University of Southern California Dissertations and Theses

Molecular and genomic studies of sex determination in Drosophila

(USC Thesis Other)

Molecular and genomic studies of sex determination in Drosophila

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
MOLECULAR AND GENOMIC STUDIES OF
SEX DETERMINATION IN DROSOPHILA

by

Laura Elizabeth Sanders

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MOLECULAR BIOLOGY)

December, 2008

Copyright 2008 Laura Elizabeth Sanders

ii

Dedication

This thesis is dedicated to my family. My dad and mom, Richard and Andrea Sanders,
have supported me in all of my decisions, and encouraged me when I needed it the most.
Their kindness, intelligence, and unique outlooks on life have given me something to aim
for, and their love and support has made me into the person I am. I am lucky to have
them as role models.

My brothers Daniel and Elliot made sure that from an early age on, I was strong enough
to take life’s bruises, and no one could have done it quite like them.

And finally, this thesis is also dedicated to my grandfathers, Robert A. Sanders and
Walter J. Schwen, and especially, my grandmothers Sara Palmer Sanders and Sarah Jean
Anderson Schwen. Both of these women know the value of a good education, but more
importantly, the value of being a good person, and I am proud to be a descendent of these
strong, smart women.

iii

Acknowledgements

I would first like to thank my mentor, Michelle Arbeitman, for all of her guidance and
support. Her dedication and ambition has been an example for me; she has taught me a
great deal about molecular biology and more importantly, how to do thorough science. I
have learned a lot from her, and I thank her for the opportunity to work in her lab.

I’d like to acknowledge my committee, both past and present. Oscar Aparicio for his
support and always inviting me to his lab’s barbecues; John Tower for all of his insightful
comments on my projects and for reading very long drafts of papers and never
complaining, but instead, offering good advice; David McKemy and Amy Barrios for
listening to my qualifying exam, and Hanna Reisler for graciously serving on my
committee.

I’d like to thank my outstanding labmates, first and foremost, Matt Lebo, with whom I
collaborated extensively on work presented in Chapters 2 and 3. He patiently taught me
more than I thought I could learn about computational biology, and has become one of
my best friends while doing it. I’d also like to thank Tom Goldman for being someone
really fun to come down the line with; Justin Dalton for fixing my problems faster than I
could see them coming myself; Emma Peebles for her good Glaswegian cheer and loyal
offers of support with a 2x4 and a rusty nail; and JP Masly for dispensing all of his wise
iv

chestnuts of wisdom. I could not have asked for more supportive, creative and caring
people to be surrounded by, and I will truly miss their company.

I’d also like to recognize the wonderful undergraduates that have passed through the lab,
in particular, Brandon Ishique for help coding behavioral assays until we worried for his
sanity, Jason Portillo for help dissecting fru mutant brains, and Allison Nishitani for her
hard work on a microarray dataset that is not presented here.

I’d like to thank Orianna Bretschger, Rebecca Nugent, Meghna Yadav, and Alison
Kraigsley for dressing me up every so often and taking me out. The wonderful friends I
have made at USC filled my time here with fun.

And lastly, I’d like to thank Chris Viggiani, who encouraged me, listened to me, and
always saved me the best bite. His support and love have meant so much to me.
v

Table of Contents

Dedication ii
Acknowledgements iii

List of Tables vii

List of Figures ix

Abstract xi

Chapter 1: Introduction to somatic sex determination 1
in Drosophila
Chapter 2: Somatic, Germline and Sex Hierarchy-Regulated 19
Gene Expression During Drosophila Metamorphosis
2.1 Overview 19
2.2 Introduction 19
2.3 Results and Discussion 22
2.4 Materials and Methods 65

Chapter 3: Ecdysone Receptor mediates courtship behaviors 75
in Drosophila by acting in the fruitless circuit in an isoform-
specific manner.
3.1 Overview 75
3.2 Introduction 75
3.3 Results 79
3.4 Discussion 110
3.5 Materials and Methods 112

Chapter 4: Chapter 4. Doublesex establishes sexual dimorphism 117
in the Drosophila central nervous system in an
isoform-dependent manner by directing cell number
4.1 Overview 117
4.2 Introduction 118
4.3 Results 121
4.4 Discussion 159
4.5 Materials and Methods 164

Chapter 5: Conclusions and future directions 167

References 173
vi

Table of Contents, continued

Appendices 184
Appendix A: Supplemental Data for Chapter 2 184
Appendix B: Supplemental Data for Chapter 3 192
Appendix C: Supplemental Data for Chapter 4 194
vii

List of Tables

Table 2.1: Sex differentially expressed genes in the 28
somatic tissues during metamorphosis

Table 2.2: Microarray experimental design used to 52
identify genes regulated downstream of DSX

Table 2.3: Genes expressed downstream of DSX at 55
48 hour APF

Table 3.1: Microarray experimental design and 80
numbers of genes identified

Table 3.2: Genes identified by microarray experiment 81
in the whole body and dissected CNS experiments as
being differentially expressed as a result of fru P1
(q≤0.15, and Fold Change ≥2)

Table 3.3: Genes that are differentially expressed in 87
male and female CNS tissue

Table 3.4: Genes that are regulated by ecdysone that 92
are also regulated by fru P1 in whole body and
dissected CNS tissue

Table 4.1: DSX-expressing cells in the pC1 123
and pC2 regions of the brain

Table 4.2: DSX-expressing cells in the TN1 124
region of the VNC

Table 4.3: Genotypes interrupting apoptosis 139

Table A1: Genes with sex-differential expression 188
across metamorphosis, as identified by a two-way
ANOVA analysis with sex and time as the independent factors.

Table A2: Additional oligonucleotide sequences for 189
sex determination hierarchy array elements on the microarrays.

viii

List of Tables, continued

Table A3: Correlation among microarray replicates for 190
the time course analysis of gene expression in male and
female tud progeny, and male and female wild type
pupae at five time points during metamorphosis (0,
24, 48, 72, and 96 hour APF).

Table A4: Number of genes expressed at each 191
of the five metamorphosis stages examined (0,
24, 48, 72, and 96 hour APF), in male and female
tud progeny and male and female wild type pupae.

Table B1: List of genes regulated by ecdysone, 192
as identified by the literature.

Table C1: Number of DSX-expressing cells at 195
distinct timepoints across development in
wild type males and females, and mutants.

Table C2: DSX-GAL4 characterization results 196

ix

List of Figures

Figure 1.1: Male courtship behavior 3

Figure 1.2: Drosophil-a sex determination hierarchy 5
Figure 2.1: Clusters of genes with similar expression 26
profiles in somatic tissues during metamorphosis

Figure 2.2: The number of genes with somatic sex- 31
differential transcript levels differs across metamorphosis
Figure 2.3: Expression profiles of genes expressed 47
in the male and female germline during metamorphosis

Figure 2.4: Cluster of global expression profiles for 50
Drosophila transcripts across metamorphosis
Figure 2.5: Model of DSX modes of regulation 63
Figure 3.1: EcR encodes three isoforms which 96
have different spatial and temporal expression patterns

Figure 3.2: FRU
M
and EcRA or EcRB1 are co- 99
expressed in sub-sets of cells at distinct stages

Figure 3.3: Males with reduced EcR levels display 102
high levels of male-male courtship behavior

Figure 3.4: Preference assays and activity levels 106
of males carrying fru P1-GAL4/UAS-EcRA show
wild type levels

Figure 3.5: FRU
M
and BR are co-expressed in 109
the CNS at three distinct times during development

Figure 4.1: The DSX antibody is specific 121

Figure 4.2: ELAV, a protein expressed in the 122
nuclei of neurons, co-localizes with DSX

Figure 4.3: DSX expression is sexually dimorphic 126

Figure 4.4: The number of DSX-expressing cells 132
depends on the dose and isoform of DSX present

x

List of Figures, continued

Figure 4.5: DSX-expressing cells in the TN1 137
cluster in females undergo cell death during
the first day of pupal development

Figure 4.6:DSX-expressing cells co-localize 149
with BrdU in males, but not in females

Figure 4.7: fru

P1-expressing cells and 154
DSX co-localize in the CNS

Figure 4.8: DSX and fru P1 expressing cells 155
co-localize at 48 hours APF, and in additional
regions in the adult CNS

Figure 4.9: DSX expression in fru P1 null 157
animals and FRU
M
expression in dsx null animals

Figure A1: Temporal expression levels of 185
highly correlated (≥0.8) gene clusters

Figure C1: DSX expression in imaginal discs 196
and CNS of third instar larvae

Figure C2: DSX is expressed in the male testis, 197
but not the female ovary

Figure C3: DSX-expressing cells in ix mutant 197
females and wild type females

Figure C4: BrdU incorporation in the CNS of 198
males and females from 8-12 hours after puparium
formation

Figure C5: DSX-expressing cells are in close 199
proximity to fru P1-expressing cells’ synapses

Figure C6: DSX-GAL4 expression pattern 202
for five fly strains

xi

Abstract

In the fruit fly Drosophila melanogaster, males perform an innate and stereotyped
courtship ritual to successfully mate with females. Male courtship behavior and the
development of sex-specific morphologies are controlled by an alternative pre-mRNA
splicing cascade called the sex determination hierarchy, which ultimately results in the
production of sex-specific transcription factors encoded by doublesex (dsx) and fruitless
(fru). DSX regulates most of the morphological differences between males and females.
The male-specific isoform of FRU (FRU
M
) functions in the central nervous system
(CNS) and is necessary for most aspects of male courtship behavior. Although the
regulatory interactions within the sex determination hierarchy are well characterized, the
activity and targets of DSX and FRU
M
, the terminal effectors of sex determination,
remain poorly understood. To explore how the genome is sex-specifically deployed, an
analysis of the transcriptional regulation of sex-specific development during
metamorphosis, the period in which the fly transitions from a larva to an adult, was
performed. Genes were identified that likely function at distinct times during
metamorphosis to regulate somatic sex-specific development, as well as genes that
function in germline development. To identify genes that are downstream of the sex
determination hierarchy, transcriptional profiles of mutants for sex determination genes
were analyzed at mid-metamorphosis, and downstream targets of FRU
M
and DSX were
identified. The set of genes regulated by fru included a preponderance of genes that are
known to be regulated by the steroid hormone ecdysone. This led to functional
characterization of the Ecdysone Receptor (EcR) specifically in the fru neural network.
xii

The data presented here demonstrate that one particular isoform of EcR, EcRA, is
required in the fru neural circuit for appropriate male courtship behavior. To examine the
role of DSX in the CNS, its sexually dimorphic expression pattern was characterized
across development. It is shown here that DSX specifies a sex-specific neural circuit in
an isoform-specific manner through distinct mechanisms in different regions of the CNS.
In addition to describing the transcriptional changes underlying sex-specific development
during metamorphosis, the data presented here detail how DSX and FRU may function to
pattern sex-specific neural circuits in the CNS.

1

Chapter 1: Introduction to somatic sex determination in Drosophila

The task of explaining a behavior in genetic terms is extremely difficult. Neurogenetics,
the field that attempts to associate genes and their products with behaviors, has advanced
greatly with the study of the fruit fly. The first report of a gene which influences
behavior came from Margaret Bastock in 1955, in which she ascribed sub-normal male
courtship behavior to a mutation in the yellow gene, originally identified because the flies
have a golden cuticle (Bastock, 1956). This work came at a time when the nature of the
gene was still under debate. A decade later, groundbreaking work from Seymour
Benzer’s laboratory showed that single genes in Drosophila could be mutated to result in
flies that could not sleep, could not remember, and could not court (reviewed in Vosshall,
2007). Several of the first studies on the genetic basis for courtship behavior used genetic
mosaics to map the regions of the brain necessary for courtship behavior (Hotta and
Benzer, 1976); (Hall, 1979). The bodies of these sexually mosaic flies were male and
part female, which allowed the assessment of behavior and the mapping of regions in the
central nervous system (CNS) that are necessary for specific courtship behaviors. The
information gleaned from these painstaking experiments laid a foundation for future work
on the neural circuits required for courtship behavior, and the genes that specify them.

Male Courtship Ritual
Drosophila males perform an innate, sequence-specific courtship ritual that, when
everything goes according to plan, culminates in the male copulating with the female
2

(reviewed in Greenspan and Ferveur, 2000). The first report describing the male
courtship ritual came in 1955 (Bastock and Manning, 1955). This robust behavior begins
when the male orients his body towards the female. The male fly views the female and
receives visual confirmation that she is an appropriate target. Next, gustatory information
is exchanged as he taps her abdomen with his leg, as males have gustatory receptors on
their front legs. Then, the male extends his wing and vibrates it, effectively singing a
species-specific love song, which has a characteristic series of pulses interrupted by
periods of humming—known as the ‘sine song.’ The production of a wing song is not
necessary for successful mating; wingless males will eventually copulate, but it takes
them substantially longer (Von Schilcher, 1976). After singing the wing song, the male
licks the female genitalia and receives gustatory cues. Finally, the male curls his
abdomen and attempts to copulate with the female. If successful, copulation lasts on
average twenty minutes; if unsuccessful, the male must start again from the beginning,
because the courtship ritual is a fixed-action pattern. When raised in complete isolation,
males will perform the courtship ritual perfectly, even if the female is his first encounter
with another organism. The easily identifiable components, the stereotyped sequence,
and the fact that it is innate make the courtship ritual a behavior well-suited to
examination; however, the greatest advantage of this system comes from its known
genetic basis (reviewed in Manoli et al., 2006).
3

Figure 1.1. A schematic of the Drosophila male courtship ritual (Adapted from Greenspan and
Ferveur, 2000)

Sex Determination and Dosage Compensation
The male courtship ritual in Drosophila is an excellent system in which to study how a
complex behavior is specified by genes, largely because it is known to be under almost
complete genetic control. A hierarchical cascade of alternatively spliced transcripts,
called the sex determination hierarchy (Figure 1.2), regulates somatic sexual
differentiation, including the sex-specific neural circuits required for male courtship
behavior, in Drosophila. The primary determinant of sex is the X chromosome to
autosome (X:A) ratio (Bridges, 1921). Upon assessment of this ratio, one of the first
responses is the activation of the sex lethal (sxl) locus in females and not in males (Cline,
1978). sxl is a master-regulator of both somatic and germline sex determination: its
activities are sufficient to direct an entire developmental program.

In addition to the sex determination hierarchy, SXL also controls dosage compensation,
the process by which males upreguate the transcription of their single X chromosome to
roughly equal transcriptional levels reached by females, who possess two X
chromosomes (reviewed in Cline and Meyer, 1996). Dosage compensation is achieved
4

through the Dosage Compensation Complex (DCC), which is comprised of one of two
non-coding RNAs, rox1 or rox2, and six proteins: MSL1, MSL2, MSL3, a
helicase/ATPase (MLE), a histone deacetylase (MOF) and histone kinase (JIL1)
(reviewed in Cline and Meyer, 1996). MSL2 is required for the formation of the protein
complex; ectopic expression of MSL2 in females is sufficient for assembly of
components of the DCC on the X chromosome, which leads to early lethality in females
(Kelley et al., 1995). In females, SXL inhibits translation of MSL2 mRNA by blocking
its interaction with the ribosome (Gebauer et al., 2003). In the absence of SXL in males,
the DCC is permitted to form. The complex modifies chromatin structure on the male X
chromosome; images of the X in males show a diffuse open structure, indicative of
upregulated transcription (Aronson et al., 1954). SXL has no known function in males;
when sxl is genetically ablated, males remain phenotypically normal (Salz et al., 1987).
However, in males that have a gain-of-function mutation in which SXL protein is
expressed, the phenotype is lethal, due to the lack of dosage compensation (Yanowitz et
al., 1999).

SXL regulates somatic sex determination, which is the main focus of the research
presented here. In females, the early transcripts resulting from sxl transcription both
auto-regulate the splicing of its own Sxl transcripts and regulate the splicing of its
downstream target, the transcript encoded by transformer (tra) (Belote et al., 1989;
Sosnowski et al., 1989). As is the case for SXL, TRA is present only in females (Baker
and Ridge, 1980); the transcript that is produced in males via a default splicing event
5

contains an early stop codon, which results in a putative non-functioning peptide (Boggs
et al., 1987). Interestingly, this non-functional peptide accounts for approximately half of
the transcripts encoded by tra in females because SXL is inefficient as a splice factor
(Boggs et al., 1987). TRA, along with the constitutively-expressed product encoded by
the transformer-2 (tra2) locus, regulate the splicing of the terminal transcription factors
encoded by the doublesex (dsx) and fruitless (fru) loci. The molecular mechanisms
through which the sex determination hierarchy functions have been extensively
characterized (reviewed in Baker et al., 1989); however, the transcriptional targets
thorough which DSX and FRU
M
ultimately function are largely uncharacterized.

Figure 1.2 The Drosophila melanogaster sex determination hierarchy.
The Drosophila sex hierarchy consists of a cascade of sex-specific alternatively spliced pre-mRNAs
culminating in the production of sex-specific transcription factors encoded by doublesex (dsx) and fruitless
P1 (fru P1). The primary determinate of sex is the X chromosome to autosomal chromosome (A) ratio. In
females (X:A =1), Sex Lethal (SXL) is produced and regulates the splicing of the pre-mRNA of
transformer (tra), resulting in the production of TRA. TRA acts in conjunction with constitutively
expressed TRA-2, and regulates the alternative splicing of the pre-mRNAs of dsx and fru P1, leading to the
production of the female-specific protein DSX
F
. In males (X:A = 0.5), SXL is not produced and dsx and
fru P1 pre-mRNAs undergo default splicing, resulting in the production of the male sex-specific proteins
DSX
M
and FRU
M
. In addition, in females SXL represses dosage compensation, the process by which
transcription of genes on the single X chromosome in males is up-regulated to roughly equal that of the two
X chromosomes in females.

6

fruitless
fru is a complex locus: it contains four promoters and three DNA binding domains that
span 140 kilobases (Ryner et al., 1996). The fru locus encodes transcripts that are both
sex-specifically and non-sex-specifically spliced (Ryner et al., 1996). The transcript
resultings from the usage of the most distal promoter, called the P1 promoter, is sex-
specifically spliced, which results in a protein that possesses a unique 101 amino acid
region at the N-terminus in males. Production of this transcript is achieved in males
through default splicing in the absence of TRA and TRA-2 (Heinrichs et al., 1998; Ryner
et al., 1996). Female-specific proteins have not been identified using
immunohistochemical techniques (Lee et al., 2000a). The trancripts resulting from TRA
and TRA-2 splicing in females contain a complete open reading fame, and therefore has
the potential to produce a functional protein, although some data suggests that the fru P1
transcript may undergo post-transcriptional repression in females (Usui-Aoki et al.,
2000). fru encodes a transcription factor that contains a domain with high homology to
BTB (Bric-A-Brac, Tram-Track, Broad) and Zinc Finger domains (Ryner et al., 1996;
Zollman et al., 1994). In flies, the BTB domain is known to mediate versatile protein-
protein interactions including cell cycle regulation, ubiquitination, and actin dynamics
(reviewed in Perez-Torrado et al., 2006). Several of the sex-nonspecific classes of fru
proteins are essential for viability; males and females that lack them die early during
development (Anand et al., 2001; Ryner et al., 1996), which may be partially a result of
the neuronal pathfinding defect that occurs during embryogenesis in these fru mutants
7

(Song et al., 2002) . The fru P1 transcript, which produces FRU
M
in males, is necessary
for male courtship behavior, but not for viability (Ryner et al., 1996).

Although it was not cloned until 1996, the fruitless locus was first implicated in courtship
behavior in 1963 by the identification of an inversion mutation (fru
1
) that caused
abnormal mating behavior and mapped to the fru region (Gill, 1963). Since then, much
work has been done on characterizing FRU
M
as a master regulator of courtship behavior.
FRU
M
is expressed in approximately 2% of neurons in the mid-pupal male CNS (Lee et
al., 2000a), as well as regions of the peripheral nervous system (Manoli et al., 2005;
Stockinger et al., 2005), and is necessary for appropriate male courtship behavior
(Villella et al., 1997). Hypomorphic allelic mutations show varying degrees of courtship
phenotypes, including high levels of male-male courtship, failure to properly perform the
wing song component of the male courtship ritual, and in some severe cases, a complete
absence of any courtship behavior (Ryner et al., 1996); (Villella et al., 2005). Several
lines of evidence point to FRU
M
’s role in directing the development of male-specific
structures. First, a small population (roughly twenty-five) of fru P1-expressing neurons
in females undergoes apoptosis, while the homologous neurons in males survive (Kimura
et al., 2005). Second, differentiation of seretonurgic neurons in the abdominal ganglion
of the ventral nerve cord (VNC) is FRU
M
-dependent (Billeter et al., 2006). Third, FRU
M

controls the development of the male-specific Muscle of Lawrence, which is thought to
aid the abdomen bending necessary for copulation (Gailey et al., 1991). And finally,
FRU
M
is required to form male-specific neuronal projections onto a glomerulus, a region
8

of the brain in which olfactory information is transmitted (Datta et al., 2008). Although
FRU
M
’s role in development is clearly established, more recent evidence suggests that it
also has a continuous role in the adult. Studies in which fru P1-expressing neurons were
synaptically silenced show that their activity is necessary for wild type levels of male
courtship (Stockinger et al., 2005), which suggests that fru P1-expressing neurons are
continually necessary for appropriate courtship behavior in adults, and that this is distinct
from the requirement for FRU
M
protein.

Recent studies which made it possible to visualize and manipulate fru P1-expressing
neurons have determined the location of the fru P1 transcript in males and females
(Manoli et al., 2005; Stockinger et al., 2005). Interestingly, neurons that express fru P1
are located in homologous positions in both males and females, and further, fru P1-
expressing neurons in the female may play a role in female behavior. When fru P1-
expressing neurons were synaptically silenced, females were largely unreceptive to male
courtship advances and showed egg-laying defects (Kvitsiani and Dickson, 2006). No
gross morphological differences between the fru P1 neural circuit in males and females
have been identified, although several small-scale differences exist (see above) (Manoli
et al., 2005; Stockinger et al., 2005). There are no obvious morphological differences
between males and females in the CNS, and further, genes that function downstream of
FRU
M
during development are largely unknown; thus, the question of how male-specific
neuronal circuitry, and in particular, the fru P1 neural circuitry, ultimately specifies
courtship behavior is open.
9

doublesex
While FRU
M
is thought to predominantly act in the nervous system to influence behavior,
dsx, the gene that encodes the other terminal transcription factors in the sex determination
hierarchy, is known to specify almost all aspects of somatic sexual differentiation outside
of the nervous system. DSX was first identified in 1965, and named for the phenotype of
flies mutant for DSX, which display characteristics of both males and females (Hildreth,
1965). Like other proteins in the sex determination hierarchy, dsx transcripts are
alternatively spliced to produce male- (DSX
M
) and female (DSX
F
)-specific proteins. The
sex-specific regions are found at the proteins’ carboxyl termini: DSX
M
has 152 unique
amino acids and DSX
F
has 30 unique amino acids (Burtis and Baker, 1989). The
common portion of DSX proteins contain a zinc-finger DNA binding domain—known as
the DM domain (Erdman and Burtis, 1993; Raymond et al., 1998)—which is conserved
among diverse taxa, including worms and mammals. In C. elegans, the male-specific
Mab-3 (DSX’s homolog) directs certain neuroblasts to develop into peripheral neurons
only found in males (Raymond et al., 1998). Surprisingly, a human protein that contains
this DM domain, DMRT1, was found to be expressed in human testis, and an inversion of
the chromosomal location of DMRT1 has been associated with human XY sex-reversal
(Raymond et al., 1998). Thus, of all the genes in the sex determination hierarchy, it
appears that DSX is the most functionally conserved, which has led to the hypothesis that
the terminal effectors in any system are under the most evolutionary constraint (reviewed
in Graham et al., 2003).

10

In Drosophila, DSX’s effect on somatic sex-specific development has been well-
characterized in many distinct tissues. DSX regulates sex-specific pigmentation patterns
(males have darker pigmentation on their abdomens than females), development of the
sex-combs, which are male-specific bristles on the forelegs, and the development of
external genitalia (Hildreth, 1965). Additionally, DSX has been shown to directly
regulate the expression of Yolk protein-1 (Yp1) and Yolk protein-2 (Yp2) (Burtis et al.,
1991) by DSX
F
upregulating transcription, and DSX
M
downregulating transcription.
Because DSX
M
and DSX
F
share a common DNA binding domain, they likely bind to the
same targets, but regulate them in different manners. As a result, Yp1 and Yp2 are
expressed only in females and function in egg development (Burtis et al., 1991). Flies in
which DSX
M
is the only isoform of DSX present are phenotypically almost
indistinguishable from wild type males, regardless of chromosomal sex (Duncan and
Kaufman, 1975). Interestingly, flies that possess both DSX
F
and DSX
M
display
intersexual characteristics of all of the above sex-specific morphologies. It has been
shown that DSX functions as a homodimer, and in the case of ‘intersexual’ flies—flies
that have both isoforms of DSX—it is thought that erroneous heterodimers comprised of
DSX
F
and DSX
M
can form and disrupt normal DSX function (Erdman et al., 1996). With
a few key exceptions (Goldman and Arbeitman, 2007) (Arbeitman et al., 2004), the
majority of genes that may function downstream of DSX have not been identified, which
leaves the actual effectors of these dramatically different sex-specific morphologies an
open question.
11

DSX
F
has an obligate co-factor encoded by intersex (ix); female ix mutants are
phenotypically similar to dsx null mutants (Garrett-Engele et al., 2002). IX is present in
both sexes, but only functions as a co-factor of DSX
F
(Chase and Baker, 1995; Garrett-
Engele et al., 2002). Recent work in which the DSX
F
protein was crystallized and
examined structurally demonstrated that IX binds directly to DSX
F
in the female-specific
C-terminus (Yang et al., 2008). As the case with IX demonstrates, the two isoforms of
DSX likely exert their different transcriptional effects through different sets of co-factors.

While the somatic morphological effects of DSX have been well-studied, DSX’s role in
the central nervous system has been more difficult to examine, largely because no overt
differences between male and female CNS tissue exist. Early behavioral studies showed
that male dsx mutants perform courtship behavior in a qualitatively sub-normal manner
(Villella and Hall, 1996), but these studies were confounded by the dramatic phenotypes
of these dsx null males, which appear completely intersexual (Villella and Hall, 1996).
Chromosomally female flies which express DSX
M
—pseudomales—do not perform any
male courtship behavior, suggesting that DSX is not sufficient to establish the neuronal
basis for courtship (Taylor et al., 1994). These conflicting results attest to the inherent
difficulties with behavioral assays performed with dsx mutants; as a result, more cellular
approaches have been undertaken in recent studies, with greater success. DSX was found
to be expressed in a sexually dimorphic pattern in the adult brain; males had many more
DSX-expressing cells than females in a region of the posterior brain (Lee et al., 2002),
which suggests that there may be a role for DSX in patterning a sexually dimorphic brain.
12

Intriguingly, a group of neuroblasts in the abdominal ganglion of the ventral nerve cord
were found to undergo more cell divisions in males, and these divisions were DSX-
dependent (Taylor and Truman, 1992). Additionally, FRU
M
and DSX
M
have been shown
to be co-expressed in subsets of neurons in the CNS of males (Billeter et al., 2006;
Rideout et al., 2007), both proteins function to direct the potential for one component of
the male fly’s wing song (Rideout et al., 2007), and animals transheterozygous for dsx
and fru P1 alleles show a reduction in male courtship behavior (Shirangi et al., 2006).
However, the mechanism by which DSX acts to establish sex-specific neural circuitries
necessary for courtship behavior remain unclear.

Metamorphosis of Drosophila
Drosophila progresses through four major stages of life: embryo, larva, pupa, and adult.
Metamorphosis is the time in which a fly transitions from a larva to an adult, which is the
duration of the pupal stage and lasts approximately 100 hours at 25 degrees Celsius (for a
summary of developmental changes, see Demerec, 1994). During this time, massive
structural changes occur, which is expected because the fly is changing from a larva that
crawls through the food into an adult that can fly, see, and mate. The morphological
events of metamorphosis have been well characterized: most larval tissues degrade, and
small groups of cells, called imaginal discs, which were formed during embryogenesis
and remained idle throughout the larval instars, begin to differentiate into almost all of
the adult fly’s structures, including eyes, brain, wings, legs, and genitalia.

13

The steroid hormone 20-hydroxyecdysone, called ecdysone, triggers the transitions
through the developmental stages by binding to its receptor, a heterodimer comprised of
the Ecdysone Receptor (EcR), and Ultraspiricle (USP) (Koelle et al., 1991; Yao et al.,
1993). Early studies of salivary glands by Ashburner, et al., which allowed visual
inspection of gene activation, showed that an ‘ecdysone hierarchy’ exists, in which “early
genes,” all of which were thought to be transcription factors, were transcriptionally
activated by the presence of ecdysone and its receptor. The products of the early genes
then started their own transcriptional cascades in which “late genes” would become
transcribed and early genes repressed (reviewed in Ashburner, 1972; Huet et al., 1995).
In line with this model, the EcR:USP heterodimer acts as a transcription factor, and has
been shown to both activate transcription upon ligand binding and regulate gene
expression through de-repression of transcription upon ligand binding (Brown et al.,
2006; Koelle et al., 1991). EcR mutations are typically lethal, depending on several
factors, including isoform and severity of the mutation (Bender et al., 1997). EcR
encodes three isoforms that are produced from both alternative promoters and alternative
splicing: EcRA, EcRB1, and EcRB2 (Talbot et al., 1993). The isoforms share DNA
binding and ligand-binding domains, but differ in their amino-terminal regions (Talbot et
al., 1993). Using isoform-specific antibodies (Talbot et al., 1993), it has been shown that
the EcR isoforms have distinct temporal and spatial expression patterns (Truman et al.,
1994), and isoform-specific mutations have shown that they also have distinct functions
(see below).

14

While the morphological events of metamorphosis, many of which are regulated by
ecdysone, have been studied in great detail, the genetic basis of these events is largely
unknown. Interestingly, a previous study has shown that the vast majority of genes are
re-used during discrete stages of development. The embryonic stage and the pupal stage
have similar transcriptional profiles, while the larval and adult stages are similar in
transcript usage (Arbeitman et al., 2002). This is intuitive because these stages are
segregated into immobile, developing animals (embryos and pupae), and mobile, feeding
animals that display behavior (larvae and adults). However, little is known about the sex-
differential deployment of the genome that directs the development of sexually dimorphic
adults.

Profound changes in the central nervous system occur during metamorphosis (reviewed
in Truman, 1990). The drastic changes in the CNS include the disassembly of the larval
nervous system and the construction (although not from scratch) of an adult nervous
system. One of the first events of this remodeling is the removal of larval elements,
which occurs within the first 12-14 hours after pupariation. Many neurons die, and those
that survive experience degeneration of their dendritic and axonal projections (Truman
and Bate, 1988). Interestingly, it has been shown that sets of neurons that contain
biogenic amines or neuropeptides, including populations of serotonergic neurons (Valles
and White, 1988), catecholamine-containing cells (Budnik and White, 1988), and
peptidergic cells (White et al., 1986) survive during metamorphosis. Larval motoneurons
also survive, but change their targets (Consoulas et al., 2002). These data suggest that the
15

bulk of surviving neurons form the neurotransmitter and motor systems of the fly. By 24
hours after pupariation, adult-typical projection patterns are seen, which consists of
extensive outgrowth of arbors. The brain itself shows a drastic increase in size, as does
the ventral nerve cord. Metamorphosis is the time during which the potential for sex-
specific behaviors are likely patterned in to the CNS. FRU
M
and DSX levels are high
during mid-metamorphosis, and given their known roles in patterning the CNS, likely act
during this window of time to specify neuronal development.

As is the case for morphological changes in the body of the fly during metamorphosis,
many of the developmental changes in the CNS are regulated by ecdysone (reviewed in
Riddiford et al., 2000). Particular isoforms of EcR have been shown to regulate different
aspects of CNS development. For example, the EcRA isoform is necessary for adult
arbor outgrowth (Truman et al., 1994) and programmed cell death of about 300 neurons
in the ventral nerve cord (Robinow et al., 1993). The EcRB isoforms regulate many
aspects of neuronal pruning of larval-specific arbors, mushroom body neurite outgrowth
(Lee et al., 2000b), and the programmed cell death of a distinct group of neurons in the
VNC (Choi et al., 2006). While the role of EcR in the development of the CNS has been
explored, much less is known about how the ecdysone hierarchy, and its subsequent brain
remodeling, integrates signals that allows for the formation of sex-specific neural circuits.

16

Overview of Chapters

In Chapter 2, transcriptional changes that occur during metamorphosis, with emphasis on
the genes that underlie somatic sex determination, will be described. Although
expression studies have been performed on all life stages of Drosophila, little is known
about the genes that underlie somatic sex determination. With this goal in mind, I along
with my collaborator Matt Lebo, assessed transcript abundance using microarrays at five
time points spanning metamorphosis, in both wild type males and females and males and
females that lack a germline. This work resulted in a description of somatic- and
germline-specific gene expression in both sexes, as well as functional classifications. To
further characterize the genes that may be involved in somatic sex determination, we
analyzed sex hierarchy mutants at the mid-point of metamorphosis, which provides
insight into the transcriptional effects of these key regulators. Targets (both direct and
indirect) of DSX were identified, and a novel mode of transcriptional regulation for DSX
is presented.

In the third chapter, a study that identifies putative targets of FRU
M
at mid-
metamorphosis will be described. Along with Matt Lebo, transcript abundance was
measured in fru P1 null males and compared to wild type males for both whole body
tissue and CNS tissue. These results showed that a significant proportion of the genes
regulated by FRU
M
are also involved in the ecdysone hierarchy, suggesting that the two
developmental programs may intersect. To investigate the possible interaction between
17

FRU
M
and EcR, immunohistochemical studies were performed in which it is shown that
FRU
M
and particular isoforms of EcR are expressed in the same neurons at distinct times
during metamorphosis, which suggests that they could be functioning together in the
same neurons to specify a male-specific neural circuit. To functionally assess the
consequences of removing EcR levels specifically in the fru P1 circuit, behavioral
courtship assays were conducted that show removing EcR in fru P1-expressing cells
results in a robust increase in male-male courtship behavior, which is a known phenotype
of fru P1 null mutants. Additional isoform-specific experiments show that one of the
EcR isoforms, EcRA, is likely sufficient for the male-male courtship phenotype.
Additional experiments here demonstrate that a gene known to directly respond to
ecdysone levels, broad (br), is also expressed in a sub-set of fru P1-expressing cells.
Thus, functional assays confirm that EcR and FRU
M
collaborate to specify the neural
substrates necessary for appropriate male courtship behavior.

In the fourth chapter, the mechanisms through which DSX acts to specify sexual
dimorphisms in the CNS will be described. An antibody was raised against a common
portion of DSX proteins to explore DSX’s expression pattern in the CNS during
metamorphosis and at the adult stage. The results show that DSX is present in a sexually
dimorphic pattern in the ventral nerve cord and in the posterior brain. The region of
sexually dimorphic DSX-expressing cells in the VNC is indistinguishable in males and
females from 0-12 hours after puparium formation. After 12 hours, however, this group
of neurons continues to proliferate in males, but undergoes cell death in females. This
18

cell death is DSX
F
-dependent, as DSX
F
-expressing cells persist when DSX
F
’s obligate
co-factor IX is genetically removed (see Figure C3). The sexual dimorphism in the
posterior brain is established earlier than the VNC region by a different mechanism, as
differences are present from the third larval instar stage onwards. Males have
significantly more DSX-expressing cells than females at all stages examined, and I show
that this is due in part to increased cell divisions in males from 8-12 hours after pupal
formation, as assed by BrdU incorporation. Thus, the results presented here demonstrate
how a sex-specific transcription factor functions through distinct mechanisms in different
portions of the CNS to pattern sex-specific neural circuitry.

19

Chapter 2. Somatic, Germline and Sex Hierarchy-Regulated Gene Expression
During Drosophila Metamorphosis

2.1. Overview
Drosophila melanogaster undergoes a complete metamorphosis, during which time the
larval male and female forms transition into sexually dimorphic, reproductive adult
forms. To understand this complex morphogenetic process at a molecular-genetic level,
whole genome microarray analyses were performed. Genes were identified that were
expressed during metamorphosis in both somatic and germline tissues of males and
females. Additionally, genes were identified that display sex-specific differences in
abundance in both of these tissues at discrete times during metamorphosis. In somatic
tissues, genes regulated downstream of the dsx branch of the sex determination hierarchy
were identified. For this set of genes, the modes of dsx-regulated gene expression were
determined. The data and analyses presented here provide a comprehensive assessment
of gene expression during metamorphosis in each sex, in both somatic and germline
tissues. These results provide a framework for further investigation into regulation of
sex-specific development. This work was performed in collaboration with Matt Lebo
(University of Southern California).

2.2. Introduction
In Drosophila melanogaster, metamorphosis is the period in development when the male
and female larval forms, which display little morphological sexual dimorphism, are
20

transformed into the reproductive male and female adult forms, which display large
differences between the sexes. This complete transformation is the result of several
discrete processes (reviewed in Riddiford, 1993): the degeneration of somatic larval
structures; the generation of adult structures from cells that are found within the larva
(imaginal discs, imaginal rings and histoblast nests); remodeling, death and neurogenesis
of the cells in the larval nervous system; and the development of the adult gonads through
interactions of both germline and the somatic tissues. Identifying the genes that underlie
and orchestrate this transformation, and understanding how sex-specific gene regulation
is integrated into the process, will provide insight into how large-scale changes in
morphology are directed at a molecular-genetic level.

Metamorphosis initiates at the end of the third larval instar by a pulse of the steroid
hormone ecdysone (reviewed in Riddiford, 1993). In response to this pulse of ecdysone,
the larva ceases movement and initiates pre-pupal development. Progression through the
subsequent pupal stages is mediated by an additional pre-pupal pulse of ecdysone that
triggers pupal formation, and finally by a large pulse of ecdysone that triggers adult
development (reviewed in Riddiford, 1993). While much is known about the
morphological changes that occur during these pupal stages (reviewed in Bodenstein,
1950), an outstanding question is what are the gene expression changes that occur
specifically in somatic and germline tissues that underlie this process and how sex-
specific regulation of gene expression is incorporated into the development decisions
being made.
21

Insight into Drosophila somatic sexual development is provided by the study of the sex
determination hierarchy (see Figure 1.1 in Chapter 1), a genetic regulatory hierarchy that
consists of a sexually dimorphic pre-mRNA splicing cascade that terminates with the
production of sex-specific transcription factors encoded by doublesex (dsx) and fruitless
(fru) (reviewed in Christiansen et al., 2002). dsx controls all morphological differences
between the sexes (reviewed in Christiansen et al., 2002), whereas fru is necessary for all
aspects of male courtship behaviors (reviewed in Manoli et al., 2006). While much is
known about adult, sex-specific phenotypes caused by mutations in dsx and fru (Hildreth,
1965; Ryner et al., 1996), how dsx and fru direct sexual development at the level of gene
expression during metamorphosis is still an open question.

Previous studies have examined gene expression during metamorphosis, but this study
differs from previous studies in two aspects. First, previous studies only examined
approximately a third of the Drosophila open reading frames (Arbeitman et al., 2002;
White et al., 1999), whereas in this study a microarray-based approach was used to
examine expression from all predicted genes. Second, this study examines male and
female transcriptional profiles separately in both wild type flies and flies that lack
germline tissues during metamorphosis, where as our previous study did not distinguish
between male and female gene expression profiles, in somatic and germline tissues
during pupal stages. Additionally, both the role of dsx in establishing somatic sex
differences in gene transcript levels and the modes of how dsx regulates gene expression
were determined.
22

2.3. Results and Discussion
Here, genes that underlie developmental changes that occur during metamorphosis in
both germline and somatic tissues were identified by assaying gene expression in male
and female wild type animals and animals that lack germline tissues. The sex differences
in transcript abundances that are downstream of the dsx branch of the somatic sex
determination hierarchy were also determined. All of the experiments were performed
using a two-color, glass-slide microarray platform spotted with ~15,100 oligonucleotides;
there is at least one array element that detects transcripts from each of the 13,820
predicted Drosophila genes, based on release 4.1 of the Drosophila genome (see
Materials and Methods for details on microarray production (Crosby et al., 2007)).

Gene expression was assayed at five time-points, in animals collected every 24 hours,
ranging from 0 hours after puparium formation (APF; 0 hour APF is the white pre-pupal
stage) to 96 hour APF (pharate adults). These analyses were performed using male and
females from both the Canton S (CS) wild type strain and from animals lacking germline
tissues. The germline-deficient animals are the progeny of female flies homozygous for
the maternal-effect, recessive mutation tudor (tud), hereafter referred to as tud progeny
(Boswell and Mahowald, 1985). These time point experimental samples were compared
to a common reference sample, consisting of RNA derived from male and female pupae
collected from all stages of metamorphosis; this approach facilitated comparisons across
all the experiments.

23

Additional microarray comparisons were performed using RNA derived from animals at
the 48 hour APF stage, to identify genes regulated downstream of the dsx branch of the
somatic sex determination hierarchy (Figure 1.1). Accordingly, microarray comparisons
using RNA from the following genotypes were performed: wild type males and females
from two different strains (CS and Berlin), male and female tud progeny, wild type
females and tra pseudomales, and wild type females and dsx
D
pseudomales. tra and dsx
D

pseudomales are chromosomally XX animals that produce DSX
M
, the male-specific
isoform of DSX, and as a result look phenotypically similar to wild type males(Duncan
and Kaufman, 1975; Sturtevant, 1945); the analysis of two distinct mutant genotypes that
produce DSX
M
in a chromosomally XX background facilitated the identification of genes
that are sex-differentially expressed downstream of DSX, and reduced the identification
of genes for which sex-differential expression is due to differences in sex-chromosome
composition, strain, or genes acting upstream of dsx and/or tra in the sex hierarchy.
Additionally, gene expression was compared between intersexual male and female flies
that do not produce DSX (dsx null; dsx
d+r3
/ dsx
m+r15
(Hildreth, 1965)) and wild type
males and females, respectively, to examine the modes of dsx-regulated gene expression.
24

Gene expression during metamorphosis
One of the primary goals of this study was to identify genes that direct the patterning and
morphogenesis of somatic tissues that display sexual dimorphisms. Therefore, we
performed hierarchical clustering, an algorithm that groups genes based on the similarity
of their expression profiles (Eisen et al., 1998), on the five time points of tud progeny
expression data. This generated hierarchical clusters — groups of genes with similar
temporal patterns of expression, in somatic tissues, during metamorphosis. Thirty-eight
clusters that each had a greater than 0.80 average Pearson’s correlation in gene
expression profiles and contained 15 genes or more were identified (Figure 2.1). When
wild type expression data was incorporated into the cluster analyses, gene expression
profiles for these genes from the wild type data were similar to those observed when
using the tud progeny data (Figure A1.1). Therefore, for most genes examined, gene
expression profiles in somatic tissues do not appear to be influenced by the presence of
the germline.

25

Figure 2.1. Clusters of genes with similar expression profiles in somatic tissues during
metamorphosis.
Clusters were generated using gene expression data from male and female tud progeny at five time points
during metamorphosis, indicated at top of clustergram. Expression profiles for each cluster were generated
by averaging the gene expression data at each time point for every gene in the cluster in both sexes.
Yellow and blue indicates high and low levels of expression compared to a common reference,
respectively. To the right of the clustergram, black and grey indicates a cluster in which gene expression is
at a peak (enriched) or at a trough (depleted), respectively, relative to the average expression value across
all time points. Clusters are at a peak or trough of expression if average expression was ⅔ of a standard
deviation above or below the mean expression value of the five time points, respectively. Functional
annotation represents functional categories that were over-represented among the genes in the cluster, as
determined by the program DAVID (P<0.05 (Dennis et al., 2003)).

26

Figure 2.1

27

To functionally analyze these clusters, the sets of genes from each cluster were examined
using the program DAVID, which identifies overrepresented functional groups among
the genes in each cluster, compared to all the genes represented on the array platform (see
Materials and Methods; DAVID is the Database for Annotation, Visualization and
Integrated Discovery (Dennis et al., 2003)). The gene sets from each cluster were
analyzed to determine if they are statistically overrepresented for additional features,
including chromosome location and if a gene’s exons are among the 2,500 most highly
conserved sequences across three Drosophila species and one mosquito (Anopheles
gambiae; see Materials and Methods (Siepel et al., 2005)).

Diverse temporal patterns of somatic gene expression were identified, but we were
unable to readily identify large clusters of genes (>15) that showed somatic sex-
differential expression in the hierarchical cluster. We reasoned that if a small fraction of
the genome is sex-differentially expressed in somatic tissues, hierarchical clustering
might not identify these genes, especially if gene expression values change more
substantially due to time than due to sex. Indeed, when t-test analyses were performed,
comparing male and female gene expression at each time point on the tud progeny
expression data, 414 genes with significant somatic sex-differential expression were
identified (q<0.05; Table 2.1, for all analyses with multiple testing, lists of P values were
converted to q values, an estimate of the false discovery rate (Storey, 2003)).

28

Table 2.1. Sex differentially expressed genes in the somatic tissues during metamorphosis. Significant
sex-differential expression determined using a t-test of means comparing expression data from male and
female tud progeny. All genes listed had at least a 2-fold difference in gene expression between the sexes.
*
Annotation symbol from FlyBase.
†
Chromosomal arm on which the gene is located.
‡
Fold change
difference between the sexes.
ξ
GO annotations for each gene as listed in Flybase.

29

Table 2.1, continued

However, when analyzing the tud progeny expression data using two-way analysis of
variance (ANOVA, sex and time as the independent factors), only 35 genes were
identified with significant somatic sex-differential expression (q<0.05; Table A1). If a
gene is not sex-differentially expressed throughout all the time-points assayed, sex-
differential expression will not be readily identified by ANOVA, but will be by t-test
comparisons, further confirming the idea that gene expression changes more substantially
due to time, than due to sex, during metamorphosis.

Overall, the two-way ANOVA analysis on the tud progeny expression data showed that
7,182 genes (~50% of genes represented on the array) had significant changes in somatic
30

gene expression between the five time-points examined (time as independent factor;
q<0.05). Similar two-way ANOVA analyses on the wild type animal expression data
showed that 8,654 genes (~63% of genes represented on our array) had significant
differential expression between the five time-points examined (time as the independent
factor; q<0.05), suggesting that some 1,400 additional genes vary in expression over time
due to the presence of the germline in wild type males and females. Our previous study
examining gene expression in wild type flies found a similar percentage of genes (~48%
of genes represented on arrays) with significant expression changes across
metamorphosis (1929/4028 genes, q < 0.05; see Materials and Methods for how q-values
were determined; (Arbeitman et al., 2002)). Here, 8,482 and 9,725 genes of the 13,820
genes examined had expression data in the tud and wild type experiments, respectively,
for the ANOVA analyses, demonstrating that approximately 70% of the predicted
Drosophila genes are expressed during metamorphosis (see Materials and Methods for
details).

Stage-specific, somatic sex-differential gene expression
For each of the five time points examined, a large range of number of genes were
identified that showed somatic, sex-differential expression: four time points (0, 48, 72
and 96 hour APF) all have similar numbers of genes with somatic, sex-biased transcript
levels (39, 41,36 and 19 genes, respectively; Figure 2.2).
31

The 24 hour APF time point, on the other hand, contains substantially more genes (291
genes) with somatic, sex-differential expression. The higher number of somatic, sex-
biased genes at 24 hour APF is not due to experimental biases, because at all five time
points, expression data for a similar number of genes was present in both sexes.
Additionally, when comparing microarray replicates within each experiment, high
correlation was seen among the microarray replicates at each time point (see Materials
and Methods). Overall, the number of genes with somatic, sex-differential expression
during pupal stages are similar to the number of somatic, sex-differentially expressed
genes as previously identified at adult stages (287 genes in (Arbeitman et al., 2004), 273
genes in (Parisi et al., 2003), and 152 genes in (Parisi et al., 2004)). In contrast, there are
thousands of genes with sex-differences in transcript levels in the male and female
Figure 2.2. The number of genes with somatic sex-differential transcript levels differs across
metamorphosis.
The abscissa indicates the five time points during metamorphosis examined (0, 24, 48, 72, and 96
hour APF). The ordinate indicates the number of sex-differentially expressed genes in the somatic
tissues, as identified by a t-test of the means on the gene expression data of male and female tud
progeny (q<0.05). Female- and male-biased genes are shown in grey and black, respectively.

32

germline tissues, at both pupal and adult stages (see below; (Arbeitman et al., 2002;
Parisi et al., 2003)).

Overall, 414 genes were identified with somatic sex-differential transcript levels. Most
of the genes (405 genes) displayed somatic, sex-differential expression at only one time
point, suggesting that most of these genes function at a specific time point and thus, are
likely to mediate discrete, sex-specific, developmental changes. Eight genes are sex-
differentially expressed at two time points, while roX1 is the only gene that was sex-
differentially expressed at three or more time points; roX1 was sex-differentially
expressed all time points examined. roX1 produces a non-coding RNA that is a
component of the dosage compensation macromolecular structure (Franke and Baker,
1999). Dosage compensation is the process in which genes on the single X chromosome
in males undergo increased transcription, which results in roughly equal amounts of
mRNA product produced by the two X chromosomes in females (reviewed in Bayer et
al., 1996). None of the transcripts that encode the MSL and MLE protein components of
the dosage compensation complex were detected by our array platform or showed sex-
differential expression in the tud progeny comparisons. However, somatic, sex-
differential expression is only expected for msl-2 transcripts and these differences are not
substantial (reviewed in Bayer et al., 1996). roX2 transcript, however, was significantly
higher in tud males than tud females only at the 24 hour APF stage, which was one of the
two time points for which there was enough data for statistical analyses. At the other
33

time points, expression was only detected in male, but not female, tud progeny data,
which prevented statistical analyses.

Gene expression during early metamorphosis (0 hour APF)
At the earliest stage examined, the white pre-pupal stage (wpp; 0 hour APF), the fly is
transitioning from a wandering larva, which scavenges for food, into an immobile pre-
pupa. While a pre-pupa, the animal initiates the major larval-to-adult transition. During
metamorphosis, this occurs in several discrete ways: 1) strictly larval tissues – including
the larval epidermis and musculature – are destroyed and replaced by corresponding adult
tissues (reviewed in DiBello et al., 1991), 2) imaginal discs and rings – physically distinct
primordia that are composed of cells that have been fated for their function during the
embryonic stage and proliferate during the larval stages – begin to undergo
morphogenesis to give rise to adult structures including eyes, antennae, wings, legs, and
genitalia (reviewed in Cohen, 1993), 3) histoblast nests – groups of cells fated early in
embryogenesis – proliferate in number and give rise to non-imaginal disc derived adult
epidermal structures (Madhavan and Schneiderman, 1977), and 4) the larval central
nervous system is remodeled through the destruction of some larval neurons, proliferation
of neuroblasts to generate new neurons, and remodeling of some larval neuronal
projections (reviewed in Truman, 1990). During the early stages of puparium formation,
strictly larval tissues, like the salivary gland, begin histolysis, the imaginal discs begin to
evert, imaginal histoblasts begin to divide, and the nervous system begins to be
remodeled.
34

First, genes were identified that show similar expression patterns in both male and female
somatic tissues, based on the hierarchical cluster analyses (Figure 2.1). Clusters were
identified that contained genes that either had a peak or trough of their transcript
abundance at the 0 hour APF time point (see Materials and Methods). Eleven clusters
that contain 1,201 genes in total had a peak of expression at the 0 hour APF stage.
Cluster 5 is the largest (608 genes) and was enriched for genes that encode proteins that
function in the proteosome, have cell death or peptidase activities and thus likely function
in the histolysis of larval tissues. In addition, proteases are required to initiate imaginal
disc eversion, which occurs at this stage (reviewed in DiBello et al., 1991). Cluster 36 is
the second largest (158 genes) and is enriched for genes whose products function in
metabolism and biosynthesis, including the production of glutamine, amino acids,
carboxylic acids, and ATP. Thus, at the onset of metamorphosis there is an abundance of
transcripts from genes that may function in the production of precursors necessary for the
synthesis of larger molecules and an increase in energy sources.

Nine clusters that contain 1,313 genes in total had a trough of transcript abundance at the
0 hour APF stage (Figure 2.1). Cluster 31 (762 genes), the largest cluster in a trough and
Cluster 38 (43 genes) are both enriched for genes whose products function in the
mitochondria, suggesting that at this stage new mitochondrial proteins are not being
synthesized. Cluster 21 (265 genes) is enriched for genes whose products function in
development, differentiation, and cell communication, suggesting that a large fraction of
35

genes that function in these patterning and developmental processes are at low transcript
levels immediately after pre-pupal formation.

At the 0 hour APF stage, 39 genes display somatic, sex-differential transcript abundance,
with 23 and 16 genes showing male- and female-biased levels, respectively. Aside from
the male-specific roX1 gene (described above), the transcript with the highest fold-
change between males and females was the female-biased Larval serum protein 1 alpha
(Lsp1a; fold change [FC] is 2.34). Lsp1a has previously been shown to be fat body
specific and highly expressed in females in third instar larvae (Powell et al., 1984;
Roberts and Evans-Roberts, 1979). In addition to Lsp1a, the female-biased genes
Transferrin 1, CG15369, CG14629, Chitinase-like, and Alcohol dehydrogenase all have
high transcripts levels in the larval fat body (Chintapalli et al., 2007) and/or encode for
proteins secreted into the larval hemolymph, the circulatory fluid of the fly (all have > 1.5
fold higher expression in females compared to males (Karlsson et al., 2004; Vierstraete et
al., 2003)). The larval fat body tissue, present in third instar larva, persists throughout
metamorphosis, and undergoes histolysis during early adult stages (Deutsch et al., 1989).
The fat body is involved in energy storage, nutrient sensing, degradation of metabolites,
and immune response (Deutsch et al., 1989; reviewed in Mirth and Riddiford, 2007), all
of which are important during late larval and early pupal stages.

Two of the genes with the greatest male-biased expression at the 0 hour APF stage are
putative G-protein coupled receptors (FC >1.6; GPCRs; CG12290 and metabotropic
36

GABA-B receptor subtype 3 (GABA-B-R3) (Brody and Cravchik, 2000)). CG12290 is an
orphan receptor with no associated ligand, while GABA-B-R3 is a member of an insect-
specific, sub-class of putative GABA receptors (Mezler et al., 2001). GPCRs bind
ligands, including hormones and neuropeptides, and the binding of ligand initiates signal
transduction cascades (reviewed in McCudden et al., 2005). The sex-differential
expression of GPCRs could lead to a sex-biased response to the presence of a hormone or
neuropeptide, thus affecting sex-differential development.

Gene expression during the early stages of metamorphosis (24 hour APF)
The 24 hour APF time point is between the pre-pupal pulse of ecdysone that peaks at 12
hour APF and triggers head eversion and the large pupal pulse of ecdysone that initiates
around 24 hour APF (reviewed in Riddiford, 1993). By 24 hr APF the majority of larval-
specific tissues are degraded and adult development is triggered (reviewed in Baehrecke,
1996).

The largest cluster identified in the entire hierarchical cluster, Cluster 13, contained 1,193
genes, which showed a peak in transcript abundance at 24 hour APF (Figure 2.1). This
set is over-represented with genes annotated as functioning in imaginal disc
morphogenesis (73 genes), neurogenesis (58 genes), programmed cell death (62 genes),
nervous system development (122 genes), and nucleic acid binding (331 genes),
demonstrating that at about 24 hour APF, many genes that drive morphogenesis and
patterning have reached a peak in their transcript abundance, marking this period as
37

critical for patterning and morphogenesis. On the other hand, Clusters 31 (762 genes)
and 38 (43 genes), which are enriched with genes whose products function in the
mitochondria, are still in a trough of expression at 24 hour APF.

In addition, the greatest number of somatic sex-biased genes was expressed at 24 hour
APF, with more male- than female-biased genes at this stage (185 and 106 genes,
respectively; Figure 2.2). One possibility for this high number of somatic, sex-biased
genes relative to the other stages examined is that somatic, sex-specific developmental
processes are occurring at this stage. Around this time, the structures derived from the
genital disc are fusing with gonad tissues to form the adult reproductive systems
(reviewed in Bodenstein, 1950); this might require different gene expression patterns in
somatic tissues of males and females. Alternatively, the observation that the cluster with
the most genes from our cluster analyses peaked in expression at the 24 hour APF stage
(Figure 2.1, Cluster 13) suggests that in general many genes reach peak transcript
abundance at this time. Interestingly, the 24 hour APF sex-biased, somatic set was over-
represented with genes that are among the most highly conserved across three
Drosophilids and one mosquito (P<0.05; 59 genes; see Materials and Methods and
(Siepel et al., 2005)). The evolutionary constraint on the nucleotide sequence for genes
in this set suggests that the sex-specific processes that these genes underlie may also be
evolutionarily conserved.

38

The set of male-biased genes (185 genes) expressed in somatic tissues was enriched for
genes that encode proteins with the insect cuticle domain (9/185 genes) – an insect-
specific domain by which some cuticular proteins bind chitin (Cuticular protein 100A,
97Eb, 12A, 30F, 64Ad, 97Ea, 51A, 49Ag, 92F (Rebers and Willis, 2001)). In addition,
one of the male-biased genes with the highest fold-change (149-fold higher in males than
females) is dusky, a gene encoding a predicted transmembrane protein containing a zona
pellucida domain, a domain that is found in extracellular matrix proteins (DiBartolomeis
et al., 2002). dusky is highly expressed in cuticle-secreting epithelia of the embryo and
pupal imaginal discs, suggesting it plays a critical role in cuticle formation. Between 18
and 36 hour APF the apolysis, or disintegration, of the pupal cuticle occurs and adult
cuticle begins to be deposited (reviewed in DiBello et al., 1991). The adult cuticle
contains several layers, composed of proteins, chitin, and cuticular hydrocarbons that are
cross-linked in distinct ways in different layers (reviewed in DiBello et al., 1991). While
much attention has focused on sex differences in cuticular hydrocarbon composition,
given the role of these hydrocarbons as pheromones (reviewed in Ferveur, 2005), sex
differences in cuticular proteins have not been examined as extensively. Our results
suggest that sex-differences in cuticular proteins are present as early as the 24 hour APF
stage. These differences may underlie differences in the morphogenesis of sex-specific
adult cuticular structures, like the external genitalia, or differences in the protein
composition of the cuticle that might contribute to differences in pheromonal profiles.

39

The set of female-biased genes (106 genes) expressed in somatic tissues at 24 hour APF
was enriched for genes whose products are involved with myosin activity, suggesting that
there is a sexual dimorphism in pupal muscle development at this stage. The female-
biased list includes myosin alkali light chain 1, myosin heavy chain, Troponin C at 73F,
and bent, a gene whose product has myosin light chain kinase activity. At this stage in
metamorphosis, most of the larval abdominal muscles have histolyzed, myocites first
appear in the legs as the leg muscles form, and the wing muscles begin to differentiate
(reviewed in Bodenstein, 1950). Although much remains to be determined about
sexually dimorphic muscle development, these expression data suggest a difference
present at 24 hour APF.

Gene expression halfway through metamorphosis (48 hour APF)
During the time between the 24 to 48 hour APF stages, the imaginal discs are still
undergoing morphogenesis, but are close to their final adult form. The wings, leg
muscles, abdominal bristles, abdominal muscles and internal genital ducts are all well
formed, while further development of the eyes, legs, wings, thorax, and abdomen is
occurring (reviewed in Bodenstein, 1950). Compared to the other four time points, the
48 hour APF time point has the fewest total number of genes with a peak or a trough in
transcript abundance (717 genes and 370 genes, respectively). At the 48 hour APF stage,
nine clusters have a peak of transcript abundance. The two largest clusters, Cluster 18
and Cluster 21, contain 141 and 265 genes, respectively, and show peak levels only at
this stage (Figure 2.1). Cluster 18 is enriched with genes whose products function in cell
40

organization, biogenesis, appendage morphogenesis and development, suggesting that
although the rudimentary adult structures are formed, there are still many structural
changes taking place. Consistent with this idea, Cluster 21 is enriched with genes whose
protein products function in cell communication, morphogenesis, and female gamete
development. Since the clusters were generated using the tud progeny expression data,
genes that function in the gamete development must have additional roles at this stage in
somatic tissues. Of the genes with a trough of transcript levels at this stage, more than
two-thirds are in Cluster 36 and Cluster 3 (158 and 102 genes, respectively). Both
clusters are over-represented with genes whose protein products function in protein,
amino acid, and carboxylic acid metabolism. It is known that oxygen consumption is low
around 48 hour APF (Poulson, 1935), and thus it is likely that metabolism in general is
low at this stage, consistent with what is seen in our expression data.

At the 48 hour APF stage, 41 genes showed somatic, sex-differential transcript levels,
with 35 and 6 genes showing female- and male-biased expression, respectively (Figure
2.2). Of the female-biased genes that display the largest difference in transcript levels,
two are annotated with mitochondrial ribosome function (CG31450 and CG5479),
suggesting that mitochondrial functions are increasing in females at this stage of
metamorphosis. Only one gene with known function, in the male-biased set, G protein
alpha49B, is suggested to have a role in a developmental process. More extensive
analyses of somatic sex-differential expression at 48 hour APF are presented below with
our analyses of dsx regulated expression.
41

Gene expression during late morphogenesis (72 hour APF)
During the later stages of metamorphosis, many of the tissues and structures developing
in the pupae are close to their final adult form (reviewed in Truman, 1990; Williams and
Carroll, 1993). At the 72 hour APF stage, 8 clusters containing 947 genes peak in
transcript abundance. Genes in Cluster 31 (762 genes, Figure 2.1), which are in a trough
at the 0 and 24 hour APF stages, quickly increase in transcript levels to ultimately peak at
72 hour APF. Cluster 31 is over-represented with genes that encode products that
function in the mitochondria (153 genes out of the 762 genes in this set). This is
consistent with what is known about respiration during insect metamorphosis; respiration
is at its lowest levels for most of metamorphosis and then sharply increases prior to
eclosion (Poulson, 1935). Cluster 5 (608 genes), which is enriched for genes that encode
products that function in proteolysis and that showed a peak of transcript levels early in
metamorphosis, is now in a trough. Although it is possible that these proteolytic proteins
may persist longer than their transcripts, the reduction in transcript abundance for many
genes that function in proteolysis suggests that the histolysis of larval-specific structures
is complete by 72 hour APF.

The 72 hour APF sex-biased, somatic set contained 36 genes, more of which are female-
biased genes (23 genes) than male-biased (13 genes). By 72 hour APF, the structure of
the nervous system largely resembles the adult tissue and even contains the same
neurotransmitter release sites (reviewed in Truman, 1990). Female-biased genes at the 72
hour APF time point include PAK-kinase, snap-25, and vegetable, all of which have been
42

shown to function in nervous system development or physiology (Albin and Davis, 2004;
Prokopenko et al., 2000; Rao et al., 2001). In addition, the female-biased set contained
rutabaga, which encodes an adenylate cyclase that functions in several adult behavioral
and physiological processes, including memory and female courtship behavior, and has a
developmental role in bristle formation (Duerr and Quinn, 1982; Kyriacou and Hall,
1984; Norga et al., 2003). Higher transcript levels of these genes in females at 72 hour
APF may help establish the potential for adult female-specific behavior or physiology.

Gene expression at the end of metamorphosis (96 hour APF)
By 96 hour APF, the pupa is within a few hours of eclosion, or emergence of the adult fly
(reviewed in Ashburner et al., 2005). There are additional morphological and gene
expression changes that occur shortly after eclosion (Arbeitman et al., 2002; reviewed in
Bodenstein, 1950) and genes that are expressed at the 96 hour APF stage likely underlie
these developmental processes. This includes expression of genes that are required for
adult physiological functions of structures like the eye, wings, legs and genitalia.
Overall, ten clusters containing 1,460 genes in total show a peak of transcript levels at the
96 hour APF stage. One large cluster, Cluster 32 (286 genes; Figure 2.1) contained genes
that showed a sharp rise in transcript levels at 96 hour APF, but no peaks early in
metamorphosis. This cluster is enriched for genes that function in the response to light
stimulus, visual perception, and rhabdomere function, all of which are critical for proper
vision and development of the adult eye. On the other hand, the largest cluster, which is
enriched for genes that function in developmental processes and was in a peak at 24 hr
43

APF, is now in a trough of transcript levels (Cluster 13, 1193 genes), consistent with the
idea that morphogenesis is largely complete by the pharate adult stage.

Nineteen genes showed somatic sex-differential levels at the 96 hour APF stage, 17 and 2
genes, displayed male- and female-biased levels, respectively. Due to the small number
of sex-biased genes identified at this stage, we reexamined our data to determine if genes
that display somatic sex-differential expression would be missed by our statistical
approach, if expression levels were below detection in the other sex. At the 96 hour APF
time point, 32 and 20 female- and male-biased genes, respectively, were identified that
had high transcript levels in one sex, but that had no transcript detected in the other sex,
and thus would not be identified by our statistical approach. Of the 20 male-biased
genes, 15 are expressed in the adult male accessory gland or ejaculatory bulb (Arbeitman
et al., 2002; Chintapalli et al., 2007; Dyanov and Dzitoeva, 1995; Holloway and Begun,
2004; Mueller et al., 2005; Saudan et al., 2002; Simmerl et al., 1995), suggesting that just
prior to eclosion, gene expression in somatic structures required for male reproductive
physiology initiates. In addition, two of the remaining five male-biased genes encode
odorant binding proteins; these proteins may be important for sex-specific recognition of
pheromones in adults. In contrast, more than half of the female-genes (17/32 genes) have
no known function. This set also includes two genes that encode for proteases, a class of
genes for which we find many members displaying sex-biased expression during
metamorphosis. In addition, there is one female-biased gene that encodes for an odorant
binding protein. Given our observations from the 96 hour APF data, we reexamined the
44

data from the other four time points to determine if we missed genes with somatic, sex-
biased expression because their transcript was present in only one sex, at a particular time
point, and found small numbers of genes.

Sex-biased gene expression in germline tissues during metamorphosis
Next, genes expressed during metamorphosis in either male or female germline tissues
were determined by identifying genes with sex-differential levels of transcript between
wild type male and females (ANOVA, sex as independent factor, q<0.05), and for which
transcripts are also significantly more abundant in wild type male and females as
compared to tud progeny male and females, respectively (ANOVA, genotype as
independent factor, q<0.05). To avoid false negatives, genes were included for which
there is no expression data in the tud experiments, as it is expected that tud progeny lack
germline tissues and thus would lack germline gene expression. Gene sets of 1174 and
349 male- and female-biased germline genes, respectively, were identified. Both the
male- and female-biased pupal germline sets had significant overlap with genes
previously identified as highly expressed in adult male and female germline tissues,
respectively (P<0.05, hypergeometric test for both sets; (Parisi et al., 2003)). Nine
hundred seventy-nine genes identified in the pupal male germline set were on the
previous study’s array platform (Parisi et al., 2003). Of those 979 genes, 756 were highly
expressed in the adult male germline. Similarly, 298 genes identified here as being in the
pupal female germline set were present on the previous study’s array platform. Forty-
five of those 298 genes were also highly expressed in the adult female germline.
45

In males, the total number of genes expressed in the germline during metamorphosis
(1174 genes) was similar to the total number previously identified at the adult stage (1951
genes). However, in females, we found many fewer genes expressed in the female
germline during metamorphosis (349 genes) as compared to genes expressed in the
female germline at the adult stage (1113 genes) (Parisi et al., 2003). It is known that
spermatocyte differentiation initiates as early as larval stages, whereas oocytes are not
observed until late metamorphosis (reviewed in Fuller, 1993; Williamson and Lehmann,
1996), consistent with our observation of greater numbers of genes identified in male, as
compared to female, germline tissues. Furthermore, the large overlap of genes expressed
in the pupa and adult male germline tissues is consistent with early spermatocyte
differentiation, while the later onset of oocyte development may explain why there are
fewer genes that are expressed in both the pupal and adult female germline tissues.

Ninety-two of the 1174 genes in the pupal male germline set function in the mitochondria
(P<0.05; DAVID analysis), consistent with the essential role for mitochondria in
spermatid development and adult function (reviewed in Fuller, 1993). 223 genes were
identified that are expressed in the male germline during metamorphosis, but are not in
the previously identified adult male-germline set (present on platform of previous study,
but not significantly differentially expressed (Parisi et al., 2003)), suggesting that there is
pupal-specific, male-germline gene expression that might underlie male germline
development.
46

Previously, it was observed that most genes expressed in the female-germline showed the
first post-embryonic peak of transcript abundance during adult stages (Arbeitman et al.,
2002). However, because our previous study did not have data from pupal stages
examining gene expression in each of the sexes separately, or in male and female tud
progeny, we were unable to definitively identify the genes expressed in the female
germline at pupal stages. The data presented here demonstrate that there are a substantial
number of female-biased germline genes that are expressed during pupal stages.
Between the 48 and 72 hour APF stages, the structures derived from the female genital
disc establish connections with female gonadal tissues to form the female reproductive
system (reviewed in Spralding, 1993). The development of female reproductive
structures likely requires expression in both somatic and germline tissues. This idea is
consistent with the functions of genes with pupal female germline expression, as this set
is overrepresented with genes that function in organ development (P<0.05; DAVID
analysis).

Interestingly, the transcript levels of genes in the pupal female-germline set is at peaks in
both wild type males and females and tud progeny males and females at the early stages
of metamorphosis (Figure 2.3), suggesting they also play a non-sex-specific role in pupal
somatic tissues.

47

Several genes annotated as functioning in the female germline (found in Cluster 21), peak
in transcript abundance in male and female somatic tissues at 48 hour APF (Figure 2.1).
However, by the later stages of metamorphosis, the levels of these transcripts remain high
Figure 2.3. Expression profiles of genes expressed in the male and female germline during
metamorphosis.
Expression profiles were generated by averaging the gene expression data of all genes that have high
expression in the male or female germline, at each time point. The data for these genes for following
genotypes were averaged separately: male tud progeny, female tud progeny, wild type (CS) males, and
wild type (CS) females. Yellow and blue indicates high and low levels of expression compared to a
common reference, respectively.

48

only in wild type females and drop to trough levels in wild type males and tud male and
female progeny.

The chromosomal distribution of genes with sex-biased expression in the male and
female germlines was additionally analyzed. Genes expressed in the pupal male germline
are underrepresented on the X chromosome and over-represented on the left arm of the
second chromosome (P<0.01, hypergeometric test), both of which have been shown for
genes expressed in adult male germline tissues (Parisi et al., 2003). Interestingly genes
expressed in the pupal male germline are also overrepresented on the right arm of the
third chromosome (P<0.01, hypergeometric test). The female-biased germline set was
over-represented on the X chromosome (73 genes; P<0.05, hypergeometric test), similar
to what has been previously seen for genes expressed in female adult germline tissues
(Parisi et al., 2003).

Global transcriptional profiles during metamorphosis
Hierarchical clustering was performed using all the data from each microarray
experiment from the time course study, rather than using the data from each gene, to
determine how similar global expression patterns are between males and females. When
the tud progeny expression data was analyzed, the global expression profiles of males and
females from each pupal time point were the most similar to each other (Figure 2.4), as
expected given that we identified very few genes with somatic, sex-differential
expression (see above). A clear distinction between overall gene expression at early
49

stages (0-48 hour APF pupae), as compared to late stages (72-96 hour APF pupae) was
observed (Figure 2.4A). This is consistent with our cluster analyses (Figure 2.1), where
many genes appear co-regulated at either early or late stages of metamorphosis, but not at
both early and late stages: 1818 genes shared either peaks or troughs of transcript levels
at multiple early stages (0-48 hour APF) or late stages (72-96 hour APF), while only 475
genes shared peaks or troughs of transcript levels at an both an early and a late time point.

When the wild type expression data and the tud progeny expression data were analyzed
together, a clear distinction between early and late metamorphosis was seen (Figure
2.4B). As expected, male-germline gene expression has a large effect on how the global
transcriptional profiles cluster, with wild type male expression data always clustering
separately from wild type females and from the male and female tud progeny. This effect
appears to be less substantial at 0 hour APF, as the global expression profile of wild type
males clusters closest to the other three genotypes, suggesting that at the start of
metamorphosis many genes expressed in the male germline are not as abundant, as at
later time points. The wild type female array experiments also cluster closely to, but
separately from, the tud progeny array data, with the largest differences seen at 72 and 96
hour APF. This is consistent with gene expression in the female germline increasing at
the end of metamorphosis as we observed (Figure 2.3).

50

Figure 2.4. Cluster of global expression profiles for Drosophila transcripts across metamorphosis.
Dendrogram shows the similarity across transcriptional profiles at five time points during metamorphosis
(0, 24, 48, 72, and 96 hour APF) for (A) male tud progeny (blue) and female tud progeny (red) and (B)
male tud progeny, female tud progeny, wild type (CS) males (green), and wild type (CS) females (violet).
Hierarchical clustering and Pearson correlation distance measure was used to group experiments based on
their global expression profile using all expression data for annotated genes from each array.

51

Sex-hierarchy regulated somatic sex-differential expression in 48 hour APF pupae
Next, genes regulated by the sex hierarchy during pupal developmental stages were
identified. Nearly all of the sexually dimorphic tissues are either patterned or undergoing
morphogenesis to bring about the adult sexual dimorphisms during pupal stages. The 48
hour APF pupal stage was chosen, as previous studies showed that FRU
M
peaks at this
stage (Lee et al., 2000a) and DSX shows high expression at this time (see Chapter 4).
For these experiments, the array hybridizations were performed as direct comparisons
using RNA from the two genotypes (see Materials and Methods). Genes were first
identified that had sex-biased transcript levels between wild type males and females, by
analyzing expression data from two different wild type strains, Canton S and Berlin
(q<0.15), and from tud progeny males and females (one-tailed t-test, q<0.15). This
resulted in a set of 420 genes (320 and 100, female- and male-biased genes, respectively).
This is substantially more than was identified in the time course analysis for this time
point (41 genes, see above); however this discrepancy is likely due to the increased
number of replicates and power in this analysis (four replicates for each comparison
versus three replicates in time course) and the decreased statistical error by directly
comparing gene expression on the same array, versus using a common reference RNA
sample (see Table 2.2 for overview of experiments).

52

Experiment Rationale Number of Genes
†

Wild type male and
female whole pupae

Identify genes with sex-biased
expression in whole pupae

7972
tud progeny male and
female (germline
minus)

Identify genes with sex-biased
expression in somatic tissues
421
tra pseudomales and
wild type females

Identify genes with sex-biased
expression regulated by TRA

95
dsx
D
pseudomales and
wild type females

Identify genes with sex-biased
expression regulated by DSX
M

and/or DSX
F

173*
XX dsx null and wild
type females

XY dsx null and wild
type males

Identify genes regulated by DSX
F

in females

Identify genes regulated by DSX
M

in males

154

155

Given the larger number of somatic sex-differentially expressed genes identified by this
approach, it could be determined if there was a bias for chromosomal positions in these
gene sets. It has previously been shown that genes with somatic male-biased transcript
levels at the adult stage are underrepresented on the X-chromosome (Parisi et al., 2003).
There was not a similar bias for the 100 genes with somatic, male-biased transcript levels
Table 2.2. Microarray experimental design used to identify genes regulated downstream of DSX.
†

The number of genes in each experiment that is significantly differentially expressed. * The DSX set was
defined with an additional statistical test to eliminate false-negatives resulting from the TRA set;
therefore, a gene’s inclusion in the DSX set is not necessarily contingent on its presence in the TRA set.
For a description of the definition of the DSX set, see text.

53

at the 48 hour APF pupal stage, but rather, there was an equal distribution across all
chromosomes (hypergeometric test, p>0.05). Interestingly, the genes with somatic
female-biased transcript levels at this pupal stage were over-represented with genes
located on the X chromosome (90 genes, hypergeometric test, P<0.001). The previous
study did not find any significant over- or underrepresentation on any chromosome for
genes with adult somatic female-biased expression in the adult (Parisi et al., 2003).

Genes differentially expressed as a consequence tra
Next, genes were identified that were differentially expressed as a consequence of tra, a
gene in the sex hierarchy that encodes a pre-mRNA splicing factor required for the
production of the sex-specific dsx mRNA splice variants (McKeown et al., 1988).
Transcript levels in chromosomally XX flies mutant for tra (hereafter called tra
pseudomales) was compared to wild type female flies. The tra pseudomales produce
DSX
M
and look very similar to wild type males. Of the 420 genes that showed somatic
sex-biased transcript levels, 95 genes were identified (72 female-biased and 23 male-
biased) that are also significantly different between tra pseudomales and wild type
females (q <0.15 for each test). As a validation of our experimental approach, Sxl, tra,
roX1, and roX2 are all sex-differentially expressed in the somatic tissues, as expected
(wild type and tud progeny comparisons; Figure 1.1). Only tra is differentially expressed
in the tra mutant comparisons; this is expected as Sxl, roX1, and roX2 are not regulated
downstream of tra in the sex-determination hierarchy (Figure 1.1; (reviewed in Mendjan
and Akhtar, 2007)).
54

Of the 326 genes that are not regulated by TRA, a large portion (169) may be false
negatives as they are significant or close to significantly different (q<0.30) in microarray
experiments identifying genes regulated by dsx (see below). A gene regulated
downstream of dsx should also be regulated downstream of tra. Another 19 have a q-
value close to the cutoff for significance in the tra microarray expression data (q<0.30).
Removing these 188 genes from consideration still leaves a large number of genes (138
genes) that are sex-differentially expressed independently of tra. A significant number of
these 138 genes (45 genes; P<0.05, hypergeometric test) are located on the X
chromosome. It is possible that differences in transcript levels of these genes is due to
differences downstream of Sxl, or X chromosome composition in males and females,
which gives evidence to the notion that the dosage compensation process does not
completely normalize expression between males and females for all genes on the X
chromosome.

Genes differentially expressed as a consequence dsx
Next, gene expression between pupae that are transheterozygous for the dsx
D
allele
(Duncan and Kaufman, 1975), dsx
D
only produces the male-specific isoform (DSX
M
),
and a dsx null deletion allele (dsx
m+r15
) was compared to gene expression in wild type
females. These chromosomally XX, dsx
D
/dsx
m+r15
pseudomales look very similar to wild
type males, as they only produce DSX
M
. Of the 95 genes that are sex-differentially
expressed in somatic tissues and downstream of tra, 66 genes were identified as being
55

regulated downstream of dsx. Forty-six and 20 genes are more highly expressed in
females and males, respectively, downstream of tra and dsx (Table 2.3).

Table 2.3. Genes expressed downstream of DSX at 48 hour APF.
All genes listed had at least a 2-fold difference in gene expression when comparing XX dsx
D
pseudomales to wild type females at 48 hour APF.
*
Annotation symbol from FlyBase.
†

Chromosomal arm on which the gene is located.
‡
Fold change difference between the sexes.
ξ
GO
annotations for each gene as listed in Flybase.

56

Table 2.3, continued

57

Table 2.3, continued

58

Aside from the tra gene itself, 26 genes were in the TRA gene set but were not
differentially expressed between dsx
D
and wild type females, although 25 of these genes
had data from three of the four dsx
D
comparisons. If genes that are close to the
significance level (q<0.30; 6 genes) and the gene with expression data in only two dsx
D

comparisons are removed, 19 genes remain that are downstream of tra, but not dsx.
Interestingly, none of these genes are significantly differentially expressed in similar
experiments examining FRU
M
regulation at this stage (data not shown). This suggests
either an alternate branch of the sex-hierarchy downstream of tra, possibly through
dissatisfaction (Finley et al., 1997), or the possibility of additional genes on which TRA
acts to sex-specifically splice their pre-mRNAs, leading to differential abundance of
transcripts due to differences in mRNA stability.

Requiring a gene to show statistical differences in expression in all of our direct
microarray experiments yields a high confidence set of true positives regulated
downstream of dsx, but will likely generate false negatives. To identify additional dsx
regulated genes that might have been missed because of the stringency of having to pass
multiple tests, genes were included that showed sex-differential, somatic expression, but
which were not differentially expressed in the tra microarray comparisons. These genes
were required to be significantly differentially expressed in the dsx
D
comparisons at a
more stringent level (q<0.05) to avoid false positives. This yielded an additional 107
genes, with 75 and 32 showing female- and male-biased expression, respectively (Table
59

2.3). This study thus identified 173 genes regulated as a consequence of dsx (DSX set;
121 and 52 female- and male-biased genes, respectively).

Several of the male- and female-biased genes in the DSX set with the highest fold change
include those that might be involved in epithelial morphogenesis, imaginal disc
morphogenesis or cuticle formation, based on their sequence identity. The male-biased
set contains seven such genes, including ecdysone inducible ImpE1 (FC=4.8), miniature
(FC=4.5), and dusky-like (FC=3.1). Among the 15 female-biased genes there are four
genes that encode proteins with cuticular domains (Cuticular protein 97Eb, 50Ca, 97Ea,
and 51A; FC=5.9, 3.5, 3.3, and 2.8, respectively), as well as obstructor-A (FC=2.5) and
abdominal A (FC=1.6). While it has long been recognized that cuticle deposition is tied
to tissue morphogenesis, and both are developmental events occurring during the middle
of metamorphosis, the identification of several genes likely involved in sex-specific
aspects of this process had not been determined until this study.

Another class of genes that showed differential expression downstream of dsx at this
stage is involved in proteolysis. The gene with the greatest expression in females as
compared to dsx
D
pseudomales (CG6337; FC=15.2) is a predicted cysteine-type
endopeptidase, and six additional female-biased genes are annotated as functioning as
peptidases or in proteolysis (CG14218, FC=3.8; CG6592, FC=2.9; CG9850, FC=2.4;
omega, FC=2.4; CG8358, FC=2.1; and CG11771, FC=1.3). Interestingly, one of the
greatest male-biased genes in the DSX set (CG1342; FC=6.0) is predicted to act as an
60

endopeptidase inhibitor, as is a second male-biased gene (CG32354; FC=2.7). A third
gene with high levels of male biased expression (CG4386; FC=4.5) is a predicted serine
protease. This is consistent with our observations of sex-biased gene expression in the
somatic tissues during metamorphosis, where genes involved in proteolysis showed high
levels of expression differences between male and female flies (Table 2.1). This suggests
the importance of regulating sex-specific processes through the cleavage of specific
proteins.

Six female-biased genes regulated downstream of DSX at 48 hour APF have functions in
the muscle or muscle differentiation (flightin, Limpet, CG31781, Tropomyosin 1,
abdominal-A and Sarcoplasmic calcium-binding protein (Arbeitman et al., 2002; Basi et
al., 1984; Kelly et al., 1997; Kim et al., 2004; Michelson, 1994; Vigoreaux et al., 1993)).
This suggests that aspects of pupal muscle development occur in temporally distinct
manner between males and females, and that this differential timing is regulated by DSX.
It is not clear is if this is due to the development of sex-specific muscles or due to
differences in the developmental rate of non-sex-specific muscles, between males and
females.

vrille (vri), a DSX-regulated male-biased gene (FC= 3.0), encodes a transcription factor
that functions at many stages, including playing an essential embryonic role, underlying
bristle morphogenesis, and imaginal disc-derived wing hair organization (Szuplewski et
al., 2003). In adult stages, vri encodes a molecular component of the circadian clock and
61

thus plays a role in locomotor activity (Blau and Young, 1999). The sex-differential
expression at pupal stages is consistent with the known roles of vri in imaginal disc
organization, though further analyses are required
.

Characterization of the modes of DSX regulation
In our previous microarray study examining modes of DSX-regulated gene expression at
the adult stage in head tissues, a large number of sex-biased genes was found that were
either activated or repressed as consequence of dsx activity in both males and females,
but that the extent of activation or repression was sex-specific (Arbeitman et al., 2004;
Goldman and Arbeitman, 2007). This mode of regulation was distinct from the previous
descriptions of DSX-regulated gene expression based on the only known direct targets of
DSX, Yolk protein 1 (Yp1) and Yolk protein 2 (Yp2). DSX
F
activates Yp1 and Yp2
expression in the female fat body and DSX
M
represses Yp1 and Yp2 expression in the
male fat body tissues (Burtis et al., 1991).

To determine the modes of regulation by DSX in pupal stages, gene expression was
compared between chromosomally XX and XY dsx null flies and wild type females and
males, respectively (hereafter called dsx null comparisons); we note that in this study
regulation by DSX might be direct or indirect. Data was examined for the 173 genes we
identified here as being downstream of dsx (DSX set; Table 2.3). There were 15 genes
for which there was not enough data for statistical analysis or no significant differential
62

expression in either dsx null comparison; these 15 genes were therefore not considered
for further analysis. Of the remaining 158 genes, 151 show significant differential
expression (q<0.15) in both dsx null comparisons, suggesting regulation by both DSX
F

and DSX
M
. The remaining seven genes of the 158 genes only showed significant
differential expression in one of the dsx null comparisons; these seven genes may
possibly be regulated by one isoform of DSX, a method of DSX regulation that was
previously posited for some genes with sex-differential expression in the adult
(Arbeitman et al., 2004; Goldman and Arbeitman, 2007).

Of the 151 genes regulated as a consequence of dsx in both sexes, 104 of the genes were
female-biased and were more highly expressed in wild type females and males as
compared to dsx null females and males, respectively. This suggests that these genes are
activated by DSX in both females and males, but that DSX
F
is a more potent activator.
40 of the 151 genes were male-biased and more highly expressed in male and female dsx
null flies than in wild type males and females, suggesting these genes are repressed as a
consequence of DSX activity in both males and females. Thus, the majority of genes that
are regulated as a consequence of dsx are not regulated in the Yp-like mode of regulation,
but rather are regulated similarly in both sexes, with one isoform acting as a more potent
regulator, as previously described in our studies of adult head tissues (Figure
2.5)(Goldman and Arbeitman, 2007). Interestingly, Yp-like regulation was observed for
only seven genes in our pupal dataset: the male biased genes CG8086 and CG14995 and
the female-biased genes abdominal-A (abdA), LpR1, CG10802, CG1441, and CG9485.
63

abdA, which appears to be activated downstream of DSX
F
in females and repressed
downstream of DSX
M
in males, is a well-characterized homeotic selector gene that was
shown to be important for specifying segment identity (Sanchez-Herrero et al., 1985;
Tiong et al., 1985). Previous research examining 40-45 hour APF, which is close in time
to the stage examined here (48 hour APF), suggested that ABD-A and DSX, along with
Abdominal-B, act to regulate the expression level of a downstream target, bric-a-brac,
and lead to differential abdominal pigmentation between males and females (Kopp et al.,
Figure 2.5. Model of DSX modes of regulation.
(A) Model depicting the mode of DSX regulation of Yolk protein 1 (Yp1) as described in (Burtis et al.,
1991). In females, DSX
F
binds to the upstream region of Yp1 and activates its expression, while in
males DSX
M
binds to the same region and represses Yp1 expression. This mode of regulation was
seen for seven genes in our DSX-regulated set. (B) Proposed model for female-biased genes. DSX
F

activates the target gene in females and DSX
M
activates the target gene in males, but DSX
F
is a more
potent activator. 104 genes in the DSX-regulated set followed this mode of regulation. (C) Proposed
model for male-biased genes in which there is a basal gene expression level independent of DSX.
DSX
F
represses the target gene in females, and DSX
M
represses the target gene in males. However,
DSX
F
is the more potent repressor, which results in these genes having male-biased expression. 40
genes in the DSX-regulated set followed this mode of regulation. (D) Model depicting the mode of
regulation where only one isoform of DSX regulates gene expression levels. In the example shown,
DSX
F
activates gene expression in females, while DSX
M
does not regulate gene expression in males.
This mode of regulation includes genes that are either activated or repressed by only one isoform of
DSX; this was true for seven genes in our DSX-regulated set.

64

2000). In that study it was shown by in situ analyses that abdA transcript levels in the
abdominal epidermis do not vary between dsx null animals and wild type animals, thus
suggesting that abdA is not regulated by DSX in this tissue. The dsx-dependent
differential expression of abdA that we observed could be due to expression in other
tissues, since here whole pupae were analyzed. Indeed, abdA has been shown to be
expressed and functional in several distinct tissues and cell types, including in the
abdominal neuroblasts and the female genital disc (Bello et al., 2003; Freeland and Kuhn,
1996).

The proposed modes of regulation were validated by additional microarray experiments
in which the male and female isoforms of DSX were over-expressed (data not shown).
Of the 157 genes in the DSX-regulated set for which DSX modes of regulation was
examined, 46 did not show significant differential expression in the experiments when we
either over-expressed DSX
F
in females or DSX
M
in males. Of the remaining 111 genes,
35 were male-biased; these genes showed decreased expression when DSX was over-
expressed, either in one or both of the DSX isoform over-expression experiments.
Similarly, of the remaining 76 female biased genes, 74 showed increased expression
levels when DSX was over-expressed, either in one or both of the DSX isoform over-
expression experiments. Only two female-biased genes (CG4484 and CG3244) showed
decreased expression levels when DSX
F
was over expressed in females compared to
control females, opposite of the predicted effect from our model (Figure 2.5).
65

Here, we provide an analysis of the gene expression changes that underlie the
developmental transitions during metamorphosis in wild type and germline deficient
males and females. We have identified many genes that likely function in sex-specific
patterning in somatic tissues, at discrete times during development. For one time point,
we have examined gene expression downstream of tra and dsx, sex determination
hierarchy regulatory genes. Our analyses also identified genes expressed in the both the
male and female germline during pupal stages. Finally, the data presented here extend
the idea that the most well-characterized mode of dsx-regulated gene expression, in
which DSX
M
and DSX
F
regulate a given gene as either an inducer or repressor, is the
exception, rather than the rule.

2.4. Materials and Methods
Fly collections and strains
Male and female flies were collected at the white pre-pupal stage between Zeitgeber time
(ZT) 1 and ZT 4 and aged at 25
◦
C for the following hours: 0, 24, 48, 72, and 96 (time
course) or 48 (sex hierarchy mutants), and then snap frozen in liquid nitrogen. Wild type
flies were Canton S (CS) and Berlin. Animals that lack germline tissues (called tud
progeny) were the progeny of female tud
1
bw
1
sp
1
and male y,w, P [w
+cM
UBI-GFP];ID-1
P[FRT(w
hs
)]101. Female tud progeny had GFP expression and could be distinguished
from males by fluorescence microscopy. Below, chromosomal sex for sex hierarchy
mutant flies is indicated in parentheses. Chromosomally XX sex hierarchy mutants were
identified based on the GFP marker on the X chromosome, derived paternally. tra
66

pseudomales were the genotype y,w, P[w
+cM
UBI-GFP]/+;tra
1
/ Df(3L)st-j7 (XX) and
were compared to CS females. dsx pseudomales were y,w ,P[w
+cM
UBI-GFP]/+;dsx
D
,
Sb
1
, e
1
/ dsx
m+r15
(XX), and compared to CS females. For the dsx null analyses, P[w
+cM

UBI-GFP]/+; dsx
d+R3
/ dsx
m+r15
(XX) flies were compared to CS females and dsx
d+r3
/
dsx
m+r15
(XY) flies were compared to CS males. All flies were kept at 25
◦
C in a 12:12
hour light-dark cycle and grown using standard food media.

Microarray experiments
All time course microarray experiments were conducted with three replicates; for every
experiment, cDNA from the experimental genotype contained incorporated Cy3-labeled
dUTP and cDNA from the reference sample contained incorporated Cy5-labeled dUTP.
The common reference was comprised of mRNA from CS flies of both sexes from all
pupal stages. Microarray comparisons using RNA derived from sex-determination
hierarchy mutants were conducted with four replicates, with a dye-swap design; i.e.
cDNA from each genotype contained incorporated Cy3-labeled dUTP in two experiments
and contained incorporated Cy5-labeled dUTP in the other two experiments. RNA was
isolated from ~30 pupae by homogenization and extraction using the TRIzol
®
protocol
(Invitrogen, Carlsbad, CA) and resuspended in 20µL diethylpyrocarbonate (DEPC)-
treated H
2
O.

cDNAs were directly labeled with Cy5 or Cy3 during the reverse transcription reaction;
30µg of total RNA was used as a starting template. The reverse transcription reaction
67

was performed for two hours at 42
◦
C using the following reagents (values in parentheses
are final molarity or final concentration): oligo dT primer (Operon, 3.75 µM),
dithiothreitol (Invitrogen, 10 mM), First Strand Buffer (Invitrogen, 1×), dNTPs minus
dTTP (Invitrogen, 0.5 mM), dTTP (Invitrogen, 50 µM), Cy-labeled dUTP (Perkin-Elmer,
0.625 nM), and Superscript II reverse transcriptase (Invitrogen, 10 U/µL). The reaction
was stopped and RNA was hydrolyzed by a 20 minute incubation at 65
◦
C with NaOH
(167 mM) and EDTA (83 mM). After neutralizing by adding HEPES buffer (pH8.0; 294
mM) and sodium acetate (pH5.2; 228 mM), cDNA samples were purified (Gel-
Purification Kit, Qiagen, Valencia, CA). cDNA samples were then dried and
resuspended in formamide (56%), sodium citrate buffer (SSC, 3.37×), SDS (1.12%),
Denhardts (5.62×), and Polyadenylic acid potassium salt (0.9 mg/ml; Sigma-Aldrich, St.
Louis, MO). Samples were boiled for 2 minutes and then applied to microarray slides
underneath a LifterSlip (Erie Scientific, Portsmouth, PA). Microarrays were hybridized
at 42
◦
C for 14-18 hours, then washed in a solution of 1.5% SDS and 1X SSC for 5
minutes, a solution of 0.20X SSC for 5 minutes, and two solutions of 0.05X SSC for 10
minutes each.

Microarray production and analysis
The oligonucleotide set that was printed on the glass slides consisted of 15,156
oligonucleotides, representing the full predicted set of transcribed regions of the D.
68

melanogaster genome, including 14,454 known and predicted open reading frames and
an additional 702 control spots. The 14,454 oligonucleotides represent 13,820 unique
genes. The oligomer set was designed by the International Drosophila Array Consortium
(INDAC; http://www.indac.net/) based on release 4.1 of the D. melanogaster genome
using a custom implementation of OligoArray2 (Rouillard et al., 2003). The
oligonucleotides were designed with sizes ranging between 65–69 nucleotides, a minimal
Tm window, bias towards the 3′-ends of transcripts, and minimal sequence similarity to
other genes (Brown et al., 2006). The oligonucleotides were synthesized by Illumina
(San Diego, CA); sequences can be downloaded from Flymine:
http://www.flymine.org/release-5.0/aspect.do?name=INDAC. Additionally, microarrays
used for the time course analyses and the dsx null analyses contained additional array
elements for Sxl, tra, the female-specific splice form of dsx (dsx
F
), the male-specific
splice form of dsx (dsx
M
), the fru transcript that is sex-specifically spliced to produce
FRU
M
(fru P1), and each of three of the fru DNA-binding domains (fru
A
, fru
B
, and fru
C
).
Sequences for these oligonucleotides can be found in Table A2. All microarrays were
printed in the laboratory of Dr. Eric Johnson at the University of Oregon (Eugene, OR)
using slides coated with aldehyde chemistry and were postprocessed using the Nunc
SuperChip Aldehyde protocol (Thermo Fisher Scientific, Waltham, MA).

All arrays were scanned using the GenePix 4100A scanner and GenePix Pro 5.0 software
from Axon Instruments (Molecular Diagnostics, Sunnyvale, CA). Visual inspection of
the microarray images filtered out florescence most likely not due to labeled cDNA
69

binding; the data from these array elements was discarded. Array elements were only
considered for further analysis if at least one channel (Cy3 or Cy5) had greater than 75%
of the pixels with intensity values one standard deviation above background levels. All
microarray normalization and statistical analyses were performed using the limma
package of BioConductor in the program R (Gentleman et al., 2004; Smyth, 2005; Smyth
and Speed, 2003). Global-loess normalization was used for all arrays, and significant
genes were found using a t-test and q-value application for R (Storey, 2003).

Microarray data can be accessed at NIH GEO database at the following web link:
http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=dnclzosomqyyedy&acc=GSE113
16. The GEO accession number is GSE11316.

Analyses of time course microarray study
Two-way ANOVA analyses on the wild type and tudor time course data used the
logarithm of the ratio data to find genes that varied significantly with (1) sex and time or
(2) genotype and time.

To determine experimental reproducibility of the microarray data for the time course
study, we calculated correlation values for each experiment using Pearson’s correlation
on the logarithm of the ratio values for every oligonucleotide on the microarray. For each
experiment, pairwise Pearson correlation comparisons were performed on the three
microarray replicates. All comparisons between replicates for the tud progeny
70

experiments of the time course microarray study had correlations >0.75 and 27 of the 30
replicate comparisons had correlations >0.80. See Table A3 for values.

To identify the number of genes expressed during metamorphosis, we required a gene to
have expression values for two of three microarray replicates in at least one time point in
either of the sexes.

To identify genes with expression only in one sex and only at one stage of development
in the somatic tissues, for each of the five time points, and each of the sexes, we required
the following criteria to be satisfied in the tud progeny data: 1) the gene must be
significantly differentially expressed between the experimental samples of interest and
the reference sample, 2) the gene must have 5-fold greater expression in the experimental
sample of interest and the reference sample, and 3) the gene must not have enough
expression data for analysis in each of the remaining nine sex and time point
experiments.

Clustering of time course microarray expression data
Clusters were generated using the tud progeny data and the program Cluster (Eisen et al.,
1998). Genes in the cluster must have had expression values for at least three of the five
time points for both tud progeny females and males. The logarithms of the ratio values
for each sex were then median-centered for each gene and clustered. For clustering, we
used Pearson correlation as the distance measure and defined similarity between clusters
71

using average-linkage clustering. The cluster files outputted were then imaged and
analyzed using Java TreeView (Saldanha, 2004). Co-regulated clusters were defined as
any group of genes for which the average correlation was at least 0.80 and contained at
least 15 members. In Figure A1.1, the average expression for the genes in each cluster in
the wild type male and female experiments is included; this data is also median-centered
separately for each sex, but is not used to derive the clusters.

The graphical images in Figure A1.1 were generated by averaging the logarithm of the
ratio data for all member genes of each cluster at each time point. The cluster intensities
represent the combined average for male and female tud progeny. Enrichment and
depletion for gene expression in each cluster were determined as follows: For each
cluster, the mean of gene expression at all time points was calculated. A time point was
considered to be enriched if the average gene expression at a particular time point was
greater than 2/3 of a standard deviation above the mean. Conversely, a time point was
considered to be depleted if the average gene expression at a particular time point was 2/3
of a standard deviation below the mean.

The dendrograms shown in Figure 2.4 were produced using the logarithms of the ratio
data for every annotated gene spotted on the microarray, not including control spots.
Clusters were generated using the program Cluster (Eisen et al., 1998), Pearson
correlation as the distance measure, and average-linkage clustering to define similarity
72

between clusters. The cluster files outputted were then imaged using Java TreeView
(Saldanha, 2004) and the resulting dendrograms were exported.

Analysis of over-represented features
For all of the below statistical analyses, significance was declared if P<0.05. Significant
over-representations of functional annotations were generated with the program DAVID
(Dennis et al., 2003). Unique GenBank accession number identifiers for the gene list and
the whole set of unique GenBank accession numbers for all possible transcripts in the full
array set (13,614 genes total) were used. When searching for over-represented functional
annotations using DAVID, we selected the following categories: all levels (or only Level
4) of each of the three Gene Ontology (GO) categories, Uniprot Sequence Features,
Swiss-Prot Keywords, KEGG metabolic pathways, InterPro domains, PIR superfamily
names, and SMART domains (Ashburner et al., 2000; Kanehisa et al., 2008; Letunic et
al., 2004; Mulder et al., 2007; UniProt Consortium, 2008; Volff et al., 2003).

Additionally, we wrote our own program to analyze over-representation of chromosomal
location and conserved genes (personal communication, Matt Lebo). This code used a
gene’s Flybase number as the unique identifier and the background set consisted of the
whole set of unique Flybase numbers for all possible transcripts in the full array set
(13558 genes total). Code is written in Java and is available upon request. Significance
was declared using the binomial approximation of the hypergeometric test on the list of
73

unique identifiers, with all possible unique identifiers in the full array set as background.
The 2500 most evolutionarily conserved DNA elements represent highly conserved
Drosophila sequences ranging from hundreds to thousands of base pairs in length. Genes
whose coding regions overlap with the 2500 most conserved DNA elements, as generated
by PhastCons (Siepel et al., 2005), were identified as over-represented conserved genes.
In the previous study (Siepel et al., 2005), multiple alignments of the genomic sequences
of four insect species were used to determine conservation.

Identification of genes expressed in germline tissues during metamorphosis
To identify genes that are expressed in the male and female germlines during
metamorphosis, we used two-way ANOVA on the gene expression data across the five
time points sampled (0, 24, 48, 72, and 96 hour APF). We first identified genes with
male- or female-biased expression in the wild type tissues using a two-way ANOVA
analysis with sex and time as the independent factors. Genes with sex differential
expression had significant differences with sex as the independent factor (q<0.05). To
identify genes expressed in the male germline, we used two-way ANOVA on the wild
type male and male tud progeny data, with genotype and time as the independent data.
Genes expressed in the male germline were those with male-biased expression in the wild
type ANOVA analysis and wild type male-biased expression in the wild type/tud progeny
male ANOVA analysis (genotype as independent factor; q<0.05). To avoid false
negatives, we also include genes that display significant, male-biased expression in the
wild type experiments (sex as the independent factor; q<0.05), but for which there is no
74

expression data in the male tud progeny experiments. The set of genes with expression in
the female germline was defined in a similar manner using the female wild type and tud
progeny expression data.

Identification of genes expressed downstream of the sex-determination hierarchy
To identify genes with sex-differential expression in the somatic tissues, we initially
identified genes showing sex-differential expression in wild type flies by combining the
gene expression data from both wild type experiments (CS and Berlin) and selecting
genes with significant sex-differential expression (q<0.15). We next refined this list
using the expression data from male and female tud progeny and a one-tailed t-test
(q<0.15) by assuming the sex-differential expression should be in the same direction for
the wild type and tud progeny experiments. To identify genes that sex-differentially
expressed in the somatic tissues and that are downstream of tra, we refined the list of
genes with somatic sex-biased expression using a one-tailed t-test on the tra expression
data (q<0.15), assuming female-biased genes have higher expression in wild type females
and male-biased genes have higher expression in tra pseudomales. Finally, to get a list of
genes expressed downstream of dsx in somatic tissues, we refined the list of genes
expressed downstream of tra by using a one-tailed t-test on the dsx
D
expression
comparison (q<0.15), assuming female-biased genes have higher expression in wild type
females and male-biased genes have higher expression in dsx
D
pseudomales.
Chapter 3. Ecdysone Receptor mediates courtship behaviors in Drosophila by acting
in the fruitless circuit in an isoform-specific manner
75

3.1. Overview
In Drosophila, male-specific FRU (FRU
M
) is required to establish the potential for
courtship behaviors, but the downstream effectors of FRU
M
during development are
largely unknown. A microarray-based approach identified genes that are differentially
expressed as a consequence of FRU
M
in pupae, in both whole body and CNS tissues.
Genes were also identified that are sex-differentially expressed in CNS tissues. The
FRU
M
-regulated gene sets were significantly overrepresented with genes also regulated
by the ecdysone regulatory pathway. Two EcR isoforms (EcRA and EcRB1) are
expressed in FRU
M
-expressing neurons during distinct periods of metamorphosis. Males
with abrogated EcRA function in FRU
M
-expressing neurons aggressively court other
males. These results demonstrate a novel role for EcR in specifying male courtship
behavior through its actions specifically in the FRU
M
neural circuitry.

3.2. Introduction
In Drosophila, the male courtship ritual is an innate, sequence-specific behavior required
for successful copulation with a female fly (reviewed in Manoli et al., 2006). Using
olfactory, gustatory, and visual cues, mature wild type males robustly court females, but
do not court other mature males (Jallon and Hotta, 1979). The master regulatory genes
necessary for proper male courtship behavior have been identified (see Figure 3.1)
(reviewed in Greenspan and Ferveur, 2000), which provides a tractable system to study
the molecular-genetic basis of a complex behavior. However, little is known about how
76

the master regulatory genes function during development to establish the potential for the
male-courtship ritual.

fruitless (fru) is required for all aspects of male courtship behaviors through the actions of
male-specific BTB-zinc-finger transcription factor isoforms (FRU
M
) that are encoded by
transcripts generated from one of four fru promoters, the P1 promoter (Ryner et al., 1996;
Villella et al., 1997). fru P1 transcripts are sex-specifically spliced by splicing factors in
the Drosophila sex-hierarchy, the regulatory cascade that controls all aspects of somatic
sexual development (reviewed in Christiansen et al., 2002)(see Figure 3.1). FRU
M
is
expressed in approximately 2% of neurons in the mid-pupal male CNS (Lee et al.,
2000a), as well as regions of the peripheral nervous system (Manoli et al., 2005;
Stockinger et al., 2005). However, overt differences between the fru P1-expressing
neural circuit in males and females were not apparent (Manoli et al., 2005; Stockinger et
al., 2005), leaving how FRU
M
specifies the potential for male behaviors an open
question. It is clear that FRU
M
plays a major role in the circuit, as when it is produced in
fru P1-expressing neurons in females, it is sufficient to provide the potential for females
to perform early steps of male-specific courtship behavior (Manoli et al., 2005;
Stockinger et al., 2005).

Studies have shown that FRU
M
regulates dimorphism in neuron number, as
approximately 25 fru P1 neurons undergo apoptosis in females, due to the absence of
FRU
M
(Kimura et al., 2005). However, because this is a small number of fru P1 neurons,
77

directing cell death is unlikely to be the primary role of FRU
M
. In addition, fru P1-
dependent differences in neuronal connectivity were observed in projections from
antennal lobe (the DA1 glomerulus) to the protocerebrum (Datta et al., 2008),
demonstrating that FRU
M
directs differences in fine-scale connectivity. Several
downstream targets of FRU
M
at the adult stage have been functionally characterized,
including takeout, neuropeptide F, yellow, and defective proboscis extension response
(Dauwalder et al., 2002; Drapeau et al., 2003; Goldman and Arbeitman, 2007; Lee et al.,
2006), consistent with FRU
M
regulating diverse processes in which particular sub-sets of
cells are involved. However, these genes were all shown to be regulated by fru P1 at the
adult stage; little is known about the transcriptional changes regulated by FRU
M
earlier in
development.

During metamorphosis, the developmental period during which the larval nervous system
is remodeled for adult functions, the CNS undergoes a massive reorganization during
which subsets of neurons undergo apoptosis, adult-specific neurons are born, and larval-
specific connections recede and are replaced with adult-specific aborizations (reviewed in
Truman, 1990). Developmental expression studies and molecular genetic studies have
suggested that the primary action of FRU
M
occurs during metamorphosis. Temperature
shift experiments in which the fru P1 transcript is forced into the female splice form
using a tra-2 temperature sensitive allele show that metamorphosis is a critical period for
patterning male courtship behavior (Belote and Baker, 1987); furthermore, FRU
M

expression is highest at the mid-pupal stage (Lee et al., 2000a).
78

Metamorphosis is triggered by release of the steroid hormone ecdysone. The functional
ecdysone receptor is a heterodimer of two nuclear receptors EcR and Ultraspiricle (USP),
the Drosophila RXR homologue (Koelle et al., 1991; Yao et al., 1993). EcR, the subunit
that binds ecdysone, encodes three isoforms (EcRA, EcRB1, and EcRB2; see Figure 1B
for schematic), which share common DNA- and hormone-binding domains, but differ in
their amino-terminal regions (Talbot et al., 1993). EcRA and EcRB1, the two most
studied isoforms, have distinct expression patterns (Talbot et al., 1993; Truman et al.,
1994) and have been shown to be functionally distinct based on differences in
transcription factor activities and abilities to regulate downstream target genes (Bender et
al., 1997; Mouillet et al., 2001; Schubiger et al., 2003). Recent studies on a hypomorphic
EcR allele that contains a mutation that affects all isoforms demonstrated role for EcR in
establishing the potential for male courtship behaviors (Ganter et al., 2007). However, the
role of EcR in the fru P1 circuit is not known.

Here, genes are identified that are expressed downstream of FRU
M
, at the 48-hours after
puparium formation (APF) stage, using a microarray approach. Additionally, genes that
are sex-differentially expressed in the CNS were identified. A significant number of the
FRU
M
-regulated genes have previously been shown to be regulated in response to
ecdysone, suggesting FRU
M
and EcR function to regulate similar downstream targets.
Indeed, FRU
M
and EcR show overlapping expression patterns in the CNS during
metamorphosis, with each EcR isoform showing different patterns of overlap,
79

suggestions that EcR isoforms have distinct functions in the fru P1 circuit. Additionally,
reducing the levels of EcRA in the fru P1 neural circuitry leads to significant increases in
the levels of male-male courtship, while reducing EcRB1 had no significant effect on
courtship. To extend this investigation, microarray experiments examining CNS tissue
that specifically reduced EcRA in FRU
M
-expressing neurons at three distinct time points,
were performed and genes were identified that likely mediate the formation of
appropriate male-specific neural circuitry downstream of FRU
M
and EcRA. To extend
these analyses, Broad, a transcription factor known to respond directly to EcR, is shown
to co-localize with FRU
M
at three discrete times during development.

3.3. Results
To identify genes that function downstream of FRU
M
during metamorphosis, a
microarray-based approach was undertaken using arrays that contain elements that detect
all known and predicted Drosophila genes (Genome Release 4.1; see Chapter 1).
Transcript abundance differences between fru P1 mutant males and wild type males were
examined in whole-body pupae and CNS tissues at the 48-hour APF stage. This period
has been shown to have peaks of FRU
M
protein and fru P1 mRNA expression levels (Lee
et al., 2000a). Two fru P1 mutant genotypes (fru
440/P14
and w; fru
w12/ChaM5
) were each
compared to wild type males (Canton S and w Berlin, respectively; see Table 3.1). The
microarray data from these two comparisons were analyzed together to eliminate genes
that show expression differences due to strain variation.

80

Table 3.1 Microarray experimental design and numbers of genes identified
Probe Pairs Rationale Number of
Genes

fru P1 mutant
male vs. wild type
male

Identify genes expressed downstream of
FRU
M
in mid-pupal male whole body

236
fru P1 mutant
male CNS vs.
wild type male
CNS

Identify genes expressed downstream of
FRU
M
in mid-pupal male CNS
94
Wild type male
CNS vs. female
CNS
Identify genes with sex-differential
expression in the CNS in mid-pupae
104

From the whole body pupae analyses, 236 genes that show a significant difference
between wild type males and fru P1 mutant males were identified (q<0.15; Table 3.2 lists
genes with a fold-difference >2). In these experiments, P-values were converted to q-
values, which are a measure of the false discovery rate (see Materials and Methods;
(Storey, 2003)). Of those 236 genes, 122 and 114 show higher transcript abundance in
wild type males and fru P1 mutant males, respectively. This gene set is referred to as the
whole-body FRU
M
-regulated set.

81

Table 3.2 Genes identified by microarray experiment in the whole body and dissected CNS
experiments as being differentially expressed as a result of fru P1 (q≤0.15, and Fold Change ≥2)

Gene Name Symbol CG # Chr FC Functional Annotation
Whole
Immune induced molecule 23 IM23 CG15066 2R 4.43 defense response
Body
Glutathione S transferase D3 GstD3 CG4381 3R 3.90 glutathione transferase activity

kinase suppressor of ras ksr CG2899 3R 2.98 protein kinase activity

CG8997 CG8997 CG8997 2L 2.87 unknown

GST-containing FLYWCH zinc-finger
protein
gfzf CG33546 3R 2.64 glutathione transferase activity

CG10814 CG10824 CG10824 2R 2.64 vitamin biosynthesis

drosomycin-6

dro6 CG32268 3L 8.55 defense response

Cuticular protein 72Eb Cpr72Eb CG12255 3L 3.53 structural constituent of cuticle

CG17352 CG17352 CG17352 3L 3.53 unknown

CG7738 CG7738 CG7738 3.42 unknown

CG2233 CG2233 CG2233 X 3.01 unknown

CG30080 CG30080 CG30080 2R 2.91 transcription factor activity

CG11350 CG11350 CG11350 3L 2.76 unknown

Ecdysone-dependent gene 78E Edg78E CG7673 3L 2.65 structural constituent of cuticle

Odorant-binding protein 56a Obp56a CG11797 2R 2.60 odorant binding

CG1702 CG1702 CG1702 X 2.59 glutathione transferase activity

pancreatic eIF-2alpha kinase PEK CG2087 3R 2.58 protein kinase activity

CG41136 CG41136 CG41136 2.57 unknown

Eig71Ej Eig71Ej CG7588 3L 2.56 autophagic cell death

CG16712 CG16712 CG16712 2L 2.56 serine-type endopeptidase
inhibitor

CG12231 CG12231 CG12231 X 2.53 unknown

Dissected
CG2177 CG2177 CG2177 4 3.16 unknown
CNS
U snoRNA host gene 2 Uhg2 CR32873 2L 2.79 unknown

CG30447 CG30447 CG30447 2R 2.57 unknown

CG15281 CG15281 CG15281 2.57 unknown

Cuticular protein 47Ee Cpr47Ee CG13222 2R 4.48 structural constituent of cuticle

Ecdysone-induced gene 71Ef Eig71Ef CG7599 3L 3.97 autophagic cell death

CG13841 CG13841 CG13841 3R 3.34 Unknown

Ecdysone-induced gene 71Eg Eig71Eg CG7336 3L 3.15 unknown

CG33468 CG33468 CG33468 2R 3.15 unknown

CG8564 CG8564 CG8564 3L 3.09 proteolysis

diadenosine tetraphosphate hydrolase Apf CG31713 2L 3.03 pyrophosphotase activity

Lysozyme X LysX CG9120 3L 2.75 lysozyme activity

CG30339 CG30339 CG30339 2R 2.67 unknown

White indicates an expression value that is higher in fru P1 null males (a transcript which is repressed by
fru P1) and gray indicates an expression value that is higher in wild type males (a transcript which is
activated by fru P1).

82

To characterize the 236 genes in the whole-body FRU
M
-regulated set, functional analyses
were carried out that identified overrepresented functional categories by using both the
program DAVID (Database for Annotation, Visualization and Integrated Discovery;
(Dennis et al., 2003) and our own analysis program (for details see Materials and
Methods).

The FRU
M
-regulated set is significantly overrepresented with three proteins containing
the Calycin domain (P=0.008), which is present in proteins in other organisms involved
in diverse processes including immune response, cell adhesion, ligand binding, receptor
binding and the formation of complexes with other macromolecules (reviewed in Flower
et al., 2000). The three Calycin Domain-containing proteins found here are CG6783,
Neural lazarillo (NLaz), and the unannotated Karl. Both NLaz and CG6783 have been
shown to be expressed in the developing nervous system and to be circadian-regulated
(Ceriani et al., 2002; Kearney et al., 2004; Sanchez et al., 2000). This suggests a
functional integration of FRU
M
neurons and a circadian clock, which is not unexpected,
given that male courtship behavior is known to be circadian regulated (Kadener et al.,
2006).

A group of seven genes that is also significantly overrepresented (P=0.009) in the FRU
M
-
regulated set are annotated as insect cuticle proteins. Two of these proteins (CG10112
and Ecdysone-dependent gene 78E) are also regulated in neurons as a consequence of the
transcription factor escargot (Hekmat-Scafe et al., 2005). This gene set may function in
83

the production and maintenance of the sex-specific pheromones which act as cuticular
hydrocarbons, which are expressed on the surface of the fly.

The FRU
M
-regulated set also contains two of the three Drosophila genes annotated with
the ninjurin domain (NijA and CG14394) ; in mammals, proteins with the ninjurin
domain are up-regulated in response to nerve injury and are known to function in cell-
adhesion (P=0.05) (Araki and Milbrandt, 1996). A role for ninjurin proteins in cell-
adhesion has also been shown in Drosophila (Zhang et al., 2006). These results suggest
that ninjurin-related proteins may play a role in the proper development of FRU
M
-
expressing neurons.

Other interesting FRU
M
-regulated genes were observed when genes were examined
individually. Odorant-binding protein 56a (Obp56a) is positively regulated as a
consequence of FRU
M
. This gene has also previously been shown to have higher
expression in female lines that mate more slowly (Mackay et al., 2005). Another gene
that may function downstream of FRU
M
is astray (aay), which has been identified in
several high-throughput screens as functioning in the peripheral nervous system
(Prokopenko et al., 2000). Additionally, aay is differentially expressed between male
flies that are fast to copulate and male flies that are slow to copulate (Mackay et al.,
2005); these results suggest that both Obp56a and aay may play a role in specifying male
courtship behavior downstream of FRU
M
.

84

When chromosomal location of genes in the whole-body FRU
M
-regulated set was
examined, three groups of 5-6 genes were identified that are located within 250kb of one
another on three different chromosomes. All three chromosomal regions were
significantly overrepresented with genes from the FRU
M
-regulated set (hypergeometric
test, p<0.05). FRU
M
may bind to enhancers in these regions and may coordinately
regulate the expression of nearby genes. Many of the genes in each of these regions are
have no annotated functions and from the genes which are annotated, no clear overall
function could be ascertained for any of the regions. It should also be noted that the third
gene region, of which 5 of the 18 genes are in the whole-body FRU
M
-regulated set, is
located upstream of the fru locus itself (and that both the fru
440
and fru
ChaM5
alleles are
large deletions overlapping with this region (Anand et al., 2001; Gailey and Hall, 1989).
Thus, these five genes’ expression changes could be due to the large deletions and
inversion used to generate the fru P1mutant flies and not due to regulation by FRU
M
.

Expression differences in the CNS between fru P1 mutant males and wild type males
Microarray analyses were also carried out comparing transcript abundance differences
between fru P1 mutant male and wild type male CNS tissues. Seventy-one and 23 genes
were more highly expressed in fru P1 mutants and wild type males, respectively (q<0.15;
Table 3.2, called FRU
M
CNS set). At 48 hours APF, substantially more genes are down-
regulated downstream of FRU
M
in the CNS

than up-regulated (71 and 23, respectively).
However, when we examine transcript abundance differences in the entire pupae, a
similar number of genes are down- and up-regulated downstream of FRU
M
(114 and122,
85

respectively). No overrepresented Gene Ontology functional groups were found in the
CNS set using the program DAVID, likely because 55 of the 94 genes in this set have no
assigned annotation or function.

The FRU
M
CNS set contained many genes that are expressed in the CNS, including
Resistant to dieldrin (Rdl), capability (capa), snapin, and CG10617 (Kearney et al.,
2004). Interestingly, both Rdl and capa were identified as being regulated by FRU
M
in
the adult CNS (Goldman and Arbeitman, 2007). Rdl encodes a putative GABA receptor
that is expressed in many regions of the brain known to be important for sensory input,
including the mushroom bodies and the central complex (Ffrench-Constant et al., 1991;
Harrison et al., 1996); GABA is an inhibitory neuropeptide. The CAPA peptide is
cleaved to produce three neuropeptides, one of which belongs to the Drm-PK-1 family;
members of this family are known to be involved in sex pheromone production (reviewed
in Altstein, 2004). Both snapin and CG10617 are annotated as functioning in synaptic
vesicles. CG8709 was found to be expressed in the developing CNS during
embryogenesis, although its function is unknown (Kearney et al., 2004).

Comparison of genes regulated downstream of FRU
M
in pupae and adults
A recent microarray studied identified 90 and 26 genes regulated downstream of FRU
M
in
adult head and CNS tissues, respectively (Goldman and Arbeitman, 2007). Twenty-four
genes were found to be regulated by FRU
M
in both pupae and adults, 23 of which show
the same direction of regulation at both developmental stages. Seven of these genes or
86

their protein products were shown to be present in the hemolymph, expressed in
hemocytes, or annotated as secreted proteins (Asha et al., 2003; Karlsson et al., 2004)
(UniProt Consortium, 2008), and six are annotated as responding to a stimulus (The Gene
Ontology Consortium, 2000). This suggests FRU
M
has a role in regulating gene
expression outside of the CNS, particularly in the peripheral nervous system and the fat
body. FRU
M
regulates the gene takeout in the fat body (Dauwalder et al., 2002), and has
been shown to regulate a larger set of genes in the adult head (Goldman and Arbeitman,
2007).

Transcriptional differences between male and female CNS tissue are predominantly
male-biased
To determine the transcript abundance differences in male and female CNS tissues,
additional microarray experiments were performed that compared transcript abundance in
CNS tissues of wild type (CS) males and females at 48-hours APF. One hundred and four
sex-differentially expressed genes were identified (q-value<0.15; Table 3.3).
Surprisingly, 97 genes were more abundant in the male CNS, whereas only seven genes
were more abundant in the female CNS. Fifteen of the 94 FRU
M
-regulated CNS genes
overlapped with the wild type sex-biased CNS gene set, 13 of which show higher
expression in wild type males than in wild type females and also higher expression in fru
P1 mutants than in wild type males. This suggests that most genes in the CNS are sex-
differentially expressed independent of fru P1 regulation, perhaps by dsx or another level
of the sex hierarchy (see Figure 1.2). For the fifteen that are sex-differentially expressed
87

and downstream of fru P1, the direction of change is opposite one would expect if fru P1
was specifying the sex-biased expression pattern in males. This suggests that the bulk of
these genes have multiple regulatory mechanisms: one that establishes sex-differential
expression and then in males, additional regulation downstream of fru P1, consistent with
our previous observations (Goldman and Arbeitman, 2007).

Table 3.3 Genes that are differentially expressed in male and female CNS tissue

88

Table 3.3, continued

89

Table 3.3, continued

DAVID analysis showed that ten of these male-biased genes are expressed in the
mitochondria. Mitochondria are required at synapses for proper utilization of a reserve
pool of vesicles (Verstreken et al., 2005) and are involved in neurodegeneration of the
adult CNS, which occurs in Drosophila models of Parkinson’s disease, Alzheimer’s
disease, and aging (reviewed in Bossy-Wetzel et al., 2003; Dodson and Guo, 2007).
During metamorphosis, but not in adults, electron-dense mitochondria (EDMIT) are seen
(Singh and Singh, 1999); EDMITs are thought to be dying mitochondria and this suggests
a possible turnover of larval and pupal mitochondria. Understanding how mitochondria
might play a role in sex-specific neural development is worth further consideration, since
90

increased energy production in the CNS of one sex may be a sign of sex-specific neural
development.

Ninety-two of the male-biased genes were included in a microarray study examining gene
expression in various tissues in the fly (Chintapalli et al., 2007); of these 92, 77 were
identified as having high expression in the male testes. This is not thought to be due to
contamination of the male testes in the dissected CNS samples; 70 of these 77 genes were
also identified as being present in the larval fat body in at least three of four arrays in that
study. This suggests that the majority of the male-biased genes identified in the dissected
CNS set are likely due to sex-biased expression in the fat body. Previous reports have
identified sex-differential gene expression in the fat body (Dauwalder et al., 2002), as
well as many genes that are expressed in both in the testes and the fat body (Jiang et al.,
2005).

The majority of genes with high levels of sex-differential expression in the CNS
identified above appear to be due to expression in the larval fat body. To identify more
genes with sex-differential expression due to CNS expression, less stringent criteria
(P<0.05) were used. This led to the identification of an additional 169 male-biased genes
and 149 female-biased genes. Along with dsx and Sxl, this set is over-represented with
four additional genes that are involved in developmental growth (ultrabithorax, merlin,
notch, and eukaryotic initiation factor 4a); these genes may be involved in regulating
sex-specific development of the CNS. In addition, this set is over-represented with four
91

genes that function in hormone binding and that are all members of the takeout gene
family; takeout has been previously shown to be sex-differentially expressed and
regulated downstream of the sex-determination hierarchy (Dauwalder et al., 2002).
Interestingly, this set also includes 20 genes annotated as having alternative splice forms;
this suggests a possible extension to the mode of sex-specific regulation used in the sex-
determination hierarchy.

Examining the list of sex-differentially expressed genes in the CNS at 48 hour APF also
identified one neuropeptide (Neuropeptide-like precursor 1) and seven genes encoding
for sensory or neuropeptide receptors, including three Odorant-binding proteins (Odp18a,
56a, and 57d), Octbeta3R, nicotinic Acetylcholine Receptor alpha 96Ab, Allatostatin
Receptor, and CG33639. Differential expression of hormones and receptors may lead to
sex-specific responses to sensory and hormonal cues necessary for adult behaviors. Also
sex-differentially expressed at this stage are six genes whose protein products are
involved in neuroblast proliferation of nervous system development and seven involved
in transmission of nerve impulse or neurotransmitter secretion. Synaptic activity is
known to regulate axonal connections (reviewed in Zhang and Poo, 2001), and these
results may indicate a role for synaptic activity in sex-specific development of the CNS.

92

Ecdysone hierarchy genes are regulated by fru P1 during metamorphosis
The FRU
M
pupal and CNS gets sets contain several genes that are annotated as
functioning in the ecdysone regulatory pathway (Table 3.4).

Table 3.4 Genes that are regulated by ecdysone that are also regulated by fru P1 in whole body and
dissected CNS tissue. FC is the fold change.

Whole-body pupae microarray
Gene Name Gene
Symbol
Genotype with higher
expression
FC
Cytochrome P450-18a1 Cyp18a1 fru P1 male 1.82
Nedd2-like caspase Nc fru P1 male 1.29
Ecdysone-dependent gene 78E Edg78E Wild type male 2.65
Eig71Ej Eig71Ej Wild type male 2.56
Ecdysone-inducible gene L2 ImpL2 Wild type male 1.45

Dissected mid-pupal CNS microarray

Gene Name Gene
Symbol
Genotype with higher
expression
FC
Inhibitor of apoptosis 2 Iap2 fru P1 male 1.80
Ecdysone-induced gene 71Ef Eig71Ef Wild type male 3.97
Ecdysone-induced gene 71Eg Eig71Eg Wild type male 3.15

To determine if ecdysone regulated genes are significantly overrepresented in the FRU
M

sets, a list of genes that were previously characterized as functioning in the ecdysone
pathway was compiled (61 genes). The ecdysone-regulated genes are significantly over-
represented in both the FRU
M
-regulated pupal and CNS gene sets (P≤0.0006,
hypergeometric test, see materials and methods). In additional experiments where FRU
M

proteins were overexpressed, ecdysone-regulated genes are also overrepresented
(P≤0.0002, hypergeometric test, unpublished data ML, LS and MA). This suggests that
the ecdysone regulatory hierarchy may function in coordination with FRU
M
to establish
the neural circuitry that underlies male courtship behaviors.
93

Identification of EcR binding sites in fru P1-regulated genes
To determine if EcR and FRU
M
have similar downstream targets, the presence of EcR
binding sites was determined for the genes regulated by fru P1. Using the imperfect
palindrome EcR consensus sequence (Antoniewski et al., 1993) and the motif/cis-
Regulatory Module discovery tool SUPRfly (personal communication, Matt Lebo, Tom
Goldman, Michelle Arbeitman, and Fengzhu Sun, unpublished data), 79/223 genes and
28/90 genes from whole animal and CNS FRU
M
sets, respectively, have at least one EcR
binding site present in the genes’ regulatory regions. The 107 genes with an EcR binding
site is a significant over-representation compared to all genes in the whole Drosophila
genome that have at least one identified EcR binding site (hypergeometric test, P=
0.021). This suggests that the fru P1-regulated gene sets may be enriched for genes that
respond to ecdysone directly through EcR.

EcRA and EcRB1 are present in the fru P1 neural circuitry at the beginning of
metamorphosis
To determine if EcR is functioning in the FRU
M
-circuit, co-localization of either EcRA or
EcRB1 with FRU
M
was examined. Using antibodies specific for EcRA and EcRB1
(Talbot et al., 1993), EcR expression patterns and levels were examined at three stages of
development (wpp, 48 hour APF, and 0-24 hour adult). Our results for the expression
patterns of the EcR isoforms are largely consistent with previous reports, with high levels
of EcRB1 detected at the wpp stage, but then very low levels detected at the 48 hours
94

APF stage, and none detected in the adult, while EcRA expression is high at all three
stages examined (Figure 3.1) (Truman et al., 1994).

95

Figure 3.1. EcR encodes three isoforms which have different spatial and temporal expression
patterns.

(A) Diagram of Drosophila sex determination hierarchy, which results in the production of sex-specific
transcription factors encoded by doublsex (dsx) and fruitless (fru). For a detailed description, see Figure
1.2. (B) Schematic of EcR isoforms adapted from (Talbot et al., 1993). Region 1 is the isoform-specific
region, Region 2 is the shared DNA-binding domain, and Region 3 is the ligand-binding domain. Black
boxes represent regions against which RNAi transgenes were targeted. EcR Common RNAi’s were
directed against the ligand-binding domain of EcR (Colombani et al., 2005); EcRA-specific RNAi was
directed against EcRA specific exon, and the EcRB1-specific RNAi was directed against B1-specific exon
(Roignant et al., 2003). Schematic is not to scale. Asterisk represents the location of mutated amino acids
in the DN form of EcR proteins (aa 645 and 650) (Cherbas et al., 2003). (C-V) EcRA isoforms show
different expression patterns at distinct stages in the male CNS. EcRA (left panels) and EcRB1 (right
panels) antibody staining (green) are shown for white pre-pupal larvae (C-F), 48 hour APF pupae (G-K,
O,Q,S-V) and adult (K-N, P,R). For all images, anterior is denoted by A and posterior by P at the bottom
of the image. 20X confocal sections (~1 µm thick) are shown. (A-F), and 40X confocal sections of ~0.5
µm are shown (C-R). 40X confocal sections of ~0.5 µm are shown (S-V) for 48 hour brain and VNC
regions to illustrate the presence of cells at a higher magnification.

96

Figure 3.1

97

When co-localization between EcR isoforms and FRU
M
was examined, both EcR
isoforms are present in FRU
M
-expressing cells in the anterior brain at the wpp stage
(Figure 3.2). However, by 48 hours APF no co-localization is observed between EcRB1
and FRU
M
; EcRB1 shows very low expression in the CNS generally at this stage. In
contrast, nearly all FRU
M
-expressing cells also express EcRA at both 48 hours APF and
the 0-24 hour adult stage (Figure 3.2). The expression patterns of EcRA and FRU
M
do
not completely overlap; EcRA and FRU
M
are present in cells in which the other is not.
For example, cells in the abdominal ganglion and the dorsal region of the VNC show
FRU
M
but not EcRA staining (Figure 3.2A and F). In addition, cells with very high
levels of EcRA staining, presumably Type II neurons, which undergo programmed cell
death early in adulthood (Robinow et al., 1993), largely do not overlap with FRU
M
,
although a small number of Type II neurons were observed in the abdominal ganglion
and the pro-thoracic ganglion of the VNC that also expressed FRU
M
(Figure 3.2O and
W). Given these different patterns of overlap of EcRB1 and EcRA with FRU
M
, if EcR
plays a role in remodeling the cells in the fru P1 circuit for adult functions, then the
isoforms likely have distinct roles.

98

Figure 3.2. FRU
M
and EcRA or EcRB1 are co-expressed in sub-sets of cells at distinct stages.
For all panels, FRU
M
expression is shown in red; yellow is co-localization. Panels A-B show EcRB1 in
green during the wpp stage. Whole wpp CNS (A) and higher magnification of a brain lobe (B) show co-
localization of FRU
M
and EcRB1. Whole wpp CNS (C) and higher magnification of a brain lobe (D) show
co-localization of FRU
M
and EcRA. FRU
M
and EcRA (green) co-localize in 48 hour APF pupal brain (E)
and VNC (F). Inset of (F) shows higher magnification of FRU
M
-expressing cells in the abdominal ganglion
of (F). FRU
M
(G and J) and EcRA (H and K) colocalize (I and L) in subsets of cells in the protocerebrum
(G-J) and anterior medial brain (J-L). FRU
M
(M) and EcRA (N) colocalize (O) in a Type II neuron in the
pro-thoracic ganglion of the VNC. FRU
M
(red) and EcRA (green) co-localize (yellow) in a 0-24 hour adult
brain (P) and VNC (Q). FRU
M
(R) and EcRA (S) co-localize (T) in the anterior medial brain. Type II
neurons express high levels of EcRA (U) and FRU
M
(V); co-localization is denoted by arrowhead in (W).
Note that some Type II neurons do not express FRU
M
(asterisk in U and W). Some FRU
M
-expressing cells
in the dorsal VNC do not express EcRA (red cells in X). 20X confocal sections (~1 µm thick) are shown
(A,C,E,F,P,Q) and 40X confocal sections of ~0.5 µm are shown (B,D,G-O,R-X).

99

Figure 3.2

100

Figure 3.2, continued

EcR function is not required in the fru P1 neural circuit for male courtship behavior
towards a female
EcR function was abrograted in the fru P1 neural circuit to determine if EcR is necessary
for wild type male courtship behavior directed towards a female. Both the courtship
index (CI) and wing extension index (WEI) were used to measure courtship activity (see
materials and methods). To abrograte EcR function in the fru P1 circuit either EcR RNAi
101

or EcR dominant negative (DN) transgenes were expressed via activation of the UAS
element by fru P1-GAL4 (Manoli et al., 2005). The RNAi lines target common (RNAi.1
and RNAi.2; Colombani, 2005 #656}) and isoform-specific regions of EcR (EcRA RNAi
or EcRB1 RNAi; (Roignant et al., 2003)). The dominant negative EcR transgenes each
carry a point mutation that affects the ligand binding domain of EcR; the products are
thought to dimerize with USP, but do not allow for ligand-dependent gene expression,
thus acting as isoform-non-specific dominant negative effectors (Cherbas et al., 2003)
(see Figure 3.1B for regions of EcR targeted by RNAi lines and region of point
mutation).

No significant difference in male-female courtship was observed between wild type and
single transgene control males as compared to males with abrogated EcR function in the
fru P1 circuit (Figure 3.3, left). In some cases the experimental flies showed significantly
lower courtship as compared to wild type males, but not compared to single transgene
EcR RNAi or EcR DN males, suggesting that the transgenes might have low-level
expression independent of the fru P1-GAL4 driver.

102

Figure 3.3. Males with reduced EcR levels display high levels of male-male courtship behavior.
Courtship Indices (light bars) and Wing Extention Indices (dark bars) are shown for male-female courtship
behavior (left panels) and male-male courtship behavior (right panels). Males carrying common EcR RNAi
transgenes are shown in (A), males carrying dominant negative EcR transgenes are shown in (B), and
males carrying isoform-specific RNAi transgenes are shown in (C). Stars indicate statistical significance
over all three controls (CS Heb, fru P1 single transgene, and respective EcR single transgene) at P<0.05.
#
in 3.3B, left panel, indicates significance over the fru P1 single transgene and CS Heb, but not over the
respective EcR single transgene (for comparison between fru P1-GAL4/EcRB1.1 DN and EcRB1.1 DN/+,
P=0.06, which is just above the statistical threshold used here to denote significance)

Abrogation of EcR function in the fru P1 neural circuit leads to increased male-male
courtship behavior
Some hypomorphic combinations of fru P1 mutant alleles cause males to fail to
discriminate between males and females as appropriate mate choices and as a result, these
103

fru P1 mutant males will court other males. A previous report examined male-male
courtship using a temperature sensitive EcR allele carrying a mutation in the ligand
binding domain of EcR, thus affecting all three isoforms in a spatially unrestricted
manner (Ganter et al., 2007). When these EcR mutant flies were raised at the non-
permissive temperature, they demonstrated elevated levels of male-male courtship
behavior (Ganter et al., 2007).

To determine if EcR activity specifically in the fru P1 circuit is required for mate
discrimination, male-male courtship behavior was assayed in males that have abrogated
EcR in the fru P1 circuit. fru P1-GAL4; UAS-EcR RNAi.1 males displayed significantly
higher male-male CI as compared to all controls (Figure 3.3A, right; all P≤0. 0.00012). In
addition, these experimental flies showed high levels of attempted copulations, with 75%
of flies attempting copulation. In contrast, no control males attempted copulation. fru
P1-GAL4 EcR RNAi.2 males also showed significantly higher male-male courtship than
wild type controls and controls carrying the EcR RNAi.2 transgene alone (P=0.015 and
0.00015, respectively), but not compared to controls carrying fru P1-GAL4 alone
(P=0.061). fru P1-GAL4; EcR RNAi.2 did not show high male-female courtship and so it
is expected that courtship towards males also would not be as high as fru P1-GAL4; UAS-
EcR RNAi.1 males.

Males bearing one of four different EcR DN transgenes driven by fru P1-GAL4 also
showed higher than control male-male courtship activity (Fig 3.3B). All DN
104

experiemental groups showed significantly higher CI than the three control groups
(Canton S (wild type), fru P1-GAL4 single transgene, and the respective DN single
transgene), with the exception of the EcRB1.1 DN experimental group, which displayed
significantly higher male-male courtship behavior than the CS Heb control group (P=7.6
x 10-5) and the fru P1-GAL4 single transgene control group (P=4.0 x 10-4), but was not
significantly higher than the group carrying the single UAS-EcRB1.1 DN transgene alone,
although the P-value was just outside of the significance test used here (P=0.06).
Surprisingly, fru P1-GAL4; UAS-EcRB1.2 DN showed a male-male CI of 0.64, a level
approaching wild type male-female CIs, which under these experimental conditions, is
approximately 0.7. Numerous copulation attempts were also seen when each EcR DN
transgene was expressed in the fru P1 circuit (Figure 3.3B). Over 40% of experimental
flies attempted copulation at least once for each EcR DN transgene; in contrast, most
control lines showed no attempted copulation.

Abrogating EcRA, but not EcRB1, in fru P1 circuit causes male-male courtship
behavior
To determine if there are EcR isoform-specific roles in the fru P1 circuit, isoform-
specific UAS-RNAi transgenes that specifically abrogate either EcRA or EcRB1 levels
(Roignant et al., 2003) were used to assess functional consequences on male-male
courtship behavior. fru P1-GAL4;EcRA RNAi males showed significantly increased
male-male courtship and wing-extension as compared to control males (Figure 3.3C,
right). In contrast, fru P1-GAL4; EcRB1 RNAi males did not show significantly higher
105

male-male courtship behavior than control males carrying only the fru P1-GAL4 single
transgene (P=0.63), although male-male courtship was higher than both wild type males
and EcRB1 RNAi single transgene males (P<0.023 for both). fruP1-GAL4;EcRA RNAi
males also showed high levels of attempted copulations toward male flies (63% of
experimental flies attempted copulation at least once), while 33% of fruP1-GAL4;EcRB1
RNAi males attempted copulation and 0% or 10% of males attempted copulation for all
controls. These behavioral data together with our immunohistochemistry results suggests
that the EcRA isoform is the predominant isoform functioning in the fru P1 circuit.

To determine if male-male courtship behavior by fru P1-GAL4;EcRA RNAi males was
because they could no longer discriminate between males and females, a preference assay
was carried out in which mutant males were presented with both a male of their same
genotype and a virgin female. All male flies tested, including fru P1-GAL4;EcRA RNAi
males, overwhelmingly preferred females (experimental genotype and controls all
showed a preference index >95%, Figure 3.4A, which suggests that the ability to
discriminate between males and females remains intact. fru P1-GAL4;EcRA RNAi males
showed normal locomotor activity (Figure 3.4B), suggesting that the increase in male-
male courtship seen here is due to a specific courtship defect. These results demonstrate
that the increased male-male courtship behavior observed here and in the previous study
(Ganter et al., 2007) are likely do to a deficit of EcRA specifically in the fru P1 neural
circuit.
106

Figure 3.4: Preference assays and activity levels of males carrying fru P1-GAL4/UAS-EcRA show
wild type level. The preference index is shown for fru P1-GAL4/UAS-EcRA RNAi (last column) and
controls in (A). The x-axis is the genotype of fly tested, and the y-axis is the preference index (for
description of calculations, see materials and methods in this chapter). For each genotype, n≥10. Activity
levels of flies (B) show that the fru P1-GAL4/UAS-EcRA flies (pink) show wild type levels of activity, as
assayed by line crossings. X-axis shows hours, with the black indicating when flies were kept in darkness.
Y axis shows number of line crossings.

Broad, a protein directly regulated by EcR, is expressed in a sub-set of FRU
M
-
expressing cells
broad (br) is the only one of the four early genes in the ecdysone regulatory hierarchy
that has been shown to encode a protein necessary for several aspects of CNS
development and remodeling during metamorphosis, including the correct fusion of left
and right brain regions, the positioning of the developing visual system, and the
separation of the subesophogeal ganglion from the thoracic ganglion (Restifo and White,
107

1991). br has been shown to be an “early” gene, which responds quickly to ecdysone
levels and initiates a transcriptional cascade (Chao and Guild, 1986) and encodes a family
of zinc-finger DNA binding proteins, in which four protein isoforms (BR-Z1, -Z2, -Z3
and -Z4) are produced from alternative splicing of the br pre-mRNA (Bayer et al., 1996;
DiBello et al., 1991). In our previous microarray analysis using cDNA microarrays, br
was identified as being regulated by FRU
M
(LS, ML and MA data not shown); however,
br was not present in the set of genes that differed in abundance between wild type males
and fru P1 null males using the oligonucleotide microarrays presented here. Both BR
and FRU
M
are members of the BTB-zinc finger family of transcription factors (Ryner et
al., 1996; Zollman et al., 1994); therefore, it is possible that BR and FRU
M
may be
heterodimeric partners (reviewed in Albagli et al., 1995). Previous reports show that all
isoforms of BR are present in the pupal CNS (Crossgrove et al., 1996; Emery et al.,
1994). As expected, extensive BR-Core, which is common to all BR isoforms,
expression is observed in the brain and VNC at the three time points examined, wpp, 48
hours APF, and 0-24 hour adult, when using an antibody generated against the common
portion of BR isoforms (Emery et al., 1994). To test if BR is present in the fru P1 neural
circuitry at these discrete stages, co-labeling experiments were performed using
antibodies against FRU
M
and BR-Core (Figure 3.5). Cells were identified that co-express
FRU
M
and BR at the wpp stage in both the brain lobes and the VNC (Figure 3.5A-C). By
48 hours APF, substantial overlap was present, including regions in the anterior mid-
brain and posterior brain, as well as the PrMs region of the VNC, all of which have been
implicated as regions critical for aspects of the male courtship ritual, as shown through
108

early mosaic studies (Hall, 1977); Figure 3.5D-J). This pattern of overlap continues
through the 0-24 hour adult stage (Figure 3.5K-Q). Thus, br may be a common target of
both EcR and FRU
M
, which act in conjunction to pattern the male-specific CNS.

109

Figure 3.5. FRU
M
and BR are co-expressed in the CNS at three distinct times during development.
For all panels, FRU
M
expression is shown in green, and BR expression is shown in red. FRU
M
and BR are
co-expressed in regions of the wpp brain (A). Arrowheads indicate regions in which FRU
M
and BR are co-
expressed. Higher magnification of wpp brain lobe (B) and VNC (C) are shown. FRU
M
and BR are co-
expressed in 48 hour APF anterior (D) and posterior (I) brain and VNC (H). Regions for which higher
magnification images are described in schematic. FRU
M
and BR are co-expressed in 0-24 hour adult
anterior (K) and posterior (P) brain and VNC (O). Regions for which higher magnification images are
described in schematic. 20X confocal sections (~1 µm thick) are shown (A,D,I,H,K,P,O) and 40X confocal
sections of ~0.5 µm are shown (B,C,E-H,J,L-N,Q). Images A-C were provided by Justin Dalton,
University of Southern California.
110

3.4. Discussion
Because targets of FRU
M
remain largely unknown, particularly during metamorphosis,
elucidating the mediators of FRU
M
-directed neural circuit formation has been difficult.
We identified many FRU
M
targets with diverse functions during the mid-pupal stage in
the body of males. When examining sex-differential expression in the CNS at the mid-
pupal stage, we unexpectedly find a much greater number of male-biased genes, possibly
underlying the importance of male-specific neural connectivity required for the potential
for male behaviors. However, we find little overlap between genes regulated downstream
of FRU
M
and sex-differentially expressed genes in the CNS, suggesting the presence of
secondary regulators, an observation also seen in previous studies (Goldman and
Arbeitman, 2007).

Twenty-four genes identified here as being regulated by FRU
M
in pupae were previously
shown to be regulated by FRU
M
adult heads, 23 of which are regulated in the same
manner at both time points (Goldman and Arbeitman, 2007). These genes may be
regulated outside of the CNS, particularly in the peripheral nervous system and in the fat
body. Additionally, FRU
M
has many more target genes in the CNS at the mid-pupal
stage than at the adult stage (Goldman and Arbeitman, 2007), suggesting that its major
function in the CNS is likely developmental. Additional support for a developmental role
of FRU
M
in neurons comes from the fact that the perturbation of EcRA levels in FRU
M
-
expressing neurons yields many transcriptional changes at the early stages of
metamorphosis (0 hour and 48 hour APF); however, by the time the adult stage is
111

reached, many fewer genes change in response to decreased EcRA levels in FRU
M
-
expressing neurons, which could be interpreted to mean that the major functions of FRU
M

and EcRA, as assessed by transcriptional changes, are largely completed by this stage.

As EcRA is prevalent in the fru P1 circuit, it may have a critical role in shaping the fine-
scale neural connectivity of FRU
M
-expressing neurons from an early point during
metamorphosis; this possibility is currently being tested. EcRA has been shown to be
required for neuronal arbor outgrowth (Brown et al., 2006). Additionally, EcR and BR
are required for proper sensory neuron development in the wing (Schubiger et al., 2005).
Flies showing high levels of male-male courtship may lack the ability to detect necessary
inhibitory cues from the male (e.g., cVA and Z-7-tricosene (Kurtovic et al., 2007;
Lacaille et al., 2007)), which could be explained by a neural developmental defect.

The ability of a hormone receptor to specify and influence courtship behavior is well
documented in diverse species (Bodo, 2008; Fusani, 2008), including Drosophila (Finley
et al., 1997). The abrogation of EcRA, one particular isoform of a steroid hormone
receptor, in just the FRU
M
-expressing neurons results in a robust, aberrant behavioral
phenotype. The interconnectivity of developmental programs, and in particular the sex
determination cascade and the ecdysone regulatory cascade, emphasizes the complex and
precise nature of development and the formation of sex-specific neural circuits.

112

3.5. Experimental Procedures
Fly collections and strains
Wild type flies were Canton S (CS) and w Berlin, which is a wild type strain carrying a
mutation in the white gene. fru P1 loss of function genotypes are transheterozygous for
the fru alleles fru
P14
and fru
4-440
, and fru
w12a
and w; fru
ChaM5
. The fru P1-GAL4 driver
used for behavioral analyses was previously described (Manoli et al., 2005), as was the
fru P1-GAL4 driver used in the EcRA RNAi microarray experiments (Stockinger et al.,
2005). EcR Common RNAi lines (stock numbers 9326, referred to here as RNAi.1 and
9327, referred to here as RNAi.2; (Colombani et al., 2005)), EcR Dominant Negative
lines (stock numbers 9451,6869,6872, and 9450, (Cherbas et al., 2003)), and isoform-
specific RNAi lines (9328 and 9329, (Roignant et al., 2003)) were obtained from the
Bloomington Stock Center, and outcrossed for five generations into a w CS background
(obtained from the Heberlein lab).

All flies were kept in a 12:12 hour light-dark cycle and grown using standard food media.
For FRU
M
and wild type microarray experiments, flies were collected at the white pre-
pupal (wpp) stage between Zeitgeber time (ZT) 1 and 4, aged at 25
◦
C for 48 hours, and
then either snap frozen in liquid nitrogen (for whole body experiments) or immediately
dissected (for CNS experiments.

RNA isolation and amplification and Cy-labeling for microarray experiments
113

RNA isolation, cDNA reverse transcription, and Cy-dye labeling for FRU
M
whole body
microarray experiments were conducted as in Chapter 2. Four independent samples
comparing CS and fru
P14/4-40
and four independent samples comparing w Berlin and w;
fru
w12a/ChaM5
were used, with a dye-swap design for both comparisons. For the FRU
M

CNS microarray experiments, three microarrays of independent biological samples
compared Cy-5-labeled cRNA from CS males and Cy-3-labeled cRNA from fru
P14/4-40

males and three microarrays of independent biological samples compared Cy-3-labeled
cRNA from w Berlin males and Cy-5-labeled cRNA from w; fru
w12a/ChaM5
males.

Microarray hybridization, production, and analyses
Microarray hybridization, production, and analyses were conducted as described in
Chapter 2. Microarray data presented in Chapter 3 can be accessed at NIH GEO database
at the following web link:
http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=rlyfdaemawciwvw&acc=GSE117
52.

Analysis of over-represented features
For all of the below statistical analyses, significance was declared if P<0.05. Significant
over-representations of functional annotations were generated with the program DAVID
(Dennis et al., 2003) and our own program as described in Chapter 2.

114

Identification of ecdysone-regulated genes
A gene was included in the ecdysone-regulated set if its protein or mRNA levels were
shown in a primary reference to respond to the presence of ecdysone in Drosophila cell
lines, tissues, or whole animals. Sixty-one genes were thus identified; references are
listed in Table B1. This list does not include genes identified in microarray studies.
Hypergeometric tests were used to determine the significance of over-representation of
ecdysone-regulated genes in the FRU
M
-regulated sets as compared to all genes present in
our microarray study.

Analysis of EcR binding sites
For the discovery of EcR binding sites, an 11-nt sequence was identified as a binding site
if it matched the known EcR consensus sequence GKTSANTGMMY (Antoniewski et al.,
1993). To identify EcR binding sites, the bioinformatics tool SUPRfly (ML, Tom
Goldman, MA, and FS, unpublished data) was used. For each gene, the region searched
started from 10,000bp upstream of the transcription start site or the distance to the
neighboring upstream gene, whichever was shorter. Similarly, the search region included
10000bp downstream from the transcriptional stop site or the distance to the neighboring
downstream gene. The search region also included all introns. Over-representation of
genes whose regulatory regions contained an EcR binding site was determined using a
hypergeometric test and the population of genes from the whole Drosophila genome that
contained an EcR binding site in their regulatory regions.

115

Immunohistochemistry
Whole mount immunochemistry experiments were performed as previously described
(Lee et al., 2000a). The primary polyclonal FRU
M
antibody was generated as previously
described (Lee et al., 2000a) and used at a 1:100 dilution in TNT (0.1M Tris-HCl, 0.3M
NaCl, 0.5% Triton X-100, pH 7.4). EcRA, EcRB1, and Broad-Core antibodies were
obtained from the Iowa hybridoma bank and used at a 1:10 dilution, except for Broad,
which was used at 1:100. Secondary antibodies were purchased from Molecular Probes
and used at 1:500 dilutions in PBS. Confocal microscopy was performed on a Zeiss
Pascal LSM5.

Courtship Assays
To eliminate strain-specific effects, all fly lines used in courtship assays were outcrossed
to a w CS stock for five generations (from Heberlein lab). Adult male flies were
collected at 0-24 hours post-eclosion under CO
2
anesthesia and raised individually in
vials and aged 4–7 days in a 12:12 hour light-dark cycle at 25 °C, with the exception of
flies carrying RNAi transgenes and their appropriate controls, which were raised at 29°C.
For courtship assays, a single male fly of each genotype was paired with a 4–7 day old w
CS virgin female or male in a 10-mm diameter chamber and video recorded for 10 min.
Courtship assays were performed at 25 °C, ~60% humidity, and at ZT 5–9. Assays for
control and experimental flies were performed on the same day. All courtship recordings were
analyzed using Noldus software (Wageningen, Netherlands). For these analyses, the
116

Courtship Index (CI) and the Wing Extension Index (WEI) measure the ratio of the time
the experimental male fly spent courting or performing a wing-song, respectively,
towards a w CS female or male to the total time of the observation. Preference assays
were performed by placing two males of the same genotype in a chamber with a 4–7 day
old w CS virgin female. Courtship was scored for the male that initiated courtship
behavior first. Preference Index was calculated as the total time spent courting the
female/total time spent courting.

Statistical analyses
All data are represented as mean ± standard error of the mean. Error bars indicate
standard error of the mean. Statistical significance was calculated by a one- and two-
tailed Mann-Whitney u-test for male-male and male-female courtship behavior,
respectively. Tests were performed in R and the resulting P-value is reported.

117

Chapter 4. Doublesex establishes sexual dimorphism in the Drosophila central
nervous system in an isoform-dependent manner by directing cell number

4.1. Overview
doublesex (dsx) encodes sex-specific transcription factors (DSX
F
in females and DSX
M
in
males) that act at the bottom of the Drosophila somatic sex determination hierarchy. dsx,
which is conserved among diverse taxa, is responsible for directing all aspects of
Drosophila somatic sexual differentiation outside the nervous system. The role of dsx in
the nervous system remains minimally understood. Here, the mechanisms by which DSX
acts to establish dimorphism in the central nervous system were examined. This study
shows that the number of DSX-expressing cells in the central nervous system is sexually
dimorphic during both pupal and adult stages. Additionally, the number of DSX-
expressing cells depends on both the amount of DSX and the isoform present. One
cluster of DSX-expressing neurons in the ventral nerve cord undergoes female-specific
cell death that is DSX
F
-dependent. Another DSX-expressing cluster in the posterior
brain undergoes more cell divisions in males than in females. Additionally, early in
development, DSX
M
is present in a portion of the neural circuitry in which the male-
specific product of fruitless (fru) is produced, in a region that has been shown to be
critical for sex-specific behaviors. This study demonstrates that DSX
M
and FRU
M

expression patterns are established independent of each other in the regions of the central
nervous system examined. In addition to the known role of dsx in establishing sexual
118

dimorphism outside the central nervous system, the results demonstrate that DSX
establishes sex-specific differences in neural circuitry by regulating the number of
neurons using distinct mechanisms.
4.2. Introduction
Sex–specific differences in neural circuitry contribute to differences in male and female
reproductive behaviors (reviewed in Ball and Balthazart, 2004; Simerly, 2002). In
Drosophila, sex-specific reproductive behaviors are specified through a genetic
regulatory cascade called the sex determination hierarchy (see Figure 1.1; (reviewed in
Manoli et al., 2006)). This hierarchy consists of a pre-mRNA splicing cascade that
culminates in the production of sex-specific transcription factors. doublesex (dsx) is at
the bottom of one branch of the sex hierarchy and has been shown to specify all aspects
of sex-specific development outside the nervous system (Hildreth, 1965). fruitless (fru)
is at the bottom of another branch of the sex hierarchy and has been shown to encode
male-specific transcription factors (FRU
M
, encoded by fru P1 transcripts; (Ryner et al.,
1996)) that underlie the potential for male courtship behaviors (reviewed in Manoli et al.,
2006). Recent studies have shown that both fru and dsx collaborate in the central nervous
system (CNS) to bring about the potential for one step in the male courtship ritual, the
production of courtship song (Rideout et al., 2007). Despite this progress, the
mechanisms by which dsx establishes differences in neural circuitry are largely unknown.

dsx encodes both male (DSX
M
) and female (DSX
F
) transcription factors (reviewed in
Christiansen et al., 2002). DSX isoforms share a common amino terminal region that
119

contains the DNA binding domain, but differ in their carboxyl terminal region (Figure
1A; (Burtis and Baker, 1989)). dsx specifies nearly all aspects of Drosophila somatic sex
determination outside the nervous system, as dsx null animals display an intersexual
phenotype (Hildreth, 1965). Furthermore, if DSX
M
is the only DSX isoform produced in
chromosomally XX animals, these animals look almost identical to wild type males
(hereafter called pseudomales), suggesting that dsx is sufficient to specify most aspects of
sex-specific somatic development (Duncan and Kaufman, 1975).

The role of dsx in directing sex-specific nervous system development and physiology has
been more difficult to examine, given that there are no overt morphological differences
between the sexes in the nervous system. Behavioral analyses on dsx mutants provide
conflicting results. Initially, it was thought that dsx was unable to direct the nervous
system to a male fate, given the observation that pseudomales do not display any male-
specific behaviors (Taylor et al., 1994). However, it was also shown that dsx null males
perform courtship in a quantitatively subnormal manner (Villella and Hall, 1996), and a
population of abdominal neuroblasts undergoes more dsx-dependent divisions in males
(Taylor and Truman, 1992). Recent studies suggest a key role for dsx in specifying
reproductive behaviors, including a demonstration that DSX
M
and FRU
M
are co-
expressed in subsets of neurons in the CNS (Billeter et al., 2006; Rideout et al., 2007),
that DSX and FRU collaborate to bring about the potential for wing song (Rideout et al.,
2007), and that animals that are transheterozygous for dsx and fru P1 alleles show a
reduction in male courtship behaviors (Shirangi et al., 2006).
120

DSX is expressed in a sexually dimorphic pattern in the adult CNS (Lee et al., 2002). In
this study, the mechanisms responsible for generating this dimorphism were determined.
DSX expression was examined during metamorphosis, in both males and females and the
pattern is very similar to that which was previously described, with some differences (Lee
et al., 2002). Additionally, this study shows that the DSX-expressing cell number is
established during a small window of time during early stages of metamorphosis. DSX
directs the number of DSX-expressing cells in the CNS in an isoform-specific and dose-
dependent manner. The sexually dimorphic number of DSX-expressing cells in one
region of the ventral nerve cord (VNC) is a result of DSX
F
-dependent cell death that
occurs during metamorphosis in females and not males. Additionally, a population of
DSX-expressing neurons in the posterior brain of males undergoes more cell divisions
than in females. Given that the sex-specific transcription factors DSX
M
and FRU
M
have
overlapping expression patterns, we examine if these transcription factors are responsible
for establishing differences in each other’s expression patterns in the CNS, and find that
they are not inter-dependent. Furthermore, in females, overlap between DSX
F
and fru
P1-expressing cells during development does not occur in the two regions examined that
have a dimorphism of number of neurons. Taken together, this work demonstrates a role
for DSX in forming sex-specific differences in cell number that underlie differences in
neural circuitry in the CNS.
121

4.3. Results
DSX expression in the CNS is sexually dimorphic across development
To determine the mechanism that underlies the sexual dimorphism in the number of
DSX-expressing cells, a polyclonal rat anti-serum specific to a common portion of DSX
was generated and DSX-expressing cells in the CNS during development were analyzed
(see Figure 4.3A; see Figures C1 and C2 for DSX expression patterns in other tissues ).
The antibody is specific to DSX, as signal is detected in wild type animals, but not in dsx
null animals (Figure 4.1A,B). Using antibodies against DSX and the nuclear, neuron-
specific ELAV protein (Robinow and White, 1991), we demonstrate that nearly all DSX-
expressing cells detected here are neuronal and that DSX is localized to the nucleus
(Figure 4.2). When DSX is over-expressed using a constitutive promoter, signal is
detected throughout the brain using the DSX antibody (Figure 4.1C).

Figure 4.1. The DSX antibody is specific. There is no DSX expression in dsx null tissue. Shown is the
brain (A) and VNC (B) of a transheterozygote of the dsx alleles dsx
M+R15
and dsx
D+R3
. (C) DSX expression
pattern is shown in a male brain in which DSX
F
has been overexpressed under the control of a heat shock
122

promoter. DSX expression in wild type female (D) and male (E) aDN cells are shown. See schematic in
Figure 4.3 for location of aDN cells.

Figure 4.2. ELAV, a protein expressed in the nuclei of neurons, co-localizes with DSX. Neurons in the
posterior midbrain (A) and TN1 region of the VNC (B) co-express DSX (red) and ELAV (green). Note
that ELAV is expressed in small, highly localized regions in the nucleus of some cells. 40X confocal
sections of ~0.5 µm are shown.

Previous studies have shown that DSX is present in the CNS and sexually dimorphic at
the adult stage, but not at the 48-hour after puparium formation (APF) stage (Lee et al.,
2002). DSX expression 48-hours APF was examined here and a sexual dimorphism in
several regions of the CNS was observed. To determine how the dimorphism in DSX-
expressing cells is both established and maintained, the number of DSX-expressing cells
in the CNS were quantified in 48-hour APF pupae and 0-24 hour adults (see Tables 4.1
and 4.2). At 48-hours APF, the combined number of cells for pC1 and pC2 (see Figure
4.3 for description of nomenclature) is 88±1.9 and 13±0.8, in males and females,
respectively (all cell counts are represented as the mean ± standard error of the mean). A
123

sexual dimorphism in the TN1 cluster of the ventral nerve cord (VNC) was also
observed, with 16±1.3 and 0±0 cells, in males and females, respectively (Figure 4.3).

A similar overall pattern of DSX-expressing cells was observed in the 0-24 hour adult
CNS, with a dimorphism in the pC1, pC2 and TN1 regions, as was observed in 48-hour
APF pupae (Tables 4.1 and 4.2, Figure 4.3, and previously described (Lee et al., 2002)).
Two DSX-expressing neurons are present on the anterior dorsal side of the brain (aDN
cluster) in both males and females (Figure 4.1D,E); the aDN neurons were previously
thought to be male-specific (Lee et al., 2002). Here, the mechanisms by which the sexual
dimorphism in DSX-expressing cells in the pC1, pC2 and TN1 clusters are generated
were examined.

Table 4.1. DSX-expressing cells in the pC1 and pC2 regions of the brain
Male (XY) Female (XX)
cells (n) cells (n)
Wild type Wild type
0-24 hour adult 83+/-3.1 (17) 0-24 hour adult 16+/-0.9 (7)
48 hours APF 88+/-1.9 (7) 48 hours APF 13+/-0.8 (12)
0 hours APF 64+/-2.7 (10) 0 hours APF 13+/-0.9 (8)

dsx mutants dsx mutants
0-24 hour adult, dsx
D
/dsx
M+R15
64+/-1.7 (24) 0-24 hour adult, dsx
D
/dsx
M+R15
56+/-2.4 (15)
0-24 hour adult, dsx
M+R15
/+ 56+/-3.8 (9) 0-24 hour adult, dsx
M+R15
/+ 15+/-0.8 (16)
0-24 hour adult, dsx
D
/+ 83+/-2.2 (15) 0-24 hour adult, dsx
D
/+ 41+/-1.3 (13)

ix mutant ix mutant
0-24 hour adult, ix
3
/Df(2R)en-B 75+/-2.9 (12) 0-24 hour adult, ix
3
/Df(2R)en-B 29+/-1.3 (14)

fru mutant
0-24 hour adult, fru
4-40
/fru
P14
93+/-4.8 (8)

124

Table 4.1, continued

Cells that co-express DSX
M

and FRU
M

48 hour APF, UAS-nlsGFP; fru
P1-GAL4/+ 18+/-0.6 (12)

Table 4.2. DSX-expressing cells in the TN1 region of the VNC
Male (XY) Female (XX)
cells (n) cells (n)
Wild type Wild type
0-24 hour adult 15+/-0.5 (12) 0-24 hour adult 0+/-0 (12)
48 hours APF 16+/-1.3 (6) 48 hours APF 0+/-0 (8)
0 hours APF 0+/-0 (12) 0 hours APF 0+/-0 (8)

dsx mutants dsx mutants
0-24 hour adult, dsx
D
/dsx
M+R15
10+/-0.5 (8) 0-24 hour adult, dsx
D
/dsx
M+R15
10+/-0.8 (10)
0-24 hour adult, dsx
M+R15
/+ 11+/-0.6 (10) 0-24 hour adult, dsx
M+R15
/+ 0+/-0 (16)
0-24 hour adult, dsx
D
/+ 17+/-0.6 (14) 0-24 hour adult, dsx
D
/+ 5+/-0.3 (16)
Table 4.2, continued
ix mutant ix mutant
0-24 hour adult, ix
3
/Df(2R)en-B 18+/-0.7 (12) 0-24 hour adult, ix
3
/Df(2R)en-B 6+/-0.5 (10)

fru mutant
0-24 hour adult, fru
4-40
/fru
P14
15+/-0.6 (5)

Cells that co-express DSX
M

and FRU
M

48 hour APF, UAS-nlsGFP; fru
P1-GAL4/+ 9+/-0.5 (8)

125

Figure 4.3. DSX expression is sexually dimorphic.

(A) Schematic of sex-specific DSX isoforms. The common region includes the DNA binding domain
(region 1). The sex-specific regions (region 2) are 152 and 30 amino acids, in the male and female
isoforms, respectively (Burtis and Baker, 1989). The polyclonal antibody was generated against amino
acids 290-398 (bar). (B) Schematic of posterior surface of adult male CNS, indicating where DSX-
expressing clusters pC1 (blue) and pC2 (white) reside in the brain. (C) Schematic of anterior surface of
adult male CNS, indicating where clusters of DSX-expressing cells aDN, SN, and SLN reside in the brain,
and TN1 (purple) and TN2 cells reside in the VNC. DSX-expressing cell groups are indicated by dots on
both sides of the schematic, but only labeled on the left, as DSX-expressing cells are bilaterally
symmetrical. (D) Schematic of male white-pre-pupal CNS, indicating where clusters of DSX-expressing
cells reside (white). DSX-expressing cells in the posterior midbrain (E-J) and ventral nerve cord (K-N) in
0-24 hour adults (E, F, K and L), 48-hour pupae (G, H, M and N), and 0-hour white pre-pupae (I and J;
only one hemisegment is shown for each brain). Arrow and arrowhead (E) indicates pC2 and pC1cluster,
respectively. Arrowheads in (K) and (M) indicate TN1 region. Magnified DSX-expressing cells in the
TN1 region of males (O) and females (P). Male and female genotypes are indicated by XY and XX,
respectively. 20X confocal sections (~1 µm thick) are shown.

126

Figure 4.3

127

DSX regulates the number of cells in the adult TN1 cluster in an isoform- and dose-
dependent manner
Given the sexual dimorphism in DSX-expressing cells, we sought to determine whether
the difference in DSX isoforms in males and females is responsible for the difference in
the number of DSX-expressing cells, or if sex-specific regulation of dsx transcription in
males and females is responsible for the difference in DSX-expressing cell number. The
TN1 cluster was examined in the adult stage, as the dimorphism in cell number in this
cluster is the most dramatic, and therefore the most straightforward to analyze: 15±0.5
and 0±0 cells are in adult males and females, respectively. If the absence of DSX-
expressing cells in the TN1 cluster in females is because dsx transcription is not
activated, then the number of DSX-expressing cells should be independent of the DSX
isoform. Alternatively, if DSX regulates cell number in an isoform-dependent manner,
then the presence of a DSX isoform will influence the number of DSX-expressing cells in
the TN1 cluster.

Accordingly, the number of DSX-expressing cells in chromosomally XX animals that
only produce the DSX
M
isoform was determined. These animals are transheterozygous
for a dsx allele (dsx
D
), whose product can only be spliced to produce the male-specific
isoform, and a dsx null allele (dsx
M+R15
); hereafter these animals will be called dsx
D

pseudomales, as they are phenotypically male. The number of DSX-expressing cells in
both chromosomally XX and XY dsx
D
/dsx
M+R15
transheterozygous flies was examined;
128

the chromosomally XY flies served as a control to determine the number of DSX-
expressing cells when only one copy of dsx can produce DSX product.

In chromosomally XX and XY dsx
D
/ dsx
M+R15
flies, 10±0.8 and 10±0.5 cells were
observed, in TN1 regions, respectively. Given that both chromosomally XX and XY
dsx
D
/dsx
M+R1
flies have the same number of DSX-expressing cells in this region, this
demonstrates that the DSX isoform directs the number of cells, rather than differences in
dsx transcription in males and females.

Both chromosomally XX and XY dsx
D
/ dsx
M+R15
flies had significantly fewer DSX-
expressing cells (10±0.5 and 10±0.8 cells) in the TN1 region than wild type males
(15±0.5 cells) (P=1.9 x 10
-5
and 7.4 x 10
-7
, respectively), which suggests a dsx dose-
dependency for the number of DSX-expressing cells, since the dsx
D
/ dsx
M+R15
mutants
examined only had one copy of dsx that can make DSX
M
product. Chromosomally XY
flies that are transheterozygous for the dsx
D
allele and a wild type dsx allele had 17±0.6
cells in the TN1 cluster, which is more than wild type males (15±0.5 cells). Taken
together, these results suggest that the dose of dsx is important, and the difference in
number in dsx
D
/ dsx
M+R15
mutants is not due to the dsx
D
allele, as if that were the case,
fewer DSX-expressing cells would be expected in flies that are transheterozygous for the
dsx
D
allele and a wild type allele, as compared to wild type males.

129

To further confirm this, DSX-expressing cells were quantified in males that are
transheterozygous for one dsx null allele and one wild type dsx allele (dsx
M+R15
/TM6B;
hemizygotes), and thus have one dsx allele that can make DSX
M
product. These flies had
11±0.6 DSX-expressing cells in the TN1 region, which is similar to the number of DSX-
expressing cells we observed in the dsx
D
/ dsx
M+R15
pseudomales and males (10±0.5 and
10±0.8 cells, P>0.05 for both comparisons). This observation suggests that there is a
threshold amount of DSX
M
activity required to specify the wild type number of DSX-
expressing neurons in the TN1 region. In animals that contain only one dsx allele
producing functional product, the amount of DSX
M
may be close to the threshold needed
to establish or maintain the wild type number of neurons.

DSX
M
and DSX
F
have antagonistic roles in establishing adult DSX cell number in
the TN1 cluster
To determine if DSX
M
and DSX
F
have antagonistic roles in establishing the number of
TN1-region, DSX-expressing cells, we examined chromosomally XX flies that contain
both DSX
M
and DSX
F
. These flies are heterozygous for the allele (dsx
D
) that only
produces DSX
M
and a wild type allele of dsx, which in chromosomally XX flies produces
DSX
F
. Hereafter these transheterozygotes are called dsx
D
intersexual flies, as they have
morphological features of both sexes. Previous studies have shown that DSX functions as
a homodimer, and the presence of DSX
F
interferes with DSX
M
activity by forming a
heterodimer (Erdman et al., 1996). Also, since both DSX isoforms bind the same
130

enhancer element DNA, DSX
F
homodimers can compete with DSX
M
homodimers for
DNA (Erdman et al., 1996).

When the TN1 cluster in dsx
D
intersexual flies was examined, 5±0.3 cells were present
(Table 4.2 and Figure 4.4A); this is intermediate to the number in wild males (15±0.5
cells) and females (0±0 cells), and significantly fewer than the number in dsx
D

pseudomale animals (10±0.5 cells; P=5.1 x 10
-5
) or male dsx hemizygotes (11±0.6 cells;
P=9.3 x 10
-7
); these latter two genotypes have one copy of dsx that produces DSX
M
.
These results suggest that DSX
M
and DSX
F
isoforms have antagonistic roles with respect
to establishing the TN1 population of DSX-expressing cells. DSX
F
may reduce the
number of cells in dsx intersexual animals indirectly, by interfering with DSX
M
activity
and effectively reducing the dose of DSX
M
. Alternatively, DSX
F
may directly regulate the
TN1 region cell number, perhaps by inducing cell death or blocking cell division.
Clearly, DSX
M
and DSX
F
establish the number of DSX-expressing cells in a context-
dependent way, as DSX-expressing cells in females persist in other regions of the CNS at
this same developmental time (Figure 4.3).

131

Figure 4.4. The number of DSX-expressing cells depends on the dose and isoform of DSX present.
(A) Dark blue and light blue bar indicates genotypes with two or one copy of dsx that can produce DSX
M
,
respectively. Light red and dark red bar indicates genotypes with one and two copies of dsx that can
produce DSX
F
, respectively. Chromosomal sex of the animal is shown in parentheses. (B and C) The
number of DSX-expressing cells in male and female CNS from 0 hour white pre-pupae to 0-24 hour adults,
in the TN1 cluster (B) and posterior brain clusters (C) in males and females. The bar in (B) represents the
period during development when the TN1-region, DSX-expressing cell number is isoform-independent.
Six to thirty individuals were analyzed at each time point (B and C). The abscissa and ordinate show the
developmental stage and number of DSX-expressing cells, respectively. Male (blue circles) and female (red
triangles) data are indicated. Error bars indicate the standard error of the mean.

132

Figure 4.4

133

Male and female TN1 regions have detectable DSX-expressing cells early in pupal
development, but DSX-expressing cells in the female TN1 region are not detectable
in later stages
To determine how the number of DSX-expressing cells is established by the DSX
isoform present, the TN1 cluster was examined during development. In females, the
absence of DSX
M
or the presence of DSX
F
may result in death of those cells.
Alternatively, DSX
M
may drive additional cell divisions. The number of DSX-expressing
cells was examined during metamorphosis, with the rationale that the developmental
profile would provide insight into the mechanism underlying the resultant sex-specific
cell number (see Figure 4.4B, Table C1).

In the TN1 cluster, in 0-hour white pre-pupae, no DSX-expressing cells were present in
males or females (Figure 4.4B, Table 4.2). Twelve hours later (12-hours APF), 11±0.4
and 11±0.6 cells in males and females were observed, respectively. In females, at 14-
hours APF, the cell numbers decreased to 5±0.4. By 48-hours APF, no DSX-expressing
cells were observed in the TN1 region in females. In males, the number of cells steadily
rose with time, with 12±0.5 and 16±1.3 cells observed at 14- and 48-hours APF.

Females initially have a substantial number of cells, located in a homologous position to
those observed in males, and none of these cells are detected by 48-hours APF. This
suggests that sex-specific cell death may be the mechanism by which the sexually
dimorphic number of cells is established. After 12-hours APF, males have more DSX-
134

expressing cells than were observed in females, which suggests that the loss of DSX-
expressing cells in females is not sufficient to explain the entire difference in cell
numbers. An additional cell division of the DSX-expressing cells may be occurring in
males, and DSX
M
may either permit or direct this cell division.

Sex-specific cell death leads to absence of DSX-expressing neurons in the TN1
cluster in females
To address if cell death is responsible for the absence of detectable TN1-region, DSX-
expressing cells in females, the number of DSX-expressing cells showing molecular signs
of cell death was determined. In one of the stages of cell death, cellular endonucleases
cleave nuclear DNA into small fragments that are detectable by the TUNEL (Terminal
deoxynucleotidyl transferase (TdT)- mediated dUTP nick-end labeling) assay. Female
and male TN1 regions were examined starting at 8 hours APF, every two hours, until 22
hours APF (Figure 4.5). These time points were chosen because the number of DSX-
expressing cells declines substantially in females during the 12-24-hour APF period
(Figure 4.4B). Animals were staged by starting with the white pre-pupae stage, which
lasts about 20-30 minutes, so each subsequent time point examined represents about 40
minutes to one hour of development.

Overlap in TUNEL staining and DSX was observed at all time points examined between
10- and 22-hours APF, in female, but never in male, TN1-region VNC tissue. During the
12 hour and 14 hour APF stages, in which the largest decline in TN1-region, DSX-
135

expressing cell number is observed (Figure 4.4B), 28% of female samples examined had
at least one TUNEL-positive DSX-expressing cell (n=36), whereas males had none
(n=16). Typically, only one cell with overlap in TUNEL staining and DSX was observed
per female VNC. However, given the difficulty of detecting a nuclear transcription factor
that is being stripped from the nucleus during cell death (Robinow et al., 1993), the
changes in nuclear morphology as a cell dies, and the limited window in which it is
possible to detect TUNEL staining in a dying cell, this result is not unexpected.
Furthermore, the overall number and distribution of TUNEL-positive cells in male and
female VNC tissue is very similar at all stages examined, which suggests that the overlap
observed with DSX and TUNEL in the female VNC is specific, and not due to a greater
number of dying cells in females which would then make it more likely that DSX-
expressing cells would randomly overlap with TUNEL staining in females. We thus
conclude that female-specific cell death in the TN1 region establishes the sexual
dimorphism in cell number.

136

Figure 4.5. DSX-expressing cells in the TN1 cluster in females undergo cell death during the first day
of pupal development.

For panels A-D, DSX-expressing cells (red), TUNEL staining (green), and co-localization (yellow) in the
TN1 regions at 14-hours APF is shown. A-C show female tissues; D shows male tissue. (A and B)
TUNEL-positive cells are also DSX-expressing cells. Arrowheads indicate a TUNEL-positive cell with
faint DSX signal. (C) TUNEL-positive cells in close proximity to remaining DSX-expressing cells. DSX-
expressing cells in the TN1 region in 0-24 hour adult wild type females (E), XX ix
3
/Df(2R)en-B flies (F),
and wild type males. 40X confocal sections of ~0.5 µm are shown for A-G. (G) The number of DSX-
expressing cells in the CNS of dsx and ix mutants from 0 hour white pre-pupae to two-day old pupae, in the
TN1 cluster (H) and posterior brain clusters (I). Legend indicates genotypes examined. Six to twenty
individuals were analyzed at each time point. The abscissa and ordinate show the developmental stage and
number of DSX-expressing cells, respectively. Chromosomal sex of animals is indicated in all panels and
legend. Error bars indicate the standard error of the mean.

137

Figure 4.5

138

Figure 4.5, continued

To further determine the mechanism that establishes female-specific, TN1-region cell
death, female adult transgenic animals that over-express cell death inhibitors (including
bacculovirus p35 and the caspase inhibitor DIAP) and adult animals that are mutant for
139

cell death effectors (transheterozygotes for a deletion of rpr, hid, and grim and a deletion
of rpr) were examined. None of the female animals examined had TN1 region, DSX-
expressing cells, leaving the cell death effectors in the TN1 region an open question
(Table 4.3).

Table 4.3
Genotypes interrupting apoptosis

Genotype n
appl-GAL4; UAS-p35 14
tubulin-GAL4; UAS-p35*
actin-GAL4;UAS-p35*
elav-GAL4; UAS-p35 14
elav-GAL4/UAS-p35 8
heat shock-GAL4;UAS-p35 18
appl-GAL4; UAS-diap 18
tubulin-GAL4; UAS-diap 8
actin-GAL4;UAS-diap 18
elav-GAL4; UAS-diap 16
elav-GAL4/UAS-diap 18
heat shock-GAL4;UAS-diap 14
Df(3L)XR38/Df(3L)H99 17
*lethal

DSX
F
function is necessary, but not sufficient, for female-specific apoptosis that
results in the absence of cells in the adult TN1 region
To determine if DSX
F
is required for female-specific cell death in the TN1 region, dsx
function was abrogated in females via mutation of an obligate DSX
F
heterodimer partner
encoded by intersex (ix); this allows for the detection of DSX-expressing cells with the α-
DSX antibody in cells in which DSX is not functional. ix encodes an obligate
140

heterodimer partner of DSX
F
, but not DSX
M
, and is required for the establishment of all
female-specific morphological features under the control of DSX
F
that have been
examined (Garrett-Engele et al., 2002). Therefore, if DSX
F
is required for the absence of
DSX-expressing cells in the female TN1 region, than in ix mutant females DSX-
expressing cells should be observed, as DSX
F
is not functional in the absence of IX.
Accordingly, a strong ix hypomorphic allele combination (ix
3
/Df(2R)en-B) was
examined. DSX-positive cells (6±0.5 cells) were observed in the TN1 region of ix adult
females (n=10, Figure 4.5F), which is significantly more than the zero observed in wild
type adult females (P = 4x10
-7
). Therefore, DSX
F
is necessary for the absence of TN1-
region, DSX-expression cells. Because the ix allele combination is not a null (Garrett-
Engele et al., 2002), some DSX
F
function may remain to direct cell death in the DSX-
expressing, TN1-region neurons, and that could explain why the number of TN1 region
cells are less than the number observed in wild type males. Alternatively, DSX
M
may be
required for establishing the final cell number in males.

Given the observation that DSX
F
activity is required for the absence of cells in the TN1
region in females, we next tested if it is sufficient to induce cell death. DSX
F
was over-
expressed with the fru P1-GAL4 driver, and the number of FRU
M
-expressing cells was
assayed using an antibody directed against FRU
M
. We and others (see below and Rideout
et al., 2007) have shown that there are a small number of DSX-expressing, TN1-region
cells that overlaps with fru P1 in males. The activity of dsx is context dependent, based
on the observation that in some cells in females, dsx is required for cell death, whereas in
141

other populations, DSX-expressing cells persist. Thus, it was important to test if DSX is
sufficient to drive cell death in a cell population most similar to the DSX-expressing,
TN1 region female cells. Accordingly, the anterior surface fru P1-expressing PrMs
region cells were assayed; the fru P1 anatomical nomenclature is used, as fru P1-
expressing cells were examined (anterior surface FRU
M
-PrMs cells overlap DSX-TN1
cells in males; (Lee et al., 2000a)). There is no statistical difference in number of anterior
region PrMs cells between control males (fru P1-GAL4/+; 72±2.3 cells) and males that
expresses DSX
F
under the fru P1-GAL4 promoter (fru P1-GAL4; UAS-DSX
F
flies;
72±2.6 cells) (P = 0.4). Therefore, over-expression of DSX
F
in the male PrMs cells is not
sufficient to induce cell death, suggesting that DSX
F
acts in a specific context in females
to direct cell death. Alternatively, the PrMs cells may not homologous to the TN1 region
cells in females, or the over-expression may not produce enough DSX
F
to induce cell
death.

Increase in cell number in TN1 region between 0 hr wpp to 12 hr pupal stages is
DSX isoform independent, but dsx dependent
To determine when in development dsx is required to establish the TN1-region, DSX-
expressing cell number, DSX–expressing cells were examined in dsx mutant
backgrounds. Between the 0 hour wpp (0 cells in both sexes) and 12 hour pupal stages
(~11 cells in both sexes), the increase in DSX-expressing cell number in males and
females is indistinguishable and shows a similar trajectory (see bar in Figure 4.4B). The
following genotypes were examined: male and female animals heterozygous for a null
142

and a wild type allele of dsx and chromosomally XX, ix (ix
3
/Df(2R)en-B) flies. In these
genotypes, dsx function is reduced, which allows us to determine if the increase in cell
number between 0 and 12 hour APF, in males and females, requires wild type levels of
dsx product and function.

At 8 hours APF, wild type males and females have 7±0.8 and 8±0.9 TN1-region cells
respectively, whereas no DSX-expressing cells were detectable in male and female dsx
hemizygotes (dsx
M+R15
/TM6B), and significantly fewer cells were observed in ix females
(2±0.7 cells), as compared to wild type females (8±0.9 cells) (P = 7x10
-6
; Figure 4.5H,
Table C1). This suggests that dsx is required for establishing the number of cells in both
males and females during the 0 hour to 8 hour APF stages.

By 12 hours APF, the last time point examined before wild type males and females
diverge in TN1-region cell number, there is no statistical difference between
chromosomally XY animals dsx
M+R15
/TM6B and wild type male animals (P = 0.12),
although there is a slight, but statistically significant decrease in chromosomally XX,
dsx
M+R15
/TM6B animals (8±0.7 cells), as compared to wild type females (11±0.6 cells) (P
= 0.002) (Figure 4.5H). Chromosomally XX, ix mutants (ix
3
/Df(2R)e-B) were also
examined at 12 hours APF, and a slight but significant decrease in cell number was
observed as compared to wild type females (9±0.9 and 11±0.6 cells, respectively; P =
0.05; Figure 4.5H). Taken together, these results suggest that the wild type amount of
143

dsx activity is required to establish DSX-expressing TN1-region cell number, between 0
and 12 hour pupal stages.

To determine if the increase in the number of cells during the 0 hour wpp to 12-hour APF
stage is DSX-isoform independent, the number of DSX-expressing, TN1-region cells
were examined in chromosomally XX, dsx
D
intersexual flies that produce both DSX
M
and
DSX
F
. If the isoforms have distinct activities, the presence of both isoforms would
interfere with wild type activity. The number of DSX-expressing cells in dsx
D

intersexual flies at 8- and 12-hour APF stages in chromosomally XX animals was
indistinguishable from that observed in wild type males and females (Figure 4.5H). At 8
hours APF, dsx
D
intersexual flies have 6±0.5 cells; wild type males and females have
7±0.8 and 8±0.9 (for both, P>0.05). At 12 hours APF, dsx
D
intersexual flies have 11±0.6,
wild type males have 11±0.4 (P >0.05), and wild type females have 11±0.6 cells (P
>0.05) (Figure 4.5H). This suggests that the establishment of cell number during this
period of development is DSX-isoform independent, and that the DSX transcription
factors do not have antagonistic functions with respect to establishing cell number. Our
previous studies have shown that DSX
M
and DSX
F
can activate and repress many of the
same target genes, but that the extent of activation or repression is isoform specific
(Goldman and Arbeitman, 2007). These results are consistent with that observation and
suggest that in the context of the TN1-region, DSX-expressing cells during early pupal
stages, the amount of DSX activity in both males and females is sufficient to establish the
final cell number.
144

Sex-specific differences in DSX-expressing TN1 region cells after 12 hour APF stage
is DSX-isoform dependent
In wild type animals, after the 12-hour APF stage, the number of DSX-expressing, TN1-
region cells is sexually dimorphic. The absence of TN1-region, DSX-expressing cells in
females is due to female-specific cell death that initiates after the 12 hour APF stage and
requires DSX activity, as when ix mutant females were examined, 6±0.5 cells were
evident at the adult stage. At the 16-hour APF stage, ix females had roughly the same
number of TN1 region, DSX-expressing cells as they had at 12 hours, demonstrating that
the trajectory of cell loss is very different in ix females than that observed in wild type
females, where the number of TN1-region DSX-expressing cells has dropped about two-
fold (Figure 4.5H). This demonstrates that DSX
F
is required for directing the cell death
process during early pupal development and not just at late stages. When female dsx
hemizygotes (dsx
M+R15
/TM6B) were examined, the cell death trajectory looked very
similar to that observed in wild type females, suggesting that ~2-fold dose reduction does
not affect cell number at this stage.

After the 12-hour APF stage, TN1-region cell number in males continues to increase,
demonstrating that cell death in females is not sufficient to account for the sexual
dimorphism in cell number observed in adults. The number of DSX-expressing, TN1-
region cells in male dsx hemizygotes (dsx
M+R15
/TM6B) at the 16-hour APF stage was
examined, and a slight (13±0.5 cells), but significant (P = 0.03), decrease in cell number
145

compared to wild type males (15±0.6 cells ) was observed (Figure 4.5H). However, at
adulthood, dsx heterozygotes only have 11+0.6 cells, compared to 15±0.5 cells in wild
type males (P = 2.5 x 10
-6
), suggesting that the dose of dsx is critical for the increase of
cells observed after the 16 hour-APF stage. This is also consistent with the observations
of chromosomally XX, dsx
D
intersexual animals, in which the TN1-region cell number is
similar to wild type males at the 16-hour APF stage (15±2 and 14.5+0.6 cells in wild type
males), but subsequently there is a decline to 5±0.3 cells, as compared to 15±0.5 cells in
adult wild type males (Figure 4.5H). Taken together, this suggests that both the dose and
isoforms of DSX are critical for the increase in cell number observed in wild type males
after the 12-hour APF stage.

The difference in DSX cell number in the adult posterior brain clusters is dependent
on the amount of DSX activity and the DSX-isoform present
To determine if the DSX isoform and the dose of dsx present also determines sex-
specific, DSX-expressing cell numbers in other regions of the CNS, the combined
number of neurons in the posterior brain pC1 and pC2 clusters in dsx
D
pseudomales was
examined. In adult chromosomally XX and XY, dsx
D
/ dsx
M+R15
pseudomales and males
(only produce DSX
M
), that only have one allele of dsx that produces DSX
M
, 56±2.4 cells
and 64±1.7 cells, respectively, were observed; this is significantly lower than the wild
type male numbers of DSX-expressing cells (83±3.1 cells; P =6.1 x 10
-8
and 3.7 x 10
-6
,
respectively). However, the number of DSX-expressing cells in XX and XY dsx
D
/
146

dsx
M+R15
animals

are much closer to each other, and to the number of cells observed in
wild type males, (83±3.1 cells, Table 4.1), than wild type females (16±0.9 cells, Table
4.1), consistent with the idea that the DSX isoform establishes the number of DSX-
expressing cells and that the difference in number is not due to sex-specific differences in
expression of dsx.

To determine if the fewer cells in the adult dsx
D
/ dsx
M+R15
animals, as compared to adult
wild type males, is because there is only one allele that can produce DSX
M
product,
males transheterozygous for a wild type dsx allele and a dsx null allele, dsx
M+R15
, were
examined. These males had 56±3.8 cells in the pC1 and pC2, which is also significantly
fewer than what is observed in wild type males (P =1.8 x 10
-5
), consistent with the idea
that the dose of dsx establishes cell number.

We also determined if the difference in number between dsx
D
pseudomales (XX; 56±2.4
cells) and dsx
D
/ dsx
M+R15
(XY; 64±1.7 cells) males, as compared to wild type males
(83±3.1 cells), is due to the reduced dose of dsx, and not the presence of the dsx
D
allele,
by examining animals that contain both the dsx
D
allele and a wild type allele. In these
animals that have two copies of dsx that can make DSX
M
product, the number of cells in
the pC1 and pC2 clusters is 83±2.3 (Table 4.1), which is indistinguishable from wild type
males (P>0.05). This demonstrates that the dsx
D
allele is not responsible for differences
in cell numbers observed in dsx
D
/ dsx
M+R15
animals, as fewer cells would have been
expected also in dsx
D
/+ animals, but rather, the dose of dsx present is the crucial factor.
147

To investigate if DSX
F
and DSX
M
also have antagonistic functions in establishing the
number of cells in the posterior brain clusters, adult dsx
D
intersexual flies that produce
both DSX
F
and DSX
M
were examined. An intermediate number of DSX-expressing cells
was observed (41±1.3 cells), as compared to wild type males (83±3.1) and females
(16±0.9 cells), but much fewer than that observed in dsx
D
pseudomales (64±1.7 cells;
Table 4.1), suggesting that the DSX isoforms have antagonistic functions in establishing
pC1 and pC2, DSX-expressing cell number. Taken together, the results from both the
TN1 region and the posterior brain cluster analyses are consistent with both the DSX
isoform and the dose of dsx establishing the adult number of DSX-expressing cells
(Figure 4.4A).

DSX-expressing posterior brain cells undergo more cells divisions during
metamorphosis in males as compared to females
The pC1 and pC2, DSX-expressing clusters were examined during development (Figure
4.4C). At the 0-hour white pre-pupal stage, males already display significantly more
DSX-expressing cells than females (64±2.7 and 13±0.8, respectively; P =1.3 x 10
-9
).
DSX-expressing cell numbers in females remain fairly constant throughout the 8-to 48-
hour APF pupal stages, while the DSX-expressing cell numbers in males increase rapidly
between the 8- and 12-hour APF time points (from 68.4 ±2.7 to 93.6±3.2 cells). Thus, in
the pC1and pC2 clusters, DSX
M
may act to promote cell division. Alternatively, the
148

presence of DSX
F
may block additional cell division during metamorphosis, and DSX
M

activity may simply be permissive for cell division that is driven by other pathways.

To address if the sex-specific difference in DSX-expressing pC1 and pC2 cell number is
due to more cell division in males during metamorphosis, as opposed to more cells
expressing dsx in males, molecular hallmarks of cell division were examined in these
cells. We reasoned that if the increase in cell number in males is due to cell division,
then we should be able to detect molecular markers of cell division in males, and not in
females. Here the incorporation of BrdU, a thymidine analog that is incorporated into
dividing cells in S phase and can be detected using immunocytochemistry, was examined
during the period in metamorphosis in which cell number increases substantially in males
and not in females (see bar in Figure 4.6A).

Male and female CNS tissues were dissected at the eight-hour stage, incubated with BrdU
for four hours, fixed and stained for BrdU and DSX. Although in other parts of the CNS
the pattern of BrdU incorporation in male and female brain is largely indistinguishable
during this time period (see Figure C4), cells that co-label with the DSX and BrdU
antibody were detected only in males (17/20 samples examined showed co-labeling), but
never in females (0/26 samples examined showed co-labeling) (Figure 4.6), consistent
with the idea that the increase of cell number in males is due to increased cell division.
The maximum number of cells that co-label with the antibody to BrdU and DSX in a
given male brain was eight cells per hemisphere, and none were ever detected in females.
149

Although this does not fully account for the approximately twenty cell increase per
hemisphere in males during this time (see Figure 4.4C), it demonstrates that DSX-
expressing cells in males are undergoing cell division in the region and during the time
period expected. One possibility is that some of the DSX-expressing pC1 and pC2 cells
are arrested in the G2 phase of the cell cycle and thus would be in the cell division cycle,
but would be beyond the stage when BrdU incorporation occurs.

Figure 4.6. DSX-expressing cells co-localize with BrdU in males, but not in females.
(A) Adaptation from Figure 2C shows a schematic of the four-hour BrdU pulse from 8-hours APF to 12-
hours APF. For all panels, anti-DSX is shown in green, anti-BrdU is shown in red, and co-localization is
yellow. (B and C) Cells in female posterior brain regions do not show co-localization with BrdU and DSX,
while cells in male posterior brains (D and E) show co-localization.
150

Consistent with DSX
M
driving additional cell divisions, when the TN1 region was
examined, there was a slight, but statistically significant increase in FRU
M
-expressing
cells when DSX
M
was over-expressed in the fru P1 circuitry, which suggests that DSX
M

directs additional cell divisions in this region (fru P1-GAL4; UAS-DSX
M
flies had 80±2.4
FRU
M
-expressing cells, while fru P1-GAL4/+ controls had 72±2.6; P = 0.02). These two
lines of evidence suggest that DSX
M
may cause additional cell divisions in both TN1 and
posterior brain regions of the CNS, where DSX expression is sexually dimorphic.

Sex-specific differences in DSX-expressing, posterior region brains cells that are
established during metamorphosis depend on the DSX isoform and dose of dsx
To examine how the sex-specific differences in DSX-expressing posterior brain cell
numbers are established during metamorphosis by dsx, the following genotypes in which
dsx function is reduced were examined: male and female animals heterozygous for a null
and a wild type allele of dsx and chromosomally XX, ix (ix
3
/Df(2R)en-B) flies. At the 8-
hour APF time point male dsx heterozygotes (dsx
M+R15
/TM6B) had the same number of
posterior brain region DSX-expressing cells, as wild type males (Figure 4.5I). However,
at the 12-hour APF time point, which is after the rapid increase in pC1 and pC2 cells in
males, dsx male hemizygotes display a significant decrease in cell number (P = 8 x 10
-5
),
as compared to wild type males, suggesting that the dose of dsx is important for the
increase in cell number in males. The difference in cell number between male dsx
hemizygotes and wild type males is maintained at the 16-hour APF time point, and by the
151

adult stage the difference is more substantial (dsx male hemizygotes have only 56±3.8
cells, whereas wild type males have 83±3.1 cells cell numbers (Figure 4.5I), suggesting
that the wild type dose of dsx is required not only to drive the early divisions, but to also
maintain the population of DSX-expressing cells after the 16-hour APF stage.

At the 12-hour APF time point, dsx female hemizygotes also display a significant
decrease in cell number, as compared to wild type females (dsx female hemizygotes have
14±1.2 cells, whereas wild type females have 19±1.3 cells; P = 0.004; Figure 4.5I). This
difference in cell number between dsx female hemizygotes and wild type females is
maintained at the 16-hour and adult stage (at 16-hours APF, dsx females have 13±0.7
cells, whereas wild type females have 16±0.6 cells; P = 0.003; Figure 4.5I), suggesting
that the initial establishment of DSX-expressing cell number in pupae is sensitive to the
dose of dsx in females. However, when female ix mutants were examined, more female
posterior-region DSX-expressing cells were observed, as compared to wild type females,
at all four stages examined (Figure 4.5I), suggesting that at a certain point in
development, DSX
F
prevents cell division in this region. Consistent with this idea is the
observation that during the pupal stages examined, chromosomally XX, dsx intersexual
animals have significantly fewer cells than dsx male and wild type males, suggesting that
DSX
F
and DSX
M
have antagonistic activities with respect to establishing cell number.

DSX
M
and FRU
M
are co-expressed in a subset of neurons in adults and pupae, but
DSX
M
expression profile does not depend on FRU
M

152

Given the sexual dimorphism in DSX-expressing cells, we wanted to determine if FRU
M

plays a role in males in establishing the number of DSX-expressing cells. A recent report
has shown that DSX
M
and FRU
M
co-localize in a subset of neurons in the mid-pupal CNS
(Rideout et al., 2007). Given the observations that demonstrate that DSX-expressing cell
number is established during metamorphosis, we wanted to determine if FRU
M
plays a
functional role in specifying that developmental fate by analyzing the developmental
profile of DSX
M
and FRU
M
co-localization. To detect FRU
M
-expressing cells, a GAL4
driver that is expressed in the same cells as the fru P1 transcript was used (hereafter
called fru P1- GAL4; (Stockinger et al., 2005) to drive the expression of nuclear localized
GFP (UAS-nlsGFP) in males.

In 0-24 hour old adult males overlap was observed between the DSX-expressing cells and
the fru P1-expressing cells in the pC1, pC2 and TN1 clusters (Figure 4.7 and Tables 4.1
and 4.2), which is similar to previous reports of overlap at mid-pupae stages (Rideout et
al., 2007). Extensive co-expression of DSX and FRU
M
was detected in the abdominal
ganglion region of the VNC (Billeter et al., 2006). Cells were identified that co-express
fru P1 and DSX
M
in the subesophogeal region of the brain, the SN and SLN clusters, and
several cells located in the TN2 region of the VNC (Figure 4.8). Given that only a subset
of DSX-expressing cells overlap in expression with fru P1-expressing cells (Figure 4.7,
Table 4.1; (Rideout et al., 2007)), DSX-expressing cells are likely functionally distinct,
despite residing in close proximity.

153

Figure 4.7. fru

P1-expressing cells and DSX co-localize in the CNS.
DSX-expressing (red) and fru P1-expressing (green) cells in adult (A-L), 8 hour APF (M-O), and 16 hour
APF (P-U) male CNS. fru P1 -expressing cells are visualized using a fru P1-GAL4 transgene, driving
expression of nuclear GFP. (A) DSX (red) and (B) fru P1-expressing cells (green) co-localize (C) (yellow)
in the posterior midbrain. Higher magnification (40X) of DSX (D) and fru P1-expressing cells (E) in the
pC1 and pC2 clusters of the midbrain; (F) is merged image. (G) DSX and (H) fru P1-expressing cells co-
localize (I) in the VNC. Higher magnification (40X) of (J) DSX and (K) fru P1-expressing cell co-
localization (L) in the TN1 region of the VNC. (M) DSX and (N) fru P1-expressing cells co-localize (O) in
male a brain 8 hours APF. (P) DSX and (Q) fru P1-expressing cells co-localize (R) in a male brain 16
hours APF. (S) DSX and (T) fru P1-expressing cells co-localize in the TN1 clusters of a male VNC 16
hours APF. Both TN1 clusters are shown. Unless otherwise indicated, 20X confocal sections (~1 µm thick)
are shown.

154

Figure 4.7

155

Figure 4.8. DSX and fru P1 expressing cells co-localize at 48 hours APF, and in additional regions in
the adult CNS.
(A) DSX and (B) fru P1-expressing cells co-localize (C; yellow) in the posterior midbrain of 48 hour APF
males. (D) DSX and (E) fru P1-expressing cells co-localize (F) in the VNC of 48 hour APF males. (G)
DSX and (H) fru P1-expressing cells (I) co-localization in the SN neurons, located near the subesophogeal
ganglion of the brain, in adults. (J) DSX and (K) fru P1-expressing cells (L) co-localization in TN2 cells in
the VNC in adults. 20X confocal sections (~1 µm thick) are shown. (A-F), and 40X confocal sections of
~0.5 µm are shown (G-L).

Expression patterns of fru P1and DSX at 8-, 12-, 16-, and 48-hours APF were next
examined in males in the pC1, pC2 and TN1 and TN2 regions. At 8-hours APF, co-
expression was observed in brain regions that appear to be the cells fated to become the
pC1 and pC2 clusters, as well as the TN2 region of the VNC, but not the TN1 region
(Figure 4.7). However, by 12-hours APF, cells in the TN1 region of the VNC, and the
pC1 and pC2 regions of the brain co-express DSX and fru P1. Males from 12-, 16-, 48-
hour APF and adults had similar patterns. Overlap of DSX and fru P1-expressing cells in
156

the TN1 and TN2 regions in the VNC, and the pC1 and pC2 regions in the brain were
observed (Figure 4.7, Figure 4.8; (Rideout et al., 2007)). Given that the number of DSX-
expressing cells in males is established during pupal stages and co-localization of FRU
M

and DSX
M
are observed during these stages, we next asked if FRU
M
plays a role in
establishing DSX
M
cell numbers.

DSX expression was examined in 0-24 hour adult fru P1 mutant males that do not
produce FRU
M
. We characterized the three clusters that show sexually dimorphic
numbers of DSX-expressing cells (pC1, pC2 and TN1) and observe no statistically
significant differences between wild type males and fru P1 mutant males (Tables 4.1 and
4.2, Figure 4.9A and B), demonstrating that FRU
M
does not establish the sexual
dimorphism in the number of DSX-expressing cells in these regions in 0-24 hour adults.

157

Figure 4.9. DSX expression in fru P1 null animals and FRU
M
expression in dsx null animals.
DSX-expressing cells in the posterior brain (A) and TN1 cluster (B) of fru P1 null mutant males. FRU
M
-
expressing cells in the PrMs region of the VNC in dsx null males (C) and wild type control males (D). 20X
confocal sections (~1 µm thick) are shown.

DSX and fru P1 do not extensively co-localize in pC1, pC2 and TN1 region neurons
in females
A recent observation has shown that the fru P1-expressing cell circuit is important for
female behaviors (Kvitsiani and Dickson, 2006). In addition, it has been shown that DSX
is required to establish the male-specific number of fru P1-expressing cells (Rideout et
al., 2007). Given the observation of female-specific cell death in the DSX-expressing,
TN1 cluster cells, we wanted to determine if DSX plays a role in establishing the number
of fru P1-expressing cells in females. If DSX
F
is responsible for a difference between
158

males and females in fru P1-expressing cell number previously reported (Rideout et al.,
2007), we would expect to see about ten cells per hemisegment with overlap between
DSX
F
and fru P1 in the TN1 region, as that is the difference in fru P1-expressing cell
number reported to be established by DSX (Rideout et al., 2007).

Accordingly, the fru P1-GAL4 driver that produces GAL4 in homologously positioned
cells in both males and females was used to assess the overlap between fru P1-expressing
cells and DSX in females. Females at 8-, 12-, and 16-hour APF were examined, when
DSX-expressing TN1 cells are present in females, as well as 48-hour APF and 0-24 hour
adults. No cells were consistently observed expressing both fru P1-GAL4 and DSX in the
TN1, pC1 or pC2 regions in females at the 8-, 12-, 16-, -48-hour APF stages and 0-24
hour adults. Taken together, the absence of overlap between fru P1-expressing cells and
DSX
F
during development suggests that DSX
F
does not direct the number of fru P1-
expressing cells in females in the TN1 region or posterior brain in a cell autonomous
manner.

DSX
M
does not direct the number of FRU
M
-expressing cells in adults
To determine if DSX
M
is required to maintain fru P1-expressing cells in the TN1 region
in males, FRU
M
expression was examined in dsx mutant males in the PrMs region of the
VNC (nomenclature from Lee, et al., 2000). The FRU
M
-PrMs cluster overlaps the
DSX
M
-TN1 region. Using a FRU
M
polyclonal antibody, no significant difference was
found between FRU
M
-expressing cell number in dsx null male XY (57±1.7 cells) and
159

wild type males (56±1.7 cells), in 0-24-hour adults (Figure 4.9C and D), on the anterior
surface of the PrMs region of the VNC where DSX and FRU
M
co-localize (n=16 and
n=18, respectively). Furthermore, the total number of FRU
M
-expressing cells in the
entire PrMs region showed no statistically significant difference in FRU
M
-expressing
cells, between dsx null males (98±5.5 cells) and wild type males (105±3.1 cells). The
number of FRU
M
-expressing cells detected was comparable to previous reports (Lee et
al., 2000a). This suggests that at the 0-24 hour stage, the number of FRU
M
-expressing
cells is not regulated by DSX
M
. DSX
M
may be required later in adult life to maintain
parts of the fru P1-circuitry. Evidence for this hypothesis comes from the recent report
that shows 5-7 day old dsx null males have roughly 20 fewer FRU
M
-expressing cells in
the PrMs region of the VNC than wild type males (Rideout et al., 2007). However, when
5-7 day old adult flies were examined in this study, dsx null flies had on average 78±2.6
cells on the anterior surface of the entire PrMS region, and wild type flies had 73±1.3
cells; these values do not differ significantly (P = 0.9). The discrepancy between the two
studies could perhaps be explained by differences in strain or reagents employed.
Nonetheless, the data presented here suggest that FRU
M
-expressing cell number in males
can be established by a mechanism that is independent of DSX.

4.4. Discussion
The identity of dsx has been known for many years (Burtis and Baker, 1989), but how
DSX specifies sexually dimorphic neural circuits has only begun to be investigated.
Here, we report a sexual dimorphism in the number of DSX-expressing cells in the TN1
160

cluster in the VNC and in the pC1 and pC2 clusters in the posterior brain, during both
pupal and adult stages, that is established by the DSX isoform and dose of dsx present.
Males have consistently higher numbers of DSX-expressing cells in the pC1, pC2 and
TN1 clusters than females, and for the pC1 and pC2 regions, we show that DSX-
expressing cells undergo more divisions in males than in females. Our results from
examining the brain clusters pC1 and pC2 and VNC TN1 clusters are consistent with
previous analyses that demonstrated that DSX
M
promotes additional neuroblast divisions
in male abdominal ganglion neuroblasts (Taylor and Truman, 1992). We propose that
DSX
M
may promote additional neuroblast divisions in these clusters, given that the DSX-
expressing pC1 and pC2 cells are in very close proximity, and thus might be derived from
the same progenitor (Pereanu and Hartenstein, 2006).

We also demonstrate that in males, fru P1 is not required to establish the number of
DSX-expressing cells in adults. Rather, our results are consistent with the sex-specific
DSX isoform establishing the dimorphism in number of DSX-expressing cells in these
clusters. Furthermore, in adult males, DSX
M
does not appear to be required to establish
the number of fru P1-expressing cells, as was suggested in a previous study (Rideout et
al., 2007).

The regions of the CNS where DSX is detected in a sexually dimorphic pattern have been
implicated in underlying the potential for male courtship behaviors. Early
gynandromorph studies, in which animals were mosaic for male and female tissues,
161

showed that certain regions of the CNS needed to be genetically male or female for
normal male courtship behavior to occur (Hall, 1977). Those studies showed that the
posterior midbrain, where the pC1 and pC2 clusters reside, were important for
reproductive behaviors in both sexes (Hall, 1977); (Tompkins and Hall, 1983).
Interestingly, the posterior brain region contains DSX-expressing cells in both male and
females, suggesting dsx function may underlie these early observations. Additionally, the
mesothoracic ganglion of the ventral nerve cord, where the TN1-region, DSX-cluster is
located, was shown to underlie male wing song (von Schilcher, 1979).

It was recently shown that DSX
M
and FRU
M
collaborate to establish the neural circuitry
that underlies wing song formation (Rideout et al., 2007), and DSX
M
is expressed in
neurons that comprise the fru neural circuit, which is necessary and sufficient for the
early steps of the male courtship ritual. Here, we observe extensive overlap of fru P1-
expressing cells and DSX expression in males in both the posterior brain and ventral
nerve cord (Figure 2; (Rideout et al., 2007)), as well as overlap in the abdominal
ganglion, as previously reported (Billeter et al., 2006). This overlap begins in certain
CNS regions as early as 8 hours APF and is maintained to the 0-24 hour adult stage.
Thus, the anatomical position of these sexually dimorphic neurons in both in the posterior
midbrain and the mesothoracic ganglion, as well as their presence in the fru neural
circuit, suggests that they may participate in establishing the neural circuits underlying
sex-specific behaviors.

162

fru P1 was shown to be sufficient to specify early, but not late, steps of the male
courtship ritual, when FRU
M
was expressed in females in homologously-positioned cells
in which it is normally expressed in males (Demir and Dickson, 2005; Manoli et al.,
2005). Additionally, gynandromorph studies showed that if the CNS was genetically
male, animals that contain female tissues in other body regions could perform all male
courtship steps (Hall, 1977). This suggests there are additional genes, other than fru P1,
that are sex-differentially utilized in the CNS that specify the correct neural circuitry for
male behavior. dsx is an excellent candidate, given the role dsx plays in establishing
sexually dimorphic numbers of neurons.

Our results may reconcile the observation that dsx is necessary, but not sufficient, for
specifying aspects of male courtship behaviors (Taylor et al., 1994; Villella and Hall,
1996). Because the male courtship ritual is a sequence-dependent series of sub-
behaviors, in which performance of late steps required completion of early steps, if
FRU
M
is required for establishing early male courtship steps, then in the absence of fru, it
would not be possible to assess if DSX
M
is sufficient to establish the potential for late
male behaviors. The majority of DSX-expressing cells do not reside in regions previously
mapped as important for the early steps, consistent with this idea (reviewed inGreenspan
and Ferveur, 2000).

The neural patterning that underlies the potential for female behaviors remains minimally
understood. In females, very few fru P1- expressing cells also express DSX
F
, with the
163

exception of the abdominal ganglion of the VNC. The observation that fru P1-expressing
cells play a role in female behaviors indicates that fru P1-expressing cells function in
female reproductive behaviors (Kvitsiani and Dickson, 2006). The observation that there
are very few DSX-expressing cells in the female brain, and they do not overlap the fru
P1-expressing brain cells, suggests DSX
F
is most likely not sufficient to specify female
behaviors, as was suggested by earlier results (Waterbury et al., 1999), but that DSX
F
-
and fru P1-expressing cells collaborate to bring about the potential for female behaviors.

Here, we have shown that DSX
F
-dependent, sex-specific cell death in the TN1 region is
the mechanism used to reach the correct number of DSX-expressing cells in females. It
has been shown that in females, a small set of fru P1-expressing neurons undergoes cell
death, thereby eliminating neurons that would go on to make male-typical projections
(Kimura et al., 2005). Additionally, DSX
F
has been shown to control cell death in the
developing embryonic gonad, which ultimately results in a sexual dimorphism in gonadal
tissues (DeFalco et al., 2003). Our results, along with other studies undertaken in C.
elegans and mammals (Conradt and Horvitz, 1999); (Davis et al., 1996), underscore the
importance of cell death as a mechanism by which differences between the sexes are
established.

164

4.5. Materials and Methods
Generation of polyclonal antibodies specific to DSX and FRU
M

We produced a recombinant glutathione S-transferase (GST) DSX fusion protein that
contained ~100 amino acids that are common to both DSX isoforms, by PCR amplifying
the region and cloning it into the pGEX 4T1 vector (GE Healthcare). The PCR primers
used are: 5’ CTCGAGCTCTTCGATTCGATTCGCCGGGAAGCCTCTTCAAT (Xho1
site is engineered at 5’ end) and 5’
GAATTCAGGTCATCGGGAACATCGGTGATCACTAGC (EcoR1 site is engineered
at 5’ end). The GST-DSX fusion protein was produced in E. coli, purified and used to
immunize a rat host (Josman, Napa Valley). The serum was affinity purified on a column
containing the recombinant GST-DSX fusion protein covalently coupled to amino resin
(Pierce). The FRU
M
antibody was generated as described in (Lee et al., 2000a), and was
shown to be specific to FRU
M
.

Immunohistochemistry
Whole mount immunochemistry experiments were performed as previously described
(Lee et al., 2000a). The primary polyclonal DSX antibody was diluted 1:50-1:100 in TNT
(0.1M Tris-HCl, 0.3M NaCl, 0.5% Triton X-100, pH 7.4). Secondary antibodies were
purchased from Molecular Probes and used at recommended dilutions. Confocal
microscopy was performed on a Zeiss Pascal. ELAV antibody was obtained from the
Iowa hybridoma bank. Cell counts of FRU
M
-expressing cells in 5-7 day old dsx null and
wild type males were performed blind by two independent members of the lab.
165

TUNEL assay for cell death.
The Apoalert DNA Fragmentation Assay kit (Clontech) was used with modifications (as
described in Firth et al., 2005).

Drosophila strains
Drosophila were grown on standard media containing cornmeal, dextrose, and yeast at
25°C. The wild type strain is Canton S. The dsx null genotype is transheterozygous for
the dsx alleles dsx
M+R15
and dsx
D+R3
. Generally, flies homozygous for the ix
3
allele were
not viable, although a small number of homozygous ix
3
flies escaped to the adult stage.
The loss-of-function ix genotype used here is transheterozygous for the ix
3
allele and a
Df(2R)en-B deficiency. The fru P1 null genotype is transheterozygous for the fru alleles
fru
P14
and fru
4-40
. The fru-P1-GAL4 driver was previously described (Stockinger et al.,
2005). The UAS-DSX
F
and UAS-DSX
M
lines were provided by Gyunghee Lee.
Genotypes of transgenic flies and flies mutant for cell death effectors are described in
Table 4.3; briefly, they include Df(3L)XR38 and Df(3L)H99 deficiencies (referred to as
XR38 and H99, respectively), described in (Peterson et al., 2002), α- amyloid protein
precursor-like (appl)-Gal4, tubulin -Gal4, actin-Gal4, elav-Gal4, heat shock-Gal4, UAS-
diap1, and UAS-p35. With the exception of the XR38, H99, appl-Gal4, UAS-diap, and
UAS-p35 strains, all were obtained from the Bloomington Drosophila Stock Center
(Indiana University, Bloomington, Indiana, United States). Pupal animals were sex-
sorted at the white pre-pupal stage, and then aged at 25º C.
166

BrdU labeling
Pupae were sex-sorted at the white pre-pupal stage and aged for eight hours. CNS tissues
were dissected in cold PBS, incubated for four hours in 15 µg/ml of BrdU in PBS at room
temperature, washed once in PBS, and then fixed in 3% paraformaldehyde in PBS for 20
minutes at room temperature. Samples were washed three times for five minutes in PBS,
three times for five minutes in TNT, and blocked for one hour in 4% normal goat serum.
To expose the BrdU antigen, samples were briefly boiled (4.5 minutes) at 100°C.
Samples were washed in PBS and incubated in mouse anti-BrdU (1:200 dilution, GE
Healthcare) and rat anti-DSX (1:50 dilution) in TNT overnight at 4°C on a rocker.
Samples were washed six times for 15 minutes each with TNT and then incubated in
secondary anti-rat and anti-mouse antibodies (Molecular Probes) diluted 1:500 in TNT
overnight at 4°C. Samples were washed six times in TNT for 15 minutes each and
mounted in VectaShield before imaging.

Statistical analyses
All cell counts are represented as mean ± standard error of the mean. Statistical
significance was calculated by unpaired, two-tailed Student t-tests and performed in
excel. The resulting P-value is reported.

167

Chapter 5: Conclusion and Future Directions
The goal of the research presented here is to understand the difficult but important
question of how sex-specific morphologies and behaviors are specified. The data and
analyses presented in Chapter 2 demonstrate that many distinct functional classes of
genes contribute to the development of males and females during metamorphosis.
Furthermore, the different stages during metamorphosis show distinct transcriptional
profiles, all of which underscore the exquisitely precise regulation that is necessary to
obtain the correct development of an organism. With the identification of somatic, sex-
differentially-expressed genes at distinct stages of development, it has been possible to
determine how sex-specific transcription might direct developmental processes.

To further characterize the transcriptional basis of somatic sex determination during
metamorphosis, more detailed studies of genes and gene classes could be performed.
Information such as the spatial location of the genes identified here could lead to a more
thorough understanding of somatic sex determination. Mutational analyses including the
abrogation or overexpression of these genes in particular tissues would provide insight
into the wild-type function. Although many genes and gene groups were identified as
being sex-differentially expressed in the soma, several warrant special consideration for
further analyses. The first gene group consists of the mitochondrial genes that are more
highly expressed in female somatic tissue than in male somatic tissue at the mid-pupal
stage. It would be worthwhile to test the hypothesis that mitochondria contribute to sex-
specific differences in somatic tissues. Secondly, the identification of two G-protein
168

coupled receptors, typically involved in signal transduction cascades affecting diverse
processes, as being expressed sex-differentially in the soma at the 0 hour wpp stage
warrants further investigation. These two genes, CG12290 and metabotropic GABA-B
receptor subtype 3 (GABA-B-R3), could be involved in mediating sex-specific
developmental processes, which could be characterized through more detailed studies.
And finally, several genes that are known to function in the nervous system, including
PAK-kinase, snap-25, rutabaga and vegetable, are more highly expressed in female
somatic tissue at 72 hours APF than in males. Currently, little is known about the neural
circuits that mediate female-specific behavior, and the characterization of these genes
may uncover a role for them in the development of female-specific neural networks.

Although much of the work presented here focused on somatic sex determination, the
findings presented in Chapter 2 concerning genes expressed in the male- and female-
germline could be used as a basis for further studies on the genetic specification of the
germline, and in particular, these genes’ role in speciation, since it is known that genes
that function in the germline are often under much less evolutionary constraint and
contribute to speciation. The dataset generated here could also be used to perform cross-
species comparisons to test if particular classes of genes are well-conserved among other
species, and thus, likely functionally constrained. Because genes that encode critical
products are more likely to be under high levels of constraint, their identification could
lead to insights about basic processes required for correct functioning of an organism.

169

Additionally, the genes identified here as functioning downstream of DSX at a mid-pupal
stage could be further characterized. Genetic analyses (i.e., loss-of-function and gain-of-
function) could be carried out to test if any of the targets may be in the same genetic
pathway as DSX, and whether they have any phenotypic effects on the animal that
partially recapitulate the dsx null phenotype. The finding that Abd-A, a well-studied
homeotic gene necessary for segmental identity, is regulated positively by DSX
F
and
negatively by DSX
M
, in contrast to the regulation of the vast majority of genes identified
here, warrants further investigation. It is known that DSX acts in concert with Abd-B in
abdominal pigmentation specification; thus it is possible that DSX and Abd-A may
similarly interact. Analyzing the relationship between these two proteins in both sexes
may uncover a novel interaction in several tissues, as it is known that both ABD-A and
DSX are expressed in the CNS and the genital discs. This fits well with the known
integration of patterning genes and the sex determination hierarchy.

The results presented in Chapter 3 show that a steroid hormone receptor, EcR, acts in the
fru P1 neural circuit in the CNSand/or PNS to specify correct male courtship behavior.
The transcriptional profiles of fru P1 mutants as compared to wild type males include
many genes that may be targets of fru P1 in the whole body and/or the CNS, many of
which could be further explored to extend the analyses presented here. As above, spatial
and functional assays would provide insight into the effector genes that have been
identified. In particular, the role of Broad, a direct target of EcR that regulates many
aspect of development including the specification of the CNS, and its relationship to fru
170

P1 could be assessed with more fine-scale studies in which individual neurons are
examined for morphological or functional phenotypes. The spatial overlap of FRU
M
and
Broad is particularly intriguing, as it is only in a sub-set of neurons. It is noteworthy that
both FRU
M
and Broad contain the BTB and Zinc Finger domains. The BTB domain has
been hypothesized to mediate protein-protein interactions. The presence of both FRU
M

and Broad in these neurons may confer unique properties onto them and give them an
important role in male-specific neural development.

One of the major results described in Chapter 3 is that males with diminished levels of
EcRA in fru P1-expressing neurons court other males vigorously; however, the basis for
their increased levels of male-male courtship is unknown. More detailed studies on the
neuronal architecture that underlies this phenotype would provide mechanistic insight
into the process of mate recognition. For example, neuronal survival, connectivity, or
activity could be altered in these mutants. It is known that a population of EcRA-
expressing cells undergoes cell death within the first 24 hours of adult life, and it is
shown here that several of these cells also express fru P1; it could be speculated that the
EcRA-dependent cell death could shape the fru P1-circuit through the programmed
demise of these cells. The identification of both the neuronal basis of this mutant
phenotype, and the sensory and/or integration systems that are affected, would expand
our understanding of mate-recognition, and more generally, the process by which
behavioral decisions are made.

171

The data presented in Chapter 4 demonstrates that DSX is able to pattern the sexually
dimorphic CNS through distinct mechanisms in a context-dependent manner. The next
level of research could focus on the molecular pathways through which DSX acts to
specify cell death in the VNC in females and cell division in the posterior brain of males.
The microarray data from DSX mutants generated in Chapter 2 provides gene lists on
which further investigations may be based. Additionally, it would be interesting to
examine how early the sexual dimorphism of DSX-expressing cells in the posterior brain
is established. The data presented here shows that the difference is present at the third
instar larval stage (Figure C4), but the point in development at which that dimorphism is
first specified is unknown.

Because DSX has such a context-dependent role in specifying cell number, it seems
likely that DSX-expressing cells themselves also have diverse functional roles. The
spatial pattern alone, with small, tightly-clustered populations of neurons, some of which
also express FRU
M
in males, leads to the hypothesis that individual cells likely have
distinct functions. To test this, a driver which specifically targets sub-groups of DSX
cells would be required (currently, we have been unable to make a transgenic fly that
possessed a DSX driver; see Appendix C, Table C2 and Figure C6). One of the most
intriguing extensions of the present research on DSX is its functional role in courtship
behavior. Do DSX-expressing cells in the CNS contribute to sex-specific behaviors, and
can certain sub-behaviors of the courtship ritual be ascribed to distinct populations of
these neurons? For example, are the neurons in the TN1 region of the VNC required for
172

the male wing song, and if so, are the neurons that express both DSX
M
and FRU
M
more
functionally important than the rest? In females, there are many fewer DSX-expressing
neurons than in males, and their functionality is unknown. The effectors of female
behavior have been much less studied, but are nonetheless important to understand.

Currently, data resulting from microarray experiments in which DSX and FRU
M
were
overexpressed (performed in collaboration with Matt Lebo and Tom Goldman) is being
analyzed for enriched DSX- and putative FRU
M
-binding sites using a computational
algorithm that searches for overrepresented binding sites (SUPRFly, created by Tom
Goldman and Matt Lebo, University of Southern California). The resulting data will
contribute to the understanding of how DSX and FRU
M
exert transcriptional control of
the genome to ultimately pattern male- and female-specific tissues.

With the sequencing of genomes, it was thought that the blueprints for the construction of
a fly, or a man, are known. However, identifying genetic regulators is only the first step
in a long and convoluted process to the true understanding of any biological system. The
interconnectivity and interdependence of diverse regulatory mechanisms confound
straightforward conclusions about developmental patterning, and make the attempt to
disentangle individual genes’ functions difficult. The work presented here will contribute
to understanding the genes, the cells, and the tissues that ultimately comprise an organism
and provide it with the potential for complex behavior.

173

References
Albagli, O., et al., 1995. The BTB/POZ domain: a new protein-protein interaction motif
common to DNA- and actin-binding proteins. Cell Growth Differ. 6, 1193-8.
Albin, S. D., Davis, G. W., 2004. Coordinating structural and functional synapse
development: postsynaptic p21-activated kinase independently specifies glutamate
receptor abundance and postsynaptic morphology. J Neurosci. 24, 6871-9.
Altstein, M., 2004. Role of neuropeptides in sex pheromone production in moths.
Peptides. 25, 1491-501.
Anand, A., et al., 2001. Molecular genetic dissection of the sex-specific and vital
functions of the Drosophila melanogaster sex determination gene fruitless. Genetics. 158,
1569-95.
Antoniewski, C., et al., 1993. Structural features critical to the activity of an ecdysone
receptor binding site. Insect Biochem Mol Biol. 23, 105-14.
Araki, T., Milbrandt, J., 1996. Ninjurin, a novel adhesion molecule, is induced by nerve
injury and promotes axonal growth. Neuron. 17, 353-61.
Arbeitman, M. N., et al., 2004. A genomic analysis of Drosophila somatic sexual
differentiation and its regulation. Development. 131, 2007-21.
Arbeitman, M. N., et al., 2002. Gene expression during the life cycle of Drosophila
melanogaster. Science. 297, 2270-5.
Aronson, J. F., et al., 1954. A comparison of giant X chromosomes in male and female
Drosophila Melanogaster by cytophotometry in the ultraviolet. J Histochem Cytochem. 2,
458-459.
Asha, H., et al., 2003. Analysis of Ras-induced overproliferation in Drosophila
hemocytes. Genetics. 163, 203-15.
Ashburner, M., 1972. Puffing patterns in Drosophila melanogaster and related species.
Results Probl Cell Differ. 4, 101-51.
Ashburner, M., et al., 2000. Gene ontology: tool for the unification of biology. The Gene
Ontology Consortium. Nat Genet. 25, 25-9.
Ashburner, M., et al., 2005. Drosophila : a laboratory handbook. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, N.Y.
Baehrecke, E. H., 1996. Ecdysone signaling cascade and regulation of Drosophila
metamorphosis. Arch Insect Biochem Physiol. 33, 231-44.
174

Baker, B. S., et al., 1989. Molecular genetic aspects of sex determination in Drosophila
melanogaster. Genome. 31, 638-45.
Baker, B. S., Ridge, K. A., 1980. Sex and the single cell. I. On the action of major loci
affecting sex determination in Drosophila melanogaster. Genetics. 94, 383-423.
Ball, G. F., Balthazart, J., 2004. Hormonal regulation of brain circuits mediating male
sexual behavior in birds. Physiol Behav. 83, 329-46.
Basi, G. S., et al., 1984. Alternative splicing of a Drosophila tropomyosin gene generates
muscle tropomyosin isoforms with different carboxy-terminal ends. Mol Cell Biol. 4,
2828-36.
Bastock, M., 1956. A Gene Mutation Which Changes a Behavior Pattern. Evolution. 10,
421-439.
Bastock, M., Manning, A., 1955. The Courtship of Drosophila Melanogaster. Behaviour.
8, 85-110.
Bayer, C. A., et al., 1996. A switch in broad-complex zinc-finger isoform expression is
regulated posttranscriptionally during the metamorphosis of Drosophila imaginal discs.
Dev Biol. 177, 1-14.
Bello, B. C., et al., 2003. A pulse of the Drosophila Hox protein Abdominal-A schedules
the end of neural proliferation via neuroblast apoptosis. Neuron. 37, 209-19.
Belote, J. M., Baker, B. S., 1987. Sexual behavior: its genetic control during development
and adulthood in Drosophila melanogaster. Proc Natl Acad Sci U S A. 84, 8026-30.
Belote, J. M., et al., 1989. Molecular genetics of transformer, a genetic switch controlling
sexual differentiation in Drosophila. Dev Genet. 10, 143-54.
Bender, M., et al., 1997. Drosophila ecdysone receptor mutations reveal functional
differences among receptor isoforms. Cell. 91, 777-88.
Billeter, J. C., et al., 2006. Isoform-specific control of male neuronal differentiation and
behavior in Drosophila by the fruitless gene. Curr Biol. 16, 1063-76.
Blau, J., Young, M. W., 1999. Cycling vrille expression is required for a functional
Drosophila clock. Cell. 99, 661-71.
Bodenstein, D., The postembryonic development of Drosophila. In: M. Demerec, (Ed.),
Biology of Drosophila. Cold Spring Harbor Laboratory Press, New York, 1950, pp. 275-
367.
175

Bodo, C., 2008. A role for the androgen receptor in the sexual differentiation of the
olfactory system in mice. Brain Res Rev. 57, 321-31.
Boggs, R. T., et al., 1987. Regulation of sexual differentiation in D. melanogaster via
alternative splicing of RNA from the transformer gene. Cell. 50, 739-47.
Bossy-Wetzel, E., et al., 2003. Mitochondrial fission in apoptosis, neurodegeneration and
aging. Curr Opin Cell Biol. 15, 706-16.
Boswell, R. E., Mahowald, A. P., 1985. tudor, a gene required for assembly of the germ
plasm in Drosophila melanogaster. Cell. 43, 97-104.
Bridges, C. B., 1921. Triploid Intersexes in Drosophila Melanogaster. Science. 54, 252-
254.
Brody, T., Cravchik, A., 2000. Drosophila melanogaster G protein-coupled receptors. J
Cell Biol. 150, F83-8.
Brown, H. L., et al., 2006. Use of time-lapse imaging and dominant negative receptors to
dissect the steroid receptor control of neuronal remodeling in Drosophila. Development.
133, 275-85.
Budnik, V., White, K., 1988. Catecholamine-containing neurons in Drosophila
melanogaster: distribution and development. J Comp Neurol. 268, 400-13.
Burtis, K. C., Baker, B. S., 1989. Drosophila doublesex gene controls somatic sexual
differentiation by producing alternatively spliced mRNAs encoding related sex-specific
polypeptides. Cell. 56, 997-1010.
Burtis, K. C., et al., 1991. The doublesex proteins of Drosophila melanogaster bind
directly to a sex-specific yolk protein gene enhancer. Embo J. 10, 2577-82.
Ceriani, M. F., et al., 2002. Genome-wide expression analysis in Drosophila reveals
genes controlling circadian behavior. J Neurosci. 22, 9305-19.
Chao, A. T., Guild, G. M., 1986. Molecular analysis of the ecdysterone-inducible 2B5
"early' puff in Drosophila melanogaster. Embo J. 5, 143-50.
Chase, B. A., Baker, B. S., 1995. A genetic analysis of intersex, a gene regulating sexual
differentiation in Drosophila melanogaster females. Genetics. 139, 1649-61.
Cherbas, L., et al., 2003. EcR isoforms in Drosophila: testing tissue-specific requirements
by targeted blockade and rescue. Development. 130, 271-84.
Chintapalli, V. R., et al., 2007. Using FlyAtlas to identify better Drosophila melanogaster
models of human disease. Nat Genet. 39, 715-20.
176

Choi, Y. J., et al., 2006. Programmed cell death mechanisms of identifiable peptidergic
neurons in Drosophila melanogaster. Development. 133, 2223-32.
Christiansen, A. E., et al., 2002. Sex comes in from the cold: the integration of sex and
pattern. Trends Genet. 18, 510-6.
Cline, T. W., 1978. Two closely linked mutations in Drosophila melanogaster that are
lethal to opposite sexes and interact with daughterless. Genetics. 90, 683-98.
Cline, T. W., Meyer, B. J., 1996. Vive la difference: males vs females in flies vs worms.
Annu Rev Genet. 30, 637-702.
Cohen, S. M., Imaginal Disc Morphogenesis. In: A. Martinas-Arias, M. Bates, Eds.), The
Development of Drosophila melanogaster. Cold Spring Harbor Press, New York, 1993,
pp. 747–841.
Colombani, J., et al., 2005. Antagonistic actions of ecdysone and insulins determine final
size in Drosophila. Science. 310, 667-70.
Conradt, B., Horvitz, H. R., 1999. The TRA-1A sex determination protein of C. elegans
regulates sexually dimorphic cell deaths by repressing the egl-1 cell death activator gene.
Cell. 98, 317-27.
Consoulas, C., et al., 2002. Dendritic remodeling and growth of motoneurons during
metamorphosis of Drosophila melanogaster. J Neurosci. 22, 4906-17.
Crosby, M. A., et al., 2007. FlyBase: genomes by the dozen. Nucleic Acids Res. 35,
D486-91.
Crossgrove, K., et al., 1996. The Drosophila Broad-Complex early gene directly regulates
late gene transcription during the ecdysone-induced puffing cascade. Dev Biol. 180, 745-
58.
Datta, S. R., et al., 2008. The Drosophila pheromone cVA activates a sexually dimorphic
neural circuit. Nature. 452, 473-7.
Dauwalder, B., et al., 2002. The Drosophila takeout gene is regulated by the somatic sex-
determination pathway and affects male courtship behavior. Genes Dev. 16, 2879-92.
Davis, E. C., et al., 1996. The role of apoptosis in sexual differentiation of the rat
sexually dimorphic nucleus of the preoptic area. Brain Res. 734, 10-8.
DeFalco, T. J., et al., 2003. Sex-specific apoptosis regulates sexual dimorphism in the
Drosophila embryonic gonad. Dev Cell. 5, 205-16.
Demerec, M., 1994. Biology of Drosophila. CSHL Press. 632.
177

Demir, E., Dickson, B. J., 2005. fruitless splicing specifies male courtship behavior in
Drosophila. Cell. 121, 785-94.
Dennis, G., Jr., et al., 2003. DAVID: Database for Annotation, Visualization, and
Integrated Discovery. Genome Biol. 4, P3.
Deutsch, J., et al., 1989. Larval fat body-specific gene expression in D. melanogaster.
Dev Genet. 10, 220-31.
DiBartolomeis, S. M., et al., 2002. Molecular analysis of the Drosophila miniature-dusky
( m-dy) gene complex: m-dy mRNAs encode transmembrane proteins with similarity to
C. elegans cuticulin. Mol Genet Genomics. 267, 564-76.
DiBello, P. R., et al., 1991. The Drosophila Broad-Complex encodes a family of related
proteins containing zinc fingers. Genetics. 129, 385-97.
Dodson, M. W., Guo, M., 2007. Pink1, Parkin, DJ-1 and mitochondrial dysfunction in
Parkinson's disease. Curr Opin Neurobiol. 17, 331-7.
Drapeau, M. D., et al., 2003. A gene necessary for normal male courtship, yellow, acts
downstream of fruitless in the Drosophila melanogaster larval brain. J Neurobiol. 55, 53-
72.
Duerr, J. S., Quinn, W. G., 1982. Three Drosophila mutations that block associative
learning also affect habituation and sensitization. Proc Natl Acad Sci U S A. 79, 3646-50.
Duncan, I. W., Kaufman, T. C., 1975. Cytogenic analysis of chromosome 3 in Drosophila
melanogaster: mapping of the proximal portion of the right arm. Genetics. 80, 733-52.
Dyanov, H. M., Dzitoeva, S. G., 1995. Method for attachment of microscopic
preparations on glass for in situ hybridization, PRINS and in situ PCR studies.
Biotechniques. 18, 822-4, 826.
Eisen, M. B., et al., 1998. Cluster analysis and display of genome-wide expression
patterns. Proc Natl Acad Sci U S A. 95, 14863-8.
Emery, I. F., et al., 1994. Differential expression of Broad-Complex transcription factors
may forecast tissue-specific developmental fates during Drosophila metamorphosis.
Development. 120, 3275-87.
Erdman, S. E., Burtis, K. C., 1993. The Drosophila doublesex proteins share a novel zinc
finger related DNA binding domain. Embo J. 12, 527-35.
Erdman, S. E., et al., 1996. Functional and genetic characterization of the oligomerization
and DNA binding properties of the Drosophila doublesex proteins. Genetics. 144, 1639-
52.
178

Ferveur, J. F., 2005. Cuticular hydrocarbons: their evolution and roles in Drosophila
pheromonal communication. Behav Genet. 35, 279-95.
Ffrench-Constant, R. H., et al., 1991. Molecular cloning and transformation of cyclodiene
resistance in Drosophila: an invertebrate gamma-aminobutyric acid subtype A receptor
locus. Proc Natl Acad Sci U S A. 88, 7209-13.
Finley, K. D., et al., 1997. dissatisfaction, a gene involved in sex-specific behavior and
neural development of Drosophila melanogaster. Proc Natl Acad Sci U S A. 94, 913-8.
Firth, L. C., et al., 2005. Analyses of RAS Regulation of Eye Development in Drosophila
melanogaster. Methods Enzymol. 407, 711-21.
Flower, D. R., et al., 2000. The lipocalin protein family: structural and sequence
overview. Biochim Biophys Acta. 1482, 9-24.
Franke, A., Baker, B. S., 1999. The rox1 and rox2 RNAs are essential components of the
compensasome, which mediates dosage compensation in Drosophila. Mol Cell. 4, 117-
22.
Freeland, D. E., Kuhn, D. T., 1996. Expression patterns of developmental genes reveal
segment and parasegment organization of D. melanogaster genital discs. Mech Dev. 56,
61-72.
Fuller, M. T., Spermatogenesis in Drosophila. In: M. Bate, A. Martinez-Arias, Eds.), The
Development of Drosophila melanogaster. Cold Spring Harbor Laboratory Press, New
York, 1993, pp. 71–147.
Fusani, L., 2008. Testosterone control of male courtship in birds. Horm Behav.
Gailey, D. A., Hall, J. C., 1989. Behavior and cytogenetics of fruitless in Drosophila
melanogaster: different courtship defects caused by separate, closely linked lesions.
Genetics. 121, 773-85.
Gailey, D. A., et al., 1991. Elements of the fruitless locus regulate development of the
muscle of Lawrence, a male-specific structure in the abdomen of Drosophila
melanogaster adults. Development. 113, 879-90.
Ganter, G. K., et al., 2007. Increased male-male courtship in ecdysone receptor deficient
adult flies. Behav Genet. 37, 507-12.
Garrett-Engele, C. M., et al., 2002. intersex, a gene required for female sexual
development in Drosophila, is expressed in both sexes and functions together with
doublesex to regulate terminal differentiation. Development. 129, 4661-75.
179

Gebauer, F., et al., 2003. Drosophila sex-lethal inhibits the stable association of the 40S
ribosomal subunit with msl-2 mRNA. Mol Cell. 11, 1397-404.
Gentleman, R. C., et al., 2004. Bioconductor: open software development for
computational biology and bioinformatics. Genome Biol. 5, R80.
Gill, K. S., 1963. A mutation causing abnormal courtship and mating behavior in males
of. Drosophila melanogaster. American Zoology. 3, 507.
Goldman, T. D., Arbeitman, M. N., 2007. Genomic and Functional Studies of Drosophila
Sex Hierarchy Regulated Gene Expression in Adult Head and Nervous System Tissues.
PLoS Genet. 3, e216.
Graham, P., et al., 2003. Masters change, slaves remain. Bioessays. 25, 1-4.
Greenspan, R. J., Ferveur, J. F., 2000. Courtship in Drosophila. Annu Rev Genet. 34,
205-232.
Hall, J. C., 1977. Portions of the central nervous system controlling reproductive
behavior in Drosophila melanogaster. Behav Genet. 7, 291-312.
Hall, J. C., 1979. Control of male reproductive behavior by the central nervous system of
Drosophila: dissection of a courtship pathway by genetic mosaics. Genetics. 92, 437-57.
Harrison, J. B., et al., 1996. Immunocytochemical mapping of a C-terminus anti-peptide
antibody to the GABA receptor subunit, RDL in the nervous system in Drosophila
melanogaster. Cell Tissue Res. 284, 269-78.
Heinrichs, V., et al., 1998. Regulation of sex-specific selection of fruitless 5' splice sites
by transformer and transformer-2. Mol Cell Biol. 18, 450-8.
Hekmat-Scafe, D. S., et al., 2005. Seizure suppression by gain-of-function escargot
mutations. Genetics. 169, 1477-93.
Hildreth, P. E., 1965. Doublesex, Recessive Gene That Transforms Both Males and
Females of Drosophila into Intersexes. Genetics. 51, 659-78.
Holloway, A. K., Begun, D. J., 2004. Molecular evolution and population genetics of
duplicated accessory gland protein genes in Drosophila. Mol Biol Evol. 21, 1625-8.
Hotta, Y., Benzer, S., 1976. Courtship in Drosophila mosaics: sex-specific foci for
sequential action patterns. Proc Natl Acad Sci U S A. 73, 4154-8.
Huet, F., et al., 1995. Sequential gene activation by ecdysone in Drosophila
melanogaster: the hierarchical equivalence of early and early late genes. Development.
121, 1195-204.
180

Jallon, J. M., Hotta, Y., 1979. Genetic and behavioral studies of female sex appeal in
Drosophila. Behav Genet. 9, 257-75.
Jiang, Z., et al., 2005. Pattern profiling and mapping of the fat body transcriptome in
Drosophila melanogaster. Obes Res. 13, 1898-904.
Kadener, S., et al., 2006. Neurotoxic protein expression reveals connections between the
circadian clock and mating behavior in Drosophila. Proc Natl Acad Sci U S A. 103,
13537-42.
Kanehisa, M., et al., 2008. KEGG for linking genomes to life and the environment.
Nucleic Acids Res. 36, D480-4.
Karlsson, C., et al., 2004. Proteomic analysis of the Drosophila larval hemolymph clot. J
Biol Chem. 279, 52033-41.
Kearney, J. B., et al., 2004. Gene expression profiling of the developing Drosophila CNS
midline cells. Dev Biol. 275, 473-92.
Kelley, R. L., et al., 1995. Expression of msl-2 causes assembly of dosage compensation
regulators on the X chromosomes and female lethality in Drosophila. Cell. 81, 867-77.
Kelly, L. E., et al., 1997. Identification of a gene family from Drosophila melanogaster
encoding proteins with homology to invertebrate sarcoplasmic calcium-binding proteins
(SCPS). Insect Biochem Mol Biol. 27, 783-92.
Kim, Y. O., et al., 2004. A functional genomic screen for cardiogenic genes using RNA
interference in developing Drosophila embryos. Proc Natl Acad Sci U S A. 101, 159-64.
Kimura, K., et al., 2005. Fruitless specifies sexually dimorphic neural circuitry in the
Drosophila brain. Nature. 438, 229-33.
Koelle, M. R., et al., 1991. The Drosophila EcR gene encodes an ecdysone receptor, a
new member of the steroid receptor superfamily. Cell. 67, 59-77.
Kopp, A., et al., 2000. Genetic control and evolution of sexually dimorphic characters in
Drosophila. Nature. 408, 553-9.
Kurtovic, A., et al., 2007. A single class of olfactory neurons mediates behavioural
responses to a Drosophila sex pheromone. Nature. 446, 542-6.
Kvitsiani, D., Dickson, B. J., 2006. Shared neural circuitry for female and male sexual
behaviours in Drosophila. Current Biology. 16, R355-R356.
Kyriacou, C. P., Hall, J. C., 1984. Learning and memory mutations impair acoustic
priming of mating behaviour in Drosophila. Nature. 308, 62 - 65.
181

Lacaille, F., et al., 2007. An inhibitory sex pheromone tastes bitter for Drosophila males.
PLoS ONE. 2, e661.
Lee, G., et al., 2006. Sex- and clock-controlled expression of the neuropeptide F gene in
Drosophila. Proc Natl Acad Sci U S A. 103, 12580-5.
Lee, G., et al., 2000a. Spatial, temporal, and sexually dimorphic expression patterns of
the fruitless gene in the Drosophila central nervous system. J Neurobiol. 43, 404-26.
Lee, G., et al., 2002. Doublesex gene expression in the central nervous system of
Drosophila melanogaster. J Neurogenet. 16, 229-48.
Lee, T., et al., 2000b. Cell-autonomous requirement of the USP/EcR-B ecdysone receptor
for mushroom body neuronal remodeling in Drosophila. Neuron. 28, 807-18.
Letunic, I., et al., 2004. SMART 4.0: towards genomic data integration. Nucleic Acids
Res. 32, D142-4.
Mackay, T. F., et al., 2005. Genetics and genomics of Drosophila mating behavior. Proc
Natl Acad Sci U S A. 102 Suppl 1, 6622-9.
Madhavan, M. M., Schneiderman, H. A., 1977. Histological analysis of the dynamics of
growth of imaginal discs and histoblast nests during the larval development of Drosophila
melanogaster. Wilhelm Roux's Archives. 183, 269-305.
Manoli, D. S., et al., 2005. Male-specific fruitless specifies the neural substrates of
Drosophila courtship behaviour. Nature. 436, 395-400.
Manoli, D. S., et al., 2006. Blueprints for behavior: genetic specification of neural
circuitry for innate behaviors. Trends Neurosci. 29, 444-51.
McCudden, C. R., et al., 2005. G-protein signaling: back to the future. Cell Mol Life Sci.
62, 551-77.
McKeown, M., et al., 1988. Ectopic expression of the female transformer gene product
leads to female differentiation of chromosomally male Drosophila. Cell. 53, 887-95.
Mendjan, S., Akhtar, A., 2007. The right dose for every sex. Chromosoma. 116, 95-106.
Mezler, M., et al., 2001. Cloning and functional expression of GABA(B) receptors from
Drosophila. Eur J Neurosci. 13, 477-86.
Michelson, A. M., 1994. Muscle pattern diversification in Drosophila is determined by
the autonomous function of homeotic genes in the embryonic mesoderm. Development.
120, 755-68.
182

Mirth, C. K., Riddiford, L. M., 2007. Size assessment and growth control: how adult size
is determined in insects. Bioessays. 29, 344-55.
Mouillet, J. F., et al., 2001. Differential control of gene activity by isoforms A, B1 and
B2 of the Drosophila ecdysone receptor. Eur J Biochem. 268, 1811-9.
Mueller, J. L., et al., 2005. Cross-species comparison of Drosophila male accessory gland
protein genes. Genetics. 171, 131-43.
Mulder, N. J., et al., 2007. New developments in the InterPro database. Nucleic Acids
Res. 35, D224-8.
Norga, K. K., et al., 2003. Quantitative analysis of bristle number in Drosophila mutants
identifies genes involved in neural development. Curr Biol. 13, 1388-96.
Parisi, M., et al., 2004. A survey of ovary-, testis-, and soma-biased gene expression in
Drosophila melanogaster adults. Genome Biol. 5, R40.
Parisi, M., et al., 2003. Paucity of genes on the Drosophila X chromosome showing male-
biased expression. Science. 299, 697-700.
Pereanu, W., Hartenstein, V., Neural Lineages of the Drosophila Brain: A Three-
Dimensional Digital Atlas of the Pattern of Lineage Location and Projection at the Late
Larval Stage. Vol. 26, 2006, pp. 5534-5553.
Perez-Torrado, R., et al., 2006. Born to bind: the BTB protein-protein interaction domain.
Bioessays. 28, 1194-202.
Peterson, C., et al., 2002. reaper is required for neuroblast apoptosis during Drosophila
development. Development. 129, 1467-76.
Poulson, D. F., 1935. Oxygen consumption of Drosophila pupae. Journal of Comparative
Physiology A: Neuroethology, Sensory, Neural, and Behavioral Physiology. 22, 466-472.
Powell, D., et al., 1984. Regulation of synthesis of the larval serum proteins of
Drosophila melanogaster. Dev Biol. 102, 206-15.
Prokopenko, S. N., et al., 2000. Mutations affecting the development of the peripheral
nervous system in Drosophila: a molecular screen for novel proteins. Genetics. 156,
1691-715.
Rao, S. S., et al., 2001. Two distinct effects on neurotransmission in a temperature-
sensitive SNAP-25 mutant. Embo J. 20, 6761-71.
Raymond, C. S., et al., 1998. Evidence for evolutionary conservation of sex-determining
genes. Nature. 391, 691-5.
183

Rebers, J. E., Willis, J. H., 2001. A conserved domain in arthropod cuticular proteins
binds chitin. Insect Biochem Mol Biol. 31, 1083-93.
Restifo, L. L., White, K., 1991. Mutations in a steroid hormone-regulated gene disrupt
the metamorphosis of the central nervous system in Drosophila. Dev Biol. 148, 174-94.
Riddiford, L. M., 1993. Hormone receptors and the regulation of insect metamorphosis.
Receptor. 3, 203-9.
Riddiford, L. M., et al., 2000. Ecdysone receptors and their biological actions. Vitam
Horm. 60, 1-73.
Rideout, E. J., et al., 2007. The Sex-Determination Genes fruitless and doublesex Specify
a Neural Substrate Required for Courtship Song. Curr Biol.
Roberts, D. B., Evans-Roberts, S., 1979. The X-linked alpha-chain gene of Drosophila
LSP-1 does not show dosage compensation. Nature. 280, 691-2.
Robinow, S., et al., Programmed cell death in the Drosophila CNS is ecdysone-regulated
and coupled with a specific ecdysone receptor isoform. Vol. 119, 1993, pp. 1251-1259.
Robinow, S., White, K., 1991. Characterization and spatial distribution of the ELAV
protein during Drosophila melanogaster development. J Neurobiol. 22, 443-61.
Roignant, J. Y., et al., 2003. Absence of transitive and systemic pathways allows cell-
specific and isoform-specific RNAi in Drosophila. Rna. 9, 299-308.
Rouillard, J. M., et al., 2003. OligoArray 2.0: design of oligonucleotide probes for DNA
microarrays using a thermodynamic approach. Nucleic Acids Res. 31, 3057-62.
Ryner, L. C., et al., 1996. Control of male sexual behavior and sexual orientation in
Drosophila by the fruitless gene. Cell. 87, 1079-89.
Saldanha, A. J., 2004. Java Treeview--extensible visualization of microarray data.
Bioinformatics. 20, 3246-8.
Salz, H. K., et al., 1987. Functional changes associated with structural alterations induced
by mobilization of a P element inserted in the Sex-lethal gene of Drosophila. Genetics.
117, 221-31.
Sanchez-Herrero, E., et al., 1985. Genetic organization of Drosophila bithorax complex.
Nature. 313, 108-13.
Sanchez, D., et al., 2000. Characterization of two novel lipocalins expressed in the
Drosophila embryonic nervous system. Int J Dev Biol. 44, 349-59.
184

Saudan, P., et al., 2002. Ductus ejaculatorius peptide 99B (DUP99B), a novel Drosophila
melanogaster sex-peptide pheromone. Eur J Biochem. 269, 989-97.
Schubiger, M., et al., 2005. Ligand-dependent de-repression via EcR/USP acts as a gate
to coordinate the differentiation of sensory neurons in the Drosophila wing.
Development. 132, 5239-48.
Schubiger, M., et al., 2003. Isoform specific control of gene activity in vivo by the
Drosophila ecdysone receptor. Mech Dev. 120, 909-18.
Shirangi, T. R., et al., 2006. A double-switch system regulates male courtship behavior in
male and female Drosophila melanogaster. Nat Genet. 38, 1435-9.
Siepel, A., et al., 2005. Evolutionarily conserved elements in vertebrate, insect, worm,
and yeast genomes. Genome Res. 15, 1034-50.
Simerly, R. B., 2002. Wired for reproduction: organization and development of sexually
dimorphic circuits in the mammalian forebrain. Annu Rev Neurosci. 25, 507-36.
Simmerl, E., et al., 1995. Structure and regulation of a gene cluster for male accessory
gland transcripts in Drosophila melanogaster. Insect Biochem Mol Biol. 25, 127-37.
Singh, K., Singh, R. N., 1999. Metamorphosis of the central nervous system of
Drosophila melanogaster Meigen (Diptera : Drosophilidae) during pupation. Journal of
Biosciences. 24, 345-60.
Smyth, G. K., Limma: linear models for microarray data. In: V. C. R. Gentleman, S.
Dudoit, R. Irizarry, W. Huber, (Ed.), Bioinformatics and Computational Biology
Solutions using R and Bioconductor. Springer, New York, 2005, pp. 397-420.
Smyth, G. K., Speed, T., 2003. Normalization of cDNA microarray data. Methods. 31,
265-73.
Song, H. J., et al., 2002. The fruitless gene is required for the proper formation of axonal
tracts in the embryonic central nervous system of Drosophila. Genetics. 162, 1703-24.
Sosnowski, B. A., et al., 1989. Sex-specific alternative splicing of RNA from the
transformer gene results from sequence-dependent splice site blockage. Cell. 58, 449-59.
Spralding, A. C., Developmental genetics of oogenesis. In: A. Martinas-Arias, M. Bates,
Eds.), The Develpment of Drosophila. Cold Spring Harbor Press
New York, 1993, pp. 1-70.
Stockinger, P., et al., 2005. Neural circuitry that governs Drosophila male courtship
behavior. Cell. 121, 795-807.
185

Storey, J. D., 2003. The positive flase discovery rate: A Bayesian interpretation and the
q-value. Annals of Statistics. 2013-2035.
Sturtevant, A. H., 1945. A Gene in Drosophila Melanogaster That Transforms Females
into Males. Genetics. 30, 297-9.
Szuplewski, S., et al., 2003. The Drosophila bZIP transcription factor Vrille is involved
in hair and cell growth. Development. 130, 3651-62.
Talbot, W. S., et al., 1993. Drosophila tissues with different metamorphic responses to
ecdysone express different ecdysone receptor isoforms. Cell. 73, 1323-37.
Taylor, B. J., Truman, J. W., 1992. Commitment of abdominal neuroblasts in Drosophila
to a male or female fate is dependent on genes of the sex-determining hierarchy.
Development. 114, 625-42.
Taylor, B. J., et al., 1994. Behavioral and neurobiological implications of sex-
determining factors in Drosophila. Dev Genet. 15, 275-96.
Tiong, S., et al., 1985. Recessive lethal mutations within the bithorax-complex in
Drosophila. Mol Gen Genet. 200, 335-42.
Tompkins, L., Hall, J. C., 1983. Identification of Brain Sites Controlling Female
Receptivity in Mosaics of DROSOPHILA MELANOGASTER. Genetics. 103, 179-195.
Truman, J. W., 1990. Metamorphosis of the central nervous system of Drosophila. J
Neurobiol. 21, 1072-84.
Truman, J. W., Bate, M., 1988. Spatial and temporal patterns of neurogenesis in the
central nervous system of Drosophila melanogaster. Dev Biol. 125, 145-57.
Truman, J. W., et al., 1994. Ecdysone receptor expression in the CNS correlates with
stage-specific responses to ecdysteroids during Drosophila and Manduca development.
Development. 120, 219-34.
UniProt Consortium, T., 2008. The universal protein resource (UniProt). Nucleic Acids
Res. 36, D190-5.
Usui-Aoki, K., et al., 2000. Formation of the male-specific muscle in female Drosophila
by ectopic fruitless expression. Nat Cell Biol. 2, 500-6.
Valles, A. M., White, K., 1988. Serotonin-containing neurons in Drosophila
melanogaster: development and distribution. J Comp Neurol. 268, 414-28.
Verstreken, P., et al., 2005. Synaptic mitochondria are critical for mobilization of reserve
pool vesicles at Drosophila neuromuscular junctions. Neuron. 47, 365-78.
186

Vierstraete, E., et al., 2003. Proteomics in Drosophila melanogaster: first 2D database of
larval hemolymph proteins. Biochem Biophys Res Commun. 304, 831-8.
Vigoreaux, J. O., et al., 1993. Flightin, a novel myofibrillar protein of Drosophila stretch-
activated muscles. J Cell Biol. 121, 587-98.
Villella, A., et al., 2005. Functional analysis of fruitless gene expression by transgenic
manipulations of Drosophila courtship. Proc Natl Acad Sci U S A. 102, 16550-7.
Villella, A., et al., 1997. Extended reproductive roles of the fruitless gene in Drosophila
melanogaster revealed by behavioral analysis of new fru mutants. Genetics. 147, 1107-
30.
Villella, A., Hall, J. C., 1996. Courtship anomalies caused by doublesex mutations in
Drosophila melanogaster. Genetics. 143, 331-44.
Volff, J. N., et al., 2003. Evolutionary dynamics of the DM domain gene family in
metazoans. J Mol Evol. 57 Suppl 1, S241-9.
Von Schilcher, F., 1976. The behavior of cacophony, a courtship song mutant in
Drosophila melanogaster. Behav Biol. 17, 187-96.
von Schilcher, F. a. H., J.C., 1979. Neural topography of courtship song in sex mosaics of
Drosophila Melanogaster. J Comp Physiol. 129, 85-95.
Vosshall, L. B., 2007. Into the mind of a fly. Nature. 450, 193-7.
Waterbury, J. A., et al., 1999. Analysis of the doublesex female protein in Drosophila
melanogaster: role on sexual differentiation and behavior and dependence on intersex.
Genetics. 152, 1653-67.
White, K., et al., 1986. Neuropeptide-FMRFamide-like immunoreactivity in Drosophila:
development and distribution. J Comp Neurol. 247, 430-8.
White, K. P., et al., 1999. Microarray analysis of Drosophila development during
metamorphosis. Science. 286, 2179-84.
Williams, J. A., Carroll, S. B., 1993. The origin, patterning and evoltion of insect
appendages. Bioessays. 15, 567-577.
Williamson, A., Lehmann, R., 1996. Germ cell development in Drosophila. Annu Rev
Cell Dev Biol. 12, 365-91.
Yang, Y., et al., 2008. Doublesex and the regulation of sexual dimorphism in Drosophila
melanogaster: structure, function, and mutagenesis of a female-specific domain. J Biol
Chem. 283, 7280-92.
187

Yanowitz, J. L., et al., 1999. An N-terminal truncation uncouples the sex-transforming
and dosage compensation functions of sex-lethal. Mol Cell Biol. 19, 3018-28.
Yao, T. P., et al., 1993. Functional ecdysone receptor is the product of EcR and
Ultraspiracle genes. Nature. 366, 476-9.
Zhang, L. I., Poo, M. M., 2001. Electrical activity and development of neural circuits. Nat
Neurosci. 4 Suppl, 1207-14.
Zhang, S., et al., 2006. An MMP liberates the Ninjurin A ectodomain to signal a loss of
cell adhesion. Genes Dev. 20, 1899-910.
Zollman, S., et al., 1994. The BTB domain, found primarily in zinc finger proteins,
defines an evolutionarily conserved family that includes several developmentally
regulated genes in Drosophila. Proc Natl Acad Sci U S A. 91, 10717-21.

188

Appendices
Appendix A: Supplemental Data for Chapter 2
This Appendix includes one figure and three tables.

Figure A1. Clusters were generated using gene expression data from male and female tud
progeny at five time points during metamorphosis. For each cluster, the abscissa
indicates the five time points during metamorphosis examined (0, 24, 48, 72, and 96 hour
APF, and the ordinate represents the average expression value for each genotype
examined. Expression profiles for each cluster were generated by averaging the gene
expression data at each time point for every gene in the cluster in both sexes. Average
expression values are represented by teal for wild type males, blue for wild type females,
green for tud males, and red for tud females. Functional categories that were over-
represented among the genes in the cluster, as determined by the program DAVID
(P<0.05 (Dennis et al., 2003)), are described in Figure 2.1.

189

Figure A1

190

Figure A1, continued

191

Figure A1, continued

192

Table A1. Genes with sex-differential expression across metamorphosis, as identified by
a two-way ANOVA analysis with sex and time as the independent factors.

193

Table A2. Additional oligonucleotide sequences for sex determination hierarchy array
elements on the microarrays.

Gene Oligo Sequence
Sxl agacactcactgactcttaagatagtatgtagtttttatttgcacgggggggcaaacgcgccacgtccgc
tra gagaagctaggacaataggactctcaactgcgcattacgtggattccgtctccgacgatgcgccaaacac
dsx
F
gccaaatatgtcgatgtgtgacagccgttctacgcgtcagctttcttcaatcaacattaccccgtgctga
dsx
M
ctcgagtggaaataaatcgcactgtagcccagatctactacaactactacaccccgatggccctggtgaa
fru
P1 tgttttagtcggtcctttcgcgcttgacttgttttgcaactgtgtgcgtacgtttgagtgtgcgagtgcc
fru
A
atgcagtcgccagcagcacatgatgtcccactattcgccgcatcatccgcaccatcgatccctcatagat
fru
B
ccacctatacgcgcagcgacaatttgcgcacccactgcaagttcaagcatcccatgtacaatccggatac
fru
C
agcaactggagcagctggccataattcgcatcacaccatgtcgtaccacaacatgttcacgccgtcccgc

194

Table A3. Correlation among microarray replicates for the time course analysis of gene
expression in male and female tud progeny, and male and female wild type pupae at five
time points during metamorphosis (0, 24, 48, 72, and 96 hour APF).

Genotpye Sex
Hours
APF Slide 1-2 Slide 2-3 Slide 1-3
Wild type Male 0 0.961116 0.886478 0.860469
24 0.735882 0.912683 0.685898
48 0.914332 0.916945 0.948516
72 0.906889 0.819678 0.850662
96 0.892545 0.836798 0.910705

Wild type Female 0 0.899597 0.942938 0.922723
24 0.907994 0.894382 0.862571
48 0.855780 0.905179 0.852510
72 0.839065 0.904321 0.803556
96 0.922572 0.933136 0.919125

tud Male 0 0.943728 0.946911 0.953675
24 0.946276 0.943197 0.936378
48 0.885300 0.827935 0.854405
72 0.819428 0.889879 0.806995
96 0.829370 0.784110 0.753727

tud Female 0 0.958525 0.926813 0.945192
24 0.929511 0.929742 0.915770
48 0.899306 0.863057 0.785371
72 0.820626 0.824227 0.894708
96 0.867795 0.901358 0.856139

195

Table A4. Number of genes expressed at each of the five metamorphosis stages
examined (0, 24, 48, 72, and 96 hour APF), in male and female tud progeny and male and
female wild type pupae.

Genotype Sex
Hours
APF
# of
Genes
Wild type Female 0 7463
24 7125
48 6768
72 8555
96 5628

Wild type Male 0 6084
24 6766
48 5320
72 5354
96 7352

tud Female 0 6698
24 5601
48 4878
72 5085
96 5474
tud Male
0 6286
24 5574
48 6060
72 7151
96 4404

196

Appendix B: Supplemental data for Chapter 3

This appendix contains one table.

Table B1. List of genes that are identified from the literature to be regulated by
ecdysone.

197

Table B1, continued

198

Appendix C: Supplemental Data for Chapter 4
This Appendix includes two tables and six figures.

Table C1. DSX-expressing cells in wild type males and females and mutants at various
timepoints during development. Hours are shown on the left side of the table, and
genotypes are listed at the top. Cell counts are shown first, followed by the standard error
of the mean, indicated by “SE”.
199

Table C1
200

Figure C1. DSX expression in imaginal discs and CNS of third instar larvae. For all
panels, chromosomal sex is indicated in the bottom right. DSX has expression in the leg
discs (A and B), brain and ventral nerve cord (C), wing disc (D), and genital discs (E and
F).

201

Figure C2 (below). DSX is expressed in the male testis, but not the female ovary. DSX
expression is not present in the ovary (A) of a 5-7 day old adult female, but is present in
the testis (B) of a 5-7 day old adult male. Co-labeling of DSX (red) and an antibody
made to detect VASA, a protein that is specifically expressed in germ cells, (green) show
no co-localization (C).

Figure C3 (below). ix
3
/Df(2R) female brain (left) has more DSX-expressing cells than a
wild type female (right). This suggests that DSX
F
is required to negatively regulate the
number of DSX-expressing cells in the poseterior brain in females.

202

Figure C4. BrdU incorporation in the CNS of females (A and C) and males (B and D)
between 8 and 12 hours after puparium formation. No distinct differences between the
sexes are present; highest levels of cell division are observed in the anterior optic lobes.

203

Figure C5. fru P1-expressing cells may synapse onto DSX-expressing cells, although
further studies are needed to reach this conclusion. DSX expression (red) and synapses of
fru P1-expressing cells (green) are in close proximity. fru P1-GAL4 driving GFP at
synapses as marked by NSF in the brain (A), and the synaptic marker RAB5 in the pC1
region of the brain (B) and TN1 region of the VNC (C). This proximity may be a result
of the GFP protein being sequestered in the nucleus of cells that co-express DSX and fru
P1, and may not represent a true synapse. Further studies are required to understand if
these are true functional synapses.

204

Table C2. Analysis of DSX-GAL4 lines that were generated by cloning the 5 kb region
upstream of DSX into the pPT-GAL4 vector (performed by Michelle Arbeitman,
University of Southern California). Lines were balanced and tested for the DSX
expression pattern by crossing them to flies carrying the reporter gene UAS-nls-GFP,
which drives nuclear localized GFP under the control of the GAL4 element. None of the
lines recapitulated the endogenous DSX expression pattern. All data is from 0-24 hour
adult Canton S (wild type) strains.

Line number
(for
identification)
Chromosome
Location of
P-element
Expression pattern Expression in the
CNS
Expression in
DSX-expressing
cells in the CNS
(as tested with
DSX antibody)
4 3 No No -
11 - Very bright staining in
proboscis, antennae,
abdomen, eye and genitalia
Yes, evenly dispersed
cells predominantly on
outside of CNS tissue
No
14 3 Slight expression in head. Yes, evenly dispersed
cells in brain and VNC,
looks glial. Not as bright
as other lines’ CNS
expression.
No
20 X No No -
41 X No No -
61 3 Very bright expression in
head and abdomen.
Yes, very high expression
in CNS, looks glial.
No.
75 2 No No -
77 2 Bright staining in head,
eyes, legs and abdomen.
Slight expression in
genitalia.
No -

205

Table C2, continued

90 3 No No -
100 3 No 3-5 regions show non-
nuclear expression in the
meso-thoracic ganglion
of VNC. No expression
in the brain.
No
138 2 Very bright, globular
staining in head, likely fat
body expression, no
expression in genitalia
Yes, evenly dispersed
cells, similar expression
pattern to lines 11 and 61.
No
141 3 Bright abdominal
expression, no expression
eye or proboscis, slight
expression in genitalia
No* -
170 2 Expression in head and
upper abdomen.
No* -

* Indicates faint expression in the mushroom body region of the brain, although no cell
bodies were visible. Grey indicates lines for which CNS tissue images are shown in
Figure C6.

206

Figure C6. DSX-GAL4 lines’ expression patterns in the male brain (A-F) and VNC (G-
L). Fly lines carrying the 5KB region upstream of the DSX open reading frame inserted
in front of a GAL4 coding sequence were crossed to flies carrying a GFP reporter
construct. The CNS tissue from these DSX-GAL4; UAS-nls-GFP progeny were
dissected and stained with an anti-GFP antibody alone (all panels except F and J), or an
anti-GFP antibody with the antibody against DSX (F is posterior brain region and J is the
TN1 region). The results indicate that none of the lines have expression in DSX-
expressing cells; the staining observed seems to be an artifact of the vector pPT-GAL4
alone, and is not a true recapitulation of the DSX protein expression pattern. Line
numbers are indicated in the bottom right of each panel, and correspond to the line
numbers in Table C2.

207

Figure C6

Abstract (if available)

Abstract In the fruit fly Drosophila melanogaster, males perform an innate and stereotyped courtship ritual to successfully mate with females. Male courtship behavior and the development of sex-specific morphologies are controlled by an alternative pre-mRNA splicing cascade called the sex determination hierarchy, which ultimately results in the production of sex-specific transcription factors encoded by doublesex (dsx) and fruitless (fru). DSX regulates most of the morphological differences between males and females. The male-specific isoform of FRU (FRUM) functions in the central nervous system (CNS) and is necessary for most aspects of male courtship behavior. Although the regulatory interactions within the sex determination hierarchy are well characterized, the activity and targets of DSX and FRUM, the terminal effectors of sex determination, remain poorly understood. To explore how the genome is sex-specifically deployed, an analysis of the transcriptional regulation of sex-specific development during metamorphosis, the period in which the fly transitions from a larva to an adult, was performed. Genes were identified that likely function at distinct times during metamorphosis to regulate somatic sex-specific development, as well as genes that function in germline development. To identify genes that are downstream of the sex determination hierarchy, transcriptional profiles of mutants for sex determination genes were analyzed at mid-metamorphosis, and downstream targets of FRUM and DSX were identified. The set of genes regulated by fru included a preponderance of genes that are known to be regulated by the steroid hormone ecdysone. This led to functional characterization of the Ecdysone Receptor (EcR) specifically in the fru neural network. The data presented here demonstrate that one particular isoform of EcR, EcRA, is required in the fru neural circuit for appropriate male courtship behavior.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Computational and experimental approaches for the identification of genes and gene networks in the Drosophila sex-determination hierarchy

PDF

Sex-specific effects of drosophila p53 on adult life span

PDF

Chorion gene amplification in Drosophila melanogaster

PDF

Molecular misreading in Drosophila melanogaster

PDF

Transgenic biomarkers of aging in drosophila melanogaster

PDF

Characterization of Drosophila longevity and fecundity regulating genes

PDF

Over-expression of superoxide dismutase in Drosophila melanogaster

PDF

The effect of microbial load and autophagy on drosophila immunity and life span

PDF

Robustness and stochasticity in Drosophila development

PDF

Biological interactions on the behavioral, genomic, and ecological scale: investigating patterns in Drosophila melanogaster of the southeast United States and Caribbean islands

PDF

Exploring three-dimensional organization of the genome by mapping chromatin contacts and population modeling

PDF

Fluorescent protein turnover in free-moving Drosophila melanogaster

PDF

Phylogeography, reproductive isolation, and the evolution of sex determination mechanisms in the copepod Tigriopus californicus

PDF

Tissue interactions & molecular pathways in specification of the ectomesenchyme from cranial neural crest

PDF

Forkhead transcription factors regulate replication origin firing through dimerization and cell cycle-dependent chromatin binding in S. cerevisiae

PDF

Unraveling the molecular mechanisms of heterochromatic double strand break repair in Drosophila cells

PDF

C. elegans topoisomerase II regulates chromatin architecture and DNA damage for germline genome activation

PDF

New tools for whole-genome analysis of DNA replication timing and fork elongation in saccharomyces cerevisiae

PDF

The role of histone H4 lysine 20 monomethylation in gene expression and differentiation

PDF

To adapt or not to adapt: the age-specific and sex-dependent differences in the adaptive stress response

Asset Metadata

Creator Sanders, Laura Elizabeth (author)

Core Title Molecular and genomic studies of sex determination in Drosophila

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Molecular Biology

Publication Date 10/23/2008

Defense Date 07/24/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Development,Drosophila melanogaster,microarray,neural circuits,OAI-PMH Harvest,sex determination

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Arbeitman, Michelle Nina (committee chair), Aparicio, Oscar Martin (committee member), Reisler, Hannah (committee member), Tower, John G. (committee member)

Creator Email laura.elizabeth.sanders@gmail.com,lesander@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1691

Unique identifier UC1312712

Identifier etd-Sanders-2342 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-123344 (legacy record id),usctheses-m1691 (legacy record id)

Legacy Identifier etd-Sanders-2342.pdf

Dmrecord 123344

Document Type Dissertation

Rights Sanders, Laura Elizabeth

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu