r. 2.3 4 . 2 . . . u nwdﬂm {Pm—N.“ . .
. . . 3A. awn-L5
. .r: .. FWMMnﬁa wwn
4.4.4? . 44.4..“ T,
{as 0 “We?

a
t. .
w

1... 1hr.

 

 

 

 

 

 

 

1.14 “3?.ng s4. {.29.
4 4444444444

. t. .1. (a I.

3.311.3hhm . .5 3mm.

. . . ,_ . . , ,. 2 .: 1 .. 1a.; 1.4
444432 .. : 2 _ 2 , 2. , .2 444444“ :234

:

 

A 3.1 . t. .

5W3:-

n n
égifOés)

LIBRARIES
MICHIGAN STATE UNIVERSITY
EAST LANSING, MICH 48824-1048

This is to certify that the
dissertation entitled

THE ROLE OF TESTOSTERONE IN SEXUAL

DIFFERENTIATION AND ADULT PLASTICITY IN A LIZARD:

THE COPULATORY NEUROMUSCULAR SYSTEM OF
GREEN ANOLES (ANOLIS CAROUNENSIS)

presented by

Melissa M. Holmes

has been accepted towards fulﬁllment
of the requirements for the

PhD. degree in Neuroscience

 

 

 

Major Professor’s Signature
May 5, 2005

 

Date

MSU is an Afﬁnnative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 a/CIRCIDateDtn—Jndd-pJS

 

THE ROLE OF TESTOSTERONE IN SEXUAL DIFFERENTIATION AND ADULT
PLASTICITY IN A LIZARD: THE COPULATORY NEUROMUSCULAR SYSTEM
OF GREEN ANOLES (ANOLIS CAROLINENSIS)

By

Melissa M. Holmes

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Neuroscience Program

2005

ABSTRACT
THE ROLE OF TESTOSTERONE IN SEXUAL DIFFERENTIATION AND ADULT
PLASTICITY IN A LIZARD: THE COPULATORY NEUROMUSCULAR SYSTEM
OF GREEN ANOLES (ANOLIS CAROLINENSIS)
By

Melissa M. Holmes

The neuromuscular system controlling copulation in green anoles (Anolis
carolinensis) is a powerful system for investigating sexual differentiation and
adult plasticity in the nervous system. Green anoles are native to the
southeastern United States, they are strict seasonal breeders (i.e. they only
breed during long days with warm temperatures), and their copulatory
neuromuscular system is remarkably sexually dimorphic. Adult males possess
two intromittant copulatory organs called hemipenes, each independently
controlled by two muscles and separate populations of associated motoneurons.
Adult females do not possess any of these structures. In male anoles the gonads
regress during the transition to winter conditions (shorter days and cooler
temperatures) and this regression results in decreased secretion of testosterone
(T), resulting in the cessation of reproduction. The present experiments clearly
demonstrate several conclusions about T action on the anole copulatory system.
1) Seasonal cues have limited effects on neuromuscular morphology of
gonadally intact males. 2) Exogenous T has trophic effects on copulatory
muscles and hemipenes in adulthood. 3) The soma size of motoneurons is not
responsive to T in adulthood, demonstrating a dissociation between the periphery

and central nervous system. 4) T treatment regulates expression of the androgen

receptor (AR) protein in a manner parallel to its morphological effects. 5) T has
trophic effects on the periphery, but not motoneurons, in juvenile males; no
effects are seen in juvenile females, suggesting sexual differentiation is complete
and non-reversible at this developmental stage. Finally, 6) T or the non-
aromatizable androgen dihydrotestosterone rescues this system in female
embryos, and administration of estradiol causes regression of the system in male
embryos. These results demonstrate that T is involved in both the development
and maintenance Of the lizard copulatory system. Comparing these data to T
effects in dewlap neuromuscular systems, which is involved in male courtship
behavior, reveals an intriguing selective sensitivity to T in green anoles. T has
very speciﬁc periods of action, as well as particular targets, even between two
sexually dimorphic neuromuscular systems in the same animal. Understanding
additional morphological and physiological parameters that respond to T in both
of these systems will further help identify the speciﬁc mechanisms via which T
mediates reproductive behaviors, and the neural circuits that underlie them, in

this species.

Copyright by
MELISSA MARIE HOLMES
2005

To Ashley and my family,
with endless love and gratitude.

ACKNOWLEDGMENTS

I would like to thank my graduate advisor and mentor, Dr. Juli Wade. It
has been a pleasure to work with you and I have Ieamed an immeasurable
amount. Your dedication and passion are truly inspirational. I also thank my
committee members, Drs. Marc Breedlove, Kay Holekamp, and Cynthia Jordan.
Your guidance, advice, and support are very much appreciated.

A very special thank you to both Matthew Lovem, who was instrumental in
developing my appreciation for herpetology and is a dear friend, and Camilla
Peabody, who is one of the most caring and generous people that l have had the
pleasure to know.

Working in the Wade lab was both intellectually stimulating and a great
deal of fun. Thank you to the many people that contributed: David Bailey, Casey
Bartrem, Laurel Beck, Matthew Burke, Carly Fuller, Katie Grausam, Amanda
Grimm, Jennifer Neal, Erin O’Byrant, Nancy Oberg, Claudia Ruiz, Jenny
Stynoski, Lace Svec, YuPing Tang, Sean Veney, and Malik Williams.

Finally, thank you to the faculty and staff of the Neuroscience Program,
and to my friends and colleagues in the Neuroscience Program, Department of

Psychology and Department of Zoology.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. x
LIST OF FIGURES .............................................................................. xii
KEY TO ABBREVIATIONS ................................................................... xvii
INTRODUCTION .................................................................................. 1
I) Sexual dimorphisms in brain and behavior ...................................... 1
ll) Steroid hormone effects on the nervous system: organization vs.
activation .................................................................................... 2
III) Neuromuscular systems and structure/function relationships ............. 3
IV) Comparative approaches for investigating sexually dimorphic
neuromuscular systems ................................................................. 4
Communication in ﬁsh .......................................................... 4
Vocalization in frogs ............................................................. 6
Vocalization in song birds ...................................................... 7
Copulation in mammals ........................................................ 9
V) The green anole as a reptilian model for neuroendocrine
investigations of neuromuscular systems ........................................ 11
The model system: Anolis carolinensis ................................. 11
The dewlap neuromuscular system ...................................... 13
The hemipenis neuromuscular system .................................. 14
Prospectus ..................................................................... 1 5
CHAPTER ONE
Characterization of projections from a sexually dimorphic motor nucleus in the
spinal cord of adult green anoles ............................................................ 17
INTRODUCTION ....................................................................... 18
MATERIALS AND METHODS ...................................................... 20
Animals and housing ......................................................... 20
Treatment, tissue collection, and analyses ............................. 21
RESULTS ................................................................................ 25
DISCUSSION ........................................................................... 31
Summary ........................................................................ 31
Organization of T17-S1 motoneuron projections ...................... 31
Implications for structure/function relationships
within this system ............................................................. 34
Comparison to organization of other dimorphic
neuromuscular systems ..................................................... 35
CHAPTER TWO
Seasonal plasticity in the copulatory neuromuscular system of green anole
lizards: a role for testosterone in muscle but not motoneuron morphology ....... 38
INTRODUCTION ........................................................................ 39

vii

MATERIALS AND METHODS ...................................................... 41

Housing conditions and care ................................................ 41
Treatment and tissue collection ............................................ 42
Tissue analyses ................................................................ 44
Statistical analyses ............................................................ 45
RESULTS ................................................................................. 46
Experiment 1 .................................................................... 46
Experiment 2 .................................................................... 49
DISCUSSION ........................................................................... 54
Summary ........................................................................ 54
Plasticity in motoneuron soma size ....................................... 55
Plasticity in androgen sensitivity in hemipenes and RPM
muscle ﬁbers .................................................................... 56
Different mechanisms in the copulatory vs. courtship
neuromuscular systems ...................................................... 59
Comparisons to other species ............................................. 59
CHAPTER THREE
Testosterone regulates androgen receptor immunoreactivity in the copulatory,
but not courtship, neuromuscular system in adult male green anoles ............. 61
INTRODUCTION ........................................................................ 62
MATERIALS AND METHODS ...................................................... 65
Housing conditions and care ................................................ 65
Treatment and tissue collection ............................................ 66
Tissue analyses ................................................................ 68
Statistical analyses ............................................................ 72
RESULTS ................................................................................. 72
Experiment 1: Gonadally intact males .................................... 72
Experiment 2: Testosterone manipulation .............................. 79
DISCUSSION ........................................................................... 82
Summary ........................................................................ 82
Comparison to testosterone effects on morphology .................. 82
Comparison to other species: muscle .................................... 84
Comparison to other species: central nervous system ............... 85
Conclusions ..................................................................... 87
CHAPTER FOUR

Effects of testosterone on the development of neuromuscular systems and their
target tissues involved in courtship and copulation in green anoles (Anolis

carolinenesis) ..................................................................................... 89
INTRODUCTION ........................................................................ 90
MATERIALS AND METHODS ...................................................... 95

Animals and housing ......................................................... 95
Experimental design .......................................................... 96
Tissue collection and analysis ............................................. 97
Radioimmunoassay ......................................................... 100

viii

Data analysis ................................................................. 100

RESULTS .............................................................................. 101
Body length and mass ...................................................... 101
Treatment effects on T ..................................................... 101
Dewlap system ............................................................... 104
Hemipenis system ........................................................... 104

DISCUSSION ......................................................................... 109
Summary ...................................................................... 109
Questions raised by the present data .................................. 110
77ming of T-responsiveness: implications for the classic theory
of organizational effects of hormones. . . .1 .............................. 1 13

CHAPTER FIVE

Sexual differentiation of the copulatory neuromuscular system in green
anoles (Anolis carolinensis): normal ontogeny and manipulation of steroid

honnones ........................................................................................ 119
INTRODUCTION ..................................................................... 120
MATERIALS AND METHODS .................................................... 123

Animals and housing ....................................................... 123
Treatment, tissue collection, and analyses ........................... 124
RESULTS .............................................................................. 130
DISCUSSION ......................................................................... 137
Summary ...................................................................... 137
Dissociation between neural and muscular components ......... 138

Steroid hormone effects on copulatory neuromuscular
structures ...................................................................... 140
DISCUSSION .................................................................................. 143
LITERATURE CITED ........................................................................ 157

ix

LIST OF TABLES
CHAPTER ONE

Table 1. Counts of fast blue-labeled motoneurons in T17-S1 following injection
into either one transversus penis (T PN), one retractor penis magnus (RPM), one
caudifemoralis (CF), or the cloacal sphincter (SC). Percentages of labeled
motoneurons, based on counts from thionin-stained alternate sections, are also
indicated. Means for each muscle are outlined in grey. All cells labeled following
TPN, RPM, and CF injections were unilateral, so the percent labeled cells was
based on thionin counts from the same side of the spinal cord. With SC
injections, labeled cells were detected on both sides of the cord, so they were
compared to the total number of thionin-stained motoneurons found bilaterally;
SC counts represent an average of the two sides of the spinal cord. All data
points come from separate individuals (see Materials and Methods) .............. 29

CHAPTER TWO

Table 1. Mean (+/- SEM) renal sex segment (RSS) cell height, body mass and
snout-vent length (SVL) for gonadally intact in Experiment 1. * indicates
statistically different from both other groups on same measure .................... 48

Table 2. Mean (+/- SEM) renal sex segment (RSS) cell height, body mass and
snout-vent length (SVL) for castrated males (GDX = with blank implant; GDX+T
= with testosterone implant) in Experiment 2. For comparisons across
experimental conditions, * indicates a signiﬁcant difference from BS-X, and **
indicates signiﬁcant differences from both other conditions. 3 indicates a
signiﬁcant difference between GDX and GDX+T groups within same
experimental condition ........................................................................ 52

CHAPTER THREE

Table 1. Mean (+/-SEM) percent of AR+ nuclei from hematoxylin and eosin
stained tissue. Nuclei were classiﬁed as in the center, on the edge, or outside of
the muscle ﬁbers for both the CH and RPM muscles. Asterisks indicate a
signiﬁcant main effect of testosterone (T) in castrated males (GDX) compared to
GDX males treated with vehicle ............................................................. 74

CHAPTER FOUR

Table 1. Mean (SEM) plasma androgen levels and epithelial cell heights of the
renal sex segments of juvenile male and female green anoles. * p < 0.05. ** p <
0.001 ............................................................................................. 103

CHAPTER FIVE

Table 1. Mean (+/- SEM) motoneuron count and soma size (umz) in spinal
segments trunk 17 and sacral 1 (T17-S1). Approximately half of the cells in this

region project to muscles controlling the hemipenes in adult male green
anoles .............................................................................................. 133

Table 2. Mean (+/- SEM) renal sex segment cell heights (pm), body weight (mg),
and motoneuron count and soma size (pmz) in Spinal segments tnInk 17 and
sacral 1 (T17-S1) for embryos treated on incubation days 10 and 13. Overall,
steroid treatment signiﬁcantly increased motoneuron number and size although
no signiﬁcant pairwise comparisons were detected. Androgen treatment
signiﬁcantly increased renal sex segment cell height in both males and females.
No differences in weight were detected for either sex or treatment ............... 136

xi

LIST OF FIGURES
CHAPTER ONE

Figure 1. Photomicrographs of muscles and motoneurons from adult male anoles
labeled with ﬂuorescent tract tracers. (A) FB+ motoneuron and (C) corresponding
labeled ﬁbers from FB-injected RPM. (B) CT-FITC+ motoneuron and (D)
corresponding labeled ﬁbers from CT-FlTC-injected TPN. Photomicrographs (A-
D) were taken from the same animal and the motoneurons were located in
adjacent tissue sections (in spinal segment T17). TPN and RPM muscle ﬁbers
run perpendicular to each other, so cross-sections of the tail produce cross-
sections of RPM ﬁbers and longitudinal sections of TPN ﬁbers. (E)
Photomicrograph of a coronal section through spinal segment T17 of an adult
male anole. The India ink used to mark eachsegment is visible on the external
surface of spinal cord, on the left side of the photomicrograph, and the central
canal is visible in the top right corner. T17 motoneurons are located in the ventral
horn (identiﬁed with arrow). (F) Fluorescent-labeled cells on one side of an
adjacent section to that depicted in panel (E), demonstrating FB+ and DY+
motoneurons in T17. Panels (E) and (F) are from a different animal than panels
(A-D). Scale bar = 10pm for (A) and (B); 95 pm for (C-E); 20pm for (F) ........... 26

Figure 2. Photomicrographs of F3 injection sites into (A) TPN, (B) RPM, (C) CF
and (D) SC of adult male anoles. Corresponding FB-I- motoneurons in the T17-S1
nucleus are shown in the insets. Scale bar = 360um for (AD) and 30pm for
insets ................................................................................................ 28

Figure 3. Total number (mean +/- SEM) of Nissl-stained motoneurons in the T17-
SI nucleus from one side of the Spinal cord. Asterisk indicates a statistically
signiﬁcant difference between males and females ...................................... 30

CHAPTER TWO

Figure 1. Effects of experimental condition on (A) T17-S1 motoneuron soma size,
(B) T16 soma size, (C) RPM muscle ﬁber size, (D) CF ﬁber size, and (E)

hemipenis cross sectional area in gondally intact males. ln panel A, a main effect
was detected without any signiﬁcant pairwise comparisons (see Results) ........ 47

Figure 2. Effects of experimental condition and testosterone manipulation on (A)
T17-S1 motoneuron soma size, (B) T16 soma size, (C) RPM muscle ﬁber size,
(D) CF ﬁber size, and (E) hemipenis cross sectional area in castrated males
(GDX = with blank implant; GDX+T = with testosterone implant). * indicates
signiﬁcantly different from other experimental groups (among BS, BS—X and
NBS). 3 indicates a signiﬁcant difference between GDX and GDX+T males within
an experimental group. In panel D, a main effect of testosterone manipulation
was detected without any Signiﬁcant pairwise comparisons; an effect of

xii

experimental condition stems from a signiﬁcant difference between the BS and
BS-X groups only (see Results) ............................................................... 50

Figure 3. Photomicrographs of cross-sections of RPM muscle ﬁbers (A, B, C) and
hemipenes (D, E, F) taken from BS males. A and D are from a gonadally intact
male, B and E are from castrated males with blank capsules, and C and F are
from a castrated male with a testosterone implant. Arrows point to the dorsal
edge of each hemipenis, where it meets the caudifemoralis muscle. Scale bar =
25pm (A, B, C) and 800nm (D, E, F) ........................................................ 51

CHAPTER THREE

Figure 1. Photomicrograph of a cross-section through CH muscle ﬁbers from a
gonadally intact male housed in non-breeding environmental conditions
illustrating expression of AR+ nuclei (A). (B) is the same section under
ﬂuorescent illumination to allow visualization of the DAPI nuclear label.
Comparison of the number of nuclei in (A) and (B) provides an estimate for the
total number of AR+ nuclei. Black arrows point to examples of AR+ nuclei and
white arrows point to DAPI labeled nuclei. Scale bar = 30pm ........................ 69

Figure 2. Mean (+/- SEM) percent of AR+ nuclei in the RPM (A and B) and CH (C
and D). These values were obtained from DAPI labeled tissue. Panels (A) and
(C) are from gonadally intact males (grey bars) and panels (B) and (D) are from
castrated males (GDX) treated with testosterone (T; black bars) or empty
capsules (white bars). Animals were housed either in breeding (BS) or non-
breeding (NBS) environmental conditions. Asterisks indicate a statistically
signiﬁcant main effect of testosterone treatment on the percentage Of AR+

nuclei ................................................................................................ 73

Figure 3. Photomicrographs of cross-sections through the ventral horn in spinal
segment T17 from a gonadally intact male (A), castrated male (B), and castrated
male treated with testosterone (C), all housed in breeding environmental
conditions. The majority of motoneurons have nuclei that exhibit dark AR
immunoreactivity; the percentage of AR+ cells not differ between experimental
groups. Scale bar = 50pm ..................................................................... 76

Figure 4. Photomicrographs of cross-sections through one hemipenis of a
gonadally intact male housed in breeding conditions. Panel (A) is a low
magniﬁcation photo of the entire cross-sectional area of the hemipenis. Panels
(B-D) are high magniﬁcation photos of the lobe epithelium (B), lobe body (C), and
clavula (D) from the hemipenis in (A). Scale bar = 600nm for (A) and 50pm for
(3-D) ................................................................................................. 77

Figure 5. Photomicrographs of cross-sections through the three cartilages
involved in dewlap extension: the second ceratobranchials (A), ﬁrst
ceratobranchials (B), and ceratohyals (C). Androgen receptor immunoreactivity is

xiii

present and dark in many of the nuclei in (A) and (C) but is absent or light in the
nuclei in (B). The ribbon-like appearance of the ﬁrst ceratobranchials in (B)
reﬂects the ossiﬁed nature of this component of the dewlap system; the second
ceratobranchials (A) and ceratohyals (C) are hyaline cartilage (13, 14). All tissue
is from a castrated male treated with testosterone and housed in breeding
environmental conditions. Scale bar = 30pm .............................................. 78

Figure 6. Mean (+l- SEM) cell height (pm) of epithelial cells in the renal sex
segment in (A) gonadally intact males (grey bars) and (B) castrated males
(GDX) treated with testosterone (T; black bars) or an empty capsule (white bars).
Animals were housed either in breeding (BS) or non-breeding (NBS)
environmental conditions. Asterisks indicate a statistically signiﬁcant difference
between BS and NBS males in (A) and a main effect of testosterone treatment in
(B). A statistically signiﬁcant main effect of season, as well as a signiﬁcant
season x testosterone interaction, is also present in (B) ............................... 80

CHAPTER FOUR

Figure 1. Treatment effects in juvenile female and male green anoles on
morphological components responsible for dewlap extension. (A) Length of the
2"d ceratobranchial cartilage; (B) cross-sectional area of the ceratohyoid muscle
ﬁbers; (C) cross-sectional areas of AmleNllmv (solid bars) and AmbX (hatched
bars) somata. Data were sampled 90 days after hatching, following 60 days of
treatment, and are presented as mean + SE (corrected for snout-vent length,
SVL). Sample sizes are given at the bottom of the bars and are not identical for
each tissue type due to histological artifact. Male and female data sets were
analyzed separately by two-sample t-tests. *P < 0.05 ................................. 105

Figure 2. Photomicrographs of ceratohyoid muscle ﬁbers collected from juvenile
green anoles 90 days after hatching, following 60 days of treatment. (A) Blank-
implanted females had signiﬁcantly smaller ﬁbers than (B) testosterone-
implanted females; (C) sham-castrated and (D) castrated males did not differ
signiﬁcantly. Scale bar = 20 um ............................................................. 106

Figure 3. Photomicrographs of the tail region in juvenile green anoles where
hemipenes are located, collected 90 days after hatching, following 60 days of
treatment. (A) Blank-implanted and (B) testosterone-implanted females did not
develop hemipenes; (C) sham-castrated males had Signiﬁcantly larger hemipene
cross-sectional areas than (D) castrated males. Scale bar = 20 pm. CF =
caudifemoralis; HP = hemipene; TPN = transversus penis .......................... 107

Figure 4. Treatment effects in juvenile female and male green anoles on
hemipenes and associated neuromusculature sampled 90 days after hatching,
following 60 days of treatment. (A) Hemipene and (B) retractor penis magnus
(RPM) ﬁber cross-sectional areas in males; females did not develop hemipenes
or RPMs. (C) Cross-sectional area of T17-S1 motoneuron somata in females and

xiv

males. Data are presented as mean + SE, corrected for snout-vent length, SVL.
Sample sizes are given at the bottom of the bars and are not identical for each
tissue type due to histological artifact. Male and female data sets were analyzed
separately by two-sample t-tests. *P < 0.05 .............................................. 108

CHAPTER FIVE

Figure 1. Photomicrographs Of cross sections through the gonads of untreated
developing green anole embryos. Gonads differentiate around incubation days 9
and 10 (total incubation time is 34 days on average). At days 7-9, gonadal tissue
in most individuals did not have deﬁning characteristics of either testes or
ovaries. By day 13, ovaries each had a well-developed cortex (noted with white
arrows) that contained developing follicles, and testes had well-developed
medullas with organized seminiferous tubules (noted with white asterisks). Scale
bar = 50 um ...................................................................................... 126

Figure 2. Photomicrographs of cross sections through the pelvic region of
untreated developing female (A, C, and E) and male (B, D, and F) embryos. The
dorsal surface of each embryo is on the top of each panel and the ventral surface,
with hemipenes protruding in most cases, is on the bottom. For orientation, the
Spinal cords are denoted with white arrows and one hemipenis in each panel
except (E) is indicated with a white asterisk. Kidneys (k), cloacal vent (c) and
caudifemoralis (cf) muscles are also marked. At incubation day 10 (A and B) both
males and females have everted hemipenes. At incubation day 13 (C and D), the
hemipenes are decreasing in size in females and increasing in size in males. By
incubation day 19 (E and F), hemipenes are absent in females. They remain
everted in males at this time. RPM ﬁbers project into the center of the hemipenes
when everted and can be seen in (F), indicated with a black arrow. Additional
tissue visible in some panels (as in the bottom right corner of (B)) is from the
lower limbs of the embryos. Scale bar = 215 um (A-D), 300 um (E, F) ........... 128

Figure 3. Photomicrographs of cross sections through the tail of two untreated
individuals, a female (A and C) and a male (B and D) hatchling. (A) and (B) are at
the level of the hemipenes, which are located on the ventral surface of the tail at
the junction of the caudifemoralis (cf) and ischiocaudalis (ic) muscles. (C) and (D)
are at the level of the RPM, which attaches to the base of the hemipenis in males
and is more caudal. Note that hemipenis and RPM are completely absent in the
female hatchling. H = hemipenis; RPM = retractor penis magnus. Scale bar = 100
um .................................................................................................. 131

Figure 4. Photomicrographs of cross sections through spinal segment trunk 17 in
untreated male embryos on incubation days 10 (A), 13 (B), 19 (C), 22 (D), 25 (E),
and at hatching (F). White arrows point to motoneurons in the ventral horn, some
of which project to the RPM and TPN muscles. Note the distinct conﬁguration of

the motoneurons that becomes apparent around incubation days 19 to 22. These

cells are not sexually dimorphic in size or number at the time of hatching. Scale
bar = 100 pm ..................................................................................... 132

Figure 5. Photomicrographs of cross sections through the tail (A-D) and pelvis
(E-H) of treated hatchlings. (A) and (B) are from a control female and male and
(C) and (D) are from an E-treated female and male, respectively. (A-D) are at the
level of the hemipenes, which are located on the ventral surface of the tail at the
junction of the caudifemoralis (of) and ischiocaudalis (ic) muscles. Note that the
hemipenis (marked with asterisks in which they are depicted) is completely
absent in the control female as well as both E treated animals. (E) and (F) are
the same animals as in (A) and (B) while (G) and (H) are from a DHT treated
female and male, respectively. For (E-H) the dorsal surface of each embryo is on
the top of each panel and the ventral surface is on the bottom. For orientation,
kidneys (k), cloacal vent (c) and cf muscles are noted. Note that hemipenes are
absent in the control female and are everted in both the DHT treated female and
male. The retracted hemipenes in (B) and (F) appear different because they are
at different levels; (B) is approximately midway through the rostral-caudal extent
of the hemipenis while (F) is at the level of the cloacal vent, right at the top of the
hemipenes. Scale bar = 150 um (A-D) and 300 um (E-H) ........................... 135

xvi

AmleNllmv
AmbX
ANOVA
AR

BC

BS
BSX
CF

CH
CT-FlTC
DAPI
DHT
DY

E

EOD
FB
GDX
ID

LA
NBS
nXllts
OD
PBF
PBS
RIA
RPM
SC
SNB
SVL

T

T1 7-S1
TPN

KEY TO ABBREVIATIONS

ventral motor nucleus of the facial nerve
vagal portion of the nucleus ambiguus
analysis of variance

androgen receptor

bulbocavemosus muscle

breeding season

breeding season with regressed testes
caudifemoralis

ceratohyoideus muscle

cholera toxin-B subunit conjugated to ﬂuorescein
4',6-DiamidinO-2-Phenylindole

dihyd rotestosterone

diamidino yellow

estradiol

electric organ discharge

fast blue

gonadectomized; castrated

inner diameter

levator ani muscle

non-breeding season

tracheosyringeal portion of the hypoglossal nucleus
outer diameter

phosphate buffered formalin
phosphate buffered saline
radioimmunoassay

retractor penis magnus muscle

cloacal sphincter

spinal nucleus of the bulbocavemosus
snout-vent length

testosterone

spinal segments trunk 17-sacral 1
transversus penis muscle

INTRODUCTION

l) Sexual dimorphisms in brain and behavior

Men and women are different. This truism is widely understood, even by
children at a very young age. We understand that males and females, humans or
otherwise, are often different in the way they look and in the way they behave.
That is, there are sexual dimorphisms between males and females in both
anatomy and behavior. Danrvin (1871) described three types of sex-related traits
that differ between the sexes: primary, secondary, and ‘eoological’. Primary sex
traits are deﬁned as those structures directly related to reproduction (i.e. the
gonads, the associated ducts and copulatory organs). Secondary sex traits are
anatomical features that indirectly participate in reproduction by increasing the
probability of copulation, including armaments and omamentations that are
involved in attraction of or competition for mates. Ecological sex traits may have
no bearing on reproductive ﬁtness, but arise as a consequence of males and
females of a species experiencing different ecological pressures and/or
occupying different niches (Andersson 1994). For example, in Centurus striatus
(a woodpecker species) males and females have very dimorphic bill morphology,
which reduces food competition between the sexes and ultimately expands the
food niche of the breeding pair (Ligon 1968; Selander 1966; as described in
Andersson 1994). Although Danrvin's concept of sexual dimorphism elegantly
categorizes gross morphology and behavior, it is not always clear how it applies
to the nervous system. Therefore, understanding neural dimorphisms in simple

behavioral systems (eg. neuromuscular systems, see below) that are

ecologically relevant, may help elucidate relationships among dimorphisms in
brain, behavior, reproduction, and ecology.

Despite the ubiquity of described gross anatomical and behavioral sex
differences throughout the animal kingdom, it is only relatively recently that
dimorphisms in the nervous system have received widespread attention. In fact,
dimorphisms have been identiﬁed in the nervous systems of all vertebrate
classes. Furthermore, within the nervous system, sex differences can occur at a
variety of levels, ranging from molecular to electrophysiological to gross
morphological. Of great interest is how these sex differences develop, how they
are maintained in adulthood, and how they relate to behavior. Given the strong
connection that neural dimorphisms often share with reproductive behavior and
strategy, it is also not surprising that there is a role for reproductive hormones in
the development and maintenance of sex differences in the central nervous

system.

ll) Steroid hormone effects on the nervous system: organization vs. activation
The classic template for describing the effects of gonadal hormones on
brain and behavior is to categorize effects as either a) organizational or b)
activational. This dichotomy, ﬁrst proposed by Phoenix and colleagues (1959),
describes how steroid hormones may act early in development to permanently
organize neural circuits or may act in adulthood to transiently activate behavior.
The criteria that distinguish these effects are the duration of the effect, the life

stage during which the hormone acts, and whether the effect occurs during a

critical period (Phoenix et al., 1959; Arnold & Breedlove, 1985). While the
heuristic is very useful, a growing body of literature reveals multiple exceptions to
these rules (see Arnold & Breedlove, 1985). Nevertheless, there are substantial
empirical data to support the hypothesis that testicular steroids (androgens) in
particular permanently organize the developing nervous system. However, the
emerging story is anything but simple. Testosterone (T) is the primary androgenic
steroid hormone that is secreted by the testes. However, T per se is often not
directly responsible for sexual differentiation and/or adult neuroplasticity. T often
serves as a pro-hormone when metabolism to 5-a-dihydrotestosterone (DHT; via
reduction by the 5-d-reductase enzyme) or aromatization to estradiol-17B (E; via

the aromatase enzyme) is required for biological effects (Nelson, 2000).

III) Neuromuscular systems and structure/function relationships

Because dimorphisms in the nervous system are often directly or indirectly
related to behavior, it is valuable to understand both how they develop (sexual
differentiation) and how they may change in adulthood (adult plasticity).
Neuromuscular systems are comprised of spinal or brainstem motoneurons and
their target musculature. Therefore these systems are particulariy conducive to
the study of both sexual differentiation and adult plasticity due to the direct link
between neuromuscular activity and behavior. Given that the behavioral
functions of neuromuscular systems are often obvious (Breedlove et al., 2002),
the adaptive signiﬁcance of dimorphisms is often obvious, particularly in systems

directly involved in reproduction. Indeed, it has been proposed that

understanding the mechanisms of steroid hormone action on neuromuscular
systems will assist with the understanding of hormone action on more

complicated neural systems (Breedlove et al., 2002).

IV) Comparative approaches for investmting sexually dimorphic neuromuscular
systems

In order to understand general principles of both sexual differentiation and
adult neuroplasticity. one must use a comparative approach. A brief review of
what is presently known about sexually dimorphic neuromuscular systems
reveals that their development and maintenance involve disparate mechanisms
among vertebrate groups. Yet similarities between model systems do exist,
suggesting that some mechanisms of hormonal action on differentiation and

plasticity may be more conserved than others.

Communication in ﬁsh

A vocalizing (sonic) species of ﬁsh (Midshipman ﬁsh, Porichthys notatus)
has two classes of sexually mature males - Type I and Type II (Brantley et al.
1993a). These two classes have different reproductive strategies and can be
distinguished by a number of traits. Type I males are larger (than Type II males
or females), build nests, guard eggs, and generate vocalizations or “hums,”
which are generated by the contraction of a single pair of muscles attached to the
lateral walls of the swimbladder. The Type I male vocalizations are performed
during the breeding season, to attract gravid females to their nests. In contrast,

Type II males sneak-spawn, and neither Type II males nor females perform these

“hums'. The frequency of the sound generated is determined by the synchronous
activity of central motoneurons that are located in the caudal brainstem. Sexual
dimorphisms are present at several levels of this system including number and
size of motoneurons, as well as total muscle and muscle ﬁber size (Bass and
Marchaterre 1989a; Bass and Marchaterre 1989b; Bass and Baker 1990). In all
of these features, Type I males are larger than either Type II males or females
(which are comparable to each other).

While sonic muscle size is larger in adult Type I males, it is not dimorphic
in juveniles (Bass and Marchaterre 1989a) which is likely due to the relatively low
levels of androgen at this developmental stage (Brantley et al. 1993a). In fact,
this muscle is androgen-sensitive in all 3 phenotypes (Type I males, Type II
males, and females), whereby androgen exposure increases muscle mass and
ﬁber structure toward the Type I male phenotype (Brantley et al. 1993b). Type I
and Type II males have different androgen proﬁles: 11-Ketotestosterone is the
predominant androgen in Type I males while T is the most prevalent androgen in
Type II males (Brantley et al. 1993c). Androgen proﬁles in Type II males more
closely resemble that of females. While additional data are required to
conclusively state which particular androgens are responsible, it is likely that
there is a role for androgen in the development and/or maintenance of these
inter- and intra-sexually dimorphic structures.

Another group of teleosts, weakly electric ﬁsh, communicate via an electric
organ discharge (EOD). Production of the EOD occurs via activation of a central

motor pathway that excites an electric organ, comprised of modiﬁed muscle cells,

located in the tail. Several species use sexually dimorphic EOD frequencies to
recognize the gender of a conspeciﬁc (Dunlap et al. 1998), but the direction of
this behavioral dimorphism varies across species. Similar to the sonic system in
midshipman, the electric organ system is functionally responsive to androgens
and estrogens whereby treatment with E will feminize the signal, while non-
aromatizable androgens masculinize it. Effects of androgens on morphology of
the system are species-speciﬁc whereby some species exhibit larger electrocytes
in response to T while others do not (6.9. Mills et al. 1992). Therefore, species
differences in the direction of functional and morphological dimorphisms can be
explained by species-speciﬁc responses to these steroids (see Zakon 8 Dunlap,

1999).

Vocalization in frogs

Adult male African clawed frogs (Xenopus Iaevis) use vocalizations to
court females, and the neuromuscular system underlying this behavior is sexually
dimorphic at multiple levels. Motoneurons located in the caudal medulla send
axons through cranial nerves lX-X to the larynx (Wetzel et al. 1985). These
axons synapse on the laryngeal dilator muscle, and contraction of this muscle
causes vocalizations to be produced (Tobias and Kelley 1995). The motoneurons
are larger and more numerous in males than in females (Hannigan & Kelley
1981; Kelley & Fenstemaker 1983). The laryngeal muscle is also dimorphic in
ﬁber type composition (males have one ﬁber type while female muscles are

heterogeneous), as well as ﬁber size and number (favoring males; Sassoon &

 

 

Kelley 1986). Also, the larynx is 2-3 times larger overall in males than in females
due to increased muscle and cartilage mass (Sassoon & Kelley 1986).
Production of male speciﬁc courtship vocalizations depends on gonadal
androgens (T or DHT) but not E (Wetzel & Kelley 1983), although treatment with
T will not induce a masculine call in adult females (Hannigan 8. Kelley 1986). This
lack of behavioral plasticity in adult females is likely a result of androgen-
regulated masculinization during development (Kelley 1986). Either T or DHT
(but not E) is capable of inducing laryngeal myogenesis in juveniles of either sex
(Sassoon et al. 1986). Furthermore, while the number of motoneurons is not
plastic in adulthood, soma size is - androgen treatment can shift females in the
masculine direction (Hannigan & Kelley 1983). Of particular interest is that
implantation with male testicular tissue is more successful at masculinizing
females than prolonged T treatment. Adult females can produce male typical
vocalizations as a result of these testicular implants but the effect is more striking
when implantation occurs while the females are still juveniles (Watson & Kelley
1992), suggesting that complete masculinization requires exposure during a

critical period and/or testicular secretions other than T.

Vocalization in song birds

A signiﬁcant body of literature exists on mechanisms regulating sexual
differentiation and adult plasticity in the adult passen'ne brain. In particular, the
zebra ﬁnch is a popular model for investigating hormonal and non-hormonal

effects on sexual differentiation (see Wade 2001 for review). Despite this fact,

surprisingly little work has been done on the neuromuscular system that actually
controls the output of song. This system consists of motoneurons located in the
tracheosyringeal portion of the hypoglossal nucleus (nXllts) and their target
muscles: two of which are the ventralis and dorsalis muscles of the syrinx.
Dimorphisms exist at both the neural and muscular levels of this system. nXllts
volume, syrinx mass and both dorsalis and ventralis muscle ﬁber size are larger
in males than females (Nottebohm and Arnold 1976; Springer and Wade 1997;
Wade et al. 1999; Wade and Buhlman 2000). Treatment with T in adulthood can
increase syrinx weight and muscle ﬁber size in adult females, although the
effects of T on nXllts volume, motoneuron soma size or number are variable
(Arnold 1980; Gurney 1981; Wade and Buhlman 2000). Similarly, T treatment
slightly increases syrinx weight and DHT increases nXllts motoneuron soma size
in juvenile females (Wade et al. 2002). Further supporting a role for androgens,

adult males have greater androgen receptor (AR) mRNA expression than

females in syrinx muscle and cartilage (Veney and Wade, 2004), although no sex

difference in AR mRNA was detected at days 3, 10, and 17 post-hatching (Veney

and Wade, 2005). However, E may also mediate the differentiation of some
components of this system. E appears to actively feminize the syrinx as E-
treatment decreases muscle ﬁber size and overall syrinx weight in juvenile males

(Wade et al., 2002).

Copulation in mammals

Perhaps the best characterized sexually dimorphic neuromuscular system
is the spinal nucleus of the bulbocavemosus (SNB) system in rodents. This
system is responsible for controlling penile reﬂexes that are required for
successful fertilization. Motoneurons in the lower lumbar spinal cord project to
the bulbocavemosus (BC) and levator ani (LA) muscles that attach to the base of
the penis. In adult male lab rats, SNB motoneurons are larger and more
numerous than in females, and the BC/LA musculature is either vestigial or
absent in adult females (Breedlove and Arnold 1980). These dimorphisms
develop shortly after birth as both males and females are born with BC/LA
muscles attached to the base of the penis and/or clitoris and these muscles are
already innervated by SNB motoneurons (Rand and Breedlove 1987). During this
perinatal period the BC/LA muscles atrophy and the corresponding motoneurons
die in females (Nordeen et al. 1985).

The development and maintenance of this system is remarkably sensitive
to gonadal androgens. Perinatal T treatment can rescue SNB motoneurons as
well as the BC/LA musculature in females, and perturbation of androgen
signaling (via pharmacological manipulation or castration) causes regression of
these structures in developing males (Breedlove 8. Arnold 1983a 8. b). These
effects are due to the biological activity of the androgen receptor, which was
elegantly shown by the use of androgen insensitive rats. Testicular feminization

mutation (tfm) results in genotypic (XY) males possessing non-functional

androgen receptors. As such, these animals are insensitive to circulating levels
of androgen (either endogenous or exogenous) and develop a female phenotype
(Breedlove & Arnold 1981 ).

The SNB system is also sensitive to circulating steroids in adulthood.
Castration of adult males results in decreased SNB motoneuron soma and
nucleus size, dendritic arborization, neuromuscular junction (NMJ) size and
BCILA muscle mass (e.g. Balice-Gordon et al 1990; Breedlove and Arnold 1981;
Rand and Breedlove 1995). All of these effects can be reversed by T
replacement. T acts directly on the BCILA muscles to increase muscle ﬁber Size
and motoneuron dendrites but acts on the motoneurons themselves to inﬂuence
soma size.

The SNB system can also Show signiﬁcant naturally occurring (and
therefore potentially functionally relevant) adult plasticity in seasonally breeding
species (eg. the white-footed deer mouse, Peromyscus Ieucopus, and the
Siberian hamster, Phodopus sungorus). Speciﬁcally, SNB motoneuron somata
and nuclei size, neuromuscular junction size, and BCILA muscle mass are larger
in breeding males (Forger & Breedlove 1987; Hegstrom and Breedlove 1999;
Hegstrom et al. 2002). The alterations in soma and nucleus size, as well as
muscle mass and ﬁber size, are a result of corresponding changes in circulating
androgens (decreased in non-breeding males due to regression of the testes).
However, T is not directly responsible for all changes in this system; for example,
NMJ size appears to be sensitive to photoperiod independent of circulating levels

of T (Hegstrom et al. 2002).

10

V) The green anole as a reptilian model for neuroendocrine investigations of
neuromuscular systems

 

Recent work, primarily in the laboratory Of Juli Wade, has focused on
establishing the green anole (Anolis carolinensis) as a reptilian model for
investigations of sexual differentiation and adult neuroplasticity. Green anoles
differ from more traditional laboratory species in that they possess both a
sexually dimorphic copulatory neuromuscular system and a dimorphic
neuromuscular system involved in courtship. As such, they provide a unique
opportunity to compare sexual differentiation and adult plasticity of two dimorphic

neuromuscular systems within the same species.

The model system: Anolis carolinensis

One advantage of the use of green anoles as a comparative model
system for neuroendocrine research is that there is a remarkable amount of ﬁeld
and laboratory data describing the ecology, behavior and physiology of this
species. The social organization is polygynous with males competing for and
establishing large territories that encompass the smaller territories of multiple
females (Ruby 1984; Jenssen et al. 1995a; Jenssen and Nunez 1998). Males
defend their territories in an attempt to maintain exclusive reproductive access to
these females. Larger males establish and defend larger territories and, as such,
have access to more females; smaller males may not establish a territory at all,
and instead become “ﬂoaters” in the habitat trying, to evade detection by other

males.

11

Interestingly, precopulatory mate choice by females is not evident in this
species. Females establish their territories independent of male location and then
mate with the male that establishes the overlapping territory (e.g. Jenssen et al.
2001). Despite the apparent lack of direct female mate choice, males do court
females by performing a suite of head-bobbing displays coupled with the
extension of a bright red throat fan, called a dewlap. If the female is receptive
She will remain stationary (often bobbing her head in return but not extending her
dewlap) while the male approaches and mounts her, biting the back of her neck
while clasping her torso with his forelegs. lntromission occurs when the male ﬂips
his tail under that of the female and inserts one of his two copulatory organs
(called hemipenes) into the female cloaca. Copulation can last from less than ﬁve
to more than sixty minutes (Greenberg and Noble 1944; Jenssen and Nunez
1998)

Reproduction in this species is regulated by both environmental and
physiological conditions. Green anoles are strictly seasonal breeders,
reproducing in the ﬁeld from about April to July when days are relatively long and
temperatures are relatively high. T levels in males are about twice as high during
the breeding season as during the non-breeding season (Lovem et al. 2001).
This T is important for the expression of male sexual behaviors, as castration
decreases these behaviors while T can prevent or reverse this effect (Adkins and
Schlesinger 1979; Crews et al. 1978; Mason and Adkins 1976; Rosen and Wade
2000; Winkler and Wade 1998). Males housed in non-breeding conditions in the

12

laboratory do not exhibit copulatory behavior, even if treated with high levels of T
(O’Bryant and Wade 2002a).

The dewlap neuromuscular system

Both the dewlap and the neuromuscular system controlling its extension
are sexually dimorphic (Jenssen et al. 2000; O'Bryant and Wade 1999; O'Bryant
and Wade 2002b; Wade 1998). The dewlap system consists of paired
ceratohyoid muscles located in the throat that are innervated by motoneurons in
the caudal brainstem via the ramus pharyngo—laryngeus nerve (Wade 1998).
When they contract, the ceratohyoid muscles cause the cartilage to bow out to
extend the dewlap from the ventral surface of the neck. Both male and female
anoles possess and use dewlaps, but male dewlaps are about 7-fold larger in
surface area and are used more frequently than those of females (Greenberg
and Noble 1944; Jensson et al. 2000; Nunez et al. 1997). Furthermore, both
males and females will extend their dewlaps in aggressive encounters, but only
males use the dewlap in courtship (Greenberg and Noble 1944). Dimorphisms
are present at multiple levels of this system including motoneuron soma size,
nerve cross-sectional area, neuromuscular junction (NMJ) size, cartilage length
and muscle ﬁber size, number and density (all favoring males with the exception
of muscle ﬁber density; Wade 1998; O'Bryant and Wade 1999; O'Bryant and
Wade 2002b). Interestingly, the morphology of this system is not plastic in
adulthood (O'Bryant and Wade 1999); it does not vary naturally across season

nor does it respond to changes in circulating levels of T (castration vs. T

13

replacement). This may not be surprising given that, while only males use the
dewlap for courtship during the breeding season, both males and females
continue to use their dewlaps across seasons for aggressive interactions (albeit
at much lower frequencies during the non-breeding season).

The dewlap neuromuscular system is not sexually dimorphic at the time of
hatching but is fully differentiated by adulthood. It is not yet known whether
differentiation of the dewlap system is mediated by T. However, the time course
of differentiation as reported by O’Bryant and Wade (2001) does parallel a
developmental rise in T that occurs only in juvenile males (Lovem et al. 2001).
Chapter 5 (see below) directly addresses whether T masculinizes the dewlap

system during the period of its differentiation.

The hemipenis neuromuscular system

Like the sexually dimorphic dewlap system, the copulatory neuromuscular
system in green anoles is dimorphic on multiple levels (Ruiz and Wade 2002).
Male anoles possess bilateral copulatory organs (located in the tail) called
hemipenes that are independently controlled by two muscles that work in an
antagonistic manner (see Figure 1). The transversus penis (TPN) wraps over the
hemipenis with the muscle ﬁbers running in a medial to lateral orientation; when
the TPN contracts the hemipenis is everted through the cloaca (Arnold 1984).
The retractor penis magnus (RPM) is attached to the base of the hemipenis with

the ﬁbers running in a rostral to caudal orientation; contraction of the RPM

14

retracts the hemipenis back into the tail (Arnold 1984). Adult females do not
possess hemipenes or either TPN or RPM muscles.

Injection Of the retrograde neuronal tracer biocytin into the TPN revealed
that the motoneurons projecting to this muscle are located in the pelvic region of
the spinal cord in spinal segments trunk 17 and sacral 1 (T17-S1). Anoles
possess 17 trunk segments analogous to thoracic and lumbar segment in
mammals as well as 2 sacral segments (Ruiz and Wade 2002). The TPN
motoneurons are located in the ventrolateral portion of the ventral horn and they

are larger and more numerous in males than in females (Ruiz and Wade 2002).

Prospectus

Given that green anoles are seasonal breeders and we know that
copulatory systems can be plastic in seasonally breeding mammals, it may be
that their copulatory neuromuscular system is a novel system with which to
investigate naturally occurring and functionally relevant adult neuroplasticity. The
experiments described below are essential not only for the characterization of the
green anole copulatory neuromuscular system per se, but also for the
establishment of this system for comparative investigations of sexual
differentiation and adult neuroplasticity both within this species (courtship vs.
copulation) and across vertebrate groups. Speciﬁcally, in order to bring present
understanding of this model system up to the level of other sexually dimorphic
neuromuscular systems, we need to know (a) how the system is organized by

characterizing motoneuron projections, (b) whether this system is plastic at

15

different life stages, including adulthood, and if putative plasticity is due to
changes in circulating androgen levels, (c) how androgen receptors (AR) are
distributed in these tissues in adulthood, and (d) when this system differentiates
and if differentiation is caused by androgens. Taken together, these experiments
will elucidate the role of androgen in the copulatory neuromuscular system of
male green anoles and will help determine whether mechanisms of steroid action
on copulatory systems (both in development and adulthood) are conserved
across species.

The experiments described in Chapter 1 identify the location of RPM
motoneurons and characterize the relative distribution of TPN and RPM
motoneurons within spinal cord segments T17-S1. The experiments in Chapter 2
investigate whether the sexually dimorphic copulatory neuromuscular system is
plastic in adulthood, and the role that testosterone plays. Chapter 3 examines
the distribution of androgen receptor (AR) in the copulatory tissues (including the
spinal motoneurons, musculature, and hemipenes) and dewlap tissues (including
muscles and cartilage) of adult males. Chapter 4 investigates the effects of
juvenile testosterone in sexual differentiation of the dewlap and copulatory
systems. Finally, Chapter 5 examines (a) the developmental time course of
differentiation of the copulatory neuromuscular structures, and (b) the effects of

steroid hormones on this process during embryonic development.

16

CHAPTER ONE

Holmes MM, Wade J (2004). Characterization of projections from a sexually
dimorphic motor nucleus in the spinal cord of adult green anoles. J Comp Neurol
471: 180-187.

17

INTRODUCTION

Neuromuscular systems are exceptionally well suited to investigations of
structure/function relationships in the central nervous system. The direct link
between neural activity and muscle contraction, as well as the often obvious
behavioral function of the target structure, make it easier to identify the adaptive
signiﬁcance of dimorphisms in these compared to more complicated neural
systems (Breedlove et al., 2002). The function of structural differences is
particularly evident in sexually dimorphic neuromuscular systems involved in
reproductive behaviors, and examples of such systems exist in all vertebrate
classes. In species that exhibit male vocal courtship behavior (including frogs,
songbirds and some ﬁsh), the neuromuscular system mediating production of
vocalizations contains motoneurons and/or muscles that are larger in males than
females (e.g. Sassoon and Kelley, 1986; Simpson et al., 1986; Bass and
Marchaterre, 1989a,b; Wade and Buhlman, 2000). The system underlying
mammalian copulation shows striking dimorphisms in motoneuron and muscle
ﬁber number as well as size, again favoring males (reviewed in Breedlove et al.,
2002)

The green anole lizard (Anolis carolinensis) has similar sexually dimorphic
neuromuscular systems, one involved in courtship and the other in copulation.
Adult males court females by performing head bobbing displays coupled with
extension of a red throat fan, called a dewlap. Females have much smaller
dewlaps than males, and extend them far less frequently, only in agonistic

encounters (Greenberg and Noble, 1944; Nunez et al., 1997; Jenssen et al.,

18

2000). The motoneurons controlling dewlap extension are in the caudal
brainstem and innervate the paired ceratohyoid muscles, located in the throat
(Font and Rome, 1990; Wade, 1998); contraction of these muscles causes the
second ceratobranchial cartilage to bow away from the ventral surface of the
throat, extending the dewlap (Bels, 1990). Similar to the systems outlined above,
many components, including motoneuron somata and ceratohyoid muscle ﬁbers,
are larger in males than females (Wade, 1998; O’Bryant and Wade, 1999;
O’Bryant and Wade, 2002b).

More recently, the neuromuscular system underlying copulation in green
anoles has been investigated. Male squamate reptiles (lizards and snakes) each
possess two penises called hemipenes. These hemipenes are located bilaterally
in the tail, and only one is extended through the cloaca during each copulation.
Two muscles mediate extension (transversus penis, TPN) and retraction
(retractor penis magnus, RPM) of each hemipenis (Arnold, 1984). Adult female
anoles do not have hemipenes or either of these muscles (Ruiz and Wade,
2002). Injection of the retrograde tracer biocytin into the TPN revealed that its
corresponding motoneurons are located in segments trunk 17 and sacral 1 (T17-
31) in the caudal spinal cord. These cells are larger and more numerous in
males than females (Ruiz and Wade, 2002).

Before one can appreciate the relationships between structure and
function in a given neuromuscular system, one needs to understand how the
motoneurons, nerves, and muscle(s) that comprise it are organized. While this

work has been largely completed for other sexually dimorphic neuromuscular

19

systems (e.g. Nottebohm et al., 1976; Schroder, 1980; Bass, 1985; McKenna
and Nadelhaft, 1986; Simpson et al., 1986; Wade, 1998), organization of the
anole copulatory system has not yet been fully characterized. lmportantly, the
motoneurons projecting to the RPM must be located. The ﬁrst goal of the
present investigation was to ﬁnd these cells in males (Study 1). Additional goals
of this investigation were to identify other muscle targets for T17-S1 motoneurons
and to estimate the relative proportions of T17-S1 motoneurons that project to
each in both males (Study 2a) and females (Study 2b). In addition to the TPN,
the caudifemoralis (CF) receives projections from T17-S1 motoneurons (Ruiz
and Wade, 2002). The cloacal sphincter (SC) was also considered a likely target
as SC motoneurons are located in a similar pelvic region of the spinal cord in
birds (Ohmori and Watanabe, 1989; Seiwert and Adkins-Regan, 1998).
Furthermore, external sphincter motoneurons are interdigitated with copulatory
motoneurons and share the pudendal nerve in rats (Schroder, 1980; McKenna
and Nadelhaft, 1986), and cloacal muscles receive projections via the pudendal

nerve in Iguanas (Akita, 1992).

MATERIALS AND METHODS
Animals and housing
Adult (>50 mm snout-vent length) male and female green anoles were
purchased from Charles Sullivan Co (Nashville, TN). Prior to the
commencement of each experiment, animals were group housed (1 male with 3-

6 females) in 110-liter glass aquaria (76 x 30 x 48 cm). Lizards were housed in a

20

14:10 hour lightzdark cycle with ﬂuorescent and full-spectrum lights, as well as a
heat lamp placed on one end of each aquarium. Temperature ranged from 28'C
(ambient) to 38'C (directly under heat lamp) during the day to 18'C at night.
These conditions approximate those the animals experience in the spring in the
ﬁeld and stimulate breeding in the lab. Aquaria were sprayed daily to help
maintain 70% relative humidity, and water was provided ad Iibitum. Animals
were fed crickets or mealworrns three times a week. Following injection of tract
tracers (see below), all animals were housed individually in 21-liter glass aquaria
(42 x 20 x 24 cm) and cared for as above. All procedures were approved by the
Michigan State University All University Committee on Animal Use and Care and

conform to NIH guidelines.

Treatment, tissue collection, and analyses

St_udy1. Seven adult male lizards were anesthetized with isoﬂurane,
placed on ice, and incisions were made in the ventral surface of the tail. The
retrograde tracer Fast Blue (FB; llling Plastics, Bergfeld, Germany; 0.5 pl at 2.5-
5% in 0.9% saline) was bilaterally injected into the RPMS. Additionally, the
retrograde tracer Cholera Toxin-B subunit conjugated to ﬂuorescein (CT-FITC;
List Biological Laboratories; 0.5 pl at 2% in 0.9% saline) was bilaterally injected
into the TPNS of the same individuals. To determine whether the projections
were ipsilateral and to conﬁrm that there were no double-labeled motoneurons,
which would indicate tracers leaking into a muscle not intentionally injected or an

unusual situation of a motoneuron projecting to more than one muscle, an

21

additional six males were injected as above with FB into one RPM and Diamidino
Yellow (DY; 0.5 pl at 3% in 0.9% saline; Sigma) into the ipsilateral TPN. CT-FITC
and F3 both label the cytoplasm of the cell (Kuypers et al., 1980; Dederen et al.,
1994) whereas DY primarily labels the nucleus (Keizer et al., 1983). As such, we
were better able to identify possible double-labeled motoneurons when F8 was
used with DY. To decrease possible contamination, excess tracer was removed
with a cotton swab and gel foam prior to suturing the incision with silk.

Three days following injection, lizards were overdosed with Brevital
Sodium (0.03 ml) and perfused with 0.1M phosphate buffered saline (PBS; pH =
7.4) and 4% paraformaldehyde in PBS. Spinal segments were marked with India
ink to permit identiﬁcation of individual segments. Cords were extracted
(segments T15-S2), embedded in gelatin, post-ﬁxed in 4% paraformaldehyde for
2.5 hours and then transferred to 20% sucrose in PBS overnight. Hemipenis,
TPN, and RPM were also extracted from the rostral tail in one block (which also
included the CF) and processed as above. Cross sections of all tissue were cut
frozen at 30 pm into PBS. Two series of alternate sections of spinal cord were
mounted onto gelatin-coated slides from 9:1 dH20:PBS. Two series of every
fourth section of muscle/hemipenis tissue was mounted from the same solution.
One series each of spinal cord and muscle/hemipenis tissue was used for
ﬂuorescent analyses (see below). They were soaked in 0.1% sodium borohydride
in PBS for 10 minutes to decrease auto-ﬂuorescence (Clancy and Cauller, 1998)
and rinsed twice with PBS and once with dH20 prior to dehydration. Alternate

Spinal cord sections were stained with thionin. In some cases, series of tail tissue

22

were stained using the trichrome method to facilitate localization of speciﬁc
muscles. However, leakage of ﬂuorescent dyes into muscles neighboring those
injected was obvious even in unstained tail sections. All tissue was dehydrated,
cleared in xylene and coverslipped using either DPX (for ﬂuorescent tissue) or
Perrnount (for thionin- and trichrome-stained tissue).

Spinal cord and muscle tissue was analyzed using a microscope
(Olympus BX60) capable of simultaneous visualization of the multiple
ﬁuorochromes. Spinal segments T15 through 82 were analyzed for the presence
of FB+, CT-FITC+, and DY+ motoneurons; sections containing labeled cells were
compared to thionin-stained tissue in which the ink marking the segments was
obvious. In addition, it was noted whether the positively labeled motoneurons
were ipsilateral or contralateral to the injection, and whether any double-labeled
motoneurons were detected. In all cases, muscle tissue was analyzed to conﬁrm
appropriate and accurate injection sites.

Study 2a. To identify other muscle targets for T17-S1 motoneurons, and to
determine the proportions of cells projecting to each muscle, adult male lizards
received a single tracer injection in one of four muscles: TPN, RPM, CF or SC.
Injections were unilateral in the TPN, RPM, and CF. The SC wraps around the
cloaca (Arnold, 1984) and was injected at the midline in both rostral and caudal
ﬁbers. To control for possible differences in transport efﬁcacy between tracers, all
males were injected with FB (as above). For estimation of percentages of
motoneurons projecting to each muscle (see below), three males were used per

muscle. While more were injected, a few were not analyzed either because

23

analysis of the tail tissue indicated the tracer had leaked outside the intended
muscle or too much auto-ﬂuorescence existed in the Spinal cord tissue. All tissue
(spinal cord and muscle) was collected and processed as described for Study 1
for animals with TPN, RPM, or CF injections. When 805 were injected, the
muscle dissection included the entire cloacal region.

Study 2b. To characterize T17-$1 motoneuron projections in females, ﬁve
adult female anoles received a single injection of F8 into either the CF (unilateral;
n=3) or the CS (n=2). Tissue processing and analyses were performed as

outlined for Study 2a.

For all animals in Study 2, FB+ neurons were counted in spinal segments
T15-SZ. All cells labeled following TPN, RPM, and CF injections were unilateral,
so total counts from the alternate sections were multiplied by two to provide an
estimate for one side of the spinal cord. With SC injections, labeled cells were
detected on both sides of the cord, so they were compared to the total number of
thionin-stained motoneurons found bilaterally; the left and right SC counts were
then averaged to provide an estimate for one side of the spinal cord. To avoid
double counting of cells across sections, motoneurons were assessed using the
physical dissector technique (as in Ruiz and Wade, 2002; Gunderson, 1986),
whereby cells were counted if the nuclei come into focus and disappear within a
given section. This technique was not used for counting of ﬂuorescent
motoneurons as their nuclei are not always identiﬁable due to the homogenous

nature of the FB label. However, double-counting of FB+ motoneurons was

24

unlikely given that cells were counted (a) in alternate 30pm sections (and
diameter of these motoneurons is approximately 20pm) and (b) only if they were
large and bright (i.e. not cell fragments or split nuclei). AS with Study 1, muscle
tissue was analyzed to conﬁrm appropriate and accurate injection sites. For each
animal, the FB+ total was divided by the total motoneuron count to provide an
estimated percentage of the number of T17-S1 motoneurons that project to that
muscle. In order to conﬁrm that T17-S1 motoneurons were sexually dimorphic in
number (Ruiz and Wade, 2002) in the present sample, total motoneuron counts
(from thionin stained tissue) were compared between males (n=12) and females
(n=5) using an unpaired t-test.

Fluorescent photomicrographs were obtained with an Olympus 35mm
camera and Kodak Elite Chrome 400 ﬁlm for color slides. Images were then
digitally scanned (Polaroid Sprint Scan 35 Plus). Nissl photomicrographs were
obtained using a Q Imaging MicroPublisher 5.0 digital camera. All images were
sized, compiled, and labeled in Adobe Photoshop 7.0. Brightness and/or contrast
were adjusted slightly as necessary to produce images matching those visible

through the microscope. No other types of modiﬁcations were made.

RESULTS
M. Injecting FB into the RPM and either CT-FITC or DY into the TPN
revealed that TPN and RPM motoneurons are interdigitated in the T17-S1
motoneuron nucleus (Figure 1). Positively labeled motoneurons (FB, CT-FITC,

and DY) were located throughout the rostral-caudal extent of T17 and SI but

25

 

Figure 1. Photomicrographs of muscles and motoneurons from adult male anoles
labeled with ﬂuorescent tract tracers. (A) FB+ motoneuron and (C) corresponding
labeled ﬁbers from FB-injected RPM. (B) CT-FITC+ motoneuron and (D)
corresponding labeled ﬁbers from CT-FlTC-injected TPN. Photomicrographs (A-
D) were taken from the same animal and the motoneurons were located in
adjacent tissue sections (in spinal segment T17). TPN and RPM muscle ﬁbers
run perpendicular to each other, so cross-sections of the tail produce cross-
sections of RPM ﬁbers and longitudinal sections of TPN ﬁbers. (E)
Photomicrograph of a coronal section through spinal segment T17 of an adult
male anole. The India ink used to mark each segment is visible on the external
surface of spinal cord, on the left side of the photomicrograph, and the central
canal is visible in the top right corner. T17 motoneurons are located in the ventral
horn (identiﬁed with arrow). (F) Fluorescent-labeled cells on one Side of an
adjacent section to that depicted in panel (E), demonstrating FB+ and DY+
motoneurons in T17. Panels (E) and (F) are from a different animal than panels
(A-D). Scale bar = 10pm for (A) and (B); 95 pm for (C-E); 20pm for (F).

26

were not found in any other spinal segment (T15, T16 or 82); no labeled cells
were found outside of the ventral horn. Unilateral injection of FB into the RPM
with DY into the ipsilateral TPN (a) did not reveal any double-labeled
motoneurons (Figure 1F) and (b) demonstrated that both the TPN and RPM
receive only ipsilateral motoneuron projections.

Study 2a. In adult males, unilateral injection of FB into either the TPN,
RPM, or CF revealed only ipsilateral labeling of motoneurons, whereas SC
injections resulted in bilateral labeling of motoneurons. As with Study 1, labeled
motoneurons projecting to each muscle were located throughout spinal segments
T17-S1 with roughly equivalent distributions. Approximately one third of T17-S1
motoneurons each projected to the TPN and CF, whereas the remaining third
projected to either the RPM or the SC. These estimates account for 96% of the
motoneurons in the T17-S1 nucleus in males (Figure 2 and Table 1). The TPN
counts with PE in the present study are comparable to those Obtained with
biocytin as reported in Ruiz and Wade (2002).

Study 2b. In adult females, unilateral injection of FB in the CF resulted in
ipsilateral labeling, and SC injections resulted in bilateral labeling of
motoneurons. Almost half f the T17-S1 motoneurons projected to the CF while
approximately one third projected to the SC. These estimates account for 70%
of the motoneurons in the T17-S1 nucleus in females (Table 1).

Comparison of the number of T17-S1 motoneurons (on one Side of the

Spinal cord) between males (Study 2a) and females (Study 2b) demonstrates that

27

 

Figure 2. Photomicrographs of F8 injection sites into (A) TPN, (B) RPM, (C) CF
and (D) SC of adult male anoles. Corresponding FB+ motoneurons in the T17-S1
nucleus are Shown in the insets. Scale bar = 360pm for (AD) and 30pm for
insets.

28

TPN RPM CF SC

Count f/_o Count ﬂ/g Count % Count _°_/g

Males 52 31 32 13 66 26 30 8
90 44 40 19 82 32 46 9

92 39 y 46 38

 

Females N/A N/A

 

Table 1. Counts of fast blue-labeled motoneurons in T17-S1 following injection
into either one transversus penis (T PN), one retractor penis magnus (RPM), one
caudifemoralis (CF), or the cloacal sphincter (SC). Percentages of labeled
motoneurons, based on counts from thionin-stained alternate sections, are also
indicated. Means for each muscle are outlined in grey. All cells labeled following
TPN, RPM, and CF injections were unilateral, so the percent labeled cells was
based on thionin counts from the same side of the spinal cord. With SC
injections, labeled cells were detected on both sides of the cord, so they were
compared to the total number of thionin-stained motoneurons found bilaterally;
SC counts represent an average of the two sides of the spinal cord. All data
points come from separate individuals (see Materials and Methods).

29

300

250 4

200 -

150-

100 4

50-

 

   

Number of T17-S1 motoneurons

Males Females

Figure 3. Total number (mean +/- SEM) of Nissl-stained motoneurons in the T17-
81 nucleus from one side of the spinal cord. Asterisk indicates a statistically
signiﬁcant difference between males and females.

30

adult males have more motoneurons in this region than adult females (t = 5.77; p
< 0.001; Figure 3).
DISCUSSION

Summary

In adult male anoles, the sexually dimorphic T17-S1 nucleus is comprised
of motoneurons projecting to muscles directly involved in copulation (T PN and
RPM) as well as two other muscles (CF and SC; present data; Ruiz and Wade,
2002). More than half of this motoneuron population projects to the copulatory
muscles (Table 1). Adult female anoles do not possess either TPN or RPM
muscles, but like males, their T17-S1 nucleus projects to the CF and SC.
Unilateral Injection of a retrograde tracer into the TPN, RPM, and CF all resulted
in ipsilateral labeling of motoneurons. Injection of the SC, which wraps around
the cloaca, results in bilateral labeling of motoneurons. For all muscles (T PN,
RPM, CF, and SC), labeled motoneurons were located in the lateral ventral horn
and were interdigitated throughout spinal segments T17-S1; speciﬁc sub-
populations were not Obviously segregated by medial-lateral or rostral-caudal

position.

Organization of T17-S1 motoneuron projections

Retrograde tracer injections into the TPN, RPM, CF, and SC resulted in
labeling of the vast majority of T17-S1 motoneurons in males (Table 1),
suggesting that these four muscles comprise all major muscle targets for this

motoneuron population. However, in adult females, while FB injections into the

31

CF and SC labeled approximately two-thirds of the T17-SI motoneurons, not all
were accounted for (T able 1). This discrepancy between males and females may
be attributed to variability between the sexes in either the degree to which tracer
ﬁlled the respective muscles or in transport efﬁcacy. However, analysis of the
injection sites revealed no obvious sex difference in the extent to which each
muscle (CF and SC) was ﬁlled with FB. Also, a sex difference in transport
efﬁcacy may exist, but if so, it is likely due to a factor other than circulating
testosterone levels; gonadal androgens do not affect axonal transport of another
retrograde tracer, cholera toxin-conjugated horseradish peroxidase (Leslie et al.,
1991)

Another possibility is that females have motoneurons in the T17-S1
nucleus that project to an additional muscle(s), as yet unidentiﬁed. Other leg and
tail muscles are located near the CF and SC, including the retractor medialis, the
retractor lateralis, and the protractor commissurae (Arnold, 1984). Perhaps the
best candidate is the iliocaudalis tail muscle (ILC; also called the ﬂexor caudae
extemus; Akita, 1992). This muscle is present in both sexes and lies adjacent to
the CF (in both sexes) and TPN (in males), and the ILC and TPN are both
supplied by the same nerve in male Iguanas (Akita, 1992). Additional projections
in T17-S1 in females would be particularly interesting. AS virtually all (96%) of
these motoneurons have been identiﬁed in males with TPN, RPM, CF, and SC
injections, it would mean that females have T17-S1 projections that are minimal

or non-existent in males.

32

Alternatively, female anoles may have residual motoneurons in the T17-
S1 nucleus that do not actually project to muscles. Female lizards do develop
hemipenes and the corresponding musculature embryonically, however, in some
species, including anoles, these structures regress by hatching (Dufaure and
Hubert, 1961; Raynaud and Pieau, 1985; Holmes and Wade, in press; Chapter
5). It would be exciting if a subset of the corresponding motoneurons were
maintained into adulthood without target musculature (perhaps via glial support),
which may be consistent with previous reports in mammals (Glicksman et al.,
1998). In at least one other reptile species, leopard geckos (Eublepharis
macularius), adult females can develop fully evertable hemipenes when treated
with testosterone (Rhen et al., 1999). The degree to which this hemipenis
plasticity is accompanied by corresponding development of TPN and RPM
musculature and motoneurons is currently under investigation. One possibility is
that female leopard geckos maintain copulatory motoneurons into adulthood,
which then support function of the remainder of the system when it develops (as
opposed to having adult motoneuron genesis, which would be exciting but quite
surprising).

Finally, given that both females and males have motoneurons projecting to
the CF and SC, it appears that the sex difference in T1 7-81 motoneuron number
(present data; Ruiz and Wade, 2002) is largely due to females not having TPN or
RPM motoneurons (but see above). One prediction that arises from this fact is
that females should have higher percentages of cells projecting to CF and SC

given their lower number of total T17-S1 motoneurons, and indeed they do,

33

although the effect is certainly more striking for SC (27% in females versus 9% in

males) than CF (43% in females versus 32% in males) projections (T able 1).

Implications for structure/function relationships within this system

The ipsilateral projections to the TPN and RPM are consistent with the
organization of copulatory behavior in this species. Male anoles use only one
hemipenis per copulation (Crews, 1978), suggesting that each is independently
controlled by its own neuromuscular system. Furthermore, over relatively short
periods of time, male anoles alternate hemipenis use in order to maximize sperm
transfer (T okarz, 1988; Tokarz and Slowinski, 1990; Tokarz and Kirkpatrick,
1991 ); green anoles in particular rarely use individual hemipenes more than two
times in succession (Crews, 1978). While there is evidence to suggest that
sensory feedback from the hemipenes and/or testes mediates this effect in part
(Crews, 1978), it may be that muscular fatigue also participates in this behavioral
pattern.

The bilateral labeling of motoneurons observed following injection of the
SC is consistent with at least two avian species (Japanese quail and Brown
Leghorn fowl), in which, similar to the present data, motoneurons projecting to
the SC are located in the pelvic region of the spinal cord (Ohmori and Watanabe,
1989; Seiwert and Adkins-Regan, 1998). This bilateral labeling may be because
the SC receives bilateral projections with axons crossing the midline.

Alternatively, the SC may receive only ipsilateral projections, but they could not

34

be detected because the muscle wraps around the cloaca and all injections ﬁlled
ﬁbers on both sides.

Finally, interdigitation of T17-S1 motoneurons projecting to different
muscles is consistent with nerve organization in other reptilian and avian species.
In male Iguanas, RPM, CF, and SC muscles all receive projections via the same
nerve trunk, which branches into the pudendal nerve (serving SC and RPM) and
the nerve serving the CF (Akita, 1992). Furthermore, SC and CF motoneurons
are interdigitated and may share the connexus caudalis nerve in domestic fowl
(Ohmori and Watanabe, 1989). This organization of the nerve is not surprising
given the close proximity of these muscles within the lizard tail as well as the fact
that, arguably, all four muscles are used (either directly or indirectly) in
copulation. While the TPN and RPM are directly involved in the control of the
hemipenis during lntromission, the CF is likely involved in postural positioning
during mounting and the SC may also be involved in facilitating lntromission,
either by passively or actively permitting the hemipenis to extend through the

cloaca.

Comparison to organization of other dimorphic neuromuscular systems

In male anoles, more than half of T17-S1 motoneurons project to the TPN
and RPM, both muscles critical for male copulatory behavior. The remaining
motoneurons project to muscles that are not directly involved in copulation: the
CF (a leg muscle) and SC. This composition is reminiscent of spinal nuclei in

laboratory rats in which the dorsolateral nucleus and spinal nucleus of the

35

bulbocavemosus, both sexually dimorphic, have motoneurons projecting to both
copulatory and non-copulatory muscles (e.g. Schroder, 1980; McKenna and
Nadelhaft, 1986), including external sphincters involved in waste elimination.
Interestingly, in the neuromuscular systems underlying vocalization in ﬁsh, frogs,
and songbirds, while populations of motoneurons within brain nuclei can project
to different muscle targets, those projecting to sexually dimorphic muscles are
usually clustered, rather than interdigitated, populations (Nottebohm et al., 1976;
Bass, 1985; Simpson et al., 1986).

Sex differences in neuromuscular systems are often not limited to
motoneuron somata and muscle size. For example, they can include
dimorphisms in dendritic arborization, neuromuscular junction size,
neurotransmitter release, and synaptic efﬁcacy (reviewed in Breedlove et al.,
2002). Indeed, in the neuromuscular system that extends the anole dewlap
during courtship, not only do adult males have larger motoneurons and muscle
ﬁbers, they also have longer cartilage, more muscle ﬁbers, larger nerve cross
sectional area, and larger neuromuscular junctions than adult females (O’Bryant
and Wade, 1999; O’Bryant and Wade, 2002b). Now that organization of the
anole copulatory neuromuscular system is better understood, we are primed to
uncover the full extent of dimorphisms in this set of structures and the
mechanisms regulating them. This work can then be related to that from
copulatory systems of mammalian species. lmportantly, it will also allow direct
comparisons between copulatory neuromuscular systems and those regulating

courtship, traditionally studied in non-mammalian vertebrates. In anoles, both

36

types of dimorphic neuromuscular systems are required for the full suite of

masculine reproductive behaviors.

37

CHAPTER TWO

Holmes MM, Wade J (2004). Seasonal plasticity in the copulatory neuromuscular
system of green anole lizards: a role for testosterone in muscle but not
motoneuron morphology. J Neurobiol 60: 1-11.

38

INTRODUCTION

Sexually dimorphic neuromuscular systems exhibit sensitivity to
androgens in adulthood in diverse vertebrate Species. For example, vocal
courtship behavior and morphology of the neural and muscular components
underlying it are regulated by androgens in ﬁsh, frogs, and birds (e.g. Kelley and
Pfaff, 1976; Hannigan and Kelley, 1986; Sassoon et al., 1987; Tobias et al.,
1991; Brantley et al., 1993b; Brenowitz and Lent, 2002; Girgenrath and Marsh,
2003). Muscles and motoneurons in the copulatory system of laboratory rodents
also enlarge in response to androgens (Breedlove and Arnold, 1981; Balice-
Gordon et al., 1990; Rand and Breedlove, 1992, 1995) and, in parallel, are
plastic in species exhibiting naturally occurring seasonal ﬂuctuations in
testosterone (T; Forger and Breedlove, 1987; Hegstrom and Breedlove, 1999;
Hegstrom et al., 2002).

Green anoles (Anolis carolinensis) are lizards that breed seasonally, from
about April to July, in the southeastern United States. Males court females by
performing a suite of head bobbing displays coupled with the extension of a
bright red throat fan, called a dewlap. Copulation occurs when a male ﬂips his tail
under a female's and inserts one of his two copulatory organs (called hemipenes)
into her cloaca. T in males is about twice as high during the breeding than the
non-breeding season (Lovem et al., 2001), and is important for the expression of
reproductive behaviors; castration of males decreases their frequency and T

replacement prevents or reverses this effect (Mason and Adkins, 1976; Crews et

39

al., 1978; Adkins and Schlesinger, 1979; Winkler and Wade, 1998; Rosen and
Wade, 2000).

The dewlap and the neuromuscular system controlling its extension are
sexually dimorphic. Male dewlaps are approximately 7-fold larger in extended
area compared to those of females (Jenssen et al., 2000). Males extend their
dewlaps more frequently than females (Nunez et al., 1997), even if females are
given doses of T equivalent to males (Winkler and Wade, 1998). Only males use
the dewlap in courtship, although both sexes will extend their dewlaps in
aggressive encounters (Greenberg and Noble, 1944; Jenssen and Nunez, 1998;
Nunez et al., 1997). Parallel to the behavioral dimorphism, the cartilage,
motoneurons, nerve, neuromuscular junctions, and muscle ﬁbers required for
dewlap extension are all larger in males than females (Wade, 1998; O'Bryant and
Wade, 1999; O'Bryant and Wade, 2002b). The morphology of this system does
not appear to vary across season or respond to changes in T (O'Bryant and
Wade, 1 999).

Like the dewlap system, the copulatory neuromuscular system in green
anoles is dimorphic on multiple levels (Ruiz and Wade, 2002). Male anoles
possess bilateral hemipenes in the tail, which are controlled ipsilaterally by two
sets of muscles. The transversus penis (T PN) wraps over the hemipenis and its
contraction everts the hemipenis through the cloaca; the retractor penis magnus
(RPM) is attached to the base of the hemipenis and retracts it back into the tail
(Arnold, 1984). Adult females do not possess hemipenes or either of the two

muscles (Ruiz and Wade, 2002). The motoneurons projecting to the TPN and

40

RPM are located in the pelvic region of the spinal cord in segments trunk 17 and
sacral 1 (T 1 7-S1; Ruiz and Wade, 2002; Holmes and Wade, 2004a; Chapter 1),
and these cells are smaller in size and number in females compared to males
(Ruiz and Wade, 2002).

It was not known whether the copulatory neuromuscular system of male
anoles is plastic in adulthood and which, if any, components of this system are
sensitive to changes in environmental condition and/or androgens. Therefore, the
goals of the present investigation were to determine whether, parallel to the
changes in reproductive behavior, (Experiment 1) seasonal changes in the
morphology of the copulatory system exist in gonadally intact male green anoles

and (Experiment 2) whether such changes are mediated via gonadal androgens.

MATERIALS AND METHODS

Housing conditions and care

Adult male green anoles were purchased from Charles Sullivan Co
(Nashville, TN) at three different times of year and were individually housed in
21-Iiter glass aquaria (42 x 20 x 24cm) for a minimum of 14 days (Experiment 1)
or 7 days (Experiment 2). Lizards purchased during April (the beginning of the
breeding season in the ﬁeld) were housed under environmental conditions
conducive to breeding. Animals were exposed to a 14:10 hour light/dark cycle
using ﬂuorescent, full-spectrum, and heat lamps. Temperature ranged from 28°C
(ambient) to 38°C (directly under the heat lamp over each cage) during the day

and was 18'C at night. These animals all had reproductive testes (see below)

41

and were classiﬁed as breeding season (BS) males. Lizards purchased in
September (after the ﬁeld breeding season) were housed in the same conditions
but had regressed testes (BS-X). This combination of breeding environmental
conditions but regressed testes does not occur naturally in this species, but
allows independent analysis of environmental conditions and reproductive status.
Finally, lizards purchased in November were housed in simulated non-breeding
conditions including a 10:14 hour light/dark cycle, with temperature ranging from
24'C (ambient) to 30'C (directly under heat lamp) during the day and 150 at
night. These animals had regressed testes and were classiﬁed as non-breeding
season (NBS) males. Thus we compared three experimental conditions: BS, BS-
X, and NBS. In all cases, aquaria were sprayed daily to help maintain 70%
relative humidity and water was provided ad Iibitum. Lizards were fed crickets or
mealworrns three times a week (BS and BS-X) or twice a week (NBS) as outlined

in Lovem et al. (2004a).

Treatment and tissue collection

Experiment 1- Gonadally Intact Males. Adult males (n=8 per group) were
overdosed with Brevital Sodium, weighed, measured snout to vent (SVL), and
perfused with phosphate buffered saline (PBS) and 10% phosphate buffered
formalin (PBF). Spinal column tissue was post-ﬁxed in PBF for one hour, and
segments were marked with ink to permit identiﬁcation. The cords were then
extracted (segments T15-82), embedded in gelatin, post-ﬁxed overnight in 20%

sucrose in PBF, and sectioned frozen in the coronal plane at 30pm. Every other

42

section was mounted out of 9:1 dHZOzPBS onto gelatin-coated slides, stained
with thionin, dehydrated and coverslipped using Perrnount.

The kidneys and rostral tail (containing the hemipenes, TPNS and RPMS)
were post-ﬁxed in Bouin’s ﬁxative for seven days. To provide better exposure of
the tissues of interest to ﬁxative and parafﬁn (see below), only the ventral half of
the tail was maintained. The tissues were then soaked in 70% ethanol overnight,
dehydrated, cleared in xylene, embedded in parafﬁn, and sectioned at 10pm. The
tail tissue was stained with the trichrome method, and kidneys were stained with
hematoxylin and eosin. In lizards, renal “sex segments” perform functions similar
to the mammalian prostate and enlarge in response to androgen, thus they
provide a bioassay for T exposure (e.g. Cueller et al., 1972; Crews, 1980;
Winkler and Wade 1998).

Experiment 2 — Testosterone manipplation. Lizards were anesthetized using
isoﬂurane and were bilaterally gonadectomized (GDX) while on ice. Condition of
the testes (size, color, and vascularization) and vasa deferentia was noted.
During the same surgery, each lizard was implanted subcutaneously with a
Silastic implant (7mm long x 0.76mm ID x 1.65mm OD) containing either T
proprionate (5mm packed) or left empty (n=8 per group) as in O'Bryant and
Wade (1999). Twenty-one days following castration, lizards were overdosed with
Brevital Sodium, body mass and SVL were noted, and they were perfused with
PBS and PBF. Conﬁrmation of the implant (including presence/absence of
steroid) was done at the time of perfusion. Spinal cord, tail, and kidney tissue

were all processed as described for Experiment 1.

43

Tissue Analyses

All measurements were conducted in the same manner for Experiments 1
and 2 without knowledge of experimental group or hormone manipulation
(Experiment 2 only).

Motoneurons were analyzed as in Ruiz and Wade (2002). Brieﬂy, the
soma size of twenty randomly selected motoneurons was measured through the
rostral-caudal extent of T17-S1 on each of the right and left sides using NIH
Image. As a control, motoneuron soma Size was measured for twenty
motoneurons using the same technique in spinal segment T16; these
motoneurons do not project to the TPN or RPM (Ruiz and Wade, 2002; Holmes
and Wade, 2004a; Chapter 1). The cross-sectional area of twenty-ﬁve randomly
selected RPM ﬁbers was measured per Side (dispersed throughout the muscle).
Caudifemoralis (CF) ﬁber size was measured using the same technique as a
control; it is a leg muscle lying largely in the tail and is not speciﬁcally involved in
copulation. Cross-sectional area of each hemipenis was determined in ten
sections, approximately 50pm apart through the rostral ~500pm of the structures.
The TPN was not analyzed as the ﬁbers run perpendicular to those of the RPM
and CF, so their cross-sectional area could not be determined in this tissue.
Finally, the height of four renal sex segment epithelial cells was measured from _
each of 10 tubules randomly selected across the two kidneys (resulting in 40

measures; Winkler and Wade, 1998).

For all tissues except kidneys, a mean was calculated separately for each
side in each individual. As the left and right sides did not differ in any case (data
not shown), analyses reported below reﬂect an average of both sides for each
structure (i.e., from a total of 40 motoneurons for both T17-$1 and T16, 50

muscle ﬁbers for both RPM and CF, and 20 sections for hemipenes).

Statistical Analyses

Experiment 1. Each measure was analyzed by one-way ANOVA with
experimental condition as the independent variable. Post hoc analyses were
performed using Tukey-Kramer tests. One male was excluded from the analyses
because, although his testes appeared reproductive (i.e., large, white and
vascularized), his renal sex segment cell height was more than two standard
deviations below the mean and his vasa deferentia were regressed at the time of
perfusion, suggesting he was not responsive to circulating T. Thus, the ﬁnal
sample size in the BS group was 7, compared to n=8 in the BS-X and NBS
groups.

Experiment 2. Each measure was analyzed by two-way ANOVA with
experimental condition and androgen treatment as independent variables. Post
hoc analyses, to test effects among the three conditions (BS, BS-X, and NBS),
were performed using Tukey-Kramer tests; planned comparisons were
performed to test effects of T within each experimental condition. Four males
were excluded from analyses: one BS+GDX male because his renal sex

segment cell height was more than two standard deviations above the mean,

45

 

suggesting that his T levels were not representative of a castrated male; one
BS+GDX+T male because the implant fell out one week prior to perfusion; and
two males died prior to perfusion (one NBS-I-GDX and one NBS+GDX+T). Thus,
measurements were obtained from 7 or 8 animals per group. All analyses were

performed using Statview (SAS Institute).

RESULTS
Experiment 1
Motonepron soma Size. T17-S1 motoneuron somata were smaller in NBS males
than in BS or BS-X males (F2, 20 = 3.64; p = 0.045; Figure 1A), although Tukey-
Kramer tests failed to detect statistically signiﬁcant differences. In the control
region (T16), the effect of experimental condition was not statistically signiﬁcant.

However, the same pattern was detected (F2, 20 = 3.22; p = 0.061; Figure 1B).

Peripheral copplatory structures and general measures. Experimental condition
did not affect RPM (F2, 20 = 0.40; p = 0.677; Figure 1C) or CF ﬁber size (F2, 20 =
2.42; p = 0.115; Figure 10), nor did it affect hemipenis cross sectional area (F219
= 0.80; p = 0.463; Figure 1E). However, renal sex segment cell height was
increased in BS compared to BS-X and NBS males (F2, 20 = 148.95; p < 0.001;
both Tukey-Kramer p < 0.05; Table 1). No Signiﬁcant differences among groups
were detected for body mass (F2, 20 = 0.01; p = 0.994) or SVL (F2, 20 = 0.33; p =
0.724).

46

 

 

 

A (g 300 B gr 300
3 250 3' 250
g 200 g zoo
a 150 n 150
II I!
g 100 g 100
U! W
5 50 o 50
l- 0 F 0
BS Bs-x NBS "' Bs 33.x "as
C i 300 D A 3000
1 250 2500
,3 200 0 2000
2 150 a 1500
h
.8 100 1000
E II. 500
E o 0 0
BS Bs-x NBS as Bs-x NBS
E 5E 1.0
1.4
512
a .
g 1.0
g 0.9
'E 0.6
a 0.4
E 0.2
o 0.0
I as Bs-x NBS

Figure 1. Effects of experimental condition on (A) T17-S1 motoneuron soma size,
(B) T16 soma size, (C) RPM muscle ﬁber Size, (D) CF ﬁber size, and (E)
hemipenis cross sectional area in gondally intact males. In panel A, a main effect
was detected without any signiﬁcant pairwise comparisons (see Results).

_3_S_ 5.3.95 N_BS

 

 

RSS height (um) 31.30 (1.36)* 10.52 (0.83) 11.53 (0.57)
Body mass (9) 5.30 (0.29) 5.31 (0.21) 5.34 (0.25)
SVL (mm) 64.57 (0.65) 65.0 (0.93) 63.94 (1.14)

Table 1. Mean (+/- SEM) renal sex segment (RSS) cell height, body mass and
snout-vent length (SVL) for gonadally intact in Experiment 1. * indicates
statistically different from both other groups on same measure.

48

Experiment 2

Motoneuron soma size. As in Experiment 1, experimental condition affected T17-
81 motoneuron soma size (F2, 33 = 8.87; p < 0.001; Figure 2A). NBS and BS
(both Tukey-Kramer p < 0.05) males had smaller somata than BS-X males.
Androgen treatment did not affect T17-S1 soma size (F1, 33 = 0.10; p = 0.757),
and the interaction between experimental condition and T-treatment was not
statistically signiﬁcant (F2, 39 = 0.32; p = 0.730). Experimental condition affected
T16 motoneuron soma size (F2, 33 = 10.09; p < 0.001; Figure 23), although
androgen treatment did not (F1, 39 = 1.83; p = 0.184). As with T17-SI
motoneurons, NBS and BS (both Tukey-Kramer p < 0.05) males had smaller T16
somata than did BS-X males. Again, no signiﬁcant interaction between condition

and androgen was present (F2, 38 = 0.82; p = 0.446).

Muscle ﬁber size. RPM ﬁber size was unaffected by experimental condition (F2, 37
= 1.07; p = 0.354; Figure 20), and this variable did not signiﬁcantly interact with
androgen treatment to alter RPM size (F2, 37 = 0.98; p = 0.387). However, T-
treatment did Signiﬁcantly increase RPM ﬁber size (F1, 37 = 14.52; p < 0.001;
Figure 3). Planned comparisons reveal that GDX+T males had larger RPM
muscle ﬁbers than GDX males in the BS (p = 0.025) and BS-X (p = 0.002)
conditions but not in the NBS condition (p = 0.387). Both androgen treatment (F1,
37 = 5.68; p = 0.022) and experimental condition (F2, 37 = 4.47; p = 0.018)
signiﬁcantly affected CF ﬁber size (Figure 2D), but they did not interact (F2, 37 =
0.05; p = 0.951 ). BS males had smaller CF muscle ﬁbers than BS-X males

49

 

350 gr 350 * EOE]
GDX+T
300 g 300 [-—
250 3 250
200 5 200
150 g 150
100 8 100
so ,9 50
‘-
0 I- 0 :
BS Bs-x N38 88 BS-X NBS
GDX —_
a a 2 EST

9

N
O
O

a
0
CF ﬁber size (pm’)
9
O

0

BS BS-X NBS BS BS-X NBS

:1 GDX
— GDXOT

M

or
D)
D)

2.0

O-lfl
"IO“

Hemipenis area (mm2)rrl RPM fiber size (pmz) O T17 soma size (pmz)>
O u
'° 8

BS BS-X NBS

Figure 2. Effects of experimental condition and testosterone manipulation on (A)
T17-S1 motoneuron soma size, (B) T16 soma size, (C) RPM muscle ﬁber size,
(D) CF ﬁber size, and (E) hemipenis cross sectional area in castrated males
(GDX = with blank implant; GDX+T = with testosterone implant). ' indicates
signiﬁcantly different from other experimental groups (among BS, BS-X and
NBS). ‘ indicates a signiﬁcant difference between GDX and GDX+T males within
an experimental group. In panel D, a main effect of testosterone manipulation
was detected without any signiﬁcant pairwise comparisons; an effect of
experimental condition stems from a signiﬁcant difference between the BS and
BS-X groups only (see Results).

50

 

Figure 3. Photomicrographs of cross-sections of RPM muscle ﬁbers (A, B, C) and
hemipenes (D, E, F) taken from BS males. A and D are from a gonadally intact
male, B and E are from castrated males with blank capsules, and C and F are
from a castrated male with a testosterone implant. Arrows point to the dorsal
edge of each hemipenis, where it meets the caudifemoralis muscle. Scale bar =
25pm (A, B, C) and 800pm (D, E, F).

51

.coEucoo .mEoEtoaxo 0E8 555, 3.65

._.+xoO ucm x00 comics cocoaoEu 2.85ch a $2865 a .mcoEucoo .050 53 Eat moococoEo 9:85:96
mow—855 9. new .x.wm Eo: cococotﬁ ESE—.9» o $52.35 .. .mcoaucoo .mEoEtoaxo «cocoa mcomtmano
Lon. .N EOEtoaxw _.__ ccmaﬁ 0:998an 59> u ._.+xoO ucm_aE_ {.53 59> u x009 mo_mE c.2958

.2 3>wv £95. Eoiaocm new moon. Son .290; =8 Awwmv EoEmoxm B» .95. 22mm 4+9 coo—2 .N .2an

598$ 80988 45938 2.5.98.8 239:8 8998.8 “15191.5. m.
E934 $98.4. 8092.... $934. 98.934 28.984 jams 8m

LSESR 6498.2 .8988 63:42 3993.94 9.92.3 gqlalle E a; mmm
IIIITxoo Ixolo El+onO xIIoo IIIEBO IlIxoo

m2 XIwm mm

52

(I’ukey-Kramer p < 0.05) but not NBS males (p > 0.05). GDX+T males overall
had slightly, but signiﬁcantly, smaller CF muscle ﬁbers than controls. Despite this
main effect of androgen treatment, no signiﬁcant effects of T on CF ﬁber size

were detected within any of the 3 conditions with planned comparisons (all p >

0.183).

Hemipenis cross-sectional area. A trend for breeding season conditions to
increase hemipenis size was detected (F2, 33 = 2.86; p = 0.070; Figure 2E), as
was a trend for an interaction between experimental condition and androgen (F2,
33 = 3.02; p = 0.061 ). While overall GDX+T males had larger hemipenes than
GDX controls (F1. 33 = 35.18; p < 0.001; Figure 3), planned comparisons
demonstrate that this is true for the BS (p < 0.001) and BS-X (p < 0.001) groups,
but not the NBS group (p = 0.179).

General measures. Androgen treatment (F 1, 37 = 480.17; p < 0.001) and breeding
environmental conditions (F2, 37 = 24.53; p < 0.001) increased renal sex segment
cell height (Table 2). A signiﬁcant interaction between experimental condition and
androgen treatment was also detected (F2, 37 = 9.58; p < 0.001) with T-treatment
increasing cell height more in breeding (BS and BS—X) than non-breeding (NBS)
conditions. Despite random selection of individuals, those in the BS-X condition
had signiﬁcantly increased SVL (F2, 33 = 10.43; p < 0.001) compared to either BS
or NBS (both Tukey-Kramer p < 0.05) males. T treatment did not affect SVL (F1,

33 = 1.99; p = 0.166), and no statistically signiﬁcant interaction was detected (F2,

53

33 = 2.38; p = 0.106; Table 2). T-treatment (F1, 33 = 4.50; p = 0.041) and breeding
conditions (trend: F2, 33 = 3.10; p = 0.057) both increased body mass without a
signiﬁcant interaction (F2, 33 = 1.25; p = 0.299; Table 2). BS-X males were heavier
than BS (T ukey-Kramer p < 0.05) but not NBS males. These group differences in
body mass were also present prior to surgery (data not shown) demonstrating

that they are not due to the treatment.

DISCUSSION

Summary

The results from Experiment 1 demonstrate that environmental condition
can alter soma size in sexually dimorphic spinal motoneurons (T17-S1); males
housed in non-breeding conditions had smaller somata than did males housed in
breeding conditions. A trend for the control motoneurons (T 16) to be smaller in
this group also existed, suggesting that the effect of non-breeding conditions on
soma size may not be speciﬁc to copulatory motoneurons. No seasonal variation
was detected in any other measure in gonadally intact animals, with the
exception of renal sex segment cell height, conﬁrming that circulating androgen
levels vary across season (Lovem et al., 2001). Results from Experiment 2
reveal that neither T17-S1 nor T16 motoneurons are sensitive to changes in
circulating T in adults. However, while both hemipenis and RPM ﬁber size
increased in response to T, this effect was only detected in males housed under
conditions simulating the breeding season. In contrast, T slightly, but

signiﬁcantly, decreased CF muscle fiber size across the three experimental

conditions, indicating that the trophic effect of T on copulatory stmctures was
relatively speciﬁc. Taken together, the data demonstrate that adult morphological
plasticity is regulated by different mechanisms in copulatory motoneurons
compared to the hemipenes and RPMs. While modest changes in motoneuron
soma size might occur seasonally on a general level, environmental cues

mediate T-induced changes in peripheral copulatory structures.

Plasticity in motoneuron soma size

The decrease in copulatory motoneuron soma size during the non-
breeding season is consistent with reports in mammals (Forger and Breedlove,
1987; Hegstrom et al., 2002). However, in contrast to the two mammalian
species studied to date (the white-footed deer mouse, Peromyscus Ieucopus,
and the Siberian hamster, Phodopus sungorus), seasonal plasticity in the soma
size of anole copulatory motoneurons appears independent of T. That is, while
circulating T in the green anole is approximately twice as high during the
breeding compared to the non-breeding season (Lovem et al., 2001), T-
treatment did not increase T17-S1 soma size in castrated males housed in either
condition. As season may also affect non-copulatory motoneurons (T 1 6), the
somewhat smaller soma size in the non-breeding season is likely a relatively
general effect of variations in environmental condition. Indeed, physical activity,
which is higher during the breeding season in the anole (Jenssen et al., 1995;

Jenssen et al., 1996), mediates motoneuron soma size; decreased activity can

55

be associated with smaller somata in some neural populations (e.g. Nakano et
al., 1997; Nakano and Katsuta, 2000; but see Breedlove, 1997a).

In Experiment 2, both BS and NBS males had smaller motoneurons than
BS-X males. While the decrease in NBS soma size may relate to seasonal
factors (as in Experiment 1), the difference between BS and BS-X males is likely
due to smaller SVL and body weight (Table 2). These differences, not seen in
Experiment 1, are an unfortunate artifact of random assignment of males to
experimental groups. Indeed, if T17-S1 soma size is corrected for body weight or
SVL, then the signiﬁcant difference between BS and BS-X groups is no longer
present (soma size/body weight, Tukey-Kramer p > 0.05; soma size/SVL, Tukey-
Kramer p > 0.05), whereas the difference between BS-X and NBS groups is
maintained (soma size/body weight, Tukey-Kramer p < 0.05; soma size/SVL,
Tukey-Kramer p < 0.05). In any case, the point of interest in the motoneuron data
from Experiment 2 is that androgen manipulation failed to alter soma size
regardless of environmental condition, suggesting that, in contrast to mammalian
species (both seasonal breeders and laboratory rats), the soma size of
copulatory motoneurons is relatively insensitive to androgen in adult anoles.
However, other structural (e.g., dendritic morphology; Rand and Breedlove,
1995) or functional (e.g., neurotransmitter release; Tobias and Kelley, 1995)
properties that were not evaluated in the present experiment, may be altered by

T.

Plasticity in androgen sensitivity in hemipenes and RPM muscle ﬁbers

56

While the size of peripheral copulatory structures does not change
seasonally in gonadally intact male anoles, the RPMs and hemipenes exhibit
seasonal plasticity in their sensitivity to androgen manipulation. That is, T-
treatment dramatically increased hemipenis and RPM ﬁber size, but only when
animals were housed in environmental conditions conducive to breeding (long
days and warm temperatures). One possible explanation for the seasonal
variation in responsiveness to T is that it serves to protect the copulatory system
from morphological regression during periods of lower circulating androgens.
During the breeding season (BS), the reduction in peripheral structures in GDX
compared to intact males is consistent with this idea (compare Figures 16 and E
to Figures 20 and E). However, unmanipulated males still have approximately
10ng/ml of plasma T during the non-breeding season (compared to ~20ng/ml
during the breeding season; Lovem et al., 2001), which may be sufﬁcient to
maintain the morphology (consistent with results from the gonadally intact males
in Experiment 1). Alternatively, the increased dynamic range under breeding
environmental conditions (BS and BS-X) may facilitate the response of the
system to even relatively small T ﬂuctuations that may occur with particular
experiences (e.g., exposure to females; reviewed in Hart, 1983).

The dose of T used in the present experiments is likely supraphysiological
given that renal sex segment cell height is larger in males treated with T than
gonadally intact breeding males (compare Tables 1 and 2). While physiological
variations in T may not be substantial enough to induce morphological changes

in the RPMs and hemipenes of gonadally intact males, the potential for

57

substantial plasticity cleady exists in these structures. And, perhaps more
interesting is the manner in which this potential for response to T varies across
environmental conditions. This pattern is similar to alterations in sensitivity to T in
the rodent brain. Neuron soma size in the medial amygdala of Siberian hamsters
increases in response to T-treatment only when animals are housed in long
photoperiods (Cooke et al., 2002). The mechanism for the changes in sensitivity
is presently unknown but may be mediated by differences in androgen receptors;
their distribution in the copulatory structures of anoles must be investigated.
Similar to seasonal changes in sensitivity to T, lizards show reduced
responsiveness to adrenocorticotropin hormone (Carsia and John-Alder, 2003)
and melatonin (Bertolucci and Foa, 1998) during the non-breeding season.
Indeed, melatonin is essential for normal circadian and seasonal rhythmicity in
lizards (e.g., Bertolucci and Foa, 1998; Bertolucci et al., 2002; Foa et al., 2002),
including green anoles (Hyde and Underwood, 1995; Hyde and Underwood,
2000), and may mediate activity of other hormones or independently create
morphological change. For example, in the Siberian hamster, increased
melatonin release facilitates gonadal regression and decreased androgen
production (Kelly et al., 1994). This pathway appears to induce shrinkage of
copulatory motoneuron soma and muscle ﬁber size (Hegstrom and Breedlove,
1999; Hegstrom et al., 2002). However, photoperiod decreases neuromuscular
junction size independent of T levels, consistent with a more direct role for

melatonin (Hegstrom et al., 2002).

58

Different mechanisms in the copulatory vs. courtship neuromuscular systems

In adulthood, the sexually dimorphic neuromuscular system controlling
dewlap extension exhibits little or no plasticity in response to season or androgen
manipulation (O’Bryant and Wade, 1999), at ﬁrst glance suggesting that
morphological changes in this system do not parallel seasonal changes in
behavior. However, while the frequency of extension is greatly reduced, both
males and females do use their dewlaps in the non-breeding season occasionally
(mostly for aggressive interactions; Jenssen et al., 1996). It is possible that this
limited use requires or facilitates maintenance of the stmctures. The seasonal
changes in copulatory motoneurons and sensitivity to androgen in peripheral
copulatory structures also relates to behavior; male anoles do not copulate under
short days and lower temperatures even when treated with T (O’Bryant and

Wade, 2002a).

Comparisons to other species

Neuromuscular systems involved in male reproductive behaviors are
sexually dimorphic in diverse vertebrate species, and gonadal steroids often
mediate the maintenance of these dimorphisms (e.g. Kelley, 1986; Brantley et
al., 1993b; Wade, 1999; Breedlove et al., 2002). For example, adult T-
manipulation modulates morphology of the neuromuscular portion of the avian
song system (e.g. Arnold, 1980; DeVoogd et al., 1991; Wade and Buhlman,
2000). Similarly, the mass of external oblique muscles of gray tree frogs (used in

sound production) is larger in breeding males than in females and non-breeding

59

males and is increased by T (Girgenrath and Marsh, 2003). Auditory sensitivity in
female midshipman ﬁsh changes seasonally to match differences in male sound
production (Sisneros and Bass, 2003), and may be due to gonadal steroids
(Fonano et al., 2003). It is unknown, however, whether the structures underlying
vocal production exhibit seasonal morphological plasticity. Finally, photoperiod
and androgens enhance the copulatory neuromuscular system of seasonally
breeding mammalian species (Forger and Breedlove, 1987; Hegstrom et al.,
2002). Collectively these data demonstrate that sex and seasonal differences in
morphology often parallel function and, furthermore, that T often mediates
morphological changes in adulthood. Likewise, components of the anole
copulatory system and the behaviors they facilitate are highly sensitive to T, but
the responsiveness depends on environmental cues. It will be important to
determine the speciﬁc mechanisms underlying this effect and whether they are

involved in other sexually dimorphic vertebrate systems.

60

CHAPTER THREE

Holmes MM, Wade J. Testosterone regulates androgen receptor
immunoreactivity in the copulatory, but not courtship, neuromuscular system in
adult male green anoles. J Neuroendo: submitted.

61

INTRODUCTION

Androgens commonly regulate the expression of masculine reproductive
behaviors in a wide variety of vertebrate groups, including ﬁsh (Brantley et al.,
1993c), frogs (Wetzel and Kelley, 1983), birds (Arnold, 1975; Harding et al., 1983),
and mammals (Hart, 1967). While the motivation for such behaviors is mediated by
hypothalamic and preoptic brain regions (Hull et al., 2002), the execution of
courtship and copulatory behaviors requires neuromuscular systems, which are
frequently sexually dimorphic (Breedlove et al., 2002). Androgens serve not only to
activate the behaviors in adulthood, but also often mediate the development and
maintenance of the underlying neuromuscular components (Breedlove et al., 2002).
These effects on motoneuron and muscle morphology in particular usually occur by

direct effects of androgenic steroids on the androgen receptor (AR) (Breedlove et al.

2002). As such, determining the distribution and regulation of AR is critical for
understanding mechanisms that regulate structural and functional change in
reproductive neural circuits.

Green anoles (Anolis carolinensis) are a particularly useful model species
with which to study steroid hormone effects on structure and function because they
possess two sexually dimorphic neuromuscular systems. One is involved in
courtship behavior, which includes extension of a red throat fan called a dewlap.
While both sexes possess dewlaps, they are approximately seven-fold larger in
males than females (Jenssen et al., 2000), and males extend their dewlaps far more
frequently than females (Greenberg and Noble, 1944; Nunez et al., 1997; Jenssen

and Nunez, 1998). Paralleling the sex difference in size and overall use of the

62

dewlap, the neuromuscular components that mediate its extension are larger in
males than females. Specifically, extension of the dewlap is caused by contraction of
the bilateral ceratohyoideus (CH) muscles, which lie on the ventral surface of the
throat and between the ﬁrst ceratobranchial and ceratohyal cartilages (see
schematic in Wade, 2005). Contraction of the CH causes the cartilages to act as a
lever, ultimately bowing out the second ceratobranchial cartilages, which are found
at the midline just under the skin on the ventral surface of the throat (Bels, 1990;
Font and Rome, 1990). CH muscle ﬁbers are larger in size and number and the
second ceratobranchial cartilages are longer in males than females (O’Bryant and
Wade, 1999; Lovem et al., 2004b; Chapter 4). In addition, the CH muscles receive
projections from two pools of motoneurons in the caudal brainstem: the vagal portion
of the nucleus ambiguus (AmbX) and the glossopharyngeal region of nucleus
ambiguus and the ventral motor nucleus of the facial nerve (AmbIXNllmv) (Font,
1991; Wade, 1998). These motoneurons are larger in size, but not number, in males
than females (O’Bryant and Wade, 1999).

The other sexually dimorphic neuromuscular system is involved in male
copulatory behavior. Male anoles possess two intromittant organs called hemipenes,
which lie inside the ventral surface of the tail, immediately caudal to the cloaca. Each
hemipenis is controlled primarily by two muscles: the transversus penis (which
causes aversion of the organ), and the retractor penis magnus (RPM; which retracts
the hemipenis back into the tail; Arnold, 1984). After hatching, only males possess
hemipenes and the associated muscles (Ruiz and Wade, 2002; Lovem et al., 2004b;

Chapter 4; Holmes and Wade, in press; Chapter 5). Contraction of the muscles

63

occurs via the innervation of ipsilateral motoneurons located in spinal segments
trunk 17 and sach 1 (T17-S1). The motoneurons in this region of the spinal cord
are a mixed pool, projecting to both copulatory (in males only) and non-copulatory
(caudifemoralis and cloacal sphincter in both males and females) muscles (Holmes
and Wade, 2004a; Chapter 1). The cells in this population are greater in both size
and number in males than females (Ruiz and Wade, 2002; Holmes and Wade,
2004a; Chapter 1).

As in other species, androgens are important for the expression of both
courtship and copulatory behaviors in male green anoles. These lizards breed
seasonally, from approximately April to July in the Southeastern United States and
testosterone levels are about twice as high in the breeding than non-breeding
season (Jenssen et al., 1995; Jenssen et al., 1996; Lovem et al., 2001).
Testosterone is the most potent hormone for activating male courtship and
copulatory behaviors. Castration reduces their expression, while treatment with
testosterone prevents or reverses these effects (Mason and Adkins, 1976; Crews et
al., 1978; Adkins and Schlesinger, 1979; Winkler and Wade, 1998; Rosen and
Wade, 2000). As dihydrotestosterone also contributes to the expression of male
reproductive behaviors (Crews et al., 1978; Adkins and Schlesinger, 1979; Rosen
and Wade, 2000), while estrogens have little or no effect (Mason and Adkins, 1976;
Crews et al., 1978; Winkler and Wade, 1998), a direct role for AR is likely.

Testosterone can also affect morphology in tissues important for
reproduction, but it does so selectively. The hormone increases RPM ﬁber size and

hemipenis cross-sectional area in males housed in breeding, but not non-breeding,

64

environmental conditions (Holmes and Wade, 2004b; Chapter 2). In contrast, neither
season nor testosterone appears to alter CH ﬁber size (O’Bryant and Wade, 1999).
Of particular interest is that testosterone does not affect motoneuron soma size in
either the courtship or copulatory systems (O’Bryant and Wade, 1999; Holmes and
Wade, 2004b; Chapter 2), which may relate to the fact that there is limited
expression of AR protein and mRNA in the courtship motoneurons (Rosen et al.,
2002). To help determine whether the differential effects of testosterone on the
dewlap and copulatory neuromuscular structures are mediated by AR, we had two
experimental goals. The ﬁrst was to identify which of the other courtship and
copulatory'tissues, including the targets of both systems (cartilages and hemipenes),
express AR. The second was to determine whether AR immunoreactivity changes in
response to season or testosterone manipulation. We evaluated AR
immunoreactivity in adult males housed in either breeding or non-breeding
conditions that were either gonadally intact (Experiment 1) or castrated and treated

with testosterone or vehicle (Experiment 2).

MATERIALS AND METHODS
Housing conditions and care
Wild-caught adult male green anoles were purchased from Charles
Sullivan Co. (Nashville, TN) at two different times of year and were individually
housed in 21-Iiter glass aquaria (42 x 20 x 24cm) for approximately 3 weeks prior
to tissue collection (Experiment 1) or treatment (Experiment 2; see below).

Lizards purchased during May (during the breeding season in the ﬁeld) were

65

housed under environmental conditions conducive to breeding. Animals were
exposed to a 14:10 hour light/dark cycle using ﬂuorescent, full-spectrum, and
heat lamps. Temperature ranged from 28'C (ambient) to 38C (directly under a
heat lamp over each cage) during the day and was 18'C at night. These animals
all had reproductive testes (large and vascularized, see below) and were
classiﬁed as breeding season males. Lizards purchased in October were housed
in simulated non-breeding conditions including a 10:14 hour light/dark cycle, with
temperature ranging from 24'C (ambient) to 30'C (directly under heat lamp)
during the day and 150 at night. These animals had regressed testes (darker,
small, not vascularized) and were classiﬁed as non-breeding season males. For
both breeding and non-breeding season animals, aquaria were sprayed daily to
help maintain 70% relative humidity and food and water was provided as outlined

in (Lovem et al., 2004a).

Treatment and tissue collection

Experiment 1- Gonadally Intact Males. Adult males (n=8 per group) were rapidly
decapitated and spinal columns, tails, throats, and kidneys were extracted, ﬂash
frozen in isopentane chilled on dry ice, and stored at -80 ’0. Condition of the
testes (size, color, and vascularization) and vasa deferentia was noted. All
tissues were frozen sectioned in cross-section in 6 series at 20pm, thaw mounted
onto Superfrost Plus glass slides (Fisher Scientiﬁc) and stored at —80 'C. A
single series of each spinal column, tail, and throat was processed for AR

immunohistochemistry similar to (Rosen et al., 2002). Brieﬂy, slides were

66

warmed to room temperature and ﬁxed for 5 min in 4% paraformaldehyde in
0.1M phosphate buffered saline (PBS). Following a 30 minute incubation in 0.5%
H202 to remove endogenous peroxidase and a 2 hour incubation in 10% normal
donkey serum in 0.1 M PBS with 0.2% Triton X-100, all tissue was incubated for
36 hours in PG21 rabbit polyclonal antibody (1.5 ug/ml for spinal cord tissue and
1.75 ug/ml for throat and tail tissue; Upstate) in 0.1 M PBS with 0.2% Triton X-100
at 4’C. A biotinylated donkey anti-rabbit secondary antibody (1:500 for 90
minutes; Jackson Laboratories), Elite ABC peroxidase reagents (Vector
Laboratories), and nickel-enhanced diaminobenzidine were used to visualize AR.
Slides were rinsed with 0.1M PBS between all steps. Spinal column tissue was
dehydrated and coverslipped with Pennount and muscle tissue (throats and tails)
was stained for 5 minutes with 4’,6-Diamidino—2-Phenylindole (DAPI;1ug/ml;
Sigma) to visualize nuclei, dehydrated and coverslipped with DPX mounting
medium (Fluka). Following quantiﬁcation of AR immunoreactive nuclei (see
below), slides containing spinal cords were placed in xylene overnight to remove
coverslips, and stained with thionin to permit visualization of all motoneurons. For
muscle tissue, adjacent series were stained with hematoxylin and eosin for
conﬁrmation of DAPI estimates for number of AR immunoreactive nuclei (see
below).

Kidneys were stained with hematoxylin and eosin. In lizards, renal “sex
segments” perform functions similar to the mammalian prostate and enlarge in

response to androgen, thus they provide a bioassay for relative levels of

67

androgen exposure (e.g. Winkler and Wade, 1998; Cueller et al., 1972; Crews,
1980).

Qtperimentg— Testosterone manipulation. Lizards were anesthetized using
isoﬂurane and were bilaterally gonadectomized (GDX) while on ice. Condition of
the testes and vasa deferentia was noted. During the same surgery, each lizard
was implanted subcutaneously with a Silastic implant (7mm long x 0.76mm ID x
1.65mm OD) containing either testosterone proprionate (5mm packed) or left
empty (n=8 per group) as in (O'Bryant and Wade, 1999; Holmes and Wade,
2004b; Chapter 2). Twenty-one days following castration, lizards were rapidly
decapitated and tissues were collected as above. Conﬁrmation of the implant
(including presence/absence of steroid) and completeness of castration was also
noted at this time. Spinal column, tail, throat, and kidney tissue were all

processed as described for Experiment 1.

Tissue Analyses

All measurements were conducted in the same manner for Experiments 1
and 2 without knowledge of experimental group.

To estimate the percentage of AR+ nuclei in the CH and RPM muscles,
the number of AR+ and AR- nuclei were determined in 30 randomly selected
muscle ﬁbers (per animal). First, photomicrographs of muscle ﬁbers were taken
using ﬂuorescent illumination to visualize DAPI labeled nuclei. The same ﬁbers
were then photographed using transmitted light to visualize AR+ nuclei.

Comparison of these images (demonstrated in Figure 1) permitted the total

68

 

Figure 1. Photomicrograph of a cross-section through CH muscle ﬁbers from a
gonadally intact male housed in non-breeding environmental conditions
illustrating expression of AR+ nuclei (A). (B) is the same section under
ﬂuorescent illumination to allow visualization of the DAPI nuclear label.
Comparison of the number of nuclei in (A) and (B) provides an estimate for the
total number of AR+ nuclei. Black arrows point to examples of AR+ nuclei and
white arrows point to DAPI labeled nuclei. Scale bar = 30pm.

69

number of nuclei and the total number of AR+ nuclei to be determined. Dividing
total number of AR+ nuclei by the total number of DAPI labeled nuclei provided
an estimate of the percentage that was AR+. In reptilian muscles, nuclei are
located both near the center and on the edge of the ﬁbers (see Figure 1). We
only included centralized nuclei in these analyses to be conﬁdent that we were
evaluating myonuclei (Monks et al., 2004); the luminescence from the ﬂuorescent
DAPI labeling extends slightly beyond the nuclei, and the edge of ﬁbers cannot
be conﬁdently determined because they are so dark. Hematoxylin and eosin was
then used to conﬁrm the DAPI counts and more accurately identify nuclei that
were close to the edge of a given ﬁber as well as those in other cell types
between the ﬁbers. In this case, the number of muscle ﬁbers as well as the total
number of hematoxylin stained and AR+ nuclei were estimated using
Stereoinvestigator (Microbrightﬁeld, Inc) in a single cross-section for each muscle
with approximately 15 sampling areas per section. Nuclei were divided into three
categories: centralized (myonuclei), edge (possibly myonuclei, satellite cells or
ﬁbroblasts), or those clearly outside the muscle ﬁbers (most likely in ﬁbroblasts).
Dividing the number of AR+ nuclei by the total number of hematoxylin-stained
nuclei provided estimates for the percentage of AR+ nuclei for each category. AR
immunoreactivity was not evaluated in transversus penis muscle ﬁbers because
they mn perpendicular to those of the RPM and therefore could not be viewed in
cross-section in this tissue.

To estimate the percentage of AR+ T17-S1 motoneurons, the number of

motoneurons with AR+ nuclei was counted. Following staining with thionin (see

70

above), the total number of motoneurons was estimated using the physical
dissector technique (Gunderson, 1986; Ruiz and Wade, 2002; Holmes and
Wade, 2004a; Chapter 1) where a cell is counted if its nucleus comes into focus
and disappears within a given section. Estimates of AR+ motoneurons and total
cell number were obtained for both sides of the spinal cords (through the rostro-
caudal extent of T17-S1) and averaged. As with muscle nuclei, dividing the total
number of AR+ motoneurons with the total number of motoneurons provided an
estimate of the percentage of AR+ cells. We did not measure AR expression in
the courtship motoneurons because they have previously been shown to express
little or no AR (Rosen et al., 2002) and do not respond morphologically to
androgen or environmental (season) manipulations (O'Bryant and Wade, 1999).

The neuromuscular target structures were also evaluated for AR
immunoreactivity. Presence or absence of detectable immunoreactivity was
evaluated for the three bilateral pairs of cartilage that control dewlap extension:
the ﬁrst ceratobranchials, the second ceratobranchials, and the ceratohyals and
in each of three components of the hemipenes: the lobe epithelium, the body of
the lobe, and the clavulae (which provide structural support for the lobes during
intromission).

Finally, the height of four renal sex segment epithelial cells was measured
from each of 10 tubules randomly selected across the two kidneys (resulting in
40 measures; Winkler and Wade, 1998) to evaluate relative levels of circulating

androgens.

71

Statistical Analyses

For Experiment 1, percent AR+ myonuclei from the DAPI analyses,
percent AR+ motoneurons, and total number of muscle ﬁbers were each
analyzed in intact animals (between breeding and non-breeding males) with
independent t-tests. Estimates of percent AR+ nuclei using hematoxylin and
eosin was analyzed with a one-way repeated-measures ANOVA with season as
the between subjects variable and nucleus type (centralized, edge, or outside of
muscle ﬁber) as a within-subjects measure. For Experiment 2, percent AR+
myonuclei from the DAPI analyses, percent AR+ motoneurons, and total number
of muscle ﬁbers were analyzed by separate two-way ANOVAs with season and
androgen treatment as independent variables. Estimates of percent AR+ nuclei in
the muscles using hematoxylin and eosin were analyzed using two-way
repeated-measures ANOVAs with season and androgen-treatment as between
subjects variables and nucleus type (centralized, edge, or outside of muscle
ﬁber) as a within-subjects measure. Post hoc analyses were performed using

Tukey-Kramer tests. All analyses were performed using Statview (SAS Institute).

RESULTS
Experiment 1: Gonadally intact males
The percent of AR+ nuclei in the RPM muscle did not differ between
breeding and non-breeding males based in either the DAPI (t = 0.43; p = 0.68;

Figure 2) or hematoxylin and eosin (F = 0.24; p = 0.63; Table 1) analyses.

72

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

RPM

A 1°°i RPM B we ,, ,, 99:.
5 so 5
g . g 80‘
r: c
+ ‘0. 'l- 00*
5 9%
E 40‘ g 404
§ 2... in.
o. 3

o- o .

BS NBS NBS

9 CH 0 «oo- 23..
2 . '3
3 ’° 7» 0°:
F 2
+ 604 4
9 2 ‘°
g 20‘ g I
o. 3 2°

0- o

 

 

 

 

 

 

 

NBS

 

Figure 2. Mean (+/- SEM) percent of AR+ nuclei in the RPM (A and B) and CH (C
and D). These values were obtained from DAPI labeled tissue. Panels (A) and
(C) are from gonadally intact males (grey bars) and panels (B) and (D) are from
castrated males (GDX) treated with testosterone (T ; black bars) or empty
capsules (white bars). Animals were housed either in breeding (BS) or non-
breeding (NBS) environmental conditions. Asterisks indicate a statistically
signiﬁcant main effect of testosterone treatment on the percentage of AR+ nuclei.

73

.o_oEo> 59> 9980.: 3.9: x00 2 umﬁano 0809 «.29: 99858 5 AC 2998283 he «come 58:. 28558
a 9855 98:22 80.83.: Sim can 10 05 £3 .8 23¢ o_om:E 35 he 32850 .0 done 9: co 93:8 65 5
mm 3538.0 995 5.032 .338: ems—Em :88 can 5.886.:2 Eat Eon: +m< .6 E833 Aﬁmwiv 2822 .F 338...

.889 8.88
8.9 58
.89 8.8
8.9 8.88

8.9 8.3
8.89 8.9.

m. .e._l..s=o

8.89935
8. F 9 8.88

.8. F 9 «.8
8.9 8.8

8.89 8.2
88.9 8.2

rule 8

mm

.8898:

8.89 8.8.
.89 4.8
:89 :8

8.89 98
3.49 8.:

.1250

8.9 8.3
5.9 «.8
8.9 8.8
8.89 o. E

6.3 «.mm
8.8 $60

0 elleso

8.89 8.?
8.9 4.2
8.9 8.83

8. F 9 8.8

8.9 8.8
8.89 8.5

ale 8

I

8.89 8.:
8.9 8.2
8.9 2K

8.5 8.8

8.89 8.88
8.9 8.88

5. Eco.

k+xomvlwmz
xowlwmz
P+X0mvlwm
xowuwm

«lscoetmdmm

wmz
mm

8 Ewe... ._._. o.m.|xm

74

Interestingly, AR immunoreactivity signiﬁcantly differed among different types of
nuclei (F = 4.70; p = 0.02; Table 1), where the percentage was higher in
centralized and edge nuclei compared to those completely outside of the muscle
ﬁbers (both p < 0.05), although no signiﬁcant interaction was detected between
season and nucleus type (F = 0.29; p = 0.75). The number of RPM muscle ﬁbers
did not differ between breeding and non-breeding males (331-404 ﬁbers on
average; t = 1.00; p = 0.33). AR immunoreactivity in T17-S1 motoneurons also
did not differ between breeding and non-breeding males (on average, 75.32% for
breeding males and 75.98% for non-breeding males; t = 0.23; p = 0.82; Figure 3),
and while AR immunoreactivity was present in all three components of the
hemipenes (Figure 4), the intensity of staining did not obviously differ across
season.

Similar to the RPM, AR expression in the CH muscle did not signiﬁcantly
differ between breeding and non-breeding males in either the DAPI (t=2.11; p =
0.06; Figure 2) or hematoxylin and eosin (F = 2.63; p = 0.13; Table 1) analyses,
and the interaction between nucleus type and season was not statistically
signiﬁcant (F = 0.77; p = 0.47). However, unlike the RPM, AR immunoreactivity
was equivalent across nucleus type (centralized, edge, and outside of ﬁbers) (F =
1.25; p = 0.30; Table 1). CH muscle ﬁber number did not differ between breeding
and non-breeding males (295-300 ﬁbers on average; t = 0.22; p = 0.83). AR
labeling was distinct in the nuclei of both the ﬁrst and second ceratobranchial
cartilages, but was very faint in the ceratohyals (Figure 5) and did not obviously

change across season in any case.

75

 

Figure 3. Photomicrographs of cross-sections through the ventral horn in spinal
segment T17 from a gonadally intact male (A), castrated male (B). and castrated
male treated with testosterone (C), all housed in breeding environmental
conditions. The majority of motoneurons have nuclei that exhibit dark AR
immunoreactivity; the percentage of AR+ cells not differ between experimental
groups. Scale bar = 50pm.

76

 

Figure 4. Photomicrographs of cross-sections through one hemipenis of a
gonadally intact male housed in breeding conditions. Panel (A) is a low
magniﬁcation photo of the entire cross-sectional area of the hemipenis. Panels
(B-D) are high magniﬁcation photos of the lobe epithelium (B), lobe body (C). and
clavula (D) from the hemipenis in (A). Scale bar = 600um for (A) and 50pm for
(B-D).

77

 

Figure 5. Photomicrographs of cross-sections through the three cartilages
involved in dewlap extension: the second ceratobranchials (A), ﬁrst
ceratobranchials (B), and ceratohyals (C). Androgen receptor immunoreactivity is
present and dark in many of the nuclei in (A) and (C) but is absent or light in the
nuclei in (B). The ribbon-like appearance of the ﬁrst ceratobranchials in (B)
reﬂects the ossiﬁed nature of this component of the dewlap system; the second
ceratobranchials (A) and ceratohyals (C) are hyaline cartilage (13, 14). All tissue
is from a castrated male treated with testosterone and housed in breeding
environmental conditions. Scale bar = 30pm.

78

Renal sex segment cell height was larger in BS than NBS males (t =
14.26; p < 0.01) suggesting that these males indeed had higher levels of

circulating androgens (Figure 6).

Experiment 2: Testosterone manipulation

Testosterone treatment signiﬁcantly increased the percent of AR+ RPM
nuclei in DAPI-labeled tissue (F = 9.99; p < 0.01 ), although this measure did not
differ across season (F = 1.19; p = 0.29), and these variables did not interact (F =
0.23; p = 0.64; Figure 2). Similarly, testosterone treatment signiﬁcantly increased
the percent of AR+ nuclei in hematoxylin-labeled tissue (F = 14.58; p < 0.01)
while the effect of season (F = 0.25; p = 0.63) and the treatment x season
interaction (F = 3.24; p = 0.08) failed to reach signiﬁcance (Table 1). As in
Experiment 1, the percent of AR+ nuclei was signiﬁcantly different among
nucleus types (F = 11.20; p < 0.01); centralized and edge nuclei had higher
percentages of AR+ nuclei compared to those outside RPM ﬁbers (both p <
0.05). Total number of RPM muscle ﬁbers did not differ between groups (287-388
ﬁbers on average; effect of season: F = 3.79; p = 0.06; effect of testosterone: F =
0.55; p = 0.47; interaction: F = 0.0001; p = 0.99). The percent of AR+
motoneurons did not differ across season (F = 0.72; p = 0.41) or with
testosterone treatment (F = 0.17; p = 0.69), and these variables did not
signiﬁcantly interact (all means between 71.6 — 77.7%; F = 1.55; p = 0.23; Figure
3). As with Experiment 1, AR immunoreactivity was present in all components of

the hemipenes (Figure 4) but did not obviously differ across groups.

79

 

 

 

 

    

 

 

 

 

A A *
E40: 340 253:“
a
5 .r:
a *
£330. '83"
g 1:
52M * 82°
3 3
310. 310.
u? 3. u‘} 0-
BS NBS

   

Figure 6. Mean (+l- SEM) cell height (pm) of epithelial cells in the renal sex
segment in (A) gonadally intact males (grey bars) and (B) castrated males
(GDX) treated with testosterone (T; black bars) or an empty capsule (white bars).
Animals were housed either in breeding (BS) or non-breeding (NBS)
environmental conditions. Asterisks indicate a statistically signiﬁcant difference
between BS and NBS males in (A) and a main effect of testosterone treatment in
(B). A statistically signiﬁcant main effect of season, as well as a signiﬁcant
season x testosterone interaction, is also present in (B).

80

In contrast to the RPM, testosterone treatment did not alter the percent of AR+
DAPI labeled nuclei in the CH muscle (F = 0.17; p = 0.90; Figure 2). Season also
did not alter this measure (F = 0.003; p = 0.96), and the treatment x season
interaction was not statistically signiﬁcant (F = 0.24; p = 0.63). These results
were replicated with hematoxylin and eosin stained tissue (Table 1), where
neither testosterone treatment (F = 1.71; p = 0.20) nor season (F = 0.51; p =
0.48) altered AR immunoreactivity in the CH. The treatment x season interaction
also failed to reach signiﬁcance (F = 0.24; p = 0.63). While the percent AR+
nuclei did not differ among CH nucleus types in Experiment 1, the percent of AR+
nuclei was signiﬁcantly different among nucleus types (F = 4.30; p = 0.02) in
Experiment 2; the percent of AR+ centralized nuclei was higher than AR+ nuclei
outside of CH muscle ﬁbers (p < 0.05). Total number of CH muscle ﬁbers did not
differ between groups (300-318 ﬁbers on average; effect of season: F = 0.73; p =
0.40; effect of testosterone: F = 0.08; p = 0.78; interaction: F = 0.04; p = 0.85). As
in Experiment 1, AR immunoreactivity was present in both the ﬁrst and second
ceratobranchial cartilages, was very faint in the ceratohyals (Figure 5), and did
not obviously differ between groups.

Renal sex segment cell height was signiﬁcantly larger in breeding than
non-breeding males (F = 6.55; p = 0.02) and in testosterone-treated males than
castrated controls (F = 54.89; p = 0.001). Season and hormone treatment did
signiﬁcantly interact (F = 4.38; p = 0.048) whereby testosterone increased cell

height more in breeding than non-breeding males (Figure 6).

81

DISCUSSION

Summary

The present results demonstrate that, in addition to the target tissues
(dewlap cartilages and hemipenes), a large percentage of nuclei in the dewlap
(CH) and copulatory (RPM) muscles express AR. Expression was consistent
across season in both muscles in gonadally intact males (Experiment 1) as well
as in castrated males treated with testosterone or vehicle (Experiment 2).
Testosterone increased the percentage of AR+ nuclei in the RPM, but not the
CH, regardless of season. In contrast to the dewlap motoneurons, which express
relatively little AR protein or mRNA (approximately 44% of the cells in Amb X and
none in Amb lXNllmv; Rosen et al., 2002), the majority of T17-S1 motoneurons
exhibit AR immunoreactivity. Yet, neither season nor testosterone manipulation

altered the percentage of AR+ cells in this population.

Comparison to testosterone eﬁects on morphology

While testosterone is the most potent hormone in the activation of both
male courtship and copulatory behaviors in this species (Mason and Adkins,
1976; Crews et al., 1978; Adkins and Schlesinger, 1979; Winkler and Wade,
1998; Rosen and Wade, 2000), it has disparate effects on the underlying
sexually dimorphic structures, the muscles and peripheral tissues in particular.
We do not yet know what factors contribute to sex differences in the
motoneurons in either system. However, testosterone increases RPM ﬁber and

hemipenis size in the adult copulatory neuromuscular system (Holmes and

82

Wade, 2004b; Chapter 2), and androgens stimulate the survival of these
structures before hatching (Holmes and Wade, in press; Chapter 5). In the
dewlap system, testosterone manipulation does not appear to alter CH muscle
ﬁber size in adult males (O’Bryant and Wade, 1999). However, androgens are
involved in the masculinization of the CH muscles and second ceratobranchial
cartilages in juveniles (Lovem et al., 2004b; Chapter 4). AR expression should
be investigated during development, but the present results from adults are
consistent with the idea that the morphological changes may involve a
testosterone-induced increase in AR expression. Testosterone only increases
RPM ﬁber size and hemipenis size when animals are housed in breeding
environmental conditions (Holmes and Wade, 2004b; Chapter 2), although the
hormone increases AR expression in both breeding and non-breeding animals,
suggesting that other variables likely also contribute to the morphological effects
(see below).

Taken together, these data suggest a variety of possible mechanisms that
might mediate the effects of testosterone on the expression of courtship and
copulatory behaviors in this species. In addition to the role the hormone plays in
enhancing or maintaining structure of the peripheral copulatory system (likely via
up-regulation of AR), it may also act on the receptors that are abundant in the
limbic forebrain (Rosen et al., 2002) to facilitate motivation to display the
behaviors (Morgentaler and Crews, 1978; Greenberg et al., 1984). Testosterone
may also directly inﬂuence the execution of courtship and/or copulation by

modulating the functions of the motoneurons in either system as well as the CH

83

muscle, in the absence of morphological changes in these particular structures.
Electrophysiological and neurochemical studies would begin to address these
issues. As the motoneurons, hemipenes and RPM all express AR, it will also be
important to determine the site(s) of action for testosterone-induced changes in
morphology of the copulatory muscles and target structures. While testosterone
may act directly on RPM ﬁbers to exert its trophic effects, other mechanisms may
be involved. For example, this growth may instead be a response to testosterone
directly increasing the size of the hemipenes, or it may be that the motoneurons
somehow exert an anterograde effect on the muscles even though they do not

show an increase in soma size themselves.

Comparison to other species: muscle

AR is commonly expressed in vertebrate muscle ﬁbers, particulariy in
muscles that are sexually dimorphic (Monks et al., 2004; Dorlochter et al., 1996;
Fischer et al., 1995; Veney and Wade, 2004). The present data demonstrate that
testosterone treatment increases AR expression in lizard muscle ﬁbers that
exhibit morphological sensitivity to testosterone. While this testosterone-induced
increase in the percent of AR+ nuclei in adulthood is consistent with previous
reports in rats (Monks et al., 2004) and humans (Sinha-Hikim et al., 2004), a
direct link between AR expression and morphological responsiveness is less
clear across species. In male Xenopus, widespread AR immunoreactivity is
detected in skeletal muscle ﬁbers, including those that do not change in size in

response to testosterone manipulation (Dorlochter et al., 1996). Furthermore, in

the sexually dimorphic and androgen sensitive larynx, three weeks of androgen
treatment in adulthood causes a decrease in AR mRNA expression (Fischer et
al., 1995), although expression of AR protein and mRNA are not always
complementary (e.g., Rosen et al., 2002). Finally, while testosterone does
increase AR immunoreactivity in rat muscles that exhibit morphological
responsiveness to androgens, it also does so in those that do not (Monks et al.,
2004). For both the levator ani (increases in size in response to testosterone)
and extensor digitomm Iongus (does not change in size in response to
testosterone), the percent of AR+ nuclei is increased by testosterone relative to
castrated control males in both myonuclei and ﬁbroblasts (Monks et al., 2004).
However, as in rats, the effects of testosterone were consistent across different
types of nuclei for both CH and RPM muscles in green anoles in the present
experiment. AR expression did not change in any of the three classes of nuclei in
the CH muscle but was increased by testosterone in all of them in the RPM
muscle, suggesting that multiple cells types are responsive to androgens in this
muscle and may contribute to the structural and/or functional changes induced by
these steroids. Taken together, these data suggest that sensitivity of a given
muscle to testosterone is not simply dictated by the distribution of AR. While
functional AR are clearly required for androgen-induced effects (reviewed in
Breedlove et al., 2002), presence of the AR protein does not necessarily confer

morphological responsiveness to testosterone to a given cell.

Comparison to other species: central nervous system

85

AR is also commonly expressed in both the brain and spinal cord, and, as
in muscle, androgens often increase AR expression in reproductive neural
circuits. For example, testosterone increases AR expression in the medial
preoptic nucleus, medial amygdala, lateral septum, and bed nucleus of the stria
terminalis (posteromedial subdivision) in hamsters (Meek et al., 1997), in limbic
system extracts (detected in Western blots) in mice (Lu et al., 1999), as well as in
sexually dimorphic motoneurons in frogs (Perez and Kelley, 1996) and rats
(Freeman et al., 1995). In the present report, a large portion (approximately 70-
80%) of the T17-S1 motoneurons were AR+, although testosterone manipulation
did not alter the percentage of these cells expressing the protein. As stated
above, this lack of change in immunoreactivity parallels the fact that testosterone
does not alter the soma size of these cells in this species. Either this population
of motoneurons is tmly unresponsive to testosterone, which seems unlikely, or
the receptors are involved with functions in these cells other than changes in
soma size. This dissociation between AR expression and morphological
responsiveness to testosterone is similar to what occurs in a subset of the
motoneurons in the spinal nucleus of the bulbocavemosus and dorsolateral
nucleus of the rat copulatory neuromuscular system. Both nuclei are mixed
populations in which a portion of cells exhibits testosterone-induced increases in
soma size while the remaining cells do not (Collins et al., 1992). Motoneurons
projecting to the bulbocavemosus, levator ani, and ischiocavemosus muscles
increase in size in response to testosterone (Collins et al., 1992) and

approximately 95% of these cells are AR+ (Jordan, 1997). However, the

86

remaining motoneurons project to the external anal sphincter and external
urethral sphincter and these cells do not exhibit morphological responsiveness to
testosterone, yet the majority of them also express AR (70% for the external anal
sphincter and 90% for the external urethral sphincter; Jordan, 1997).

While we did not see any effect of season on AR in lizard copulatory
motoneurons, some precedent exists for neural AR expression to ﬂuctuate in
concert with naturally occurring changes in testosterone levels. AR mRNA is
distributed in several brain regions important for reproductive behaviors
(including the medial preoptic area, external nucleus of the amygdala, and the
ventromedial hypothalamus) in male and female leopard geckos (Rhen and I
Crews, 2001). Testosterone mediates both male and female reproductive
behaviors in this species (Rhen et al., 1999), and AR mRNA expression in the
female brain is increased during late vitellogenesis, when females are receptive
and plasma testosterone levels are more than an order of magnitude greater
than previtellogenesis (Rhen et al., 2000; Rhen et al., 2003). Similarly, increased
androgen receptor expression is seen in the brains of goldﬁsh (Pasmanik and
Callard, 1988), Atlantic croaker (Sperry and Thomas, 1999), and Gambel's white-
crowned sparrows (Soma et al., 1999), corresponding to time points with

naturally occurring increased circulating testosterone.
Conclusions

AR immunoreactivity can, but does not necessarily, correspond to the

effects of androgens on morphology within a given tissue. Testosterone up-

87

regulates AR expression in RPM muscle ﬁbers, which increase in size in
response to this steroid (Holmes and Wade, 2004b; Chapter 2). Yet, a high
percentage of T17-S1 motoneurons (approximately 70—80%), as well as CH
muscle ﬁbers (approximately 60-70%), which do not appear to change in size in
response to testosterone (O’Bryant and Wade, 1999), express AR. These
structures may certainly be sensitive to testosterone in additional physiological
and/or morphological capacities that have not yet been investigated, but it is
clear that the expression of AR does not alone determine whether morphological
change will occur in response to testosterone. Additional factors must contribute
to the effects of androgens on these systems in adulthood. Taken together, not
only are the morphological effects of androgens often complex both within and
across species, but understanding the site of action of these hormones involves
more than simply characterizing the distribution of AR. Ultimately, insight into the
general principles of androgenic effects on the structure and function of sexually
dimorphic circuits will come both from characterizing the distribution of AR in the
tissues of interest and identifying how additional transcription factors, including

steroid hormone receptor co-activators, are involved.

88

CHAPTER FOUR

Lovem MB, Holmes MM, Fuller CO, Wade J (2004). Effects of testosterone on

the development of neuromuscular systems and their target tissues involved in
courtship and copulation in green anoles (Anolis carolinensis). Horm Behav 45:
295-305.

89

INTRODUCTION

Sexual dimorphism in behavioral function and associated morphological
structure is a hallmark of reproduction in gonochoristic vertebrates. For example,
in species such as plainﬁn midshipman ﬁsh (Porichthys notatus; Brantley et al.,
1993a; Brantley and Bass, 1994; Knapp et al., 1999), African clawed frogs
(Xenopus Iaevis; Sassoon and Kelley, 1986; Kelley and Dennison, 1990; Kelley
and Tobias, 1999), and zebra ﬁnches (Taeniopygea guttata; Bottjer et al., 1985;
Wade et al., 2002), adult males produce courtship vocalizations that are not
produced by females, and the neuromuscular systems controlling them show
male-biased dimorphisms that begin differentiating in development. Similarly, the
neuromuscular system controlling penile function during copulation in rodents
shows a number of male-biased dimorphisms that become apparent shortly after
birth (Cihak et al., 1970; Breedlove and Arnold, 1980; Nordeen et al., 1985).

Often, gonadal steroids play a direct role in regulating the differentiation of
sexually dimorphic neuromuscular systems as well as their peripheral, target
tissues (e.g., reviewed in Cooke et al., 1998; Breedlove et al., 2002). Androgen
production by juvenile male African clawed frogs increases after metamorphosis,
which results in masculinization of components of the male song system
including increased motoneuron and muscle ﬁber number (Kelley, 1986;
Sassoon and Kelley, 1986; Kay et al., 1999). Similarly, just after birth, male rats
produce androgens that masculinize the copulatory neuromuscular system by
preventing cell death of the motoneurons and muscle ﬁbers targeting the penis

(Breedlove and Arnold, 1980, 1983a, b; Nordeen et al., 1985). However, early

90

exposure to testosterone (T) or its metabolites may not be necessary or sufﬁcient
for the sexual differentiation of some neuromuscular systems. The role of
steroids in sexual differentiation of the zebra ﬁnch song system appears quite
complicated (reviewed in Arnold, 1997; Wade, 2001). For example, during
development androgens can masculinize motoneuron soma size and syrinx (the
vocal organ) weight in females, and estrogens can feminize the syrinx in males
(Wade et al., 2002). However, the antiandrogen ﬂutamide does not prevent
masculinization, and the aromatase inhibitor fadrozole does not prevent
feminization (Wade et al., 2002). Thus, the neuromuscular systems involved in
courtship and copulation are frequently sexually differentiated in development,
but comparative study suggests that the underlying mechanisms that regulate
this differentiation may be diverse.

In the present study, we examine the development of neuromuscular
systems and target tissues involved in courtship and copulation in the green
anole, Anolis carolinensis. This lizard has a polygynous social organization,
breeding seasonally from approximately April through July, with females laying
single-egg clutches over the entire four month breeding season (see Lovem et
al., 2004a for natural history overview). Hatchlings emerge during the summer
and early fall months and reach sexual maturity in the following breeding season,
approximately 6-9 months later. Both adult males and females communicate via
stereotyped head bobbing displays that may be given with or without the
extension of a red throat fan called a dewlap (DeCourcy and Jenssen, 1994;

Jenssen et al., 2000). Males use these head bobbing displays with dewlap

91

extension in territorial advertisement and in aggressive and sexual interactions,
whereas females use them in aggressive (with dewlap extension) and sexual
(without dewlap extension) interactions and overall at much lower rates
(Greenberg and Noble, 1944; Nunez et al., 1997; Jenssen and Nunez, 1998;).
During courtship, the male approaches the female while head bobbing and
prominently displaying his extended dewlap. If the female is receptive, she
remains stationary and gives head bob displays in return — but without dewlap
extension — often adopting a characteristic neck-bend posture (e.g., Crews,
1980). The male than bites the female on the back of the neck as he mounts and
intromits one of two bilateral, independently controlled penises (called
hemipenes), characteristic of lizards and snakes (Pough et al., 2001).

The anatomy involved in dewlap extension, like the courtship behavior
described above, is sexually dimorphic. Two pools of motoneurons in the caudal
brainstem regulate dewlap extension. One is located in the vagal portion of
nucleus ambiguus (AmbX), and the other in the glossopharyngeal region of
nucleus ambiguus and the ventral motor nucleus of the facial nerve
(AmbIXNllmv) (Font, 1991; Wade, 1998). These motoneurons innervate the
bilaterally symmetrical ceratohyoid muscles, which upon contraction cause the
2"" ceratobranchial cartilage to bow out, extending the dewlap (Bels, 1990; Font
and Rome, 1990). Extended dewlap area is approximately 7-fold greater in adult
males than in adult females (Jenssen et al., 2000). Underlying this dramatic sex
difference are male-biased dimorphisms in 2"" ceratobranchial cartilage length,

ceratohyoid muscle weight and ﬁber cross-sectional area and number,

92

neuromuscular junction size, motoneuron soma size, and nerve cross-sectional
area (Wade, 1998; O’Bryant and Wade, 1999, 2002b).

Similar to the anatomy involved in dewlap extension, the hemipenis
neuromuscular system is highly sexually dimorphic. Each hemipenis is controlled
by motoneurons in the ipsilateral trunk segment 17 and sacral segment 1 of the
caudal spinal cord (T 1 7-S1) (Ruiz and Wade, 2002; Holmes and Wade, 2004a;
Chapter 1). These motoneurons innervate the paired transversus penis (T PN)
muscles, which evert the hemipenes, and the retractor penis magnus (RPM)
muscles, which retract them (Arnold, 1984; Ruiz and Wade, 2002; Holmes and
Wade, 2004a; Chapter 1). In adults, the hemipenes, TPN, and RPM are present
only in males (Ruiz and Wade, 2002). The T17-S1 motoneurons are present in
both males and females, due to the fact that a subset of the cells in this nucleus
innervates targets possessed by both sexes (a leg muscle and the cloacal
sphincter“, Ruiz and Wade, 2002; Holmes and Wade, 2004a; Chapter 1).
Nevertheless, overall the motoneurons are larger and more numerous in males
(Ruiz and Wade, 2002; Holmes and Wade, 2004a; Chapter 1).

The dewlap system (i.e., the 2"d ceratobranchial cartilage, the ceratohyoid
muscles, and AmbX and AmleNlImv motoneurons) is sexually monomorphic at
hatching. However, between 60 and 90 days post-hatching, both cartilage length
and muscle ﬁber cross-sectional area have become larger in males, although
motoneuron soma size has not (O’Bryant and Wade, 2001). This period of
differentiation occurs when plasma T levels in juvenile males are elevated

compared to females (Lovem et al., 2001), suggesting a potential causal

93

relationship. Less is known about sexual differentiation of the hemipenis system
(i.e., the hemipenes, TPN and RPM muscles, and T17-S1 motoneurons). In
lizards, both males and females develop hemipenes as embryos, but they
regress in females prior to hatching (e.g., Dufaure and Hubert, 1961 ). The
muscles and motoneurons have not been examined during development.

The objective of the present study was to determine the effect of juvenile T
manipulation on the dewlap and hemipenis systems in green anoles. Post-
hatching males ﬁrst have plasma T levels that are signiﬁcantly higher than those
of females at approximately 30 days of age (Lovem et al., 2001). Thus, we
manipulated juvenile T exposure at this time by giving gonadally intact females
either blank or T implants, and by castrating or sham-castrating juvenile males.
Then, at 90 days post-hatching, we euthanized all juveniles and assessed
treatment effects on both the dewlap and hemipenis systems. We predicted that
females given T implants would develop larger, more masculine, dewlap systems
compared to females given blank implants, and that castrated males would
develop smaller, more feminine, dewlap systems compared to sham-castrated
males. As described above, the hemipenes are sexually differentiated by
hatching, much earlier than the dewlap system. However, whether this
differentiation is reversible was unknown, and the associated muscles and
motoneurons had not been examined during development. Therefore, in addition
to investigating the effects of T on the sexual differentiation of the dewlap
neuromusculature, we addressed the degree to which steroid-mediated plasticity

may exist in the hemipenis system.

94

MATERIALS AND METHODS

Animals and housing

Our established laboratory breeding colony provided us with juveniles for
this study. Brieﬂy, adult (3 50 mm snout-vent length, SVL) A. carolinensis were
purchased in the breeding season (Charles Sullivan Co., Nashville, TN) and kept
in the laboratory in cages consisting of one male and 3-7 females under
conditions conducive to breeding (e.g., Lovem et al., 2004a). Breeders were kept
on a 14:10 h lightzdark cycle using a combination of ﬂuorescent, ultraviolet, and
incandescent lights. Ambient temperature ranged from 18°C at night to 28-38°C
during the day depending on location in each cage relative to a basking light.
Relative humidity was set at 70%. Nest boxes were placed into each cage and
checked daily for eggs, which were incubated individually in sealed plastic cups
containing a 1:1 vermiculitede2O mixture. Incubation temperature averaged
28°C (range 27-29°C); hatching occurred at a mean of 34.1 days (SE=0.6).

Hatchlings were sexed by post-anal scale size (males have two enlarged
post-anal scales, females do not), their SVL and mass were recorded, and they
were toe-clipped for permanent individual identiﬁcation. Groups of up to 20
juveniles were housed in 110 L aquaria furnished with a peat moss substrate and
multiple dowels, rocks, and artiﬁcial vegetation for perching, basking, and hiding.
The environmental conditions were the same as above and approximate those
found in the ﬁeld during the summer. Juveniles had access to water in shallow
dishes ad Iibitum. their cages were misted daily, and they were fed daily a diet of

vitamin-dusted crickets and vestigial-winged fruit ﬂies.

95

Experimental design

Manipulations were conducted according to NIH guidelines and with
approval of the Michigan State University All-University Committee on Animal
Use and Care. At 30 days post-hatching, juveniles were again measured and
weighed, after which they were treated in one of four ways. Females were given
either blank or T implants and males were either castrated or sham-castrated.
Implants consisted of T (testosterone propionate; Steraloids, Inc.) uniformly
mixed into silicone sealant (Type A, Dow Coming). This mixture was expelled
through a 1-cc syringe in a straight line onto wax paper and allowed to dry over
night, after which it was cut into 3 mm lengths. Each implant contained
approximately 1 mg of T. Blank implants were made in the same fashion, minus
the addition of T.

Surgeries were performed following isoﬂurane anesthesia. In females, a
small (~ 3 mm) dorsolateral incision was made in the skin, just rostral to the hind
legs. Either a T or a blank implant was then inserted subcutaneously, and the
incision was closed with Vetbond tissue adhesive (3M). In males, a dorsolateral
incision (~ 3 mm) was made through the skin and body wall adjacent to where
the testes were located. In castrates, the testes were removed; in sham-
castrates, the testes were probed and left intact. These incisions were closed
with silk sutures.

Following surgeries, the lizards were placed into cages identical to those
in which they were housed for the ﬁrst 30 days. They were kept in groups of up to

16 individuals, fed and watered as described above, and were mixed in the

96

enclosures across sex and treatment. Lizards were housed in this fashion until
they reached 90 days post-hatching (see below). Every 14 days, we checked
whether each female still had her implant, and replaced it when necessary (9 of
19 females required re-implantation over the 60 days of treatment). Survival to
day 90 in females was 87% (11 of 13 T-implanted and 9 of 10 blank-implanted
survived), and in males was 77% (10 of 15 castrated and 10 of 11 sham-
castrated survived). One female with a T implant was not included in the study;
her mass at 90 days was > 4 SEs below the female mean and her overall

physical appearance clearly suggested that she was unhealthy.

Tissue collection and analysis

At 90 days post-hatching, SVL and mass were recorded for the third time
for each juvenile, after which they were decapitated. Blood samples were
collected from the trunk, and the plasma fractions obtained following
centrifugation were measured to the nearest uL and individually stored at —80°C
until analysis for T content (see below). The kidneys were removed and placed in
Bouin’s ﬁxative. The epithelial cells of the renal “sex segments”, analogous to the
mammalian prostate, increase in height in the presence of androgens, and thus
serve as a T bioassay (Cuellar et al., 1972). In males, we used a dissecting
microscope to conﬁrm that castrates had no testicular tissue, and that sham-
castrates still had their testes. In all lizards, the 2"d ceratobranchial cartilage was
exposed by removing the dewlap skin and measured with calipers to the nearest

0.1 mm. The lower jaw, containing the ceratohyoid muscles, was placed in

97

Bouin’s ﬁxative. The cranium, containing the brain tissue including the AmbX and
AmleNllmv motoneurons, was placed in 10% phosphate-buffered formalin. To
assess potential effects in the hemipenis system, the ventral half of the rostral
third of the tail (the region where the hemipenes and RPM muscles are located)
and the tmnk-sacral region of the spinal cord containing the T17-S1 motoneurons
were placed in Bouin’s ﬁxative. The tail tissue also contained the caudifemoralis
(CF), 3 leg muscle not directly involved in copulation, which was measured to
assess the speciﬁcity of T action.

After ﬁxing for ﬁve days, the brain was removed from the skull and, along
with the other tissues, was soaked in 70% ethanol overnight, dehydrated, and
cleared in xylene. Tissues were then individually embedded in paraffin and
sectioned at 10 um. Brains and spinal cords were stained with thionin, throats
and tails were stained with trichrome, and kidneys were stained with hematoxylin
and eosin.

Measurements were taken, blind to treatment, using NIH Image software
on a PC and an Olympus BX-60 light microscope. In the dewlap system, cross-
sectional area was assessed in 15 ceratohyoid muscle ﬁbers for each animal,
randomly selected across the left and right muscles. Soma size was measured in
10 AmbX and 10 Amle/Vllmv motoneurons from each side (20 measurements
per animal) when possible, although in a few individuals fewer or no
measurements were obtained from one side. These cells are easily distinguished
based on not only their location, but also on the size relative to neighboring

neurons and speciﬁc morphological characteristics (Wade, 1998). Hemipenis

98

cross-sectional area was assessed in 10 sections per animal, spaced
approximately 50pm apart, for each the left and right hemipenes (20
measurements total). RPM muscle ﬁber cross-sectional area was measured in
25 ﬁbers in each the left and right muscle for every animal. The same procedure
was followed for the CF muscle. Cross-sectional areas of TPN muscle ﬁbers
could not be analyzed because those ﬁbers run parallel to plane of sectioning
(i.e., perpendicular to those of the RPM and CF; Ruiz and Wade, 2002); however
their presence/absence was noted. Twenty measures of cross-sectional area,
from each the left and right pools of T17-S1 motoneurons, were collected for
each animal. As for the dewlap motoneurons, the anatomical location and
particular characteristics of size, shape and orientation make these neurons easy
to identify (Ruiz and Wade, 2002). T17-S1 motoneuron number was also
estimated because it is increased in adult males compared to females (dewlap
motoneuron number is sexually monomorphic). Motoneurons with an in-focus
nucleolus were counted in every ﬁfth section throughout the rostro-caudal extent
of the T1 7-S1 nucleus and multiplied by the sampling ratio. Finally, epithelial cell
height was measured in four cells in each of four tubules in the renal sex
segment, for a total of 16 measurements per animal. A mean was computed for
each size measure in each animal for statistical analysis (see below). The
estimated total motoneuron number was analyzed after computing a mean for
the left and right sides. Sample sizes vary from 7 to 10 across tissue type (see

table and ﬁgures) due to histological artifact.

99

Radioimmunoassa y

Plasma levels of androgen were measured by ether extraction and
subsequent radioimmunoassay (RIA). Samples were thawed and mixed with 0.5
mL dH2O to provide sufﬁcient volume for extraction, and were equilibrated
overnight at 4°C with 1000 cpm of 3H-T (NET-370, 95 Ci/mmol) from Perkin
Elmer for individual recovery determinations. The following day, samples were
extracted twice with 3 mL diethyl ether, dried under nitrogen gas, and
reconstituted in 0.1 M phosphate buffered saline. They were then stored
overnight at 4°C. A competitive binding RIA was performed using 3H-T and
antibody from Wien Laboratories (T -3003). Because this antibody reacts
approximately equally with T and dihydrotestosterone, our measure is an
assessment of circulating androgen rather than strictly T. The standard curve
ranged from 1 to 250 pg and was run in triplicate. Samples were run in duplicate,
averaged, and adjusted for individual recovery and initial sample volume. The
intra-assay coefﬁcient of variation, based on four aliquots from a standard pool

included in the assay, was 4.2%.

Data analysis

Data were analyzed separately for males and females because the
treatments were not parallel across the sexes. Nonparametric statistics were
used to analyze total androgen (Wilcoxon rank sum and Spearrnan rank
correlation tests) because these data were not normally distributed. Most of the

measures of neuromuscular structures and their and peripheral targets showed

100

signiﬁcant positive relationships with SVL (Pearson correlations, P<0.05).
Therefore, prior to further statistical analysis we divided each of these data points
by the SVL of the animal from which it was collected (motoneuron number data
were not corrected). The data followed a normal distribution and were analyzed
using ANOVAs or two-sample t-tests. The residuals from regression plots of SVL
vs. motoneuron and muscle ﬁber area, cartilage length and hemipenis area were
also analyzed with t-tests. The statistical signiﬁcance of the results differed in
only one case. It is indicated below; otherwise, for simplicity, only the analyses of
the ratios are reported. Hypothesis tests were two-tailed with an overall a = 0.05;

data are presented as means 1 SE.

RESULTS

Body length and mass

Body length and mass increased steadily and signiﬁcantly for both
females and males. On average, females ranged from 23mm and 273mg on the
day of hatching to 35mm and 947mg at day 90 (SVL: F2,53=467.7, P < 0.0001;
mass: F2,53=187.6, P < 0.0001 ). Similarly, males grew from 23mm and 284mg to
35mm and 986mg (SVL: F2,33=238.1, P < 0.0001; mass: F2,53=142.9, P <
0.0001). However, there was no effect of treatment on either SVL or mass, nor
were there treatment x SVL or treatment x mass interactions, in either females or
males.

Treatment effects on T

101

Total androgen levels were signiﬁcantly higher, and epithelial cells in the
renal sex segments were signiﬁcantly larger, in juvenile females given T implants
than in those given blank implants (androgen level: W=61, P=0.02; epithelial cell
height: t=5.39, P<0.001; Table 1). Furthermore, a signiﬁcant positive relationship
between androgen level and epithelial cell height was detected in females
(Speannan’s r=2.48, P=0.01). Similarly, mean androgen level was higher and
mean epithelial cell height was greater in sham-castrated than in castrated
juvenile males (Table 1). Although these mean differences were in the expected
direction, they were not statistically different (androgen level: W=125, P=0.14;
epithelial cell height: t=1.60, P=0.13), and the positive relationship between
androgen level and epithelial cell height only approached statistical signiﬁcance
(Speannan’s r=1.66, P=0.09). Nevertheless, it is clear that androgen differences
in males were sufﬁcient to produce treatment effects (see below), even if group
means for measures of T exposure were too close to be statistically

distinguished.

102

 

§e_x Androgen (ng/ml)
Treatment (N)
Female
Blank (9) 0.77 (0.23)
T (10) 18.34 (5.5)*
Male
Sham (10) 1.87 (0.47)
Castrated (10) 1 .02 (0.21 )

Epithelial cell height (um)

14.1 (0.54)
28.1 (2.6)"

16.0 (0.65)
14.3 (0.88)

Table 1. Mean (SEM) plasma androgen levels and epithelial cell heights of the
renal sex segments of juvenile male and female green anoles. * p < 0.05. ** p <

0.001.

103

Dewlap system

Females with T implants had longer 2“d ceratobranchial cartilages (85.47,
P<0.001; Figure 1) and larger ceratohyoid muscle ﬁbers (divided by SVL: 82.62,
P=0.02; residuals: t=1.90, P=0.08; Figures 1 and 2) than did females with blank
implants. Sham-castrated males had longer 2"" ceratobranchial cartilages
(t=2.33, P=0.03) than did castrated males (Figure 1), but an effect of treatment
on ceratohyoid muscle ﬁbers was not detected (Figures 1 and 2). Soma sizes of
AmbX and AmleNllmv motoneurons were not affected by treatment in either

females or males (Figure 1).

Hemipenis system

Females did not develop hemipenes in response to T exposure, nor did
they develop RPM or TPN muscles (Figure 3). In males, castration resulted in
signiﬁcantly smaller hemipenes (t=3.08, P=0.007; Fig. 3 and 4), but had no effect
on RPM muscle ﬁber size (Figure 4). The CF muscle, measured as a control,
contained signiﬁcantly smaller ﬁbers in T-implanted than in blank-implanted
females (t=3.18, P=0.006), but showed no size differences across male
treatment groups (data not shown). A decrease in CF muscle ﬁber size also
occurs in response to T treatment in adult males (Holmes and Wade, 2004), but
the mechanisms are unknown. As in the dewlap system, treatment did not affect
soma size for T17-S1 motoneurons in females or males (Figure 4). It also did not
alter motoneuron number in females (blank [n=9]: 232.5180; T[10]: 24731124)
or males (castrated [10]: 23681135; sham [10]: 24551104).

104

0.30

 

0.25 .

0.20 «

0.15 <

0.10 -

 

Cartilage length (mm) I SVL (mm)

 

 

 

 

Flber size (m2) I SVL (mm)

 

 

 

Soma size (umz) I SVL (mm)
-I N

O

 

 

 

 

Sham Castrate

Figure 1. Treatment effects in juvenile female and male green anoles on
morphological components responsible for dewlap extension. (A) Length of the
2"d ceratobranchial cartilage; (B) cross-sectional area of the ceratohyoid muscle
ﬁbers; (C) cross-sectional areas of AmleNllmv (solid bars) and AmbX (hatched
bars) somata. Data were sampled 90 days after hatching, following 60 days of
treatment, and are presented as mean + SE (corrected for snout-vent length,
SVL). Sample sizes are given at the bottom of the bars and are not identical for
each tissue type due to histological artifact. Male and female data sets were
analyzed separately by two-sample t-tests. *P < 0.05.

105

 

Figure 2. Photomicrographs of ceratohyoid muscle ﬁbers collected from juvenile
green anoles 90 days after hatching, following 60 days of treatment. (A) Blank-
implanted females had signiﬁcantly smaller ﬁbers than (B) testosterone-
implanted females; (C) sham-castrated and (D) castrated males did not differ
signiﬁcantly. Scale bar = 20 pm.

106

 

Figure 3. Photomicrographs of the tail region in juvenile green anoles where
hemipenes are located, collected 90 days after hatching, following 60 days of
treatment. (A) Blank—implanted and (B) testosterone-implanted females did not
develop hemipenes; (C) sham-castrated males had signiﬁcantly larger hemipene
cross-sectional areas than (D) castrated males. Scale bar = 20 pm. CF =
caudifemoralis; HP = hemipene; TPN = transversus penis.

107

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

'5 0.004 *

g A T

g 0.003.

“E 1'

g 0.002.

5

I

i 0.001 «

5 0.000 10 10
Sham Cestrate

A 0.8

E

.E. B l

.l

> 0.0. i

a:

“E 0.4 .

E

a 0.2.

2’

if. 0.0 1° 9
Sham Cutrate

10
C Females Males

 

 

 

 

Soma size (me) I SVL (mm)
&

 

 

 

 

 

 

10 10
Blank T Shem Outfit.

 

Figure 4. Treatment effects in juvenile female and male green anoles on
hemipenes and associated neuromusculature sampled 90 days after hatching,
following 60 days of treatment. (A) Hemipene and (B) retractor penis magnus
(RPM) ﬁber cross-sectional areas in males; females did not develop hemipenes
or RPMs. (C) Cross-sectional area of T17-S1 motoneuron somata in females and
males. Data are presented as mean + SE, corrected for snout-vent length, SVL.
Sample sizes are given at the bottom of the bars and are not identical for each
tissue type due to histological artifact. Male and female data sets were analyzed
separately by two-sample t-tests. *P < 0.05.

108

DISCUSSION

Summary

Our data suggest that steroid-mediated plasticity is present during juvenile
development in neuromuscular systems and their target tissues (cartilage and
hemipenes) involved in courtship and copulation in green anoles. As predicted,
juvenile females given T implants developed longer dewlap cartilages and larger
dewlap muscle ﬁbers than did those given blank implants. Similarly, castrated
juvenile males developed shorter dewlap cartilages than intact males, although
dewlap muscle ﬁber size was not affected by treatment. The effect of T
manipulation did not extend to dewlap motoneuron soma size in either males or
females. Collectively, these results suggest that the peripheral tissues, but not
the motoneurons in the dewlap system, differentiate during juvenile development
and that T likely mediates this process. With regard to the hemipenis system,
none of the juvenile females in either the T or control group possessed
hemipenes or the underlying muscles controlling them, nor were there any
effects on motoneuron soma size or number. These results suggest that sexual
differentiation of the hemipenis system is largely complete by post-hatching day
30, when our treatments began. Consistent with that idea, castration of males did
not affect muscle ﬁber size or motoneuron soma size or number, but it caused a
reduction in hemipenis size. Thus, from post-hatching days 30 to 90, beyond the
period of sexual differentiation, testis-mediated plasticity exists in the copulatory
peripheral tissue, but its neuromuscular effectors are less or not responsive

during this period.

109

Questions raised by the present data

Two general classes of questions are suggested by the present data. The
ﬁrst set involves mechanisms of action for sexual differentiation; the second
concerns the timing of responsiveness to testosterone and how the
morphological changes in the dewlap and hemipenis systems relate to
organizational vs. activational theories of steroid hormone action.

Mechanisms of semi differentiation. Juvenile T clearly masculinizes the
dewlap system cartilage. Administering the hormone to females enlarged the
structure, and castrating males reduced it. However, the muscle ﬁbers were
affected more modestly and only in females. Why? The diminished response
might be due to a difference in any number of factors inﬂuencing androgen action
(e.g., steroid metabolism, receptor distribution). Likely, however, the differential
effect between the sexes is due to the manner in which T was manipulated in
males and females (castration vs. implants). Plasma T was approximately 20
times higher in treated compared to control females (and similar to that of
breeding adult males; Jenssen et al., 2001), whereas castrated males had
approximately half the plasma T of sham-castrates (Table 1). The implants may
have provided sufﬁcient T exposure to induce dewlap muscle growth in females,
while castrations did not sufﬁciently reduce T to prevent muscle growth in males.
The relative contributions of the testes and other organs to circulating androgen
levels are unknown for developing anoles; substantial adrenal output is a

reasonable possibility. Future work could circumvent this issue by treating

110

juvenile males with anti-androgens, which should inhibit muscle growth if T is
responsible.

The hemipenis system is more completely differentiated than the dewlap
system. Adult males have hemipenes and associated musculature; females do
not. The sex difference in hemipenes at least is obvious by the day of hatching
(Holmes and Wade, unpublished), and T treatment from post-hatching days 30-
90 does not reverse this effect. When in embryonic development does
differentiation occur and is it mediated by T? Little work in this area has been
done in any lizard species. In Lacertids, estradiol feminizes development of
genital structures, but opposite to anoles, these animals exhibit female
heterogamity (reviewed in Adkins-Regan, 1981; Raynaud and Pieau, 1985).
Treatment of “post-hatching” (timing undeﬁned) or adult male anoles with
estradiol causes atrophy of the hemipenes (Forbes, 1964), but that could be due
to decreased testicular function caused by negative feedback. The
developmental time course needs to be worked out and the steroid environment
of embryonic anoles needs to be manipulated in future experiments.

Does T inﬂuence sexual differentiation of either the copulatory or dewlap
motoneurons? Although dewlap motoneurons are larger in adult males than
females (Wade, 1998; O'Bryant and Wade, 1999), their size has not yet
differentiated by post-hatching day 90 (O’Bryant and Wade, 2001). It is possible
that T inﬂuences this feature later in development, perhaps as the animals enter
their ﬁrst breeding season the following spring and males initially experience

adult levels of T. Alternatively, T may not directly inﬂuence development of

111

dewlap motoneuron soma size; these cells may become dimorphic only after the
dewlap has been used more by males (unlike adults, juvenile males and female
extend their dewlaps equivalently; Lovem and Jenssen, 2001). Both T and T-
mediated behavior affect motoneuron soma size in the copulatory system of rats,
although increased use appears to decrease this measure (Breedlove, 1997a;
but see Raouf et al., 2000). It is presently unknown when the copulatory
motoneurons differentiate in unmanipulated green anoles, but it is likely after
post-hatching day 90. While not completely appropriate to compare between
these groups due to different types of surgical manipulations, soma size and cell
number were equivalent in sham-castrated males and blank-implanted females in
the present study. Thus, either scenario could apply to them as well. That is, the
soma size and/or number of cOpulatory motoneurons may differentiate just prior
to or during their ﬁrst breeding season. All of the previous experiments on
morphology of the dewlap and copulatory systems have been conducted in
adults of unknown ages, collected from the ﬁeld sometime after they have
entered their ﬁrst breeding season, so we have no way at present of addressing
this issue.

Given that juvenile T has at least some masculinizing effects, where
speciﬁcally does it act? Dewlap motoneurons (AmbX) contain androgen receptor
(AR) in adulthood (Rosen et al., 2002), but it is unknown whether any of the other
dewlap or copulatory tissues assessed in this study do, and AR distribution has
not yet been investigated in development. Adult leopard geckos express AR in

hemipenes (Rhen and Crews, 2001), and given the robust response of

112

hemipenis morphology to castration in the present experiment (even stronger
than our standard renal sex segment bioassay), it is likely that juvenile anoles do
as well. In theory, T could act on any of the structures to induce morphological
change. Although there is some variability due to developmental stage, AR are
expressed in both muscles and motoneurons in other sexually dimorphic
courtship and copulatory systems (Arnold, 1980; Fishman et al., 1990; Jordan et
al., 1991, 1997; Godsave et al., 2002; Breedlove et al., 2002; Veney and Wade,
2004). It is also possible that the androgen acts on the more peripheral
stmctures only (dewlap cartilage and hemipenes) and the muscles and
motoneurons are subsequently masculinized due to enhanced mechanical
stimulation and/or retrograde chemical signals. Future work on the distribution of
AR will be critical to addressing this question of where androgen acts.
Additionally, that work may provide insight as to how the dewlap and copulatory

systems differentially respond to T at several life stages.

Timing of T-responsiveness: Implications for the classic theory of organizational
effects of hormones.

Steroid effects on early sexual differentiation and subsequent behavioral
expression have classically been conceptualized by theories of ‘organization’ and
“activation” (Phoenix et al., 1959). However, it is increasingly clear that they
represent extremes on a continuum of effects on nervous system and behavior
(e.g., Arnold and Breedlove, 1985; Breedlove et al., 1999). Male-biased sex

differences in courtship and copulatory behavior in green anoles are clearly

113

activated by T (e.g., Winkler and Wade, 1998; Rosen and Wade, 2000). But what
do our data suggest about the role of T in the organization of morphology?

Some aspects of the dewlap system seem to ﬁt a traditional
organizational action of T. In particular, the cartilage and muscle at least appear
to be permanently organized during development. Male-biased dimorphisms in
neuromuscular structures of the dewlap exist at many levels (see Introduction),
and these sex differences are quite stable; neuromuscular structure varies little
between breeding and non-breeding seasons, and androgen manipulations have
no effect on the sexual dimorphisms in this system in adults (O’Bryant and Wade,
1999, 2002b). Similarly, dewlap coloration is sexually dimorphic and ﬁxed in adult
tree lizards (Urosaurus omatus), but like the anole dewlap structures, it is
sensitive to androgen exposure in both juvenile males and females until around
60 days post-hatching (Hews and Moore, 1995, 1996; Hews et al., 1994).
However, these effects on dewlaps of both species are quite late, especially
considering that lizards hatch in very precocial states. They may be similar to
what occurs during puberty in mammals (e.g., Romeo et al., 2002), when
dramatic sex differences appear in a second, key organizational window during
development, well beyond the perinatal period. A similar idea is suggested for the
copulatory motoneurons of Mongolian gerbils (Men'ones unguiculatus), which
appear to develop later than those of other rodents (Fraley and Ulibarri, 2001).
However, the late differentiation of the dewlap system may not be entirely like
“classic” mammalian puberty, in which reproductive behavior and fertility

immediately follow pubertal maturation, given appropriate environmental

114

conditions. In anoles, the increase in T to levels typical of breeding males, and
consequent reproduction, does not occur until the spring following the summer of
hatching, 6-9 months later. Perhaps it is therefore more appropriate to think of
the dewlap and copulatory systems in anoles as having two distinct
organizational periods prior to reproductive maturation.

The fact that juvenile T manipulations do not induce masculinization of the
hemipenis system in females is consistent with the idea that it is also
permanently organized, albeit much earlier in development. It is, of course,
possible that factors other than testosterone could induce growth (or prevent
regression) of this system in juvenile females. However, data from at least one
other lizard species suggest that T is important in some of these structures. Adult
female leopard geckos (Eublepharis macularius) retain the capacity to develop
hemipenes in response to as little as two weeks of androgen treatment (Rhen et
al., 1999). This result suggests that, in contrast to green anoles, the hemipenis
system in leopard geckos may not be permanently organized, which may be
fundamentally related to their mode of sex determination (leopard geckos have
temperature-dependent sex determination, in contrast to anoles which have
genotypic sex determination; Viets et al., 1994).

While juvenile female anoles apparently do not grow hemipenes, this
system shows some plasticity in adulthood in response to T, similar to the
copulatory system of rats in which both muscles ﬁbers and motoneurons enlarge
in response to adult T (reviewed in Breedlove et al., 2002). However, the

sensitivity to androgen in adult anoles depends on the environmental conditions

115

(Holmes and Wade, 2004b; Chapter 2). That is, hemipenis and RPM ﬁber size
are not different in intact males in the breeding and non-breeding seasons. But,
castrated males housed in breeding conditions (long days and warm
temperatures) respond to T treatment with dramatic increases in hemipenis and
RPM ﬁber sizes, whereas males housed in non-breeding environmental
conditions do not, indicating that sensitivity to T is greater in the breeding
season.

This differential adult plasticity between the dewlap and copulatory
systems parallels behavior to some degree. Like dewlap morphology, the
frequency of its extension shows strong male-biased sex differences (e.g.,
Jenssen et al., 2000; Nunez et al., 1997). But, unlike the underlying structure, the
behavior shows dramatic seasonal differences in A. carolinensis (Jenssen et al.,
1995, 1996) and the congeneric A. sagrei (T okarz et al., 2002), as well as effects
due to androgen manipulation (Adkins and Schlesinger, 1979; Winkler and
Wade, 1998; O’Bryant and Wade, 2002a; Tokarz et al., 2002) that are consistent
with endogenous T proﬁles (Lovem and Wade, 2001; Lovem et al., 2001). Both
males and females use their dewlaps more during the breeding season, when
androgen levels are higher (approximately 20 and 0.25 ng/ml for males and
females respectively), than during the non-breeding season, when androgen
levels are lower (10 and 0.1 ng/ml, respectively). lmportantly, however, while the
behavior is displayed far less frequently, the dewlap is extended by adults of both
sexes in the non-breeding season (Jenssen et al., 1996). That maintenance of

behavior may require (or facilitate) stability of the underlying morphology. In

116

parallel, T—induced plasticity in the RPM and hemipenes only under breeding
environmental conditions mirrors the fact that, unlike dewlap use, T can facilitate
copulation during the breeding, but does not appear to in the non-breeding,
season (O'Bryant and Wade, 20023). As all juveniles in the present study were
reared under summer environmental conditions (those that are conducive to
breeding in adults), it is unclear whether temperature and/or photoperiod might
affect responsiveness to T at this age.

Relationships between structure and function and between the role of T in
behavioral facilitation and masculinization of morphology (in terms of both sexual
differentiation and adult plasticity) are common (see Cooke et al., 1998;
Breedlove et al., 2002). However, the selective plasticity across the two
neuromuscular systems in green anoles indicates perhaps even more broad
deﬁnitions of organizational and maybe activational effects of steroid hormones
than previously considered. That is: (1) while sexual differentiation of both the
dewlap and copulatory systems appears permanent, it is far more extreme in one
than the other (females lack the hemipenis system, but have a reduced dewlap
system); (2) sexual differentiation of the two systems occurs at quite different life
stages (T in juvenile females masculinizes dewlap, but not the hemipenis,
structures); and (3) the hemipenis, but not the dewlap, system shows plasticity in
adulthood, but it is tempered by environmental conditions. The opportunity to
simultaneously examine the effects of identical treatments on two neuromuscular
systems required for the full suite of masculine reproductive behaviors in anoles

will allow us to further characterize the possibilities for modulating structures and

117

their functions, and importantly to uncover the mechanisms involved in mediating

speciﬁc aspects of the changes.

118

CHAPTER FIVE

Holmes MM, Wade J (2005). Sexual differentiation of the copulatory
neuromuscular system in green anoles (Anolis carolinensis): normal ontogeny
and manipulation of steroid hormones. J Comp Neurol: in press.

 

119

 

 

INTRODUCTION

Gonadal steroid hormones commonly control differentiation of sexually
dimorphic neuromuscular systems. Androgens masculinize both the muscles and
motoneurons in the rodent copulatory neuromuscular system during perinatal
development (e.g., Breedlove et al., 1982; Breedlove and Arnold, 1983a, 1983b;
Nordeen et al., 1985), while estradiol (E) increases both the soma size
(Breedlove, 1997b) and dendritic outgrowth (Goldstein and Sengelaub, 1994) of
motoneurons in this system. Similarly, androgens masculinize (increase) neuron,
axon, and muscle ﬁber number in the laryngeal system of juvenile Afn'can clawed
frogs (Sassoon and Kelley, 1986; Robertson et al., 1994; Kay et al., 1999),
muscle ﬁber number in the sonic (vocal) muscle of juvenile plainﬁn midshipman
ﬁsh (Brantley et al., 1993b), and they can affect syrinx weight and associated
motoneuron soma size in zebra ﬁnches (Wade et al., 2002). Interestingly, E
appears to have an active role in the feminization of some dimorphic
neuromuscular systems as it decreases syrinx weight and muscle ﬁber size in
juvenile zebra ﬁnches (Wade et al., 2002) and feminizes the neuromuscular
junction (by increasing synaptic strength) in juvenile African clawed frogs (Tobias
and Kelley, 1995).

Androgens and estrogens often also inﬂuence sex differences in both the
structure and function of these systems in adulthood. This is particularly evident
in the rodent copulatory system where treatment with testosterone (T) or
dihydrotestosterone (DHT) typically increases the size of both muscles and

motoneurons and is required for the expression of the associated behavior

120

(penile reﬂexes) (e.g., Hart, 1967; Breedlove and Arnold, 1981; Rand and
Breedlove, 1991). However, in adulthood, E does not alter the size of these
copulatory muscle ﬁbers (Fargo et al., 2003) or motoneurons (Forger et al.,
1992), although it does increase excitability of the muscles (Fargo et al., 2003).
In adult African clawed frogs, prolonged T-treatment induces complete
masculinization of laryngeal muscles in females (Tobias et al., 1991), and adult
ovariectomy decreases quantal content at the synapse (Tobias et al., 1998). T
also masculinizes the syrinx muscles, but not associated motoneurons, in adult
female zebra ﬁnches (Wade and Buhlman, 2000).

Similar to other reproductive neuromuscular systems, the components that
underlie copulation in lizards are also sexually dimorphic. Adult male green
anoles possess two intromittant copulatory organs called hemipenes. Each
hemipenis is independently controlled by two muscles, the transversus penis
(T PN; everts the hemipenis) and the retractor penis magnus (RPM; retracts the
hemipenis) (Arnold, 1984). These muscles receive projections from ipsilateral
populations of motoneurons found in spinal segments trunk 17 and sacral 1 (T 17-
S1) (Ruiz and Wade, 2002; Holmes and Wade, 2004a; Chapter 1). As in many
others, androgens have stimulating effects on this neuromuscular system; T-
treatment increases size of hemipenes and copulatory muscles (but not
motoneurons) in adult male green anoles (Holmes and Wade, 2004b; Chapter 2)
and castration decreases these measures (only statistically signiﬁcant for
hemipenes) in juvenile males (Lovem et al., 2004b; Chapter 4). In contrast, E can

induce atrophy of hemipenes and copulatory muscles in adult males (Holmes

121

and Wade, unpublished observations). Adult females do not possess hemipenes
or the associated muscles and have fewer and smaller T17-S1 motoneurons
(Ruiz and Wade, 2002; Holmes and Wade, 20043; Chapter 1). Juvenile females
also do not possess male-typical copulatory neuromuscular structures, and
treatment with T at that age does not cause them to develop (Lovem et al.,
2004b; Chapter 4), consistent with the idea that sexual differentiation of this
system is not reversible once it occurs.

It is clear that the adult lizard copulatory neuromuscular system exhibits
responsiveness to steroid hormones similar to other species described above.
However, while signiﬁcant work has focused on ﬁsh, frogs, birds, and mammals,
little is known about the effects that these hormones have during sexual
differentiation of reptilian neuromuscular systems. Androgen treatment has
masculinized the hemipenes in embryos of some lizard species and estrogens
can inhibit the development of these structures (reviewed in Raynaud and Pieau,
1985), but this work has been done in few species (not the green anole) and the
role of these hormones in the differentiation of the underlying muscles and
motoneurons was unknown. To address this issue in particular, in Experiment 1
we characterized the time course of normal ontogeny of copulatory
neuromuscular morphology in green anoles by collecting embryos at a variety of
developmental stages. Using the developmental time points determined, in
Experiment 2 we treated embryos with T, DHT, E or vehicle control after gonadal

differentiation but prior to differentiation of secondary copulatory structures.

122

Animals were then collected on the day of hatching, allowing us to determine the

effects of hormone treatment on morphology of the neuromuscular components.

MATERIALS AND METHODS

Animals and housing

All embryos and hatchlings came from our breeding colony. Adult male
and female green anoles were purchased from Charies Sullivan Co (Nashville,
TN). Animals were group housed (1 male with 3-6 females) in 110-liter glass
aquaria (76 x 30 x 48 cm) that each contained a nest box (0.5 to 1.0 L plastic
containers with one hole cut in the top, 213 full of peat moss dampened with
distilled water). Lizards were housed in a 14:10 hour lightzdark cycle with
ﬂuorescent and full-spectrum lights, as well as a heat lamp placed on one end of
each aquarium. Temperature ranged from 28'C (ambient) to 38'C (directly under
heat lamp) during the day to 18'C at night. These conditions approximate those
the animals experience in the spring in the ﬁeld and stimulate breeding in the lab.
Aquaria were sprayed daily to help maintain 70% relative humidity, and water
was provided ad Iibitum. Animals were fed crickets or mealwon'ns three times a
week. Nest boxes were checked daily for eggs, which were then placed
individually in plastic cups containing a 1:1 (mass) vermiculite:dH20 mixture and
sealed with plastic wrap. Eggs were incubated at 28’C and hatched
approximately 34 days after laying. All procedures were approved by the
Michigan State University All University Committee on Animal Use and Care and

conform to NIH guidelines.

123

Treatment, tissue collection, and analyses

Exgriment 1. To determine the time course of differentiation of copulatory
neuromuscular structures, embryos were collected at one of several points
during incubation: days 10, 13, 16, 19, 22, 25, 28, or 31 or on the day of hatching
(day 0 was the day of laying). The number of eggs collected at each stage
ranged from 5 to 19. Eggs were opened and embryos were removed, weighed
and decapitated. Bodies were placed in Bouin’s ﬁxative for seven days. Tissue
was then soaked in an ascending series of alcohols and xylene, embedded in
parafﬁn, and sectioned rostral to caudal through the abdomen, pelvis and tail at
10 um. All tissue was stained with hematoxylin and eosin. For each animal,
tissue was analyzed for a) classiﬁcation of gonad (ovary vs. testes), b)
presence/absence of hemipenes and associated muscles, and c) renal sex
segment cell size (see below). A gonad was classiﬁed as an ovary if it had a well-
developed cortex with developing follicles and a regressing medulla and as a
testis if it contained a well-developed medulla with organized seminiferous
tubules with no distinct cortex or follicular cells (see Figure 1). Hemipenes and
muscles were identiﬁed by their locations on the ventral surface of the tail; they
lie immediately inside the skin at the junction of the caudifemoralis (CF),
ischiocaudalis (IC) and iliocaudalis (ILC) muscles (see diagram in Ruiz and
Wade, 2002). The height of four renal sex segment epithelial cells was measured
from each of 4 randomly selected kidney tubules (resulting in 16 measures). The

renal sex segments serve as a bioassay for androgen exposure as they function

124

similar to the mammalian prostate and enlarge in response to these hormones
(Cueller et al., 1972; Crews 1980; Winkler and Wade, 1998).

An additional 16 embryos were collected on incubation day 7, 8, or 9 to
determine the timing of gonadal differentiation. Tissue was collected and treated
as above. Gonads were classiﬁed as ovaries or testes as above or as
“undifferentiated” if they did not possess distinguishing characteristics (Figure 1).

Motoneurons could not be evaluated in the above tissue as it was stained
with hematoxylin and eosin, which does not provide a distinct outline of the
somata. Therefore, 11 additional hatchlings (5 female and 6 male) were collected
for this purpose. The tissue was collected and processed as above but was
stained with thionin. Motoneurons were analyzed as in Lovem et al. (2004b;
Chapter 4). To avoid double-counting of cells across sections, motoneuron
number was estimated using a modiﬁed version of the physical dissector
technique (Gunderson, 1986; Ruiz and Wade, 2002; Holmes and Wade, 20043;
Chapter 1) in which a cell was counted if it possessed an in-focus nucleolus.
Motoneurons were counted in every ﬁfth section throughout the rostro-caudal
extent of the T17-Si nucleus (from the rostral onset of the T17 dorsal nerve to
the caudal extent of the S1 dorsal nerve). Counts were multiplied by the sampling
ratio, providing an estimate for total motoneuron number for the left and right
sides of the spinal cord. Average motoneuron soma size was estimated by
measuring the cross-sectional area of 20 motoneurons from each of the left and

right sides of the spinal cord in segments T17-S1. A mean was computed for

125

Undifferentiated

 

Hatching

Figure 1. Photomicrographs of cross sections through the gonads of untreated
developing green anole embryos. Gonads differentiate around incubation days 9
and 10 (total incubation time is 34 days on average). At days 7-9, gonadal tissue
in most individuals did not have deﬁning characteristics of either testes or
ovaries. By day 13, ovaries each had a well-developed cortex (noted with white
arrows) that contained developing follicles, and testes had well—developed
medullas with organized seminiferous tubules (noted with white asterisks). Scale

bar = 50 pm.

126

each side but as there were no effects of laterality (data not shown), analyses
below reﬂect an average of both sides for both motoneuron size and number.

Body weight and renal sex segment cell height were analyzed by two-way
ANOVA with age and gonadal sex as independent variables. Posthoc analyses
were performed using Tukey-Kramer tests. Motoneuron soma size and number
in male and female hatchlings were analyzed with unpaired t-tests. Statistical
analyses were not performed on hemipenes or RPMs as the size of these
structures could not be reliably measured; hemipenes were everted until shortly
before hatching, which makes it difﬁcult to discern the edge of each hemipenis
from the wall of the cloaca, and RPMs extend into the hemipenes when everted,
which additionally distorts their shape and size (Figure 2).

Experiment 2. To determine whether steroid hormones control
differentiation of copulatory neuromuscular structures, eggs were treated with
95% ethanol containing T (100ug), DHT (100ug), E (10pg) or no steroid (number
of animals per group is listed in Table 2). All treatments were a 5pl bolus applied
topically to the vascularized portion of the shell (Crews et al., 1991). Eggs each
received two treatments: one on incubation day 10 and one on incubation day
13. These time points were chosen because gonadal differentiation is complete
by day 10 but regression of the copulatory neuromuscular structures is not fully
undenrvay (see Results for Experiment 1 below). Eggs remained in the incubator
until hatching at which point tissue was collected and processed as for

Experiment 1 with one exception. Slides containing the pelvic region of the spinal

127

 

Figure 2. Photomicrographs of cross sections through the pelvic region of
untreated developing female (A, C, and E) and male (B, D, and F) embryos. The
dorsal surface of each embryo is on the top of each panel and the ventral
surface, with hemipenes protruding in most cases, is on the bottom. For
orientation, the spinal cords are denoted with white arrows and one hemipenis in
each panel except (E) is indicated with a white asterisk. Kidneys (k), cloacal vent
(c) and caudifemoralis (cf) muscles are also marked. At incubation day 10 (A and
B) both males and females have everted hemipenes. At incubation day 13 (C
and D), the hemipenes are decreasing in size in females and increasing in size in
males. By incubation day 19 (E and F), hemipenes are absent in females. They
remain everted in males at this time. RPM ﬁbers project into the center of the
hemipenes when everted and can be seen in (F), indicated with a black arrow.
Additional tissue visible in some panels (as in the bottom right corner of (B)) is
from the lower limbs of the embryos. Scale bar = 215 pm (A-D), 300 pm (E, F).

128

 

cord were stained with thionin while slides containing abdominal tissue (gonads)
and tail tissue (hemipenes and muscles) were stained with hematoxylin and
eosin. These techniques permitted accurate analyses of motoneuron size and
number while also allowing for evaluation of the gonads, hemipenes, and
muscles in the same animals.

Analyses of gonads, hemipenes, muscles, renal sex segments, and
motoneurons were performed as described for Experiment 1. As genetic sex of
the animals could not be determined, animals were classiﬁed as males if they
possessed testes and as females if they possessed ovaries. There were no
cases in which the identiﬁcation of these organs was ambiguous. Body weight,
renal sex segment cell height, and motoneuron number and size were analyzed
by two-way ANOVA with steroid treatment and gonadal sex as the independent
variables. Posthoc analyses were performed using Tukey-Kramer tests. As in
Experiment 1, size of hemipenes and muscles could not be analyzed as
hemipenes were everted in hormone treated but not control animals (see below).

All photomicrographs (Experiments 1 and 2) were obtained using an
Olympus BX51 microscope with a Macroﬁre (Optronics) digital camera. All
images were sized, compiled, and labeled in Adobe PhotoShop 7.0 (San Jose,
CA). In addition, brightness and contrast were adjusted as necessary and
retouching was performed to remove background artifacts; no modiﬁcations were

made to images of the tissue.

129

RESULTS

Experiment 1. Gonadal differentiation occurred around incubation days 9
and 10 (Figure 1). Gonads were undifferentiated (lacked clear cortex and/or
organized seminiferous tubules) in all day 7 (n=5) and day 8 (n=4) embryos and
in 4 of 7 day 9 embryos; 3 of 7 day 9 embryos had begun to develop ovarian
cortices. By day 10, only 2 of 7 embryos had undifferentiated gonads; of the
others, 4 had ovaries and 1 had testes. All animals possessed Wolfﬁan ducts
while only animals with ovaries possessed oviducts. All animals possessed
hemipenes and rudimentary RPM ﬁbers on incubation day 10 (Figure 2).
Hemipenes were everted prior to hatching and increased in size throughout
development in gonadal males (Figure 2). RPM ﬁbers organized into distinct,
ensheathed muscles by incubation day 19, but may not have been functional
until shortly before hatching as that is when hemipenes were retracted (Figure 3).
Hemipenes and muscles degenerated in concert in females; they were clearly
regressing in all day 13 gonadal females (n=6) and were completely absent in
day 19 females (n=5) (Figure 2). Cells in spinal segments T17-Si differentiated
into recognizable multipolar motoneurons between days 10 and 19 and were
cleariy organized in the ventral horn by days 19-22 (Figure 4). In hatchlings, T17-
S1 motoneurons were not sexually dimorphic in size (t = 0.054, p = 0.958) or
number (t = 0.934, p = 0.375) (Table 1).

Renal sex segment cells were larger in males (overall average = 12.822
+I- 0.204 um) than females (overall average = 11.241 +l- 0.239 urn) (F = 18.317,
p < 0.0001), but age did not affect this measure (F = 1.175, p = 0.323). A

130

 

Figure 3. Photomicrographs of cross sections through the tail of two untreated
individuals, a female (A and C) and a male (B and D) hatchling. (A) and (B) are at
the level of the hemipenes, which are located on the ventral surface of the tall at
the junction of the caudifemoralis (cf) and ischiocaudalis (ic) muscles. (C) and (D)
are at the level of the RPM, which attaches to the base of the hemipenis in males
and is more caudal. Note that hemipenis and RPM are completely absent in the
female hatchling. H = hemipenis; RPM = retractor penis magnus. Scale bar = 100
pm.

131

\

 

Figure 4. Photomicrographs of cross sections through spinal segment trunk 17 in
untreated male embryos on incubation days 10 (A), 13 (B), 19 (C), 22 (D), 25 (E),
and at hatching (F). White arrows point to motoneurons in the ventral horn, some
of which project to the RPM and TPN muscles. Note the distinct conﬁguration of
the motoneurons that becomes apparent around incubation days 19 to 22. These
cells are not sexually dimorphic in size or number at the time of hatching. Scale
bar = 100 pm.

132

T17-S1 count T17-S1 soma size
224.20 (11.20) 149.35 (14.39)

 

Female (n=5)
Male (n=6) 249.17 (22.40) 150.41 (12.15)

 

Table 1. Mean (+/- SEM) motoneuron count and soma size (pmz) in spinal
segments trunk 17 and sacral 1 (T17-S1). Approximately half of the cells in this
region project to muscles controlling the hemipenes in adult male green anoles.

133

statistically signiﬁcant interaction between gonadal sex and age was detected (F
= 2.715, p = 0.010) as the sex difference appeared only after incubation day 16.
Embryos increased in weight as they aged (F = 64.877, p < 0.0001) from 75.0
(+/-2.35) g on day 10 to 274.86 (+/- 7.25) g on the day of hatching, but this
measure did not differ between the sexes (F = 0.014, p = 0.9077). The interaction
between age and sex also was not statistically signiﬁcant (F = 0.554, p = 0.791 ).
Experiment 2. All control animals (8 males and 7 females) had copulatory
morphology consistent with their gonadal sex. Only animals with testes
possessed hemipenes and associated muscles (Figure 5). These structures were
absent in all animals treated with E (14 gonadal females and 4 gonadal males)
(Figure 5). All androgen-treated males had hemipenes, RPMs, and TPNs (11 for
DHT and 5 for T), and most androgen-treated females also had hemipenes and
associated muscles (9 of 10 for DHT and 8 of 10 for T) (Figure 5). Similar to the
untreated hatchlings in Experiment 1, T17-S1 motoneurons were not sexually
dimorphic in size (F = 0.852, p = 0.360) or number (F = 2.060, p = 0.157) in this
experiment (Table 2). There was a main effect of steroid treatment on
motoneuron size (F = 2.823, p = 0.047) and number (F = 2.918, p = 0.042), and a
signiﬁcant interaction between gonadal sex and steroid treatment was detected
for motoneuron number (F = 3.476, p = 0.022) but not for motoneuron size (F =
0.782, p = 0.509). However, no statistically signiﬁcant pairwise comparisons
between treatment groups were detected on either measure. There was a main
effect of steroid treatment on renal sex segment cell height (F = 16.825, p <

0.0001), but no signiﬁcant main effect of sex was detected (F = 0.662, p = 0.419)

134

 

Figure 5. Photomicrographs of cross sections through the tail (A—D) and pelvis
(E-H) of treated hatchlings. (A) and (B) are from a control female and male and
(C) and (D) are from an E-treated female and male, respectively. (A-D) are at the
level of the hemipenes, which are located on the ventral surface of the tail at the
junction of the caudifemoralis (cf) and ischiocaudalis (ic) muscles. Note that the
hemipenis (marked with asterisks in which they are depicted) is completely
absent in the control female as well as both E treated animals. (E) and (F) are
the same animals as in (A) and (B) while (G) and (H) are from a DHT treated
female and male, respectively. For (E-H) the dorsal surface of each embryo is on
the top of each panel and the ventral surface is on the bottom. For orientation,
kidneys (k), cloacal vent (c) and cf muscles are noted. Note that hemipenes are
absent in the control female and are everted in both the DHT treated female and
male. The retracted hemipenes in (B) and (F) appear different because they are
at different levels; (B) is approximately midway through the rostral-caudal extent
of the hemipenis while (F) is at the level of the cloacal vent, right at the top of the
hemipenes. Scale bar = 150 um (A-D) and 300 um (E-H).

135

.EoEEo: 8 x8 .058 .8 3833 903 £903 5 moocozoEo oz 89¢th new 8.9: 53 5

E90; =8 EoEmcm xow Eco. 33905 22.8556 Eocene... comets/x .3693 903 mcomtquoo 3323a
E8556 a: 5:053 05m .Em LBEE coSocoEE oommmzos zucmoEcmﬁ EoEEm: 299m .__Eo>O .8. can

or gnu cos—3:05 co nouns: mobnEo L8 CWCC F .903 can t 3.ch mEoEmom 55% c. AuEav mum mEom

new ano cosocouoE new .BEV E903 >08 .AEe $290... =8 EoEmom x3 .29 3mm .3 coo—2 .N mick

5.9 m3:
$0.9 Raﬁ

6i 8.»:
awe 9.82

63: 8.3
see «$3

63: aims
is 28$

mum Yarn E.

63.: 93mm
Ammo: on. .2

am. E 8.93
an»: Eda

Emmy 0909
39m: odeN

83: 2.88
8%: 5.3

=58 leh E.

23...: 8.3%..
$3: 8.3m

and: amok
ave: 8.8m

5.88 86mm
sod: «team

33: 28mm
and: 83mm

j a; 8m

3.: 8.2
an: 3:

28. : 3:
so: no.

Go: Zr
64.9 a:

see 8.9
3.3 2:

mwm

: c was.
a: swam“. :5
E was.
a: mass“. e
E was.
$5 mass“. m
av was.
E 0.2.0". .380
: xow allllllcoEEoF

136

(Table 2). The interaction between treatment and gonadal sex was not
statistically signiﬁcant (F = 1.424, p = 0.244). Post hoc comparisons demonstrate
that T- and DHT-treated animals, regardless of gonadal sex, had larger cells than
control animals (both p < 0.05). Neither steroid treatment (F = 0.191, p = 0.902)
nor gonadal sex (F = 2.143, p = 0.148) nor the interaction between these

variables (F = 0.734, p = 0.535) affected body weight (T able 2).

DISCUSSION

Summary

Differentiation of the copulatory neuromuscular system in green anoles
occurs during embryonic development but after gonadal differentiation, which is
largely complete by incubation day 10 (total incubation time at 28'C is 34 days on
average). Both male and female embryos possess hemipenes and the
associated musculature eariy in development, but these stmctures completely
regress in females by incubation day 19. The motoneurons projecting to
copulatory muscles are not sexually dimorphic in size or number in hatchlings. As
these cells are both larger and more numerous in adult males than females, this
result suggests a dissociation in the time course of differentiation in central
versus peripheral neuromuscular components in this system. Treating embryos
with androgens (either T or DHT) rescued the hemipenes and muscles in female
embryos, whereas treatment with E caused the system to regress in males,
suggesting that both types of steroid may be actively involved in sexual

differentiation in this system. The effects of androgenic and estrogenic hormones

137

on hemipenis development in green anoles are generally consistent with those
reported for the few other lizard species investigated (reviewed in Raynaud and
Pieau, 1985). In addition, the present work demonstrates that the associated
muscles are also responsive to both classes of steroid. Overall, steroid treatment
increased the size and number of associated motoneurons, although as no
signiﬁcant pairwise comparisons were detected it is difﬁcult to determine the
roles of the speciﬁc hormones. In sum, embryonic androgens may masculinize
and estrogens may feminize peripheral components of the lizard copulatory
neuromuscular system, while other mechanisms (or the same mechanisms at a

different time point) likely regulate differentiation of the associated motoneurons.

Dissociation between neural and muscular components

The timing of differentiation of neuromuscular components differs across
species. For example, dimorphisms in the central nervous system precede
differentiation of the muscles in the vocal neuromuscular systems of plainﬁn
midshipman ﬁsh (Brantley et al., 1993a; Knapp et al., 1999) and African clawed
frogs (Kelley and Dennison, 1990). In contrast, in the rodent copulatory system,
dimorphisms in the muscles develop ﬁrst and mediate differentiation of the
associated motoneurons likely via release of one or more trophic factors
(reviewed in Breedlove et al., 2002). The present data suggest that differentiation
of the lizard copulatory system is in some ways similar to that in rodents as the
hemipenes and RPM and TPN muscles cleariy differentiate prior to the

motoneurons. However, a particulariy intriguing aspect of the present data is the

138

 

large dissociation in timing between differentiation of the motoneurons compared
to the muscles and target stnrctures (hemipenes), which is not seen in other
species. While the motoneurons are clearly organized in the ventral horn by the
time of hatching, they are not sexually dimorphic in size or number; males had
more cells than females in Experiment 1 and in three of four groups in
Experiment 2 but the difference was not statistically signiﬁcant. This lack of a sex
difference in cell number exists despite a complete regression of the target
musculature in females more than 15 days earlier (Figure 3). Even more striking
is the fact that these motoneurons are still not dimorphic in size or number in
juvenile anoles at 3 months after hatching (Lovem et al., 2004b; Chapter 4).
Comparing cell counts in hatchling (present data) or juvenile (Lovem et al.,
2004b; Chapter 4) with adult anoles (Ruiz and Wade, 2002; Holmes and Wade,
2004a; Chapter 1) suggests that the sex difference in cell number may arise from
a decrease in motoneurons in adult females rather than motoneuron genesis in
adult males. One hypothesis is that the motoneurons differentiate as the animals
enter their ﬁrst breeding season, the ﬁrst time point at which males exhibit adult
levels of T (Lovem et al., 2001). However, the motoneurons do not appear to be
sensitive to steroid hormones in later adulthood (at least in regards to soma size;
Holmes and Wade, 2004b; Chapter 2), so an androgenic mechanism may not be
involved. Another possibility is that differentiation of the cells is inﬂuenced by sex
differences in use that occur once the animals reach reproductive maturity. Thus,

the motoneurons may exhibit an as yet undeﬁned “pubertal” period of sensitivity

139

to T at the beginning of reproductive maturity, or differentiation may be mediated

by a non-steroidal mechanism.

Steroid hormone effects on copulatory neuromuscular structures

Differentiation of sexually dimorphic neuromuscular systems is typically
mediated by gonadal steroids. Juvenile androgens masculinize components of
the vocal motor systems of African clawed frogs (Sassoon and Kelley, 1986;
Robertson et al., 1994; Kay et al., 1999), plainﬁn midshipman ﬁsh (Brantley et al.,
1993b), and possibly zebra ﬁnches, although exogenous estrogens also feminize
other components in this species (Wade et al., 2002; see below). Perhaps the
closest comparison to the present data, however, is the rodent copulatory
neuromuscular system. Here the neural and muscular tissues develop in both
sexes but regress in females shortly after birth; androgens clearly mediate
differentiation of the muscles, which in turn control survival of the motoneurons
(e.g., Breedlove et al., 1982; Breedlove and Arnold, 1983a, 1983b; Nordeen et
al., 1985). Differentiation of the green anole copulatory neuromuscular system
parallels this pattern in that hemipenes, RPMs, TPNs, and associated
motoneurons are present in both sexes embryonically, and the peripheral
components regress early in development in females. As treatment with
androgens can in both cases rescue these structures in females, it is clear that a
masculinizing role for androgens appears to be conserved.

Of particular interest, though, is the active feminization induced by E in

anoles that appears less conserved across vertebrates. E can play a role in the

140

feminization of the sexually dimorphic passerine song system as it decreases
syrinx weight and muscle ﬁber size in juvenile zebra ﬁnches (Wade et al., 2002).
In contrast, E has masculinizing properties in the developing rodent copulatory
system as it increases both the soma size (Breedlove, 1997b) and dendritic
outgrowth (Goldstein and Sengelaub, 1994) of motoneurons. It will be important
in the future to solidify the role of exogenous androgens and estrogens in anoles
by inhibiting exposure to these hormones during embryonic development.

It is also worth exhibiting some caution in interpreting the results for
another reason. In contrast to mammals, steroid hormones strongly inﬂuence
differentiation of both avian and reptilian gonads. Inhibiting E synthesis induces
testicular development in female birds (Elbrecht and Smith, 1992; Wade and
Arnold, 1996) while E treatment disrupts testicular morphology in males (Wade et
al., 1997). Similariy, treating reptilian eggs with E causes embryos to develop
ovaries in some species (Bull et al., 1988; Crews et al., 1991; Wibbels and
Crews, 1992; Tousignant and Crews, 1994; Wibbels and Crews, 1995), whereas
T induces testicular development in at least one other (Ganesh and Raman,
1995). As there is no method at present to determine the genetic sex of green
anoles, we cannot be sure that the gonads were not affected in our steroid-
treated animals. In particular, one might be concerned that the low number
animals with testes following E treatment (4 of 18) is a result of E feminizing the
gonads. However, it seems unlikely that our treatments altered gonadal
development as they commenced after the time point of differentiation and we

saw no ambiguity in their structure in Experiment 2. More likely is our intended

141

scenario of the lipophillic steroids providing exogenous hormone exposure to the
animal until hatching, probably following absorption into the yolk. Consistent with
this idea is that androgen treated animals had larger renal sex segment cells
(indicates androgenic activity) at the time of hatching - two weeks after the
second (and ﬁnal) steroid treatment. An additional possibility is that steroid
treatment might have altered production of endogenous gonadal hormones,
resulting in an indirect effect on copulatory neuromuscular morphology.

It has been hypothesized that mechanisms of sex determination may dictate
which sex is “default", that the homogametic sex is neutral and easier to sex
reverse whereas the heterogametic sex possesses the dominant gonad that
produces hormones controlling differentiation (Adkins-Regan, 1981). The present
data challenge these ideas. While estrogenic steroids actively feminize genitalia
in Lacerta viridr’s, a lizard with female heterogametic sex determination (ZZ/ZW,
as in birds), E has the same effect in green anoles, which, like mammals, exhibit
male heterogametic sex determination (XX/XY, as in mammals) (Adkins-Regan,
1981). As such, it may be that the role of steroid hormones in sexual
differentiation is not determined by mechanism of sex determination per se.
Indeed, the present data suggest that androgenic and estrogenic hormones both
actively contribute to the differentiation of male and female green anoles,

implying that neither may be “default” in this species.

142

DISCUSSION

Species used in the study of sexually dimorphic neuromuscular systems
involved in reproduction typically are useful for examining either courtship
systems (e.g., in ﬁshes, frogs, and birds; see Chapter 1) or copulatory systems
(e.g., in mammals; see Chapter 1), but generally not both. Many non-mammalian
species do not possess intromittant organs for reproduction, and most mammals
do not exhibit courtship behaviors that are easily measurable in the laboratory.
Thus, green anoles provide the opportunity to simultaneously examine
mechanisms regulating two neuromuscular systems required for the full suite of
masculine reproductive behaviors: one controlling the use of their copulatory
organs and another controlling extension of a throat fan (dewlap) used in
courtship. The present work largely focused on the green anole copulatory
neuromuscular system because it exhibits a remarkable degree of sexual
dimorphism. Adult males possess two hemipenes and their associated muscles
and motoneurons while adult females do not (Ruiz 8: Wade, 2002; Holmes and
Wade, 2004a; Chapter 1). Given its critical role in male reproduction, it is not
surprising that components of this system are responsive to testosterone (T), the
most potent steroid activator of male-typical reproductive behaviors in this
species (Mason & Adkins, 1976; Crews et al., 1978; Adkins & Schlesinger, 1979;
Winkler & Wade, 1998). What is surprising, however, is the pattern of

responsiveness that these structures exhibit. The results presented in the

143

preceding chapters clearly demonstrate several conclusions about T action on

the anole copulatory system.

1) Seasonally-induced morphological plasticity is limited

Precedence for naturally occurring neural and muscular plasticity exists in
seasonally breeding species. For example, the song control system in the
passerine brain exhibits remarkable structural and functional plasticity
(Brenowitz, 2004; Park et al., 2005) associated with periods of song learning.
Similariy, the copulatory neuromuscular system of seasonally breeding rodents
changes in parallel with reproductive behavior. Motoneurons in the spinal
nucleus of the bulbocavemosus and their associated muscles are larger in males
housed in breeding than non-breeding environmental conditions (Forger and
Breedlove, 1987; Hegstrom and Breedlove, 1999). The effects of season are
typically due to direct activity of T; levels of the hormone increase corresponding
to, or in anticipation of, breeding conditions. However, photoperiod can also
induce plasticity in some neuromuscular measures independent of androgens
(e.g., neuromuscular junction size; Hegstrom et al., 2002).

Green anoles are strict seasonal breeders, exhibiting courtship and
copulatory behaviors only when days are long and temperatures are warm
(Jenssen et al., 1995; 1996). T levels rise and fall in parallel with reproduction
and are approximately twice as high during the breeding than non-breeding
season (Lovem et al., 2001). The muscles and motoneurons that control dewlap

extension do not appear to respond morphologically to changes in environmental

144

I I

 

conditions (O’Bryant and Wade, 1999). However, as both males and females
occasionally use their dewlaps in the non-breeding season, it is possible that a
lack of seasonal effect on morphology is due to required maintenance of the
system. Comparing copulatory neuromuscular morphology in gonadally intact
male anoles housed in either breeding or non-breeding environmental conditions
(Holmes and Wade, 2004b; Chapter 2) demonstrated that seasonal cues also do
not induce morphological changes in this system. The only measure that differed
between seasons in gonadally intact animals was a very small change in
motoneuron soma size, which was not limited to copulatory motoneurons. One
possibility is that copulatory neuromuscular morphology is insensitive to T and/or
other seasonal cues. More likely, however, is that endogenous ﬂuctuations in T
are not robust enough to induce morphological change. Although the copulatory
system may not exhibit remarkable naturally occurring ﬂuctuations in
morphology, it does have the capacity for plasticity. Indeed, seasonal cues have
indirect effects on copulatory neuromuscular morphology by mediating the

responsiveness of these structures to exogenous T (see below).

2) menops T has trophic effects on the whery in adulthood

While T is the most potent steroid for the activation of both courtship and
copulatory behaviors in adult green anoles (Mason and Adkins, 1976; Crews et
al., 1978; Adkins and Schlesinger, 1979; Winkler and Wade, 1998), it does not
induce morphological change in the adult dewlap neuromuscular system. Dewlap

motoneurons and CH muscle ﬁbers do not appear to differ between castrated

145

 

and T-treated males (O’Bryant and Wade, 1999). Yet T-treatment increases RPM
muscle ﬁber size and hemipenis size in the copulatory system. Interestingly, this
effect is only seen when animals are housed in breeding environmental
conditions (Holmes and Wade, 2004b; Chapter 2), suggesting that seasonal cues
can inﬂuence responsiveness to exogenous T.

The trophic effect of T on the muscles and hemipenes of the copulatory

 

system is consistent with other species; exogenous T commonly inﬂuences
peripheral components in sexually dimorphic neuromuscular systems of a wide I . 1
variety of vertebrates. For example, adult T-treatment increases muscle ﬁber size
in the syrinx of female zebra ﬁnches (Wade and Buhlman, 2000). Similariy,
androgens increase the mass of external oblique muscles of gray tree frogs
(used in sound production; Girgenrath and Marsh, 2003) and sonic muscles in
Type II male plainﬁn midshipman ﬁsh (Lee and Bass, 2005). Testicular implants
increase muscle ﬁber number (Watson et al., 1993) and long term T-treatment
masculinizes muscle ﬁber type (Sassoon et al., 1987; Tobias et al., 1991) in the
larynx of adult female Xenopus. Finally, T increases ﬁber size (Rand and
Breedlove, 1991) and twitch capacity (Souccar et al.,1992) in the muscles of the
adult rodent copulatory system.

Given that exogenous T, but not normally occurring T ﬂuctuation, alters
copulatory morphology, it is important to address how these data are relevant to
the reproductive function of green anoles. The difference in circulating T between
castrated males and those that received T implants in the studies in this

dissertation was substantially greater than the difference between intact males in

146

the simulated breeding and non-breeding seasons. These two approaches, of
comparing gonadally intact males as well as castrated males with and without T-
treatment, allows us to explore both what naturally occurs in these animals, and
importantly, what the possibilities or limits are for the system. In regards to the
copulatory system, the data suggest that sufﬁcient T is present in the non-
breeding season to maintain copulatory neuromuscular morphology in gonadally
intact males. Alternatively, decreased responsiveness to T in non-breeding
conditions, as demonstrated in castrated males treated with T, may serve to

protect copulatory structures during periods of low circulating hormones.

3) Morpholmy of copulatory muscles, but not the motoneurons, is responsive to
I

The data presented above demonstrate a dissociation between T
responsiveness in the periphery of the copulatory system versus the central
nervous system. While the hemipenes and muscles appear sensitive to
exogenous T in adulthood, the corresponding T17-S1 motoneurons are not, at
least in terms of soma size (Holmes and Wade, 2004b; Chapter 2). This
insensitivity is consistent with the dewlap motoneurons in green anoles, which
also fail to exhibit changes in soma size following androgen manipulation
(O’Bryant and Wade, 1999), as well as the syrinx motoneurons in the zebra ﬁnch
song system (Wade and Buhlman, 2000) and sonic motor nucleus in Type II

male plainﬁn midshipman ﬁsh (Lee and Bass, 2005). In contrast, motoneuron

soma size in the adult female canary song system increases in response to T-

147

treatment (DeVoogd et al., 1991), as do motoneurons in the rodent copulatory
system (Breedlove and Arnold, 1981). Taken together, responsiveness in the
muscles and/or target tissues does not necessarily parallel sensitivity in the
corresponding motoneurons. Of course, there are many additional morphological
and functional parameters that T could be manipulating. As such, a true
dissociation between motoneuron and muscle sensitivity has not yet been

determined.

4) T mp1 have its morpmlgqical effects via mulation of AR

T increases AR expression in the anole copulatory neuromuscular system
in a manner parallel to its effects on morphology, whereby the percent of AR+
nuclei is greater in RPM muscle ﬁbers following T-treatment (Holmes and Wade,
submitted; Chapter 3). Interestingly, T increased AR expression in the RPMs of
males housed in either breeding or non-breeding conditions, although
morphological effects are only seen during the breeding season. While AR is
expressed in a high percentage of cells that do not exhibit morphological
responsiveness to T, including dewlap muscle ﬁbers (approximately 60-70%) and
copulatory motoneurons (approximately 70-80%), T-treatment did not alter the
percentage of these cells that were AR+ (Holmes and Wade, submitted; Chapter
3). Furthermore, treatment with estradiol causes atrophy of the hemipenes and
muscles in adulthood (Holmes and Wade, unpublished observations),
demonstrating that the trophic effects of T are not a result of aromatization to

estradiol. As AR expression is relatively high in structures that do not respond

148

 

morphologically to T, the data also demonstrate that presence of AR does not
alone confer morphological responsiveness to T, although it is critical to
acknowledge that additional morphological and/or physiological parameters that
we have not yet measured may be sensitive to this hormone. This dissociation
between AR expression and morphological responsiveness to testosterone is
similar to the dorsolateral nucleus of the rat copulatory neuromuscular system.
Approximately 90% of the motoneurons projecting to the external urethral
sphincter are AR+ (Jordan, 1997), but these cells to not change in size following
T manipulation (Collins et al., 1992).

These types of data have interesting implications for potential sites of
action of this hormone. Does T act directly on the muscle to increase ﬁber size or
does the trophic response occur indirectly due to increased hemipenis size?
Although unlikely, T could also be acting on AR in the copulatory motoneurons
and having an anterograde effect on muscles and hemipenes. Given the
increased expression of AR in RPM muscle ﬁbers following T-treatment (Holmes
and Wade, submitted; Chapter 3), as well as the high expression in the
hemipenes, it is likely that testosterone can act directly on both muscles and
hemipenes, causing their respective sizes to change in concert with each other.
It is likely that a minimum hemipenis size is required for male reproductive
success and it is possible, although purely speculative, that increased hemipenis
size increases fertility (perhaps facilitating sperm transfer, sperm competition,
etc). Perhaps larger hemipenes would need larger muscles to maneuver them,

explaining why the effects of T are typically parallel in both structures. But why is

149

 

there no effect of T on the size of the corresponding motoneurons? The
functional signiﬁcance of the relationship between motoneuron and muscle ﬁber
size in this system is less clear. While muscle and motoneuron are typically size-
matched (reviewed in Gordon et al., 2004) and respond in parallel to T (reviewed
in Breedlove et al., 2002) in mammalian neuromuscular systems, it is not
apparent that a similar relationship exists in the lizard copulatory neuromuscular

system.

5) Testosterone has trophic effects on whery during juvenile development
The dewlap neuromuscular system begins to differentiate in juveniles
between 30 and 90 days post-hatching. The second ceratobranchial cartilages
and CH muscles become larger in males than females during this period
(O’Bryant and Wade, 2001). Treating juvenile females with T increases cartilage
length and CH muscle ﬁber size, demonstrating that the dewlap system is
sensitive to T during development (Lovem et al., 2004b; Chapter 4). In contrast,
hemipenes or RPM muscle ﬁbers were not evident in these females, suggesting
that sexual differentiation of the copulatory neuromuscular system is complete by
this time, and is not reversible by T-treatment once ﬁnished (Lovem et al., 2004b;
Chapter 4). Paralleling T-manipulation in adulthood, castrating juvenile males
decreased hemipenis size, and to a lesser extent, RPM ﬁber size. Both courtship
and copulatory motoneurons were not sexually dimorphic in size and number on
the day of hatching, and neither population responded to T-manipulation (Lovem

et al., 2004b; Chapter 4). Therefore, the dissociation between T sensitivity in

150

 

peripheral copulatory structures and their associated motoneurons is also
present in juveniles. Furthermore, a dissociation between the timing of sexual
differentiation in the dewlap versus copulatory systems clearly exists. While
periphery of copulatory system is sensitive to T manipulation in juveniles,
differentiation of these structures is complete by day 90 post-hatching, illustrating
that the time point of differentiation in this system is much earlier than the dewlap
system (see below).

Juvenile development is also a period for steroid sensitivity in other
species. T and 11-ketotestosterone increase sonic muscle mass and ﬁber
phenotype in juvenile male and female plainﬁn midshipman ﬁsh, as well as
muscle ﬁber number in males only (Brantley et al., 1993b). T masculinizes
muscle ﬁber type and size in the larynx of juvenile Xenopus (Sassoon et al.,
1987) and DHT masculinizes motoneuron number in female Xenopus tadpoles
(Kay et al., 1999). Post-hatching DHT-treatment increases nXllts motoneuron
soma size and overall syrinx weight in female zebra ﬁnches, and E decreases
syrinx weight and muscle ﬁber size in juvenile males (Wade et al., 2002). Finally,
peripubertal steroids inﬂuence motoneuron morphology in the rodent copulatory
neuromuscular system. T, DHT, and E increase dendritic length in juvenile male
rats, but only T and DHT increase motoneuron soma size (Goldstein and
Sengelaub, 1990; Goldstein and Sengelaub, 1994). Interestingly, T may increase
motoneuron number in the copulatory system of pubertal male Mongolian gerbils

(Siegford and Ulibarri, 2004).

151

 

 

6) Testosterone has trophic effects on periphery during embryonicﬁdgrelopment
The present data demonstrate that the hemipenes and associated
muscles are present in both sexes during early embryonic development but
completely regress in female embryos by the time of hatching (Holmes and
Wade, in press; Chapter 5). They also suggest that androgenic mechanisms are
likely involved to some extent in the differentiation of this system. Treating
embryos with androgens (either T or DHT) rescued the hemipenes and
associated muscles in almost all embryos (Holmes and Wade, in press; Chapter
5). Androgens also mediate differentiation of sexually dimorphic neuromuscular
systems in other species, although differences in the details cleariy exist.
Androgens masculinize components of the vocal motor systems of African
clawed frogs (Sassoon and Kelley, 1986; Robertson et al., 1994; Kay et al.,
1999), plainﬁn midshipman ﬁsh (Brantley et al., 1993b), and possibly zebra
ﬁnches, although exogenous estrogens also feminize other components in this
species (Wade et al., 2002; see below). In the rodent copulatory neuromuscular
system, the neural and muscular tissues develop in both sexes but regress in
females shortly after birth; androgens clearly mediate differentiation of the
muscles, which in turn control survival of the motoneurons (e.g., Breedlove et
al., 1982; Breedlove and Arnold, 1983a, 1983b; Nordeen et al., 1985). As such,
differentiation of the green anole copulatory neuromuscular system most closely
parallels the mammalian pattern in that hemipenes, RPMs, TPNs, and
associated motoneurons are present in both sexes embryonically, and the

peripheral components regress eariy in development in females (Holmes and

152

 

Wade, in press; Chapter 5). As treatment with androgens can rescue these
structures in females, it is clear that a masculinizing role for androgens in
peripheral structures is somewhat conserved across species.

There may also be a role for estrogens in differentiation of the anole
copulatory system as treating embryos with estradiol caused a regression of
hemipenes and associated muscles (Holmes and Wade, in press; Chapter 5).

This is of particular interest as the active feminization by E appears less

‘ my "q-l'ff ‘u

conserved across vertebrates. E can play a role in the feminization of the

 

sexually dimorphic passerine song system as it decreases syrinx weight and
muscle ﬁber size in juvenile zebra ﬁnches (Wade et al., 2002). In contrast, E has
masculinizing properties in the developing rodent copulatory system as it
increases both the soma size (Breedlove, 1997b) and dendritic outgrowth
(Goldstein and Sengelaub, 1994) of motoneurons. It will be important in the
future to better characterize the role of exogenous androgens and estrogens in
anoles by inhibiting exposure to these hormones during embryonic development.

The time point of differentiation of the copulatory motoneurons, and the
mechanism(s), has not yet been identiﬁed. They have not differentiated by the
time of hatching (Holmes and Wade, in press; Chapter 5) or by 90 days post-
hatching (Lovem et al., 2004b; Chapter 4), but are clearly sexually dimorphic in
both size and number in adults (Ruiz and Wade, 2002; Holmes and Wade,
2004a; Chapter 1). These cells also fail to respond to hormone manipulation at
all life stages that have been investigated. The estimated number of

motoneurons at both developmental time points demonstrates that the masculine

153

number of cells is present in both sexes, suggesting that differentiation of
motoneuron number is due to a loss of cells (or cell death) in females rather than
an increase in cells (motoneuron genesis) in males. If this is indeed the case, it is
perplexing that these cells are surviving in females at least 3 months after their

target muscles have completely regressed.

Taken together, it is clear that peripheral components of the anole
copulatory neuromuscular system, the hemipenes and RPM muscle ﬁbers,
exhibit trophic responses to T during embryonic (Holmes and Wade, in press;
Chapter 5) and juvenile (Lovem et al., 2004b; Chapter 4) development as well as
in adulthood (Holmes and Wade, 2004b; Chapter 2). This lifelong sensitivity to
testosterone is consistent with the hypothesis that androgens are necessary for
both the differentiation and maintenance of these sexually dimorphic structures.
The data also suggest that the effects of testosterone are likely due to activation
of AR. Further understanding the degree to which components of the system
respond to T in regards to additional morphological and functional measures will
help to elucidate the extent of this maintenance, as well as shed light on the
signiﬁcance that this T responsiveness has on the reproductive success of male
green anoles.

Prior to the work on the green anole, a role for androgens had been
demonstrated in sexually dimorphic neuromuscular systems in all vertebrate
classes (including ﬁsh, frogs, birds, and mammals) except for reptiles. As such,

these results, in addition to work demonstrating androgenic mediation of the

154

 

leopard gecko copulatory system (Holmes et al., 2005), ﬁll a critical gap in the
comparative approach to understanding steroid hormone effects on dimorphic
neuromuscular systems.

Comparing across species, some relationship between structure and
function clearly exists in all sexually dimorphic neuromuscular systems. But,
collectively, the work with the anole has made it clear that this relationship
between morphology and behavior is not nearly as simple as it ﬁrst appears. This
idea is illustrated with two points: 1) both the dewlap and copulatory systems
experience a huge seasonal variation in use but do not exhibit naturally occurring
morphological plasticity and 2) differentiation and adult plasticity in the copulatory
muscles does not parallel the corresponding motoneurons. In addition, the
present work illustrates selective sensitivity to T within and between
neuromuscular systems. Directly comparing the dewlap and copulatory
neuromuscular systems is particularly powerful because they are presumably
exposed to the same circulating hormones at the same developmental time
points. Given that the systems differentiate at independent stages and have
dissimilar responses to T in adulthood, it is clear that sensitivity to this hormone
is tissue speciﬁc and can be turned on and off at different times. This speciﬁcity
is likely mediated by factors in addition to the presence of AR; this protein is
expressed in neuromuscular effectors that respond to T as well as those that do
not (Holmes and Wade, submitted; Chapter 3). Candidate mechanisms for
conferring selective hormone sensitivity include steroid metabolizing enzymes

(e.g. 5-a-reductase and aromatase) as well as steroid hormone receptor co-

155

activators. Understanding how additional morphological, neurochemical, and
electrophysiological parameters are inﬂuenced by T in diverse species will help

identify these mechanisms and how they are conserved across vertebrates.

156

LITERATURE CITED

Adkins E, Schlesinger L (1979). Androgens and the social behavior of male and
female lizards (Anolis carolinensis). Horm Behav 13: 139-152.

 

Adkins-Regan E (1981). Early organizational effects of hormones: an
evolutionary perspective. In: Adler NT, editor. Neuroendocrinolqay of
Reproduction: Pimiology and Behavior. New York: Plenum Press. p 159-228.

Andersson M (1994). Sexual Selection. Princeton University Press, Princeton
USA.

Akita K (1992). An anatomical investigation of the muscles of the pelvic outlet in
iguanas (Iguanidae Iguana iguana) and varanus (Varanidae Varanus (dumerillir))
with special reference to their nerve supply. Ann Anat 174:119-129.

Arnold AP (1975). The effects of castration and androgen replacement on song,

courtship, and aggression in zebra ﬁnches (Poephila guttata). J Exp Zool 191:
309-326.

Arnold AP (1980). Effects of androgens on volumes of sexually dimorphic brain
regions in the zebra ﬁnch. Brain Res 185: 441-444.

Arnold AP (1997). Sexual differentiation of the zebra ﬁnch song system: positive

evidence, negative evidence, null hypotheses, and a paradigm shift. J Neurobiol
33: 572-584.

 

Arnold AP, Breedlove SM (1985). Organizational and activational effects of sex
steroids on brain and behavior: a reanalysis. Horm Behav 19: 469-498.

 

Arnold EN (1984). Variation in the cloacal and hemipenial muscles of lizards and
its bearing on their relationships. Syngool Soc London 52: 47-85.

Balice-Gordon RJ, Breedlove SM, Bernstein S, Lichtman JW (1990).
Neuromuscular junctions shrink and expand as muscle ﬁber size is manipulated:

in vivo observations in the androgen-sensitive bulbocavemosus muscle of mice.
J Neurosci 10: 2660-2671.

Bass AH (1985). Sonic motor pathways in teleost ﬁshes: A comparative HRP
study. Brain Behav Evol 27:1 15-131.

Bass AH, Marchaterre MA (1989a). Sound-generating (sonic) motor system in a
teleost ﬁsh (Porichthys notatus): sexual polymorphism in the ultrastructure of
myoﬁbrils. J Comp Neurol 286: 141-153.

157

 

Bass AH, Marchaterre MA (1989b). Sound-generating (sonic) motor system in a
teleost ﬁsh (Porichthys notatus): sexual polymorphisms and general synaptology
of sonic motor nucleus. J Comp Neurol 286: 154-69.

Bass AH, Baker R (1990). Sexual dimorphisms in the vocal control system of a
teleost ﬁsh: morphology of physiologically identiﬁed neurons. J Neurobiol 21:
1 155-1 168.

Bels VL (1990). The mechanism of dewlap extension in Anolis carolinensis
(Reptilia: Iguanidae) with histological analysis of the hyoid apparatus. J Mor_'ph
206:225-244.

Bertolucci C, Foa A (1998). Seasonality and role of SCN in entrainment of lizard
circadian rhythms to daily melatonin injections. Am J Physiol 274: R1004-1014.

Bertolucci C, Foa A, Van't Hof TJ (2002). Seasonal variations in circadian
rhythms of plasma melatonin in ruin lizards. Horm _Behav 41: 414-419.

Bottjer SW, Glaessner SL, Arnold AP (1985). Ontogeny of brain nuclei controlling
song learning and behavior in zebra ﬁnches. J Neurosci 5: 1556-1562.

Brantley RK, Bass AH (1994). Alternative male spawning tactics and acoustic
signals in the plainﬁn midshipman ﬁsh Porichthys notatus Girard (T eleostei,
Batracoididae). Etholpgy 96: 213-232.

Brantley RK, Tseng J, Bass AH (1993a). The ontogeny of inter- and intrasexual
vocal muscle dimorphisms in a sound-producing ﬁsh. Brain Behav Evol 42: 336-
349.

Brantley RK, Marchaterre MA, Bass AH (1993b). Androgen effects on vocal

muscle structure in a teleost ﬁsh with inter- and intra-sexual dimorphism. J Morph
216: 305-318.

Brantley RK, Wingﬁeld JC, Bass AH (1993c). Sex steroid levels in Porichthys
notatus, a ﬁsh with alternative reproductive tactics, and a review of the hormonal
bases for male dimorphism among teleost ﬁshes. Horrn ghav 27: 332-347.

 

Breedlove SM (1997a). Sex on the brain. Nature 389: 801.

Breedlove SM (1997b). Neonatal androgen and estrogen treatments masculinize
the size of motoneurons in the rat spinal nucleus of the bulbocavemosus. Qe_ll
Mol Neurobiol 17: 687-697.

Breedlove SM, Arnold AP (1980). Hormone accumulation in a sexually dimorphic
motor nucleus of the rat spinal cord. Science 210: 564-566.

158

 

 

Breedlove SM, Arnold AP (1981). Sexually dimorphic motor nucleus in the rat
lumbar spinal cord: response to adult hormone manipulation, absence in
androgen-insensitive rats. Brain Res 225: 297-307.

Breedlove SM, Arnold AP (1983a). Hormonal control of a developing
neuromuscular system. |. Complete demasculinization of the male rat spinal

nucleus of the bulbocavemosus using the anti-androgen ﬂutamide. J Neurosci 3:
417-423.

Breedlove SM, Arnold AP (1983b). Hormonal control of a developing
neuromuscular system. ll. Sensitive periods for the androgen-induced

masculinization of the rat spinal nucleus of the bulbocavemosus. J Neurosci 3:
424-432.

Breedlove SM, Jacobson CD, Gorski RA, Arnold AP (1982). Masculinization of
the female rat spinal cord following a single neonatal injection of testosterone
proprionate but not estradiol benzoate. Brain Res 237: 173-181.

Breedlove SM, Cooke BM, Jordan CL (1999). The orthodox view of brain sexual
differentiation. Brain Behav Evol 54: 8-14.

 

Breedlove SM, Jordan CL, Kelley DB (2002). What neuromuscular systems tell
us about hormones and behavior. In Hormones, Brain and Behavior, Pfaff DW,
Arnold AP, Etgen AM, Fahrbach SE, Rubin RT, editors. Academic Press: USA.

Brenowitz EA (2004). Plasticity of the adult avian song control system. Ann N Y
Acad Sci 101 6: 560-585.

Brenowitz EA, Lent K (2002). Act locally and think globally: lntracerebral
testosterone implants induce seasonal-like growth of adult avian song control
circuits. Proc Natl Acad Sci 99: 12421-12426.

Bull JJ, Gutzke WHN, Crews D (1988). Sex reversal by estradiol in three reptilian
orders. Gen Coy Endo 70: 425-428.

Carsia RV, John-Alder J (2003). Seasonal alterations in adrenocortical cell
function associated with stress-responsiveness and sex in the eastern fence
lizard (Sceloporus undulatus). Horm Behav 43: 408-420.

 

Cihak R, Gutmann E, Hanzllkova V (1970). lnvolution and honnone-induced
persistence of the M. sphincter (levator) ani in female rats. J Anat 106: 93-110.

Clancy B, Cauller LJ (1998). Reduction of background autoﬂuorescence in brain
sections following immersion in sodium borohydride. J Neurosci Methods 83: 97-
102.

 

159

 

Collins WF, Seymour AW, Klugewicz SW (1992). Differential effects of castration

on the somal size of pudendal motoneurons in the adult male rat. Brain Res 577:
326-330.

Cooke B, Hegstrom CD, Villeneuve LS, Breedlove SM (1998). Sexual
differentiation of the vertebrate brain: principles and mechanisms. Front
Neuroendocrinol 19: 323-362.

Cooke BM, Hegstrom CD, Breedlove SM (2002). Photoperiod-dependent
response to androgen in the medial amygdala of the Siberian hamster, Phodopus
sungonrs. J Biol Rhyth 17: 147-154.

Crews D (1978). Hemipenile preference: Stimulus control of male mounting
behavior in the lizard Anolis carolinensis. Science 199:195-196.

 

Crews D (1980). Interrelationships among ecological, behavioral, and
neuroendocrine processes in the reproductive cycle of Anolis carolinensis and
other reptiles. In Advances in the study of behavior, Rosenblatt JS et al., editors.
Academic press: New York.

Crews D, Bull JJ, Wibbels T (1991). Estrogen and sex reversal in turtles: a dose-
dependent phenomenon. Gen Comp Endo 81: 357-364.

Crews D, Traina V, Wetzel FT, Muller C (1978). Hormonal control of male
reproductive behavior in the lizard, Anolis carolinensis: role of testosterone,
dihydrotestosterone, and estradiol. Endocrinology 103: 1814-1821.

Cuellar HS, Roth JJ, Fawcett JD, Jones RE (1972). Evidence for sperm
sustenance by secretions of the renal sexual segment of male lizards, Anolis
carolinensis. Herpetolpgica 28: 53-57.

Danrvin C (1871 ). The Descent of Man, and Selection in Relation to Sex. Murray:
London.

DeCourcy KR, Jenssen TA (1994). Structure and use of male territorial headbob
signals by the lizard Anolis carolinensis. Anim Behav 47: 251-262.

 

Dederen PJ, Gribnau AA, Curfs MH (1994). Retrograde neuronal tracing with
cholera toxin B subunit: comparison of three different visualization methods.
Histochem J 26: 856-862.

DeVoogd TJ, Pyskaty DJ, Nottebohm F (1991). Lateral asymmetries and

testosterone-induced changes in the gross morphology of the hypoglossal
nucleus in adult canaries. J ComLNeurol 307: 65-76.

160

 

Doriochter M, Astrow SH, Herrera AA (1996). Effects of testosterone on a
sexually dimorphic frog muscle: repeated in vivo observations and androgen
receptor distribution. J Neurobiol 25: 897-916.

Dufaure JP, Hubert J (1961). Table de developpement du lezard vivipare:
Lacerta (zootoca) vivipara jacquin. Arch Ant Micros Morphol EQ 50: 309-327.

Dunlap KD, Zakon HH (1998). Behavioral actions of androgens and androgen
receptor expression in the electrocommunication system of an electric ﬁsh,
Eigenmannia virescens. Horm Behav 34: 30-38.

Dunlap KD, McAnelly ML, Zakon HH (1997). Estrogen modiﬁes an
electrocommunication signal by altering the electrocyte sodium current in an
electric ﬁsh, Stemopygus. J Neurosci 17: 2869-2875.

Dunlap KD, Thomas P, Zakon HH (1998). Diversity of sexual dimorphism in
electrocommunication signals and its androgen regulation in a genus of electric
ﬁsh, Apteronotus. J Comp Physiol [A] 183: 77-86.

Elbrecht A, Smith RB (1992). Aromotase enzyme activity and sex determination
in chickens. Science 255: 467-70.

 

Fargo KN, Foster AM, Harty MW, Sengelaub DR (2003). Estrogen alters
excitability but not morphology of a sexually dimorphic neuromuscular system in
adult rats. J Neprobiol 56: 66-77.

Fischer LM, Catz D, Kelley DB (1995). Androgen-directed development of the
Xenopus laevis larynx: control of androgen receptor expression and tissue
differentiation. Dev Biol 170: 1 15-126.

Fishman RB, Chism L, Firestone GL, Breedlove SM (1990). Evidence for
androgen receptors in sexually dimorphic perineal muscles of neonatal male rats:

Absence of androgen acuumulation by the perineal motoneurons. J Neurobiol 21:
694-704.

 

Foa A, Magnone C, Bertolucci C (2002). Circadian organization in ruin lizards:
phase response curve for melatonin changes with season. J Commersiol A
Neuroethol Sens Neural Behav Physiol 188:141-145.

Font E (1991). Localization of brainstem motoneurons involved in dewlap
extension in the lizard, Anolis equestris (Iguanidae). Behav Brain Res 45: 171-
1 76.

Font E, Rome LC (1990). Functional morphology of dewlap extension in the
lizard Anolis equestris (Iguanidae). J Mogph 206: 245-258.

161

Forbes TR (1964). Intersexuality in reptiles. In: C.N. Armstrong and A.J.
Marhsall, editors. Intersexuality in Vertebratesﬁncludipq Man. Academic Press:
New York, p. 273-283.

Forger NG, Breedlove SM (1987). Seasonal variation in mammalian striated
muscle mass and motoneuron morphology. J Neurobiol 18: 155-165.

Forger NG, Fishman RB, Breedlove SM (1992). Differential effects of
testosterone metabolites upon the size of sexually dimorphic motoneurons in
adulthood. Horm Behav 26: 204-213.

Forlano PM, Sisneros JA, Knapp R, Bass AH (2003). Reproductive periodicity of
gonadal steroid hormone levels in the vocal plainﬁn midshipman, Porichthys
notatus. Horm Behav 44: 49.

Fraley GS, Ulibarri C (2001). Sexual dimorphism in the number and size of SNB
motoneurons: delayed development during normal ontogeny. Dev Brain Res 126:
57-64.

Freeman LM, Padgett BA, Prins GS, Breedlove SM (1995). Distribution of
androgen receptor immunoreactivity in the spinal cord of wild-type, androgen-
insensitive and gonadectomized male rats. J Neurobiol 27: 51-59.

 

Ganesh 8, Raman R (1995). Sex reversal by testosterone and not by estradiol or
temperature in Calotes versicolor, the lizard lacking sex chromosomes. J Exp
Zool 271: 139-144.

Girgenrath M, Marsh RL (2003). Season and testosterone affect contractile
properties of fast calling muscles in the gray tree frog Hyla chrysoscelis. Am J
Physiol 284: R1513-1520.

Glicksman MA, Chiu AY, Dionne CA, Harty M, Kaneko M, Murakata C,
Oppenheim RW, Prevette D, Sengelaub DR, Vaught JL, Neff NT (1998). CEP-
1347/KT7515 prevents motor neuronal programmed cell death and injury-
induced dedifferentiation in vivo. J Neurobiol 35:361-370.

 

Godsave SF, Lohmann R, Vloet RPM, Gahr M (2002). Androgen receptors in the
embryonic zebra ﬁnch hindbrain suggest a function for maternal androgens in
perihatching survival. J Comp Neurol 453: 57-70.

Goldstein LA, Sengelaub DR (1990). Hormonal control of neuron number in
sexually dimorphic spinal nuclei of the rat: IV. Masculinization of the spinal
nucleus of the bulbocavemosus with testosterone metabolites. J Neurobiol 21:
719-730.

162

 

Goldstein LA, Sengelaub DR (1994). Differential effects of dihydrotestosterone
and estrogen on the development of motoneuron morphology in a sexually
dimorphic rat spinal nucleus. J Neurobiol 25: 878-892.

Gordon T, Thomas CK, Munson JB, Stein RB (2004). The resilience of the size
principle in the organization of motor unit properties in normal and reinnervated
adult skeletal muscles. Can J Physiol Phannacol 82: 645-661.

Greenberg B, Noble GK (1944). Social behavior of the American chameleon
(Anolis carolinensis Voigt). Physiol Zool 17: 392-439.

Greenberg N, Scott M, Crews D (1984). Role of the amygdala in the reproductive
and aggressive behavior of the lizard, Anolis carolinensis. Physiol Behav 32:
147-1 51 .

Gunderson HJG (1986). Stereology of arbitrary particles. A review of unbiased
number and size estimators and the presentation of some new ones, in memory
of William R. Thompsen. J Microsc 143: 3-45.

Gurney ME (1981). Hormonal control of cell form and number in the zebra ﬁnch
song system. J Neurosci 1: 658-673.

Hannigan P, Kelley DB (1986). Androgen-induced alterations in vocalizations of
female Xenopus laevis: modiﬁability and constraints. J Comp Physiol [A] 158:
517-527.

Harding CF, Sheridan K, Walters MJ (1983). Hormonal speciﬁcity and activation
of sexual behavior in male zebra ﬁnches. Horrn Behav 17: 111-133.

Hart BL (1967). Testosterone regulation of sexual reﬂexes in spinal male rats.
Science 155: 1282-1284.

Hart BL (1983). Role of testosterone secretion and penile reﬂexes in sexual
behavior and sperm competition in male rats: A theoretical contribution. Physiol
Behav 31: 823-827.

Hegstrom CD, Breedlove SM (1998). Short day lengths delay development of the
SNB neuromuscular system in the Siberian hamster, Phodopus sungoms. J
Neurobiol 35: 355-360.

Hegstrom CD, Breedlove SM (1999). Seasonal plasticity of neuromuscular
junctions in adult male Siberian hamsters (Phodopus sungonrs). Brain Res 819:
83-88.

Hegstrom CD, Jordan CL, Breedlove SM (2002). Photoperiod and androgens act
independently to induce spinal nucleus of the bulbocavemosus neuromuscular

163

 

plasticity in the Siberian hamster, Phodopus sungorus. J Neuroendo 14: 368-
374.

 

Hews DK, Moore MC (1995). Inﬂuence of androgens on differentiation of
secondary sex characters in tree lizards, Urosaurus omatus. Gen Comp Endo
97: 86-102.

 

Hews DK, Moore MC (1996). A critical period for the organization of alternative
male phenotypes of tree lizards by exogenous testosterone? Physiol Behav 60:
425—429.

Hews DK, Knapp R, Moore MC (1994). Early exposure to androgens affects
adult expression of alternative male types in tree lizards. Horm Behav 28: 96-
1 15.

Holmes MM, Wade J (20043). Characterization of projections from a sexually
dimorphic motor nucleus in the spinal cord of adult green anoles. J Comp Neurol
471: 180-187.

 

Holmes MM, Wade J (2004b). Seasonal plasticity in the copulatory
neuromuscular system of green anole lizards: a role for testosterone in muscle
but not motoneuron morphology. J Neurobiol 60: 1-11.

Holmes MM, Putz O, Crews D, Wade J (2005). Normally occurring intersexuality
and testosterone induced plasticity in the copulatory system of adult leopard
geckos. Horrn Behav 47: 439-445.

Holmes MM, Wade J (2005). Sexual differentiation of the copulatory
neuromuscular system in green anoles (Anolis carolinensis): normal ontogeny
and manipulation of steroid hormones. J ComLNeurol: in press.

Holmes MM, Wade J. Testosterone regulates androgen receptor
immunoreactivity in the copulatory, but not courtship, neuromuscular system in
adult male green anoles. J Neuroendo: submitted.

 

Hull EM, Meisel RL, Sachs BD (2002). Male sexual behavior. In: Hormonesy
Brain and Behavior, Pfaff DW, Arnold AP, Etgen AM, Fahrbach SE, Rubin RT,
editors. New York: Academic Press, p. 3-137.

 

Hyde LL, Underwood H (1995). Daily melatonin infusions entrain the Iocomotor
activity of pinealectomized lizards. Physiol Behav 58: 943-951.

Hyde LL, Underwood H (2000). Effects of melatonin administration on the

circadian activity rhythm of the lizard Anolis carolinensis. Physiol Behav 71: 183-
1 92.

164

 

Jenssen TA, Nunez SC (1998). Spatial and breeding relationships of the lizard,
Anolis carolinensis: evidence of intrasexual selection. Behaviour 135: 981-1003.

Jenssen TA, Greenberg N, Hovde KA (1995). Behavioral proﬁle of free-ranging
male lizards, Anolis carolinensis, across breeding and post-breeding seasons.
Herpetol Monogg 9: 41-62.

Jenssen TA, Congdon JD, Fischer RU, Estes R, Kling D, Edmands S, Berna H
(1996). Behavioural, thermal, and metabolic characteristics of a wintering lizard
(Anolis carolinensis) from South Carolina. Func Ecol 10: 201-209.

Jenssen TA, OrreII KA, Lovem MB (2000). Sexual dimorphisms in aggressive

signal structure and use by a polygynous lizard, Anolis carolinensis. Copeia 1:
140-149.

Jenssen TA, Lovem MB, Congdon JD (2001). Field-testing the protandry-based
mating system for the lizard, Anolis carolinensis: does the model organism have
the right model? Behav Ecol Sociobiol 50: 162-172.

Jordan CL (1997). Androgen receptor (AR) immunoreactivity in rat pudendal
motoneurons: implications for accessory proteins. Horm Behav 32: 1-10.

 

Jordan CL, Breedlove SM, Arnold AP (1991). Ontogeny of steroid accumulation
in spinal lumbar motoneurons of the rat: Implications for androgen’s site of
action during synapse elimination. J Corpp Neurol 313: 441-448.

Jordan CL, Padgett G, Hershey J, Prins G, Arnold A (1997). Ontogeny of
androgen receptor immunoreactivity in lumbar motoneurons and in the sexually
dimorphic levator ani muscle of male rats. J Comp Neurol 379: 88-98.

Kay JN, Hannigan P, Kelley DB (1999). Trophic effects of androgen:
development and hormonal regulation of neuron number in a sexually dimorphic
vocal motor nucleus. J Neurobiol 40: 375-385.

Keizer K, Kuypers HG, Huisman AM, Dann O (1983). Diamidino yellow
dihydrochloride (DY.2HCI); a new ﬂuorescent retrograde neuronal tracer, which
migrates only very slowly out of the cell. Exp Brain Res 51: 179-191.

Kelley DB (1986). Neuroeffectors for vocalization in Xenopus Iaevis: hormonal
regulation of sexual dimorphism. J Neurobiol 17: 231-248.

Kelley DB, Pfaff DW (1976). Hormone effects on male sex behavior in adult
South African clawed frogs, Xenopus laevis. Horm Behav 7: 159-182.

 

165

 

Kelley DB, Dennison J (1990). The vocal motor neurons of Xenopus Iaevis:
development of sex differences in axon number. J Neurobiol 21: 869-882.

Kelley DB, Tobias ML (1999). The vocal repertoire of Xenopus Iaevis. In: II_1_e_
Design of Animal Communication, Hauser M, Konishi M, editors. MIT Press:
Cambridge, p. 9-35.

Kelley DB, Fenstemaker S, Hannigan P, Shih S (1988). Sex differences in the
motor nucleus of cranial nerve lX-X in Xenopus laevis: a quantitative Golgi study.
J Neurobiol 19: 413-429.

Kelly KK, Goldman BD, Zuckerl (1994). Gonadal growth and hormone P
concentrations in photoregressed Siberian hamsters: pinealectomy versus
photostimulation. Biol Reprod 51 : 1 046-1050.

Knapp R, Marchaterre MA, Bass AH (1999). Eariy development of the motor and '
premotor circuitry of a sexually dimorphic vocal pathway in a teleost ﬁsh. J
Neurobiol 38: 475-490.

 

Kuypers HG, Bentivoglio M, Catsman-Berrevoets CE, Bharos AT (1980). Double
retrograde neuronal labeling through divergent axon collaterals, using two
ﬂuorescent tracers with the same excitation wavelength which label different
features of the cell. Exp Brain Res 40: 383-392.

Lee JS, Bass AH (2005). Differential effects of 11-ketotestosterone on dimorphic

traits in a teleost with alternative male reproductive morphs. Horm Behav 47:
523-531.

Leslie M, Forger NG, Breedlove SM (1991). Does androgen affect axonal
transport of cholera toxin HRP in spinal motoneurons? Neurosci Lett 126: 199-
202.

Ligon JD (1968). Sexual differences in foraging behavior in two species of
Dendrocopus woodpeckers. A_u_lg 85: 203-215.

Lovem MB, Jenssen TA (2001). The effects of context, sex, and body size on
staged social interactions in juvenile male and female green anoles (Anolis
carolinensis). Behaviour 138: 1 1 17-1 135.

Lovem MB, Wade J (2001). Maternal plasma and egg yolk testosterone
concentrations during embryonic development in green anoles (Anolis
carolinensis). Gen Comp Endo 124: 226-235.

Lovem MB, McNabb FMA, Jenssen TA (2001). Developmental effects of

testosterone on behavior in male and female green anoles (Anolis carolinensis).
Homﬁehav 39: 131-143.

166

Lovem MB, Holmes MM, Wade J (20043). The green anole (Anolis carolinensis):
a reptilian model for laboratory studies of reproductive morphology and behavior.
ILAR J 45: 54-65.

Lovem MB, Holmes MM, Fuller CO, Wade J (2004b). Effects of testosterone on
the development of neuromuscular systems and their target tissues involved in

courtship and copulation in green anoles (Anolis carolinensis). Horrn Behav 45:
295-305.

Lu S, Simon NG, Wang Y, Hu S (1999). Neural androgen receptor regulation:
effects of androgen and antiandrogen. J Neurobiol 41: 505-512.

Mason P, Adkins EK (1976). Hormones and social behavior in the lizard, Anolis
carolinensis. Horrn Behav 7: 75-86.

McKenna KE, Nadelhaft I (1986). The organization of the pudendal nerve in the
male and female rat. J Comp Neurol 248: 532-549.

 

Meek LR, Romeo RD, Novak CM, Sisk CL (1997). Actions of testosterone in
prepubertal and postpubertal male hamsters: dissociation of effects on

reproductive behavior and brain androgen receptor immunoreactivity. Horm
Behav 31: 75-88.

Mills A, Zakon HH, Marchaterre MA, Bass AH (1992). Electric organ morphology
of Stemopygus macrurus, a wave-type, weakly electric ﬁsh with a sexually
dimorphic EOD. J Neurobiol 23: 920-932.

Monks DA, O’Bryant EL, Jordan CL (2004). Androgen receptor immunoreactivity

in skeletal muscle: enrichment at the neuromuscular junction. J Comp Neurol
473: 59-72.

Morgantaler A, Crews D (1978). Role of the anterior hypothalamic-preoptic area
in the regulation of reproductive behavior in the lizard, Anolis carolinensis:
implantation studies. Horrn Behav 11: 61-73.

Nakano H, Masuda K, Sasaki SY, Katsuta S (1997). Oxidative enzyme activity
and soma size in motoneurons innervating the rat slow-tvvitch and fast-twitch
muscles after chronic activity. Brain Res Bull 43: 149-154.

 

Nakano H, Katsuta S (2000). Non-weight bearing condition arrests the
morphological and metabolic changes of rat soleus motoneurons during
postnatal growth. Neurosci Lett 290: 145-148.

Nelson RJ (2000). An Introduction to Behavioral Endocrinology. Sinauer
Associates: Massachusetts.

167

 

 

Nordeen EJ, Nordeen KW, Sengelaub D, Arnold AP (1985). Androgens prevent
normally occurring cell death in a sexually dimorphic spinal nucleus. Science
229: 671-673.

Nottebohm F, Arnold AP (1976). Sexual dimorphism in vocal control areas of the
songbird brain. Science 194: 21 1-213.

Nottebohm F, Stokes TM, Leonard CM (1976). Central control of song in the
canary, Sen'nus canaries. J Comp Neurol 165: 457-486.

Nunez SC, Jenssen TA, Ersland K (1997). Female activity proﬁle of a polygynous
lizard (Anolis carolinensis): evidence of intersexual asymmetry. Behaviour 134:
205-223.

O'Bryant EL, Wade J (1999). Sexual dimorphisms in a neuromuscular system
regulating courtship in the green anole lizard: Effects of season and androgen
treatment. J Neurobiol 40: 202-213.

O'Bryant EL, Wade J (2000). Arborization of dewlap motoneurons in the green
anole lizard (Anolis carolinensis) is not sexually dimorphic. Neurosci Letters 281:
1 15-1 18.

 

O'Bryant EL, Wade J (2001). Development of a sexually dimorphic
neuromuscular system involved in green anole courtship behavior. Brain Behav
Evol 58: 362-369.

 

O'Bryant EL, Wade J (2002a). Seasonal and sexual dimorphisms in the green
anole forebrain. Horm Behav 41: 384-395.

O'Bryant EL, Wade J (2002b). Sexual dimorphism in neuromuscularjunction size
on a muscle used in courtship by green anole lizards. J Neurobiol 50: 24-30.

 

Ohmori Y, Watanabe T (1989). Location of motoneurons innervating the
musculus sphincter cloacae in the domestic fowl. Neurosci Lett 101: 1-5.

 

Park KH, Meitzen J, Moore IT, Brenowitz EA, Perkel DJ. Seasonal-like plasticity
of spontaneous ﬁring rate in a songbird pre-motor nucleus. J Neurobiol: in press.

Pasmanik M, Callard GV (1988). A high abundance of androgen receptor in

goldﬁsh brain: characteristics and seasonal changes. Endocrinology 123: 1162-
1 171 .

Perez J, Kelley D (1996). Trophic effects of androgen: receptor expression and
the survival of laryngeal motor neurons after axotomy. J Neurosci 16: 6625-6633.

168

Phoenix CH, Goy RW, Gerall AA, Young WC (1959). Organizing action of
prenatally administered testosterone propionate on the tissues mediating mating
behavior in the female guinea pig. Endocrinology 65: 369-382.

Pough FH, Andrews RM, Cadle JE, Crump ML, Savitzky AH, Wells KD (2001).
Herpetolpgy, 2"“I edition. Prentice Hall: New Jersey.

Rand MN, Breedlove SM (1987). Ontogeny of functional innervation of
bulbocavemosus muscles in male and female rats. Brain Res 430: 150-152.

Rand MN, Breedlove SM (1992). Androgen locally regulates rat bulbocavemosus
and levator ani size. J Neurobiol 23: 17-30.

Rand MN, Breedlove SM (1995). Androgen alters the dendritic arbors of SNB
motoneurons by acting upon their target muscles. J Neurosci 15: 4408-4416.

Raouf S, Van Roo B, Seglaub F (2000). Adult plasticity in hormone-sensitive
motoneuron morphology: methodological/behavioral confounds. Horm Behav
38: 210-221.

 

Raynaud A, Pieau C (1985). Embryonic development of the genital system. In:
Biology of the Reptilia, Gans C editor. John Wiley 8: Sons: USA.

Rhen T, Crews D (2001). Distribution of androgen and estrogen receptor mRNA
in the brain and reproductive tissues of the leopard gecko, Eublephan's
macularius. J Comp Neurol 437: 385-397.

Rhen T, Ross J, Crews D (1999). Effects of testosterone on sexual behavior and
morphology in adult female leopard geckos, Eublepharis macularius. Horm
Behav 36: 119-128.

Rhen T, Sakata JT, Zeller M, Crews D (2000). Sex steroid levels across the
reproductive cycle of female leopard geckos, Eublepharr's macularius. from
different incubation temperatures. Gen Corpp Endo 118: 322-331.

Rhen T, Sakata JT, Woolley S, Porter R, Crews D (2003). Changes in androgen
receptor mRNA expression in the forebrain and oviduct during the reproductive

cycle of female leopard geckos, Eublepharis macularius. Gen Comp Endo 132:
133-141.

 

Robertson JC, Watson JT, Kelley DB (1994). Androgen directs sexual

differentiation of laryngeal innervation in developing Xenopus Iaevis. J Neurobiol
25: 1625-1636.

 

169

Romeo RD, Richardson HN, Sisk CL (2002). Puberty and the maturation of the
male brain and sexual behavior: recasting a behavioral potential. Neurosci
Biobehav Rev 26: 381-391.

Rosen GJ, Wade J (2000). The role of 5-alpha-reductase activity in sexual
behaviors of the green anole lizard. PhLSIOI Behav 69: 487-498.

 

Rosen GJ, O'Bryant EL, Matthews J, Zacharewski T, Wade J (2002). Distribution
of androgen receptor mRNA expression and immunoreactivity in the brain of the
green anole lizard. J Neuroendo 14: 19-28.

Ruby DE (1984). Male breeding success and differential access to females in
Anolis carolinensis. Hegpetolpgica 40: 272-280.

Ruiz CC, Wade J (2002). Sexual dimorphisms in a copulatory neuromuscular
system in the green anole lizard. J COI‘I'IJ Netﬂl 443: 289-297.

Sassoon D, Kelley DB (1986). The sexually dimorphic larynx of Xenopus laevis:
development and androgen regulation. Am J Anat 177: 457-472.

Sassoon D, Segil N, Kelley D (1986). Androgen-induced myogenesis and
chondrogenesis in the larynx of Xenopus Iaevis. Dev Biol 113: 135-140.

 

Sassoon DA, Gray GE, Kelley DB (1987). Androgen regulation of muscle ﬁber
type in the sexually dimorphic larynx of Xenopus Iaevis. J Neurosci 7: 3198-3206.

Seiwert CM, Adkins-Regan E (1998). The foam production system of the male
Japanese quail: Characterization of structure and function. Brain Behav Evol 52:
61 -80.

 

Selander RK (1966). Sexual dimorphism and differential niche utilization in birds.
Condor 68: 113-151.

Schroder HD (1980). Organization of the motoneurons innervating the pelvic
muscles of the male rat. J Comp Neurol 192: 567-587.

Siegford JM, Ulibarri C (2004). Increase in motoneurons in the spinal nucleus of
the bulbocavemosus of prepubertally castrated male Mongolian gerbils following
delayed treatment with testosterone. J Comp Neurol 472: 485-495.

Simpson HB, Tobias ML, Kelley DB (1986). Origin and identiﬁcation of ﬁbers in

the cranial nerve IX-X complex of Xenopus laevis: Lucifer yellow backﬁlls in vitro.
J Cm Neurol 244: 430-444.

170

Sinha-Hikim l, Taylor WE, Gonzalez-Cadavid NF, Zheng W, Bhasin S (2004).
Androgen receptor in human skeletal muscle and cultured muscle satellite cells:
up-regulation by androgen treatment. J Clin Endocrinol Metab 89: 5245-5255.

Sisneros JA, Bass AH (2003). Seasonal plasticity of peripheral auditory
frequency sensitivity. J Neurosci 23: 1049-1058.

Soma KK, Hartman VN, Wingﬁeld JC, Brenowitz EA (1999). Seasonal changes
in androgen receptor immunoreactivity in the song nucleus HVc of a wild bird. J
Comp Neurol 409: 224-236.

Souccar C, Lapa AJ, Ribeiro do Valle J (1982). Inﬂuence of castration on the
electrical excitability and contraction properties of the rat levator ani muscle. Eyp
Neurol 75: 576-588.

Sperry TS, Thomas P (1999). Characterization of two nuclear androgen
receptors in Atlantic croaker: comparison of their biochemical properties and
binding speciﬁcities. Endocrinology 104: 1602-161 1 .

Springer ML, Wade J (1997). The effects of testicular tissue and prehatching
inhibition of estrogen synthesis on the development of courtship and copulatory
behavior in zebra ﬁnches. Horm Behav 32: 46-59.

 

Tobias ML, Kelley DB (1995). Sexual differentiation and hormonal regulation of
the laryngeal synapse in Xenopus laevis. J Neurobiol 28: 515-526.

 

Tobias ML, Marin ML, Kelley DB (1991). Temporal constraints on androgen
directed laryngeal masculinization in Xenopus Iaevis. Dev Biol 147: 260-270.

Tobias ML, Tomasson J, Kelley DB (1998). Attaining and maintaining strong
vocal synapses in female Xenopus Iaevis. J Neurobiol 37: 441-448.

Tokarz RR (1988). Copulatory behavior of the lizard Anolis sagrei: alternation of
hemipenis use. Anim Behav 36: 1518-1524.

 

Tokarz RR, Slowinski JB (1990). Altemation of hemipenis use as a behavioural
means of increasing sperm transfer in the lizard Anolis sagrei. Anim Behav 40:
374-379.

 

Tokarz RR, Kirkpatrick SJ (1991). Copulation frequency and pattern of hemipenis
use in males of the lizard Anolis sagrei in a semi-natural enclosure. Anim Behav
41: 1039-1044.

 

Tokarz RR, McMann S, Seitz L, John-Alder H (1998). Plasma corticosterone and
testosterone levels during the annual reproductive cycle of male brown anoles
(Anolis sagrer). Physiol Zool 71: 139-146.

171

 

Tokarz RR, McMann S, Smith LC, John-Alder H (2002). Effects of testosterone
treatment and season on the frequency of dewlap extensions during male-male
interactions in the lizard Anolis sagrei. Horm Behav 41: 70-79.

 

Tousignant A, Crews D (1994). Effect of exogenous estradiol applied at different
embryonic stages on sex determination, growth, and mortality in the leopard
gecko (Eublepharis macularius). J Exp Zool 268: 17-21.

Veney SL, Wade J (2004). Steroid receptors in the adult zebra ﬁnch syrinx: a sex
difference in androgen receptor mRNA, minimal expression of estrogen receptor
a and aromatase. Gen CompEndo 136: 192-199.

Veney SL, Wade J (2005). Post-hatching syrinx development in the zebra ﬁnch:
an analysis of androgen receptor, aromatase, estrogen receptor alpha and
estrogen receptor beta mRNAs. J Comp Physiol A Neuroethol Sens Neural
Behav Physiol 191 : 97-104.

Viets BE, Ewert MA, Talent LG, Nelson CE (1994). Sex-deterrnining mechanisms
in squamate reptiles. J Exp Zool 270: 45-56.

Wade J (1998). Sexual dimorphisms in the brainstem of the green anole lizard.
Brain Behav Evol 52:46-54.

Wade J (1999). Sexual dimorphisms in avian and reptilian courtship: Two
systems that do not play by mammalian rules. Brain Behav Evol 54: 15-27.

Wade J (2001). Zebra ﬁnch sexual differentiation: the aromatization hypothesis
revisited. Microsc Res Tech 54: 354-363.

Wade J (2005). Current research on the behavioral neuroendocrinology of
reptiles. Horm Behav: in press.

Wade J, Arnold AP (1996). Functional testicular tissue does not masculinize
development of the zebra ﬁnch song system. Proc Natl Acad Sci 93:5264-5268.

 

Wade J, Buhlman L (2000) Lateralization and effects of adult androgen in a
sexually dimorphic neuromuscular system controlling song in zebra ﬁnches. J
Comp Neurol 426:154-164.

Wade J, Arnold AP (2004). Sexual differentiation of the zebra ﬁnch song system.
Ann NY Acad Sci 1016: 1-20.

Wade J, Gong A, Arnold AP (1997). Effects of embryonic estrogen on

differentiation of the gonads and secondary sexual characteristics of male zebra
ﬁnches. J Exp Zool 278: 405-411. .

172

 

Wade J, Swender DA, McElhinny TL (1999). Sexual differentiation of the zebra
ﬁnch song system parallels genetic, not gonadal, sex. Horm Behav 36: 141-152.

 

Wade J, Buhlman L, Swender D (2002). Post-hatching hormonal modulation of a
sexually dimorphic neuromuscular system controlling song in zebra ﬁnches.
Brain Res 929: 191-201.

Watson JT, Kelley DB (1992). Testicular masculinization of vocal behavior in
juvenile female Xenopus Iaevis reveals sensitive periods for song duration, rate,
and frequency spectra. J Comp Physiol [A] 171: 343-350.

Watson JT, Robertson J, Sachdev U, Kelley DB (1993). Laryngeal muscle and
motor neuron plasticity in Xenopus laevis: testicular masculinization of a
developing neuromuscular system. J Neurobiol 24: 1615-1625.

 

Wetzel DM, Kelley DB (1983). Androgen and gonadotropin effects on male mate
calls in South African clawed frogs, Xenopus Iaevis. Horm Behav 17: 388-404.

 

Wetzel DM, Haerter UL, Kelley DB (1985) A proposed neural pathway for
vocalization in South African clawed frogs, Xenopus Iaevis. J Comp Physiol [A]
1 57: 749-761 .

 

Wibbels T, Crews D (1992). Speciﬁcity of steroid hormone-induced sex
determination in a turtle. J Endocrinol 133: 121-129.

 

Wibbels T, Crews D (1995). Steroid-induced sex determination at incubation
temperatures producing mixed sex ratios in a turtle with TSD. Gen Comp Endo
100: 53-60.

 

Winkler SM, Wade J (1998). Aromatase activity and regulation of sexual
behaviors in the green anole lizard. Physiol Behav 64: 723-731.

Zakon HH, Dunlap KD (1999). Sex steroids and communication signals in electric
ﬁsh: a tale of two species. Brairﬁehav Evol 54: 61-69.

173