with An has no effect (van den Hurk and van
Oordt 1985). Our results suggest that interrenal cells within the
anterior kidney secrete An before the gonad does--and before the
onset of feeding, which is the time reported for gonadal differentiation
(Takashima et al. 1980; van den Hurk and Slof 1981; Le Brun et al.
1982). We did not determine the activity of 1113-hydroxylase, but
previous experiments with coho salmon yolk sac fry (Chapter IV)
demonstrated the capacity of explants containing anterior kidney and
gonads to produce cortisol which also requires 1113-hydroxylase
activity. The elevated production of T that we observed in males
compared to females may be a consequence rather than a cause of
differentiation, as T is not particularly effective in masculinizing
salmonids (Hunter and Donaldson 1983; Redding, Fitzpatrick, Feist,
and Schreck, unpublished data) and the antiandrogen, cyproterone
acetate, has no effect on sex differentiation in rainbow trout (van den
Hurk and van Oordt 1985) or tilapia, Oreochromts aureus (Hopkins et
al. 1979). Females produced detectable amounts of E2 before males
did, but only several weeks after hatching, which agrees with the
observation that ovaries bioconvert An to T and estrogens only well
after differentiation (van den Hurk et al. 1982). Interestingly, E2 is
effective in feminizing salmonids only when administered very early in
development (Hunter and Donaldson 1983; Piferrer and Donaldson
1989; Redding, Fitzpatrick, Feist, and Schreck, unpublished data).
143
Gonadal sex differentiation involves the development of both
germinal and somatic elements. Jost (1965) reported that in
mammals, somatic cells differentiate before germ cells. Similar
findings have been reported for the teleost Xiphophorus maculatus
(Kramer and Ka llman 1985). In other teleosts such as the medaka,
Oryzias latipes, germ cells show the first indications of differentiation
(Satoh and Egami 1972; Quirk and Hamilton 1973). Changes in germ
cell numbers (Le Brun et a/. 1982) or the presence of meiotic germ
cells (Takashima et a/. 1980) have been cited as the initial indication
of gonadal differentiation in salmonids. We observed meiotic germ
cells within lamellar structures at 117 DPF, which suggests
concomitant development of germinal and somatic elements. This is
in agreement with van den Hurk and Slof (1981); however, our
observations came much later than the onset of feeding, which
Takashima et at. (1980) and Le Brun et a/. (1982) reported as the time
of gonadal sex differentiation. In many mammals, germ cell survival
requires that the surrounding somatic elements share the same
genetic complement, i.e. )0( germ cells die within an XY gonad
(Haseltine and Ohno 1981). Such observations suggest that germ cell
development requires that the adjacent somatic cells must undergo
some degree of differentiation that either supports or inhibits the
development of the germ cells. On the other hand, proper germ cell
development is required for complementary development of somatic
tissue in triploid salmonids where females develop only rudimentary
gonads, but males develop substantial gonads (Thorgaard and Gall
1979). Conceivably, the early unsuccessful entry into meiosis prevents
144
the triploid ovary from developing and the delayed entry into meiosis
allows considerable development in triploid testes.
Histochemical and chemical techniques suggested an early onset
of steroidogenic capacity in embryonic vertebrate gonads, but always
after differentiation of the gonads (Chieffi 1965). The more precise
methods that have been developed for determining steroidogenic
potential have not succeeded in reversing this premise. In rabbits, E2
and T synthesis commence shortly before and after, respectively,
histological differentiation of ovaries and testes (Wilson et al. 1980),
with testicular differentiation occurring well before ovarian
differentiation. In contrast, ovarian differentiation manifests itself
before testicular differentiation in salmonids (Le Brun et al. 1982;
Piferrer and Donaldson 1989). However, similar to findings in
mammals, the current study demonstrated that androgen production
appeared earlier than estrogen production.
The sensitivity of mammalian gonads to gonadotropin (GtH)
stimulation does not occur until shortly after the onset of
steroidogenic capacity (Jost 1965). Our findings suggest that rainbow
trout gonads display a similar period of GtH-insensitivity. Van den
Hurk (1982) reported the presence of pituitary GtH in trout near the
onset of feeding which suggests that the lack of gonadal sensitivity may
be due to a dearth of GtH receptors, rather than of GtH itself. The
adult salmon pituitary glands from which SG is prepared contain
mostly "mature" GtH in contrast to the 'juvenile" GtH (Kawauchi et al.
1989). Therefore, the insensitivity may be an artifact of using a
"mature" as opposed to a 'juvenile" preparation, although both are
145
nearly equal in stimulating steroidogenesis in bioassays (Suzuki et al.
1988c).
Despite gonadal insensitivity, the interrenal appears sensitive to
GtH stimulation during this early period of development. In
vertebrates, adrenal secretion of sex steroids appears to be controlled
primarily by adrenocorticotropic hormone (ACTH) although
gonadotropic stimulation has also been reported (Kime et al. 1980).
SG is not a purified preparation; therefore, stimulation of An
production could be due to contamination with ACTH. However, at
other life stages, the salmonid interrenal produces An in response to
more purified forms of GtH (Schreck et at. 1989) and therefore, the
production of An that we observed in response to SG could truly be
due to gonadotropic stimulation. The role of interrenal
androstenedione remains unclear but other investigators have
suggested that interrenal androgen production may be important in
directing gonadal sex differentiation (van den Hurk and Leeman 1984;
van den Hurk and van Oordt 1985).
Although the sex-inverting effects of steroid treatment are well
documented (see Hunter and Donaldson 1983), little work exists on
the mechanism of action. The current study demonstrated that
dietary steroids decreased gonadal steroidogenesis. Furthermore,
steroid-treated fish did not show gonadal sensitivity to SG. The effect
appears to be gonad-specific as dietary steroids had no effect on
anterior kidney An production--either under resting or SG-stimulated
conditions. The suppression of steroidogenesis was reversible in E2treated animals which indicates no permanent damage to the
steroidogenic system. Animals treated with MT never recovered their
146
steroidogenic capabilities and considerable alteration of gonad
morphology was evident. Goetz et al. (1979) showed that prolonged
MT-treatment of coho salmon can cause sterilization. Preliminary
studies indicate that steroid treatment affected the production of sexspecific protein(s), although this effect may be transitory (Chapter VI).
The sex-inverting propensity of specific steroids suggests that
even if not the prime directors of the initial events of gonadal sex
differentiation, certain steroids may be required for maintaining
proper development. The difference between males and females in
the timing of germ cell entry into meiosis may be a key step in gonadal
differentiation. Steroids (namely progestins) carry out a critical
function in the resumption of meiosis during final maturation in adult
vertebrates (Masui and Clarke 1979) and possibly, the sexually
distinctive patterns of resting or stimulated steroidogenesis that we
observed influence the timing of the entry of germ cells into meiosis.
Acknowledgements
We would like to thank Sam Bradford, Cameron Sharpe, and
Reynaldo Patifio for providing excellent technical assistance during
the dissections of tissues. The funds for this study were partly
provided by the U.S. Department of Agriculture, under grant No. 86CRSR-2-2876.
147
VI. SEX-SPECIFIC GONADAL PROTEINS
IN SALMONIDS DURING DEVELOPMENT
Martin S. Fitzpatrick'
Cristobal L. Miranda2
Donald R. Buhler2
J. Michael Reddingl*
Carl B. Schreck1
1 Oregon Cooperative Fishery Research Unit or 2 Department of
Agricultural Chemistry, Oregon State University, Corvallis, OR 97331
* current address: Department of Biology, Tennessee Technological
University, Cookeville, TN
148
Abstract
The mechanism of action by which steroids cause sex inversion
and the relation of steroid action to natural sex differentiation remain
unclear. A complicated process such as differentiation probably
involves the differential production of proteins. Native protein gel
electrophoresis and sodium dodecylsulfate-polyacylamide gel
electrophoresis (SDS-PAGE) of gonadal homogenates from yearling
coho salmon (Oncorhynchus kisutch) revealed several sex-specific
bands. In particular, a prominent band of about 50,000 daltons was
apparent in the ovaries but not the testes after SDS-PAGE. A band of
similar molecular weight was detected in ovaries of rainbow trout (0.
mykiss). Treatment of trout with dietary steroids to cause sex
inversion resulted in the reduction or absence of this female-specific
protein in six-month-old female trout from a monosex population, but
the band was present in estradiol-fed, 13-month-old female trout from
a bisexual population. Thus, the effect of estradiol on this protein may
be transitory. Unreduced gonadal homogenates subjected to SDSPAGE had the female band in the same position as reduced gonadal
homogenates, which suggests that the female-specific protein is the
complete protein and not a subunit of some larger peptide.
149
Introduction
The physiology of sex differentiation requires a complicated
cascade of biochemical events. The details of these events remain
poorly understood. The nature of the inducing substance(s) has and
continues to be a rich area of inquiry. Witschi (1934) hypothesized
the existence of two hormone-like inducing substances responsible for
setting off the chain of events leading to female and male gonadal
development. Although some investigators postulated that sex
steroids were the endogenous sex inducers (Burns 1961; Yamamoto
1969), Witschi (1967) argued that the substances were proteins and
not steroids based on the inability of steroids to induce functional sex
inversion in mammals. The concept of dual inducing substances has
been deemed unnecessarily complicated; instead, sexual dimorphism
is believed to be the presence or absence of a single state termed the
dominant sex rather than two opposite states (McCarrey and Abbott
1979).
The discovery and further study of H-Y antigen--a protein
seemingly invariably associated with the heterogametic sex of many
vertebrates--offered hope that the inducer substance had been found
(Wachtel and Koo
However, sexual differences in H-Y antigen
were lacking in several species of teleosts (Muller and Wolf 1979) and
1981).
the antigen was absent in a laboratory strain of testes-bearing rats
(McLaren et al. 1984). Recently, the H-Y gene has been mapped to a
location outside that of the purported sex determining region of the Y
chromosome (Simpson
et al. 1987).
Such evidence has made the H-Y
150
antigen theory problematic--indeed, it seems to preclude H-Y antigen
from being the primary inducer.
The identification of the testes determining region of the Y
chromosome (ZFY--for "Zinc Finger protein on Y") in several
mammalian species has again raised hopes that an inducer substance
may shortly be found (Page et al. 1987). However, attempts to probe
for ZFY in a number of reptiles (Bull et al. 1988) and two species of
fish (Ferreiro et al. 1989) have revealed no differences in probe
hybridization between males and females.
The study of sexual differentiation in teleosts may benefit from a
different approach than those used in studying mammals. Instead of
initially searching for the differentiation switch, i.e. the inducer
substance(s), our strategy has been to look for early characteristic
effects of the switch being thrown, i.e. the early biochemical changes
associated with gonadal differentiation. We have previously observed
sex-specific in vitro steroid secretion from the gonads of rainbow trout
during early development (Chapter V). A complex phenomenon such
as differentiation likely involves the differential production of proteins.
The following experiments were undertaken to determine 1) if sex-
specific gonadal proteins could be detected during early development
of salmonid fish; 2) the location of such proteins within the cell; and
3) the effects of steroid treatment on the expression of such proteins.
The developmental plasticity of salmonids allowed us to make use of
techniques that create monosex populations and that cause functional
sex inversion. In this way, we were able to conduct our studies on
experimental animals whose destined gender was known even before
the onset of gonadal differentiation without the need for karyotyping
151
each animal. The studies were carried out on animals that had already
undergone gonadal sex differentiation to ascertain the feasibility of the
detection techniques.
Materials and Methods
Animals. Yearling coho salmon (Oncorhynchus kisutch) were
maintained at the Smith Farm Experimental Hatchery at 110C. These
animals had been obtained as eggs from the Oregon Aqua Foods
hatchery. The fish were immersed in 17a-methyltestosterone (500
pg/liter) for four hours twice weekly for a total of 8 times between
hatching and the initiation of feeding. To better define the effects of
exogenous steroids on sexual development (i.e. to separate the effects
of any steroid on a particular sex), monosex populations of rainbow
trout were created. Shasta strain rainbow trout (0. mykiss) eggs were
obtained from Dr. Jerry Hendricks, Food Sciences Laboratory, Oregon
State University and shipped to Dr. Gary Thorgaard, Washington State
University, for production of monosex populations. The techniques
used for creating all-male and all-female populations and holding
conditions have been described previously (see Chapter V). In
addition, a control population was maintained under similar
conditions. After hatching, the trout were separated into groups and
fed estradiol (5 µg /kg of food), 17a-methyltestosterone (3 µg /kg), or
control diets for the purposes of sex inversion. The water
temperature in which the fish were maintained varied between 8 and
110C.
152
Tissue sampling. Coho salmon were sampled at 11 and 13
months of age (time from fertilization). Rainbow trout from monosex
populations were sampled at 6 months and from the control
population at 17 months of age. At sampling, fish were killed by
overdose with the anesthetic tricaine methansulfonate (MS-222) and
held on ice. Gonads, livers, and head kidneys were removed to icecold Tris-acetate buffer (0.1 M Tris-acetate pH 7.5, 0.1 M KC1, 1 mM
EDTA, and 0.1 mM PMSF). The tissues were pooled by sex and tissue
type (e.g. all livers of males were pooled separately from livers of
females), dried, weighed, and then homogenized with a motor-driven
teflon piston. Each homogenate was stored frozen at -800C in 1.5 mL
microfuge tube after a determination of volume. The protein content
of each pooled homogenate was calculated using the method of Lowry
et al. (1951).
Subcellular fractionation. Gonad homogenates from rainbow
trout were fractionated using the method of Perdew et al. (1986).
Homogenates were centrifuged at 2,000 x g for 30 min at 40C in a
Beckman TJ6 refrigerated centrifuge. The pellets which contained
nuclear and membrane material were resuspended in Tris-acetate
buffer and stored frozen at -800C. The supernatants were centrifuged
at 12,000 x g for 30 min at 40C in a Beckman L8-60M Ultracentrifuge.
The mitochondrial pellets were resuspended in Tris-acetate buffer
and stored as described above. The supernatants were centrifuged at
100,000 x g for 1 hr. The cytosolic supernantants were stored frozen.
The pellets were resuspended in Tris-acetate buffer, gently
homogenized using a pasteur pipet and then centrifuged at 100,000 x
153
g for 1 hr. The supernatants were discarded and the microsomal
pellets were resuspended and stored as above.
Electrophoresis. Homogenates or subcellular fractions were
subjected to electrophoresis using the methods of Laemmli (1970) for
SDS-PAGE and Hames (1981) for native-protein electrophoresis. Both
8% and 12% polyacrylamide separating gels were used with 3%
stacking gels. The apparatus and chemicals used for electrophoresis
were from Bio-Rad. Briefly, homogenates were diluted at least by one
half with sample buffer (20% [v/v] of 10% SDS, 12.5% of 0.5 M TrisHCI pH= 6.8, 10% of glycerol, 5% 2J3- mercaptoethanol, and 2.5% of
0.1% Bromophenol Blue for SDS-PAGE; 12.5% of 0.5 M Tris-HC1 pH=
6.8, 10% of glycerol, and 2.5% of 0.1% Bromophenol Blue for native
protein electrophoresis) to obtain a final concentration of 2.5 µg /µL of
protein. Samples for SDS-PAGE were boiled for 4 min and then held
on ice. Samples were applied to the gels in 50 pi, aliquots when
possible (some testes samples were low in protein content and
therefore up to 100 p.1, was applied to the gel). Every SDS-PAGE gel
was calibrated with molecular weight standards ranging from 12,400
to 97,500 daltons. Separation was performed for 5-6 hr at 25 mA/gel.
The gels were stained with a solution of 0.15% Coomassie Blue, 50%
methanol, and 10% acetic acid for at least 1 hr and then destained
with 10% acetic acid and 5% methanol. The gels were fixed in 10%
acetic acid and 1% glycerol, photographed, and then dried.
154
Results
Sex-specific proteins in coho salmon. Sexually distinctive
patterns of banding were apparent from SDS-PAGE of gonadal
homogenates in 11 and 13 month old coho salmon. An extremely
prominent band was detected in ovarian but not testicular
homogenates (Fig. 38). This peptide had an estimated molecular
weight of 49,000-51,000 daltons. Both testes and ovaries had a
similarly prominent band corresponding to about 47,000 daltons.
These bands were present after SDS-PAGE in non-reducing
conditions. Treatment with 17a-methyltestosterone (MT) resulted in
fish with either testes or ovotestes. The banding pattern from
homogenates of testes from MT-treated individuals was similar to that
of control testes. Homogenates of ovotestes from MT-treated fish had
a banding pattern similar to that of control ovaries except the staining
was less dense. The banding patterns of liver and head kidney (not
shown) homogenates were different from that of the gonadal
homogenates in positions and density of staining.
The electrophoresis of native proteins in a high pH buffer
system also resulted in sex-specific banding (Fig. 39). A prominent
testicular band at a position different from several prominent ovarian
bands was readily apparent. All gonadal patterns were different from
those from male and female liver homogenates.
Sex-specific proteins in rainbow trout. Sex-specific banding
patterns of gonadal homogenates were detected in rainbow trout by
SDS-PAGE. A prominent band with a molecular weight between
49,000-50,000 daltons appeared in ovarian homogenates of 6 and 17
155
Figure 38. SDS-PAGE of tissue homogenates from coho salmon. Lane
1: molecular weight standards: Lane 2: testicular homogenates from
control diet fish; Lane 3: testicular homogenates from
methyltestosterone (MT) diet fish; Lane 4: homogenates of ovotestes
from MT diet fish; Lane 5: ovarian homogenates from control diet fish;
Lane 6: molecular weight standards; Lane 7: liver homogenates from
control diet females. Arrows indicate female-specific protein of about
50,000 daltons molecular weight.
156
Figure 38.
157
Figure 39. Native gel electrophoresis of tissue homogenates from coho
salmon. Lanes 1-3: ovarian homogenates from control diet fish; Lanes
4-6: testicular homogenates from control diet fish; Lane 7: liver
homogenates from control diet females; Lane 8: liver homogenates
from control diet males; Lane 9: bovine serum albumin. Arrows
indicate sex-specific proteins.
158
I
Figure 39.
23456789
159
month old rainbow trout from monosex and bisexual populations,
respectively (Fig. 40). Bands of about 35,000 and 53,000 daltons were
also seen in ovarian but not testicular homogenates and of about
70,000 daltons in testicular but not ovarian homogenates from the
older trout. Ovarian and testicular homogenates shared a band
corresponding to 46,000 daltons at both ages.
Subcellular fractionation of gonadal homogenates from 17 month
old trout revealed a multitude of bands from ovaries but not as many
from testes (Fig. 41), which may have been due to the low protein
content in the latter. The 49,000 dalton ovarian-specific band was
present in all ovarian fractions, but had the highest density of staining
in the microsomal fraction (followed closely by the mitochondrial
fraction). The 35,000 dalton ovarian band was most prominent in the
cytosolic fraction and the 53,000 dalton band was equally prominent
in mitochondrial and microsomal fractions. The 70,000 dalton
testicular band was observed in all testicular fractions. Two very
prominent testicular bands appeared at 27,000 and 30,000 daltons in
microsomal but not the other fractions, although this difference might
be due to low protein content in the other fractions.
Both MT and estradiol (E2) dietary treatment of all male and all
female populations resulted in the absence or reduction of the 49,000
dalton band in gonad homogenates from both sexes at 6 months of age
(Fig. 40). In contrast, E2-treated females from the bisexual population
had the 49,000 dalton band in ovarian homogenates at 17 months of
age. Visual inspection of the gonads of MT-treated gynogenetic
females at 2 years of age revealed well-developed testicular tissue in a
majority of the animals although some gonads also contained oocytes.
160
Figure 40. SDS-PAGE of tissue homogenates from rainbow trout. A.
Six-month old trout. Lane 1: molecular weight standards: Lane 2:
gonad homogenate from control-diet females; Lane 3: gonad
homogenate from control-diet males; Lane 4: gonad homogenate from
estradiol (E2)-diet females; Lane 5: gonad homogenate from E2-diet
males; Lane 6: gonad homogenate from methyltestosterone (MT)-diet
females; Lane 7: gonad homogenate from MT-diet males; Lane 8:
molecular weight standards; Lanes 9-10: anterior kidney homogenates
from control-diet females and males, respectively; Lanes 11-12: liver
homogenates from control-diet females and males, respectively. B.
17-month old trout. Lane 1: molecular weight standards: Lane 2:
ovarian homogenate from control-diet fish; Lane 3: testicular
homogenate from control-diet fish; Lane 4: ovarian homogenate from
E2-diet fish; Lane 5: testicular homogenate from E2-diet fish; Lane 6:
ovarian homogenate from MT-diet fish; Lane 7: molecular weight
standards. Arrows indicate female-specific protein of about 50,000
daltons molecular weight.
A
B
3
45
6
7
89
10
11
12
2345
1
6
7
kD
95.5
55
43
36 se
29 04,
95.5
55 010
43
4
29
36..
18.4.
Figure 40.
5
162
Figure 41. SDS-PAGE of subcellular fractions from rainbow trout (17month old control-diet fish). Lane 1: molecular weight standards: Lane
2: ovarian homogenate; Lane 3: testicular homogenate; Lane 4: ovarian
nuclear and membrane fraction; Lane 5: ovarian mitochondrial
fraction; Lane 6: ovarian cytosolic fraction; Lane 7: ovarian microsomal
fraction; Lane 8: testicular nuclear and membrane fraction; Lane 9:
testicular mitochondrial fraction; Lane 10: testicular cytosolic fraction;
Lane 11: testicular microsomal fraction; Lane 12: molecular weight
standards. Arrows indicate female-specific protein of about 50,000
daltons molecular weight.
163
23456789
1*--
kD
95.5 04
55
43
36
29 de
18.4
Figure 41.
10
11
12
164
The gonads from the E2-treated all-male population at 2 years of age
had well-developed testicular tissue.
Discussion
The electrophoretic techniques used in the study detected sexspecific gonadal proteins in both coho salmon and rainbow trout. A
particularly prominent ovarian band corresponding to about 50,000
daltons was found in both species. Treatment with steroids inhibited
the expression of this protein 2 months after the removal of the fish
from the steroid diets, but this band was apparent in females 13
months after E2 administration. Although the animals used in this
study had already undergone gonadal sex differentiation, the
electrophoretic technique holds promise in studying some of the
initial changes in protein synthesis during differentiation.
Page et al. (1987) reported that the sex determining regions of
the Y chromosome in mammals had been identified and that DNA
sequence analysis revealed the product of the ZFY gene to likely be a
Zn-finger protein. The hybridization pattern of the cDNA probe for
ZFY differs between males and females in a number of mammals (Page
et al. 1987); however, in chickens (Page et a/. 1987), several reptiles
(Bull et al. 1988), and two species of fish (Ferreiro et al. 1989), males
and females have the same hybridization pattern. This suggests that
the presence of ZFY alone may be insufficient for determining sex in
non-mammalian vertebrates. If ZFY is involved in the initial
differentiation of the gonad, the gene may require regulation by other
factors.
165
Detection and eventual identification of sex-specific gonadal
proteins may provide the endpoint for tracing the biochemical steps
involved in gonadal differentiation. In salmonids, germ cell
development has been described as synchronous (Wallace and Selman
1981), although it is important to note that this term was used to
describe oocytes, i.e. germ cells that had already undergone
differentiation. We have previously observed that in some steroidtreated salmon, the gonads contained both testicular and ovarian
tissue (unpubl. observations). Such mosaic development suggests that
at any time, the gonad contains both sensitive and insensitive regions
to sex-inverting steroid treatments. Therefore, differentiation may
not be synchronous occurring instead at different times throughout
the gonads. Nevertheless, if steroids modulate the expression of the
inducer gene(s), then the use of steroid treatments may be useful for
stimulating the production of the inducer protein(s) in sufficient
concentrations for detection by electrophoretic techniques. Once
isolated, the important proteins can be quantified during early
development.
We did not attempt to identify any of the sex-specific proteins;
however, several candidates bear consideration. We have previously
demonstrated that the ovaries and testes produce different amounts of
androstenedione, testosterone, and estradio1-17j3 during early
development (Chapter V). Therefore, different amounts and perhaps
types of steroidogenic enzymes may be present. Some of the sex-
specific proteins that we detected may be enzymes responsible for
steroidogenesis. The molecular weights and cellular locations of many
of the mammalian steroidogenic cytochrome P-450s (Waterman et al.
166
1986) are consistent with that of our 50,000 dalton protein. Williams
et al. (1986) found sexual differences in kidney cytochrome P-450
levels in adult trout. We observed a reduction of the 50,000 dalton
protein in ovaries of 6-month old genetic female rainbow trout treated
with dietary steroids. At 5 months of age, E2-treated (but not MTtreated) females had in vitro E2 production comparable to controls
(Chapter V) which suggests that the protein was not aromatase. The
reduction of this protein in E2-treated females may not be a
permanent condition because 17-month old E2-treated fish (from the
bisexual population) that developed into females had this protein (with
a staining density similar to that of control females). The proteins
resulting from early steroid treatment may not be revealed until later
in life. Waxman et al. (1985) demonstrated the induction at puberty of
several sex-specific hepatic cytochrome P-450s in male rats that had
undergone neonatal gonadectomy and hormonal replacement.
Steroid action requires the presence of specific receptors
(Yamamoto 1985) and therefore, steroid receptors may be possible
candidates for proteins produced early in development. The
developing gonad may contain different types or amounts of steroid
receptors depending on the intended sex, which might provide a
means for distinguishing by electrophoresis between undifferentiated
animals destined to be males and females. However, the capacity of
estrogens to induce genetic males to become females and androgens
to induce the opposite outcome in teleosts (Hunter and Donaldson
1983) seems to indicate that the developing testis contains estrogen
receptors and the developing ovary contains androgen receptors.
167
The list of other proteins important in early development is not
limited to steroidogenic enzymes and steroid receptors. Schindler
and Muller (1984) found that in newborn female rats, testosterone
treatment induced the production of a number of testis-specific
proteins. One of the proteins induced appeared to be myosin in
myoepithelial cells (Muller and Schindler 1983). Thus, contractile
and structural proteins may also be early peptides that can be detected
by electrophoresis. Finally, one of the initial differentiation events
noted in trout (Takashima et al. 1980; van den Hurk and Slof 1981;
Le Brun et at. 1982) and even mammals (Byskov 1986) is the entry into
meiosis of germ cells destined to be oogonia. One of the functions of
the proposed inducer substance(s) may be to initiate or delay meiosis.
Murray and Kirschner (1989) recently developed a model for the
control of the cell cycle which involves several proteins--including
cyclin, p34cdc2 (probably a protein kinase), maturation-promoting
factor (which may be a complex of the two former proteins), and a
cytostatic factor. The early entry into meiosis by animals destined to
be females may require alteration in levels of these proteins, thereby
allowing the detection of differential production of cell cycle proteins.
The complete explanation of gonadal sex differentiation will
require more than the sequencing of the inducer substance gene. The
use of electrophoretic techniques on gonadal tissues in conjunction
with steroid treatment of monosex populations of salmonids offers an
exciting model system for describing some of the molecular details
involved in gonadal morphogenesis.
168
Acknowledgements
We would like to thank Sam Bradford, Grant Feist, and Cam
Sharpe for assistance in the collection of tissues.
169
VII: CONCLUSION
The analyses and conclusions of the individual studies in my
dissertation accompany each chapter (see Chapters II through VI)
The goal of my thesis was to expand the understanding of the
regulation of two distinct events within reproduction in salmonids:
final sexual maturation and sexual differentiation. In this chapter, I
will attempt to summarize the salient points of the dissertation and
expand upon previous conclusions.
Although distinct stages of development, sexual maturation and
sex differentiation comprise facets of a common theme: reproduction.
Both center on the same tissue, i.e. the gonad. Both involve endocrine
changes; both involve parts of the meiotic cell cycle. Finally, both
constitute stages of development that can be exogenously controlled
and therefore, offer potential for development of technologies that are
important to fisheries science.
Final Maturation
We found that plasma testosterone levels are strongly correlated
to the ovulatory response of coho salmon to GnRHa (Chapter II), which
raises the question: does testosterone perform a vital function during
final maturation? In salmonids, DHP is clearly the maturationinducing substance (Nagahama and Yamashita 1989), therefore,
testosterone's role in final maturation may be supportive or
preparatory. We attempted to test the explanation that the correlation
between testosterone and response to GnRHa is causal by simply
170
raising plasma testosterone (or androgen) before treatment with
GnRHa (Chapter III). We expected that such treatment would result in
an expansion in the window of sensitivity to GnRHa. The lack of
apparent effect of testosterone, methyltestosterone (MT), or the
antiandrogen, cyproterone acetate (CA), on ovulatory response (as
determined by days elapsed to ovulation from GnRHa treatment)
seemingly argues against testosterone's direct involvement in ovulatory
processes. However, this conclusion may be premature as we failed in
our initial attempt with coho salmon to attain plasma levels of
testosterone (about 200 ng/mL) that were found (in Chapter II) to be
indicative of early response to GnRHa. The second attempt with coho
salmon began when females already had attained the desired level of
testosterone, thus the maturational process may have proceeded so far
as to mask any effect of MT or CA (see below).
Our studies did not prove that testosterone is without function
in the ovulatory process, just that androgens are generally without
effect under the experimental conditions. The relation between
testosterone and ovulation may be correlative rather than causal and
the suggestion that high testosterone concentrations circulating in
females is related to behavioral changes associated with nest building
(Li ley et a/. 1986) presents an interesting hypothesis worth
consideration. Two observations from our studies point to the need
for further experimentation. First, the finding that MT:saline-treated
chinook salmon ovulated at the same rate of either MT:GnRHa or
sham:GnRHa fish (and significantly faster than sham:saline females)
warrants further investigation into the effects of androgens alone.
Second, the trend towards delay of ovulation at one week following
171
GnRHa injection in CA:GnRHa-treated coho compared to sham:GnRHa-
or MT:GnRHa-treated females coupled with the dramatic drop in
plasma testosterone after CA treatment suggest not only that
testosterone may be involved in the ovulatory process, but also that the
use of antiandrogens may be an extremely useful means for addressing
androgen function in females.
The observation that CA and MT result in lower plasma
testosterone raises questions about mechanisms of action. If negative
feedback controls are in place to maintain homeostasis, one would
expect androgen treatment to eventually lead to reduced androgen
secretion and the effect of MT would support this view. Indeed, we
observed that dietary treatment with MT resulted in reduced in vitro
steroid production in gonads from developing rainbow trout (see
Chapter V). The lowering of testosterone by CA may suggest the
presence of positive feedback controls. If CA action is mediated
through the testosterone receptor (Rastogi and Chieffi 1975), then the
drop in testosterone after CA treatment implies that under normal
conditions, testosterone amplifies its own production. In other words,
one of the functions of the high levels of testosterone in female
salmonids may be to attain even higher levels of testosterone.
However, if CA acts to lower androgens through interference with
enzymes responsible for androgen production, then the idea of a
positive feedback function for androgens would not necessarily be
supported. Another possible mechanism for CA-induced reduction of
plasma testosterone is that CA increases the clearance of testosterone
by displacement from, or prevention of binding of testosterone to
172
plasma binding proteins, thereby freeing more testosterone for liver
metabolism (Schreck 1973).
The increase in testosterone often observed after GnRHa
treatment is puzzling in that the model for the physiological processes
leading to ovulation in salmonids focuses on the elevation of
gonadotropin (GtH II) and DHP. If DHP is the only requirement for
the reinitiation of meiosis in the oocytes, why is testosterone--which
is downstream from DHP in the synthetic pathway--also elevated? One
explanation may be that we are seeing an artefact of using a
superphysiological dose of GnRHa, which is a superactive analog of
GnRH. However, if this effect is not an artefact, then consideration
should be turned to the possible roles testosterone may play in
support of DHP. In order for DHP to act, it must have a functional
receptor on the oocyte membrane. In a recent study on the kisu
(Sallago japonica), Kobayashi et al. (1988) suggested that the
availability of receptors may determine when the oocyte becomes
sensitive to the maturation-inducing steroid. Synthesis of DHP
requires an increase in 2013-hydroxysteroid dehydrogenase and 17ahydroxylase activity (Kanamori et al. 1988) and therefore, the effect of
testosterone on the activity to these enzymes should be considered
(testosterone is known to affect the activity of other cytochrome P450s [Waxman et al. 1985]). Ovulation is mediated by gonadotropin
through prostaglandins (Goetz 1983) not through DHP--testosterone
may also influence this portion of the final maturation process.
173
Sex Differentiation
Feist et al. (1990) have demonstrated that the high plasma titers
of steroids in adults are reflected in high steroid content in oocytes.
Although the levels of steroids drop rapidly after fertilization, they are
always detectable. Thus, steroids are part of a biochemical continuum
between generations. We found no effect of testosterone treatment of
adults on the sex ratio of progeny (Chapter III). However, we did
observe that in vitro steroid production begins very early in both coho
salmon (Chapter IV) and rainbow trout (Chapter V).
The in vitro production of steroids in coho salmon and rainbow
trout were similar in that media concentrations of androstenedione
were higher than any other steroid measured in common and in that
both species demonstrated tissue sensitivity to partially purified
salmon gonadotropin. The high levels of androstenedione produced
both in the interrenal and from developing testes raises a question
concerning the function of androstenedione. Previous studies have
used androstenedione as a precursor to explore biosynthetic pathways
(van den Hurk et a/. 1982), but if androstenedione is only a precursor,
why was it not converted completely in our in vitro system? It would
be interesting to look for androstenedione receptors within the
gonads and elsewhere (e.g. the brain). The development of
pronounced steroidogenic potential in the interrenal earlier than that
of the gonads has been observed in other vertebrates (Kime et al.
1980). The interrenal production of androstenedione in high
concentrations relative to that produced from the gonads raises
questions about function--van den Hurk and van Oordt (1985)
174
suggested that androstenedione is converted into 1153hydroxyandrostenedione which may be the natural masculinizing agent
of the gonad. The increase in in vitro production of steroids observed
in coho salmon just before hatching is consistent with the results of
Feist et al. (1990) in which steroids extracted from whole bodies of
coho salmon also increased at this time. This increase was not
observed in incubations from rainbow trout tissues but may have been
missed due to the time of sampling. Whole body steroid levels in
rainbow trout also demonstrate an increase around the time of
hatching (Feist, unpublished observations).
Despite establishing the sex-specific production of steroids and
potential for pituitary control of steroidogenesis early in the
development of salmonids, we have not answered the question of
whether steroids are the natural inducers of gonadal sex
differentiation. Our results certainly point to early sex differences that
may act in the process of sexual development, but these observations
could be either a result or cause of gonadal differentiation. The use of
sex-inverting steroid treatments on monosex populations of salmonids
represents an exciting paradigm for studying how steroids modulate
the expression of the sexual phenotype. The description of sexspecific gonadal proteins (Chapter VI) constitutes a first step towards
understanding the mechanism by which exogenous steroids act.
Defining this mechanism through the identification of gene products
should yield information useful in delineating the biochemical steps
during natural sex differentiation.
The cell cycle of germ cells comprises one of the themes common to
sex differentiation and sexual maturation. The control of germ cell
175
cycling by external factors suggests the potential for involvement of
the endocrine system. In the yeast Saccharomyces cerevisiae, two
mating types (a and a) exist. Each type produces a pheremone that
acts specifically on the opposite mating type to cause cell cycle arrest
just before the so-called Start point which is the prerequisite for the
initiation of DNA synthesis and subsequent cell division (Pringle and
Hartwell 1981). The arrest before Start allows the haploid yeast cells
to conjugate to form a diploid cell that can then undergo meiotic
division (Pringle and Hartwell 1981). Murray and Kirshner (1989)
suggest that the mating pheremone arrests the cell cycle by
decreasing the half-life of a cyclin-like protein thereby preventing it
from accumulating to the critical level required for the cell to pass
Start. If cyclin decreases, then maturation-promoting factor (MPF)
activity will also decrease because cyclin and the protein kinase
p34cdc2 are the proposed subunits of the MPF (Murray and Kirshner
1989). Meiotic maturation in oocytes results from progestin-induced
increase in the activity of the very same MPF (Masui and Clarke 1979).
The similarities between the yeast cell cycle and final maturation
become even more striking when one considers that the mating
pheremone from the a-type yeast has an amino acid sequence with
considerable similarity to gonadotropin-releasing hormone (GnRH),
and in fact, demonstrates activity in pituitary cultures designed as a
bioassay for GnRH activity (Loumaye et a/. 1982). Thus, although the
mating pheremone and GnRH have opposite effects on MPF activity in
yeast and oocytes, the fact that they both ultimately act on MPF activity
may suggest a highly conserved biochemical relationship. One could
devise an evolutionary scenario in which the mating pheremone-MPF
176
connection became internalized and interspersed with regulatory
mechanisms during the development of more complex multicellular
organisms, eventually evolving into the h
othalamo-pituitary-gonadal
axis characteristic of most vertebrates. Through such a scenario,
perhaps the cell cycle, or a portion of it, came to be regulated by the
endocrine system. Whether the initiation of meiosis, which appears to
be one the first sexually distinct characteristics of development in
salmonids, involves the same proteins that control the resumption of
meiosis during final maturation and whether these proteins are
influenced by steroids remain questions for future pursuit.
177
VIII: BIBLIOGRAPHY
Anti la, E. (1984). Steroid conversion by oocytes and early embryos of
Salmo gairdneri. Ann. zool. Fenn. 21, 465-476.
Austin, C.R., Edwards, R.G., and Mittwoch, U. (1981). Introduction.
In: "Mechanisms of sex differentiation in Animals and Man" (C.R.
Austin and R.G. Edwards, eds.), pp. 1-54, Academic Press,
London.
Batty, K.A., Herbert, J., Deverne, E.B., and Vellucci, S.V. (1986).
Differences in blood levels of androgens in female talapoin
monkeys related to their social status. Neuroendocrinology 44,
347-354.
Bromage, N.R., Elliott, J.A.K., Springate, J.R.C., and Whitehead, C. (1984).
The effects of constant photoperiods on the timing of spawning in the
rainbow trout. Aquaculture 43, 213-223.
Bull, J.J., Hillis, D.M., and O'Steen, S. (1988). Mammalian ZFY
sequences exist in reptiles regardless of sex-determining
mechanism. Science 242, 567-569.
Burns, R.K. (1961). Role of hormones in the differentiation of sex. In:
"Sex and internal secretions," (W.C. Young, ed.), vol. I, pp. 76158. Williams and Wilkins Co., Baltimore.
Byskov, A.G. (1986). Differentiation of mammalian embryonic gonad.
Physiol. Rev. 66, 71-117.
Chieffi, G. (1965). Onset of steroidogenesis in the vertebrate
embryonic gonads. In: "Organogenesis" (R.L. De Haan and H.
Ursprung, eds.), pp. 653-671, Holt, Rinehart and Winston, New
York.
Clayton, R.N. and Catt, K.J. (1981). Regulation of pituitary
gonadotropin-releasing hormone receptors by gonadal
hormones. Endocrinology 108, 887-895.
Clemens, H.P. and Sneed, K.E. (1962). Bioassay and use of pituitary
materials to spawn warm-water fishes. Res. Rpt. 61, U.S.
Department of the Interior: Fish and Wildlife Service.
Crews, D. and Bull, J.J. (1987). Evolutionary insights from reptilian
sexual differentiation. in: Genetic Markers of Sex Differentiation
(F.P. Haseltine, M.E. McClure, and E.H. Goldberg, eds.), pp. 1126, Plenum Press, New York.
178
Crim, L.W. and Evans, D.M. (1979). Stimulation of gonadotropin by
testosterone in juvenile rainbow trout (Salmo gairdneri).
Gen. Comp. Endocrinol. 37: 192-196.
Crim, L.W. and Evans, D.M. (1980). LH-RH-stimulated gonadotropin
release from the rainbow trout pituitary gland: an in vitro assay
for detection of teleost gonadotropin releasing factor(s). Gen.
Comp. Endocrinol. 40, 283-290.
Crim, L.W. and Glebe, B.D. (1984). Advancement and synchrony of
ovulation in Atlantic salmon with pelleted LHRH analog.
Aquaculture 43, 47-56.
Crim, L.W., Peter, R.E., and Billard, R. (1981). Onset of gonadotropic
hormone accumulation in the immature trout pituitary gland in
response to estrogen or aromatizable androgen steroid
hormones. Gen. Comp. Endocrinol. 44, 374-381.
Crim, L.W., Evans, D.M., and Vickery B.H. (1983a). Manipulation of the
seasonal reproductive cycle of the landlocked Atlantic salmon
(Salmo salar) by LHRH analogues administered at various stages
of gonadal development. Can. J. Fish. Aquat. Sci. 40, 61-67.
Crim, L.W., Sutter lin, A.M., Evans, D.M., and Weil, C. (1983b).
Accelerated ovulation by pelleted LH-RH analogue treatment of
spring-spawning rainbow trout (Salmo gairdneri) held at low
temperature. Aquaculture 35, 299-307.
Donaldson, E.M., Hunter, G.A., and Dye, H.M. (1981). Induced
ovulation in coho salmon (Oncorhynchus kisutch). II.
Preliminary study of the use of LH-RH and two high potency
analogues. Aquaculture 26, 129-141.
FAO (Food and Agriculture Organization of the United Nations). (1975).
Workshop on controlled reproduction of cultivated fishes. EIFAC
Tech. Paper 25.
Feist, G. (1988). Sex steroid profiles of coho salmon, Oncorhynchus
kisutch, during early development and sexual differentiation.
Master of Science Thesis, Oregon State University, 74 pp..
Feist, G. and Schreck, C.B. (1990). Hormonal content of commercial
fish diets and of young coho salmon (Oncorhynchus kisutch) fed
these diets. Aquaculture, in press.
Feist, G., Schreck, C.B., Fitzpatrick, M.S., and Redding, J.M. (1990).
Sex steroid profiles of coho salmon (Oncorhynchus kisutch)
during early development and sexual differentiation. Gen. Comp.
Endocrinol., in press.
179
Ferreiro, C., Medrano, J.F., and Gall, G.A.E. (1989). Genome analysis of
rainbow trout and sturgeon with restriction enzymes and
hybridization with a ZFY gene derived probe to identify sex.
Aquaculture 81, 245-251.
Fitzpatrick, M.S., Suzumoto, B.K., Schreck, C.B., and Oberbillig, D.
(1984). Luteininzing hormone-releasing hormone analogue
induces precocious ovulation in adult coho salmon
(Oncorhynchus kisutch). Aquaculture 43, 67-73.
Fitzpatrick, M.S., Van Der Kraak, G., and Schreck, C.B. (1986). Plasma
profiles of sex steroids and gonadotropin in coho salmon
(Oncorhynchus kisutch) during final maturation. Gen. Comp.
Endocrinol. 62, 437-451.
Fitzpatrick, M.S., Redding, J.M., Ratti, F.D., and Schreck, C.B. (1987).
Plasma testosterone concentration predicts the ovulatory
response of coho salmon (Oncorhynchus kisutch) to
gonadotropin-releasing hormone analog. Can. J. Fish. Aquat. Sci.
44, 1351-1357.
Frager, M.S., Pieper, D.R., Tonetta, S.A., Duncan, J.A., and Marshall,
J.C. (1981). Pituitary gonadotropin-releasing hormone
receptors. Effects of castration, steroid replacement, and the
role of gonadotropin-releasing hormone in modulating receptors
in rat. J. Clin. Invest. 67, 615-623.
Gielen, J.T., Goos, H.J.T., Peute, J., van den Bosch, R.A., and van Oordt,
P.G.W.J. (1982). The brain-pituitary-gonadal axis in the rainbow
trout, Salmo gairdneri: gonadal hormones and the maturation of
gonadotropic cells. Cell Tissue Res. 225: 45-56.
Giguere, V., Lefebvre, F.A., and Catt, K.J. (1981). Androgens decrease
LHRH binding sites in rat pituitary cells in culture.
Endocrinology 108, 350-352.
Goetz, F.W. (1983). Hormonal control of oocyte final maturation and
ovulation in fishes. In: " Fish Physiology" vol. 9B (W.S. Hoar, D.J.
Randall, and E.M. Donaldson, eds.), pp. 117-170, Academic
Press, New York.
Goetz, F.W., Donaldson, E.M., Hunter, G.A., and Dye, H.M. (1979).
Effects of estradio1-17J3 and 17a-methyltestosterone on gonadal
differentiation in the coho salmon, Oncorhynchus kisutch.
Aquaculture 17, 267-278.
Hames, B.D. (1981). An introduction to polyacrylamide gel
electrophoresis. In: "Gel electrophoresis of proteins: a practical
180
approach" (B.D. Hames and D. Rickwood, eds.), pp. 1-91, IRL
Press Ltd., Oxford, England.
Haseltine, F.P. and Ohno, S. (1981). Mechanism of gonadal
differentiation. Science 211, 1272-1278.
Hopkins, K.D., Shelton, W.L., and Engle, C.R. (1979). Estrogen sexreversal of Ti lapia aurea. Aquaculture 18, 263-268.
Hunter, G.A. and Donaldson, E.M. (1983). Hormonal sex control and
its application to fish culture. In: "Fish Physiology' (W.S. Hoar,
D.J. Randall, and E.M. Donaldson, eds.), vol. 9B, pp. 223-303,
Academic Press, New York.
Jalabert, B. (1976). In vitro oocyte maturation and ovulation in
rainbow trout ( Salmo gairdneri), northern pike (Esox Lucius),
and goldfish (Carassius auratus). J. Fish. Res. Bd. Canada 33,
974-988.
Jalabert, B. and A. Fostier. (1984). The follicular sensitivity in vitro to
maturation-inducing hormones in rainbow trout, Salmo
gairdneri: role of oestradiol -17J3. Aquaculture 43: 1-11.
Jost, A. (1965). Gonadal hormones in the sex differentiation of the
mammalian fetus. In: "Organogenesis" (R.L. De Haan and H.
Ursprung, eds.), pp. 611-628, Holt, Rinehart and Winston, New
York.
Kagawa, H., G. Young, S. Adachi, and Y. Nagahama. (1982). Estradiol1753 production in amago salmon (Oncorhynchus rhodurus)
ovarian follicles: role of thecal and granulosa cells. Gen. Comp.
Endocrinol. 47: 440-448.
Kanamori, A., Adachi, S., and Nagahama, Y. (1988). Developmental
changes in steroidogenic responses of ovarian follicles of amago
salmon (Oncorhynchus rhodurus) to chum salmon gonadotropin
during oogensis. Gen. Comp. Endocrinol. 72, 13-24.
Kawauchi, H., Suzuki, K., Itoh, H., Swanson, P., Naito, N., Nagahama, Y.,
Nozaki, M., Nakai, Y., and Itoh, S. (1989). The duality of teleost
gonadotropins. Fish Physiol. Biochem. 7, 29-38.
Kelley, C.D., Lee, C.-S., and Tamaru, C.S. (1987). Acceleration of
maturation and double spawning in one season in the grey mullet
(Mugil cephalus) following chronic hormone treatment. In:
Proceedings of the Third International Symposium on
Reproductive Physiology of Fish (D.R. Idler, L.W. Crim, and J.M.
Walsh, eds.), p. 203, Memorial University of Newfoundland, St.
John's.
181
Kime, D.E., Vinson, G.P., Major, P.W., and Kilpatrick, R. (1980).
Adrenal-gonad relationships. In: "General, Comparative and
Clinical Endocrinology of the Adrenal Cortex" (I. Chester Jones
and I.W. Henderson, eds.), vol. 3, pp. 183-264, Academic Press,
London.
Kobayashi, M., Aida, K., Furakawa, K., Law, Y.K., Moriwaki, T., and
Hanyu, I. (1988). Development of sensitivity to maturationinducing steroids in the oocytes of the daily spawning teleost,
the kisu Sallago japonica. Gen. Comp. Endocrinol. 72, 264-271.
Kobayashi, M., Aida, K., and Hanyu, I. (1989a). Involvement of steroid
hormones in the preovulatory gonadotropin surge in female
goldfish. Fish Physiol. Biochem. 7, 141-146.
Kobayashi, M., Aida, K., and Hanyu, I. (1989b). Induction of
gonadotropin surge by steroid hormone implantation in
ovariectomized and sexually regressed female goldfish. Gen.
Comp. Endocrinol. 73, 469-476.
Kramer, C.R. and Kallman, K.D. (1985). Sex differentiation of somatic
tissue in the unsexualixed gonad primordia of the embryos of
three species of poeciliid fish. J. Anat. 140, 269-277.
Laemmli, U.K. (1970). Cleavage of structural proteins during assembly
of the head of bacteriophage T4. Nature 227, 680-685.
Lam, T.J. (1982). Applications of endocrinology to fish culture. Can. J.
Fish. Aquat. Sci. 39, 111-137.
Le Brun, C., Billard, R., and Jalabert, B. (1982). Changes in the number
of germ cells in the gonads of the rainbow trout (Salmo
gairdneri) during the first 10 post-hatching weeks. Reprod.
Nutr. Develop. 22, 405-412.
Lee, C.-S., Tamaru, C.S., Banno, J.E., and Kelley, C.D. (1986a).
Influence of chronic administration of LHRH-analogue and/or
17a-methyltestosterone on maturation in milkfish, Chanos
chanos. Aquaculture 59, 147-159.
Lee, C.-S., Tamaru, C.S., Banno, J.E., Kelley, C.D., Bocek, A., and Wyban,
J.A. (1986b). Induced maturation and spawning of milkfish,
Chanos chanos Forskal, by hormone implantation. Aquaculture
52, 199-205.
Li ley, N.R., Fostier, A., Breton, B., and Tan, E.S.P. (1986). Endocrine
changes associated with spawning behavior and social stimuli in
182
a wild population of rainbow trout ( Salmo gairdneri). II. Females.
Gen. Comp. Endocrinol. 62, 157-167.
Loumaye, E., Thorner, J., and Catt, K.J. (1982). Yeast mating
pheremone activates mammalian gonadotrophs: evolutionary
conservation of a reproductive hormone?. Science 218, 13231325.
Lowry, 0.H., Rosebrough, N.J., Farr, A.L., and Randall, R.V. (1951).
Protein measurement with folin phenol reagent. J. Biol. Chem.
193, 265-275.
Masui, Y. and Clarke, H.J. (1979). Oocyte maturation. Int. Rev. Cytol.
57, 185-282.
McCarrey, J.R. and Abbott, U.K. (1979). Mechansisms of genetic sex
determination, gonadal sex differentiation, and germ-cell
development in animals. Adv. Genetics 29, 217-290.
McLaren, A., Simpson, E., Tomonari, K., Chandler, P., and Hogg, H.
(1984). Male sexual differentiation in mice lacking H-Y antigen.
Nature (London) 312, 552-555.
Muller, U. and Schindler, H. (1983). Testicular differentiation--a
developmental cascade. Differentiation 23, S99-S103.
Muller, U. and Wolf, U. (1979). Cross-reactivity to mammalian anti-HY antiserum in teleostean fish. Differentiation 14, 185-187.
Murray, A.W. and Kirschner, M.W. (1989). Dominoes and clocks: the
union of the two views of the cell cycle. Science 246, 614-621.
Nagahama, Y. and Yamashita, M. (1989). Mechanisms of synthesis and
action of 17a,205-dihydroxy-4-pregnen-3-one, a teleost
maturation-inducing substance. Fish Physiol. Biochem. 7, 193200.
Nagahama, Y., Kagawa, H., and Tashiro, F. (1980). The in vitro effects
of various gonadotropins and steroid hormones on oocyte
maturation in amago salmon Oncorhynchus rhodurus and
rainbow trout Salmo gairdneri. Bull. Jpn. Soc. Sci. Fish. 46:
1097-1102.
Nagahama, Y., Hirose, K., Young, G., Adachi, S., Suzuki, K., and
Tamaoki, B.-I. (1983). Relative in vitro effectiveness of 17a,205dihydroxy-4-pregnen-3-one and other pregnene derivatives on
germinal vesicle breakdown in oocytes of ayu (Ptecoglossus
altivelis), amago salmon (Oncorhynchus rhodurus), rainbow trout
183
(Salmo gairdneri), and goldfish (Carassius auratus). Gen. Comp.
Endocrinol. 51, 15-23.
Neumann, F., Graf, K.-J., Hasan, S.H., Schenck, B., and Steinbeck, H.
(1977). Central actions of antiandrogens. In: "Androgens and
antiandrogens" (L. Martini and M.Motta, eds.), pp. 163-177,
Raven Press, New York.
Nikolarakis, K.E., Pfeiffer, D.G., Almeida, O.F.X., and Herz, A. (1986).
Opioid modulation of LHRH release in vitro depends on levels of
testosterone in vivo. Neuroendocrinology 44, 314-319.
Page, D.C., Mosher, R., Simpson, E.M., Fisher, E.M.C., Mardon, G.,
Pollack, J., McGillivray, B., de la Chapelle, A., and Brown, L.G.
(1987). The sex-determining region of the human Y
chromosome encodes a finger protein. Cell 51, 1091-1104.
Patifio, R., Bradford, C.S., and Schreck, C.B. (1986). Adenylate cyclase
activators and inhibitors, cyclic nucleotide analogs, and
phosphatidylinositol: effects on interrenal function of coho
salmon (Oncorhynchus kisutch) in vitro. Gen. Comp Endocrinol.
63, 230-235.
Patifio, R., Redding, J.M., and Schreck, C.B. (1987). Interrenal
secretion of corticosteroids and plasma cortisol and cortisone
concentrations after acute stress and during seawater
acclimation in juvenile coho salmon (Oncorhynchus kisutch).
Gen. Comp. Endocrinol. 68, 431-439.
Perdew, G.H., Schaup, H.W., Becker. M.M., Williams, J.L., and
Selivonchick, D.P. (1986). Alterations in the synthesis of
proteins in hepatocytes of rainbow trout fed cyclopropenoid fatty
acids. Biochim. Biophys. Acta 877, 9-19.
Piferrer, F. and Donaldson, E.M. (1989). Gonadal differentiation in
coho salmon, Oncorhynchus kisutch, after a single treatment
with androgen or estrogen at different stages of ontogenesis.
Aquaculture 77, 251-262.
Pringle, J.R. and Hartwell, L.H. (1981). The Saccharomyces cerevisiae
cell cycle. In: "The molecular biology of the yeast
Saccharomyces cerevisiae: life cycle and inheritance" (J.N.
Strathern, E.W. Jones, and J.R. Broach, eds.), pp. 97-142, Cold
Spring Harbor Laboratory, New York.
Quirk, J.G. and Hamilton, J.B. (1973). Number of germ cells in known
male and known female genotypes of vertebrate embryos
(Oryzias latipes). Science 180, 963-964.
184
Rastogi, R.K. and Chieffi, G. (1975). The effects of antiandrogens and
antiestrogens in nonmammalian vertebrates. Gen. Comp.
Endocrinol 26, 79-91.
Redding, J.M., Schreck, C.B., Birks, E.K., and Ewing, R.D. (1984).
Cortisol and its effects on plasma thyroid hormone and
electrolyte concentrations in fresh water and during seawater
acclimation in yearling coho salmon, Oncorhynchus kisutch.
Gen Comp. Endocrinol. 61, 127-133.
Rosenblum, P.M. and Peter, R.E. (1989). Evidence for the involvement
of endogenous opioids in the regulation of gonadotropin
secretion in male goldfish, Carassius auratus. Gen. Comp.
Endocrinol. 73, 21-27.
Satoh, N. and Egami, N. (1972). Sex differentiation of germ cells in
the teleost, Oryzias latipes, during normal embryonic
development. J. Embryol. Exp. Morph. 28, 385-395.
Schally, A.V., Arimura, A., and Kastin, A.J. (1973). Hypothalamic
regulatory hormones. Science 179, 341-350.
Scheerer, P.D., Thorgaard, G.H., Allendorf, F.W., and Knudsen, K.L.
(1986). Androgenetic rainbow trout produced from inbred and
outbred sperm sources show similar survival. Aquaculture 57,
289-298.
Schindler, H. and Muller, U. (1984). Testis cell-specific proteins in
testosterone-treated rat ovaries. Differentiation 27, 113-119.
Schreck, C.B. (1973). Uptake of 3H-testosterone and influence of an
antiandrogen in tissues of rainbow trout (Salmo gairdneri). Gen.
Comp. Endocrinol. 21, 60-68.
Schreck, C.B., Bradford, C.S., Fitzpatrick, M.S., and Patirio, R. (1989).
Regulation of the interrenal of fishes: non-classical control
mechanisms. Fish Physiol. Biochem. 7, 259-265.
Scott, A.P., Sheldrick, E.L., and Flint, A.P. (1982). Measurement of
17a.2013-dihydroxy-4-pregnen-3-one in plasma of trout (Salmo
gairdneri Richardson): seasonal changes and response to salmon
pituitary extract. Gen. Comp. Endocrinol. 46: 444-451.
Scott, A.P., Sumpter, J.P., and Hardiman, P.A. (1983). Hormone
changes during ovulation in the rainbow trout ( Salmo gairdneri
Richardson). Gen. Comp. Endocrinol. 49, 128-134.
185
Sherwood, N., F. Eiden, M. Brownstein, J. Spiess, J. Rivier, and W.Vale.
1983. Characterization of a teleost Gn-RH. Proc. Natl. Acad. Sci.
USA 80: 2794-2798.
Simpson, E., Chandler, P., Goulmy, E., Disteche, C.M., Ferguson-Smith,
M.A., and Page, D.C. (1987). Separation of the genetic loci for
the H-Y antigen and for testis determination on human Y
chromosome. Nature 326, 876-878.
Sower, S.A., Schreck, C.B., and Donaldson, E.M. (1982). Hormoneinduced ovulation of coho salmon (Oncorhynchus kisutch) held
in seawater and fresh water. Can. J. Fish. Aquat. Sci. 39, 627632.
Sower, S.A., Iwamoto, R.N., Dickhoff, W.W., and Gorbman, A. (1984).
Ovulatory and steroidal responses in coho salmon and steelhead
trout following administration of salmon gonadotropin and DAla6, des-Gly10 gonadotropin-releasing hormone ethylamide
(GnRHa). Aquaculture 43, 35-46.
Suzuki, K., Kanamori, A., Nagahama, Y., and Kawauchi, H. (1988a).
Development of salmon GTH I and GTH II radioimmunoassays.
Gen. Comp. Endocrinol. 71, 459-467.
Suzuki, K., Kawauchi, H., and Nagahama, Y. (1988b). Isolation and
characterization of two distinct gonadotropins from chum
salmon pituitary glands. Gen. Comp. Endocrinol. 71, 292-301.
Suzuki, K., Nagahama, Y., and Kawauchi, H. (1988c). Steroidogenic
activities of two distinct salmon gonadotropins. Gen. Comp.
Endocrinol. 71, 452-458.
Swanson, P. and Dickhoff, W.W. (1988). Effects of exogenous steroids
on gonadotropins I and II in juvenile coho salmon. Amer. Zool.
28, 55A.
Swanson, P., Bernard, M., Nozaki, M., Suzuki, K., Kawauchi, H., and
Dickhoff, W.W. (1989). Gonadotropins I and II in juvenile coho
salmon. Fish Physiol. Biochem. 7, 169-176.
Takashima, F., Patirio, R., and Nomura, M. (1980). Histological studies
on the sex differentiation in rainbow trout. Bull. Jpn. Soc. Sci.
Fish. 46, 1317-1322.
Tamaru, C.S. and Lee, C.-S. (1987). Testosterone and estradio1-17J3
profiles of female milkfish, Chanos chanos undergoing chronic
LHRH-A and 17a-methyltestosterone. In: Proceedings of the
Third International Symposium on Reproductive Physiology of
186
Fish (D.R. Idler, L.W. Crim, and J.M. Walsh, eds.), p. 218,
Memorial University of Newfoundland, St. John's.
Thorgaard, G.H. (1977). Heteromorphic sex chromosomes in male
rainbow trout. Science 196, 900-902.
Thorgaard, G.H. and Gall, G.A.E. (1979). Adult triploids in a rainbow
trout family. Genetics 93, 961-973.
Thorgaard, G.H. and Allen, S.K. (1987). Chromosome manipulation
and markers in fishery management. In: "Population Genetics
and Fishery Management" (N. Ryman and F. Utter, eds.), pp.
319-331, Washington Sea Grant Program, University of
Washington Press, Seattle.
Van Der Kraak, G., Lin, H.R., Donaldson, E.M., Dye, H.M., and Hunter,
G.A. (1983). Effects of LH-RH and des-G1y10D-Ala61LH-RHethylamide on plasma gonadotropin levels and oocyte maturation
in adult female coho salmon (Oncorhynchus kisutch). Gen.
Comp. Endocrinol. 49, 470-476.
Van Der Kraak, G., Dye, H.M., and Donaldson, E.M. (1984). Effects of
LH-RH and des-Glyio[D_ A/A 1la-6,]LH-RH-ethylamide on plasma sex
steroid profiles in adult female coho salmon (Oncorhynchus
kisutch). Gen. Comp. Endocrinol. 55, 36-45.
Van Der Kraak, G., Dye, H.M., Donaldson, E.M., and Hunter, G.A.
(1985). Plasma gonadotropin, 1713-estradiol , and 17a,2013-
dihydroxy-4-pregnen-3-one levels during luteinizing hormonereleasing hormone analogue and gonadotropin induced ovulation
in coho salmon (Oncorhynchus kisutch). Can. J. Zool. 63, 824833.
Van Der Kraak, G., Donaldson, E.M., and Chang, J.P. (1986). Dopamine
involvement in the regulation of gonadotropin secretion in coho
salmon. Can. J. Zool. 64, 1245-1248.
van den Hurk, R. (1982). Effects of steroids on gonadotropic (GTH)
cells in the pituitary of rainbow trout, Salmo gairdneri, shortly
after hatching. Cell Tissue Res. 224, 361-368.
van den Hurk, R. and Slof, G.A. (1981). A morphological and
experimental study of gonadal sex differentiation in the rainbow
trout Salmo gairdneri. Cell Tissue Res. 218, 487-497.
van den Hurk, R. and Leeman, W.R. (1984). Increase of steriodproducing cells in interrenal tissue and masculinization of
gonads after long-term treatment of juvenile rainbow trout with
cyanoketone. Cell Tissue Res. 237, 285-289.
187
van den Hurk, R. and van Oordt, P.G.W.J. (1985). Effects of natural
androgens and corticosteroids on gonad differentiation in the
rainbow trout, Salmo gairdneri. Gen. Comp. Endocrinol. 57,
216-222.
van den Hurk, R., Lambert, J.G.D., and Peute, J. (1982).
Steroidogenesis in the gonads of rainbow trout fry (Salmo
gairdneri) before and after the onset of gonadal sex
differentiation. Reprod. Nutr. Develop. 22, 413-425.
van den Hurk, R., Gielen, J.Th., and Terlou, M. (1984). Accumulation
of glycoprotein gonadotropin in the pituitary of juvenile rainbow
trout in response to androgens and C21-steroids, including 11steroids. Cell Tissue Res. 235: 635-642.
Wachtel, S.S. and Koo, G.C. (1981). H-Y antigen in gonadal
differentiation. In: "Mechanisms of sex differentiation in animals
and man" (C.R. Austin and R.G. Edwards, eds.), pp. 255-299.
Academic Press, New York.
Wallace, R.A. and Selman, K. (1981). Cellular and dynamic aspects of
oocyte growth in teleosts. Amer. Zool. 21, 325-343.
Waterman, M.R., John, M.E., and Simpson, E.R. (1986). Regulation of
synthesis and activity of cytochrome P-450 enzymes in
physiological pathways. In: Cytochrome P-450, structure,
mechanisms, and biochemistry (P.R. Ortiz de Montellano, ed.),
pp. 345-386, Plenum Press, New York, N.Y.
Waxman, D.J., Dannan, G.A., and Guengerich, F.P. (1985). Regulation
of rat hepatic cytochrome P-450: age-dependent expression,
hormonal imprinting, and xenobiotic inducibility of sex-specific
isoenzymes. Biochemistry 24, 4409-4417.
Williams, D.E., Masters, B.S.S., Lech, J.J., and Buhler, D.R. (1986). Sex
differences in cytochrome P-450 isozume composition and
activity in kidney microsomes of mature rainbow trout.
Biochem. Pharmacol. 35, 2017-2023.
Wilson, J.D., Griffin, J.E., and George F.W. (1980). Sexual
differentiation: early hormone synthesis and action. Biol.
Reprod. 22, 9-17.
Witschi, E. (1934). Genes and inductors of sex differentiation in
amphibians. biol Rev. Biol. Proc. Cambridge Phil. Soc. 9, 460488.
188
Witschi, E. (1967). Biochemistry of sex differentiation in vertebrate
embryos. In: "The biochemistry of animal development" (R.
Weber, ed.), vol. 2, pp. 193-223, Academic Press, New York.
Wright, R.S. and S.M.V. Hunt. (1982). A radioimmunoassay for
17a,2013-dihydroxy-4-pregnen-3-one: its use in measuring
changes in serum levels at ovulation in Atlantic salmon(Sa/mo
salar), coho salmon (Oncorhynchus kisutch), and rainbow trout
(Salmo gairdneri). Gen. Comp. Endocrinol. 47: 475-482.
Yamamoto, K.R. (1985). Steroid receptor regulated transcription of
specific genes and gene networks. Ann. Rev. Genet. 19, 209252.
Yamamoto, T.O. (1969). Sex differentiation. In: "Fish Physiology" (W.S.
Hoar and D.J. Randall, eds.), vol. 3, pp. 117-175, Academic
Press, New York.
Yamazaki, F. (1983). Sex control and manipulation in fish.
Aquaculture 33, 329-354.
Young, G., Kagawa, H., and Nagahama, Y. (1982). Oocyte maturation in
the amago salmon (Oncorhynchus rhodurus): in vitro effects of
salmon gonadotropin, steroids, and cyanoketone (an inhibitor of
313-hydroxy-A5-steroid dehydrogenase). J. Exp. Zool. 224, 265275.
Young, G., Crim, L.W., Kagawa, H., Kambegawa, A., and Nagahama, Y.
(1983). Plasma 17a,20J3-dihydroxy-4-pregnen-3-one levels
during sexual maturation of amago salmon (Oncorhynchus
rhodurus): correlation with plasma gonadotropin and in vitro
production by ovarian follicles. Gen. Comp. Endocrinol. 51: 96105.
Zohar, Y. (1989). Endocrinology and fish farming: aspects in
reproduction, growth, and smoltification. Fish Physiol. Biochem.
7, 395-405.
APPENDIX
189
APPENDIX:
HPLC DETECTION OF METHYLTESTOSTERONE
AND CYPROTERONE ACETATE IN PLASMA
OF COHO SALMON
Introduction
The cocoa butter implants used in 1986 on coho salmon
(Chapter III) were problematic for two reasons. First, the cocoa butter
had to be kept heated to facilitate delivery by injection, but the
raceway and the nearest place for operating a hot plate were some
distance apart. The result was that only a few fish could be injected at
any one time. Second, the plasma concentrations of testosterone one
week after cocoa butter implantation were generally far below those
indicative of sensitivity to GnRHa (Chapter II). Therefore, we decided
to use microencapsulated steroids for all further experiments.
Microcapsules have the advantage of remaining fluid at lower
temperatures. We conducted the following pilot study to determine
the release characteristics of two different preparations of
microcapsules containing 17a-methyltestosterone (MT) over time.
Materials and Methods
Juvenile coho salmon (Oncorhynchus kisutch) were maintained
at Smith Farm Experimental Hatchery in 110C water. Dr. Chris
Langdon provided us with the methods for microencapsulation. The
microcapsules were made by dissolving 200 mg of MT in 3.2 g of
Menhaden Oil (gift from Zapata Haynie Corp., Reedville, VA), 1.3 g of
kaomel (gift from Durkee Foods, Inc., Lafayette, CA.), and 1.9 g of
either lecithin (in 1 mL EtOH) or tripalmitin. These ingredients were
190
mixed well at 40-60°C, and then 10.7 mL of 2% (w/v) polyvinyl alcohol
(at the same temperature) was added. The entire mixture was
immediately sonicated until frothy and well mixed, at which time 16.0
mL of ice cold H2O was added. A sample of microcapsules was
examined under the microscope (100x) to ensure that the steroid was
dissolved and that the microcapsules were uniform in shape.
Animals were anesthetized with tricaine methansulfonate (MS222), weighed, and then injected with MT microcapsules at a dose of
0.02 mg/g body weight. The fish were returned to tanks to recover
with each treatment group occupying a separate tank. On one, four,
and 10 days after treatment, five fish from each group were killed by
overdose with MS-222 (untreated controls were killed on day 1).
Blood was collected from the severed caudal vein of each individual
and the plasma from each blood sample was stored at -200C until
analysis by High Performance Liquid Chromatography (HPLC).
Plasma samples were pooled by treatment and day. Each pool
was analyzed for the presence of MT using the method described in
Feist (1988 Master's Thesis, Oregon State University). In brief, we
extracted 2 mL of plasma using a Baker-10 SPETM System. Plasma
extracts were then filtered through 0.45 µm disposable filters and
reconstituted in mobile phase (see below) for injection onto HPLC.
Each reconstituted extract was injected onto a reverse phase C 18
HPLC column and MT was eluted starting with an isocratic mobile
phase (0.4 mL/min) of water:methanol:acetonitrile:isopropanol
(62:28:5:5) followed immediately by a linear gradient of 3.3%/min of
water:methanol:butanol (35:45:20) for 30 minutes. Standards of
known amounts of MT (50, 100, and 250 ng) were run through the
191
same system interspersed with samples. Concentrations of MT in the
samples were calculated by comparing the sample peak heights with
peak heights of the standards.
To determine if we could detect cyproterone acetate (CA) in the
plasma one week after microcapsule injection, Sam Bradford extracted
plasma samples from control and CA-treated adult females (see
Chapter III for details of treatment) in the manner described above
and analyzed the reconstituted extracts by HPLC as described. No
quantification was attempted.
Results
Measurable amounts of MT were detected in plasma from the
tripalmitin (TP)-based microcapsule-treated fish on one, four, and 10
days after treatment (Fig. 42), whereas the plasma from the lecithin
(L)-based microcapsule-treated fish had detectable MT only on one day
after treatment. The concentrations of MT ranged from 90 ng/mL
(day 1) to 9 ng/mL (day 10) in fish treated with TP-microcapsules and
were 48 ng/mL (day 1) in fish treated with L-microcapsules (Fig. 43).
Based on these results, we decided to use TP-microcapsules on adult
salmon (see Chapter III).
We detected the presence of CA in plasma from adult female
coho salmon one week after treatment (Fig. 44).
192
Figure 42. HPLC chromatogram of methyltestosterone (MT) in plasma
from juvenile coho salmon at 1, 4, and 10 days after treatment with
MT microcapsules. Absorbance units at full scale, 254 nm; range,
0.001; chart speed, 0.5 cm/min.
193
40*
MT
Day 1
Day 4
E
MT
1
Day 10
MT
5
I.
r
10
I.
I
Retention time (min)
Figure 42.
,
15
20
194
Figure 43. Plasma levels of methyltestosterone (MT) in juvenile coho
salmon treated (on Day 0) with MT microcapsules made with either
tripalmitin or lecithin. Values were calculated by comparison with
authentic standards.
195
200
150
0
o
N
2
c
--e-U
100
- - 0- - -
50
2
4
6
Days after injection
Figure 43.
8
10
Control
Tripalmitin
Lecithin
196
Figure 44. HPLC chromatogram of cyproterone acetate (CA) in plasma
from adult female coho salmon at 1 week after treatment with CA
microcapsules.
CA
I
15
16
I
17
I
18
1
19
I
20
1
21
1
22
I
23
I
24
Retention time (min)
Figure 44.
25
26
27
28
29
30