{(33,} 4'; .1’
' ~ - ”.1."

 

1 .
”.1 {If};
.Ig‘v >.-

 

, .23 :
MPA‘kf..V"\!r
' ‘ "Rig“:

I
L

17:???‘(3
' 4
I

1i}?
2:

1
ch
1:

9""
".- n

:, q;
33¢.
*6

.f
‘l
1: ~i
«2-153; "‘3‘
1'7
’ x

. "’7‘

".

 

"s.
M

.\
» J

,2" .V
‘ , ..,
“pm”

I .. «L

F
'1‘]

-A
l r‘
‘ 7,1;

‘19:”.

D J .
~ - b r. :‘1 .1: ,
“If/'7. 1."
L

. a

1...". . v ' ‘
m. “-1.5: v ’ ' J Y?
. , ”15"5/h .
, . _

 

llllllllll\llllllllllIll|||ll||lllll|ll|lllllllllllll

99 5596
THE-Sis 3 1293 013

This is to certify that the

dissertation entitled

REGULATION OF ANDROGEN RECEPTOR-IWUNOREACTIVE CELLS
DURING PUBERTAL MATURATION WITHIN THE CENTRAL NERVOUS
SYSTEM OF THE MALE EUROPEAN FERRET: EFFECTS
OF GONADAL STEROIDS

presented by

Michael L. Kashon

has been accepted towards fulﬁllment
of the requirements for

Ph.D. degmin Neumsc1ence 8. Psychology

 

 

W6LM

Masor professor

Date 7‘2’0 .9 S—

 

MS U is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

*4 ﬁﬁ. ———_ﬁ ‘ L _

 

LIBRARY

Michigan State
University

 

PLACE Ii RETURN BOX to remove thi- checkout from your record.
To AVOID FINES return on or before due due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU ieAnAi'iirmetive ActioNEquei Opportunity intuition
W1

 

 

 

 

_ —___—‘

REGULATION OF ANDROGEN RECEPTOR-IMMUNOREACTIVE CELLS
DURING PUBERTAL MATURATION WITHIN THE CENTRAL NERVOUS
SYSTEM OF THE MALE EUROPEAN FERRET: EFFECTS OF GONADAL
STEROIDS
By

Michael L. Kashon

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Neuroscience Program and Department of Psychology

1995

ABSTRACT

REGULATION OF ANDROGEN RECEPTOR-IMMUNOREACTIVE CELLS
DURING PUBERTAL MATURATION WITHIN THE CENTRAL NERVOUS
SYSTEM OF THE MALE EUROPEAN FERRET: EFFECTS OF GONADAL
STEROIDS
By

Michael L. Kashon

Puberty in males is characterized by onset of gonadal maturation brought about
by an increase in the: a) frequency of episodic release of luteinizing hormone-releasing
hormone from the hypothalamus, b) frequency of release of luteinizing hormone from
the anterior pituitary, and c) testosterone from the testes. Two well known actions of
testosterone on the central nervous system are suppression of gonadotropin secretion
from the anterior pituitary, and activation of male reproductive behavior. During
pubertal maturation, interaction of testosterone with target cells in the central nervous
system is dynamic. First, there is a decline in the ability of testosterone to suppress
gonadotropin secretion as the animal undergoes sexual maturation. Second, there is an
increase in the ability of testosterone to activate male reproductive behavior. The
biological actions of testosterone are mediated by the androgen receptor, or through the
estrogen receptor following conversion of testosterone to estrogen. Experiments in this
dissertation tested the hypothesis that the pubertal increase in behavioral responsiveness
to testosterone is due to an increase in number of androgen receptor containing cells in
regions of the brain which are involved in control of reproductive behavior.

Immunocytochemical staining for androgen receptor was greatest in number of cells in

hypothalamic and limbic structures in ferrets undergoing pubertal maturation relative to
prepubertal ferrets. This increase in androgen receptor-immunoreactive cells was linked
to the increased concentrations of testosterone which accompany pubertal maturation.
Since androgens regulate the level of their own receptors, it was proposed that
prepubertal ferrets may be unable to respond to testosterone treatment with a sufﬁcient
upregulation of androgen receptors, and that this may be involved in mediating
responsiveness to the behavioral effects of testosterone. However, both prepubertal and
adult male ferrets displayed identical patterns of androgen receptor-immunoreactive cells
following 10 days of treatment with testosterone, indicating that an upregulation of brain
androgen receptors in response to testosterone treatment is not a limiting factor for the

precocious display of male reproductive behaviors.

To my wife and best friend Kris,

Thanks for all the support.

iv

ACKNOWLEDGMENTS

I extend my gratitude to my mentor and friend Dr. Cheryl Sisk for your many
contributions to my development as a scientist. You freely contributed your time,
knowledge, experience, and understanding to my professional development and I have
beneﬁted greatly from your efforts. I wish you continued success in the future.

I also thank Drs. Antonio Nunez, Laura Smale, James Ireland and John 1. Johnson
for taking a genuine interest in my career development, and for insightful contributions
to this dissertation.

I am grateful to my fellow graduate students Kevin Sinchak, Yu Ping Tang, Alan
Elliott, Colleen Novak, and Cathy Katona for your friendship, assistance, support, and
for the many good times over the years.

Thanks also to the many undergraduate students who contributed valuable time
and effort to my research including Cristina Grigorean, Rebecca Zacks, Penny Shek,
Michael Hayes, John Arbogast, Kevin Tenbrink, and Bridget Swanberg.

Thanks ﬁnally to my family for their love, understanding and support. Thanks

for always being there.

TABLE OF CONTENTS

List of Tables

List of Figures

List of Abbreviations.

Introduction

Background and Signiﬁcance
Photoperiodic Regulation of Pubertal Maturation
Activation of Hypothalamic-Pituitary-Gonadal Axis
Activation of Reproductive Behavior
Contribution of Testosterone and its Metabolites
Hypotheses and Predictions

Experiment 1. Distribution of androgen receptor containing cells in
male ferrets before and after the onset of pubertal maturation.

Methods

Results

Discussion
Experiment II. Time-course for the appearance of androgen receptor-
immunoreactive cells in the brains of male ferrets undergoing pubertal
maturation under different photoperiod conditions.

Rationale

Methods

Results

Discussion

Experiment HI. Regulation of androgen receptor-immunoreactive cells
in the brains of prepubertal male ferrets by gonadal steroids.

Rationale

vi

viii
ix

xiii

10

11
18
22

24
24
26

3O

33

Methods 35
Results 36
Discussion 44

Experiment IV. Regulation of AR-IR cell density by gonadal steroids in
pre- and postpubertal male ferrets.

Rationale
Experiment Iva: 47
Experiment Ivb: 49
Methods 53
Results 56
Experiment Iva: 58
Experiment Ivb: 64
Discussion 79
General Discussion 92
Appendices 99

Appendix A: Distribution of Androgen Receptor-Immunoreactive
Cells in the Forebrain of Male Ferrets

Introduction 99
Methods 101
Results 102
Discussion 124

Appendix B: Effects of Long-term Storage in Cryoprotectant on the
Staining of AR-IR Cells 128

List of References 130

vii

LIST OF TABLES

Table 1. Summary table for the three-way analysis of variance from Experiment IVa.

viii

LIST OF FIGURES

Figure 1. Schematic diagram of a cell indicating the cellular mechanism of steroid
hormones. Testosterone (T) can alter cellular function by binding to the androgen
receptor (AR) as itself, or as dihydrotestosterone (D) following 50- reduction. T can also
exert its actions by acting through the estrogen receptor (E) following aromatization.
Steroid-receptor complexes alter transcription of speciﬁc genes which contain androgen
(ARE) or estrogen (ERE) response elements.

Figure 2. Speciﬁcity of androgen receptor irnmunoreactivity was conﬁrmed by the lack of
immunostaining when the tissue was processed in the absence of primary antiserum (A),
or by preincubation of the primary antiserum with 10M excess of the peptide fragment
against which it was raised (B). Preincubation of the primary antiserum with a distant
peptide fragment of the human androgen receptor did not inhibit immunostaining (C).
Androgen receptor-immunoreactivity using the standard irnmunocytochemical staining
procedure is shown in D. Bar = 10 um for all panels.

Figure 3. Line drawings of representative brain sections which were anatomically matched
for quantiﬁcation of the number of AR-IR cells. The position in which the reticle grid was
placed in each region is marked by the square (I). Not all regions were examined in each

experiment.

Figure 4. Mean (iSEM) paired testis weight (g) and plasma testosterone concentration
(ng/ml) in 12 wk old and 20 wk old male ferrets. Asterisk designates statistically
signiﬁcant differences between the two groups (p < .05).

Figure 5. Mean (rhSEM) number of androgen receptor-immunoreactive (AR-IR) cells /
15,625 um2 in speciﬁc brain regions of 12 wk old and 20 wk old male ferrets. Asterisk
designates statistically signiﬁcant differences between the two groups (p < .05).

Figure 6. Low (A and B) and high (C and D) power photomicrographs of androgen
receptor-immunoreactivity in the POA ﬁ'om representative 12 wk old (A and C) and 20
wk old (B and D) male ferrets. Bar = 200nm (A and B) or 20pm (C and D).

Figure 7. Mean (iSEM) paired testis weight (g) and plasma testosterone concentration
(ng/ml) in male ferrets undergoing spontaneous pubertal maturation in short days or
photoinduced pubertal maturation in long days. Asterisk designates signiﬁcantly different
than 12 wk old prepubertal controls (p<.05).

ix

Figure 8. Mean (:tSEM) number of androgen receptor-immunoreactive (AR-IR) cells /
15,625 um2 in speciﬁc brain regions of male ferrets undergoing spontaneous pubertal
maturation in short days or photoinduced pubertal maturation in long days. Symbols: (a)
signiﬁcantly different from 12 wk old prepubertal controls (ANOVA, p<.05), (b)
significantly different from 20 wk old ferrets in short days (ANOVA, p<.05), (c)
signiﬁcantly different from 12 wk old prepubertal controls (Dunnett‘s tests, p<.05).

Figure 9. Photomicrographs of AR-IR cells taken from the periventricular preoptic area
(pvPOA; A-E), and the lateral portion of the ventromedial hypothalamus (IVMH; F -J)
from oil-injected intact animals (A,F), and those castrated and injected once daily for 10
days with either 5 mg/kg testosterone propionate (B,G), 5 mg/kg dihydrotestosterone
propionate (C,H), IOug/kg estradiol benzoate (D,I), or oil vehicle (EJ). The pvPOA is
an example of a region in which androgen treatment (T or DHT) resulted in a signiﬁcantly
greater number of AR-IR cells compared with intact males (cf A with B and C). The
lVMH is an example of a region where androgen treatment did not increase the number of
AR-IR cells above the number seen in intact animals (cf F with G and H). Arrows in E
and I point to AR-immunoreactivity in the cytoplasm, which was seen primarily only in
tissue ﬁom castrated animals. All micrographs are of tissue incubated in DAB for 60
minutes. Magniﬁcation bar = 10 um.

Figure 10. Mean (iSEM) number of androgen receptor-immunoreactive (AR-IR)
cells/15,625 pm2 in speciﬁc brain regions of intact and castrated prepubertal male ferrets
injected (sc) once daily with either 5 mg/kg testosterone propionate (T), 5 mg/kg
dihydrotestosterone propionate (D), 10 ug/kg estradiol benzoate (E), or oil vehicle
(castrate (C) and intact (I) animals) for 10 days. Animals were sacriﬁced 4 hr following
the last injection. Regions in which androgen treatment signiﬁcantly increased the number
of AR-IR cells compared with intact animals are shown in A. Regions in which androgen
treatment restored the number of AR-IR cells to that seen in intact animals are shown in
B. Asterisk (*) designates signiﬁcant difference from intact, T and DHT treatment.
Double asterisk designates signiﬁcant difference from T and DHT treated animals.

Figure 11. Percent change relative to intact prepubertal male ferrets in the number of
AR-IR cells/15,625 um2 in speciﬁc brain regions of castrated prepubertal male ferrets
treated with either 5 mg/kg testosterone propionate, 5 mg/kg dihydrotestosterone
propionate, IOug/kg estradiol benzoate, or oil vehicle (sc) once daily for 10 days.

Figure 12. Mean (iSEM) testosterone concentration (ng/ml) in prepubertal and adult
ferrets injected with oil or 5.0 mg/kg testosterone propionate for either 30 min, 4 hr, 8 hr,
or daily for 10 days. Bars with different letters are signiﬁcantly different from one another

(p< 0.05).

Figure 13. Mean (:hSEM) number of AR-IR cells / 10,000 um2 in individual brain regions
from prepubertal (open bars) and adult (closed bars) male ferrets injected daily for 10

days with either oil (intact and castrate), 10 rig/kg E, 5.0 mg/kg T, or 5.0 mg/kg DHT.

Figure 14. Photomicrogaphs of the pvPOA from prepubertal (A-E) and adult (F-J) male
ferrets that remained intact (a, f) or were injected daily for 10 days with either 5.0 mg/kg

T (b, g), 5.0 mg/kg DHT (c, h), oil (d, i) or 10 rig/kg E (e, j). Magniﬁcation bar = 30 um.

Figure 15. Photomicrogaphs of the MeA from prepubertal (A-E) and adult (F-J) male
ferrets that remained intact (a, f) or were injected daily for 10 days with either 5.0 mg/kg

T (b, g), 5.0 mg/kg DHT (c, h), oil (d, i) or 10 jig/kg E (e, j). Magniﬁcation bar = 30 um.

Figure 16. Mean (iSEM) number of AR-IR cells / 10,000 um2 collapsed across brain
region ﬁ'om prepubertal (open bars) and adult (closed bars) male ferrets injected daily for
10 days with either oil (intact and castrate), 10 ug/kg E, 5.0 mg/kg T, or 5.0 mg/kg DHT.
Asterisk indicates signiﬁcant difference from prepubertal animal in same steroid condition,
double cross indicates signiﬁcant difference from all treatment groups withi a age group,
single cross indicates signiﬁcant diffrence from I, T and D within a age goup.

Figure 17. Mean (:hSEM) number of AR-IR cells / 10,000 um2 collapsed across steroid
treatment from prepubertal (open bars) and adult (closed bars) male ferrets. Asterisk
indicates signiﬁcant difference from prepubertal animals within the same brain region.

Figure 18. Mean (iSEM) number of AR-IR cells / 10,000 tun2 within individual brain
regions collapsed across age group. Signiﬁcant differences only apply to comparisons
within a given brain region. Double asterisk indicates signiﬁcant difference from T and D,
single asterisk indicates signiﬁcant difference from T, D and I, cross indicates signiﬁcant
difference from T, solid circle indicates signiﬁcant difference from E within brain region.

Figure 19. Mean (:tSEM) immunocytochemical staining intensity from intact and T
treated prepubertal and adult male ferrets collapsed across brain region. Bars with
different letters are signiﬁcantly different from one another (Mann-Whitney U test, p <
0.05).

Figure 20. Mean (:hSEM) immunocytochemical staining intensity from prepubertal and
adult intact animals, and those castrated and injected with oil or 5.0 mg/kg T for either 30
min, 4 hr, 8 hr, or 10 days. Bars with different letters are signiﬁcantly different from one
another (Mann-Whitney U test, p < 0.05).

Figure 21. Mean (iSEM) number of AR-IR cells / 10,000 m2 collapsed across brain
region ﬁ'om adult animals castrated and injected with oil or 5.0 mg/kg T for either 30 min,
4 hr, 8 hr, or 10 days. Bars with different letters are signiﬁcantly different from one
another (p < 0.05).

Figure 22. Mean (:tSEM) number of AR-IR cells / 10,000 m2 in individual brain regions

from adult animals castrated and injected with oil or 5.0 mg/kg T for either 30 min, 4 hr, 8
hr, or 10 days. Statistical analysis was not performed in individual regions.

xi

Figure 23. Mean (:tSEM) immunocytochemical staining intensity, collapsed across brain
region, from adult animals castrated and injected with oil or 5.0 mg/kg T for either 30 min,
4 hr, 8 hr, or 10 days. Bars with different letters are signiﬁcantly different from one
another (Mann-Whitney U test, p < 0.05).

Figure 24. Photomicrographs from the PMV, and lVMH from adult animals castrated and
injected with oil (a, t) or 5.0 mg/kg T for either 30 min, (B, g), 4 hr (c, h), 8hr (d, I), or
once daily 10 days (e, j). Magniﬁcation bar = 100 um.

Figures 25-34. Maps of the distribution of androgen receptor-immunoreactive cells in the
forebrain of an adult male ferret castrated and injected with 5.0 mg/kg testosterone. Each
dot represents one AR-IR cell visible at a magniﬁcation of 70 times.

Figure 35. Photomicrogaphs of AR-IR cells within circumventricular organs of the male
ferret brain. (A), subfomical organ; (B), median eminence; (C), vascular organ of the
lamina terminalis; and (D), choroid plexus. Magniﬁcation bar = 250 um.

Figure 36. Photomicrogaphs of AR-IR cells within the anterior (A), medial and cortical
(B), arnygdaloid nuclei of an adult male ferret. Magniﬁcation bar = 250 um.

Figure 37. Photomicrogaphs of AR-[R cells within the preoptic area of an adult male
ferret. Magniﬁcation bar = 250 um.

Figure 38. Photomicrogaphs of AR-IR cells within the anterior hypothalamus (A),
retrochiasmatic area (B), ventromedial hypotlnalamic nucleus and arcuate nucleus (C and
D), of an adult male ferret. Magniﬁcation bar = 250 um.

Figure 39. Photomicrogaphs of the male nucleus of the preoptic areal andterior
hypothalamus of an adult male ferret (upper arrow panel A). The ventral nucleus is
indicate dby lower arrow. A and C are nissl stained sections adjacent to the AR-IR

stained sections in B and D. Magniﬁcation bar = 250 pm (A and B), 100 um (C and D).
Figure 40. Photomicrogaphs of AR-IR cells in (A) the caudal portion of the arcuate

nuclwus and ventral premammillary nucleus, and (B), in the medial, and lateral marnmillary
nuclei, and the posterior hypothalamus. Magniﬁcation bar = 250 um.

xii

LIST OF ABBREVIATIONS

! . l l l l . .
3V third ventricle

AA anterior arnygdaloid area

ac anterior commisure

Acb nucleus accumbens

ACO cortical arnygdaloid nucleus, anterior
AHA anterior hypothalamic area

AHi arnygdalo-hippocampal transition area
ALA lateral arnygdaloid area

APir amygdalo-piriform transition area
BMA basomedial arnygdaloid nucleus
BNST bed nucleus of the stria terminalis

Ca caudate nucleus

CA cornu ammonis

cc corpus callosum

Ce central arnygdaloid nucleus

DG dentate gyrus

DM dorsomedial hypothalamic nucleus

f fomix

FF ﬁelds of Forel

HDB nucleus of the diagonal band, horizontal limb
ic internal capsule

LD lateral dorsal nucleus of the thalamus
LHA lateral hypothalamic area

LM lateral mammillary nucleus

10 lateral olfactory tract

LOT nucleus of the lateral olfactory tract
LS lateral septa] nucleus

LV lateral ventricle

ME median eminence

MeA medial arnygdaloid nucleus, anterior
MeP medial arnygdaloid nucleus, posterior
MM medial mammillary nucleus

MnPO median preoptic nucleus

MPA medial preoptic area

IVIES Innreclirrl snarrtnrrrn

mt mammilothalamic tract

MV medioventral thalamic nucleus
mVMH ventromedial hypothalamic nucleus, medial

xiii

opt optic tract

OVLT vascular organ of the lamina terminalis
ox optic chiasm

PC paracentral thalamic nucleus

PCo cortical arnygdaloid nucleus, posterior
PH posterior hypothalamic area

Pir piri form cortex

PMV premammillary nucleus, ventral

POA preoptic area

PT paratenial thalamic nucleus

PV paraventricular thalamic nucleus

PVH paraventricular hypothalamic nucleus
RCh retrochiasmatic area

Sch suprachiasmatic nucleus

SFi septoﬁmbrial nucleus

SFO subfomical organ

SHy septohypothalamic nucleus

sm stria medularis

SOa nucleus, anterior

SOC supraoptic nucleus, caudal

SuM supramammillary nucleus

Tu olfactory tubercle

VDB nucleus of the diagonal band, vertical limb
VMH ventromedial nucleus of the hypothalamus
VP ventral pallidum

ZI zona incerta

5 l l l l . .

ANOVA analysis of variance

AR androgen receptor

AR-IR androgen receptor-immunoreactive
ATD androstendione .......

DAB diaminobenzidine

DHT dihydrotestosterone

E estradiol

LH luteinizing hormone

LHRH luteinizing hormone releasing hormone
PBS phosphate buffered saline

T testosterone

TBS tris buffered saline

xiv

INTRODUCTION

Puberty in males is characterized by the onset of gonadal maturation brought
about by an increase in the frequency of episodic release of luteinizing hormone-
releasing hormone (LHRH) from the hypothalamus, a concomitant increase in the
frequency of release of luteinizing hormone from the anterior pituitary, and testosterone
from the testes. Two well known actions of testosterone on the central nervous system
are the suppression of gonadotropin secretion from the anterior pituitary, and the
activation of male reproductive behavior. The interaction of testosterone with target cells
in the central nervous system is dynamic during pubertal maturation. First, there is a
decline in the ability of testosterone to suppress gonadotropin secretion as the animal
undergoes sexual maturation, (Le, a change in the setpoint for steroid negative feedback
regulation of gonadotropin secretion). In some species, this appears to be the mechanism
by which frequent LHRH release is initiated at the onset of puberty. Second, there is an
increase in the ability of testosterone to activate male reproductive behavior (re, the
CNS becomes more responsive to behavioral actions of gonadal steroids). The timing of
puberty is variable between species and within species depending upon environmental
and nutritional factors, however regardless of when pubertal maturation occurs, these
patterns of neuroendocrine and behavioral changes occur in virtually all mammalian

species examined. Understanding the mechanisms underlying these changes in

2

responsiveness to testosterone during sexual maturation will increase our understanding
of one of the fundamental developmental periods of an animal‘s reproductive life.

The working hypothesis of the Sisk laboratory is that target tissue responsiveness
to steroid hormones is in part mediated by the number of steroid receptor containing cells
and/or the number of steroid receptors per cell. One prediction of this hypothesis is that
pubertal changes in the responsiveness to steroid hormones are positively correlated with
changes in the number of brain cells which contain steroid receptors. Experiments in this
dissertation address one aspect of this working hypothesis, namely that changes in the
abundance of androgen receptor-containing cells during the process of pubertal
maturation is one potential mechanism for pubertal shifts in responsiveness to
testosterone. Using the European ferret as an animal model, the experiments in this
dissertation investigate l) the distribution of androgen receptor containing cells in
forebrain structures, 2) pubertal changes in the density of androgen receptor-containing
brain cells, and 3) changes in the density of androgen receptor-containing cells as a
function of steroidal milieu. Particular attention will be focused on brain regions
implicated in the control of both reproductive behaviors and negative feedback on

gonadotropin secretion.

Background and Signiﬁcance
The European ferret was used as an animal model for these experiments. In this
section, some characteristics of this model will be described and the signiﬁcance of this
research will be placed in the context of understanding the process of pubertal

maturation.

3

Photoperiodic Regulation of Pubertal Maturation

The European ferret (Mustela puroriusﬁrro) is a photoperiod-sensitive mammal
that is reproductively active in the long days of spring and early summer (Neal, Murphy,
Moger, & Oliphant, 1977). Sexual maturation of juvenile male ferrets, as measured by
increased testicular size and serum testosterone concentrations, begins in December when
ferrets are 20-25 wks old and, as daylength increases following the winter solstice.
However, this increase in daylength is not required for sexual maturation, since ferrets
maintained continuously in Short day lengths (e. g. 8 hr light/day) under artiﬁcial light
conditions also experience the onset of puberty at about 18-20 wks of age (Sisk, 1990).
Nevertheless, early onset of sexual maturation can be induced in laboratory housed male
ferrets by transferring them from short day to long day (e. g. 18 hr light/day) conditions at
12 wk of age (Sisk, 1990). Whether sexual maturation occurs spontaneously in short
days, or following a photoperiod transition to long days, quantitative characteristics of
luteinizing hormone (LH) pulse frequency and testosterone (T) levels are similar once

sexual maturation is reached (Sisk, 1987; Sisk, 1990).

Activation of Hypothalamic-Pituitary-Gonadal Axis

The transition ﬁ'om an immature (low frequency) pattern of LH release to a
reproductively mature (high frequency) pattern is the result of a shift in responsiveness of
steroid sensitive neural circuits to the negative feedback effects of T. The immature
ferret is fully capable of secreting adult levels of LH in the absence of gonadal steroids,
and a dose of T which virtually abolishes LH secretion in the castrated prepubertal male

ferret is less effective in lowering LH secretion in a castrated adult male ferret (Sisk,

4

1987). Furthermore, the pituitary gland of gonadally intact prepubertal male ferrets is
capable of secreting LH in response to exogenous administration of LHRH (Berglund &
Sisk, 1990). Therefore the pubertal increase in LH pulse ﬁequency is the result of a

reduction in the negative feedback effects of T on LHRH release.

Activation of Reproductive Behavior

Studies in the Sisk laboratory have demonstrated that pubertal maturation of
sexual behavior patterns also involves a shift in responsiveness to gonadal steroids. In
this component of puberty, steroid sensitive neural circuits in the brain become more
responsive to the activational effects of T on male sexual behavior. Prepubertal male
ferrets display almost none of the characteristic reproductive behaviors (i.e. neck
gipping, mounting and thrusting) when tested with a receptive female (Sisk, Berglund,
Tang, & Venier, 1992). Although some of these characteristic behavior patterns begin to
emerge in castrated prepubertal males when injected with high doses of T, these behavior
patterns can be activated in castrated adult males with a much lower dose of T.
Interestingly, adults that undergo pubertal maturation following a photoperiod transition
show an increased responsiveness to the behavioral actions of T on male sexual behavior
compared to adults that undergo pubertal maturation in short days. When castrated and
injected with T, adults raised in long days respond with an increase in sexual behaviors to
lower doses of T than either adults raised in short days, or prepubertal males. Adults
raised in short days in turn respond to lower doses of T than prepubertal males. These
data indicate that steroid sensitive neural circuits are more responsive to T in adults

compared with those in prepubertal males, and that the photoperiod conditions under

5

which pubertal maturation occur, and/or the photoperiod under which the animals were

tested, inﬂuence the responsiveness to the activational effects of T in adulthood.

Contribution of Testosterone and its Metabolites
Negative Feedback: The central actions of T are often mediated following its
conversion to either estrogen (E) following aromatization, or dihydrotestosterone (DHT)

following Sci-reduction (Figure 1). These steroids then bind to their respective steroid

receptors and alter transcriptional activity of genes which contain steroid response
elements (Beato, 1989; Evans, 1988; O‘Malley, Schrader, & Tsai, 1986; Truss & Beato,
1993; Yamamoto, 1985). It has been clearly demonstrated that estrogens can mediate
negative feedback on gonadotropin secretion (Ellinwood, Hess, Roselli, Spies, & Resko,
1984; Kalra & Kalra, 1980; Schanbacher, 1984; Worgul, Santen, Samojlik, Irwin, &
Falvo, 1981). Administration of the aromatase inhibitor ATD to male ferrets in breeding
season elevates the number of LH pulses and the level of circulating LH (Carroll &
Baum, 1989). Furthermore, mean LH levels, and LH pulse amplitudes were not
signiﬁcantly different in castrated males when compared to intact males given ATD
(Carroll & Baum, 1989). However, androgens themselves induce negative feedback on
gonadotropin secretion. Implantation of DHT directly into the mediobasal
hypothalamus of intact rats decreased circulating levels of LH, without increasing the
circulating level of DHT, indicating that hormone leakage into the periphery and a
possible direct negative feedback action on the pituitary are unlikely to fully account for
the decline in LH secretion (Kalra & Kalra, 1980). This is further supported by the

reduced effectiveness of peripheral implantation of DHT or T ﬁlled silastic capsules to

 

 

 

 

 

Figure 1. Schematic diagram of a cell indicating the cellular mechanism of steroid
hormones. Testosterone (T) can alter cellular function by binding to the androgen
receptor (AR) as itself, or as dihydrotestosterone (D) following Sat-reduction. T
can also exert its actions by acting through the estrogen receptor (E) following
aromatization. Steroid-rece tor cfomplexes alter transcription of speciﬁc genes
which contain androgen (A ) or estrogen (ERE) response elements.

7

induce negative feedback following peripheral administration of the androgen receptor
blocker ﬂutamide (Kalra & Kalra, 1980). In the castrated adult male ferret, analysis of
plasma LH levels following implantation of silastic capsules of DHT has shown that
DHT is a potent inhibitor of LH secretion (Tang & Sisk, 1988). Although a direct
negative feedback effect on the pituitary gland cannot be ruled out when DHT or
ﬂutamide are administered peripherally, the effectiveness of central implants of DHT in
reducing LH secretion indicate that the binding of an androgen to the androgen receptor
within the CNS may be directly involved in the negative feedback of gonadotropin
secretion.

Reproductive Behavior: A great deal of research has been directed toward the
investigation of whether T itself, or a metabolite of T such as E, is responsible for the
activation of male reproductive behavior. This work led to the ”aromatization
hypothesis” stating that the metabolic conversion ﬁ'om T to E via aromatization is
essential for the activation of male sexual behavior in a variety of species. This was
based on evidence from many species, including the ferret, indicating that androgens
subject to aromatization such as T, or androstenedione, are effective in maintaining
copulation, while androgens not subject to aromatization, such as DHT, are ineffective
(Meisel & Sachs, 1994). Furthermore, administration of E alone to castrated adult male
ferrets can restore neck gipping, mounting and pelvic thrusting (Baum, 1976), while
blocking the conversion of androgens to estrogens can inhibit male reproductive
behaviors. In intact adult male ferrets, silastic capsules containing the aromatase
inhibitor ATD signiﬁcantly reduced the occurrence of neck gipping, mounting and

intromissive behaviors compared to ferrets implanted with empty capsules (Carroll,

Weaver, & Baum, 1988).

In spite of good evidence that aromatization of T to E and, by implication, action
of the estrogen receptor, is important for the behavioral effects of T, there are several
lines of converging evidence indicating that androgens, acting on the androgen receptor,
also contribute directly to the activation of male reproductive behaviors. For example,
the combined treatment with E and DHT more closely approximates the behavior pattern
seen when castrate male ferrets receive T replacement, than when E is administered alone
(Baum & Vreeburg, 1983). In the castrated rat, DHT is effective in activating copulatory
behavior when administered in conjunction with doses of E that are ineffective when
administered alone (Baum & Vreeburg, 1973; F eder, Naﬁolin, & Ryan, 1974; Lanson,
Perez-Palacios, Morali, & Beyer, 1975; Morali, Lennus, Oropeza, Garcia, &
Perez-Palacios, 1990). Since DHT cannot be aromatized to E, these data suggest that
androgen receptors themselves are mediating these effects. In this same context,
administration of the androgen receptor blocker ﬂutamide eliminates the restoration of
both intromissions and ejaculations following T treatment (Gladue & Clemens, 1980).
Flutarnide treatment is most effective when the dose of T used activates a moderate
(Gladue & Clemens, 1980) rather than a high (Gray, 1977) level of copulatory behavior,
and is likely due to the low afﬁnity of ﬂutamide for the androgen receptor (Liao, Howell,
& Chang, 1974). Further support for this hypothesis is data indicating that a metabolite
of ﬂutamide, Sch 16423, which has a higher afﬁnity for the androgen receptor was able
to eliminate mounting in 80% of rats treated with a dose of T which restored ejaculation

in all control animals (McGinnis & Mirth, 1986).

9

Indirect Role of Androgen Receptors: The activity of the aromatase enzyme is
regulated by androgens in several species (Connolly, Roselli, & Resko, 1990; Roselli,
Ellinwood, & Resko, 1984; Steimer & Hutchison, 1981; Weaver & Baum, 1991),
presumably through an androgen receptor mechanism (Roselli, Horton, & Resko, 1987;
Roselli & Resko, 1984). The regulatory role of androgens on aromatase activity is brain
region speciﬁc. In castrated adult male ferrets the magnitude of aromatase activity is
similar in the bed nucleus of the stria terminalis, medial and lateral preoptic area, medial
and lateral amygdala, and the ventromedial hypothalamic nucleus (Weaver & Baum,
1991). However, following treatment with DHT, aromatase activity was stimulated in
the medial preoptic area, medial amygdala, and ventromedial hypothalamic nucleus, but
this same treatment did not stimulate aromatase activity in other brain regions. Further
evidence for androgen dependent regulation of aromatase activity is seen in the rat.
Castration reduces aromatase activity in the preoptic area, but not the amygdala of male
rats, while androgen administration restores aromatase activity in the preoptic area to
levels greater than intact controls (Roselli, et al., 1984; Roselli, Horton, & Resko, 1985;
Roselli & Resko, 1984). Furthermore, treatment of intact male rats with ﬂutarnide
reduces aromatase activity in the hypothalamus-preoptic area (Roselli & Resko, 1984).
Thus, even in cases where the effects of T on reproductive behaviors and gonadotropin
secretion are ultimately mediated by E, the presence of androgen receptors may be
required for this aromatization to occur. Further support for this hypothesis is the fact
that in male hamsters housed in short days, a situation in which steroid responsiveness to
the effects of T on behavioral and endocrine parameters is similar to that of a prepubertal

male, there is a reduction in hypothalamic aromatase activity compared with those

10

hamsters housed in long days, and androgen administration to short day hamsters fails to
increase hypothalamic aromatase activity compared with those housed in long days
(Callard, Mak, & Solomon, 1986). Therefore, changes in the relative abundance of
androgen receptors during the process of pubertal maturation may both directly and

indirectly modulate responsiveness to the actions of testosterone on the nervous system.

Hypotheses and Predictions

The working hypothesis of this dissertation is that target tissue responsiveness to
steroid hormones is in part mediated by the number of cells that express steroid receptor.
While it is clear that both estrogen receptors and androgen receptors are involved in
mediating the biological responses obtained following androgen administration,
experiments in this dissertation only address the possible role of androgen receptors.
Based on our working hypothesis, it was predicted that increases in androgen receptor
containing cells would occur in brain regions which are known to be involved in the
control of male sexual behavior. Second, it was predicted that a decrease in androgen
receptor containing cells would occur in brain regions involved in the control of the
negative feedback actions of T on gonadotropin secretion. Initial experiments in this
dissertation demonstrate that there is indeed an increase in density of androgen receptor
containing cells in some brain regions during pubertal maturation (Experiments I and 11).
Since the increase in circulating levels of T during pubertal maturation is one of the most
dramatic endocrine changes during puberty, and several lines of evidence indicate that
testosterone can regulate the expression of the androgen receptors, Experiments III and

IV speciﬁcally address the hypothesis that the differential behavioral responsiveness to

11

the effects of T in pre- and postpubertal animals is related to differences in the regulation
of the androgen receptor by androgens themselves. Thus if behavioral responsiveness to
T is related to the ability of T to upregulate the level of AR in speciﬁc brain regions, then
an identical dose of T to pre- and postpubertal ferrets would be predicted to result in a
different pattern of AR immunostaining. Speciﬁcally, it was predicted that androgen
treatment to prepubertal animals would result in a less robust increase in AR-IR cells

than the same treatment to adults.

Experiment 1. Distribution of androgen receptor containing cells in male ferrets
before and after the onset of pubertal maturation.

As an initial investigation of developmental changes in androgen-sensitive brain
regions during pubertal maturation, androgen receptor-immunoreactive (AR-IR) cells
were identiﬁed in the brain of male ferrets before and after puberty onset.

Methods
Animals and Housing

All animals were treated in accordance with the NIH Guide for the Care and Use
of Laboratory Animals using protocols that were approved by the Michigan State
University All-University Committee for Animal Use and Care. Thirty-six weanling (7
wk old) male ferrets were purchased ﬁ'om Marshall Farms (North Rose, NY) and housed
under short days (8 hr light per day) in pairs in stainless steel cages (51 x 60 x 38 cm) in
a temperature controlled (23° 1: 1°C) colony room. Purina Ferret Chow (Ralston Purina,

St. Louis, MO) and water were available ad libitum.

12

Procedure

At 12 wks of age, a group of ferrets (n=16) was transferred from short days to a
long day photoperiod (16 hr light per day) to induce pubertal maturation, while the
remaining ferrets (n=20) remained in short days to undergo spontaneous pubertal
maturation. Under these two photoperiodic conditions, an increase in testicular size
occurs within 2 wks following the transfer from short to long days (i.e., by 14 wk of
age), while an increase in testicular size does not occur until approximately 18 wk of age
in ferrets remaining in short days (Sisk, 1990). A group of four ferrets in short days was
sacriﬁced at 12 wk of age, and additional groups of four ferrets in botln short days and
long days were sacriﬁced at 13.5, 15, 17.5, and 20 wks of age. Animals were
anesthetized with methoxyﬂurane anesthesia (Metofane; Pittrnan—Moore; Washington
Crossing, NJ), a blood sample (3 ml) was obtained via cardiac puncture, and the testes
were removed via a mid-scrotal incision and weighed. Ferrets were then injected with
Equithesin anesthetic (2.5 ml/kg ip) and perfused intracardially with 250 ml of 0.5 M
Sorensen‘s phosphate buffer containing 0.8% NaCl, 0.8% sucrose, 0.4% B-D-glucose and
3 IU/ml heparin, pH 7.4, followed by ﬁxative containing 4% paraformaldehyde and
0.5% glutaraldehyde in 0.1 M phosphate buffered saline (PBS), pH 7.5. Brains were
removed from the skull, placed in ﬁxative containing 15% sucrose, and stored at 4°C
until sectioning. Brains were cut into 40 pm thick sections which were stored in a
cryoprotectant solution at -20°C (Watson, Weigand, Clougln, & Hoffman, 1986).
Testosterone Radioimmunoassay

Plasma concentrations of T in the terminal blood samples were measured with

reagents in the Coat-a-Count Total Testosterone Kit (Diagnostic Products, Los Angeles,

13

CA). The limit of detectability of the assay was 0.26 ng/ml, and the intra-assay
coefﬁcient of variation was 9.8%.
Androgen Receptor Immunocytochemistry

Every sixth section from each brain was processed for immunocytochemistry, and
tissue from both treatment groups was processed simultaneously. Sections were rinsed
10 times in 0.1 M PBS to remove the cryoprotectant solution, and then incubated
sequentially in 0.1 M PBS containing 0.1 M glycine for 30 min, 0.5% H202 for 30 min,
4% normal goat serum for 30 min (V ectastain ABC Elite Kit, Vector Laboratories,
Burlingame, CA), and rabbit anti-androgen receptor [PG-21 obtained from Dr. Gail
Prins, Michael Reese Hospital, Chicago, IL, 0.5 ug/ml in 0.1 M PBS containing 0.2%
Triton X-100 (PBS-TX); 36 h]. Sections were then incubated in secondary antibody
(goat anti-rabbit immunoglobulins, Vectastain ABC Elite Kit, 1:200, 4 h), avidin-biotin-
HRP complex (V ectastain ABC Elite Kit, 1:200, 2 h) and 1.0 % 3,3‘-diaminobenzidine
(DAB) containing 2% B—D(+) glucose, 0.04% NILCl, 0.038% imidazole, 50 ul/ml of
250mM NiClz, and 0.0075% I-IZO2 in 0.05 M Tris buffered saline (TBS; 10 min).
Sections were rinsed 3 times in PBS-TX between each incubation with the exception of
the rinse immediately preceding the DAB reaction which was in TBS. All incubations
were at room temperature except primary antiserum, which was at 4°C. Following the
DAB reaction, all sections were rinsed 5 times in PBS, mounted onto gelatin coated
slides, dried, dehydrated and coverslipped.

Speciﬁcity of immunostaining was analyzed by competition and deletion studies.
The primary antiserum is directed against a peptide fragment containing the ﬁrst 21

amino acids of the human androgen receptor (Prins, Birch, & Greene, 1991).

14

Preincubation of the primary antiserum with 10 M excess of the peptide ﬁ'agnent against
which it was raised virtually eliminated immunocytochemical staining (Figure 2B), while
preincubation with a distant fragnent of the human androgen receptor (amino acids 462-

478) did not inhibit immunostaining (Figure 2C). Processing tissue in the absence of the

primary antiserum resulted in no detectable immunostaining by the secondary antibody

(Figure 2A). Speciﬁc immunostaining is shown in Figure 2D.

Tissue Analyzed in Experiment I

In this laboratory, brain tissue is typically processed into several sets of equally
spaced brain sections allowing for the analysis of multiple anatomical dependent
measures. For this initial study of pubertal regulation of androgen receptors, only tissue
from 12 wk old prepubertal ferrets and 20 wk old ferrets housed in short days was
processed for androgen receptor immunocytochemistry. This was the most efﬁcient
manner to begin answering the questions posed in the above sections, since it was
important ﬁrst to establish whether there are differences in AR expression before and
after puberty onset.

Quantiﬁcation of AR-IR Cells

A detailed description of the distribution of AR-IR cells in an adult T treated
male is provided in Appendix A. As shown in Appendix A, literally thousands of cells
that stain for AR can be found in a single section under the appropriate conditions. It
was therefore not feasible to attempt to quantify the total number of AR-IR containing
cells in each brain region. Estimates of AR-IR cell abundance were made by sampling

the number of AR-IR cells within a circumscribed area at a high magniﬁcation. Thus the

 

Figure 2. Speciﬁcity of androgen receptor immunoreactivity was conﬁrmed by the
lack of immunostaining when the tissue was processed in the absence of primary
antiserum (A), or by preincubation of the primary antiserum with 10M excess of
the peptide fragment against which it was raised (B). Preincubation of the primary
antiserum with a distant peptide fragment of the human androgen receptor did not
inhibit immunostaining (C). Androgen ‘ uty using the
standard immunocytocherrnical staining procedure 18 shown in D. Bar

= 10 pm for all panels.

16

density of AR-IR cells was quantiﬁed in a given brain region by counting (at 1000x) the
number of immunopositive cells present within an eyepiece gid (125 pm by 125 um)
positioned over the region. The gid was placed over this same relative position in two
sections separated by 240 um, and the two sections were anatomically matched across
animals. Figure 3 schematically illustrates the position of the gid within each of the
brain regions analyzed. Bilateral cell counts were made for each region, and the number
of cells/grid/region was averaged across the 4 counts for each animal. Occasionally the
tissue quality in a given region was not adequate for quantiﬁcation, and in such cases,
data ﬁ'om that region for that animal were omitted from the analysis.

The density of AR-IR cells was quantiﬁed in the lateral septum (LS), bed nucleus
of the stria terminalis (BNST), medial preoptic area (MPA), periventricular preoptic area
(pvPOA), retrochiasmatic area (RCh), anterior medial amygdala (MeA), arcuate nucleus
(ARC), and the lateral portion of the ventromedial nucleus of the hypothalamus (IVMH).
The LS, MPA, BNST and MeA are brain regions which have been identiﬁed as
components of the steroid-sensitive neural circuit controlling male reproductive
behaviors (Kondo, Shinoda, Yarnanouchi, & Arai, 1990; Powers, Newman, & Bergondy,
1987). The ARC, RCh and VMH are structures known to contain androgen binding
cells (Sar & W.E., 1975; Vito, Baum, Bloom, & Fox, 1985), but they have not been
directly implicated in the control of male reproductive behavior. We have used the
anatomical nomenclature adopted by Paxinos and Watson (1986) to refer to these areas,
but cannot say whether responses observed in the precise areas examined in the present

study would generalize to previously identiﬁed subdivisions in rodent brain.

17

 

Figure 3. Line drawings of representative brain sections which were anatomically
matched for quantiﬁcation of the number of AR-IR cells. The position in which
the reticle grid was placed in each region is marked by the square (I). Not all
regions were examined in each experiment.

1 8
Data Analysis

Group differences in mean plasma T concentration, mean testis weight, and the
density of AR-IR cells within each brain region were assessed by Student‘s t-tests (2
tailed).

Results

Mean paired testis weight and plasma T concentrations were signiﬁcantly greater
in 20 wk old ferrets compared to 12 wk old ferrets (Figure 4; 1(5) = 13.96, p < .01, and
1(5) = 3.21, p < .05, respectively). Of the four 12 wk old ferrets, only one had detectable
levels of plasma T (1.8 rng/ml) at the time of sacriﬁce, whereas all of the 20 wk old
ferrets had detectable levels of plasma T (range of 54-152 ng/ml). The mean density of
AR-IR cells in the regions analyzed is shown in Figure 5. The density of AR-IR cells
was signiﬁcantly greater in 20 wk old ferrets compared to prepubertal ferrets in the ARC,
1(5) = 4.12, p < .01, MPA, 1(5) = 3.29, p < .05, pvPOA, 1(5) = 5.11, p < .01, MeA,1(5) =
4.07, p < .01, RCh, 1(5) = 3.29, p < .05, and lVMH,1(5) = 2.98, p < .05. The mean
density of AR-IR cells in the LS or the BNST was not signiﬁcantly different between 12
and 20 wk old ferrets, botln p > 0.05.

Cellular immunoreactivity in brains of both prepubertal and 20 wk old ferrets was
restricted to the nucleus (Figure 6, C&D). Nucleoli were unlabeled. The intensity of
intracellular staining was heterogeneous within most of the cell groups analyzed, ranging
from a very intense dark reaction product in some cells to a light reaction product in
others. This was particularly true of the VMH, POA and MeA. Overall, immunostaining
was noticeably more intense in 20 wk old ferrets than in prepubertal animals, however

optical density was not quantiﬁed (cf Figure 6C & D).

 

Mean (:SEM) Paired Testis Weight (g)

 

 

 

Figure 4. Mean (:SEM) paired testis weight (g) and plasma testosterone
concentration (ng/ml) in 12 wk old and 20 wk old male ferrets. Asterisk
designates statistically significant differences between the two groups (p < .05).

Mean (:tSEM) Testosterone (ng/rnl)

20

 

I 12 week
I 20 week

 

 

 

Mean (:SEM) Number of AR-IR Cells / 15625 um2

LS aBNST MPA pvPOA RCh ARC MeA lVMH

 

Figure 5. Mean (iSEM) number of androgen ‘ Ivc (AR- IR)
cells I 15, 625 um2 1n speciﬁc brain regions of 12 wk old and 20 wk old male
ferrets. Asterisk designates statistically signiﬁcant differences between the two
groups (p < .05)

 

Figure 6. Low (A and B) and high (C and D) power photomicrographs of
androgen .y in the POA from representative 12 wk old (A
and C) and 20 wk old (B and D) male ferrets. Bar: 200 pm (A and B) or 20 um
(C and D)

22
Discussion

This experiment is the ﬁrst comparison of brain AR-immunoreactivity in males at
different stages of pubertal maturation, and demonstrates that sexual maturation is
characterized not only by gonadal growth and elevated circulating T concentrations, but
also by an increase in the density of AR-IR cells in speciﬁc brain regions. The increase
in the density of AR-IR cells in behaviorally relevant brain regions such as the MPA and
MeA is consistent with the hypothesis that pubertal increases in behavioral
responsiveness to the activational effects of T on reproductive behavior may be in part
mediated by an increase in the number of cells responsive to androgens.

With respect to the neuroendocrine changes associated with pubertal maturation,
this experiment found no evidence to support the hypothesis stated in the introduction
that the pubertal decrease in responsiveness to the negative feedback effects of T may be
due to a reduction in the density of AR-IR cells in brain regions implicated in the control
of negative feedback. There is some evidence indicating that negative feedback on
gonadotropin secretion can be achieved by direct implantation of gonadal steroids into
the region of the mediobasal hypothalamus (Cheung & Davidson, 1977; Kalra & Kalra,
1980). However, this includes several nuclei and including the VMH, RCh, ARC, any or
all of which may be involved in mediating negative feedback on gonadotropin secretion.
Furthermore, these data do not rule out the possibility that other regions may also be
involved in mediating negative feedback. Given that the precise anatomical location
and/or locations which mediate the negative feedback effects of T on gonadotropin is
unclear, it is possible that there is a brain region that shows a reduction in the density of

AR-IR cells that was not quantiﬁed in the present study.

23

The results of this initial experiment leave several other issues unresolved that
form the basis for the remainder of this dissertation. First, is the increase in the density
of AR-IR cells observed in Experiment I linked to the process of pubertal maturation, or
to some other unknown developmental event associated with age? Experiment 11
addresses this issue by using a photoperiod manipulation to alter the timing of the onset
of pubertal maturation. Second, what is the mechanism responsible for the increase in
the density of AR-[R cells during pubertal maturation? Experiment III addresses the
hypothesis that the pubertal rise in circulating T is responsible for the increased density
of AR-IR cells as animals undergo pubertal maturation. Experiment IVa addresses the
hypothesis that the differential behavioral response of prepubertal and adult male ferrets
to androgen administration is due to the differential regulation of androgen receptors by
androgens themselves. Finally, Experiment Nb examines the short-term effects (30 min
to 8 hrs) of T exposure on androgen receptor immunocytochemical staining in castrated
adult ferrets. This experiment was designed to provide evidence against the possibility
that changes in AR-IR staining following steroid treatment are solely due to translocation
of receptor proteins in the presence of ligand, as well as to begin to assess the time-
course of androgenic regulation of androgen receptors. The results of these experiments
will lay the groundwork for experiments designed to test speciﬁcally whether the
increase in AR-IR cell density is functionally linked to the changes in responsiveness to

T which occur during pubertal maturation.

24
Experiment II. Time-course for the appearance of androgen receptor-
immunoreactive cells in the brains of male ferrets undergoing pubertal maturation
under different photoperiod conditions.
Rationale

Increases in gonadal size and in LH secretion have been directly linked to the
process of pubertal maturation in male and female ferrets by demonstrations that the
timing of the increase in these parameters is advanced during photostimulated pubertal
maturation relative to that during spontaneous pubertal maturation (Ryan, 1985; Ryan &
Robinson, 1985; Sisk, 1990). The present experiment tested the hypothesis that the
increase in the density of AR-IR cells in speciﬁc regions of the male ferret brain
observed in Experiment I is also directly linked with the process of pubertal maturation.
If true, then advancing the age at onset of pubertal maturation by photostimulation
should similarly advance the age at which the increase in the number of AR-IR cells in
speciﬁc brain regions occurs, and the increase in AR-IR cell number should be

temporally correlated with other indices of pubertal maturation.

Methods
Tissue utilized in Experiment I]

Since Experiment I showed that there was an increase in AR-IR cell density after
the onset of spontaneous puberty, sets of tissue from all animals described in the methods
section of Experiment I were processed to determine if the increase in AR-IR cell density
was linked directly to the process of pubertal maturation, or to a developmental process
more closely related to the chronological age of the animal. Speciﬁc protocols are

outlined in Experiment I.

25

Data analysis

A one-way analysis of variance (ANOVA) on the number of AR-IR
cells/gid/region in 12 wk old prepubertal controls housed in short days, in 20 wk old
ferrets undergoing photostimulated puberty in long days, and in 20 wk old ferrets
undergoing spontaneous puberty in short days was performed in order to determine
whether our previous experimental results showing a regional increase in the density of
AR-IR cells was replicated across 2 separate immunocytochemical runs, and to determine
if, at 20 wks of age, there was a difference in AR-IR cell density between ferrets in long
days and ferrets in short days. Signiﬁcant differences were probed with the Fisher‘s
Protected Least Signiﬁcant Difference (PLSD) test, and were considered signiﬁcant at
p<0.05. Because the group of 12 wk old ferrets in short days served as a baseline
control group against which the other groups of ferrets of different ages and in different

photoperiods were compared, Dunnett's tests (one tailed, or = 0.05) were used to

determine at what age mean testis weight, circulating T levels, and density of AR-[R
cells within each region in ferrets undergoing photostimulated or spontaneous puberty
ﬁrst differed ﬁom these measures in 12 wk old prepubertal controls (Steel & Torrie,
1980). This analysis was used to determine if transfer of ferrets from short days to long
days at 12 wks of age advanced the onset of gonadal growth and circulating T levels
relative to ferrets remaining in short days, and if advancing the onset of gonadal growth
similarly advanced the increase in the density of AR-IR cells in speciﬁc brain regions.
Finally, to determine whether chronological age, testis weight, or concentrations of
circulating T, was the best predictor of the density of AR-IR cells, a multiple regession

analysis, forcing all three variables into the model, was performed on the combined data

26

from those regions in which an increase in the density of AR-IR cells during pubertal
maturation was observed. Each independent variable in the model (i.e., age, testis
weight, and circulating T) was assessed for statistical signiﬁcance using a t-test (Steel &

Torrie, 1980).

Results

Gonadal indices

Based on Dunnett‘s tests, the photoperiod transition from short to long days
induced precocious gonadal maturation as evidenced by signiﬁcant increases in testis
weight and plasma T concentrations at younger ages in ferrets transferred to long days
compared with ferrets that remained in short days (Figure 7). Relative to 12 wk
prepubertal ferrets, testis weight signiﬁcantly increased by 15 wks of age in ferrets
transferred to long days at 12 wk of age, but did not signiﬁcantly increase until 17.5 wks
of age in ferrets that remained in short days. Mean plasma T concentration in ferrets in
long days was signiﬁcantly greater than that of 12 wk old prepubertal ferrets by 17 .5 wks
of age (p<0.01), while plasma T levels in ferrets remaining in short days did not

signiﬁcantly increase above levels seen in 12 wk old controls until 20 wks of age.

Androgen receptor immunoreactivity
Cellular AR immunoreactivity in brains of all ferrets was restricted to the nucleus
with a lack of reaction product in the nucleoli. The intensity of intracellular staining was

heterogeneous within most of the cell groups analyzed, ranging from a very intense dark

4.5
4.0

3.5

3.0

2.5

2.0

1.5

1.0

Mean (:tSEM) Paired Testis Weight (g)

0.5

27

 

 

+ Short Days
-U—Long Days 1

 

 

 

 

rlrlrlrljlgljlalslr

.35

430

.15

~10

Mean (:SEM) Testosterone (ng/ml)

O

 

 

12 13 1415 1617 18 19 20

12 1314 15 16 1718 19 20

AGE (wks)

Figure 7. Mean (iSEM) paired testis weight (g) and plasma testosterone
concentration (ng/ml) in male ferrets undergoing spontaneous pubertal maturation
in short days or photoinduced pubertal maturation in long days. Asterisk
designates signiﬁcantly different than 12 wk old prepubertal controls (p<.05).

28

reaction product in some cells to a light reaction product in others. Overall,
immunostaining increased in intensity as ferrets began undergoing pubertal maturation,
however optical density was not quantiﬁed.

One-way ANOVA between 12 wk old prepubertal ferrets housed in short days,
and 20 wk old ferrets undergoing photostimulated puberty in long days or spontaneous
puberty in short days indicated that both groups of 20 wk old ferrets had a signiﬁcantly
greater density of AR-IR cells than 12 wk old prepubertal animals in the MPA
[Fu=13.22] MeA [Fu=21.58], lVMH [FL7=13.25], and ARC [FL-38.62] (Figure 8; all
p‘s<0.05). No developmental increase in AR-IR cell density was observed in the BNST
or LS (p‘s >0.05). Furthermore, 20 wk old ferrets undergoing photoinduced pubertal
maturation had a signiﬁcantly geater density of AR-IR cells in the MeA and MPA, than
did 20 wk old ferrets undergoing spontaneous pubertal maturation in short days
(p‘s<0.05). There were no differences between photoperiod conditions at 20 wks of age
in the density of AR-IR cells in the ARC or lVMH (p's>0.05).

For those regions in which an increase in the density of AR-IR cells occurred,
Dunnett‘s tests indicated that the photoperiod transition ﬁ'om short to long days resulted
in an increased density of AR-IR cells at a younger age (Figure 8). The mean density of
AR-IR cells in the MPA, MeA, lVMH and ARC was signiﬁcantly greater than that of 12
wk old controls by 17.5 wks of age in ferrets transferred to long days, but did not
signiﬁcantly increase in ferrets that remained in short days until 20 wks of age.

Multiple regression analysis indicated that a signiﬁcant proportion of the variance
in the dependent variable (AR-IR cell density) could be accounted for by the three

independent variables (age, testis weight, and T levels) in those brain regions in which

29

 

 

 

llllllllllllllllllllllllllllllllllllll

 

Mean (:SEM) Number AR-IR Cells / 15625 um2
8

 

 

 

12131415 16 171819 20 12 13 14151617 18 19 20

AGE (wks)

 

Figure 8. Mean (:tSEM) number of androgen receptor-immunoreactive (AR-IR)
cells / 15,625 um2 in speciﬁc brain regions of male ferrets undergoing
spontaneous pubertal maturation in Short days or photoinduced pubertal maturation
in long days. Symbols: (a) Signiﬁcantly different from 12 wk old prepubertal
controls (ANOVA, p<.05), (b) signiﬁcantly different from 20 wk old ferrets in
short days (ANOVA, p<.05), (c) signiﬁcantly different from 12 wk old
prepubertal controls (Dunnett's tests, p<.05).

30

there was a pubertal increase in the density of AR-IR cells [F3,m=37.60; p<0.0001,
Rz=.482]. T-tests performed on each independent variable indicated that only testis

weight was a signiﬁcant predictor of the density of AR-IR cells [F339, p<0.01].

Discussion

This experiment has conﬁrmed our previous demonstration of an increase in the
density of AR-IR cells in the POA, MeA, VMH and ARC of male ferrets undergoing
spontaneous pubertal maturation (Kashon & Sisk, 1994), and extends this observation to
male ferrets undergoing photoinduced pubertal maturation. We have further
demonstrated that a photoperiod transition from short days to long days at 12 wks of age
advances the age at which signiﬁcant gonadal growth and increases in circulating T
levels occur. This early onset of gonadal maturation stimulated by photoperiod is also
reﬂected in the brain by an advance in the timing of the increase in the density of AR-IR
cells in POA, MeA, VMH and ARC. In addition, testicular size, but not chronological
age, is a signiﬁcant predictor of AR-IR cell density in these regions. Increases in LH
pulse frequency and circulating levels of gonadal steroids are well-known endocrine
correlates of pubertal maturation, and in seasonal breeders, the timing of increases in
these measures is inﬂuenced by environmental photoperiod (Ebling & Foster, 1989).
This experiment provides evidence that an increase in the density of AR-IR cells in
speciﬁc brain regions is-aneuroanatomical correlate of pubertal maturation that is also
inﬂuenced by environmental daylength.

The mechanism responsible for the pubertal increase in the density of AR-IR cells

in POA, MeA, VMH, and ARC is unknown. There is ample evidence that androgens can

31

increase androgen receptor immunoreactivity, nuclear androgen receptor binding, and
androgen receptor half-life (discussed in detail in Experiment IVb) (Bittrnan & Krey,
1988; Krey & McGinnis, 1990; Prins & Birch, 1993; Sar, Lubahn, French, & Wilson,
1990; Wood & Newman, 1993 a). Thus, the increase in circulating levels of T which
occurs during pubertal maturatiOn, operating through one or more mechanisms could
mediate the pubertal increase in the density of AR-IR cells in certain brain regions. In
this experiment, multiple regression analysis indicated that T concentration was not a
signiﬁcant predictor of the density of AR-IR cells in any brain region. However, the
single sample of blood collected in this experiment may not be an accurate estimate of
circulating T due to the pulsatile nature of T secretion. This analysis did indicate that
testis weight, which provides a more stable correlate of ongoing gonadal steroid
secretion, is a signiﬁcant predictor of the density of AR-IR cells in those regions of the
brain in which a pubertal increase in the density of AR-IR cells occurred (i.e., POA,
MeA, ARC, and VMH).

The density of AR-IR cells did not increase as a function of pubertal maturation
in all brain regions examined. While clear increases in the density of AR-IR cells were
seen in the POA, MeA, ARC, and VMH, tlnere were no increases in the density of AR-IR
cells in the BNST or the LS associated with pubertal maturation in either photoperiod
condition. The lack of a pubertal increase in AR-IR cell density in the LS and BNST
indicates that pubertal regulation of androgen receptor is different across brain regions.
The factors that result in a geater density of AR-IR cells in the POA, VMH, ARC, and
MeA of more reproductively mature ferrets apparently do not operate in the same fashion

in LS or BNST. Other instances of brain region differences in the regulation of androgen

32

receptor irrnmunoreactivity have been previously documented (Menard & Harlan, 1993).
Regulation of AR is also different in separate lobes of the prostate (Prins & Birch, 1993),
and regulation of estrogen receptor-immunoreactivity is also brain region speciﬁc (Koch,
1990; Sisk & DonCarlos, 1995). Differences in afferent input and in intracellular
chemical milieu are likely to be involved in brain region-speciﬁc regulation of steroid
hormone receptors.

Pubertal maturation is associated with increased responsiveness to the activational
effects of T on steroid-dependent male reproductive behaviors (Sisk, et al., 1992).
Prepubertal males display little reproductive behavior in the presence of a receptive
female, and relatively high doses of T are required to induce reproductive behavior in
prepubertal males (Beach, 1942; Sisk, et al., 1992; Sodersten, Damassa, & Smith, 1977).
Furthermore, the lowest dose of T able to activate male reproductive behaviors decreases
as animals undergo pubertal maturation, indicating a pubertal change in sensitivity to the
behavioral actions of steroid hormones (Sisk, et al., 1992). Similarly, the photoperiod in
which male ferrets undergo pubertal maturation, and/or in which they are tested as adults,
modulates responsiveness to the behavioral actions of T in adulthood (Sisk, et al., 1992).
As adults, ferrets that undergo spontaneous pubertal maturation in short days require a
higher dose of T to activate male reproductive behavior patterns, and are thus less
responsive to the activational effects of T on male reproductive behaviors, than animals
that undergo photostimulated pubertal maturation following a photoperiod transition
from short days to long days (Sisk, et al., 1992). This is true despite the fact that
measures of gonadal function and LH secretion between spontaneous and photoinduced

animals do not differ in adulthood (Sisk, 1990). The cellular basis for pubertal and

33

photoperiod modulation of behavioral responsiveness to T is unknown, but this
experiment provides evidence consistent with the hypothesis that a pubertal increase in
androgen receptor expression in brain regions involved in the control of male sexual
behavior (cg. MPA, MeA) may mediate the increased responsiveness to the behavioral
effects of T. In addition, at 20 wk of age, AR-IR cell density in these brain regions is
greater in animals undergoing photostimulated puberty compared with animals
undergoing spontaneous puberty. It is unknown if this photoperiod difference in the
density of AR-IR cells in POA and MeA will persist once pubertal maturation is
complete (by approximately 28 wks (Sisk, 1990). If it does, then photoperiod
modulation of responsiveness to behavioral actions of T may also be mediated by
photoperiod-induced changes in AR-IR cell density in brain regions important for male

reproductive behavior.

Experiment III. Regulation of androgen receptor-immunoreactive cells in the
brains of prepubertal male ferrets by gonadal steroids.
Rationale

The temporal correlation between the pubertal rise in T secretion and the
increased density of AR-IR cells in certain brain regions suggests that T may be
responsible for the pubertal increase in AR-IR cell number. There is evidence from both
immunocytochemical studies and androgen receptor binding studies that both the cellular
location of androgen receptors (nuclear vs cytoplasmic) and the intracellular
concentration of androgen receptors in brain can be altered by androgens (Bittrnan &

Krey, 1988; Krey & McGinnis, 1990; Menard & Harlan, 1993; Wood & Newman,

34

1993a). Studies in other tissues or model systems also indicate, either directly or
indirectly, that androgens may stimulate AR synthesis and/or increase AR protein
stability (Blondeau, Corpéchot, Le Goascogne, Baulieu, & Robel, 1975; Grino, Grifﬁn,
& Wilson, 1990; Handa, Stadelman, & Resko, 1987b; Kemppainern, Lane, Sar, &
Wislon, 1992; Krongrad, Wilson, Wilson, Allman, & McPhaul, 1991; Prins & Birch,
1993; Robel, Eychenne, Blondeau, Jung-Testas, Groyer, Mercier-Bodard, et al., 1983;
Syms, Norris, Panko, & Smith, 1985; Verhoeven & Cailleau, 1988). Thus, the higher
circulating levels of T, present after the onset of puberty, and operating through one or
more mechanisms, could induce the pubertal increase in the number of AR-IR cells in
brain.

This experiment was designed to examine the regulation of brain AR-IR cells by
gonadal steroids in prepubertal male ferrets, and to test the hypothesis that the pubertal
increase in AR-IR cell density is the result of the pubertal increase in circulating
androgens. It was predicted that castration would reduce nuclear AR immunoreactivity,
and that androgen treatment would reverse this effect of castration. Furthermore, if the
previously observed pubertal increases in the number of AR—IR cells in POA, ARC,
MeA, and VMH are induced by the pubertal increase in circulating T concentrations,
then treatment with a dose of T that yields adult plasma levels of hormone in castrated
prepubertal ferrets should result in a signiﬁcant increase in the density of AR-IR cells in

these brain regions compared with those in intact prepubertal controls.

35

Methods

Animals and Experimental Design

Twenty ﬁve 7-wk-old weanling male ferrets were purchased from Marshall
Farms (North Rose, NY) and housed in pairs in stainless steel cages (51 x 60 x 38 cm) in
a temperature- (23° :h 1°C) and light- (8 hr light : 16 hr dark) controlled colony room.
Under these conditions, testicular growth begins at about 18 wk of age, and therefore
ferrets in this study were still juvenile at the time of sacriﬁce (mean body weight 3: SEM
= 816.3 :t 55.9 g). At 9 wk of age, 4 groups of 5 ferrets each were castrated under
methoxyﬂurane anesthesia (Metofane; Pittman-Moore, Washington Crossing, NJ), and
one group of 5 ferrets underwent sham castration and remained intact. Following a 1 wk
recovery period animals were injected (sc) once daily with either 5 mg/kg testosterone
propionate, 5 mg/kg dihydrotestosterone propionate, long/kg estradiol benzoate, or oil
vehicle (castrated and intact animals) for 10 days. The doses and injection regimen for T
and B were chosen based on their ability to activate reproductive behavior in adult ferrets
(Baum, Carroll, Cherry, & Tobet, 1990; Sisk, et al., 1992).

Animals were perfused 4 hr following the ﬁnal steroid injection, animals were
sacriﬁced, brains were prepared for AR immunocytochemistry , and plasma was prepared
for radioimmunoassay for T as described in Experiment 1. However, several
modiﬁcations of the AR immunocytochemistry procedure were included in this
experiment. Due to the report that visualization of brain AR immunostaining in castrated
hamsters required a lengthy incubation of the tissue in the chromagen solution (Wood &
Newman, 1993a), half of the sections from each animal were incubated in the DAB

solution for 6 min, while the other half were incubated for 60 min. To further enhance

36

androgen receptor immunocytochemical staining, the incubation in secondary antibody
was increased to 24 hours, and the concentration of NiCl2 in the DAB was doubled to 50
ul/ml of SOOmM NiClz. In addition to T levels measured in intact, castrate and T treated
animals, a radioimmunoassay for E was performed using the terminal blood samples
from intact, castrated, and E treated animals with reagents from the Coat-A-Count
Estradiol Kit. All samples were run in duplicate and the minimum detectable levels of
steroid were 0.10 ng/ml for T and 10.0 pg/ml for B.
Data Analysis

For each brain region, the mean density of AR-IR cells was analyzed using a two-
way analysis of variance (steroid treatment x incubation time). Signiﬁcant differences
were probed with the Fisher’s PLSD test (Steel & Torrie, 1980). All differences were

considered signiﬁcant at ps0.05.

Results

Distribution and patterns of AR immunocytochemical staining

The general distribution of AR-IR cells throughout the forebrain and
hypothalamus of the male ferret was similar to that seen in Experiments I and II, and
previously reported in males of other species (Choate & Resko, 1992; Clancy, Bonsall, &
Michael, 1992; Menard & Harlan, 1993; Sar, et al., 1990; Wood & Newman, 1993a).
However, due to the modiﬁcations in the immunocytochemical staining procedure, AR-
IR cells were much darker in this experiment than previously seen. There were
noticeable differences in intracellular immunocytochemical staining intensity across the

regions in which AR-IR cells were quantiﬁed in this experiment. In intact animals, AR-

37

IR cells were darkest in the LS and pvPOA, appeared to be slightly less dark in the MPA,
VMH, and ARC, and were lightest in the MeA and BNST. Photomicrographs of AR-IR
cells in the pvPOA and the lVMH from representative animals within each treatment
group are shown in Figure 9. In intact and androgen-treated animals immunostaining
was restricted to the nucleus (Fig. 9A, 9B, 9C, 9F, 9G, 9h). Subjectively, androgen
treatment resulted in darker nuclear immunostaining within any given brain region when
compared with that of any of the other treatment groups. However, even in the presence
of androgen, AR-IR cells in the MeA and BNST remained more lightly stained compared
with AR-IR cells in other regions. Castration reduced the intensity of nuclear
immunocytochemical staining in all brain regions, and resulted in the appearance of
visible immunocytochemical reaction product distributed throughout the cytoplasm of
some, but not all, cells (Figure 9E and 9], arrows). The qualitative appearance of AR-
immunostaining in estrogen-treated castrates resembled that of oil-treated castrates,
except that cytoplasmic staining, while observed in estrogen-treated ferrets, did not seem

as prevalent as it was in oil-treated castrates (Fig. 9D and 91).

Effect of incubation time in DAB on AR-immunoreactivity

Incubation of tissue in DAB for 60 min resulted in darker immunocytochemical
staining intensity when compared with that in tissue incubated in DAB for only 6
minutes. However, ANOVA revealed that, of the eight brain regions examined, only in
ARC [F1,3,=6.22, p=0.017] and in MeA [F 1,39=12.731, p=0.001] was there a signiﬁcant
main effect of incubation time in DAB on the density of AR-IR cells. In these two

regions, tissue which was incubated for 60 min in chromagen had more AR-IR cells than

38

 

Figure 9. Photonnicrographs of AR-IR cells taken from the periventricular preoptic
area (pvPOA; A-E), and the lateral portion of the ventromedial hypothalamus
(lVMH; F-J) from oil-injected intact animals (A,F), and those castrated and
injected once daily for 10 days with either 5 mg/kg testosterone propionate (B,G),
5 mg/kg dihydrotestosterone propionate (C,H), louglkg estradiol benzoate (D,I),
or oil vehicle (E,J). The pvPOA is an example of a region in which androgen
treatment (T or DHT) resulted in a signiﬁcantly greater number of AR-IR cells
compared with intact males (cf A with B and C). The lVMH is an example of a
region where androgen treatment did not increase the number of AR-IR cells
above the number seen in intact animals (cf F with G and H). Arrows in E and J
point to AR-immunoreactivity in the cytoplasm, which was seen primarily only in
tissue from castrated animals. All micrographs are of tissue incubated in DAB for
60 minutes. Magniﬁcation Bar: 10pm.

39

did tissue incubated for only 6 min. The mean density of AR-IR cells in the short and
long incubation times, respectively, and collapsed across steroid treatment, were 27.5 i
2.85 and 34.2 :1: 3.2 for the ARC, and 17.0 i 3.0 and 26.2 i 3.1 for the MeA. However,
no signiﬁcant interactions existed between steroid treatment and incubation time on AR-
IR cell density in any brain region examined, indicating that the long incubation in the
chromagen did not preferentially increase the density of AR-IR cells in a given steroid
treatment group. Therefore, comparisons between steroid treatments in each region were

analyzed using data collapsed across incubation time (Keppel, 1982).

Effect of steroid treatment on the density of AR-IR cells

Two-way ANOVA revealed a signiﬁcant main effect of steroid treatment on the
density of AR-IR cells/grid in all brain regions (all p‘s < 0.01). Figure 10 depicts the
mean density of AR-IR cells in intact and castrated ferrets treated with oil or steroids in
each brain region collapsed across incubation time (Keppel, 1982). Several effects of
steroid treatment were consistent across brain regions. First, castration resulted in a
signiﬁcant decrease in the density of AR-IR cells in all regions examined. Second, T or
DHT administration to castrates resulted in a signiﬁcantly greater density of AR-IR cells
compared to treatment with oil or E in all brain regions. Finally, for all brain regions,
there were no signiﬁcant differences in density of AR-IR cells between T- and DHT-
treated castrates, or between oil- and E—treated castrates.

Irnportantly, there were brain region differences with respect to the magnitude of
the increase in AR-IR cell density following administration of T or DHT. The statistical

analysis indicated that speciﬁc brain region responses to androgen treatment of castrates

8O
70

S

30
20
10

7O

8

30

Mean (:SEM) Number AR-IR Cells / 15626 um2

10

4O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

r- A“
r , -.
i- ‘ d
. I I .
' 'l
I: * . I .‘
L i t- I
h ** -I
f *r 2
H
i- I!
1CETD iCETD rcsrn ircsrn
ARC MeA pvPOA MPA
L B.‘
. J
’ 'l
_ 1' .
. II. I .
L r .-
_ II I II , -
L * I
t
: 1 1
tests Irsrn rcrrn rcrrn
BNST LS lVMH mVMH

Figure 10. Mean (:tSEM) number of androgen receptor-immunoreactive (AR-IR)
cells/15,625 um2 in speciﬁc brain regions of intact and castrated prepubertal male
ferrets injected (sc) once daily with either 5 mg/kg testosterone propionate (T), 5
mg/kg dihydrotestosterone propionate (D), 10 ug/kg estradiol benzoate (E), or oil
vehicle (castrate (C) and intact (1) animals) for 10 days. Animals were sacriﬁced 4
hr following the last injection. Regions in which androgen treatment Signiﬁcantly
increased the number of AR-IR cells compared with intact animals are shown in A.
Regions in which androgen treatment restored the number of AR-IR cells to that
seen in intact animals are shown in B. Asterisk (*) designates significant
difference from intact, T and DHT treatment. Double asterisk designates
Signiﬁcant difference from Tand DHT treated animals.

41

could be classiﬁed as one of two types: either a restoration of the density of AR-IR cells
similar to that seen in intact animals, or a signiﬁcant increase in the density of AR-IR
cells compared with that of intact animals. These two types of response to androgen
treatment are illustrated by the photomicrographs of AR-[R cells in pvPOA and lVMH in
Figure 9, and in Figure 10, the 8 regions are separated in panels A and B on the basis of
which category of response to androgen was observed in that region. In the BNST, LS,
mVMH, and lVMH, androgen administration to castrates restored the density of AR-IR
cells to that observed in these regions in intact animals (cf 9F with 9G and 9H; Figure
10B). In contrast, in the MPA, pvPOA, MeA, and ARC, the same androgen treatment
not only reversed the effects of castration, but resulted in a signiﬁcant increase in AR-IR
cell density compared with that seen in these regions in intact animals (cf 9A with 9B
and 9C; Figure 10A). Figure 11 depicts these data as the percent of increase or decrease
in the density of AR-IR cells in steroid treatment goups relative to intact prepubertal
males, and arranges brain regions in order of increasing magnitude of response to
androgen treatment. Viewed in this way, brain region speciﬁc effects of androgen on AR
immunoreactivity may be characterized as a continuum, in contrast to the dichotomy
presented in Figure 10. Regardless of how the brain region responses to androgen are
categorized, it is clear that there are brain region differences in response to androgen
under the present experimental conditions.
The factors that contribute to regional differences in androgen receptor

immunocytochemistry are unknown. It is conceivable that regional differences in AR
expression are due to regional differences in cell and/or nuclear size, which result from

steroid treatment, and thus a bigger cell may be more likely to be detected by

42

 

an
O

_ D Castrate
Estradiol
i- Testosterone
r I Dihydrotestosterone

   
   

  
 

[04>
CO

 
 

&\\\\\\\\\\\\\\\\V
\\\\\\\\\\\\\\\\\‘C

     

Intact —

    

   

Percent Change

Ill/Illlflflle

 

 

 

 

 

'I/I/ll/I/I/I/I/I/I/I/I/IA

 

 

VII/IllIll/IlllllllllIll/Illlllllla
f/Ill/lll/I/I/I/I/I/I/I/I/I/I.

 

 

 

BNST LS lVMH mVMH ARC MeA pvPOA MPA

Figure 11. Percent change relative to intact prepubertal male ferrets in the number
of AR-IR cells/15,625 um2 in speciﬁc brain regions of castrated prepubertal male
ferrets treated with either 5.0 mg/kg testosterone propionate, 5.0 mg/kg
dihydrotestosterone propionate, 10 ug/kg estradiol benzoate, or oil vehicle (sc)
once daily for 10 days.

43

immunocytochemistry. Androgen treatment increases dendritic sprouting (Cherry,
Tobet, DeVoogd, & Baum, 1992) and increase cellular size (Tobet, Zahniser, & Baum,
1986b) in some regions of the ferret brain such as the male nucleus of the POA-AH.
Thus increases in cellular size within a given region may lead to what appears to be an
increase in cell number within that region. However, whether the regional differences in
AR found in the present experiment are positively correlated with an increase in cell
and/or nuclear size following androgen treatment remains to be determined.

Since there was an effect of incubation time on the density of AR-[R cell in the
ARC, and since the effect of androgens on the density of AR-IR cells in this region fell
approximately in the middle of the continuum depicted in Figure 11, a one way ANOVA
was performed on cell counts for the short and long incubation times separately for this
region, even though two way AN OVA did not reveal a signiﬁcant interaction between
steroid treatment and incubation time. Androgen treated groups were not signiﬁcantly
different from intact groups when short incubation time data were analyzed alone, and
when long incubation time data were analyzed alone, the number of cells/gid in the
ARC of intact animals was signiﬁcantly different from that of T treated ferrets, but not
from that of DHT treated ferrets. Therefore, under the hormone treatment paradign used
in this study, the effect of androgen on the number of AR-IR cells/gid inARC is

somewhat ambiguous.

Steroid Hormone Levels
Plasma T concentrations (mean ng/ml :t SEM) were 0.644 i 0.315, 0.116 :1:

0.014, and 23.328 :t 2.393 for intact, castrated, and T-treated ferrets, respectively. The

44

levels of T in T-treated ferrets are high, but within the normal physiologic range seen in
adult male ferrets (Sisk & Desjardins, 1986). Estradiol levels in all castrated ferrets, and
in 3 of 5 intact ferrets, were below the assay limit of detectability. Mean (:tSEM) plasma

E concentration was 43.1 :t 11.5 pg/ml in E-treated castrates.

Discussion

This experiment demonstrated that castration of prepubertal male ferrets results in
a decrease in immunocytochemical staining intensity and in the density of AR-IR cells in
speciﬁc brain regions, and that AR immunostaining can be restored by treatment of
castrates with androgens, but not estrogen. The dramatic decrease in AR
immunostaining following castration indicates that brain AR immunoreactivity is
inﬂuenced even by the very low levels of circulating T characteristic of prepubertal
males. The castration-induced decrease and androgen-induced restoration in AR
immunoreactivity in prepubertal male ferrets are in general agreement with two other
reports of the effects of castration and steroid replacement on brain AR immunoreactivity
(Menard & Harlan, 1993; Wood & Newman, 1993a). In a detailed study of intracellular
partitioning of AR immunoreactivity in adult male hamsters under different steroid
conditions, Wood and Newman (Wood & Newman, 1993a) found that nuclear staining
intensity decreased, while cytoplasmic staining increased, after castration. Furthermore,
androgen, but not estrogen, replacement reversed these effects, and the low levels of
circulating T in males in the nonbreeding condition were sufﬁcient to restrict

immunoreactivity to the nucleus. Using a quantitative analysis similar to that used in this

45

experiment, Menard and Harlan (Menard & Harlan, 1993) reported that a cocktail of
anabolic-androgenic steroids, either administered to castrated rats, or superimposed over
endogenous steroids in intact males, increased the number of AR-[R cells/unit area in
several brain regions. Thus, the ability of androgens to enhance brain AR-
irnmunoreactivity appears to be consistent across a variety of androgen treatment
paradigns and species (but see (Choate & Resko, 1992; Clancy, Whitman, Michael, &
Albers, 1994) for possible exceptions).

When a quantitative analysis is used to assess the effect of androgen on AR
immunoreactivity, as it was in this experiment and in the Menard and Harlan (1993)
study, it becomes apparent that the magnitude of the androgen-induced increase in AR-IR
cells/unit area varies across brain regions. In the present study, 10 days of androgen
treatment, designed to approximate adult circulating concentrations of steroid in castrated
prepubertal ferrets, restored the density of AR-IR cells in the BNST, LS, mVMH, and
lVMH to that seen in intact prepubertal ferrets. In contrast, the same androgen treatment
resulted in a signiﬁcantly greater density of AR-IR cells in the pvPOA, MPA, and MeA
(and possibly in ARC) compared with corresponding densities in intact animals.
Differences in the magnitude of brain region responses to androgen have also been
documented in male rats. Administration of an androgenic-anabolic steroid cocktail to
intact adult male rats increased the number of AR-IR cells/area in VMH above that seen
in untreated intact animals,.but not in the MPA or MeA (Menard & Harlan, 1993). The
data from the present study and those from the Menard and Harlan study (Menard &
Harlan, 1993) are not directly comparable because of differences in experimental

protocols (steroid treatment of castrated vs intact animals), the particular steroids and

46

doses administered, species, and reproductive status of the animals (prepubertal vs adult
males). The conclusion from both sets of data, however, is that the regulation of AR
immunoreactivity by androgens is different across brain regions. Clearly, under
experimental conditions, the speciﬁc response evoked by androgen in cells in a given
brain region will depend upon many factors, including afferent input and neurochemical
phenotype of the cells at the time of androgen action. The challenge is to understand
what accounts for cell Speciﬁcity in the regulation of AR by androgen under normal
physiological and developmental conditions.

Even under physiological conditions, regulation of AR varies across brain
regions. The ﬁrst two experiments of this dissertation showed that during normal
pubertal maturation, the number of AR-IR cells/area increases in the POA, MeA, ARC,
and VMH, but not in the BNST and LS. The fact that the regional increases in the
density of AR-IR cells were correlated with the increased levels of T that accompany
pubertal maturation prompted the present attempt to test the hypothesis that, under
experimental conditions, adult levels of T could increase AR-IR cell density in these
regions in prepubertal males. Under the replacement paradign used in the present study,
androgens increased the density of AR-IR cells above those seen in intact prepubertal
males in POA, MeA, and possibly in ARC, but not in the VMH. Thus, this experiment
suggests that the pubertal increase in AR-IR cell density in at least POA and MeA is
stimulated by .the pubertal increase in T. It is possible that a longer exposure to
androgens, a different pattern of hormone administration, or exposure to androgens at an
older age, would have simulated the pubertal increase in AR-IR cell density in VMH and

ARC as well. Because AR-IR cell density in BNST and LS have not been shown to

47

increase as a function of puberty, the lack of a signiﬁcant androgen-induced increase in
the density of AR-IR cells in these regions relative to prepubertal controls was expected.

We found no effect of E treatment on the density of AR-IR cells in any brain
region. An earlier study reported that E treatment to castrated male rats resulted in a
small increase in the amount of radiolabeled androgen binding in the cytosolic fraction of
micropunches from the MPA and the MeA, but androgen binding in the VMH, BNST,
LS, pvPOA, and ARC was not affected by E (Handa, Roselli, Horton, & Resko, 1987a).
It is possible that small estradiol-induced increases in cytosolic receptor levels in the
MPA and MeA were below the immunocytochemical limits of detection in the present
study.

In summary, this experiment has documented that androgen inﬂuences brain AR
immunoreactivity in prepubertal males, but that the magnitude of the response to
androgen varies across brain region. Regional differences in the regulation of brain AR
by androgen allow for the possibility that an increase in circulating T at the time of
puberty could sensitize cells in some brain regions to androgen action, while not

affecting cells in other regions.

Experiment IV. Regulation of AR-IR cell density by gonadal steroids in pre- and
postpubertal male ferrets.
Rationale

Experiment IVa: Experiments I and H demonstrated that after the pubertal
increase in T output from the testes, there are more AR-IR cells per unit area in several

brain regions compared to these regions in prepubertal males, and that advancing the

48

onset of increased T production by using a photoperiod manipulation, also advances the
time of increase in the density of AR-IR cells in these brain regions. Experiment III
demonstrated that treating castrated prepubertal male ferrets with doses of androgens that
result in high level of circulating steroid increases the density of AR-IR cells relative to
prepubertal intact males in the POA, MeA and perhaps ARC, indicating that the pubertal
rise in androgens may be responsible for the pubertal rise in AR-IR cell density in these
regions. Previous experiments showed that a paradign of androgen treatment to
castrated ferrets similar to that used in Experiment 111 results in the activation of
reproductive behaviors in adults as well as prepubertal animals, with a trend for
prepubertal animals to show less behavior than adults, however differences between the
groups were not statistically signiﬁcant (Sisk, et al., 1992). Up to this point,
Experiments I, II, and III have been performed on ferrets either in the prepubertal stage
of development or in an advanced peripubertal stage, but not in fully mature adults.
Indeed, the behavioral data indicating that there is a change in sensitivity to the
behavioral actions of T are based on comparisons between prepubertal male ferrets and
fully developed adult ferrets. This experiment directly compares the distribution,
density, and intensity of androgen receptor-immunocytochemical staining between
prepubertal and fully mature adult ferrets, both in gonadally intact animals, and in
castrated animals treated with steroids. Differences in androgen receptor-
immunocytochemical staining patterns between these two age groups may lead to
irnsights as to why administration of androgens to prepubertal ferrets does not activate
exactly the same pattern of reproductive behaviors as seen in adults. For example, it is

possible that differences in behavioral capacities between prepubertal and adult male

49

ferrets following identical steroid treatment are because of differences in the steroid
hormone induced regulation of steroid receptors themselves. Based on this hypothesis,
one could predict that while steroid treatment to castrated prepubertal ferrets may
increase the density of AR-IR cells above prepubertal intact levels, intact or androgen
treated adults may still have a greater density than androgen treated prepubertal ferrets in
certain regions of the brain involved in the control of male reproductive behaviors. Were
this to be the case, it is possible that some other process occurs during pubertal
maturation that modulates the way in which androgens regulate the density of androgen
receptor-immunoreactive cells. On the other hand, if there are no differences between
prepubertal and adult male ferrets in the ability of T to up—regulate AR-IR cells in the
brain, this would indicate that other processes are at work that inhibit prepubertal male
ferrets from displaying the full range of male reproductive behaviors, despite sufﬁcient
amounts of AR in behaviorally relevant brain regions.

Experiment IVb: The cellular mechanism responsible for the increase in the
density of AR-IR cells observed in Experiments 1, H, and 11] remains to be determined.
However, higher concentrations of plasma T, whether endogenous or exogenous,
probably contributed to the increased density of AR-IR cells in certain brain regions as
suggested in Experiment III. Previous immunocytochemical studies of AR provide
evidence that circulating T levels may inﬂuence AR-immunostaining by several (not
mutually exclusive) mechanisms. First, the presence of ligand may increase AR-
irnmunocytochemical staining intensity by a mechanism which does not require AR gene
expression or protein synthesis. Castration leads to greatly diminished AR-

immunoreactivity in rat prostate tissue, and staining intensity is increased within 15 min

50

of androgen replacement (Prins & Birch, 1993; Sar, et al., 1990). This effect of
androgen on AR-irnmunoreactivity is likely too quick to be due solely to an increase in
receptor protein synthesis. In brain, castration was found to reduce AR-
irnmunoreactivity in some studies (Menard & Harlan, 1993; Wood & Newman, 1993a),
but not others (Choate & Resko, 1992; Clancy, et al., 1994), however, in none of these
experiments was a rapid effect of androgen on AR-immunoreactivity assessed. In
principle, a rapid androgen-induced increase in AR-immunostaining could be
accomplished in several ways. The presence of ligand could result in a conformational
change in the receptor protein that allows an increase in the number of epitopes available
for the antibody to bind. An increase in the number of epitope sites could also occur by
the dissociation of heat-shock proteins in the presence of ligand. Alternatively, the
presence of androgens could result in a translocation or stabilization of AR witlnin the
nucleus, effectively concentrating the receptor protein in the nucleus and increasing the
intensity of nuclear staining. Indeed, some in vitro experiments have demonstrated an
increase in androgen receptor half-life in the presence of ligand (Kemppainen, et al.,
1992; Robel, et al., 1983). The increase in receptor half-life would not only result in an
increase in immunocytochemical staining without requiring protein synthesis, but could
have important physiological implications by allowing prolonged androgen action on the
genome.

Another mechanism. by which androgen inﬂuences AR-immunostaining involves
an up-regulation of the receptor protein. Following 3 days of androgen replacement to
castrated rats, immunocytochemical staining intensity in prostate tissue is increased over

and above that seen after 15 min of replacement and approximates that observed in

51

gonadally intact rats, indicating that prolonged exposure to androgen results in increased
expression of androgen receptor protein (Prins & Birch, 1993). Similarly, brain AR-
irnmunostaining in hamsters and rats is reduced by castration, while chronic androgen
administration for 8 hrs or more restores it (Menard & Harlan, 1993; Wood & Newman,
1993 a). That the increase in AR-immunoreactivity induced by prolonged exposure to
androgen is due at least in part to increased synthesis of receptor protein is corroborated
by biochemical studies in tissue homogenates, which indicate that the increase in nuclear
AR binding in hypothalamus and pituitary observed after 6 or more hr of androgen
treatment is the result of synthesis of new receptor proteins (Bittrnan & Krey, 1988;
Handa, et al., 1987b; Krey & McGinnis, 1990). In one of these studies (Krey &
McGinnis, 1990), there was an initial rapid increase in nuclear androgen-plus-receptor
complexes following administration of T that was likely due to the translocation of
receptor protein into the nucleus. However, androgen-plus-receptor complexes continued
to rise above this initial increase until they reached a plateau 16 hr following T
administration, indicating synthesis or enhanced stability of receptor proteins (Krey &
McGinnis, 1990). Other studies indicate that androgen may upregulate androgen
receptor in other tissues and in cell cultures (Blondeau, et al., 1975; Grino, et al., 1990;
Handa, et al., 1987b; Kemppainen, et al., 1992; Krongrad, et al., 1991; Robel, et al.,
1983; Syms, et al., 1985; Verhoeven & Cailleau, 1988).

Although we cannot exclude the possibility that the increase in the density of AR-
IR cells in certain brain regions of ferrets observed in the above experiments is due in
part to T-dependent enhanced immunoreactivity without an actual increase in receptor

protein, it seems doubtful that this could be the sole explanation of these results. First, if

52

it were, one would expect to observe a higher density of AR-IR cells in maturing ferrets
in all brain regions examined, but this was not the case in the lateral septum or BNST.
Second, prepubertal ferrets are not strictly comparable to castrated animals in terms of
circulating T concentrations. Based on previous determinations of LH pulse frequency in
prepubertal ferrets (Sisk, 1987), the demonstration that the testes of the prepubertal ferret
are capable of responding to LH by secreting T (Neal, et al., 1977), and the fact that there
are detectable levels of T in a small percentage of terminal blood samples obtained from
prepubertal ferrets, it is likely that prepubertal ferrets secrete a few pulses of T per day.
Given that it takes longer than 24 hr for nuclear AR-immunoreactivity in prostate tissue
to diminish after castration, it seems improbable that the relative lack of AR-[R cells in
some brain regions of intact prepubertal ferrets is comparable to the reduced AR-
immunoreactivity observed by us and others after castration. Thus, we conclude that
enhanced immunoreactivity in the presence of ligand is unlikely to be the sole
mechanism responsible for the pubertal increase in the density of AR-IR cells, and that
an increase in receptor protein expression or stability also contributes to the greater
density of AR-IR cells in older animals. This experiment will examine AR
immunoreactivity in brain regions of castrated adult ferrets after short term androgen
replacement to determine the relative contributions of receptor translocation and/or
receptor synthesis to the androgen induced pubertal increase in the density of androgen

receptor-containing cells in speciﬁc brain regions.

53
Methods

Adult animals for Experiments [Va and Nb were raised as cohorts, thus allowing
us to use castrated adult ferrets treated with oil as a control group for both the
examination of long-term (10 days) exposure to gonadal steroids, as well as for the short-
terrn (hours) exposure to T experiments. Thirty-seven male ferrets were obtained from
the MSU Mink Farm at approximately 14 wks of age and placed in a short—day lighting
environment (8 hr light/day). Short days were chosen so that photoperiod condition
would be identical between prepubertal and adult animals at the time of sacriﬁce.
Animals were weighed and testis measurements were taken at biweekly intervals. At
approximately 33 wks of age all animals were castrated under Metofane anesthesia with
the exception of one group, which was sham castrated and remained intact (n=5).
Following a one-week recovery period the animals received the following steroid
treatments: Daily T (5 mg/kg) for 10 days (n=5), DHT (5.0 mg/kg) for 10 days (n=5), E
(10 ug/kg) for 10 days (n=5), oil vehicle for 10 days (intact and castrate animals n=5 for
each group). Animals were deeply anesthetized, a blood sample was taken via heart
puncture, and then perfused transcardially approximately 4 h following the ﬁnal steroid
injection, according to the method outlined in Experiment 1. Twelve of the castrated
animals were injected with oil vehicle each day for 9 days. On day 10, each animal was
given an injection of T (5 mg/kg) and perfused either 30 min, 4 hr, or 8 hr after the
injection (n=4 per group). Every sixth section from the above animals was processed for
AR immunocytochemistry along with a set of tissue from Experiment III consisting of
prepubertal males that remained gonadally intact, or that were castrated at 8 wks of age

exposed to a steroid injection paradign beginning at 10 wks of age identical to that

54

described for the adults in this experiment. Incorporating the prepubertal animals into
this experiment enables us to make direct comparisons between pre- and postpubertal
animals on the number and staining intensity of AR—IR cells using tissue
irnmunocytochemically processed at the same time (see Appendix B for statistical
analysis indicating that no signiﬁcant loss of AR-lR staining occurred due to long term
storage of tissue in cryoprotectant). Thus, the statistical analysis for Experiment IVa
includes prepubertal and adult animals that either remained intact, or were castrated and
treated for 10 days with either oil, T, DHT, or E. The statistical analysis for Experiment
IVb includes only adults that were castrated and injected with oil, or with T for 30min,
4hr, 8hr or 10 days. Thus, two groups of animals will be included in the analysis for
each of the experiments, adults castrated treated with oil or with T for 10 days.
Steroid Radioimmunoassays

Plasma concentrations of testosterone in the terminal blood samples from intact,
castrated, and T treated animals were measured with reagents in the Coat-a-Count Total
Testosterone Kit (Diagnostic Products, Los Angeles, CA). All samples were run in
duplicate and the minimum detectable levels of T were 0.10 ng/ml.
Data Analysis
AR-IR Cell Density

Cell counts were recorded at a magniﬁcation of 1000x and consisted of counting
all cells which fell within the boundaries of a 100 by 100um ocular gid in the following
regions: MPA, pvPOA, aBNST, meNST, MeA, MeP, LS, RCh, rARC, cARC, lVMH,
and PMV. The meNST, cARC and MeP are additional brain regions that were

included in this analysis in order to extend our survey of the steroid regulation of AR-IR

55

cells to other regions which are known to contain AR. The meNST and the MeP are
areas which are active during male reproductive behavior based on staining for the proto-
oncogene cFos (Baum & Everitt, 1992; Wood & Newman, 1993b). The PMV was
included because it contains a high level of AR containing cells and shows heavy
cytoplasmic staining in castrated animals (Wood & Newman, 1993b). For Experiment
IVa, an overall three-way ANOVA (Steroid treatment by Age by Brain Region) was
performed on the mean number of AR-IR cells/gid from two matched sections counted
bilaterally for each brain region quantiﬁed. For Experiment IVb, a two-way ANOVA
(Duration of T exposure by Brain Region) was performed to compare the following
groups: castrate plus oil, and castrate plus T for 30 min, 4 hr, 8 hr, and 10 d. Signiﬁcant
effects were probed with Fisher's Protected Least Signiﬁcant Difference (PLSD) test.
All differences were considered signiﬁcant if p < 0.05.
AR-IR Staining Intensity

Intact animals and those pre- and postpubertal animals that received T treatment
were rated for intensity of immunocytochemical staining in the pvPOA, MPA, MeA,
RCh, rARC, LS, aBNST and lVMH using a rating scale modiﬁed from DonCarlos et al.,
1991. The same sections that were quantiﬁed for cell number as described above were
rated for staining intensity by two investigators blind to the treatment of the animals.
The category assigned to a given animal corresponded to the predominant cell type
within the ﬁeld of view according to the following scale: 0 = no cells stained; l= very
light: staining at threshold of detection, nucleolus not detectable within the nucleus; 2 =
light: staining witlnin the nucleus diffuse, transparent, 3 = moderate: nucleolus

detectable within the stained nucleus, staining still somewhat transparent; 4 = dark:

56

staining dense, not transparent but can still detect individual ganules of reaction product;
5 = very dark: intense stain, opaque. These estimates were made at a magniﬁcation of
1000x. All sections that contained cells stained either moderately, dark, or very dark had
a high proportion of cells which, individually scored, would be considered light or very
light. Thus ratings in the three darkest categories were based on the predominant cell
type within the region, excluding those that were light and very light. Intensity ratings
from both parts of Experiment IV were analyzed with nonparametric statistical tests. A
correlation coefﬁcient of 0.876 was calculated on the ratings of staining intensity by the
two investigators. The data used in the analysis were an average of the two investigators
ratings. In Experiment IV a, pair-wise comparisons between pre- and postpubertal intact
animals, as well as between intact and T treated animals within the same age group, were
analyzed with the Mann-Whitney U test. In Experiment IVb, the Kruskall-Wallis test
was used to determine if signiﬁcant differences in intensity ratings existed between
treatment groups. Pair-wise comparisons were then made where appropriate with Mann-

Whitrney U tests. All differences were considered signiﬁcant at p < 0.05.

Results
Circulating Testosterone
One-way ANOVA showed that there were signiﬁcant differences between
treatment groups in the concentration of circulating T [F,o,42=52.9; p<0.0001] (Figure
12). Post-hoe analysis with the Fisher’s PLSD test indicated that animals treated with T
for 30 min were not different than intact adults. Both of these groups, however, had

signiﬁcantly lower T concentrations than did animals treated with T for either 4 hr, 8 hr

57

 

3O

 

 

20

15

F-i
IL...

10

O

 

 

Mean (iSEM) Testosterone (ng/ml)

 

 

 

 

 

 

 

 

 

 

 

 

 

Pre- Intact *]‘ 0

Adult T- 30 m ‘
Adult T-4 h *
Adult T-8 h ‘

Adult T-10 d ‘
Pre- T 10 d '
Adult Intact ‘

Adult Castrate ‘

Pre- Castrate ‘ °

Figure 12. Mean (tSEM) testosterone concentration (ng/ml) in prepubertal and
adult ferrets injected with oil or 5.0 mg/kg testosterone propionate for either 30
rrnin, 4 hr, 8 hr, or daily for 10 days. Bars with different letters are Signiﬁcantly
different from one another (p < 0.05).

58

or 10 days (prepubertal and adult), but both had signiﬁcantly higher T concentrations
than prepubertal intact animals, or castrated animals (prepubertal and adults) treated with
oil. Animals treated with T for either 4 hr, 8 hr or 10 days (prepubertal and adult) were
not signiﬁcantly different from one another, nor were there any differences among
prepubertal intact animals and animals castrated and treated with oil. All comparisons
between castrated animals treated with T and those treated with oil were signiﬁcantly
different. Adults treated with DHT or E had undetectable levels of T (data not shown);
no samples from DHT or E treated prepubertal males were processed for T levels.
Experiment IVa: Steroid Regulation of AR-IR cells in Pre and Postpubertal Male
Ferrets
AR-IR cell density

The three-way AN OVA table for the number of AR-IR cells per unit area is

shown in Table 1. There were signiﬁcant main effects of age, steroid treatment and brain
region, as well as signiﬁcant interactions between age and steroid treatment, age and
brain region, and between steroid treatment and brain region. The three-way interaction
among age, steroid treatment and brain region was not signiﬁcant. Since main effects
cannot fully describe the data when interactions are present, the results of this experiment
will be presented and, discussed in terms of the interactions, and therefore data will be
collapsed across one factor for each 2-way interaction. However, so that all of the data
from each region may be viewed, the mean number of AR-IR cells per unit area within
each region for each age for each steroid treatment are illustrated in Figure 13.
Photomicrogaphs from the pvPOA, and MeA for each treatment condition are shown in

Figures 14 and 15.

Table 1. Summary table for the three-way analysis of variance

from Experiment IVa.

Age

Steroid

Age * Steroid

Region

Age * Region

Steroid * Region

Age * Steroid * Region
Residual

59

DF Sum of Squares Mean Square F—Value P-Value

 

 

 

 

 

 

 

 

 

1 853.419 853.419 17.369 <.0001
4 58092.620 14523.155 295.586 <.OOOl
4 1091.933 272.983 5.556 .0002
1] 39198.722 3563.520 72.528 <.OOOl
l 1 1241.082 1 12.826 2.296 .0097
44 10108.314 229.734 4.676 <.0001
44 2549.209 57.937 1.179 .2077
449 22060.876 49. 133

 

 

 

 

 

 

6O

 

 

 

MPA

Mean (:tSEM) Number AR-IR Cells / 10,000,111)2

 

 

 

 

o u o a o o
i in s s s s it s ‘=’
h o H 6|— : H o H be
28 b 3 o r—r a f: 8 0
U o >. U o

[-< O... E—r QB

Figure 13. Mean (:tSEM) number of AR-IR cells / 10,000 umZ in individual brain
regions from prepubertal (open bars) and adult (closed bars) male ferrets injected
daily for 10 days with either oil (intact and castrate), 10 rig/kg E, 5.0 mg/kg T, or
5.0 mg/kg DHT.

 

61

50
4o
- 3o
50
. 40
j 30
20
10

 

MeA

 

MeP
meN ST

 

 

 

 

 

 

 

 

 

 

  

aBNST

   

 

 

 

. RCh

h r b b h I n r

bl

 

 

. F DPPLL
wmm mwmmw mm

$38.2 EEO m3? eeeaez cam—me 56:

h ’
0 0 O
3

occuuumoumou
-6625

0€Ou0um0umorﬁ

cowobwm

”~55me

SSE

25.33033
-6635

occuoumOumoH

cowobmm

oembmmu

Figure 13 (continued).

62

 

Figure 14. Photomicrographs of the pvPOA from prepubertal (A-E) and adult (F-
J) male ferrets that remained intact (a, t) or were injected daily for 10 days with
either 5.0 mg/kg T (b, g), 5.0 mg/kg DHT (c, h), oil ((1, i), or 10 ug/kg E (e, j).
Magniﬁcation bar = 30pm.

63

 

Figure 15. Photomicrographs of the MeA from prepubertal (A-E) and adult (F-J)
male ferrets that remained intact (a, f) or were injected daily for 10 days with
either 5. 0 mg/kg T (b, g), 5. 0 mg/kg DHT (c, h), oil (d, i), or 10 rig/kg E (e, j).
Magniﬁcation bar: 30pm

64

Interaction Between Age and Steroid Treatment: Figure 16 shows the mean
number of AR-IR cells for each steroid treatment of pre— and postpubertal male ferrets
collapsed across brain region. Fisher's PLSD tests examining the effects of steroid
treatment within each age goup showed that botln adult and prepubertal ferrets treated
with either T or DHT had a signiﬁcantly greater density of AR-IR cells than did their age
matched counterparts in any other steroid treatment group. Furthermore, intact animals
in each age group had a signiﬁcantly greater density of AR-IR cells than did those treated
with oil or E. These differences in the density of AR-IR cells among steroid treatment
groups represent the main effects of steroid treatment. The source of the interaction
between age and steroid treatment lies in the differences between prepubertal and adult
male ferrets that either remained intact or were castrated and treated with oil or E. Adult
animals that remained intact, as well as those that were castrated and treated with either
oil or E, had signiﬁcantly more AR-IR cells than did similarly treated prepubertal ferrets.
There were no differences between adult and prepubertal male ferrets castrated and
treated with either T or DHT.

Interaction Between Age and Brain Region: Figure 17 shows the mean
number of AR-IR cells in each brain region for pre- and postpubertal male ferrets
collapsed across steroid treatment. Fisher's PLSD tests between pre- and postpubertal
male ferrets within each brain region showed that in the RCh, lVMH, pvPOA, and the
aBNST adult animals had a signiﬁcantly greater density of AR-IR cells than did
prepubertal animals. In all other regions there were no signiﬁcant differences between

pre- and postpubertal male ferrets.

65

 

35
E] Prepubertal
30 r I Adult

*

25 ‘ l'l'

 

 

 

 

 

Mean (:tSEM) Number AR-IR Cells / 10,0001un2

e—u 8 I:
0
0 L3

Testosterone
testosterone

Dihydro-

Figure 16. Mean (iSEM) number of AR-IR cells / 10,000 um2 collapsed across
brain region from prepubertal (open bars) and adult (closed bars) male ferrets
injected daily for 10 days with either oil (intact and castrate), 10 rig/kg E, 5.0
mg/kg T, or 5.0 mg/kg DHT. Asterisk indicates signiﬁcant difference from
prepubertal animal in same steroid condition, double cross indicates signiﬁcant
difference from all treatment groups within an age group, single cross indicates
signiﬁcant difference from I, T and D within an age group.

66

A
C

 

DJ
LII
l u

E] Prepubertal
I Adult

.5:
10_ . I .
- _ I II

< <:
sass

U:
r

 

 

 

Mean (:tSEM) Number AR-IR Cells / 10,00011m2
N

o o
I I

pvPOA

cARC

VPM

lVMH

rARC

MeP

aBNST

meMST

Figure 17. Mean (:tSEM) number of AR-IR cells/ 10,000 um2 collapsed across
steroid treatment from prepubertal (open bars) and adult (closed bars) male ferrets.
Asterisk indicates signiﬁcant difference from prepubertal animals within the same
brain region.

67

Interaction Between Steroid Treatment and Brain Region: The interaction
between steroid treatment and brain region was analyzed by performing a one-way
analysis of variance for each brain region using data collapsed across age (Figure 18). In
each brain region, there was a signiﬁcant main effect of steroid treatment, and thus
pairwise comparisons between steroid treatments within each brain region were made
using Fisher‘s PLSD. Across all brain regions, castrated animals treated with oil did not
differ from castrated animals treated with E. Similarly, castrated animals treated with T
did not differ from castrated animals treated with DHT except in the MeA, where DHT
treatment resulted in signiﬁcantly more AR-IR cells than T treatment (indicated by the
cross). Animals treated with either oil or E had signiﬁcantly fewer AR-IR cells
(indicated by the single asterisk) than did those animals treated with T or DHT in all
regions except in the meNST, where those animals treated with oil or E did not differ
from those animals treated with DHT. In the pvPOA, MPA, MeA, MeP, rARC, cARC,
LS, lVMH, and PMV, intact animals had a signiﬁcantly greater density of AR-IR cells
compared to castrated animals treated with oil or E, while treatment with T or DHT
resulted in a signiﬁcantly greater density of AR-IR cells compared to intact animals in all
of these regions except in the MeP, where intact animals did not differ from DHT treated
animals, and in the PMV where intact animals did not differ from those treated with
either T or DHT. Brain regions which deviated from this general pattern were the
aBNST, meNST, and RCh, where the density of AR-[R cells in intact animals fell in
between animals treated with oil or E, and those treated with T or DHT. This resulted in
signiﬁcant differences only between intact animals and T treated animals in the aBNST

and meNST, and between intact animals and those treated with E in the RCH.

68

Figure 18. Mean (iSEM) number of AR-IR cells / 10,000 um2 within individual
brain regions collapsed across age group. Signiﬁcant differences only apply to
comparisons within a given brain region. Double asterisk indicates significant
difference from T and D, single asterisk indicates significant difference from T, D
and I, cross indicates signiﬁcant difference from T, solid circle indicates
Significant difference from E within brain region.

69
Mean (:tSEM) Number AR-IR Cells / 10,000um2

N b.) 8 UI

 

 

 

 

 

 

 

     

 

      

 

 

  

 

O o o o
I I r
._. are
pvPOA Q
'—1
C
cARC S .
S
VPM O .
m
'-i
U
lVMH 5‘
m
._]
U
rARC 5‘
m
._]
U
MPA é
a m-
HW - a
LS O 3‘38? 5-
m
"1
U
MW an:
MeA n ‘3-
m .
“'1 mm
D
_ m
RCh 0 MY$13_ O
5 “WWW“:
__ W
MeP n w -+
m _.
-l mm-
C
m
aBNST 5‘ 31f -*
r11 . ,
.—1 m
U
meMST 8 __+
a m-

 

Figure 18

70

Intensity of AR-IR Staining: Mann-Whitney U tests between groups using data
collapsed across brain region indicated that AR-IR staining in adult intact animals was
signiﬁcantly darker than in intact prepubertal animals, and that AR-IR staining in both
adults and prepubertal animals treated with T for 10 days was signiﬁcantly darker than
intact animals of the same maturational state (Figure 19). Pair-wise comparisons
between intact prepubertal and adult animals within each brain region (Figure 20)
indicated that in the pvPOA, MPA, MeA, RCh, and the lVMH adults had signiﬁcantly
darker immunostaining than did prepubertal animals, while in the rARC (p=0.083), LS
(p=0.99) and the aBNST (p=0.35) the differences in staining intensity were not
statistically signiﬁcant. Mann-Whitney U tests between intact and T treated animals of
the same age within each region indicated that in prepubertal animals, T treatment for 10
days increased the intensity of immunocytochemical staining in all regions except the
aBNST and the LS. However increases in immunocytochemical staining intensity failed
to reach statistical signiﬁcance in T-treated adult animals compared to intact adults,

except in the LS.

71

 

 

3.0 - T T

 

HQ)

2.5 -

 

2.0 ..

Her

 

1.5 -

1.0 L

 

 

 

Mean (:tSEM) AR-IR Staining Intensity

 

 

 

 

 

 

 

 

Aduli Intact Pre- Intact Adult'T-10d Pre- T-lOd

Figure 19. Mean (:SEM) immunocytochemical staining intensity from intact and
T treated prepubertal and adult male ferrets collapsed across brain region. Bars

with different letters are Signiﬁcantly different from one another (Mann-Whitney

U test, p < 0.05).

72

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4.0 1"va 1 MPA I ' 4.0
3.5 . i'_l l_' 1 3.5
3.0 T J— _L F 1 3.0
7.5 : j 2.5
2.0 :- I : 2.0
1.5 :- 1 1.5
1.0 ' “ 1.0
0.5 ﬂ 1 0.5
4.0 rrARC I FL IVNIH _Ll_1 _L.. '_I_ j 4.0
3.5 .- 1 3.5
3.0 r _L -‘ 3.0
2.5 f 1 2.5
2.0 +- I 1 2.0
1.5 f 1 1,5
1.0 r 1 1.0
0.5 :- ﬂﬁ {J— r 0.5
4.0 EMCA '_-_1 LS -: 4.0
3.5 1. l_'| '4 3.5
3.0 E- _L '—L _r_. '3 3.0
2.5 r 1 2.5
10: 1 PL 1 2.0
1.5 r .. 1,5
m1 3 1.0
0.5 L l—_IJ : 0.5
’ i

4.0 . RCh “BNST .‘ 4.0
3.5 i ' ' .‘ 3.5
3.0 L I"_| 1‘ 3.0
2.5 L __]__ ‘L‘ . 2.5
2.0L : 2.0
1.5; 4 l 1.5
ml ’ J 1.0
0.5 L I— ll] II H 'l l ,‘ 0.5

0 r: .1: .c: "’ "’ “O H a

EEse§§2§ easieeé§

ﬁght—155,19: got-«14.5.5713.

O.-§§§'=' Or===b§b

gtwwvgsg ssh-3355235:

.O 3 < < <2 '2 é i <3 < < 1:

<
< '2 < 1::
<1
Figure 20. Mean (:tSEM) immunocytochennical staining intensity in individual

brain regions from prepubertal and adult intact animals, and those castrated and
injected with oil or 5.0 mg/kg T for either 30 min, 4 hr, 8 hr, or 10 days. Bars
joined by arches are Signiﬁcantly different from one another (Mann-Whitney U

test, p < 0.05), however not all comparisons were analyzed, see text.

73

Experiment IV b: Effects of Duration of T Exposure on AR-IR Cell Density in
Adults
Win

Two-way ANOVA utilizing brain region and the duration of T exposure as the
independent variables showed signiﬁcant main effects of duration of T exposure
[F4’m=4395.9; p<0.0001], and brain region [me=154l.l; p<0.0001]. There was no
interaction between brain region and duration to T exposure [F “MEI .3; p=0.15].
Figure 21 shows the mean number of AR-IR cells per grid collapsed across brain region.
Post-hoe analysis showed no differences between castrated adults and those treated with
T for 30 min, even though circulating T was signiﬁcantly higher in the latter group
(Figure 12). All other pair-wise comparisons were signiﬁcant indicating that as the
duration of T exposure increases, there is an associated increase in the number of AR-IR
cells per unit area. Examination of individual brain regions (Figure 22) indicated that
only two regions deviated from the general pattern of a steady increase in AR-IR cell
number following longer exposure to T. In the PMV, AR-IR cells reached their
maximum level after only for 4 hr of T exposure, while in the meNST, AR-IR cells
were virtually absent in all groups except those which had been treated with T for 10
days.
lR-IBSI . . I | 'I

Kruskal-Wallis tests comparing the intensity of immunocytochemical staining
among treatment groups collapsed across brain region indicated that there were
signiﬁcant differences in staining intensity (Figure 23). Mann-Whitney U tests indicated

that castrated animals and those treated with T for 30 min were signiﬁcantly different

74

 

 

 

 

 

 

 

 

 

 

 

 

 

NE 35

§ ' d

, 30 .. _T__

.9 .

\

g 25 " c

3 _I__

E 20 - b ‘

g D I

t— 15 '

.8 a

S T

z 10 a

E s —

m

i!

5 0 '

D

2 8 E J: .c: 13
g o a °9 e
s '1‘ E— E— '
U E— —3 a S
S 2:: -o '0 :3
{3’ 6" < < '2
< <

Figure 21. Mean (:tSEM) number of AR-IR cells / 10,000 um2 collapsed across
brain region from adult animals castrated and injected with oil or 5.0 mg/kg T for
either 30 min, 4 hr, 8 hr, or 10 days. Bars with different letters are signiﬁcantly
different from one another (p < 0.05).

75

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.pvPOA cARC ,
50 - «50
. 1' ‘40
40 l I
30 r ' ‘30
N I I +
E20 l r «20
. I .
8 10 - ,10
q . .
O . . . . . . f
H
3 PW lVMH ,
g 50 ’- j50
E 40 L I I r T ‘40
I
- ~30
% 30 ’ I I I ‘
E 2" l I 12°
E 10 L -10
5 - W
Z . . . , . . . v . T
A
a rARC MPA ,
m 50 ' .50
+| r t
E 4" L 4°
0 ’ ‘
30 ' ‘30
E . ,
20 ~ I I ‘20
D I T 1
IO - I ‘10
8 .c: .r: .c: “U 8 t: .r: 4: ":3
g E <r «3 2 g 'E' ‘l’ °? 53
3 8 t t E'— 53 8 *‘ "‘ E"
I v— —! H U | a: x H
:s z '2 2 {3’ '5 ... “U “U :3
'0 :3 < 13 '5' < < <
< “O < 13
< <

Figure 22. Mean (:tSEM) number of AR-IR cells I 10,000 um2 in individual brain
regions from adult animals castrated and injected with oil or 5.0 mg/kg T for either
30 min, 4 hr, 8 hr, or daily. Statistical analysis was not performed in individual
regions.

76

 

 

1111111111

 

 

 

 

 

 

. c OTB :25.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MeA
J . Ir—f—I
MeP
J—“—l
meNST

: mum. :33.»

a v4. :35.

. 58 cm? :2u<

. 39580 22%

 

 

 

 

 

 

 

 

 

 

. U OTB 239$

 

 

 

l
l
[4.1

 

 

 

 

 

 

 

 

 

 

 

1. 1m

 

 

 

 

 

 

 

[ lt—f—‘l I

. : mun. “16¢.

. a .1. “32

E8 Om? :sv<

. amuse =22

 

 

 

 

 

 

, aBNST

w m

bbbbbbbbbb

 

10»

m wmmm ,w-.w..m.wlm shown.
«838.2 3:8 m3? e352 Gamma; :32

Figure 22 (continued).

77

 

 

 

 

 

 

 

 

 

 

 

 

 

 

b * C
g 3.0 r T
g r
E” 2.5 -
.= b
U5, 2.0 - T
E . b
a; 1.5 - -—T-—
“ _ a
a 1.0 - ——1— a
m
H
E 0.5 r-
O
E 0 . . .
=3 t- s.- "U
8 g -= J: o
=9 o it °9 '7
32% "9 E“ E" E?
«9 r— '3
"D
<

Figure 23. Mean (:SEM) immunocytochemical staining intensity, collapsed
across brain region. Bars with different letters are signiﬁcantly different from one
another (Mann-Whitney U test, p < 0.05).

78

 

 

Figure 24. Photomicrographs from the PMV, and lVMH from adult animals
castrated and injected with oil (a, f) or 5 .0 mg/kg T for either 30 min (b, g), 4 hr (c,
h), 8 hr ((1, i ), or once daily for 10 days (e, j). Magniﬁcation bar: lOOum

79

ﬁnal those treated with T for either 4 hr or 8hr, while those animals treated with T for 4
hr were not different from those treated for 8 hr. Additional tests comparing the staining
intensity between animals treated with T for 10 days and animals treated with a single
dose of T for either 30 min, 4 hr or 8 hr indicated that AR-IR cells in 10d animals were
signiﬁcantly darker than cells in any of these groups. Photomicrographs of AR—IR cells

in the PMV and lVMH following T injections are shown in Figure 24.

Discussion

Regulation of AR-IR cell density following 10 days of steroid administration

The results of this experiment demonstrate that some changes in the density of
AR-IR cells following alterations in the steroidal milieu are dependent upon the
maturational state of the animal. When data are collapsed across brain regions, there is a
modest but signiﬁcant increase in the number of AR-IR cells per grid in gonadally intact
adult male ferrets compared to intact prepubertal ferrets. The magnitude of this
difference between gonadally intact pre- and postpubertal male ferrets is not as great as
observed in our previous experiments comparing prepubertal ferrets and those in the
process of pubertal maturation (Experiments I and II). However, this is likely due to
modiﬁcations in the immunocytochemical staining procedure that enhanced our ability to
detect more of the lightly stained cells in the prepubertal animals. For example, in the
MeA the number of cells in intact prepubertal animals is approximately 2 times greater in
Experiments III and IV compared to Experiments I and II. The increase in AR-IR cell
detectability across experiments did not eliminate differences between pre- and

postpubertal male ferrets. This difference is clearly evident when the intensity of the

80
staining is examined (Figures 14, 15, and 19). In most regions where intensity of
immunocytochemical staining was examined, AR-IR cells in adult ferrets were darker
than in prepubertal male ferrets. Assuming that intensity of immunocytochemical
staining is a function of the quantity of the antigen within the cell, it is likely that adult
ferrets have more AR proteins per cell than do prepubertal ferrets. This however cannot
be conﬁrmed until other experimental methods, such as quantitative androgen receptor
binding, are utilized. Based on other experiments which have quantiﬁed androgen
receptor binding under various hormonal conditions it seems likely that adult ferrets
would have increased androgen receptor binding relative to prepubertal ferrets. For
example, in brain tissue, androgen treatment has been shown to increase androgen
binding in both the hamster and the rat (Bittman & Krey, 1988; Krey & McGinnis,
1990). Thus, higher circulating androgen concentrations in adult ferrets would
presumably lead to increased androgen receptor binding. Again, it has not been
unequivocally determined if this increase in binding is a result of an increase in AR
protein synthesis, an increase in the half-life of the AR protein, or a combination of
these. An experiment where protein synthesis inhibitors are concurrently administered
with androgen treatment would help to clarify this issue.

The differential response to changes in steroidal milieu between pre- and
postpubertal male ferrets is also apparent when castrated animals treated with oil or
estrogen are examined. There is a signiﬁcantly greater density of AR-IR cells in both oil
and E treated adult ferrets compared to prepubertal ferrets (Figures 14-16). The reason
for this difference cannot be determined based on data from this experiment. These

differences may be due to differences in the output of adrenal androgens from pre- and

81

postpubertal castrated animals, or it could simply be due to the fact that adult ferrets had
a higher level of androgen prior to castration, and that the decline in AR-IR cell density
normally seen after castration takes longer than the 17 day post-castration period
examined in the present experiment. This explanation is unlikely based on data indicating
that the decline in androgen receptor binding occurs within several hours of steroid
removal (Krey & McGinnis, 1990). It is also possible that exposure to androgens during
pubertal maturation has permanently altered a subset of AR-IR cells that retain a higher
degree of AR protein in the absence of circulating androgens. It would be interesting to
determine if these additional AR-IR cells are present in prepubertal animals following
withdrawal of exogenously administered androgens. Evidence for permanent effects of T
on the CNS during pubertal maturation does exist. Speciﬁc environment-related social
interactions characteristic of adult male rats are dependent upon the presence of T during
pubertal development, and not on the presence of T during adulthood (Primus &

Kellogg, 1990). Further research has indicated that the GABA receptor may be involved
in mediating these social interactions. The presence of the gonads during pubertal
maturation inﬂuences the anxiolytic effect of diazepam on environment-related social
interactions in the adult rat (Primus & Kellogg, 1990).

The most striking effects of the present experiment are the many similarities
between adult and prepubertal ferrets in response to treatment with gonadal steroids. 0n
the one hand, relative to oil treated castrated animals, there was no effect of E treatment
on the density of AR-IR cells in either age group (Figure 16). Similarly, both pre- and
postpubertal male ferrets castrated and treated with 5 mg/kg androgens (T or DHT) for

10 days display essentially identical patterns of AR immunocytochemical staining. This

82

effect is evident in terms of both cell density, and in the intensity of
immunocytochemical staining. Examination of the data with regard to the age by steroid
interaction (i.e., irrespective of brain region), indicates that androgen treatment resulted
in a clear up-regulation in the density of AR-IR cells (Figure 16). Furthermore,
immunocytochemical staining intensity in ferrets treated with androgens is greater than in
castrated or intact animals (Figure 20). Thus, following 10 days of androgen treatment,
prepubertal and adult ferrets display similar AR-IR staining parameters. However since
adult castrated ferrets have a greater density of AR-IR cells than do castrated prepubertal
animals, it is clear that prepubertal animals required a greater increase in AR-[R cell
density to reach this point. Exactly when this particular level of AR-IR cell density is
reached, or if prepubertal and adult animals reach it at the same time point following the
beginning of androgen treatment is unclear, though based on data from the short term T-
treatrnent to adult ferrets, it apparently takes more than 8 hr.

The underlying rationale for the present experiment was to identify anatomical
substrates that may correlate with changes in behavioral responsiveness to T during
pubertal maturation. It has been demonstrated that prepubertal male ferrets require a
higher dose of T to engage in male reproductive behaviors compared to adults that
undergo pubertal maturation in either short days or long days (Sisk, et al., 1992). At
higher doses (5.0 and 10.0 mg/kg), the amount of time that prepubertal animals engage in
mounting and thrusting while paired with a female compared with that of adults raised in
short days also tended to be lower, however differences were not statistically signiﬁcant
(Sisk, et al., 1992). The evidence from this experiment indicates that any possible

differences in responsiveness to the effects of 5.0 mg/kg of T are not due to a failure of T

83

to upregulate the expression of AR in the brains of the prepubertal ferrets treated with T.
In all respects, the distribution of AR-IR cells in prepubertal animals treated with 5.0
mg/kg of T for 10 days is identical to adults treated in a similar manner. On the other
hand, it is clear from the behavioral experiment (Sisk, et al., 1992), that treatment of
prepubertal animals with 5.0 mg/kg of T will induce a signiﬁcant amount of male sexual
behavior in most of the animals. While no causal link between AR-IR cell density and
behavioral capacity can be formed based on the correlational data provided in this
experiment, the increase in the density and staining intensity of AR-IR cells in the brains
of prepubertal animals treated with T, and their increased likelihood of engaging in male
sexual behaviors is consistent with the hypothesis that an up-regulation of AR-IR cells in
the brains of male ferrets as they undergo pubertal maturation is a step which precedes
the induction of male reproductive behavior patterns. An experiment where both
reproductive behavior and a measure of AR, either immunocytochemical staining or
androgen-binding capacity, are analyzed following treatment with a lower dose of T -
one capable of activating reproductive behavior in adults, but not in prepubertal animals -
would help to clarify this issue. If an identical pattern of AR—IR staining were found in
this case, it would indicate that an upregulation of AR could not be the limiting factor in
the activation of reproductive behavior in T treated prepubertal male ferrets. On the
other hand, another outcome of such an experiment could be that the AR staining
patterns of prepubertal males was different, speciﬁcally that there are fewer AR-IR cells
in brain regions regulating reproductive behavior patterns in steroid treated prepubertal
animals compared with adults. Were this the case, it would provide support for our

initial hypothesis, that pubertal changes in responsiveness to the behavioral effects of T

84

are due to pubertal changes in the ability of T to upregulate the AR itself. However,
given that a dose of Smg/kg of T to pre- and postpubertal animals resulted in the same
pattern of AR-IR staining, it is unlikely that a lower dose of T would result in differential
staining. We know that the pattern of AR-IR staining found in the present experiment is
not required for the activation of reproductive behavior because intact adults, which have
a lower density of AR-IR cells in many regions compared with T treated adults, will
display the full range for reproductive behaviors (Sisk, et al., 1992). A more plausible
explanation would be that, if responsiveness to the behavioral actions of T is a function
of a differential response to circulating androgens, then this difference is not at the level
of steroid induced increases in androgen receptor levels. This does not eliminate possible
pubertal differences in the regulation of estrogen receptors. Pubertal changes in estrogen
receptor immunocytochemistry, and estrogen receptor binding have been reported in
female Guinea pigs, and rats within speciﬁc brain regions (Brown, Hochberg, Naﬁolin,
& MacLusky, 1994; Olster, 1994). Speciﬁcally, there is an increase in the number of
estrogen receptor-immunoreactive cells in the rostral portion of the ventrolateral
hypothalamus in adult female compared to prepubertal female guinea pigs (Olster, 1994),
consistent with the increase in responsiveness to the behavioral effects of E in these
animals as adulthood is reached. Similarly, there is an increase in nuclear estrogen
receptor binding in the POA and amygdala of female rats at the time of pubertal
maturation (Brown, et al., 1994). Whether similar changes in brain estrogen receptors
occurs during pubertal maturation in males remains to be determined.

The signiﬁcant steroid by region interaction indicates that there are differences

across brain regions in the way in which animals, irrespective of age, respond to changes

85

in steroid hormone concentrations. The source of this interaction is partly attributable to
differences in the magnitude of the response to androgen treatment. In some regions,
including the pvPOA, MPA, cARC, rARC, MeA, LS, and lVMH, androgen treatment
resulted in a signiﬁcant increase in the density of AR-IR cells above the level seen in
intact animals, while in the PMV and the RCh, androgen treatment did not produce a
signiﬁcant increase in AR-IR cell density compared with intact animals. In our previous
experiment in which we analyzed the steroid regulation of steroid receptors in
prepubertal male ferrets (Experiment III), we were left with the impression that in some
brain regions androgen treatment resulted in an increase in the density of AR-IR cells
above the level seen in intact animals while in other brain regions this treatment only
resulted in a restoration of AR-IR cell density to intact levels (Experiment III; (Kashon,
Hayes, Shek, & Sisk, 1995)). We interpreted those data as representing a regional
difference in the ability of androgens to upregulate AR-IR cells. However, in light of the
current data, a reinterpretation seems necessary. In every brain region examined, T
treatment resulted in a signiﬁcant increase in AR-IR cell density relative to castrated
animals, thus it is thus clear that cells in all regions are capable of increasing AR in
response to increasing levels of T regardless of the maturational state of the animal.
Furthermore, there are no differences in the density or the staining intensity of AR-[R
cells in any brain region between prepubertal and adult animals treated with 5.0 mg/kg of
androgen, indicating that the response to the dose of T used in the present experiment is
the same at each maturational state. Thus differences among brain regions regarding the
ability of androgens to increase the density of AR-IR cells relative to intact animals does

not seem to be due to a region speciﬁc ability for androgens to increase AR, but due to a

86

regional differences in the steady-state levels of AR under physiological levels of
androgen secretion. An increase in the density of AR-IR cells above intact levels may
thus be dependent upon the number of cells within a nucleus that express AR under
physiological conditions. In other words, endogenous androgen secretion may lead to a
near maximal level of AR-IR cells in some regions, and thus a ceiling is reached whereby
additional androgenic stimulation is not effective in increasing cell number, while in
other regions endogenous androgen secretion results in a low number of AR expressing
cells relative to the number possible. For example, in the PMV, androgen treatment
failed to increase the density of AR-IR cells above the level seen in intact animals. The
PMV contains the most intensely stained AR-IR cells within the brain, and it appears that
under physiological conditions it contains its maximum density of AR-IR cells since
androgen treatment failed to induce an upregulation in AR-IR cell number relative to
intact animals. At the other extreme is the BNST where AR-IR cell number and staining
intensity are low. Only prolonged androgen treatment (10 days) was able to produce an
increase in AR staining, thus under physiological conditions, staining for AR-TR in the
BNST also appears to be near its maximum density. The factors that contribute to
regional differences in AR-IR cell number and intensity of staining under experimental
and physiological conditions are unknown. The particular 'phenotype‘ of a given brain
cell is likely a function of developmental processes, as well as the particular afferent and
efferent connections it possesses.

The interaction between age and brain region shown in Figure 17 indicated that
there are signiﬁcant differences in some brain regions between prepubertal and adult

male ferrets on the number of AR-IR cells when collapsed across steroid treatment.

87

Although this analysis is a necessary component of the 3-way ANOVA, it may have very
little meaning in a physiological sense. Given that circulating androgen concentrations
are one of the primary factors which inﬂuences the density of AR-IR cells, it is difﬁcult
to interpret the signiﬁcance of age differences within a region when such diverse steroid
treatments contribute to the means that are being compared. Figure 17 should not be
confused as equivalent to Figure 5, since Figure 5 is based on data from intact animals
only.
Short-Term Administration of T

Following a single injection of T to castrated adult male ferrets, a gradual
increase in the number of AR-lR cells occurred between 30 min and 8 hr of exposure.
This increase in AR-IR cell density relative to castrates was even greater in animals that
had been exposed to daily injections of T for 10 days. These data support the notion that
androgens are capable of inducing an upregulation of the AR protein, and that a
translocation of receptor proteins into the nucleus, or a conformational change in the AR
receptor protein that allows the antibody to bind more readily, is unlikely to fully account
for the increase in the number of AR-IR cells in animals exposed to T over many days.
If the increase in AR immunoreactivity were merely an artifact of immunocytochemical
staining in the presence or absence of androgens we would have expected to see a single
increase in AR-IR cell number and intensity of staining. In most brain regions there was
a steady increase in AR-IR cell number and intensity of staining up to 8 hours following
steroid treatment. There was an additional increase in these parameters seen in those
animals treated for 10 days with T in most brain regions.

Although no statistical differences exist between castrated adults and those treated

88

with T for 30 min, there are differences in immunocytochemical staining patterns that
warrant discussion. Figure 24 shows photomicrographs from the PMV and lVMH of
adults treated with T for various periods of time. In the castrated animal (panels A and
F) AR-IR staining is diffuse and sometimes extended into the cytoplasmic compartment
of the cell. However, 30 min following a single injection of T, all staining is restricted to
the nucleus (panels B and G). Not all regions displayed cytoplasmic AR staining in
castrated animals. Indeed, only in those regions that display high levels of AR staining
under physiological conditions, such as the VPM, lVMH, and occasionally in the POA,
was cytoplasmic staining visible. Other experiments have reported rapid effects of
androgen treatment on AR staining patterns, however quantiﬁcation of AR-IR cells was
not performed. These experiments reported an upregulation in AR-IR staining in
prostate tissue within 15 min of an injection of T in an ethanol vehicle (Prins & Birch,
1993; Sar, et al., 1990), and a rapid increase in AR-IR staining was reported in hamster
brain 15 min following an injection of T although the vehicle was not speciﬁed (Zhou,
Blaustein, & De Vries, 1994). Thus it is possible that route of administration could
partially explain these discrepancies since T administered in an ethanol vehicle may
reach target cells more rapidly than T in an oil vehicle.

It is interesting that the number of AR-IR cells in those animals exposed to T for
8 hr was approximately double the number seen in castrated animals. This is consistent
with the view that androgens are able to increase the half-life of receptor proteins ﬁom
approximately 1.5 hr to 6 hr as reported by (Kemppainen, et al., 1992; Robel, et al.,
1983). The additional increase seen following 10 days of T exposure cannot be

explained in terms of an increase in the half-life of the receptor proteins without

89

additional information regarding transcriptional activity of the AR gene. Transcriptional
activity of the AR gene in many tissue types is complex and under the inﬂuence of
several factors including second messenger systems and autoregulation by androgens
themselves (Lindzey, Kumar, Grossman, Young, & Tindall, 1994). The autoregulatory
effect of androgens is cell and tissue speciﬁc, and appears to occur at the transcriptional,
translational and posttranslational levels (Lindzey, et al., 1994). With respect to
transcriptional regulation, the most common experimental ﬁnding is that androgens
reduce the steady state levels of AR-mRNA in kidney, brain, coagulating gland,
epididymis, and LNCaP cells, while CAMP increases AR—mRNA (Blok, Themmen,
Peters, Trapman, Baarends, Hoogerbrugge, et al., 1992; Krongrad, et al., 1991;
Quarmby, Yarbrough, Lubahn, French, & Wilson, 1990; Tan, Joseph, Quarmby, Lubahn,
Sar, French, et al., 1988). However, some experiments have shown increases in the
level of AR-mRNA in response to androgen treatment (Burgess & Handa, 1993;
Gonzalez-Cadavid, Vemet, Navarro, Rodriguez, Swerdloff, & Raj fer, 1993; Shan,
Hardy, Catterall, & Hardy, 1995). A decrease in mRNA levels is not necessarily
accompanied by a decrease in protein levels (Krongrad, et al., 1991), which suggests that
changes in protein or mRNA stability may offset changes in transcriptional activity. This
is further supported by the ﬁnding that androgens induce a two-fold increase in the half-
life of AR-mRNA in LNCap cells (a prostate carcinoma cell line), and reduce the
initiation of new AR-mRNA transcripts (Blok, et al., 1992; Wolf, Herzinger, Herrneking,
Blaschke, & Horz, 1993). Thus a general assumption that high mRNA levels lead to
high protein levels is not warranted. In a functional sense, it may not be important if

increased levels of AR protein following androgenic stimulation are due to an increase in

90
protein synthesis or an enhanced receptor stability. The bottom line is that there is more
AR present to act on the genome and either enhance or repress gene transcription.

The data on the short term effects of T on AR-IR staining also indicate that
steroid history is an important factor. A single blood sample, taken at the time of
sacriﬁce, is not necessarily a good predictor of the density or the staining intensity of
AR-IR cells. For example, animals treated with T for only 4 hours have fewer AR-[R
cells than those treated with T for 10 days, yet there are no differences in T levels taken
at the time of death. Based on data from this experiment, a longer duration of exposure
to high levels of T leads to an increase in the density of AR-TR cells and to a greater
staining intensity. It is thus clear that the effects of long-term steroid administration are
to a certain extent cumulative in nature, thus correlations between T levels and AR-IR
staining parameters could be misleading if steroid history is unknown.

As described for long term (10 day) treatment with T, there are some regional
differences in the manner in which short term administration of androgens can affect AR-
IR staining. For example in the PMV, an area which has some of the darkest staining
AR-IR cells within the brain, the number of AR—IR cells visible 4 hr after a single
injection of T to a castrated animals was equivalent to the number seen in an animal
which had been treated daily for 10 days with androgens. On the other hand, in the
meNST, there were virtually no AR-IR cells present at any time point following the
single injection of T, and were only present after 10 days of T administration. The
signiﬁcance of these differences is unknown, however it seems likely that these regional
differences in the response to short term T administration are related to regional

differences in AR-IR staining under physiological conditions and thus would impact the

91

ability of high concentrations of androgens to induce an upregulation relative to intact
animals.

In summary, this experiment has demonstrated that prepubertal male ferrets are
fully capable of responding to androgen treatment with an upregulation in the number
and/or staining intensity of AR-IR cells. The increase in AR-IR staining parameters
following androgen treatment in prepubertal ferrets is identical to what is observed in
fully mature adult males. Thus it is unlikely that changes in the responsiveness to the
behavioral actions of T which occur during pubertal maturation are due to differences in
the steroid hormone induced regulation of the androgen receptor. This does not rule out
the possibility that the transcriptional activity of the AR protein is different between
prepubertal and adult male ferrets, and indeed the possibility exists that changes in
responsiveness to steroid hormone treatments are due to pubertal changes in the ability of

the steroid receptor complex to either initiate or repress speciﬁc gene transcription.

GENERAL DISCUSSION

The present experiments have demonstrated that as animals undergo pubertal
maturation there is an increase in androgen receptor-immunoreactive staining in many
regions of the brain including those known to be involved in reproductive behavior
patterns. This increase in AR-IR staining is correlated with the pubertal increase in
responsiveness to the behavioral effects of T, and is tightly linked to the increase in
circulating androgens that occurs during puberty. Advancing the onset of pubertal
maturation, and the associated increase in T, advances the age at which AR-[R staining
begins to increase. Treating prepubertal male ferrets with androgens was able to increase
the number of AR-IR cells in several brain regions relative to intact prepubertal ferrets,
providing further support for the hypothesis that the pubertal increase in AR-IR staining
is linked to the pubertal rise in T. Since prepubertal male ferrets are less sensitive to the
behavioral effects of T than are adults (Sisk, et al., 1992), we then tested the hypothesis
that the differences in sensitivity between prepubertal and adult male ferrets may be due
to differences in the ability of androgens to induce an upregulation in AR-IR staining.
Speciﬁcally we proposed that even though androgen treatment was capable of producing
an upregulation in AR-IR staining in prepubertal males, it was possible that the
magnitude of this response to androgens would be even greater in adults treated with
androgens. However, the pattern of brain AR immunoreactivity in prepubertal males
treated with a relatively high dose of androgens was indistinguishable from adults treated
with androgens. Therefore, a failure of androgen treatment to induce a sufﬁcient

upregulation in AR-IR containing cells in prepubertal male ferrets is an unlikely

92

93

mechanism for the differential behavioral responsiveness to T. However, the effect of
lower doses of androgen that yields a greater difference in behavioral activation in pre-
and postpubertal males, and other methods, such as androgen binding which are more
sensitive than immunocytochemistry, should be utilized before this hypothesis is
completely ruled out.

There are several other means by which differential sensitivity to testosterone
could be mediated. Data from the experiments in this dissertation do not address the
cellular effects of androgen treatment outside of its apparent autoregulatory role on
androgen receptor expression. It is conceivable that the activated steroid receptor
complexes in prepubertal males are not operating in the same fashion as in adults, and
that differential responsiveness to T may be due to changes in the downstream
physiological effects of the AR. One possibility is the androgenic regulation of the
aromatase enzyme. It is known that aromatase activity in speciﬁc brain regions can be
increased by the administration of T and that this is an androgen receptor-dependent
mechanism (Roselli & Resko, 1984). Evidence from research on the effects of
photoperiod on reproductive ﬁrnctioning in male hamsters indicates that aromatase
activity is differentially active under different photoperiod conditions. Adult hamsters
that have undergone gonadal regression in short days are less responsive to the
behavioral actions of T on reproductive behavior (Morin & Zucker, 1978), and have
fewer nuclear androgen receptors in the POA, as measured by in vitro ligand binding
techniques, than reproductively active male hamsters in long days (Bittman & Krey,
1988). Furthermore, there is a signiﬁcant reduction in hypothalamic aromatase activity

in gonadally regressed hamsters housed in short days compared with those hamsters

94

housed in long days, and androgen administration to short day hamsters fails to increase
hypothalamic aromatase activity compared with those housed in long days (Callard, et
al., 1986). Whether this is related to the reduced level of AR in regressed hamsters is
unknown. However, if a similar situation exists in prepubertal male ferrets, then it is
conceivable that the reduced responsiveness to T is a function of a reduction in aromatase
activity and thus a decrease in the amount of available estrogen. This could in turn, alter
estrogen receptor levels in speciﬁc regions since the level of estrogen receptor mRNA
(Lauber, Mobbs, Muramatsu, & Pfaff, 1991; Lauber, Romano, Mobbs, & Pfaff, 1990;
Shughrue, Bushnell, & Dorsa, 1992; Simerly & Young, 1991), and estrogen receptor
binding (Krey, Lieberburg, Roy, & McEwen, 1979; Lustig, Mobbs, Bradlow, McEwen,
& Pfaff, 1989; McGinnis, Drey, MacLusky, & McEwen, 1981) are modulated by
changes in gonadal steroids.

Another potential site of action that may be involved in mediating
differential responsiveness to the behavioral effects of T is the androgenic regulation of
neuropeptides. The behavioral effects of androgens are likely mediated by the gene
products that are produced following androgen administration. Thus, even though
androgens stimulate an upregulation of the androgen receptor in both pre- and
postpubertal males, it is possible that these receptor complexes do not have the same
effect on gene transcription. This could be due to differential access to steroid response
elements, or the presence of other factors that prevent or facilitate androgen receptor
complexes from initiating gene transcription. However, there are a large number of
androgen regulated neuropeptides such as vasopressin, oxytocin, and galanin (Harlan,

1988), yet limited knowledge regarding the effects of these gene products on male

95

reproductive behavior (Doman & Malsbury, 1990), and in the mediation of negative
feedback on gonadotropin secretion. Future studies investigating how responsiveness to
the effects of T may be mediated by differential activation of androgen responsive gene
products should be ﬁrmly based on empirical research regarding the role of these gene
products on the activation of male reproductive behavior and negative feedback. When
examining the role of the androgen receptor in mediating differential sensitivity to
testosterone, the downstream effects of steroid responsive gene products must also be
addressed. If there are no differences between prepubertal and adult male ferrets in the
autoregulatory effects of androgens on the androgen receptor, this does not indicate that
these animals are equally sensitive to the actual gene products which are induced
following the transcription of both androgen and estrogen responsive gene products. As
more information becomes available regarding the biological responses mediated by
steroid responsive gene products more speciﬁc hypotheses can be generated to examine
this particular aspect of steroid hormone action on reproductive behavior and negative
feedback.

Another possibility that has not yet been addressed is that peripheral structures
such as penile musculature, or sensory pathways which provide feedback to CNS
structures may be differentially sensitive in pre- and postpubertal males. For example, it
is possible that androgens modulate sensory information from the penis and that in
prepubertal males, the feedback obtained ﬁom copulation may not be processed in the
same manner within the CNS of prepubertal males relative to adults. The same can be
said for chernosensory information encountered during precopulatory anogenital

investigation of the female, which is a necessary component for copulation to occur in

96

many species (Meisel & Sachs, 1994). It is known that interest for vaginal secretions is
increased in male hamsters that are treated with androgens (Johnston & Coplin, 1979).
With respect to negative feedback on gonadotropin secretion, non-neural targets such as
glia cells may be involved in modulating differential sensitivity to the negative feedback
actions of testosterone. Glial cells within the median eminence have been shown to
contain estrogen receptors (Langub & Watson, 1992), and the ultrastructural relationship
between non-neural elements and LHRH terminals changes with steroid milieu and
pubertal maturation (King & Rubin, 1994; Witkin, Ferin, Popilskis, & Silverrnan, 1991;
Witkin & Romero, 1995). Thus glial cells may participate in the regulation of LHRH
release, and are a potential site for steroid modulation of gonadotropin secretion.

Multiple and interconnected mechanisms are likely to be involved in modulating
changes in responsiveness to gonadal steroids. This view is further supported by the fact
that changes in responsiveness to steroid hormones, either during pubertal maturation or
during photoperiod changes in reproductive state, occur in both directions. That is, as
responsiveness to negative feedback effects decreases, the responsiveness to the
behavioral effects of steroids increases. However, these are not likely to be occurring
simultaneously since the decrease in responsiveness to negative feedback in many species
occurs prior to, and is responsible for, the increase in gonadal function that eventually
leads to an increase in reproductive behaviors. Moreover, there is evidence that the
neuroendocrine and the behavioral effects of gonadal steroids are occurring in discrete
regions of the brain. For example, implantation of T directly into the MPA of male
ferrets will activate reproductive behavior patterns, but not cause a reduction in

circulating LH levels (Tang & Sisk, 1991). Thus, it can be assumed that the MPA is not

97

a brain region which is responsible for mediating negative feedback effects on
gonadotropin secretion, at least in the ferret. Similar studies in which steroids were
directly implanted into the mediobasal hypothalamus of rodents have shown that such
implants can induce a reduction in gonadotropin secretion, however behavior was not
tested in these studies (Cheung & Davidson, 1977; Kalra & Kalra, 1980). Another
approach to this problem has been to produce lesions in speciﬁc brain regions thought to
mediate these distinct effects of gonadal steroids. In terms of the behavioral effects, a
large body of literature has determined that many regions are involved in mediating
reproductive behaviors including the POA, AHA, LS, MeA, BNST and structures in the
main and accessory olfactory pathways (reviewed in (Meisel & Sachs, 1994)). On the
other hand, lesions in the mediobasal hypothalamus, which could include RCh, ARC and
VMH, indicate that this is a likely target site for the effects of gonadal steroid induction
of negative feedback (Cheung & Davidson, 1977; Plant, Krey, Moossy, McCormack,
Hess, & Knobil, 1978). Again, comprehensive studies that simultaneously examine
mechanisms of both negative feedback and behavior are generally not performed. The
experiments in this dissertation have indicated that in all of these regions, i.e., both those
known to be involved in mediating negative feedback and those involved in the
regulation of behavior, increases in T, either during puberty or by exogenous
administration, lead to increases in AR-IR cells. Thus, the regulation of androgen
receptors in these brain regions with diverse functions is essentially the same, even
though the amount of T required to induce a behavioral response, as well as to induce
negative feedback, presumably by acting in these discrete regions, is different. Thus, the

working hypothesis of this dissertation, that changes in responsiveness to steroid

98

hormones are positively correlated with changes in steroid receptors seems to be too
simple to explain what actually may be occurring. Indeed, there are no empirical data to
suggest that a decrease in the responsiveness to steroid hormones necessarily requires that
there be a reduction in the number of steroid receptors. Other factors such as steroid
metabolizing enzymes could also prevent steroids from exerting their speciﬁc
physiological actions. It is also conceivable that an increase in steroid hormone receptors
could lead to a decreased biological response depending on which speciﬁc gene products
were induced by the interaction of the steroid hormone receptor with the DNA, and what
their effects are on the particular physiological process under investigation. In
conclusion, based on the immuncytochemical experiments in this dissertation, it cannot
be unequivocally determined if regulation of androgen receptors is a step at which
responsiveness to the effects of T is being mediated. Future studies using more sensitive
techniques of androgen receptor quantiﬁcation, and those examining the cellular effects
of gonadal steroids in discrete brain regions will help to illuminate exactly how

responsiveness to steroid hormones is mediated.

APPENDICES

Appendix A
Distribution of Androgen Receptor-Immunoreactive Cells

in the Forebrain of Male Ferrets

Testosterone and estradiol are gonadal steroids that exert profound regulatory
actions on reproductive function by altering developmental, behavioral and
neuroendocrine processes. Many of these processes are mediated by steroid hormone
regulation of gene transcription. Binding of steroid hormones to their intracellular
receptor proteins, and the subsequent binding of this steroid-receptor complex to steroid
response elements on the DNA, lead to an alteration of speciﬁc gene transcription (Beato,
1989; Evans, 1988; O‘Malley, et al., 1986; Truss & Beato, 1993; Yamamoto, 1985).
Both reproductive behavior and gonadotropin secretion are regulated by the presence of
steroid hormones and functional steroid receptors across several stages of the life span,
including during sexual differentiation, puberty, and in adulthood. Localization of
steroid receptor containing cells within the central nervous system has contributed greatly
to our understanding of ﬁrndamental neuroendocrine processes.

The localization of steroid hormone receptors has been accomplished using a
variety of techniques. Their existence was initially established using biochemical
methods which demonstrated that both androgens and estrogens bind speciﬁcally to cell
nuclei isolated from brain tissue homogenates (McEwen, 1976). However, this

anatomical speciﬁcity of this technique is limited to microdissected brain regions and

99

100

speciﬁc cell types could not be determined. Autoradiographic histological techniques
were able to achieve cellular resolution by taking advantage of the speciﬁcity of steroid
hormone binding to their receptors. Cells which contain speciﬁc binding proteins
(presumably steroid receptors) will accumulate radio-labeled hormones such as tritiated
E or T. Speciﬁc anatomical location of these cells is possible at the light microscopic
level. Using autoradiographic techniques, the distribution of androgen and estrogen
receptors has been described for the rodent brain (Sar & Stumpf, 1977; Sar & W.E.,
1975), and has provided the anatomical basis for much of the research conducted in the
ﬁeld of neuroendocrinology. More recently, in-situ hybridization histochemistry has
been utilized to extend our knowledge of the anatomical location of steroid receptor
mRNA containing cells within the CNS of the rat brain (Simerly, Chang, Muramatsu, &
Swanson, 1990). However, the use of monoclonal and polyclonal antibodies directed
against steroid receptors has greatly simpliﬁed the detection, of steroid receptor
containing cells within the brain, and has been utilized in a number of species
(DonCarlos, Monroy, & Morrell, 1991; Menard & Harlan, 1993; Wood & Newman,
1993a). Irnmunocytochemical techniques have also been utilized to determine the
neurotransmitter / neuropeptide content of steroid receptor containing cells (Harlan, 1988
2108), and the activation of steroid receptor-immunoreactive cells during steroid-
dependent behaviors (Wood & Newman, 1993b).

Within the brains of European ferrets, estrogen receptor-containing cells in
discrete regions have been described using autoradiographic histochemistry (Sisk, 1993;
Tobet & Baum, 1991), and immunocytochemistry (Sisk & DonCarlos, 1995; Tobet,

Basham, & Baum, 1993). However, only androgen receptor binding techniques in

101

microdissected brain regions has been used to describe the distribution of androgen
receptor containing cells in the ferret brain (Vito, et al., 1985) and thus a description of
the distribution of androgen receptor-containing cells with cellular resolution in the ferret
does not exist. This experiment was undertaken to describe the distribution of androgen

receptor-immunoreactive cells in telencephalic and diencephalic structures of the male

ferret brain.

Methods

Construction of AR-IR Cell Maps

The brain of an adult male ferret that was castrated and treated daily for 10 days
with an injection of 5 mg/kg testosterone propionate was used to construct the atlas of
AR-IR cells within the forebrain and hypothalamus. Work described in this dissertation
clearly demonstrates that treatment with T maximizes the number of androgen receptor-
immunocytochemically labeled cells, and the intensity of their staining. Comparisons
between T treated castrated male ferrets and intact animals at the microscopic level
indicated that intact animals had the same distribution of AR-IR cells than did T-treated
animals, but the intensity of immunocytochemical staining was reduced in intact animals.
Since AR-IR cells can be detected by microscopic examination in intact animals within
all brain regions stained following T treatment, this map generated from a T-treated adult
is a good representation of the distribution of AR-IR cells in the male ferret brain. All
AR-IR cells visible at a magniﬁcation of 70x were plotted using a projection scope
(Bausch & Lomb). These plots were scanned into a Macintosh computer using a

Silverscan II scanner and Adobe Photoshop software at a resolution of 300 dots per inch.

102

Individual scans were then converted to Canvas ﬁles, reconstructed, and reduced to a
magniﬁcation of 15x. Two adjacent sets of tissue from the mapped brain were also
utilized in the construction of these maps. One set was thionin stained, while the other
was stained for AR-IR cells and counterstained with thionin. These maps do not
accurately reﬂect the regional variation in staining intensity, therefore these differences
are described in the text.
Anatomical Nomenclature

The distribution of AR-IR cells in the ferret brain is essentially a continuum of
cells running throughout the hypothalamus, septum, and amygdala. AR-TR cells do fall
within speciﬁc nuclei, however they do not remain within these cytoarchitectonically
deﬁned regions. Thus the boundaries of classically deﬁned brain nuclei are not included
in these maps, however their location was identiﬁed by comparison with previous work
in the ferret (Lockard, 1982; Lockard, 1985; Westwood, 1962), and from previous work
describing the basal forebrain and hypothalamus of the cat (Berrnan & Jones, 1982;
Bleier, 1961). Terminology used for the current experiment is based largely on that
adopted by Paxinos and Watson (1986)(Paxinos & Watson, 1986).

Results

General Description

Figures 25-34 show the distribution of plotted AR-IR cells, visible through a
projection scope at a magniﬁcation of 70x, from a point just anterior to the OVLT
through the mammillary bodies. The most dense aggregations of AR-IR cells were
located in the preoptic area, hypothalamus, septum and amygdala. Indeed there was a

continuum of AR-IR cells which extended uninterrupted along the midline from a point

103

Section 39-4-1

 

Figures 25-34. Maps of the distribution of androgen receptor-immunoreactive
cells in the forebrain of an adult male ferret castrated and injected daily with 5.0
mg/kg testosterone for 10 days. Each dot represents one AR-IR cell visible at a
maginiﬁcation of 70 x.

104

Section 39-4-3

 

1mm

Figure 26

 

105

Section 39-5-2

 

 

 

Pir

 

Figure 27

 

Section 39-6—2

Pir

Figure 28

Ca

 

__.-.,'..;.__~ ll $03

106

BNST '

VP

       

I.. I.
‘3.’r

- I'
. ‘I _..-r
u use. I‘ a» ~'- :.

1.0 mm
l—————l

107

Section 39-7-1

Ca

3V

 

'21: “5‘, ‘4 -: iii-33"“ 7:3 3'19. AHA

Pir

 

Figure 29

108

Section 39-8-2

 

 

a
C

7
~95 .

airyﬂ
1' .

ic

 

r-
P

1.0 mm

 

Figure 30

109

Section 39-9-2

CC

 

Pir

. -~
1‘!

ACO

‘ I "
a 13'}? if; ;. '9. ; .., ‘ 2’ 'U;‘
. ,7 ,& g ’ t, .J.‘ .! . ,'. I
a“ ' " u
0‘: t L f} U
u M173] '5 '

Figure 31

Section 39-10-2

Figure 32

110

Ca t

v ..
.‘l
'3
:\ '
"\
PV‘I
l' ' . . 5:"

 

111

Section 39-11-3

 

Figure 33

112

Section 39-12-3

 

1.0 mm

“1
r...
7.
‘\
....
a
'
.J..\u
“u A... r
9......» 4
.. Q) a
a: .... c.
... l .31»...
1:. «M. 1
. ..
.. «an. - . .
. .
‘71. .
a .
. .. a; . 1 pa
. ,.... .
\. ..., . .
J. _. .
u .l n
.. ... .v
.. . Jr
. .
..

 

Figure 34

113

rostral to the OVLT through the mammillary bodies, and in lateral regions from the
olfactory tubercle and the nucleus of the diagonal band, caudally through the arnygdaloid
complex and amygdala-hippocampal transition area. AR-IR cells were located
throughout the lateral septum, but were not prevalent in the medial septum. Lightly
stained cells, detectable under microscopic examination but not visible as individual cells
under the projection scope, were found throughout the piriform cortex primarily in
lamina II and III, but also a few scattered cells in the molecular layer. Additional lightly
stained cells were visible in the cingulate cortex. Ependymal cells lining the rostral-
caudal extent of the third ventricle, as well as most portions of the lateral ventricle were
heavily stained for the AR protein. AR-IR cells were also present in the
circumventricular organs examined including the median eminence, OVLT, SF0 and
choroid plexus (Figure 35).
Telencephalon
Noncortical Nuclear Groups: The olfactory tubercle (Tu) is bounded medially

by the nucleus of the diagonal band (HDB), and laterally by the lateral olfactory tract
(10). The Tu contains a dense collection of light to moderate staining AR-IR cells within
the pyramidal and polymorphic cell layers. The anterior portions of the Tu are in close
association with the Islands of Calleja which are composed of tightly packed granule
cells containing a small proportion of lightly stained AR-IR cells. The lateral septal
nuclei, including the dorsal, intermediate and ventral subdivisions all contain AR-IR
cells. These cells are lightly stained in sections rostral to the anterior commissure, but
are stained moderate to dark proceeding caudally to the ﬁmbria of the fomix.

Moderately stained AR-IR cells are located in the subfomical organ (Figure 35). Dorsal

 

Figure 35. Photomicrographs of AR-IR cells within circumventricular organs of
the male ferret brain. (A), subfomical organ; (B), median eminence; (C),
vascular organ of the lamina terminalis; and (D), choroid plexus. Magnification
bar=100pm(A,C)and30pm(B ,D).

115

and medial to the ventral subdivision of the LS and rostral to the anterior commissure,
AR-IR cells are found in the septohypothalamic nucleus (SHy). Cells in SHy are lightly
stained and scattered. The darkest staining cells in the LS are in the ventral subdivision.
A few lightly stained AR-IR cells are located in the medial septum, primarily where it is
in continuation with the vertical limb of the nucleus of the diagonal band. Both nuclei
associated with the diagonal band (horizontal and vertical limbs) contain a large number
of AR-IR containing cells . These cells are lightly stained and are more diffuse than cells
in the nearby MPA.

Cells stained for AR in the bed nucleus of the stria terminalis were very light
throughout its rostro-caudal extent. Cells were more numerous, and slightly darker in the
anterior medial and posterior medial portions of this nucleus than in lateral components,
though still lighter than AR-IR cells in the surrounding structures such as the POA,
AHA, and LS.

Several subdivisions of the arnygdaloid complex contained AR-IR cells. A few
scattered lightly stained AR-IR cells were found in the nucleus of the lateral olfactory
tract. The highest concentration of AR-IR cells was seen in the anterior, medial, and
cortical arnygdaloid nuclei (Figure 36). These cells form a sheet stretching from the
anterior arnygdaloid area through the posterior subdivision of the medial amygdaloid
nucleus. These cells were light to moderately stained and were tightly packed. AR-IR
cells were also seen in the rostral-caudal extent of the cortical nucleus, being more dense
and intensely stained in the posterior region. A few lightly stained AR-IR cells were
detected in the central nucleus, the lateral nucleus in posterior regions, as well as the
basomedial nucleus. AR-IR cells were detected in the amygdalo-hippocampal transition

area, as well as the amygdalo-piriform transition area.

 

Figure 36. Photomicrographs of AR-IR cells within the anterior (A), medial and
cortical (B) arnygdaloid nuclei of an adult male ferret. Magniﬁcation bar = 250

pm.

117

Preoptic area: Within the preoptic area there was a dense collection of intensely
stained AR-IR cells (Figure 37). These cells are very tightly packed adjacent to the 3V,
in the periventricular nucleus of the preoptic area, and became less tightly packed in a
lateral direction through the MPN into the lateral preoptic area. As described for the
distribution of estrogen receptor-immunoreactive cells in the Guinea Pig (DonCarlos, et
al., 1991), AR-IR cells in the ferret MPA formed columns running parallel with the 3V
(Figure 37C). AR-IR cells were numerous in the MnPO (Figure 37A), a group of tightly
packed small cells located on the dorsal aspect of the 3V beginning at the level of the
OVLT and extending to the anterior commissure (Bleier & Byne, 1985). There were
also numerous AR-IR cells within the OVLT (Figure 35C), and a large number of AR-IR
cells running along the edges of the lamina terminalis itself.
Diencephalon

Hypothalamus: There is no clear demarcation between the POA and the AHA,
thus it has been arbitrarily set at the rostral edge of the anterior commissure. Like the
POA, AR-IR cells in the AHA are more tightly packed medially near the 3V and become
less dense laterally into the LHA (Figure 38A). Within the AHA there are a number of
nuclei which contain AR-IR cells including the male nucleus of the AH/POA as deﬁned
by Tobet 1986, the suprachiasmatic nucleus (SCh), and the paraventricular nucleus of the
hypothalamus (PVH). Moderately stained, AR-IR cells are also found within the MN-
POA/AH, a sexually dimorphic nucleus within the AH (Figure 39). Cells within the non-
sexually dimorphic ventral nucleus contain fewer AR labelled cells compared to the MN-

POA/AH. The SCh in the ferret is a prominent nucleus located in the ventral AHA

118

 

Figure 37. Photomicrographs of AR-IR cells within the preoptic area of the male
ferret. Magniﬁcation bar = 250 pm (A and B), and 100 pm (C).

 

 

Figure 38. Photomicrographs of AR-IR cells within the anterior hypothalamus
(A), retrochiasmatic area (B), ventromedial hypothalamic nucleus and arcuate
nucleus (C and D) of an adult male ferret. Magniﬁcation bar = 250 pm.

 

Figure 39. Photomicrographs of the male nucleus of the preoptic area/ anterior
hypothalamus of an adult male ferret (upper arrow panel A). Ventral nucleus is
indicated by lower arrow. A and C are nissl stained sections adjacent to the AR-IR
stained sections in B and D. Magniﬁcation bar = 250 pm (A and B), 100 pm (C
and D).

121
dorsal to the optic chiasm and adjacent to the 3V (Figure 38A). The SCh contains small,
round, and tightly packed cells; AR-IR cells are distributed homogeneously within this
nucleus. Within the paraventricular nucleus of the hypothalamus AR immunostaining
was detected in both magnocellular and parvocellular neurons throughout its rostro-
caudal extent.

In the area of the tuber cinereum, AR-IR cells are located within the
retrochiasmatic area (RCh), ventromedial nucleus of the hypothalamus (VMH), the
arcuate nucleus (ARC) and median eminence (ME), as well as the dorsomedial nucleus
(DM) (Figure 38 B,C,D)). AR-IR cells in the RCh are moderately stained for AR, and
extend ﬁ'om the formation of the optic tracts to the ME. Cells in the RCh are less dense
than the SCh rostrally and the ARC caudally. The ventromedial nucleus contains a high
density of heavily labeled cells which form an oval. AR-IR cells within the interior of
the VMH are less dense than the lateral component (Figure 38 CD). The lateral portion
of the VMH extends more dorsal, and contains darker staining cells than the medial
component. The dorsomedial nucleus of the hypothalamus forms between the
descending column of the fornix and the PVH. Cells in the DM are small and extend in a
ventromedial direction merging with the periventricular nucleus anteriorly and the
arcuate nucleus posteriorly (Figure 40). AR-IR cells in the DM are lightly stained and
less dense than the adjacent VMH and ARC. The ARC contains moderately stained AR-
IR cells which are of approximately the same density, but not as darkly stained as AR-IR
cells in the VMH. AR-IR cells become more tightly packed proceeding caudally to its
terminus in the posterior hypothalamus rostral to the mammillary nuclei. Several AR-IR

cells are located within the median eminence, primarily in the interior zone (Figure 35).

122

 

Figure 40. Photomicrographs of AR-IR cells in (A) the caudal portion of the
arcuate nucleus and ventral premammillary nucleus, and (B), in the medial, and
lateral mammillary nuclei, and the posterior hypothalamus. Magniﬁcation bar
=250 um. .

123
However, background staining in the ME was too dark for cells to be detected under the
projection scope and are thus not illustrated in the illustrations.

Caudal to the VMH, the ventral premammillary nucleus forms in the ventral-
lateral portion of the tuber cinereum (Figure 40A). This nucleus contains some of the
darkest staining AR-IR cells within the brain. The remainder of the posterior
hypothalamus consists of the nuclei of the mammillary bodies (Figure 40B). The lateral
and medial mammillary nuclei contain a moderate density of light to moderate stained
AR-IR cells, however in the supramammillary nucleus there is are considerably fewer
AR-IR cells than the surrounding LM and MM ventrally or the posterior hypothalamic
area dorsally between the marnmillothalarnic tract and the 3V.

Thalamus: The most notable nuclear conﬁgurations within the thalamus
containing AR-IR cells were the paraventricular thalamic nucleus, and the medioventral
nucleus. The paraventricular nucleus contains lightly stained AR-IR cells which border
the dorsal aspect of the 3V. Cells in the MV are more numerous but lightly stained.
Other thalamic nuclei containing AR-IR cells were the paracentral nucleus and the lateral
dorsal nucleus.

Comparisons between Intact and T-Treated Adult and Prepubertal Ferrets

Examination of intact and T treated adult and prepubertal males at the light
microscope level indicated a number of differences in the distribution of AR-IR cells.
These differences reﬂect what has been demonstrated in the above experiments, that
animals with lower T levels have lighter staining AR-IR cells in many regions than do
animals with higher T levels. This effect is most prominent in those regions that

contained lightly stained AR-IR cells such as the piriform cortex, BNST, SHy, and the

124
nuclei within the thalamus. AR-IR cells could be detected in intact adults in all regions
in which they were mapped in T treated adults, and prepubertal animals treated with T
were identical in all respects to adult animals treated with T. However, prepubertal
animals often lacked any detectable AR-IR cells in these lightly stained regions when
observed under lower magniﬁcations, and high magniﬁcation optics were often required
to detect any AR-IR cells in these regions. In those regions classically thought of as
containing steroid receptors, such as the POA, hypothalamus, amygdala, and septum,
AR-IR cells were readily detectable in prepubertal intact animals. In no region were
there AR-IR cells detectable in prepubertal intact animals that were not also detectable in

adults.

Discussion

Autoradiographic techniques have demonstrated the presence of steroid receptor
containing cells from every major class of vertebrate (Kim, Stumpf, Sar, &
Martinez-Vargas, 1978; Morrell & Pfaff, 1978). There are many common regions
across species that accumulate gonadal steroids including the preoptic area, septum,
hypothalamus and amygdala (Kim, et al., 1978; Morrell & Pfaff, 1978). Furthermore,
considerable overlap exists regarding the distribution of androgen and estrogen receptors
across species and across techniques, including autoradiography, immunocytochemistry,
and in situ hybridization (see (Blaustein & Olster, 1989; Simerly, et al., 1990)).
Irnmunocytochemical experiments that have described the distribution of AR in
mammals and birds have revealed a similar pattern within this collection of neural

structures (Balthazart, Foidart, Wilson, & Ball, 1992; Choate & Resko, 1992). The

125
present study adds to the comparative literature by describing the distribution of
androgen receptor immunoreactive cells in the European ferret. Not surprisingly, the
greatest concentration of AR-IR cells was detected in the within the preoptic area,
hypothalamus, septum, and amygdala. This highly conserved distribution of steroid
receptors across vertebrate species reﬂects many of the common neuroendocrine and
behavioral effects mediated by gonadal steroids across species.

The neural circuits for the activation of male sexual behavior in rodents have
been well characterized (Meisel & Sachs, 1994). The display of reproductive behavior
requires activation of this circuitry by both chemosensory stimuli and steroid hormones,
and the information encoded by these exogenous and endogenous cues is integrated
within components of the circuit. Chemosensory input is obtained during ano-genital
investigation of the female prior to copulatory behavior and is transmitted through the
main and accessory olfactory systems. This information travels along the lateral
olfactory tract to the cortical and medial amygdala (Davis, Macrides, Youngs, Schneider,
& Rosene, 1978; Scalia & Winans, 1975). The amygdala then sends olfactory
information to the mPOA (Kevetter & Winans, 1981) as well as to the BNST (Lehman &
Winans, 1983 4800), which in turn also projects to the mPOA (Chiba & Murata, 1985).
Reproductive behavior can be blocked by disrupting either the hormonal or
chemosensory input to this neural circuitry, or by disruptions to the circuitry itself. For
example, either elimination of circulating gonadal steroids (Beach & Pauker, 1949;
Whalen & DeBold, 1974), or blockage of the nares, which eliminates stimulation of the
olfactory bulbs, (Doty, Carter, & Clemens, 1971; Murphy & Schneider, 1969) results in a
decline in reproductive behavior. Reproductive behavior is also compromised by lesions

in brain regions, such as the POA, BNST and amygdala, which are part of the steroid

126
sensitive neural circuitry, or by lesions along the olfactory pathway that supplies
chemosensory information to these regions (W inans & Powers, 1974). Although the
olfactory bulbs were not analyzed in the present study, AR-IR cells in the male ferret
brain were found within all other components of the neural circuit mediating
reproductive behavior. Interestingly, relative to what has been described for the rodent
brain (Clancy, et al., 1992; Menard & Harlan, 1993; Sar, et al., 1990; Wood & Newman,
1993a), the male ferret brain contains considerably fewer AR-IR cells within the nuclei
of the BNST. Unfortunately the contribution of the BNST to reproductive behavior in
the male ferret is unknown and thus no conclusion can be drawn regarding the functional
signiﬁcance of this species difference.

Sexual differentiation of reproductive behavior patterns is a product of the steroid
environment during speciﬁc stages of development (Baum, et al., 1990). In the male
ferret, as in other species (Arnold & Gorski, 1984; Meisel & Sachs, 1994), there is a
morphological sexual dimorphism in the preoptic area-anterior hypothalamic area
continuum (T obet, et al., 1986b). In this region there is a nucleus present only in male
ferrets, the dorsal nucleus of the POA/AH (also called the male nucleus (MN) of the
POA/AH. The formation of this nucleus in the male is dependent upon exposure to T,
and aromatization to E within the brain, during a critical period of development (Cherry,
Basham, Weaver, Krohmer, & Baum, 1990; Tobet, Zahniser, & Baum, 1986a).
However, the increased size of the cells within this nucleus, is reduced in castrated
adults, but can be restored by treatment with either E , T or DHT (T obet, et al., 1986b).
Thus, androgens, acting via the androgen receptor, can directly and indirectly inﬂuence
cells within the MN-POA/AH. Small lesions restricted to this nucleus in the ferret

produce small deﬁcits in male reproductive behavior, however animals were still capable

127
of achieving intromissions (Cherry & Baum, 1990). Interestingly, when animals were
castrated and treated with estradiol benzoate, the time it took to approach a stud male in
an L-maze was lower than even E treated females suggesting that this region of the brain
may inhibit female behaviors from occurring in males. (Cherry & Baum, 1990).
Estrogen receptors in this nucleus has been identiﬁed using autoradiography (Tobet &
Baum, 1991). The presence of AR-IR cells in the MN-POA/AH of the male ferret, as
shown in the present experiment, is consistent with the regulatory role of both androgens
and estrogens in the modulation of cell size within this nucleus, and in the inhibition of
the occurrence of female reproductive behaviors. The ultimate action of T in this nucleus
may thus be directly on the AR, or through estrogen and the estrogen receptor following
androgen dependent aromatization to estrogen (Weaver & Baum, 1991), in either case
the androgen receptor is playing a critical role.

In summary, the ﬁndings of this experiment are in agreement with those of others
describing the distribution of androgen receptors within the forebrain of vertebrate
species. Several techniques have been utilized to locate putative steroid receptor
containing cells within the brain. The use of immunocytochemistry for steroid receptors
has advantages over others with regard to simplicity of procedures, and the ability to
detect multiple antigens within individual cells. Future use of double-label techniques
for steroid receptors and other neurochemicals will undoubtedly lead to an increase in

our understanding of the many actions of steroid hormones on brain function.

Appendix B

Effects of Long-term Storage in Cryoprotectant on the Staining of AR-IR Cells

Since separate sets of tissue from prepubertal animals under various steroid
treatment conditions were processed immunocytochemically and quantiﬁed for AR-IR
cells in Experiments HI and IV, statistical analyses were performed to compare the
replicability of these results across immunocytochemical runs. Only data from the short
incubation in DAB from Experiment III was analyzed so that the comparisons would not
be confounded by this experimental manipulation. A two-way ANOVA using brain
region and experiment as the independent variables was performed. There was not a
signiﬁcant main effect of Experiment [FL332=1.106, p=0.29] indicating that there was not
a signiﬁcant loss of antigen due to long-term storage in cryoprotectant at -20°C. There
was however a signiﬁcant interaction between experiment and brain region. Post hoc
analysis indicated that signiﬁcantly more cells were detected in the lVMH and the
aBNST in Experiment 111 than were detected in Experiment IV. There is no obvious
reason for the differences between the two experiments regarding the density of AR-IR
cells in these regions. In the lVMH, cell counts for Experiment III were performed by a
different experimenter than in Experiment IV. However, cell counts in the BNST were
made by the same experimenter in both studies. Further analyses were conducted in
order to determine if our ﬁnding of brain region speciﬁc increases in AR-IR cells
following androgen treatment to castrated animals, relative to intact animals, was

replicated in Experiment IV since the statistical analysis in Experiment IV did not call

128

129
for these comparisons to be made. One-way AN OVA‘s for each of the brain regions that
were analyzed in both experiments were performed. Pair-wise comparisons within each
brain region were made with the Fishers PLSD. In the ARC, MPA, and MeA these
comparisons indicated that a signiﬁcant increase in the density of AR-IR cells relative to
intact animals occurred following androgen treatment in Experiment IV as was found in
Experiment 111, however in the lVMH, and pvPOA these results were only partially
replicated in that in the pvPOA only DHT treatment resulted in a signiﬁcant increase in
AR-IR cell density relative to intact animals. In the lVMH, an area in which no
signiﬁcant up-regulation of AR-IR cells was found in our previous experiment, T
treatment again failed to statistically increase AR-IR cell number, however DHT
treatment did. In the LS, and the BNST, areas in which androgen treatment did not yield
an increase in the density of AR-IR cells relative to intact animals in Experiment III, the
results were replicated in that there were no differences between intact animals and T or
DHT treated animals in Experiment IV. Thus, although the density of cells in every
brain region across the two experiments were not all replicated in terms of absolute

numbers, the relative differences among groups was consistent.

LIST OF REFERENCES

LIST OF REFERENCES
Arnold, A. P., & Gorski, R. A. (1984). Gonadal steroid induction of structural sex

differences in the nervous system. Wm, 2, 413—442.
Balthazart, J., Foidart, A., Wilson, E. M., & Ball, G. F. (1992). Irnmunocytochemical
localization of androgen receptors in the male songbird and quail brain. The
laumaleLCemnaratixeNemlesx, 3.11, 407-420.
Baum, M. J. (1976). Effects of testosterone propionate administered perinatally on sexual

behavior of female ferrets.

 

EM), 399-410.

Baum, M. J ., Carroll, R. S., Cherry, J. A., & Tobet, S. A. (1990). Steroidal control of
behavioral, neuroendocrine and brain sexual differentiation: Studies in a
carnivore, the ferret. Wm, 2(4), 401-418.

Baum, M. J ., & Everitt, B. J. (1992). Increased expression of c-fos in the medial preoptic
area after mating in male rats: Role of afferent inputs from the medial amygdala
and midbrain central tegrnental ﬁeld. W, 5_Q(3), 627-646.

Baum, M. J ., & Vreeburg, J. T. M. (1973). Copulation in castrated male rats following
combined treatment with estradiol and dihydrotestosterone. Science, 182,
283-285.

Baum, M. J., & Vreeburg, J. T. M. (1983). Copulation in castrated male rats following
combined treatment with estradiol and dihydrotestosterone. Science, L82,

283-285.

130

131

Beach, F. A. (1942). Sexual behavior of prepuberal male and female rats treated with

gonadal hormones. W. 3.4., 285-292.
Beach, F. A., & Pauker, R. S. (1949). Effects of castration and subsequent androgen

administration upon mating behavior in the male hamster (cricetus auratus).

Endocrinology, 9.5, 211-221.
Beato, M. (1989). Gene regulation by steroid hormones. Cell, 16, 335-344.

Berglund, L. A., & Sisk, C. L. (1990). Pituitary responsiveness to luteinizing
hormone-releasing hormone in prepubertal and postpubertal male ferrets. Biology
ofRenLerrstion, $3, 335-339.

 

Wisconsin Press.

Bittrnan, E. L., & Krey, L. C. (1988). inﬂuences of photoperiod on nuclear androgen
receptor occupancy in neuroendocrine tissues of the golden hamster.
Neumendosrinolosx, 12, 61-67.

Blaustein, J. D., & Olster, D. H. (1989). Gonadal steroid hormone receptors and social
behaviors. In J. Balthazart (Eds), ‘
Physiology (pp. 31-104). Berlin: Springer-Verlag.

Bleier, R. (1961). I: -

WW. Baltimore: Johns Hopkins Press.
Bleier, R., & Byne, W. (1985). Septum and hypothalamus. In G. Paxinos (Eds), mm

W (pp. 87-118). Sydney: Academic Press.

 

 

132

Blok, L. J., Themmen, A. P. N., Peters, A. H. F. M., Trapman, J., Baarends, W. M.,
Hoogerbrugge, J. W., & Grootegoed, J. A. (1992). Transcriptional regulation of
androgen receptor gene expression in Sertoli cells and other cell types. Moleoolar;
andL‘sllrrlaLEndogrinologx, 8.8, 153-164.

Blondeau, J. P., Corpéchot, C., Le Goascogne, C., Baulieu, E. E., & Robel, P. (1975).
Androgen receptors in the rat ventral prostate and their hormonal control.
mm, 3.3, 319-344-

Brown, T. J., Hochberg, R. B., Naftolin, F ., & MacLusky, N. J. (1994). Pubertal
development of estrogen receptors in the rat brain. WW

Neuroscience, 5, 475-483.

Burgess, L. H., & Handa, R. J. (1993). Hormonal Regulation of Androgen Receptor
Messenger RNA in the Brain and Anterior Pituitary Gland of the Male Rat.
WW, 120-2), 31-38-

Callard, G. V., Mak, P., & Solomon, D. J. (1986). Effects of short days on aromatization
and accumulation of nuclear estrogen receptors in the hamster brain. BiologLQf
Ropzoouotion, 35, 282-291.

Carroll, R. S., & Baum, M. J. (1989). Evidence that oestrogen exerts an equivalent
negative feedback secretion in male and female ferrets. 1mm, 8o,

23 5-245.

Carroll, R. S., Weaver, C. E., & Baum, M. J. (1988). Evidence implicating aromatization

of testosterone in the regulation of male ferret sexual behavior. mm

mm. 512, 457-460-

133

Cherry, J. A., Basham, M. E., Weaver, C. E., Krohmer, R. W., & Baum, M. J. (1990).
Ontogeny of the sexually dimorphic male nucleus in the preoptic/anterior
hypothalamus of ferrets and its manipulation by gonadal steroids. Journalof
Neurobiology, 21(6), 844-857-

Cherry, J. A., & Baum, M. J. (1990). Effects of lesions of a sexually dimorphic nucleus
in the preoptic/anterior hypothalamic area on the expression of androgen- and
estrogen-dependent sexual behaviors in male ferrets. BMW, 522,
191-203.

Cherry, J. A., Tobet, S. A., DeVoogd, T. J., & Baum, M. J. (1992). Effects of sex and
androgen treatment on dendritic dimensions of neurons in the sexually dimorphic
preoptic/anterior hypothalamic area of male and female ferrets. Iholoomalof
WM, 323, 577-585.

Cheung, C. Y., & Davidson, J. M. (1977). Effects of testosterone implants and
hypothalamic lesions on luteinizing hormone regulation in the castrated male rat.
Endocrinology, 1m, 292-302.

Chiba, T., & Murata, Y. (1985). Afferent and efferent connections of the medial preoptic
area in the rat: a WGA-I-IRP study. W, 15, 261-272.

Choate, J. V. A., & Resko, J. A. (1992). Androgen Receptor Irnmunoreactivity in Intact
and Castrate Guinea Pig Using Antipeptide Antibodies. Moss-mob, 521(1),
5 1-59.

Clancy, A. N., Bonsall, R. W., & Michael, R. P. (1992). Irnmunocytochemical labeling

of androgen receptors in the brain of rat and monkey. Mom, 5_Q, 409-417.

134

Clancy, A. N., Whitman, C., Michael, R. P., & Albers, H. E. (1994). Distribution of
Androgen Receptor-Like Irnmunoreactivity in the Brains of Intact and Castrated
Male Hamsters. W, 33(3), 325—332.

Connolly, P. B., Roselli, C. E., & Resko, J. A. (1990). Aromatase activity in adult guinea
pig brain is androgen dependent. Bjolomﬁﬁooroooglon, 513, 698-703.

Davis, B. J ., Macrides, F., Youngs, W. M., Schneider, S. P., & Rosene, D. L. (1978).
Efferents and centrifugal afferents of the main and accessory olfactory bulbs in
the hamster. MW 3, 59-72.

DonCarlos, L. L., Monroy, E., & Morrell, J. I. (1991). Distribution of estrogen
receptor-immunoreactive cells in the forebrain of the female guinea pig. Iho
MW, 325, 591-612.

Doman, W. A., & Malsbury, C. W. (1990). Neuropeptides and male sexual behavior.
WW 13(1), 1-15.

Doty, R. L., Carter, C. S., & Clemens, L. G. (1971). Olfactory control of sexual behavior
in the male and early-androgenized female hamster. HonnonmnsLBohaxior, 2,
325-335.

Ebling, F. J. P., & Foster, D. L. (1989). Pineal melatonin rhythms and the timing of
puberty in mammals. Exmriontia, 45, 946-954.

Ellinwood, W. E., Hess, D. L., Roselli, C. E., Spies, H. G., & Resko, J. A. (1984).

Inhibition of aromatization stimulates luteinizing hormone and testosterone

secretion in adult male rhesus monkeys. Journal of Clinical Endocn'nolog}: and
Metabolism. 59(6), 1088-1096.

135

Evans, R. M. (1988). The steroid and thyroid receptor superfamily. Soionoo, 239,
889-895.

Feder, H. H., Naftolin, F., & Ryan, K. J. (1974). Male and female sexual responses in
male rats given estradiol benzoate and 5a—androstan—17B—ol-3-one propionate.
Endocrinology, 24, 136—141.

Gladue, B. A., & Clemens, L. G. (1980). Flutarnide inhibits testosterone-induced
masculine sexual behavior in male and female rats. EMnology, M,

191 7-1922.

Gonzalez-Cadavid, N. F ., Vernet, D., Navarro, A. F ., Rodriguez, J. A., Swerdloff, R. S.,
& Rajfer, J. (1993). Up-Regulation of the Levels of Androgen Receptor and Its
messenger RNA by Androgens in Smooth-Muscle Cells ﬁ'om Rat Penis.
Moleculaundﬁellularjndminolm, 29(2), 219-229-

Gray, G. D. (1977). Differential effects of the antiandrogen ﬂutamide on aspects of
sexual behavior in castrated, androgen-treated male rats.
BMW, 2, 315-320.

Grino, P. B., Grifﬁn, J. E., & Wilson, J. D. (1990). Testosterone at high concentrations
interacts with the human androgen receptor similarly to dihydrotestosterone.
Endocrinology, 12o, 1165-1172.

Handa, R. J., Roselli, C. E., Horton, L., & Resko, J. A. (1987a). The quantitative
distribution of cytosolic androgen receptors in microdissected areas of the male
rat brain: Effects of estrogen treatment. Endocrinology, 121, 233-240.

Handa, R. J ., Stadelman, H. L., & Resko, J. A. (1987b). Effect of estrogen on androgen

receptor dynamics in female rat pituitary. ﬁdmdnology, 121, 84-89.

136

Harlan, R. E. (1988). Regulation of neuropeptide gene expression by steroid hormones.

MoleculaLNeurohiolosx. 2, 183-200.

Johnston, R. E., & Coplin, B. (1979). Development of responses to vaginal secretion and
other substances in golden hamsters. Won, 25,
473-489. .

Kalra, P. S., & Kalra, S. P. (1980). Modulation of hypothalamic luteinizing
honnone-releasing hormone levels by intracranial and subcutaneous implants of
gonadal steroids in castrated rats: Effects of androgen and estrogen antagonists.
Endocrinology, .1116, 390-397.

Kashon, M. L., Hayes, M. J., Shek, P. P., & Sisk, C. L. (1995). Regulation of brain
androgen receptor immunoreactivity bu androgen in prepubertal male ferrets.
W42, 1198-1205.

Kashon, M. L., & Sisk, C. L. (1994). Pubertal maturation is associated with an increase
in the number of androgen receptor-immunoreactive cells in the brains of male
ferrets. W, 18, 237-242.

Kemppainen, J. A., Lane, M. V., Sar, M., & Wilson, E. M. (1992). Androgen receptor
phosphorylation, turnover, nuclear transport, and transcriptional activation.
Speciﬁcity for steroids and antihormones. WWW,
291(2), 968-974. ‘

Keppel, G. (1982). Wham. Englewood Cliffs, New Jersey: Prentice Hall.

Kevetter, G. A., & Winans, S. S. (1981). Connections of the corticomedial amygdala in

the golden hamster. H. Efferents of the 'olfactory amygdala‘. Loomﬂof

MW, 122, 99-11 1.

137
Kim, Y. S., Stumpf, W. E., Sar, M., & Martinez-Vargas, M. C. (1978). Estrogen and
androgen target cells in the brain of ﬁshes, reptiles and birds: phylogeny and

ontogeny. We. rs. 425-433.
King, J. C., & Rubin, B. S. (1994). Dynamic changes in LHRH neurovascular terminals

with various endocrine conditions in adults. Homomandﬂohaxior, 28(4),
349-3 56.

Koch, M. (1990). Effects of treatment with estradiol and parental experience on the
number and distribution of estrogen-binding neurons in the ovariectomized mouse
brain. Nouroondoorinology, 51, 505-514.

Kondo, Y., Shinoda, A., Yamanouchi, K., & Arai, Y. (1990). Role of septum and
preoptic area in regulating masculine and feminine sexual behavior in male rats.
Homenesaniﬁehaxior, 2A, 421-434.

Krey, L. C., Lieberburg, 1., Roy, E., & McEwen, B. S. (1979). Oestradiol plus receptor
complexes in the brain and anterior pituitary gland: quantitation and
neuroendocrine signiﬁcance. 1W, 11, 279-284.

Krey, L. C., & McGinnis, M. Y. (1990). Time-courses of the appearance/disappearance
of nuclear androgen+receptor complexes in the brain and adenohypophysis
following testosterone administration/withdrawal to castrated male rats:
Relationships with gonadotropin secretion. MW, 35,
403-408.

Krongrad, A., Wilson, C. M., Wilson, J. D., Allman, D. R., & McPhaul, M. J. (1991).

Androgen increases androgen receptor protein while decreasing receptor mRNA

in LNCaP cells. MolesulaunslﬁelhrlaLEndocrinolm, 19, 79-88.

138

Langub, J., M.C., & Watson, J., RE. (1992). Estrogen receptor-immunoreactive glia,
endothelia, and ependyma in guinea pig preoptic area and median eminence:
Electron microscopy. Endocﬂnology, ]_3_Q(1), 364-372.

Larrson, K., Perez-Palacios, G., Morali, G., & Beyer, C. (1975). Effects of
dihydrotestosterone and estradiol pretreatment upon testosterone-induced sexual
behavior in the castrated male rat. WW, 6, 1-8.

Lauber, A. H., Mobbs, C. V., Muramatsu, M., & Pfaff, D. W. (1991). Estrogen receptor
messenger RNA expression in rat hypothalamus as a function of genetic sex and
estrogen dose. Endmrinology, 122, 3180-3186.

Lauber, A. H., Romano, G. J., Mobbs, C. V., & Pfaff, D. W. (1990). Estradiol regulation
of estrogen receptor messenger ribonucleic acid in rat mediobasal hypothalamus:
An in situ hybridization study. WW, 2(5), 605-611.

Lehman, M. N., & Winans, S. S. (1983). Evidence for a ventral non-strial pathway from
the amygdala to the bed nucleus of the stria terminalis in the male golden
hamster. W, 268, 139-146.

Liao, S., Howell, D. K., & Chang, T. M. (1974). Action of a nonsteroidal antiandrogen,
ﬂutamide, on the receptor binding and nuclear retention of
5-a-dihydrotestosterone in rat ventral prostate. Endmrinology, %, 1205-1209.

Lindzey, J ., Kumar, M. V., Grossman, M., Young, C., & Tindall, D. J. (1994). Molecular
mechanisms of androgen action. W, 42, 383-432.

Lockard, B. I. (1982). Telencephalon of the Ferret (Mustela F are). In E. C. Crosby & H.

N. Schnitzlein (Eds),

Iolonooohalon (pp. 484-502). New York: Macmillian.

 

139

Lockard, B. I. (1985). The forebrain of the ferret. W, 35,
216-228.

Lusting, R. H., Mobbs, C. V., Bradlow, H. L., McEwen, B. S., & Pfaff, D. W. (1989).
Differential effects of estradiol and 16-alpha-hydroxyestrone on pituitary and
preoptic estrogen receptor regulation. Endooﬁnology, 125, 2701-2709.

McEwen, B. S. (1976). Steroid receptors in neuroendocrine tissues: Topography,
subcellular distribution, and functional implications. In F. Naftolin, K. Ryan, & J.
Davies (Eda), WWW (pp-
277-304). Amsterdam: Elsevier.

McGinnis, M. Y., Drey, L. C., MacLusky, N. J., & McEwen, B. S. (1981). Steroid
receptor levels in intact and ovariectomized estrogen-treated rats: an examination
of quantitative, temporal and endocrine factors inﬂuencing the efﬁcacy of an
estradiol stimulus. Homoenglocrinology, 33, 158-165.

McGinnis, M. Y., & Mirth, M. C. (1986). Inhibition of cell nuclear androgen receptor
binding and copulation in male rats by an antiandrogen, Sch 16423.

Neuroendocnnolm, :13, 63-68.
Meisel, R. L., & Sachs, B. D. (1994). The physiology of male sexual behavior. In E. K.

a. J. D. Neill (Eds), thoﬂhyoioloaofﬁoproouoﬁon (pp. 3-105). New York:
Raven Press.
Menard, C. S., & Harlan, R. E. (1993). Up-Regulation of Androgen Receptor

Irnmunoreactivity in the Rat Brain by Androgenic-Anabolic Steroids. Brain

Mb, 622(1-2), 226-236.

140

Morali, G., Lemus, A. E., Oropeza, M. V., Garcia, G. A., & Perez-Palacios, G. (1990).
Induction of male sexual behavior by norethisterone: Role of its A-ring reduced
metabolites. Wm, 29, 185-187.

Morin, L. P., & Zucker, I. (1978). Photoperiodic regulation of copulatory behaviour in
the male hamster. mm, 12, 249-258.

Morrell, J. I., & Pfaff, D. W. (1978). A neuroendocrine approach to brain function:
Localization of sex steroid concentrating cells in vertebrate brains. Amman
Zoologist, 18, 447-460.

Murphy, M., & Schneider, G. E. (1969). Olfactory bulb removal eliminates mating
behavior in the male golden hamster. 891m, 181, 302-303.

Neal, J., Murphy, B. D., Moger, W. H., & Oliphant, L. W. (1977). Reproduction in the
male ferret: Gonadal activity during the annual cycle; recrudescence and

maturation. Biolomfkemxmetion 11, 380-385.
O‘Malley, B. W., Schrader, W. T., & Tsai, M. J. (1986). Molecular actions of steroid

 

hormones. a . a.

Olster, D. H. (1994). Hypothalamic estrogen receptor-immunoreactivity in prepubertal vs
adult female Guinea pigs. Immunology, 6, 617-625.

Paxinos, G., & Watson, C. (1986). W (Second ed).
San Diego: Academic Press.

Plant, T. M., Krey, L. C., Moossy, J ., McConnack, J. T., Hess, D. L., & Knobil, E.
(1978). The arcuate nucleus and the control of gonadotropin and prolactin
secretion in the female rhesus monkey (Macaca malaria). Endocrinology, 192,

52—62.

141

Powers, J. B., Newman, S. W., & Bergondy, M. L. (1987). MPOA and BNST lesions in
male syrian hamsters: differential effects on copulatory and chemoinvestigatory
behaviors. W, 23, 181-195.

Primus, R. J ., & Kellogg, C. K. (1990). Gonadal hormones during puberty organize
environment-related social interaction in the male rat. WWI,
24, 31 1-323.

Prins, G. S., & Birch, L. (1993). Irnmunocytochemical Analysis of Androgen Receptor
Along the Ducts of the Separate Rat Prostate Lobes After Androgen Withdrawal
and Replacement. Endmfmology, 132(1), 169-178.

Prins, G. S., Birch, L., & Greene, G. L. (1991). Androgen receptor localization in
different cell types of the adult rat prostate. Endocrinology, 122, 3187-3199.

Quarmby, V. E., Yarbrough, W. G., Lubahn, D. E, French, F. S., & Wilson, E. M.
(1990). Autologous down-regulation of androgen receptor messenger ribonucleic
acid. MoleonlaLEndmrinology, 4, 22-28.

Robel, P., Eychenne, B., Blondeau, J. P., Jung-Testas, 1., Groyer, M. T., Mercier-Bodard,
C., Hechter, O., Roux, C., & Dadoune, J. P. (1983). Androgen receptors in rat
and human prostate. HormonoRosoaroh, 18, 28-36.

Roselli, C. E., Ellinwood, W. E., & Resko, J. A. (1984). Regulation of brain aromatase
activity in rats. Endocrinology, 114(1), 192-200.

Roselli, C. E., Horton, L., & Resko, J. A. (1985). Distribution and regulation of
aromatase activity in rat hypothalamus and limbic systems. Endoorinology, 111,

247 l-2477.

142

Roselli, C. E., Horton, L. E., & Resko, J. A. (1987). Time-course and steroid speciﬁcity
of aromatase induction in rat hypothalamus-preoptic area. Bjology_of
Reproduction, 12, 628-633.

Roselli, C. E., & Resko, J. A. (1984). Androgens regulate brain aromatase activity in
adult male rats through a receptor mechanism. Engloozinology, 114, 2183-2189.

Ryan, K. D. (1985). Maturation of ovarian function in female ferrets is inﬂuenced by
photoperiod. W, :73, 503-507.

Ryan, K. D., & Robinson, S. L. (1985). A rise in tonic luteinizing hormone secretion
occurs during photoperiod-stimulated sexual maturation of the female ferret.
Endocrinology, 1.16, 2013-2018.

Sar, M., Lubahn, D. E, French, F. S., & Wilson, E. M. (1990). Irnmunohistochemical
localization of the androgen receptor in rat and human tissues. Endoginology,
121, 3 180-3186.

Sar, M, & Stumpf, W. E. (1977). Distribution of androgen target cells in rat forebrain
and pituitary after [RH-dihydrotestosterone administration. lomngloﬁﬁtoroio

Biochemistry,§, 1131-1135.

Sar, M, & W.E., S. (1975). Distribution of androgen-concnetrating neurons in rat brain.

Anatomicallieutoendocrinolm 120-133.
Scalia, F., & Winans, s. s. (1975). The differential projections of the olfactory bulb and

accessory olfactory bulb in mammals. WW, 1.6.1,
3 1-56.

Schanbacher, B. D. (1984). Regulation of luteinizing hormone secretion in male sheep by

endogenous estrogen. Endoorinology, 115, 944-950.

143

Shan, L., Hardy, D. O., Catterall, J. F., & Hardy, M. P. (1995). Effects of luteinizing
hormone (LH) and androgen on steady state levels of messenger ribonucleic acid
for LH receptors, androgen receptors, and steroidogenic enzymes in rat Leydig
cell progenitors in viva. Mom, 156, 1686-1693.

Shughrue, P. J ., Bushnell, C. D., & Dorsa, D. M. (1992). Estrogen receptor messenger
ribonucleic acid in female rat brain during the estrous cycle: A comparison with
ovariectomized females and intact males. Endmrinology, 151(1), 381-388.

Simerly, R. B., Chang, C., Muramatsu, M., & Swanson, L. W. (1990). Distribution of
androgen and estrogen receptor mRNA-containing cells in the rat brain: An in
situ hybridization study. Iholonmnloﬂﬂommmtiyoﬂomology, 2%, 76-95.

Simerly, R. B., & Young, B. J. (1991). Regulation of estrogen receptor messenger
ribonucleic acid in rat hypothalamus by sex steroid hormones. Moloonlnr
Englmnology, 5, 424-432.

Sisk, C. L. (1987). Evidence that a decrease in testosterone negative feedback mediates
the pubertal increase in luteinizing hormone pulse ﬁequency in male ferrets.

W 12, 73-81.
Sisk, C. L. (1990). Photoperiodic regulation of gonadal growth and pulsatile luteinizing

hormone secretion in male ferrets. JonmolofﬂiologiogLRhﬁhms, 5(3), 177-186.
Sisk, C. L. (1993). Autoradiographic labeling of preoptic area (POA) cells in prepubertal

and postpubertal male ferrets after in vivo injection of 3H-estradiol. Sooiﬂyfoi:

MW, 12, 820-

144

Sisk, C. L., Berglund, L. A., Tang, Y. P., & Venier, J. E. (1992). Photoperiod modulates
pubertal shifts in behavioral responsiveness to testosterone. lonmnloﬂﬁiologioal
Rm. 1, 329-339.

Sisk, C. L., & Desjardins, C. (1986). Pulsatile release of luteinizing hormone and
testosterone in male ferrets. W, 112, 1195-1203.

Sisk, C. L., & DonCarlos, L. L. (1995). Distribution of estrogen receptor-containing
neurons in prepubertal ferrets in the presence and absence of gonadal steroids.

Sodersten, P., Damassa, D. A., & Smith, E. R. (1977). Sexual behavior in developing
male rats. Homongnngﬂahnyior, 8(320-341).

Steel, R. G. D., & Torrie, J. H. (1980). W

BiomonioaLApnroach (second ed). New York: McGraw—Hill.
Steimer, T., & Hutchison, J. B. (1981). Androgen increases formation of behaviorally

effective oestrogen in dove brain. Noting, 222, 345.
Syms, A. J., Norris, J. S., Panko, W. B., & Smith, R. G. (1985). Mechanism of
androgen-receptor augmentation Analysis of receptor synthesis and degradation
by the density-shift technique. lonmaloﬂﬁiologionlﬁhomiony, 2%), 455-461.
Tan, J .-a., Joseph, D. R., Quarmby, V. E., Lubahn, D. B., Sar, M., French, F. S., &
Wilson, E. M. (1988). The rat androgen receptor: Primary structure,

autoregulation of its messenger ribonucleic acid, and immunocytochemical

localization of the receptor protein. Moloonlnnondmn’nology, 2, 1276-1285.

145

Tang, Y. P., & Sisk, C. L. (1988). Differential effects of testosterone,

5-alpha-dihydrotestosterone and oestradiol-176 on plasma concentrations of LH

in castrated ferrets. Whom, 1.11. 461-466.
Tang, Y. P., & Sisk, C. L. (1991). Testosterone in the MPOA elicits behavioral but not

neuroendocrine responses in ferrets. W, 26, 373-378.

Tobet, S. A., Basham, M. E., & Baum, M. J. (1993). Estrogen receptor immunoreactive

neurons in the fetal ferret forebrain. W, 12, 167-180.
Tobet, S. A., Zahniser, D. J ., & Baum, M. J. (1986a). Differentiation in male ferrets of a

sexually dimorphic nucleus of the preoptic/anterior hypothalamic area requires

prenatal estrogen. WM. 45.. 299-308.
Tobet, S. A., Zahniser, D. J., & Baum, M. J. (1986b). Sexual dimorphism in the

preoptic/anterior hypothalamic area of ferrets: Effects of adult exposure to sex

steroids. W, 3%, 249-257.

Tobet, S. Z., & Baum, M. J. (1991). Estradiol binding in the sexually dimorphic nucleus
of the preoptic/anterior hypothalamic area of adult male ferrets and in the
equivalent region of females. mm, 516, 345-350.

Truss, M., & Beato, M. (1993). Steroid Hormone Receptors - Interaction with
Deoxyribonucleic Acid and Transcription Factors. W, 15(4),
459-479.

Verhoeven, G., & Cailleau, J. (1988). Follicle-stimulating hormone and androgens

increase the concentration of the androgen receptor in sertoli cells.

Endocrinology. 122, 1541-1550.

146

Vito, C. C., Baum, M. J ., Bloom, C., & Fox, T. O. (1985). Androgen and estrogen
receptors in perinatal ferret brain. Iholonmalofﬂonrosoiong, 5(2), 268-274.

Watson, R. E., Weigand, S. J ., Clough, R. W., & Hoffman, G. E. (1986). Use of
cryoprotectant to maintain long-term peptide immunoreactivity and tissue
morphology. Emilia. 1, 155-159.

Weaver, C. E., & Baum, M. J. (1991). Differential regulation of brain aromatase by
androgen in adult and fetal ferrets. Endocrinology, 128(3), 1247-1254.

Westwood, W. J. A. (1962). Anatomy of the hypothalamus of the ferret. L_Qomo,
NonroL, 118, 323-341.

Whalen, R. E., & DeBold, J. F. (1974). Comparative effectiveness of testosterone,
androstenedione and dihydrotestosterone in maintaining mating behavior in the
castrated male hamster. Engloorinology, 25, 1674-1679.

Winans, S. S., & Powers, J. B. (1974). Neonatal and two-stage olfactory bulbectomy:
Effects on male hamster sexual behavior. Behayiornlﬁiology, 1_Q, 461-471.

Witkin, J. W., Ferin, M., Popilskis, S. J., & Silverrnan, A. J. (1991). Effects of gonadal
steroids on the ultrastructure of GnRH neurons in the rhesus monkey: Synaptic
input and glial apposition. Englmrjnology, 122(2), 1083-1092.

Witkin, J. W., & Romero, M. T. (1995). Comparison of ultrastructural characteristics of
gonadotropin-releasing hormone neurons in prepubertal and adult male rats.
Nonmojonoo, QM), 1145-1151.

Wolf, D. A., Herzinger, T., Herrneking, H., Blaschke, D., & Horz, W. (1993).

Transcriptional and Posttranscriptional Regulation of Human Androgen Receptor

Expression by Androgen. MolmilaLEndmnnology, 1(7), 924-93 6.

147

Wood, R. I., & Newman, S. W. (1993a). Intracellular partitioning of androgen receptor
immunoreactivity in the brain of the male syrian hamster: Effects of castration
and steroid replacement. Jonmoloﬂﬂonzooiology, 24, 925-93 8.

Wood, R. I., & Newman, S. W. (1993b). Mating activates androgen receptor-containing
neurons in chemosensory pathways of the male Syrian hamster brain. Brain
Bomb, 613, 65-77.

Worgul, T. J., Santen, R. J., Samojlik, E., Irwin, G., & Falvo, R. E. (1981). Evidence

that brain aromatization regulates LH secretion in the male dog. Amonioan

laumalﬂhxsiglegx. 241, E246-E250.

Yamamoto, K. R. (1985). Steroid receptor regulated transcription of speciﬁc genes and

gene networks. AnnualRexislofﬁenetics 12, 209-252.

Zhou, L., Blaustein, J. D., & De Vries, G. J. (1994). Distribution of androgen receptor

immunoreactivity in vasopressin- and oxytocin-immunoreactive neurons in the

male rat. Endocrinology, 185, 2622-2627.

111011an 31an UNIV. LIBRARIES
ll”Will"llllHllllllllllllllllllllllllllll
31293013995596