Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Robustness and stochasticity in Drosophila development
(USC Thesis Other)
Robustness and stochasticity in Drosophila development
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Copyright 2020 Sammi Ali
Robustness and Stochasticity in Drosophila
Development
By
Sammi Ali
A Dissertation presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MOLECULAR BIOLOGY)
DECEMBER 2020
ii
Acknowledgements
I would like to thank my family, my advisor Dr. Sergey Nuzhdin and all of our lab mates and
undergraduate researchers. I would also like to thank Dr. Fraser’s lab and the Translational
Imaging Center, as well as other committee members, Dr. Tower, Dr. Ehrenreich, and Dr. Haas.
Lastly, I would like to thank our many collaborators especially Konstantin Kozlov and George
Courcoubetis for all their help in accomplishing this work.
iii
Table of Contents
Acknowledgements…………………………………………………………..ii
List of Figures………………………………………………………………...v
List of Tables………………………………………………………………….vii
Thesis Abstract……………………………………………………………….viii
Chapter 1 Introduction ........................................................................... 1
1.1. Balancing Robustness and Adaptability ..................................................... 1
1.2. Developmental Systems Drift ....................................................................... 3
1.3. The Compound Eye is an Ideal System ....................................................... 3
1.4. Morphogenesis: From Larval Eye Disc to Adult Compound Eye ............ 4
1.5. The Adult Eye ................................................................................................ 6
1.6. Eye Development Gene Regulatory Network ............................................. 7
1.7. Hybridization Chain Reaction (HCR) ......................................................... 9
1.8. Modeling Disease Using Eye Discs: Rough Eye Phenotype .....................11
1.9. Sexual Determination Hierarchy (SDH) ...................................................13
1.10. Dissertation Chapters Summary ..............................................................15
Chapter 2: Robustness in Eye Disc Development .............................. 19
2.1. Chapter 2 Abstract: .....................................................................................19
2.2. Chapter 2 Introduction ...............................................................................20
2.3. Chapter 2 Materials & Method ..................................................................24
2.4. Chapter 2 Results.........................................................................................28
2.5. Chapter 2 Discussion ...................................................................................36
Chapter 3: Threshold Response to Developmental Robustness ....... 40
3.1. Chapter 3 Abstract ......................................................................................40
3.2. Chapter 3 Introduction ...............................................................................41
3.3. Chapter 3 Methods ......................................................................................44
3.4. Chapter 3 Results.........................................................................................50
3.5. Chapter 3 Discussion ...................................................................................56
iv
Chapter 4: The Role of the Sexual Determination Hierarchy in Eye
Development ........................................................................................... 58
4.1. Chapter 4 Abstract ......................................................................................58
4.2. Chapter 4 Introduction ...............................................................................59
4.3. Chapter 4 Method ........................................................................................65
4.4. Chapter 4 Results.........................................................................................69
4.5. Chapter 4 Discussion: ..................................................................................73
Chapter 5: The Role of the Sexual Determination Hierarchy in
Brain Development ................................................................................ 77
5.1. Chapter 5 Abstract ......................................................................................77
5.2. Chapter 5 Introduction ...............................................................................78
5.3. Chapter 5 Methods ......................................................................................90
5.4. Chapter 5 Results.........................................................................................97
5.5. Discussion ...................................................................................................102
Concluding Remarks ........................................................................... 105
References ............................................................................................ 107
v
LIST OF FIGURES
FIGURE 1…………………………………………………………………………………..…..4
FIGURE 2…………………………………………………………………………………..…..5
FIGURE 3………………………………………………………………………………..……..6
FIGURE 4………………………………………………………………………………..……..7
FIGURE 5………………………………………………………………………………..……..9
FIGURE 6………………………………………………………………………………..……..10
FIGURE 7…………………………………………………………………………………..…..11
FIGURE 8………………………………………………………………………………..……..12
FIGURE 9……………………………………………………………………………..………..14
FIGURE 10…………………………………………………………………………..…………29
FIGURE 11……………………………………………………………………………………..32
FIGURE 12……………………………………………………………………………………..33
FIGURE 13……………………………………………………………………………………..34
FIGURE 14……………………………………………………………………………………..42
FIGURE 15……………………………………………………………………………………..44
FIGURE 16……………………………………………………………………………………..46
FIGURE 17……………………………………………………………………………………..50
FIGURE 18……………………………………………………………………………………..51
FIGURE 19……………………………………………………………………………………..52
FIGURE 20……………………………………………………………………………………..53
FIGURE 21……………………………………………………………………………………..53
FIGURE 22……………………………………………………………………………………..54
FIGURE 23……………………………………………………………………………………..55
FIGURE 24……………………………………………………………………………………..66
FIGURE 25……………………………………………………………………………………..67
FIGURE 26……………………………………………………………………………………..69
FIGURE 27……………………………………………………………………………………..71
vi
FIGURE 28……………………………………………………………………………………..72
FIGURE 29……………………………………………………………………………………..73
FIGURE 30……………………………………………………………………………………..81
FIGURE 31……………………………………………………………………………………..83
FIGURE 32……………………………………………………………………………………..87
FIGURE 33……………………………………………………………………………………..93
FIGURE 34……………………………………………………………………………………..95
FIGURE 35……………………………………………………………………………………..97
FIGURE 36…………………………………………………………………………………….98
FIGURE 37…………………………………………………………………………………….100
vii
LIST OF TABLES
TABLE 1……………………………………………………………………………………..35
TABLE 2……………………………………………………………………………………..35
TABLE 3……………………………………………………………………………………..102
viii
Thesis Abstract
Biological development requires the careful coordination of several processes, such as
transcription, replication, differentiation and patterning. These intrinsically stochastic processes
require developmental programs to be robust. This robustness is maintained through an intricate
web of interacting Gene Regulatory Networks (GRNs) and clean-up mechanisms. As species
diverge, stochasticity in GRNs can lead to developmental systems drift, whereby genetic variation
accumulates between populations without any observable phenotypic change. How can biological
systems use variation in gene expression, or transcriptional stochasticity, to balance developmental
robustness versus evolutionary adaptability depending on developmental and environmental
contexts? The Drosophila compound eye is an excellent system to study these interlinked topics
since individual eye units, or ommatidia, specified in the larval eye disc eventually form the highly-
organized, offset patterning seen in adult eyes.
We measured the quantitative and spatial relationship of 4 key genes involved in eye
development simultaneously, using the Hybridization Chain Reaction (HCR), in 3 natural
populations of D. melanogaster and D. simulans males and females respectively. Our novel
published findings show significant transcriptional stochasticity between species, genotypes and
sexes, as well as regulatory logic divergence despite the phenotypic similarity of Drosophila
melanogaster and D. simulans eyes (Ali 2019, Chapter 2).
Afterwards, we created a mathematical model to understand the observed phenotypic
robustness in eye GRNs. Our published results quantitatively assessed this threshold response,
whereby transcriptional stochasticity is buffered up to a certain threshold, after which the system
rapidly degrades (Courcoubetis 2019, Chapter 3).
ix
We further investigated the eye GRN in different sexual contexts by genetically manipulating
the Sex Determination Hierarchy (SDH) and performing HCR on several genes in male, female
and pseudomale (masculinized females) larval eye discs, which showed no differences. (Chapter
4). To understand how GRNs evolve between closely related species, we used HCR to measure
the quantitative and spatial relationships of fruitless (fru) Neuron Expression Clusters (NECs) in
the SDH between D. melanogaster and D. simulans brains, which showed transcriptional and
volumetric variation in several reproductively-relevant brain regions (Chapter 5).
Collectively, our findings improve our understanding of the genotype-to-phenotype map,
which is instrumental in elucidating the genetic basis of complex diseases.
1
1. Chapter 1 Introduction
1.1. Balancing Robustness and Adaptability
Biological development requires the careful coordination of complex processes, such as
cellular growth, replication, differentiation and tissue patterning to create a functional final form.
The interconnectedness of these processes necessitates developmental systems to be robust,
whereby perturbations and signal noise in a particular part of the system can be correctly counter-
balanced to maintain the predetermined, conserved goal. This robustness is maintained through an
intricate web of interacting Gene Regulatory Networks (GRNs).
Important developmental processes are highly robust, often relying on functionally-redundant
back-up components. While fully-redundant inputs can completely compensate for perturbations,
such as loss-of-function mutations, many enhancers tend to be partially-redundant (Frankel 2010
& Osterwalder 2018). These redundant inputs form an evolutionary basis for robustness versus
adaptability during development (Perry 2010 & Gonzalez-Blas 2020).
Variation in gene expression has been extensively documented (Albert 2015 and Marinov
2014). Numerous mechanisms can drive this variation in gene expression, including mutations,
recombination, environmental perturbations (Arsenault 2018), maternal effects (Signor 2018), and
the general burstiness of transcription between cells within a tissue (Bothma 2014 and Felix 2015).
Despite this transcriptional variation, developmental and phenotypic outcomes remain robust
(Garfield 2013, McNeil 2011, Raj 2008, and Romero 2012). On the other hand, organisms must
modify their genetic adaptability in response to selection pressures by tapping into these same
sources of expression variation, to overcome robustness and evolve (Chan 2010, Gompel 2005,
Hoekstra 2007, Jeong 2008, Pia 2014, Signor 2016, Wray 2007, and Yassin 2016). While cryptic
variation focuses on mutation-by-environment interactions (Gibson 2008 and Paaby 2014), we
2
will focus on expression variation due to mutation and recombination, termed genetic perturbations
(Paaby 2016).
The mechanisms that dictate this delicate balance between developmental stability and
evolutionary adaptability have rarely been addressed experimentally (Casci 2005, Gibson 2009,
Green 2017, Heranz 2010, Hermisson 2004 Nijhout 2017, Rutherford 2007, Stern 2000, and True
2001). This is partly due to previous limitations in molecular and microscopy technologies that
restricted studies to be quantitative or spatial, but not both. Most quantitative methods rely on
combining tissues, while spatial information is often restricted to qualitative comparisons.
Furthermore, many previous studies focus on large mutational effects in isogenic backgrounds
(Paaby 2016 and Lott 2007) or single cell methods where gene expression variation due to genetic
differences is confounded with differences in transcription. These issues often limit the ability to
replicate these studies and apply rigorous statistical models. Thus, effective comparisons of
developmental processes between conditions, such as health vs. disease, are restricted. Resolving
these issues is extremely important since there might be complex interactions between
conditionally-neutral differences in gene expression that result in disease phenotypes.
To overcome these challenges, we generated spatially-resolved, quantitative data within and
between natural populations, where expression variation due to genetic differences can be
compensated by other genes in the pathway (Estes 2011, Landry 2005, & Thompson 2015),
mitigated by robust GRN structure, such as feedback loops (MacNeil 2011), or below threshold
levels (Felix 2015).
3
1.2. Developmental Systems Drift
Populations can diverge by accumulating small expression differences in their GRNs due to
various evolutionary forces such as genetic drift and mutation. These cumulative differences can
lead to developmental systems drift. whereby hybrids between two closely-related species are less
fit than their parents due to the significantly different down- or up-regulation of genes within the
hybrid GRN. As these two species drift further apart, there is stronger selection against their
hybrids. One way to counter-balance this issue is to utilize robust GRNs that can handle
perturbations in gene expression to still achieve the developmental end goal. Thus, while the
phenotypic outcome is conserved, the underlying differences in these developmental GRNs
between populations can diverge. This hidden variation can manifest in different phenotypic
outcomes depending on genetic background effects, incomplete penetrance and other non-linear
relationships (Ali 2019).
Elucidating the evolutionary forces that maintain this delicate balance is instrumental in
understanding how heritable GRNs interpret different developmental contexts to coordinate the
complex patterning required to correctly reconstruct an organism.
1.3. The Compound Eye is an Ideal System
The Drosophila compound eye is an excellent system to investigate the complicated,
interlinked relationship of developmental processes, such as GRN expression or tissue patterning,
and biological outcomes, such robustness or developmental systems drift. Although the resulting
highly-ordered, offset eye units, or ommatidia, that make the compound eye are only fully-formed
in the adult, the role of individual cells within this repeating array are pre-determined during larval
development (Kumar 2012). The GRNs within the eye-antenna imaginal disc, hereafter referred
4
to as the eye disc, use morphogens and a physical tissue indentation, the morphogenetic furrow
(MF), to space out individual ommatidium within the array and specify the role of each cell
involved in the fully-formed eye. The eye disc to eye morphogenesis occurs primarily in 2D along
the anterior-posterior and dorsal-ventral axes since it is derived from a single epithelial layer
(Haynie 1986 and Barkai 2009). Furthermore, this transformation relies primarily on cell-cell
interactions, such as lateral inhibition and morphogen diffusion, with very limited cell-lineage
effects (Lawrence 1979).
1.4. Morphogenesis: From Larval Eye Disc to Adult Compound Eye
Cell specification in larval eye-antenna imaginal discs (eye discs), manifest into a crystal-like
adult compound eye during metamorphosis (Figure 1). The adult eye is composed of repeating
Figure 1 Drosophila Eye Development. Developing regions (antenna, head & eye) of the 3
rd
instar larval Eye-antenna disc using GFP-tagged DE-Cadherin (left, scale bars = 40µm)
compared to a developed adult eye (right). The morphogenetic furrow (MF) is circled in
yellow. Anterior is to the left. Posterior to the MF, individual eye units, or ommatidia (bright
green dots), are highly ordered with equal, offset spacing. Anterior to the MF, cells are
proliferating and undifferentiated. Within the MF, cell cycles are synchronized.
5
offset eye units, or ommatidia. This highly-ordered structure is generated by GRNs that repeat a
series of steps starting with the placement of the R8 central photoreceptor. R8 cells are specified
by diffusion of morphogens within the morphogenetic furrow (MF), an indentation in the eye disc
tissue, followed by other signals that organize the role of neighboring cells and ultimately advance
the MF. The R8 placement is crucial in determining the final spacing between ommatidia (Figure
2). Below are extensive details on the structure and development of the Drosophila compound eye.
Eye development starts with a physical indentation in the eye disc called the morphogenetic
furrow (MF). As the MF sweeps across the tissue anteriorly, the roles of individual cells within
Figure 2: Ommatidial Assembly at the Morphogenetic Furrow (MF).
(A-E) Confocal images showing the progression of the MF in wild type, 3
rd
instar larva using
F-actin (red) and ELAV (green). (F) Schematic showing the developmental specification of
each ommatidia, beginning with a refined, single R8 cell immediately posterior to the MF
and ending with an 8-cell mature cluster. The mature cluster will rearrange into the
asymmetric trapezoidal orientation as the ommatidia rotates. Anterior is to the right in all
images (from Kumar 2012).
6
the eye unit are specified. Thus, if we look at a single time snapshot of the eye disc during
development, cells more posterior to the MF are more solidified in their ommatidial role, while
cells immediately posterior to the MF are not. Conversely, cells more anterior to the MF are
undifferentiated and proliferating asynchronously, while cells near or within the MF are arrested
in G1 so they can divide in a synchronized fashion, known as the second mitotic wave, once they
receive instructions from their posterior neighbors (Wolff 1991).
1.5. The Adult Eye
As previously mentioned, the placement of each eye unit is decided much earlier, during larval
development of the eye-antenna imaginal disc, hereafter to referred to as eye disc. The placement
and specification of each unit within the larval eye disc proceeds in a posterior to anterior fashion.
The adult compound eye is composed of ~800 individual eye units, or ommatidia, that are almost
perfectly offset from each other (Figure 3). A typical adult eye contains 32-34 rows of ommatidia.
Each ommatidium is made up of 14 cells: 8 photoreceptors, named R1-R8, 4 cone cells, and 2
pigment cells (Figure 4). Ommatidia share 6 secondary pigment cells, 3 tertiary pigment cells and
Figure 3: Organization of Ommatidia Within the Adult Fly Eye.
Electron microscopy of (a) the whole Drosophila adult fly eye, (b) the organization of
individual eye units, or ommatidia, along the dorsal-ventral axis (red line), and (c) the 8
photoreceptors within each ommatidium (R1-R8), note R8 is directly above R7 & not shown.
Anterior is to the right in all images (from Kumar 2012).
7
3 bristle cells (Waddington 1960). The roles of these various cells is determined in larva, after the
refinement of atonal (ato) expression.
1.6. Eye Development Gene Regulatory Network
The MF advances anteriorly through the interaction of several genes with the eye disc GRN,
primarily Hedgehog (Hh) and Decapentapegic (Dpp). These are long-range activators of atonal
(ato). Cells that express ato inhibit their neighbors from also expressing ato (Greenwood 1999 and
Baonza 2001). Thus, as we move posteriorly from the MF, the concentration of ato-expressing
cells is slowly reduced from a line of tightly-packed ato-expressing cells, to clusters of 3 ato-
expressing neighbors, and finally to single ato-expressing cells that are evenly spaced (Voas 2004).
These evenly-spaced cells will become the R8 photoreceptor within each ommatidium. This is the
initial brick, upon which the rest of the eye unit is built. Notch/Delta (DI) lateral inhibition is used
Figure 4: Organization of Photoreceptors in Both Eyes.
Schematic showing (a) the 8 photoreceptors within ommatidium (R1-R8) & (b) dorsal-ventral
organization within each eye and the chirality between eyes (from Kumar 2012).
8
for ato cluster refinement and to specify the role of the cells surrounding the individual ato-
expressing cells while Scabrous (sca) is a short-range diffusible activator of ato (Gavish 2016).
Anterior to the MF, undifferentiated cells are expressing hairy (h), which inhibits ato and allows
cells to proliferating asynchronously (Fried 2016).
The speed at which the MF sweeps through the eye disc depends on several factors, including
the interplay between Hedgehog (Hh), which acts posteriorly as an activator, and hairy (h), which
acts anteriorly as an inhibitor. Within the MF, a band of atonal (ato) is expressed that becomes
refined posteriorly. Cells expressing high amounts of ato ultimately activate Delta/Notch (DI/N)
signaling to inhibit their neighbors. This results in evenly-spaced clusters of ato-expressing cells
in each row that are perfectly offset from the previous and following row of ato-expressing cells.
Thus, the spacing between these ato-expressing cells provides the first brick upon which the final
spacing between adult ommatidia is built (Kumar 2012).
As the MF is propagated anteriorly by several key developmental genes, or morphogens, in the
eye disc, the precise positioning of ommatidia behind the MF are determined to ultimately form
the final, fully-functional adult eye. The spacing between neighboring R8 cells is instrumental in
the final placement of the ommatidia (Gavish 2016). Defects that misplace an ommatidium can
affect neighbors which propagates further misplacement resulting in a distorted array. The
regulatory logic that guides this repeated patterning recycles the same genes within the eye
development GRN. Thus, perturbations via mutations or changes in transcription of a single gene
can affect all eye units (Kumar 2012). Many of the genes involved in the regulatory logic of the
fly eye are conserved in mammals. We will focus on hedgehog (Hh), Delta (DI), atonal (ato), and
hairy (h), which are heavily involved in the regulatory logic of fly eye development (Figure 5).
9
1.7. Hybridization Chain Reaction (HCR)
Many studies focus on 1 or 2 genes, reporting either spatial or expression data, but not both.
In order to study the impact of developmental systems drift and robustness in natural populations,
we used the Hybridization Chain Reaction (HCR) extensively to observe mRNA expression and
spatial differences within and between Drosophila species, genotypes and sexes in eye discs and
brains. HCR is an innovative technology that allows the simultaneous pattern-visualization and
quantification of mRNA transcripts within a fixed tissue.
Figure 5: Regulatory Logic of Eye Disc Developmental Genes.
Schematic showing the regulatory logic of hedgehog (Hh), Delta (DI), atonal (ato), and hairy
(h), around the morphogenetic furrow (M.F.). Anterior is to the left (from Ali 2019).
10
HCR uses small DNA probes to bind specific mRNA molecules. Afterwards, probe-
specific fluorescent hairpins can attach to amplify the signal (Figure 6). We used confocal
microscopy to generate z-stacks. The ZEN software, along with our experimental design allows
for the linear un-mixing of spectrally overlapping signals to generate higher resolution images and
more reliable colocalization data. HCR technology allows us to image several genes
simultaneously at cellular resolution in a single sample. This minimizes heterogeneity between
genotypically identical samples and avoids potential inconsistencies arising from sequential
staining. Furthermore, this high-throughput multiplexing microscopy method can assess
colocalization patterns at cellular resolution between species, thereby accounting for the possibility
Figure 6: Schematic for the Hybridization Chain Reaction (HCR).
Schematic showing (a) asterisk indicates fluorescent molecules. Hairpin 1 (H1) has an “a”
overhang that can be attacked by the complimentary “a*” on the DNA initiator sequence.
This opens up the H1 hairpin which exposes the “c*-b*” portion of H1. This allows the “c-b”
on H2 to bind to “c*-b*” and opens up the H2 hairpin which generates a new “a*-b*”
sequence that can open another H1 leading to a chain reaction. (b) At the detection stage,
small DNA probes are hybridized to the mRNA target. Unbound probes are washed off. At
the amplification stage, probe-specific fluorescent hairpin amplifiers result in a hybridized
chain reaction as described in A. (c) The experimental timeline: probes are given 14 hours to
bind to mRNA at 45
o
C, excess probes are washed then the amplifiers are given 14 hours to
bind to the probes at 25
o
C. (From Choi 2014)
11
that genes and morphological levels can be conserved while still showing divergence in cell
number, patterning, and differentiation between species (Choi 2014 and Tanaka 2008).
HCR is unique in that it produces gene expression patterns that are both quantitative and
spatial. While other approaches such as FISH can be adapted to detect individual transcripts, HCR
has a linear signal that is 20× brighter than FISH, it reduces non-specific background staining, and
it can detect 88% of single RNA molecules in a cell with an appropriately low false discovery rate
(Ma & Moses, 1995; Pan & Rubin, 1995). It is also highly repeatable, with different sets of probes
targeted to the same gene showing correlations of .93–.99 (Fraser, pers. comm.). All DNA probes
were designed and synthesized by Molecular Instruments (Choi et al., 2014). Each probe contains
two-initiator sequences (I1 and I2) that bind to a specific amplifier. Genes are multiplexed in each
preparation as orthogonally-designed hairpins allowed the simultaneous amplification of their
target sequences (Figure 6). Each target mRNA is detected using five DNA probes to annotate the
position and expression levels for each gene (Figure 7).
1.8. Modeling Disease Using Eye Discs: Rough Eye Phenotype
Problems in eye development can lead to a “rough eye” phenotype, which has served as a
model for various diseases. This interdisciplinary research has important public health
Figure 7: Multiplexed HCR Results. The mRNA expression patterns for hairy (Grey), atonal (Red), Delta
(Green), hedgehog (Blue) and Merge in 3
rd
instar larval eye discs using Hybridization Chain Reaction
(HCR), (From Ali 2019).
12
implications. Understanding these complex topics will help elucidate the genetic basis of disease,
especially the role of incomplete penetrance and genetic background effects in developmental
defects (Wolff 1991). Understanding developmental robustness is crucial for developing models
that can predict molecular and mechanistic properties from biochemical circuits (Averbukh 2017).
Numerous studies have shown that GRNs tolerate perturbations up to a certain point, after
which, the system rapidly breaks down. Although many of these developmental processes have
been investigated extensively, most studies focus on a single process, often relying on large-effect
mutations. Furthermore, many studies focus on 1 or 2 genes, reporting either spatial or expression
data, but not both.
We modified these newer models to quantitatively measure the robustness of ommatidial
ordering by simulating R8 placement and altering transcriptional noise, or stochasticity, in
Figure 8: Rough Eye Phenotype. Light microscopy showing (a) wild-type eyes with evenly-spaced, offset
ommatidia, and (b) a rough eye mutant. (From Wittman 2001).
13
different eye disc GRNs. During eye disc development, cells produce diffusible morphogens that
affect neighboring cells’ outputs in a non-linear fashion. This means the concentration of a
particular morphogen within a cell does not necessarily produce a uniform production or diffusion
rate. Thus, reliably forming the near-perfect compound eye pattern requires robustness in the
overall GRNs, despite the transcriptional noise, of individual genes. We focus on the placement of
R8 cells, which is a crucial step in the determination of the final position of ommatidia within the
compound eye to affect the organism’s visual acuity.
We found that the eye disc GRNs can tolerate parametric stochasticity in expression levels,
leading to ommatidial displacement up to a certain threshold, after which the system reaches a
point-of-no-return and rapidly degrades. Our model predicts that this sigmoidal threshold response
allows eye development to be robust.
Our mathematical modeling of eye disc robustness (see chapter 2) supported our predictions
for experimentally-verified HCR results. The high-throughput molecular and microscopy
innovations we developed allowed us to directly observe changes in gene expression,
colocalization and spatial patterning at cellular resolution between species, sexes and genotypes in
Drosophila eye discs (chapter 3 and 4) and brains (chapter 5).
1.9. Sexual Determination Hierarchy (SDH)
In Drosophila, sex-specific expression is determined on a cellular basis using a single, multi-
branch Sex Determination Hierarchy (SDH) Gene Regulatory Network (GRN) (Christiansen et al.
2002). The SDH controls X chromosome dosage compensation, somatic sexual differences, and
male-specific courtship behaviors (Yamamoto 2013).
14
The ratio of X chromosomes to autosomes (X:A ratio) dictates the sex-specific regulation of
cells. Females have an X:A of 1. This leads to the activation of Sex-lethal (Sxl) which uses an auto-
regulatory positive feedback loop to increase Sxl expression. Sxl is a master-switch which activates
transformer (tra) leading to sex-specific splicing of doublesex (dsx
F
) which is important for sexual
development, including genitalia. Males have an X:A of 0.5. Thus, Sxl is not activated in males,
which leads to the activation of male specific-lethal-2 (msl-2) for dosage compensation, and male-
specific versions of dsx
M
and fruitless (fru
M
), which is involved in male-specific courtship
behaviors (Penn 2007).
The Sex-Determination Hierarchy (SDH) in Drosophila is a pivotal pathway for somatic
sexual dimorphism (Figure 5). If the concentration of X-linked Signal Elements (XSEs) reaches a
critical threshold at the start of zygotic transcription, then Sex lethal (Sxl) will be irreversibly turned
on. In XX females, Sxl activation is maintained through a self-regulatory loop and results in a
Figure 9: Sexual Determination Hierarchy (SDH) in Drosophila.
An illustration of the Gene Regulatory Networks (GRNs) that determine sexual
developmental outcomes in males and females. In Females, the master regulator Sex-lethal
(Sxl) is turned on, leading the female-specific expression of doublesex (DSX
F
) and no
fruitless (FRU). In males, SXL is not expressed, leading to dosage compensation, and male-
specific expression of DSX
M
and FRU
M
via alternative splicing (from Robinett 2010).
15
series of splicing events via transformer 1 (Tra 1) and transformer 2 (Tra 2) to lead to the
expression of female-specific Doublesex (dsx
F
). Alternative splicing of fruitless (fru) leads to an
inactive isoform in females (Cachero 2010). In XY males, XSEs never reach the critical threshold
in time, thus Sxl remains inactive. Sxl inactivity leads to the default male development via the
expression of dsx
M
, and the sex-specific expression of fru
M
via the P1 promoter, as well as male-
specific lethal-2 (msl-2) which is important in dosage compensation (Salz 2010). Numerous studies
have concluded that fru
M
plays a critical role in the development of masculinized neurons and
ultimately male courting behavior (Rideout 2010, Christiansen 2002 and Yamamoto 2013). There
are many cells in both sexes that express the common region of fru (fru-COM), however fru
M
is
expressed only in about 2000 male neurons via alternative splicing. These fru
M
expressing neurons
form clusters only in the masculinized brain (Lee 2000). Both the default transcript, dsxM, and the
alternatively spliced transcript, dsx
F
, have multiple instructive roles in the sex-specific
development of the central nervous system, imaginal discs and body size (Arbeitman 2007 and
Lee 2009).
1.10. Dissertation Chapters Summary
Biological development requires the careful coordination of several processes, such as
transcription, replication, differentiation and patterning. These intrinsically stochastic processes
require developmental programs to be robust. This robustness is maintained through an intricate
web of interacting Gene Regulatory Networks (GRNs) and clean-up mechanisms. As species
diverge, stochasticity in GRNs can lead to developmental systems drift, whereby genetic variation
accumulates between populations without any observable phenotypic change. Thus, biological
systems use variation in gene expression, or transcriptional stochasticity, to counter-balance
16
developmental robustness versus evolutionary adaptability depending on developmental and
environmental contexts. Understanding how these topics are linked is crucial for understanding
complex diseases.
In Chapter 1, we review the interdisciplinary literature involved in these topics. The Drosophila
complex eye is an excellent system to understand these interdisciplinary topics, whereby cell
specification in larval eye-antenna imaginal discs (eye discs), manifest into a crystal-like adult
compound eye after metamorphosis. The adult eye is composed of repeating offset eye units, or
ommatidia. This highly-ordered structure is generated by GRNs that repeat a series of steps starting
with the placement of the central photoreceptor, R8, by diffusion within the morphogenetic furrow,
an indentation in the eye disc tissue, followed by other signals to organize the role of neighboring
cells and advance the MF. The R8 placement is crucial in determining the final spacing between
ommatidia. Problems in this process can lead to a “rough eye” phenotype, which has served as a
model for various diseases.
In chapter 2, we measured the quantitative and spatial relationship of 4 key genes involved in
eye development simultaneously, using the Hybridization Chain Reaction (HCR), in 3 natural
populations of D. melanogaster and D. simulans males and females respectively. HCR allows
visualization and quantification of several mRNA transcripts simultaneously. We used this multi-
plexing method to obtain expression data of 4 developmental genes (Hh, DI, ato, and h) in
Drosophila eye discs of 5 day old 3rd instar larvae. Previous studies have observed quantitative or
spatial relationships, but not both for the 4 genes that we studied. By comparing eye discs within
and between sexes, genotypes and species (D. melanogaster vs D. simulans) at cellular resolution
using state-of-the-art confocal microscopy, we found extensive underlying variation in the spatial
patterning and expression levels of these genes. We also found evolutionary differences in the
17
spatial interrelationship of some of these genes. Our published findings show significant
transcriptional stochasticity between species, genotypes and sexes, as well as regulatory logic
differences despite the phenotypic similarity of Drosophila melanogaster and D. simulans eyes
(Ali 2019)
In chapter 3, we created a mathematical model to understand the robustness we observed in
eye development by simulating morphogen diffusion, R8 placement & MF progression in this
well-studied spatiotemporal process. We quantified order within different spatial patterning
scenarios and observed a universal sigmoidal response to increasing transcriptional noise. The
simulated eye disc remains ordered, despite increasing noise, up to a certain threshold. Consistent
with previous qualitative studies, once this point-of-no-return is breached, the spacing of eye units,
or ommatidial order, rapidly degrades. Our published mathematical model quantitatively assessed
this threshold response and predicted that we would observe stochasticity in the eye development
GRNs of different Drosophila species without observing any phenotypic changes in the adults
(Courcoubetis 2019).
In chapter 4, we further investigated how the eye GRN interprets different sexual contexts by
genetically manipulating the Sex Determination Hierarchy (SDH). Although females are about
30% larger in body size in Drosophila species, female eyes are only 2% larger than males (Gasper
2020). This suggests that the SDH somehow influences the size of certain organs. We used HCR
on several genes in SDH and eye GRNs in male, female and pseudomale (masculinized females)
larval eye discs. We found no significant differences in DI and hh expression between eye discs.
In chapter 5, we further investigated the balance between developmental robustness and
evolutionary adaptability. The species differences in development from our Chapter 3 results drove
us to ask the question, how quickly can complex traits evolve between species despite stochasticity
18
and developmental robustness? We postulated that sexually dimorphic brains regions evolve faster
than monomorphic regions. We used HCR to measure the quantitative and spatial relationships of
Neuron Expression Clusters (NECs) in the SDH between D. melanogaster and D. simulans brains.
We found significant transcriptional and volumetric variation in several reproductively-relevant
brain regions. Collectively, Our findings improve our understanding of the genotype-to-phenotype
map, which is instrumental in elucidating the genetic basis of complex diseases.
19
2. Chapter 2: Robustness in Eye Disc Development
2.1. Chapter 2 Abstract:
Robustness in development allows for the accumulation of genetically based variation in
expression. However, this variation is usually examined in response to large perturbations, and
examination of this variation has been limited to being spatial, or quantitative, but because of
technical restrictions not both. Here we bridge these gaps by investigating replicated quantitative
spatial gene expression using rigorous statistical models, in different genotypes, sexes, and species
(Drosophila melanogaster and D. simulans). Using this type of quantitative approach with
molecular developmental data allows for comparison among conditions, such as different genetic
backgrounds. We apply this approach to the morphogenetic furrow, a wave of differentiation that
patterns the developing eye disc. Within the morphogenetic furrow, we focus on four genes, hairy,
atonal, hedgehog, and Delta. Hybridization chain reaction quantitatively measures spatial gene
expression, co-staining for all four genes simultaneously. We find considerable variation in the
spatial expression pattern of these genes in the eye between species, genotypes, and sexes. We also
find that there has been evolution of the regulatory relationship between these genes, and that their
spatial interrelationships have evolved between species. This variation has no phenotypic effect,
and could be buffered by network thresholds or compensation from other genes. Both of these
mechanisms could potentially be contributing to long term developmental systems drift.
20
Additional details about this chapter’s work and supplemental figures can be found in
our publication:
Ali S, Signor SA, Kozlov K, Nuzhdin SV. Novel approach to quantitative spatial gene
expression uncovers genetic stochasticity in the developing Drosophila eye. Evolution &
Development. 2019;1–15. https://doi.org/10.1111/ede.12283
2.2. Chapter 2 Introduction
2.2.1. Robustness in Biological Systems
The Drosophila complex eye is an excellent system to understand developmental robustness,
whereby cell specification in larval eye-antenna imaginal discs (eye discs), manifest into a crystal-
like adult compound eye after metamorphosis. The adult eye is composed of repeating offset eye
units, or ommatidia. This highly-ordered structure is generated by GRNs that repeat a series of
steps starting with the placement of the central photoreceptor, R8, by diffusion within the
morphogenetic furrow (MF), an indentation in the eye disc tissue, followed by other signals to
organize the role of neighboring cells and advance the MF. The R8 placement is crucial in
determining the final spacing between ommatidia. Problems in this process can lead to a “rough
eye” phenotype, which has served as a model for various diseases.
2.2.2. Robustness versus Adaptability
Important developmental processes are highly robust, often relying on functionally-redundant
back-up components. While fully-redundant inputs can completely compensate for perturbations,
such as loss-of-function mutations, many enhancers tend to be partially-redundant (Frankel 2010
21
& Osterwalder 2018). These redundant inputs form an evolutionary basis for robustness versus
adaptability during development (Perry 2010 & Gonzalez-Blas 2020).
Variation in gene expression has been extensively documented (Albert 2015 and Marinov
2014). Numerous mechanisms can drive this variation in gene expression, including mutations,
recombination, environmental perturbations (Arsenault 2018), maternal effects (Signor 2018), and
the general burstiness of transcription between cells within a tissue (Bothma 2014 and Felix 2015).
Despite this transcriptional variation, developmental and phenotypic outcomes remain robust
(Garfield 2013, McNeil 2011, Raj 2008, and Romero 2012). On the other hand, organisms must
modify their genetic adaptability in response to selection pressures by tapping into these same
sources of expression variation, to overcome robustness and evolve (Chan 2010, Gompel 2005,
Hoekstra 2007, Jeong 2008, Pia 2014, Signor 2016, Wray 2007, and Yassin 2016). While cryptic
variation focuses on mutation-by-environment interactions (Gibson 2008, Paaby 2014), we will
focus on expression variation due to mutation and recombination, termed genetic perturbations
(Paaby 2016).
2.2.3. Eye Development & The Morphogenetic Furrow
The Drosophila eye is formed from an imaginal disc, which is initially patterned by a wave of
differentiation marked by a visible indentation of the tissue, termed the morphogenetic furrow
(MF). The MF passes from the posterior to the anterior of the disc over a period of two days (90
min per adjacent row of cells), giving each disc an element of both time and space in development
(Roignant 2009). The MF is initiated by hedgehog, which represses hairy within the MF and
activates it indirectly in front of the MF through the long-range diffusion of dpp within the tissue
(Figure 1) (Felsenfeld 1995, Fu 2003, & Strutt 1997). hairy represses atonal, preventing
22
precocious neural development anterior to the MF (though this role has been recently contested)
(Bhattacharya & Baker, 2012; Brown, Sattler, Markey, & Carroll, 1991; Brown, Sattler, Paddock,
& Carroll, 1995). hedgehog activates the expression of atonal, driving the MF anteriorly (Figure
1) (Heberlein 1993; Ma 1993, & Greenwood 1999). atonal is the proneural gene in Drosophila,
establishing the competency to become photoreceptor cells (Jarman 1994). The relationship
between Delta/Notch and the other members of the pathway is more complex, but it is clear that
in cells posterior to the furrow Delta/Notch repress atonal (Figure 1) (Baker 1996, Dokucu 1996,
Firth 2005, & Gavish 2016). There is also some evidence that Delta/Notch repress negative
regulators of atonal at the furrow, such as hairy (Baonza 2001, Bhattacharya 2009, & Brown
1995).
In addition, there is some evidence suggesting that Delta/Notch are involved in the early stages of
atonal induction, and alternatively that atonal activates its own transcription (Baker 1997, Baker
1996, Jarman 1994, Li 2001, Spencer 1998, & Sun 1998). Delta/Notch are expressed in the MF
under the redundant control of hedgehog and dpp (Baonza 2001, Baonza 2005, & Parks 1995).
There are many other genes involved in the specification of the eye disc that will not be
mentioned here, in favor of focusing on the genes we have assayed. We analyze the spatial
quantitative expression of hedgehog, hairy, atonal, and Delta to understand the evolving
regulatory logic of the gene network and changes in spatial dynamics between sexes and species.
2.2.4. Significance and Innovation
Given that genetically-encoded variation in expression can accumulate due to developmental
systems drift, how does a developing organism buffer against stochastic components to achieve a
23
predetermined outcome? What are the evolutionary consequences of this robustness? What are the
consequences of sexual conflict on this robustness?
The mechanisms that dictate this delicate balance between developmental stability and
adaptability have rarely been addressed experimentally (Casci 2005, Gibson 2009, Green 2017,
Heranz 2010, Hermisson 2004 Nijhout 2017, Rutherford 2007, Stern 2000, and True 2001). This
is partly due to previous limitations in molecular and microscopy technologies that restricted
studies to be quantitative or spatial, but not both. Most quantitative methods rely on combining
tissues, while spatial information is often restricted to qualitative comparisons. Furthermore, many
previous studies focus on large mutational effects in isogenic backgrounds (Paaby 2016 and Lott
2007) or single cell methods where gene expression variation due to genetic differences is
confounded with differences in transcription. These issues often limit the ability to replicate these
studies and apply rigorous statistical models. Thus, effective comparisons of developmental
processes between conditions, such as health vs. disease, are restricted. Resolving these issues is
extremely important since there might be complex interactions between conditionally-neutral
differences in gene expression that result in disease phenotypes. To overcome these challenges,
we generated spatially-resolved, quantitative data within and between natural populations, where
expression variation due to genetic differences can be compensated by other genes in the pathway
(Estes 2011, Landry 2005, & Thompson 2015), mitigated by robust GRN structure, such as
feedback loops (MacNeil 2011), or below threshold levels (Felix 2015).
We used the hybridization chain reaction (HCR) to simultaneously measure variation in the
spatial patterning and mRNA expression levels of 4 eye development genes (hairy (h), atonal
(ato), Delta (DI), and Hedgehog (Hh)) within and between males and females of two species
(Drosophila melanogaster and D. simulans). These well-known genes drive ommatidia
24
specification (Atkins 2013, Li 1995, Raj 2008, Shah 2016, & Tsachaki 2011). We focus on the
morphogenetic furrow (MF) since cells within the MF are arrested in G1 and therefore less likely
to vary due to differences in cell division (Baker 2001, Escudero 2007, Firth 2005, & Firth 2010).
We average across cells within an ommatidial row to reduce the impact of differences in
transcriptional bursting (Bothma 2014, Fukaya 2016, & Tantale 2016).
We define gene expression differences between species, genotypes and sexes as “genetic
stochasticity” since there are no phenotypic differences between these Drosophila eyes, except for
small differences in eye size (Hilbrant 2014). We used genetic stochasticity in these 4 genes to
investigate their regulatory relationships. We performed our spatial and quantitative analysis by
creating a 2D spatial gene expression profile to compare species, genotypes and sexes. We
examined whether the regulatory relationship between these genes evolved between species or
contains variation within populations. We also analyzed whether or not the spatial differences of
these genes relative to the MF has evolved or varies within populations. We uncovered gene
expression differences across tissues that would otherwise be unnoticed by conventional
qualitative methods.
2.3. Chapter 2 Materials & Method
2.3.1. Fly Stocks
D. simulans were collected from the Zuma organic orchard in Zuma beach, CA in the spring
of 2012 (Nuzhdin 2017). They were inbred by 15 generations of full sibling crosses. D.
melanogaster were collected in Raleigh, North Carolina and inbred for 20 generations (Mackay
2012). Thus, any differences detected between species, sexes, or genotypes is due to natural
variation
25
2.3.2. Stages and Dissection of Larvae
All flies were reared on a standard medium at 25°C with a 12hr light/12-hr dark cycle. Five
replicate eye discs were isolated from each of three strains, two sexes, and two species, for a total
of 60 discs. Vials were used a single time for collection to avoid pseudo-replication, such that
every genotype and sex was collected from a separate replicate population. Density of the cultures
was controlled, as a standard number of parents were given two hours to lay their eggs. 120 hr
after hatching, wandering 3rd instar larvae were placed in phosphate buffered saline (PBS) and
separated by sex. Their guts were carefully removed posteriorly and their body was inverted
anteriorly to expose the brains, eye discs and mouth hooks. After fixation and labeling (described
below), eye discs were isolated and mounted. The MF moves one row every 90 minutes; thus, the
expectation is that the MF will be consistently located within the same approximate location.
Further, replicates were not conducted concurrently meaning that variation due to differences in
timing will be randomly distributed across genotypes and sexes. If there are systematic differences
in the position of the furrow, for example because there is some evidence that D. simulans develops
approximately half a day more slowly than D. melanogaster, we test explicitly for an effect of the
position of the morphogenetic furrow on gene expression in the following sections, for both species
and sex (Génétique, 1983; Kuntz & Eisen, 2014).
2.3.3. Hybridization chain reaction (HCR)
HCR is unique in that it produces gene expression patterns that are both quantitative and
spatial. While other approaches such as FISH can be adapted to detect individual transcripts, HCR
has a linear signal that is 20× brighter than FISH, it reduces non-specific background staining, and
26
it can detect 88% of single RNA molecules in a cell with an appropriately low false discovery rate
(Ma & Moses, 1995; Pan & Rubin, 1995). It is also highly repeatable, with different sets of probes
targeted to the same gene showing correlations of .93–.99 (S. Fraser, pers. comm.). The DNA
probes were designed and synthesized by Molecular Instruments (Choi et al., 2014). Four genes
were multiplexed in each preparation as orthogonally-designed hairpins allowed the simultaneous
amplification of their target sequences (Figures 1 and S1). Each target mRNA was detected using
five DNA probes to annotate the position and expression levels for each of the four assayed genes
(hairy, atonal, Delta and hedgehog). Each probe contained two-initiator sequences (I1 and I2) that
bound to a specific amplifier.
The protocol for HCR was modified from Choi et al. (2014) and is described briefly. The full
protocol is available in S1 File. Inverted 3rd instar larvae were fixed in 4% paraformaldehyde
diluted with PBS containing .2% Tween 20 (PBST). After fixation, larvae were washed with
PBST, then increasing concentrations of methanol (30%, 70%, and 100%) at 25°C. Larva were
stored in 100% methanol at −20°C. Methanol-fixed samples were thawed, washed with ethanol,
re-permeabilized in 60% xylene, washed with ethanol, then methanol and rehydrated with PBST.
Samples were permeabilized with proteinase K (4 mg/ml), fixed in 4% formaldehyde then washed
with PBST at 25°C. Finally, at 45°C, samples were pre-hybridized for 2 hr before the addition of
all the probes. The probe-hybridized larvae were placed in wash buffer (Molecular Instruments) at
45°C to remove excess probes. Fluorescently labeled hairpins were snap-cooled then added to the
samples at 25°C and placed in the dark to amplify the signal. Afterwards, samples were washed in
5X SSCT solution, isolated in PBST, then placed in Prolong Gold antifade mounting medium
(Molecular Probes).
27
2.3.4. Microscopy
Three dimensional images of mounted, HCR stained 3rd instar larva eye discs were acquired
using ZEN on a Zeiss LSM 780 laser scanning microscope (Carl Zeiss MicroImaging, Inc.,
Thornwood, NY) with Objective Plan-Apochromat 63×/1.40 Oil. The gain was adjusted to avoid
pixel saturation.
2.3.5. Image Processing
We performed our spatial and quantitative analysis by creating a 2D spatial gene expression
profile to compare species, genotypes and sexes. We examined whether the regulatory relationship
between these genes evolved between species or contains variation within populations. We also
analyzed whether or not the spatial differences of these genes relative to the MF has evolved or
varies within populations. The full details can be found in our publication (Ali 2019).
Morphological reconstruction and contrast mapping segmentation were performed using a
version (http://sourceforge.net/p/ prostack/wiki/mrcomas) of the MrComas method modified for
3D images to detect mRNA transcripts for each gene. We have published previous work on these
image analysis methods (Kozlov 2017 & Ali 2019).
Briefly, this approach first enhances contrast within the image and reduces noise. After a
nearest-neighbor algorithm, morphological reconstruction, using opening and erosion (Vincent
1993), was used on all images to remove very dark or bright spots, and connect them (Kozlov 2017
& Ali 2019). Using contrast mapping to obtain a pixel-by-pixel difference of the reconstructed
images, we can produce a mask for each channel. After visually inspecting the quality of
segmentation between object borders and the original image, we return masks to their original size
and quantitative measures are made, such as volume, area, and intensity.
28
2.4. Chapter 2 Results
2.4.1. Individual spatial gene expression patterns
First, to characterize the spatiotemporal dynamics of transcriptional activity along the anterior-
posterior axis, we took the spatial average of signal across the dorsal-ventral axis and compared
between genotypes, sexes, and species (Figure 11a, we note that the smoothed curves in the figures
were created using smooth.spline in ggplot2, which is not the same method for curve fitting as in
the analysis). We found abundant spatial quantitative variation in expression profiles (Fig 11 &
Fig. 12). The expression profile of hairy around the MF harbors variation between genotypes and
there is an interaction between genotype & sex (Table 1, p = 2 × 10−3, p = 0.02). There has also
been evolution between species for hairy (Table 1, p = 3 × 10−4). While atonal has not evolved
between species, there is variation in expression profile between genotypes, sexes, and there is an
interaction between genotype & sex (Figures 12a–12c, Table 1, p = 4 × 10
−4
, p = .02, p = .02).
Surprisingly, given the conservation of Delta in general, Delta harbors variation in spatial
quantitative expression behind the MF between genotypes and sexes (Table 3.1, p = 2 × 10−3, p =
7 × 10−4) and there are significant interactions between genotype & sex (Figures 21b and 21c,
Table 1, p = 2 × 10−3). There has also been evolution of Delta between species, and evolution of
the interaction between species and sex (Table 1, p = .03, p = 3 × 10−4). hedgehog is not different
between species but is significantly different between genotypes, sexes, and there is an interaction
between the two (Table 3.1, p = 5 × 10−4, p = .05, p = 1 × 10−3). There is also a significant
interaction between species and sex (Table 3.1, p = .01).
Thus, hairy and Delta have evolved different spatial quantitative expression patterns between
species, while Delta, atonal, and hedgehog harbor variation between genotypes and sexes. Given
that there are regulatory network. However, whatever the source of “buffering” of the network, be
29
it the effect of other genes or threshold effects on development, the fact that this information is not
retained within the steps of the pathway supports our supposition that this variation does not
ultimately have a phenotypic effect.
Figure 10: HCR Results for Eye Disc Development
(a) A summary of the eye patterning genes. The position of the MF is shown in red, and its
direction of movement indicated below. Regulatory relationships are illustrated either as
repression (bar) or activation (arrow). Regulatory relationships which are unclear are shown
as gray dotted lines. (b) HCR image illustrating gene expression patterns of each gene for D.
melanogaster (left) and D. simulans (right). Note there are different shapes of individual
imaginal discs. Antenna discs were not included in the analysis. The composite image
demonstrates multiplexing. (c) Illustration of the expression pattern of each of the four genes.
Anterior is to the left in all panels. (From Ali 2019).
30
2.4.2. Variation and evolution of the eye patterning gene network
The regulatory effects between genes can be discerned from the degree of quantitative co-
localization, and variation in this co-localization between species, genotypes, and sexes. There has
been evolution in the regulatory logic of hairy and its upstream regulators Delta and hedgehog
between species (Figure 13, Table 2, p = .03). There is also variation between sexes in the
regulatory logic of hairy and its upstream regulators Delta and hedgehog (Table 2, p = .03). There
has been significant evolution of the regulatory logic of atonal, in a significant interaction between
species and sex (Figure relationships between these genes, it is interesting to see that they do not
all harbor variation for the same factors. This could potentially be due to the influence of other
unmeasured regulatory factors, or to variation in the relationship between these genes and other
components in the gene regulatory network. However, whatever the source of “buffering” of the
network, be it the effect of other genes or threshold effects on development, the fact that this
information is not retained within the steps of the pathway supports our supposition that this
variation does not ultimately have a phenotypic effect.
2.4.3. Variation and evolution of the eye patterning gene network
The regulatory effects between genes can be discerned from the degree of quantitative co-
localization, and variation in this co-localization between species, genotypes, and sexes. There has
been evolution in the regulatory logic of hairy and its upstream regulators Delta and hedgehog
between species (Figure 13, Table 2, p = .03). There is also variation between sexes in the
regulatory logic of hairy and its upstream regulators Delta and hedgehog (Table 2, p = .03). There
has been significant evolution of the regulatory logic of atonal, in a significant interaction between
species and sex (p = 1 × 10−3). Furthermore, while there was no significant effect of genotype for
31
hairy, there is for atonal, indicating that there is variation segregating in the population affecting
the relationship between atonal, hedgehog, and Delta (p = 1.6 × 10−5). There is also a significant
interaction between genotype and sex (p = 1 × 10−3). Thus, the relationship between hairy and
atonal and their regulators has evolved between species and sexes in hairy, and between genotypes
and sex in atonal. We illustrate this difference between species in Figure 23, where a different
relationship between hairy and hedgehog is visible between D. melanogaster and D. simulans. In
brief, the frequency of cells with a given log transformed level of expression are plotted against
one another for hairy and hedgehog. hairy is primarily expressed anterior to the MF and hedgehog
posterior, and they have a different regulatory relationship in each region with hedgehog activating
hairy indirectly at long range (anterior) and repressing it short range (posterior). This is reflected
in the frequency of cells expressing both genes for D. melanogaster, where anterior to the MF
there is a high frequency of hairy expressing cells and a low frequency of co-occurring high
hedgehog expression. Posterior to the MF the opposite is true, with high expression of hedgehog
lacking concordance with any expression of hairy. In D. simulans, posterior to the MF, this
relationship is the same as in D. melanogaster. However, in anterior to the MF this is not the case.
Expression of hairy and hedgehog both increase as the other increases, with widespread co-
occurrence.
32
Figure 11: HCR Results for Eye Disc Development
(a) Example of fitting a curve to the gene expression profile, measured as the average in a
given row (x-axis). The hexagons are intended to represent cells with varying amounts of
hairy expression, from highest (red) to lowest (white). (b) Illustration of variation in Delta
expression between species and sexes. Curves shown are fitted to all genotypes within a sex
and species. (c) Illustration showing how Delta expression varied between species and sexes,
with lower expression in D. melanogaster females and D. simulans males. (d) Evolution of
Delta illustrated within the context of the gene network, showing that changes in Delta
expression are not perturbing the gene network and result in phenotypically normal
Drosophila. (From Ali 2019).
33
Figure 12: HCR Results for Eye Disc Development
(a) Illustration of variation in atonal expression between genotypes and sexes. Curves shown
are fitted to each genotype and sex. (b) Illustration showing how atonal expression varied
between genotypes, with lower expression in females of D. melanogaster R153 and males of
D. simulans Sz208. D. simulans Sz173 have lower expression than females of Sz208 but it is
not sexually dimorphic. (c) Evolution of atonal illustrated within the context of the gene
network, illustrating how changes in atonal expression are not perturbing the gene network
and result in phenotypically normal Drosophila. (From Ali 2019).
34
2.4.4. Variation and evolution of MF structure
The amount that the eye discs were shifted is not significant for genotype, sex, or species. This
suggests that any differences in development time are not significant between
species, genotype, or sex, and/or that the exact position of the furrow is not having an effect on
these gene expression patterns. However, the amount that they were scaled is, after accounting for
original differences in size, between species (p = 1.38 × 10−6). This suggests that the total relative
Figure 13: Variation in Regulatory Logic.
Variation in regulatory logic between D. simulans and D. melanogaster for hairy and
hedgehog. The heat map illustrates the density of points, and thus reflects the frequency of a
given co-expression profile between hairy and hedgehog. Gene expression values were log-
transformed to better illustrate lower values and split between anterior to the MF and
posterior to the MF. The split between the two regions investigates the possibility that
hedgehog had a different regulatory relationship with hairy depending upon its relationship to
the MF, given that hedgehog is thought to activate hairy long range (indirectly) and repress
hairy short range. (From Ali 2019).
35
width of the MF varies between species, but not between genotypes or sexes. This is also
suggestive of evolving interrelationships among genes that could result in broader or narrower
areas in which they enhance or suppress expression of one another.
2.4.5. MANOVA Results
TABLE 2 The results of the MANOVA for species × genotype × sex for the regulatory
relationship between hairy, Delta, and hedgehog, and atanol. Significant values are shaded
grey. (From Ali 2019).
TABLE 1 The results of the MANOVA for species × genotype × sex for each gene.
Significant values are shaded grey. (From Ali 2019).
36
2.5. Chapter 2 Discussion
We present here a replicated, quantitative, and spatially explicit analysis of the expression of
genes driving cell differentiation in the Drosophila eye. Our results summarize a complicated
pattern of variation sorting in the gene network involved in patterning the MF. For example, the
overall shape of the expression of hedgehog across the eye disc is different between genotypes,
sexes, and there is an interaction between species and sex and genotype and sex. hedgehog
upregulates hairy, but hairy has differences in expression between species (which hedgehog does
not), genotypes, and there is an interaction between genotype and sex. Thus, the differences seen
in upstream regulators, such as hedgehog, are not recapitulated in their downstream targets. In
another example, Delta/Notch is expected to repress atonal, but while Delta/Notch is significant
for all categories tested atonal is only significant for genotype, sex, and their interaction. It is
possible that this variation is being mitigated or dampened by other regulatory factors not assayed
here, or that certain aspects of genetic background are more or less sensitive to variation. For
example, fixed variation between species could dampen variation at Delta/Notch while sorting
variation remains sensitive between genotypes, which propagates to atonal .
It may be that all of this variation is within levels tolerated by the network, as it has been
shown that gene networks can have thresholds of variation, below which differences in expression
are effectively neutral. These thresholds can also be two sided, creating a sigmoidal curve the
center of which is neutral phenotypic space (Felix 2015). Many studies have shown a relative
insensitivity to variation in gene dosage, for example, in Drosophila early embryos the bicoid
gradient results in normal development at one to four dosages of the gene, but markedly abnormal
development at six or more (Liu 2013, Lucas 2013, & Namba 1997).
37
It is also possible that the “genetic stochasticity” documented in these genes is in fact
deleterious, and is being compensated for elsewhere in the network. While most deleterious
mutations are purged by selection, they may rise in frequency due to genetic drift or hitchhiking,
among other possible causes (Burch 1999, Chun 2011, Estes 2003, & McKenzie 1988). This type
of compensatory mutation has been documented in microbial and animal systems (Burch 1999,
Brown 2010, Charusanti 2010, Estes 2003, Estes 2011, Maisnier-Patin 2004, McKenzie 1993,
McKenzie 1982, Moore 2000, Stoebel 2009, & Szamecz 2014). Recently cell cycle heterogeneity
has been implicated in the appearance of widespread noise in development, however this is likely
not responsible for the genetic stochasticity observed here as the majority of MF cells assayed here
are arrested at G1 (Keren 2015 & Kumar 2013).
The differences in the co-expression patterns of hairy, atonal, and their upstream regulators
hedgehog and Delta/Notch may represent the early stages of developmental systems drift, where
pathway components can evolve different relationships while maintaining the same phenotypic
output (True 2001). This type of “quantitative developmental systems drift” has been proposed to
be a pervasive feature of evolving developmental systems (Crombach 2016; Wotton et al., 2015).
It may also be that the overall relationship between the genes has not evolved, but that variation
between the network components is beneath some critical threshold that would results in
differences in downstream regulation—that is that co-expression patterns may show evolution and
variation, but the basic inputs and outputs of the network are conserved. These are not
fundamentally different ideas, rather they may represent different points on an evolutionary
continuum. If developmental systems drift is at work, it could potentially be due to changes in
these critical thresholds, for example if the threshold is lower in one species then less of an
38
upstream gene is needed to create the same output. The evolution and maintenance of expression
thresholds and developmental systems drift is potentially an interesting area of future research.
There have been other semi-quantitative approaches to studying spatial gene expression
patterns. In another study on orthodenticle, the authors found that the spatial and temporal pattern
of gene expression was conserved but the amount of gene product was not, though this work was
not strictly quantitative given that measurements were from in situ hybridization and reporter
constructs and there was no rigorous statistical testing (Goering 2009). This is in contrast to our
results which showed significant differences in the spatial relationship between gene expression
patterns between species. Other semi-quantitative works on the Drosophila embryo using in situ
hybridization found that the regulatory relationship between genes in the anterior-posterior
blastoderm patterning network were conserved, despite differences between species in their spatio-
temporal pattern (Fowlkes 2011 & Wunderlich 2012). Here we find that the regulatory relationship
between atonal and hairy, and their regulators hedgehog and Delta, has evolved between species,
sexes, and genotypes.
One of the important messages from this work is that rigorous statistical testing can
uncover molecular variation in spatial and quantitative developmental gene expression. Using the
type of replication applied in quantitative genetics with developmental data we were able to apply
rigorous statistical models to micro-evolutionary variation in development. Despite this variation,
observed with repeatable observations of developmental patterns among natural genotypes, the
phenotypes of all flies are normal. This points to a potential abundance of hidden noise in spatial
and quantitative gene expression. The evolutionary approach to development generally targets
large changes that have occurred over broad phylogenetic distances (Ito 2013, Jeong 2008, Kopp
2000, Reed 2011, Rosenblum 2010, Signor 2016, & Yassin 2016). Accordingly, the presence of
39
abundant underlying variation is perhaps not a huge surprise. But it does have large implications,
as it suggests that developmental models should be modified so that such abundant genetic
variation is buffered from perturbing the final phenotype. In the future, application of this type of
replicated, quantitative, spatially resolved data will have unique insights into the penetrance of
disease phenotypes and the origin of developmental defects.
DATA AVAILABILITY
All data associated with the manuscript can be found at: https://doi.org/10.5061/dryad.410t5. This
work was supported by grants U01GM103804 and RO1GM102227 to S. Nuzhdin.
40
3. Chapter 3: Threshold Response to Developmental Robustness
3.1. Chapter 3 Abstract
The biological development of complex structures requires robustness to achieve a functional
form, despite the intrinsically stochastic processes that produce this high degree of order. This
robustness relies on Gene Regulatory Networks (GRNs) and clean-up mechanisms. The
Drosophila complex eye is an excellent system to understand these interdisciplinary topics,
whereby cell specification in larval eye imaginal discs manifest into a final functional adult eye
after metamorphosis. The adult eye is composed of repeating ommatidia. This crystal-like structure
is generated by GRNs that repeatedly place the central photoreceptor, R8, by diffusion within the
morphogenetic furrow, a tissue-indentation, then use other mechanisms to advance the MF. The
R8 placement is crucial in determining the final spacing between repeating eye units. Problems in
this process can lead to a “rough eye” phenotype, which has served as a model for various diseases.
We created a mathematical model to simulate morphogen diffusion, R8 placement & MF
progression in this well-studied spatiotemporal process. We quantified order within different
spatial patterning scenarios and observed a universal sigmoidal response to increasing
transcriptional noise. The simulated eye disc remains ordered, despite increasing noise, up to a
certain threshold. Consistent with previous qualitative studies, once this point-of-no-return is
breached, the spacing of eye units, or ommatidial order, rapidly degrades. Our results suggest that
this threshold response provides developmental robustness to prevent observable phenotypic
changes by buffering against the accumulation of genetic variation and transcriptional noise in
natural populations.
41
More details about this chapter’s work can be found in our publication:
Courcoubetis G, Ali S, Nuzhdin SV, Marjoram P, Haas S (2019) Threshold response to
stochasticity in morphogenesis. PLoS ONE 14(1): e0210088. https://doi.org/10.1371/journal.
pone.0210088
3.2. Chapter 3 Introduction
3.2.1. Eye Development Summary
Biological development requires robustness to coordinate complex processes, such as cellular
growth, replication, differentiation and patterning. This robustness is maintained through an
intricate web of interacting Gene Regulatory Networks (GRNs). As species diverge, the
stochasticity of expression in GRNs can lead to developmental systems drift, whereby genetic
variation accumulates between populations without any phenotypic change. The Drosophila
compound eye, made of repeating eye units, or ommatidia, is an excellent system for understanding
many of these processes.
In the developing eye-antenna imaginal disc of the larva, the precise placement of the initial
R8 photoreceptor cell will dictate the offset spacing between eye units, after which clean-up
mechanisms refine the hexagonal structure to create the near-perfect crystal lattice of the adult eye.
For the full details of this process and the evolutionary topics, see chapter 1.
Green Fluorescent Protein (GFP) was fused to DE-Cadherin to show ommatidial order around
the MF (Figure 14). Posterior to the MF (solid green line), individual eye units, or ommatidia
42
(bright green dots), are highly ordered with equal, offset spacing. Anterior to the MF, cells are
proliferating and undifferentiated. Within the MF, cell cycles are synchronized.
3.2.2. Previous Eye Disc Models and Innovation
Problems in this eye development process can lead to a “rough eye” phenotype, which has
served as a model for various diseases. This interdisciplinary research has important public health
implications. Understanding these complex topics will help elucidate the genetic basis of disease,
especially the role of incomplete penetrance and genetic background effects in developmental
defects (Wolff 1991). Understanding developmental robustness is crucial for developing models
that can predict molecular and mechanistic properties from biochemical circuits (Averbukh 2017).
Figure 14: Ommatidial Order Around The Morphogenetic Furrow (MF).
Confocal live imaging of a Drosophila eye disc from a 5-day old larva expressing DE-
Cadherin(GFP). Posterior to the MF (solid green line), individual eye units, or ommatidia
(bright green dots), are highly ordered with equal, offset spacing. Anterior to the MF, cells are
proliferating and undifferentiated. Within the MF, cell cycles are synchronized. Scale bars =
30µm, (from Courcoubetis 2019).
43
The various processes that guide the formation of the larval imaginal eye disc into the fully
functional adult compound eye have been studied extensively, thereby allowing several studies to
generate biologically-relevant models (Gavish 2016 and Lubensky 2011). While these models can
produce the highly-ordered, offset hexagonal lattice, they are limited to qualitative analysis of
spatial order when this pattern is disrupted.
The development and patterning of the Drosophila compound eye is an excellent system
for understanding many of these processes. The adult compound eye is composed of ~800
individual eye units, or ommatidia, that are almost perfectly offset from each other. Each
ommatidium is made up of 14 cells. The placement of each eye unit and role of various cells is
decided much earlier, during larval development of the imaginal eye disc. The morphogenetic
furrow (MF) is a physical indentation in the developing eye disc that sweeps anteriorly thereby
dictating the precise positioning of ommatidia to form the final, functional fly eye. Many of the
genes involved in the regulatory logic of the fly eye are conserved in mammals. Some of the most
important genes involved in fly eye development are hedgehog (Hh), Delta (DI), atonal (ato), and
hairy (h).
Given that genetically-encoded variation in expression can accumulate due to
developmental robustness, how does a developing organism buffer against stochastic components
to achieve a deterministic outcome?
To answer these questions, we simulated ommatidial formation in the Drosophila eye disc
and quantitatively assessed parametric stochasticity. We created a mathematical model to simulate
morphogen diffusion, R8 placement & MF progression in this well-studied spatiotemporal
process. We quantified order within different spatial patterning scenarios and observed a universal
sigmoidal response to increasing transcriptional noise. The simulated eye disc remains ordered,
44
despite increasing noise, up to a certain threshold. Consistent with previous qualitative studies,
once this point-of-no-return is breached, the spacing of eye units, or ommatidial order, rapidly
degrades. Our results suggest that this threshold response provides developmental robustness to
prevent observable phenotypic changes by buffering against the accumulation of genetic variation
and transcriptional noise in natural populations.
3.3. Chapter 3 Methods
3.3.1. Summary of Eye Disc Mathematical Model
We modified these newer models to quantitatively measure the robustness of ommatidial
ordering by simulating R8 placement and altering transcriptional noise, or stochasticity, in
Figure 15: Simulating Morphogenetic Furrow (MF) Movement.
Schematic showing (a) ato-expressing cells (red) in the MF (dark gray lines), with an
inhibition zone in posterior ato-expressing cells (red with gray circles), and (b) the new row
of ato-expressing cells (green) as the MF moves forward. Anterior is to the right. (from
Courcoubetis 2019).
45
different eye disc GRNs. During eye disc development, cells produce diffusible morphogens that
affect neighboring cells’ outputs in a non-linear fashion. This means the concentration of a
particular morphogen within a cell does not necessarily produce a uniform production or diffusion
rate. Thus, reliably forming the near-perfect compound eye pattern requires robustness in the
overall GRNs, despite the transcriptional noise, of individual genes.
We focus on the placement of R8 cells, which is a crucial step in the determination of the final
position of ommatidia within the compound eye to affect the organism’s visual acuity, and the
progression of the MF.
While some morphogens can diffuse, others are bound to the cell. In our model, ato is the only
non-diffusible signal. All other morphogens in our model are diffusible. When present in a cell, a
particular morphogen can either lead to activation (increase in production rate of another
morphogen) or inhibition (decrease in production rate of another morphogen). The MF is treated
as an inductor, which causes undifferentiated cells to produce ato. The MF is assumed to have a
constant velocity (V). The initial conditions are evenly spaced differentiated cells in the most
posterior portion of the simulated eye disc. There is an inhibitory signal that blocks ato above a
threshold. This results in the inhibition of neighboring cells from expressing ato, leading to an
evenly-spaced, offset pattern of ato-expressing cells (Figure 15).
46
In order the make the model more biologically relevant, we looked at ato-expressing cell
clusters and included sca as a short-range diffusible activator that is faster than the MF. This allows
the first uninhibited cells to create a circular activation region. These assumptions are supported
by several other studies that modelled the eye disc (Gavish 2016 and Barkai 2009). The regulatory
logic of our mathematical model is shown in figure 16A, where h acts as a sole activator of a, a
represent atonal which increases production of itself, leading to the activation of u and s. The
inhibitor u decreases production of h and a, while s, representing sca, is a diffusible short-range
activator of a (Figure 16).
3.3.2. Mathematical formulation
The complete details of our mathematical modeling are described in our publication
(Courcoubetis 2019). This work was an interdisciplinary collaboration with George Courcoubetis
and Dr. Stephan Haas in the physics department. Below is a brief description.
Figure 16: Mathematical Modeling of Ommatidia.
(a) Diagram showing the regulatory logic of our mathematical model, where h acts as a sole
activator of a, a represent atonal which increases production of itself, leading to the activation
of u & s. The inhibitor u decreases production of h & a, while s, representing sca, is a
diffusible short-range activator of a. (b) A simulation run showing perfectly offset, evenly-
spaced ommatidia (highly ordered). Anterior is to the right. (From Courcouebtis 2019).
47
The formation of the repeating eye unit in flies follows a stereotypical series of events.
Cells within the posterior portion of the eye-antenna imaginal disc of larva commit to the eye cell-
fate by expressing Hedgehog (hh). The posterior portion of the imaginal disc invaginates, creating
a physical indentation that moves anteriorly through the careful coordination of various
morphogens. This anteriorly sweeping wave is known as the Morphogenetic Furrow (MF). Cells
that are distantly anterior to the MF proliferate extensively. Cells directly anterior to the MF begin
to arrest in G1, which allows cell cycle of cells within the MF to be synchronized. Cells directly
posterior to the MF begin expressing atonal (ato), which forms the first brick upon which the
individual eye unit, or ommatidium, is built. Cells expressing ato inhibit neighboring cells from
expressing ato. This leads to a stripe of ato-expressing cells, followed by evenly-spaced clusters
of 3 ato-expressing cells, and finally evenly-spaced single ato-expressing cells. Each individual
ato-expressing forms the R8 photoceptor, after which Notch/Delta (DI) lateral inhibition dictates
the role of the other 13 cells within an ommatidium. The placement of the final R8 cells is
responsible for the crystal-like, hexagonal offset structure of the compound fly eye.
The speed at which the MF sweeps through the eye disc is controlled by several cellular
processes, primarily the diffusion of various morphogens. Activators, such as hh and scabrous
(sca) act on R8 cells to help move the MF anteriorly, while inhibitors, such as hairy (h), hinder
MF progression.
48
We modified mathematical models from Lubensky 2011 and Gavish 2016 to simulate MF
progression and R8 cluster refinement. The model simplifies certain aspects of eye disc
development; there’s a fixed number of cells and they’re arranged along a hexagonal grid. Thus,
every ommatidia has 6 adjacent neighbors. These assumptions are justified since eye discs develop
primarily in 2D (Lawrence 1986 and Barkai 2009).
Equations 1 through 4 describe the main morphogens involved in MF progression. Each
R8 cell is marked with a for ato. All diffusible inhibitors are represented by u. The diffusible
short-range activator sca is denoted by s. The MF, which has a constant velocity (v), is described
by h for hedgehog and its diffusivity (Dh), Production rate (Ph), and reaction time scale (τh). Each
morphogen has a unique time-scale (τ), and the −λ coefficient specifies the morphogen’s
spontaneous decay rate. These two terms introduce noise into the system by varying the average
lifetime of the morphogen. Equations 2 and 3 introduce the diffusion relationship between u, the
inhibitor of a, and s, the activator of a.
The initial concentration of a is defined by equation 5. We compared numeric simulations
with experimental imaging data to obtain biologically-plausible parameter configurations for MF
49
initiation, morphogen decay rates, and R8 cluster placement and refinement (Gavish 2016 and
Lubensky 2011).
After placement of the initial row of a, an inhibition radius is generated to simulate the
offset pattern of ommatidia found in adult fly eyes and avoid edge-effects. Using experimentally
validated noiseless starting conditions (Gavish 2016 and Lubensky 2011), we can simulate the
stochasticity in expression. For any given model parameter, the final production value is generated
by randomly assigning a value from a normal distribution centered at the mean to produce the
ordered R8 pattern (Figure 20A). A unique parameter value is obtained for each cell in the system.
The Guassian distribution’s center and width is the same for all cells, with more noise
corresponding to more width. This noise can be anywhere from 0% to 60% of the mean (Figure
18). Equation 6 describes how noise was introduced to the inhibitor u: As the MF moves anteriorly
during eye development, the numerous ato-expressing cells that are now posterior to the MF group
into evenly spaced clusters of roughly three neighboring ato-expressing cells and then refine these
clusters to a single ato-expressing cell (Kumar 2012). This single ato-expressing cell becomes the
central R8 photoreceptor, upon which the rest of the cells within the ommatidia are built, and which
dictates the spacing between neighboring ommatidia that form the final compound eye (Voas
2004). The roles of the other cells within the ommatidia are determined by Notch/Delta lateral
inhibition (Baker 1996). Cluster refinement increases eye pattern robustness (Gavish 2016). We
simulated this cluster refinement process by selecting the most central cell within the most
populated row of a given cluster, as illustrated in Figure 18.
50
3.4. Chapter 3 Results
3.4.1. Threshold Response to Increasing Stochasticity
We ran simulations at 4 noise levels, 0% (perfect lattice), 30%, 40% and 60%. For each noise
level, 50 realizations were simulated. The simulated eye disc width was chosen to allow sufficient
space to avoid issues with clusters forming on the edge. 120 lattice sites were simulated per eye
Figure 17: Patterning Disruption with Increased Stochasticity.
Simulations showing cluster positions and disorder of ato-expressing cells where
transcriptional noise ( σ/ μ) is (a) 0 (b) 30%, & (c) 40%. (From Courcoubetis 2019).
51
disc to observe how posteriorly misplaced ommatidia might propagate errors in more anterior
ommatidia. Each run ended when the MF passed through the entire eye disc.
Our simulations showed a threshold response, whereby the eye disc pattern can withstand
ommatidial misplacement due to stochasticity up to a certain point. After this threshold is crossed,
the perturbations due to stochasticity result in severe malformations. This threshold response is
observed in several quantitative studies (Jiao 2014 and Kim 2016).
3.4.2. Nearest Neighbor Distances and Angle
We refined clusters to a single ato-expressing cell (Figure 18), and calculated nearest
neighbor distances and angles for various noise levels in our simulations (Figure 19). The
histogram bars flatten as noise is increased, indicating an increasingly disordered eye disc.
Figure 18: Cluster Refinement Step.
Simulations showing (a) clusters at 40% noise and (b) cluster refinement to a single point by
taking the center of each grouped ato-cluster. (From Courcoubetis 2019).
52
Figure 19: Nearest Neighbor Distance and Angle
Histograms showing the nearest neighbor distance (a-d) and angle (e-h) as stochasticity is
increased (0%, 30%, 40% & 60%). (From Courcoubetis 2019).
53
3.4.3. Variance Plots
The variance plots in figure 20 show the robust nature of pattern formation in these
stochastic models. The patterning remains highly ordered up to a 20% noise threshold, after
which the variance increases linearly in both angle and distance.
3.4.4. Probability Distance Measures
The cluster refinement step affected some of the angle histogram results, so to rigorously
investigate our results, we used the probability distance order measure to eliminate this potential
bias (Figure 21).
Figure 21: Probability Distance Measures Show Threshold Response.
Fidelity (F) was used to recalculate the response to increased stochasticity for nearest
neighbor (a) distance and (b) angle. The results still showed a threshold response (From
Courcoubetis 2019).
Figure 20: Variance Plots.
Variance plots showing a threshold response ( σ/ μ < 20%), after which a linear increase in
variance for (a) nearest neighbor distance and (b) nearest neighbor angle is observed with
increasing stochasticity (From Courcoubetis 2019).
54
3.4.5. Universal Threshold Response
We found a universal threshold response in all model parameters, including diffusion rates
of u & s, production rate of ato (Pa), production rate due to sca activation (S), and production
rate G, which is due to MF progression (Figure 22).
Figure 22: Universal Threshold Response in All Model Parameters
Fidelity (F) was used to calculate the response to increased stochasticity for (a) diffusion rates
of u & s, (b) production rate of ato (Pa), (c) production rate due to sca activation (S), & (d)
production rate due to MF (G). All showed a consistent threshold response. (From
Courcoubetis 2019).
55
3.4.6. Propagation of Errors
We found that there were more errors in the anterior portion of disordered eye discs and an
increased likelihood of ommatidial misplacement in larger eye discs (Figure 23).
Figure 23: Error Propagation
Fidelity (F) was used to calculate order. (Top) Increased stochasticity was higher in anterior
regions (position 80-120). (Bottom) Larger eye discs (green) tolerated less stochasticity.
(From Courcoubetis 2019).
56
3.5. Chapter 3 Discussion
Biological development requires robustness to coordinate complex processes. The Drosophila
compound eye is an excellent system for understanding how patterning can be affected by
perturbations, such as transcriptional noise. We modeled R8 cell placement and refinement, two
crucial steps in the formation of regular-patterned offset ommatidia. Our simulations showed a
buffered response to parametric stochasticity in the expression of eye development GRNs (Figure
20 and Figure 21). Once this stochasticity overburdens the system, ommatidia are rapidly
displaced, as measured by nearest-neighbor distance and angle (Figure 19 and Figure 20). This
displacement can affect downstream patterning anteriorly (Figure 23 top). Consistent with
previous studies (Gavish 2016 and Lubensky 2011), this biologically-relevant sigmoidal threshold
response to increased stochasticity is a universal feature of our model (Figure 22). Our findings of
error propagation due to eye disc size suggest that adult eyes are functionally constrained (Figure
23). Despite having 30% larger body size, female flies only have 2% larger eyes. This allometric
discrepancy suggests a link between eye size in different sexual contexts. Despite the inherent
stochasticity in the noisy developmental process, fly eyes tend to be phenotypically similar among
different Drosophila species (Gasper 2020). We will focus on the interconnectedness of the Sexual
Determination Hierarchy (SDH) and eye development in chapter 4.
While this variation in gene expression is often neutral, isolated populations can undergo
developmental systems drift to diverge (Metzger 2015), at which point the transcriptional
stochasticity can become deleterious in certain genetic backgrounds (Gibson 2004). Our model
predicts that eye development in natural populations should be robust, whereby transcriptional
noise in eye development GRNs is sufficiently suppressed to form near-perfect offset arrays of
ommatidia.
57
We used mathematical modeling to simulate perturbations in ommatidial placement in the
Drosophila eye disc. We observed a threshold response to stochasticity in the simulated eye disc
GRN, whereby the ommatidia remained ordered until stochasticity pushed the system to a point-
of-no-return. Our data suggest that the developmental robustness of these GRNs can buffer against
disruptive processes such as the accumulation of deleterious alleles. This response is critical for
the integration of complex cellular processes to produce a resilient phenotype.
In chapter 3, we verify these predictions using the Hybridization Chain Reaction (HCR) in
males and females of 3 genotypes for Drosophila melanogaster and D. simulans. Our model
predicts that we would observe stochasticity in the eye development GRNs of different Drosophila
species without observing any phenotypic changes in the adults. This stochasticity should be
buffered, whereby different genotypes, sexes or species should have a similar final eye phenotype,
despite variation in the expression of the eye development GRNs.
58
4. Chapter 4: The Role of the Sexual Determination Hierarchy in Eye
Development
4.1. Chapter 4 Abstract
Phenotypic outcomes tend to be robust despite transcriptional noise due to variation in gene
expression. Our previous findings indicate significant variation in eye development between sexes
and genotypes in Drosophila melanogaster (Ali 2019). Despite being 30% bigger in body size,
female eyes are only slightly larger than males (Gasper 2020). This allometric discrepancy, along
with several studies, suggests a link between the Sex Determination Hierarchy (SDH) and eye
development (Penn 2007, Horabin 2003 & Horabin 2005), possibly through the highly-conserved
Notch(N)/Delta(DI) signaling (Gray 1999 & Artavanis-Tsakonas 1999).
Previous studies concentrated on spatial or expression differences in SDH and eye
development, but not both simultaneously.
To thoroughly investigate developmental robustness in interconnected GRNs, such as
sexual and eye development, we used HCR to measure spatial and quantitative differences in
several SDH and eye GRNs simultaneously between developmentally-synchronized 3
rd
instar
larva of pseudomale transformer mutants (tra
1
), males and females. These tra
1
pseudomales have
a larger body size but develop male genitalia and wildtype eyes.
We analyzed expression between atonal (ato), Delta (DI), fruitless (fru), and doublesex
(dsx). Afterwards, we compared the expression profiles of DI, Hedgehog (Hh), dsx, Sexl lethal
(Sxl), & male-specific lethal-2 (msl-2). Consistent with other studies (Penalva 2003), msl-2 was
expressed in all 3 genotypes. We found no significant differences in the DI and Hh expression
profiles of these genes between all 3 developmental contexts, suggesting that both the SDH and
59
eye GRNs we measured remain robust in males, females and pseudomales to generate highly-
ordered, offset ommatidia. This phenotypic plasticity could be the outcome of a buffered network
threshold or compensation from other genes. Further investigation of how these pathways are
interconnected could provide valuable insight on the delicate balance of developmental robustness
and evolutionary adaptability.
4.2. Chapter 4 Introduction
4.2.1. The Sex Determination Hierarchy (SDH)
In Drosophila, sex-specific expression is determined on a cellular basis using a single, multi-
branch Sex Determination Hierarchy (SDH) Gene Regulatory Network (GRN) (Christiansen et al.
2002). The SDH controls X chromosome dosage compensation, somatic sexual differences, and
male-specific courtship behaviors (Yamamoto 2013).
The ratio of X chromosomes to autosomes (X:A ratio) dictates the sex-specific regulation of
cells. Females have an X:A of 1. This leads to the activation of Sex-lethal (Sxl) which uses an auto-
regulatory positive feedback loop to increase Sxl expression. Sxl is a master-switch which activates
transformer (tra) leading to sex-specific splicing of doublesex (dsx
F
) which is important for sexual
development, including genitalia. Males have an X:A of 0.5. Thus, Sxl is not activated in males,
which leads to the activation of male-specific lethal-2 (msl-2) for dosage compensation, and male-
specific versions of dsx
M
and fruitless (fru
M
), which is involved in male-specific courtship
behaviors (Penn 2007).
The Sex-Determination Hierarchy (SDH) in Drosophila is a pivotal pathway for somatic
sexual dimorphism (Figure 5). If the concentration of X-linked Signal Elements (XSEs) reaches a
60
critical threshold at the start of zygotic transcription, then Sex lethal (Sxl) will be irreversibly turned
on. In XX females, Sxl activation is maintained through a self-regulatory loop and results in a
series of splicing events via transformer 1 (Tra 1) and transformer 2 (Tra 2) to lead to the
expression of female-specific Doublesex (dsx
F
). Alternative splicing of fruitless (fru) leads to an
inactive isoform in females (Cachero 2010). In XY males, XSEs never reach the critical threshold
in time, thus Sxl remains inactive. Sxl inactivity leads to the default male development via the
expression of dsx
M
, and the sex-specific expression of fru
M
via the P1 promoter, as well as male-
specific lethal-2 (msl-2) which is important in dosage compensation (Salz 2010). Numerous studies
have concluded that fru
M
plays a critical role in the development of masculinized neurons and
ultimately male courting behavior (Rideout 2010, Christiansen 2002 and Yamamoto 2013). There
are many cells in both sexes that express the common region of fru (fru-COM), however fru
M
is
expressed only in about 2000 male neurons via alternative splicing. These fru
M
expressing neurons
form clusters only in the masculinized brain (Lee 2000). Both the default transcript, dsxM, and the
alternatively spliced transcript, dsx
F
, have multiple instructive roles in the sex-specific
development of the central nervous system, imaginal discs and body size (Arbeitman 2007 and
Lee 2009).
4.2.2. Linking Eye Development and the Sex Determination Hierarchy (SDH)
Our results from Chapter 3 showed sexually-dimorphic differences in Delta (DI) expression &
genotype-by-sex interactions. Interestingly, these dimorphic differences were switched between
species. How is eye development linked to sexual development?
We studied the expression of Hedgehog (Hh) and Notch(N)/Delta (DI) to investigate the sex-
specific deployment of N and understand how phenotypic robustness in eye development can be
61
maintained. Many organisms, including humans, rely on N/DI lateral inhibition for differentiation.
This highly-conserved pathway is instrumental in regulating interactions and developmental
decisions between neighboring cells, including in humans (Gray 1999). The SDH plays a key role
in regulating body size and sexual differentiation (Arbeitman 2007 and Lee 2009). Studies have
found that Sxl directly inhibits N production (Penn 2007). As previously discussed, Hh plays a
vital role in cell specification and tissue patterning. Studies have shown that Hh promotes the
nuclear entry of Sxl in germ and somatic cells (Horabin 2003 & Horabin 2005). Sexual size
dimorphism causes males to develop more slowly (Testa 2013).
Despite being a 30% bigger in body size, females eyes are only slightly larger than males
(Gasper 2020). This suggests a link between eye development and the SDH to alter the allometric
scaling of adult eyes. Interestingly, overexpression of fruitless showed no significant differences
in eye size (Osterwalder 2018). To further investigate the role of SDH in eye development, we
performed crosses to generate transformer mutants (tra
1
) pseudo-males. These tra1 pseudomale
flies have two X chromosomes and thus a larger body size but develop male genitalia since they
express DSX
M
. Furthermore, N signaling plays a key role in integrating signals throughout
development (Artavanis-Tsakonas 1999).
4.2.3. Eye Disc Development Gene Regulatory Networks (GRNs)
The MF advances anteriorly through the interaction of several genes with the eye disc GRN,
primarily Hedgehog (Hh) and Decapentapegic (Dpp). These are long-range activators of atonal
(ato). Cells that express ato inhibit their neighbors from also expressing ato (Greenwood 1999 and
Baonza 2001). Thus, as we move posteriorly from the MF, the concentration of ato-expressing
cells is slowly reduced from a line of tightly-packed ato-expressing cells, to a clusters of 3 ato-
expressing neighbors, and finally to single ato-expressing cells that are evenly spaced (Voas 2004).
62
These evenly-spaced cells will become the R8 photoreceptor within each ommatidium. This is the
initial brick, upon which the rest of the eye unit is built. Notch/Delta (DI) lateral inhibition is used
for ato cluster refinement and to specify the role of the cells surrounding the individual ato-
expressing cells while Scabrous (sca) is a short-range diffusible activator of ato (Gavish 2016).
Anterior to the MF, undifferentiated cells are expressing hairy (h), which inhibits ato and allows
cells to proliferating asynchronously (Fried 2016).
The speed at which the MF sweeps through the eye disc depends on several factors, including
the interplay between Hedgehog (Hh), which acts posteriorly as an activator, and hairy (h), which
acts anteriorly as an inhibitor. Within the MF, a band of atonal (ato) is expressed that becomes
refined posteriorly. Cells expressing high amounts of ato ultimately activate Delta/Notch signaling
to inhibit their neighbors. This results in evenly-spaced clusters of ato-expressing cells in each row
that are perfectly offset from the previous and following row of ato-expressing cells. Thus, the
spacing between these ato-expressing cells provides the first brick upon which the final spacing
between adult ommatidia is built.
As the MF is propagated anteriorly by several key developmental genes, or morphogens, in the
eye disc, the precise positioning of ommatidia behind it is determined to ultimately form the final,
fully-functional adult eye. The spacing between neighboring R8 cells is instrumental in the final
placement of the ommatidia. Defects that misplace an ommatidium can affect neighbors which
propagates further misplacement resulting in a distorted array. The regulatory logic that guides
this repeated patterning recycles the same genes within the eye development GRN. Thus,
perturbations via mutations or changes in transcription of a single gene can affect all eye units
(Kumar 2012). Many of the genes involved in the regulatory logic of the fly eye are conserved in
63
mammals. We will focus on hedgehog (Hh), Delta (DI), atonal (ato), and hairy (h), which are
heavily involved in the regulatory logic of fly eye development (Figure 5).
4.2.4. Innovation and significance
Variation in complex traits, such as sexual behavior and disease susceptibility, in natural
populations remains poorly understood since they can be sensitive to environmental and
developmental perturbations (Falconer 1996, Lynch 1998, & Mackay 2009). Furthermore,
phenotypic outcomes tend to be robust despite transcriptional noise due to variation in gene
expression (Bothma 2014 and Felix 2015). One of the highly-coveted goals of quantitative genetics
is to innovate methods that predict developmental outcomes from the genotype-to-phenotype map,
which can aid in personalized disease risk assessment (Mackay 2009). The Drosophila Sex-
Determination Hierarchy (SDH) and eye development Gene Regulatory Networks (GRNs) provide
an ideal system to investigate these complex questions.
Problems in eye development can lead to a “rough eye” phenotype, which has served as a
model for various diseases. This interdisciplinary research has important public health
implications. Understanding these complex topics will help elucidate the genetic basis of disease,
especially the role of incomplete penetrance and genetic background effects in developmental
defects (Wolff 1991). Understanding developmental robustness is crucial for developing models
that can predict molecular and mechanistic properties from biochemical circuits (Averbukh 2017).
Numerous studies have shown that GRNs tolerate perturbations up to a certain point, after
which, the system rapidly breaks down. Although many of these developmental processes have
been investigated extensively, most studies focus on a single process and 1 or 2 genes, reporting
either spatial or expression data, but not both.
64
To thoroughly investigate developmental robustness in interconnected GRNs, such as
sexual and eye development, we used HCR to measure spatial and quantitative differences in
several SDH and eye GRNs simultaneously between developmentally-synchronized 3
rd
instar
larva of pseudomale transformer mutants (tra
1
), males and females. These tra
1
pseudomales have
a larger body size but develop male genitalia and wildtype eyes.
We analyzed expression between atonal (ato), Delta (DI), fruitless (fru), and doublesex
(dsx). Afterwards, we compared the expression profiles of (DI, Hedgehog (Hh), dsx, Sexl lethal
(Sxl), & male specific-lethal-2 (msl-2). Consistent with other studies (Penalva 2003), msl-2 was
expressed in all 3 genotypes. We found no significant differences in the DI and Hh expression
profiles of these genes between all 3 developmental contexts. Males, females and pseudomales all
generated highly-ordered, offset ommatidia. These results suggest that eye development is linked
to the SDH through other genes that we did not measure. Further investigation of how these
pathways are interconnected could provide valuable insight on the delicate balance of
developmental robustness and evolutionary adaptability in various sex-specific scenarios.
65
4.3. Chapter 4 Method
4.3.1. Fly Stocks
To generate pseudomales, we used 2 fly strains graciously provided by Dr. Michelle Arbeitman
and therefore named MA1 and MA2. The crossing scheme is depicted in Figure 24 and 25,
whereby crossing 10 female MA1 and 10 male MA2 virgins results in 3 distinct genotypes:
Wildtype XY males (tubby, male genitalia), Wildtype XX females (tubby, female genitalia), and
XX Pseudomales (no tubby, male genitalia). These pseudomales have a mutation in transformer
(tra
1
), that results in the expression of sex-specifically spliced male version of fruitless (FRU
M
),
and doublesex (DSX
M
), despite having two X chromosomes.
Table 3: ANOVA Results for NECs. Colors indicate NECs that are related.
4.3.2. Staging and Dissection
Flies were raised on standard cornmeal medium at 25 ± 1°C in a 12-h/12-h light/dark cycle.
Vials were used a single time for collection to avoid pseudo-replication, such that all imaged
samples were collected from a separate replicate population. Population density was controlled as
20 parents per vial were given 6 hours to lay their eggs. Developmentally-synchronized wandering
3rd instar larvae were placed in phosphate buffered saline (PBS) and separated by genotype. The
posterior portion of their guts was carefully removed and their body inverted anteriorly to expose
the eyes discs and mouth hooks. After performing HCR (see below), all genotype eye-antenna
discs were isolated and mounted. Replicates were not conducted concurrently such that very slight
differences in developmental time between genotypes are randomly distributed and should not
effect morphogenetic furrow (MF) position (Génétique, 1983; Kuntz & Eisen, 2014).
66
.
4.3.3. Hybridization Chain Reaction
To understand the role of SDH in eye development, we used HCR to image genes within both
GRNs. We used our validated eye disc protocol (Ali 2019) on two different experimental sets to
obtain the mRNA expression profiles. In the first set, we used HCR to visualize the patterning of
of ato, dsx, DI, fru and DAPI in 5 day old 3rd instar larva. In the second set, we used HCR to
visualize the patterning of Hh, dsx, Sxl, DI, msl-2 and DAPI in 6 day old 3rd instar larva. Probe
sequences for all genes were professionally designed by Molecular Instruments
®
.
Figure 24: Generating Pseudomales
Crossing scheme to generate pseudomales. MA1 white-eyed (X
w
) females with a transformer
(tra
1
) mutation and balancer chromosome (TM6B) are crossed with MA2 males.
67
Briefly, developmentally-synchronized 3rd instar larvae were dissected then fixed in 4%
paraformaldehyde diluted with PBS containing .2% Tween 20 (PBST). After fixation, larvae were
washed with PBST, then increasing concentrations of methanol (30%, 70%, and 100%) at 25°C.
Inverted, processed larva were stored in 100% methanol at −20°C. Methanol-fixed samples were
then rehydrated and permeabilized with proteinase K (4 mg/ml), fixed in 4% formaldehyde then
washed with PBST at 25°C. Finally, at 45°C, samples were pre-hybridized for 2 hr before the
addition of all the probes. The probe-hybridized larva were placed in wash buffer (Molecular
Instruments
®
) at 45°C to remove excess probes. Fluorescently labeled hairpins were snap-cooled
then added to the samples at 25°C and placed in the dark to amplify the signal. Afterwards, samples
were washed in 5X SSCT solution, isolated in PBST, then placed in Prolong Gold antifade
mounting medium (Molecular Probes
®
).
Figure 25: Crosses to Generate Pseudomales.
The 3 developmental outcomes from crossing 10 unmated female MA1 and male MA2 flies
are (a) pseudomales that have a large body (XX), and orange-colored eyes (ubi-GFP), (b)
females that developed tubby and orange eyes, and (c) males that developed tubby and white,
bar-eyes (see close up). Images taken with dissecting scope at 1.5X magnification.
68
4.3.4. Microscopy
Three dimensional images of mounted, HCR stained 3rd instar larva eye discs were acquired
on a Zeiss LSM 780 laser scanning microscope (Carl Zeiss MicroImaging, Inc., Thornwood, NY)
with Objective Plan-Apochromat 63×/1.40 Oil. The gain was adjusted to avoid pixel saturation.
We used similar settings from our publication to minimize heterogeneity (Ali 2019).
4.3.5. Image Processing
We performed our spatial and quantitative analysis by creating a 2D spatial gene expression
profile to compare species, genotypes and sexes. We examined whether the regulatory relationship
between these genes evolved between species or contains variation within populations. We also
analyzed whether or not the spatial differences of these genes relative to the MF has evolved or
varies within populations. The full details can be found in our publication (Ali 2019).
Morphological reconstruction and contrast mapping segmentation were performed using a
version (http://sourceforge.net/p/ prostack/wiki/mrcomas) of the MrComas method modified for
3D images to detect mRNA transcripts for each gene. We have published previous work on these
image analysis methods (Kozlov 2017, Ali 2019).
Briefly, this approach first enhances contrast within the image and reduces noise. After a
nearest-neighbor algorithm, morphological reconstruction, using opening and erosion (Vincent
1993), was used on all images to remove very dark or bright spots, and connect them (Kozlov
2017 & Ali 2019). Using contrast mapping to obtain a pixel-by-pixel difference of the
reconstructed images, we can produce a mask for each channel. After visually inspecting the
quality of segmentation between object borders and the original image, we return masks to their
original size and quantitative measures are made, such as volume, area, and intensity.
69
4.4. Chapter 4 Results
4.4.1. Pseudomales (tra
1
) had no eye defects.
To generate pseudo-males, we used 2 fly strains graciously provided by Dr. Michelle
Arbeitman and therefore named MA1 and MA2. The crossing scheme and developmental
outcomes are depicted in Figures 24 and 25. After crossing female MA1 with male MA2, we used
immunohistochemistry to stain Fasciclin III (Fas III) in third instar larva. Our results verify that
tra1 pseudo-males have a robust eye phenotype, with no noticeable defects in the offset, crystalline
array of ommatidia (Figure 3.3). Pseudomales MF moves at a similar pace to females, both of
which develop faster than males. Our finding that males have a slower MF, is consistent with other
studies. This slower male MF is due to size dimorphism between sexes (Testa 2013).
Figure 26: Pseudomales (tra
1
) Mutants Eyes Have No Defects.
Confocal imaging at 63X magnification of Fasciclin III (Fas III) antibody staining
(left) shows highly-structured ommatidial patterning, posterior to the MF (arrow), and (right)
close-up inspection shows a phenotypically wildtype, near-perfect offset ommatidial pattern
(arrows) indicating robust eye development in Pseudomales (tra
1
).
70
4.4.2. HCR Results
Studies have found that Sxl directly inhibits N production (Penn 2007) and Hh promotes
nuclear entry of Sxl in germ and somatic cells (Horabin 2003 & Horabin 2005). Females are 30%
larger in body size but only 2% larger in ommatidia size (Gasper 2020). This allometric
discrepancy suggests a link between Drosophila eye development Gene Regulatory Networks
(GRNs) and the Sexual Determination Hierarchy (SDH). Our previous findings also indicate
significant variation in eye development GRNs between sexes and genotypes in Drosophila
melanogaster (Ali 2019).
We used genetic crosses to produce pseudomales in D. melanogaster. We performed HCR on
developmentally-synchronized 3
rd
instar larva between pseudomales, males (XY) and females
(XX). Firstly, we analyzed expression between atonal (ato), Delta (DI), fruitless (fru), and
doublesex (dsx) (Figure 27).
Afterwards, we compared. the expression profiles of (DI, Hedgehog (Hh), dsx, Sexl lethal
(Sxl), & male specific-lethal-2 (msl-2) (Figure 28). ANOVA suggests no significant differences in
the DI and Hh expression profiles of these genes between all 3 developmental contexts (Figure
3.6). Despite males, females and pseudomales making the msl-2 transcripts, only the XY males
express MSL-2 protein. The msl-2 transcript was found equally in males, females and pseudo-
males consistent with other studies (Penalva 2003). Our ANOVA results suggest no differences
between expression levels of hh and DI in pseudomales, males nad females (Figure 3.6). These
findings suggest that both SDH and eye GRNs remain robust in males, females and pseudomales.
71
Figure 27: HCR results showed no differences in pseudomales for SDH and Eye GRNs.
Imaging results for ato (red), dsx (white), DI (green), fru (blue) and DAPI (cyan) showed no
differences between (A) pseudomales, (B) females and (C) males). Arrow indicates
morphogenetic furrow (MF).
72
Figure 28: HCR results showed no differences in pseudomales for SDH and Eye GRNs.
HCR aging results using DAPI (cyan), dsx (grey), DI (green), Sxl (red), Hh (blue), and msl-2
(magenta) showed no differences between (A) pseudomales, (B) females and (C) males).
Arrow indicates morphogenetic furrow.
A
C
B
73
4.5. Chapter 4 Discussion:
Variation in complex traits, such as sexual behavior and disease susceptibility, in natural
populations remains poorly understood since they can be sensitive to environmental and
developmental perturbations (Falconer 1996, Lynch 1998, & Mackay 2009). One of the highly-
Figure 29: DI and hh Expression Levels Did Not Vary.
Our ANOVA results showed no differences between Pseudomales, females and males in
delta (DI) and Hedgehog (hh) expression.
74
coveted goals for quantitative genetic studies is to innovate methods that predict developmental
outcomes from the genotype to phenotype map. This can aid in individual disease-risk assessment
(Mackay 2009). We studied the expression of Hedgehog (Hh) and Notch(N)/Delta (DI) to
investigate the sex-specific deployment of N and understand how phenotypic robustness in eye
development can be maintained. Many organisms, including humans, rely on N/DI lateral
inhibition between neighboring cells for differentiation (Gray 1999).
Studies have found that Sxl directly inhibits N production (Penn 2007). As previously
discussed, Hh plays a vital role in cell specification and tissue patterning. Studies have shown
that Hh promotes the nuclear entry of Sxl in germ and somatic cells (Horabin 2003 & Horabin
2005). Despite females having a 30% larger body size, their eyes are roughly the same size as
males (Gasper 2020).
We used genetic crosses to produce masculinized XX females, or pseudomales. We
generated developmentally-synchronized larva between pseudomales, males (XY) and females
(XX). We verified that mutants expressing tra
1
had regular eyes (Figure 26). Pseudomales MF
moves at a similar pace to females, both of which develop faster than males. Our finding that males
have a slower MF, is consistent with other studies. This slower male MF is due to size dimorphism
between sexes (Testa 2013).
Afterwards, we performed HCR to investigate quantitative and spatial gene expression profiles
for several genes involved in SDH and eye development. We first analyzed expression between
ato, DI, fru & dsx (Figure 27) , then compared the expression profiles of DI, Hh, dsx, Sxl, & msl-
2 (Figure 28). We used ANOVA and found no differences between Hh and DI levels in males,
females and pseudomales between our samples (Figure 29) Interestingly, dsx was anterior to the
MF, with a dorsal bias. This finding is consistent with other studies which suggest the dorsal-
75
deployment of dsx is a vestigial trait from close evolutionary relatives that have the ancestral trait
of in-flight copulation which was lost by various Drosophila species (Robinett 2010 & Ali 2019).
It may also play a role in finding mates (Gasper 2020). In summary, our results indicate that eye
GRNs remain robust and require phenotypic plasticity regardless of sex-specific development and
transcriptional noise.
Interestingly, alternative splicing generates sex-specific outcomes, such as a larger female
body size (Arbeitman 2007 and Lee 2009). Unfortunately, our HCR probes were not isoform
specific, leaving the possibility that a specific version of dsx is differentially expressed between
males, females and pseudomales. Future experiments can use isoform-specific methods to
investigate this possibility. Furthermore, it would be interesting to see how these SDH genes differ
in the adult brains of males, females and pseudomales.
As previously discussed, biological systems can use transcriptional stochasticity, or variation
in gene expression, to balance phenotypic robustness versus evolutionary adaptability depending
on developmental contexts. Robustness buffers development against perturbations, such as
transcriptional noise, up to a certain threshold, after which the system rapidly degrades. ). We
experimentally verified eye disc robustness by measuring the quantitative and spatial relationship
of 4 key genes in the eye development Gene Regulatory Network (GRN) simultaneously using the
Hybridization Chain Reaction (HCR). Our published findings show that the eyes of Drosophila
melanogaster and D. simulans are phenotypic similar, despite significant transcriptional
stochasticity between species, genotypes and sexes (Ali 2019, Chapter 2). We published a
mathematical model that quantitatively assessed this threshold response in the fly compound eye,
where individual eye units are specified in the larval eye disc to form the highly-organized, offset
patterning seen in adults (Courcoubetis 2019, Chapter 3). We further investigated the eye GRN in
76
different sexual contexts by genetically manipulating the Sex Determination Hierarchy (SDH) and
performing HCR on several genes in male, female and pseudomale (masculinized females) larval
eye discs, which all showed robustness (Chapter 4). To further understand the delicate balance
between developmental robustness and evolutionary adaptability, we used HCR to measure the
quantitative and spatial relationships of Neuron Expression Clusters (NECs) in the SDH between
D. melanogaster and D. simulans brains, which showed transcriptional and volumetric variation
in several reproductively-relevant brain regions (Chapter 5).
77
5. Chapter 5: The Role of the Sexual Determination Hierarchy in
Brain Development
5.1. Chapter 5 Abstract
The link between sex-specific expression patterns that lead to sexually-dimorphic
phenotypes within and between species remains unclear. How do genes, regulatory
networks, cells and tissues integrate the information necessary to generate various
developmental outcomes? Drosophila melanogaster provides an ideal system for
understanding these large-scale processes on multiple biological levels. The sex-
determination hierarchy (SDH) is involved in several dimorphic phenotypes via the sex-
specific alternative splicing of fruitless (fru) and doublesex (dsx). The sex-specific
expression of these 2 transcription factors leads to the formation of sexually dimorphic
interconnected neural circuits. The neuronal network that forms these fru and dsx neuron
expression clusters (NECs) show interspecies differences despite the conservation of the
SDH. This, along with evidence from other studies, suggests that there must be an
alternative pathway controlling the patterning of NECs via upstream fru and dsx regulators
that are independent of the SDH. We used cell-specific quantitative microscopy and
genetic manipulation, to analyze the spatial and expression relationships between NECs
that express fru and dsx.
We found a lot of variation in certain Neuron Expressing Clusters (NECs) of pupa
brains between D. melanogaster vs D. simulans.
78
Can look at noise between both halves of the brain to see internal symmetry/as a control
For genotypes that have a reduced cluster, do you see a smaller amount of that behavior?
ex. males circling females
5.2. Chapter 5 Introduction
5.2.1. Neuron Expression Clusters (NECs) in Brains
Our chapter 2 results drove us to ask how do developmental differences accumulate between
species despite this delicate balance between robustness in development versus adaptability for
evolutionary outcomes? How can variation in complex structures, such as the brain, drive
behavioral differences that lead to speciation? To understand the genes that bridge the complex
relationship between brain development and behaviors, many studies have relied on the link
between SDH in Drosophila and male courtship behavior. This provides a genetically-
manipulatable and behaviorally-observable model system, where the manipulation of individual
components leads to reproducible behavioral outputs. The SDH results in certain Neuron
Expression Clusters (NECs) which influence male courtship rituals. Although the SDH is
conserved between various Drosophila species, which NECS and cells express certain SDH genes
vary. This suggests an upstream signal that operates in a cell-specific manner to influence how
NECs combine in a particular species. How quickly can complex structures, such as brains, evolve
between species despite stochasticity and developmental robustness? How do these differences
lead to changes in courtship behavior between species? We postulated that sexually dimorphic
brains regions evolve faster than monomorphic regions. We used HCR to measure the quantitative
79
and spatial relationships of Neuron Expression Clusters (NECs) in the SDH between D.
melanogaster and D. simulans brains. We found significant transcriptional and volumetric
variation in several reproductively-relevant brain regions.
Several studies have shown that sexually dimorphic traits tend to be under stronger selection
and evolve faster than monomorphic traits, especially in males. Furthermore, male-specific genes
show more variation in expression levels within and between species compared to female-specific
and monomorphic genes (Meiklejohn 2003). For example, the size and orientation of sex combs
in males of several Drosophila sister species is highly variable, despite recent divergence (Tanaka
2008).
5.2.2. The Sex Determination Hierarchy (SDH) in Drosophila
In Drosophila, sex-specific expression is determined on a cellular basis using a single, multi-
branch Sex Determination Hierarchy (SDH) Gene Regulatory Network (GRN) (Christiansen et al.
2002). The SDH controls X chromosome dosage compensation, somatic sexual differences, and
male-specific courtship behaviors (Yamamoto 2013).
The ratio of X chromosomes to autosomes (X:A ratio) dictates the sex-specific regulation of
cells. Females have an X:A of 1. This leads to the activation of Sex-lethal (Sxl) which uses an auto-
regulatory positive feedback loop to increase Sxl expression. Sxl is a master-switch which activates
transformer (tra) leading to sex-specific splicing of doublesex (dsx
F
) which is important for sexual
development, including genitalia. Males have an X:A of 0.5. Thus, Sxl is not activated in males,
which leads to the activation of male specific-lethal-2 (msl-2) for dosage compensation, and male-
specific versions of dsx
M
and fruitless (fru
M
), which is involved in male-specific courtship
behaviors (Penn 2007).
80
The Sex-Determination Hierarchy (SDH) in Drosophila is a pivotal pathway for somatic
sexual dimorphism (Figure 5). If the concentration of X-linked Signal Elements (XSEs) reaches a
critical threshold at the start of zygotic transcription, then Sex lethal (Sxl) will be irreversibly turned
on. In XX females, Sxl activation is maintained through a self-regulatory loop and results in a
series of splicing events via transformer 1 (Tra 1) and transformer 2 (Tra 2) to lead to the
expression of female-specific Doublesex (dsx
F
). Alternative splicing of fruitless (fru) leads to an
inactive isoform in females (Cachero 2010). In XY males, XSEs never reach the critical threshold
in time, thus Sxl remains inactive. Sxl inactivity leads to the default male development via the
expression of dsx
M
, and the sex-specific expression of fru
M
via the P1 promoter, as well as male-
specific lethal-2 (msl-2) which is important in dosage compensation (Salz 2010). Numerous studies
have concluded that fru
M
plays a critical role in the development of masculinized neurons and
ultimately male courting behavior (Rideout 2010, Christiansen 2002 and Yamamoto 2013). There
are many cells in both sexes that express the common region of fru (fru-COM), however fru
M
is
expressed only in about 2000 male neurons via alternative splicing. These fru
M
expressing neurons
form clusters only in the masculinized brain (Lee 2000). Both the default transcript, dsxM, and the
alternatively spliced transcript, dsx
F
, have multiple instructive roles in the sex-specific
development of the central nervous system, imaginal discs and body size (Arbeitman 2007 and
Lee 2009).
5.2.3. Neurons expressing FRU
M
Control Male-Specific Behaviors
81
To understand the genes that bridge the complex relationship between brain development
and behaviors, many studies have relied on the link between SDH in Drosophila and male
courtship behavior. This provides a genetically-manipulatable and behaviorally-observable
model system, where the manipulation of individual components leads to reproducible
behavioral outputs. The steps involved in Drosophila male courtship follow a stereotypical
sequence, starting with visual and olfactory detection to orient the male and follow the female.
Afterwards, the male taps the female to detect pheromones and assess receptivity. The male
then vibrates his wing to generate a song with two components; sine and pulse. The sine song
is a humming sound, while the pulse song consists of inter-pulse intervals that vary between
Figure 30 Male Courtship Behavior in Drosophila Species
(A) The steps involved in male courtship behavior in Drosophila are stereotypical. (B)
Drosophila males modify their courtship rituals in a species-specific way by manipulating
their wing vibration to generate different pulse-songs (from Cande 2014).
82
species (Billeter 2006). Although these behaviors are innate, they can be modified depending
on prior experience with females (Siegel 1979). Thus, Drosophila males modify their courtship
rituals in a species-specific way by manipulating their wing vibration to generate different
pulse-songs (Fig 30).
5.2.4. Linking FRU
M
NECs and Male-Specific Behaviors
As previously discussed, the SDH uses alternative splicing to generate sex-specific
isoforms of various transcriptional regulators. The 2 primary functions of FRU is viability in
both sexes and male courtship behavior. Thus, in addition to FRU-expressing cells in both
sexes, the male-specific isoforms, FRU
M
, are generated by using the fru promoter (P1) to drive
sex-specific alternative splicing (Kohatsu 2011). Thus, FRU
M
forms Neuron Expression
Clusters (NECs) which correspond to different aspects of male-courtship behavior, such as
mate detection, following, tapping, and the wing vibrations that generate the species-specific
song (Fig 31).
Separate analyses of behavioral effects from various fru mutations, including loss-of-
function, nearly null, and mild perturbations indicate that fru
M
plays a vital role in courtship,
including mate-recognition, singing, and mating (Goodwin 1999 & Lee 2000). Furthermore,
FRU
M
is necessary to form the male-specific abdominal Muscle of Lawrence (MOL) via
masculinization of certain neurons (Lawrence 1986 & Gailey 1991). Females with
homozygous fru mutations show no sex-specific phenotypes (Villella 1997).
Alternative splicing produces several fru transcripts from at least 4 different promoters,
including fru
COM
, which is non-sex-specific and therefore “common” in both sexes (Ito 1996
& Goodwin 2000). The P1 promoter produces the sex-specific fru
M
(Ryner 1996).
83
Although the fru is conserved between different species, the fru NECs vary between them.
Studies indicate that replacing the entire fru
M
locus, including regulatory sequences, from
several closely-related species into a D. melanogaster fru-null mutant restores wild type
melanogaster male development indicating an upstream regulator that dictates the patterning
of fru NECs (Cande 2014).
Previous studies comparing fru expression patterns between Drosophila species indicate
general similarities, while cluster-by-cluster comparisons of NECs reveal some differences
between species. Although D. Yakuba is closely related to D. melanogaster and D. simulans,
they do not have the muscle of Lawrence (MOL). Some other distantly-related Drosophila
Figure 31: Neurons Expressing FRU
M
Control Male-Specific Behaviors
An illustration of the anterior fru-expressing neurons in the brain (Left) and ventral nerve
cord (Right) of Drosophila adults (from Billeter et. al. 2006).
84
species have MOL. It should be noted that D. ananassae, D. virilis and D. yakuba do not have the MOL.
Furthermore, some distantly-related Drosophila species were similar in their laminal fru
expression, while closely related species differed. This indicates that functional convergence
(Usui-Aoki 2005). Similar laminal fru expression might indicate similarities in visual
processing.
5.2.5. Sex-specific neuron clustering of fru and dsx between species.
Reproductive success is contingent on species recognition, mate attraction and discrimination
among favorable traits in the opposite sex. Thus, the genes that influence the development of the
neural circuitry that guides sexual behaviors is under selection. Sexual conflict is the selection for
different fitness optima between males and females despite shared genes (Griffin 2013). There are
several ways to resolve sexual conflict, such as changes in magnitude or direction of sex-specific
gene expression, sex-specific regulatory networks and sex-specific alternative splicing.
Drosophila melanogaster has proven to be an outstanding system for understanding the complex
interactions between genes, regulatory networks, neurons and behaviors that shape sexual
dimorphism. Mutations in sex-biased genes often display sex-specific phenotypes in this model
system (Cannollon 2011).
Surprisingly, the decision to be male or female occurs on a cell-by-cell basis in Drosohila
species. For example, the sex-specific behavior of germ cells mimics the surrounding gonad cells,
not chromosomal sex. For example, when XY germ cells are placed in female gonads, the female-
specific developmental program is activated via Sxl (Janzer 2001, Oliver 1993, & Waterbury
2000). Thus, the sex-specific trajectory of the surrounding soma has a large impact on the
development of the entire organism. Several regions in the adult brain are dimorphic. This can be
85
due to an increase in the number of neurons and a rewiring of neuronal connections to form sex-
specific connections. Increasing the number of neurons in a certain region can amplify the signal
strength of a particular cue. However, differences in dendritic arbor wiring and overlap can allow
sex-specific connections to be made. For example, rewiring fru arbors in the lateral horn allows
the pheromone cVA to be perceived as attractive for females and repulsive for males.
Several studies have shown that sexually dimorphic traits tend to be under stronger selection
and evolve faster than monomorphic traits, especially in males. Furthermore, male-specific genes
show more variation in expression levels within and between species compared to female-specific
and monomorphic genes (Meiklejohn 2003). For example, the size and orientation of sex combs
in males of several Drosophila sister species is highly variable, despite recent divergence (Tanaka
2008). Although it is well documented that developmental traits under sexual selection can evolve
very rapidly, it remains unclear if the sex-specific brain neural circuitry also evolves rapidly.
Altering the neuronal patterning can drive species differences since even complicated
developmental pathways can be highly flexible and diverge over short evolutionary time scales
(Tanaka 2008). Some dimorphisms occur through specification of precursor cells, while others
through sex-specific morphogenesis.
While there are numerous examples of rapidly-evolving dimorphic morphological
characteristics, the evolutionary mechanisms that drive species differences in dimorphic brain
regions has not been studied extensively (Sanders 2008, Cachero 2010 and Manton 2014). Sexual
selection signatures may actually be more visible at the transcriptional level, rather than the coding
sequence, since dimorphism can bypass mutational constraints by altering gene expression (Mank
2017). How do GRNs diverge despite developmental robustness? Do sexually monomorphic brain
regions evolve at the same rate as sexually dimorphic regions?
86
5.2.6. Sexually Dimorphic versus Monomorphic NECs
Several Drosophila behaviors are modified by regulating the SDH to splice various transcripts
in a sex-specific manner. Altering the expression of SDH components can feminize males or
masculinize females. For example, inducing the expression of tra
F
in males leads to reduced
courtship behavior (Lee 2006). As previously discussed, there are several fru-NECs. We focus on
14 bilaterally symmetric and 1 non-bilaterally symmetric fru-NECs. Many of these fru-NECs
overlap with sexually-dimorphic regions that are divergent between D. melanogaster and D.
simulans (Fig 32).
There are 3 anterior NECs around the Antenna Lobe (AL), including fru-AL, which are
broadly scattered cells immediately above the AL, fru-mAL, which are medially located above the
AL, fru-mcAL, which are mechanosensory and below the AL near the esophageal foramen. We
predict all 3 NECs will be highly divergent since the antenna lobe is instrumental in many
conspecific mate detection and other courtship behaviors (von Philipsborn 2011). We expect mAL
to be highly divergent since it plays a pivotal role in the species-specific pulse song (Kimura 2005)
and males have 30 mAL neurons, as opposed to only 5 in females (Ito 2008).
87
There are 2 lateral NECs. These are fru-Lv which is lateral-ventral to fru-Ld, which is
lateral-dorsal. Another important distinction is that fru-Lv is only found in the anterior, while fru-
Ld spans both regions. Within the Lv, there are 5 unique clusters. Two of these clusters are
expressed only in males. These are D1, comprised of 5 neurons, and D2, comprised of 2 neurons.
On the other had, the L1, L2 and L3 clusters are shared by both sexes, but L1 differs between
sexes in cell number (Lee 2006). Thus, within the Lv, we expect the D1, D2 and L1 clusters to
differ between species, since they are regulated by fru and involved in male-specific behaviors.
Within the Anterior Superior Protocerebrum (aSP), there are 3 NECs, abbreviated as aSP1,
aSP2, and aSP3. These regions are involved in courtship initiation, mate-following, tapping, and
Figure 32: fru neuron expression clusters (NECs) overlap with dimorphic volumetric
regions.
Top: HCR results for fru (red) expression patterns in male D. melanogaster (left) and D.
simulans (right).
Bottom: Sexually dimorphic regions that vary in volume for D. melanogaster (purple, left)
and D. simulans (green, right) from Manton 2014.
88
wing extension, including the species-specific dance. Given the high concentration of fru-
expressing P1 dendrites in the anterior superior protocerebrum, we expect the aSP1, aSP2 and
aSP3 NECs to be divergent (Kimura 2008).
Within the Posterior Superior Protocerebrum (pSP), there are 2 NECs, abbreviated as pSP1
and pSP2, which are involved in the very early steps of courtship behavior. These steps do not
differ greatly between D. melanogaster and D. simulans and thus are less likely to vary between
species (Kimura 2008 & Kohatsu 2011). Furthermore, pSP2 is less likely to be similar since the
medial-posterior portions of the Superior Protocerebrum between species also have similar cell
numbers (Usui-Aoki 2005).
Other important posterior NECs include fru-P which is involved in licking and broadly
scattered posteriorly, as well as fru-pL which is in the posterior lateral region. We predict both of
these regions will be divergent since there is high sexual dimorphism in the Posterior Cells (pC1
and pC2) regions within brains (Lee 2009). We expect pL to vary greatly between species, due to
its large volumetric divergence (Manton 2014).
The fru-SOG NEC, found in the Subesophageal Ganglion, spans both anterior and posterior
portions of the brain. It is not bilaterally symmetric and very small, making it unlikely to be
divergent (Usui-Aoki 2005). The optic lobes contain 2 NECs. These are fru-Lo, which is found in
the Lobula posteriorly, and fru-M which is very large and spans the posterior and anterior medulla.
We predict these regions will not be divergent since the adult compound eye is sexually
monomorphic as previously discussed (Chapter 4, Gasper 2020).
89
Although the SDH pathway in Drosophila sister species is conserved, these fru- and dsx-
expressing cells, hereafter referred to as Neuron Expression Clusters (NECs), are divergent,
including patterning and volumetric variation in certain regions between sexes and species. This
suggests some signal that is upstream or independent of the SDH. Since fru is involved in male
courtship behaviors, such as singing, these sexually dimorphic NECs might explain changes in
sexual behavior between Drosophila species (Cachero 2010). While there are numerous examples
of rapidly-evolving dimorphic morphological characteristics, the evolutionary mechanisms that
drive species differences in dimorphic brain regions has not been studied extensively (Sanders
2008, Cachero 2010 and Manton 2014). How are dimorphic neuron clusters differentially
regulated between sexes and closely-related species? Do these NECs evolve at the same speed as
other sexual traits?
To investigate these questions, we used HCR on late pupa males to measure the quantitative
and spatial relationships of NECs in three D. melanogaster and three D. simulans brains. We
performed an ANOVA to understand how different NECs vary between species. These images
are currently being processed. Our preliminary results indicate that some of the regions that
vary between species correspond to brain regions that have volumetric differences between
sexes (Fig 32). We plan on comparing brain symmetry, as well as monomorphic and diverged
brain regions to better understand how these sex-specific brains differences alter male
courtship, among other traits, to drive speciation. We plan on using these same methods to
90
investigate genes that are highly expressed in opposite sexes between species to understand
how evolutionary mechanisms can alter the regulatory logic of GRNs.
5.3. Chapter 5 Methods
5.3.1. Fly stocks
Fly stocks: D. melanogaster line R303 is from the Drosophila Synthetic Population Resource
(DSPR). D. simulans were collected from the Zuma organic orchard in Zuma beach, CA in the
spring of 2012 (Signor 2017). They were inbred by 15 generations of full sib crosses. D.
melanogaster were collected in Raleigh, North Carolina and inbred for 20 generations (Mackay et
al., 2012). Thus, any differences detected between species, sexes, or genotypes is due to natural
variation.
5.3.2. Staging and dissection:
All flies were reared on a standard medium at 25°C with a 12hr light/12-hr dark cycle. Three
replicate brains were isolated from each species, for a total of 6 brains. Vials were used a single
time for collection to avoid pseudo-replication, such that every genotype and sex was collected
from a separate replicate population. Density of the cultures was controlled, as a standard number
of parents were given 24 hours to lay their eggs. Virgin male flies were collected at 11 days, just
before eclosion to make dissection and HCR easier since adult brains tend to float in liquid.
Dissections were performed to carefully remove the head tissue and proboscis without damaging
the brain. A pipette was used to transfer the brain into an Eppendorf tube for HCR. Replicates were
not conducted concurrently meaning that variation due to differences in timing will be randomly
distributed. Previous studies have found that the highest amount of temporal variation and
91
strongest expression of dsx and fru is during the mid-pupal stage in both sexes. This suggests that
during this key stage of metamorphosis, dsx and fru set up several elements of the sex-specific
neuroanatomy that influences courtship and mating in adults (Lee 2009). We will use HCR to
compare neuron patterning differences in NECs between NECs in late pupa males (P15 stage).
The late pupa stage is easy to identify and dissect, thus allowing a developmentally synchronized
comparison between genotypes at higher sample sizes. More importantly, the P15 stage shows no
significant differences compared to the mature male central nervous system in NECs (Lee 2000).
Therefore, imaging at this stage will show strong species differences and the developmentally
mature male NEC circuitry.
5.3.3. Hybridization chain reaction (HCR):
We optimized our published HCR eye disc protocol (Ali 2019) for Drosophila brains. This
high-throughput multiplexing microscopy method can assess colocalization patterns at cellular
resolution between specie, thereby accounting for the possibility that genes and morphological
levels can be conserved while still showing divergence in cell number, patterning, and
differentiation between species (Tanaka 2008).
5.3.4. Microscopy
For each species, we performed 3 replicates, imaging from posterior to anterior. Three
dimensional images of mounted, HCR-stained fly brains were acquired on a Zeiss LSM 780 laser
scanning microscope (Carl Zeiss MicroImaging, Inc., Thornwood, NY) with Objective Plan-
Apochromat 20×/1.40 water. Images were obtained in ZEN using lambda mode, which improves
signal detection by measuring the full emission spectrum from a single excitation wavelength. All
92
samples were imaged with nearly identical settings such as laser intensity and gain to minimize
pixel saturation and heterogeneity between brains to improve mRNA quantification. Furthermore,
signal intensity was normalized within samples via dividing by the brightest pixel to improve
quantification inferences.
5.3.5. HCR Brain Image Processing & Data Analysis
We tailored our HCR image analysis pipeline for the analysis of brain z-stacks (Ali 2019). To
verify our results, the preliminary data and analyses below will be further verified using Imaris
®
software when the Translational Imaging Center at the University of Southern California reopens.
Aside from our own image analysis, we will upload our images to the Virtual Fly Brain database
and compare our colocalization data with others in the field. The IS2 database in the Virtual Fly
Brain Project (http://www.virtualflybrain.org) has about 1000 images of various fru clones
(Manton 2014).
To investigate the role of the SDH in brain development, we stained fru and dsx in Sz139 (D.
simulans) and R338 (D. melanogaster). After applying previous image registration methods (see
Ali 2019 and Chapter 2 Methods), brains were segmented into anterior and posterior halves (x-
axis & z-axis), followed by lateral-medial segmentation using the z-axis and x-axis respectively.
The detected signal for a gene is then assigned into 1 of 29 unique Neuron Expression Clusters
(NECs). These clusters correspond to 14 pairs of symmetric, bilaterally-expressed NECs (aSP1,
aSP2, aSP3, Lv, AL, mAL. mcAL, P, pSP1, pSP2, pL, Ld & Lo), and 1 nonbilateral NEC (SOG).
Regions bordering 2 clusters were manually separated. For each detected point within each cluster,
voxel intensity, area and volume were extracted. Intensity was averaged for each NEC, while
volume and area were calculated using the 3D envelope of particles, as opposed to plain addition,
93
from an R package. Morphological reconstruction and contrast mapping segmentation was
performed using a version (http://sourceforge.net/p/ prostack/wiki/mrcomas) of the MrComas
method modified for 3D images to detect mRNA transcripts for each gene. We have published
previous work on these image analysis methods (Kozlov 2017, Ali 2019).
Briefly, this approach first enhances contrast within the image and reduces noise. After a
nearest-neighbor algorithm, morphological reconstruction, using opening and erosion (Vincent
1993), was used on all images to remove very dark or bright spots, and connect them (Kozlov 2017
& Ali 2019). Using contrast mapping to obtain a pixel-by-pixel difference of the reconstructed
images, we can produce a mask for each channel. After visually inspecting the quality of
segmentation between object borders and the original image, we return masks to their original size
and quantitative measures are made, such as volume, area, and intensity.
Figure 33: Data Extraction from Raw Images.
NECs for fru (large dots) in D. melanogaster (top) & D. simulans (bottom) using HCR were extracted. Scale
bars = 50 µm
94
In order to increase the contrast between objects represented by gene transcripts and
background planes of 3D stack were selected by visual inspection from images obtained with
different lambda parameter. All stacks were brought to the same orientation. Object detection was
performed using a modified MrComas method developed recently (Kozlov 2017, Ali 2019). In
brief, the algorithm enhances contrast within the image and reduces noise using median filtering,
subsequent to that the image is enlarged by a factor of four with the nearest-neighbor algorithm.
The image is further enhanced using histogram normalization and a cascade of operations of
morphological opening and closing by reconstruction in order to remove extraneous bright spots
and connect bright objects (Vincent 1993). The contrast mapping operator assigns each pixel the
maximum value between the pixel-by-pixel difference of the reconstructed images and their pixel-
by-pixel product and produces the rough mask of the image in which bright pixels correspond to
gene transcripts and dark – to background. The mask is further refined by the Crimmins speckle
removal algorithm (Crimmins 1985). Subsequently, distance transform substitutes each pixel value
with the number of pixels between it and the closest background pixel. Thus, the intensity of pixels
at the border between two erroneously merged objects is reduced and the objects are split by
watershed segmentation (Meyer 1994). This method treats the whole image as a surface and
intensity of each pixel as its height and determines the watershed lines along the tops of ridges that
separates the catchment basins. Finally, quantitative measures were made of shape and intensity
characteristics such as the number of pixels, the mean and standard deviation of intensity of pixels
95
in the detected object. The procedure was implemented in the framework of previously developed
tool ProStack (Kozlov 2008).
The result of the image processing procedure is stored as a table of detected objects that
represent gene transcripts. For each object its x, y and z coordinates are recorded together with
shape and intensity characteristics. Further, the data were centered in x direction and split into
anterior and posterior parts for each image. Subsequently, objects detected in each image were
assigned into 1 of 29 unique clusters. These clusters correspond to 13 bilaterally-expressed clusters
(aSP1, aSP2, aSP3, Lv, AL, mAL. mCAL, P, pSP1, pSP2, pL, Ld & Lo), and 1 nonbilateral cluster
(SOG). Due to imaging issues, fru-M NECs were discarded. We used CloudCompare software to
Figure 34: fruitless (fru) Neuron Expression Clusters (NECs) in Adult Brains
96
perform manual clustering. Spatial characteristics of clusters such as area and volume were
calculated using function convhulln from geometry package (Habbel 2019) for R statistical
software that builds a convex hull that contains the specified 3d points. After extracting the
intensity, volume and area, analysis was performed considering NECs as pairs (“Grouped
clusters”) and independently per hemisphere (“Ungrouped clusters”), using a “zone number.” For
example, aSP1-left hemisphere is “zone 1,” while aSP1-right hemisphere is “zone 15.” Consistent
with other studies, we found no differences between symmetrical NECs in intensity. We performed
ANOVA testing on grouped clusters to understand how quantitative differences vary in relation to
volume and species in our dataset. Conducting our test on ungrouped clusters did not significantly
change our findings.
97
5.4. Chapter 5 Results
5.4.1. NEC Comparisons
Using a previously developed pipeline for HCR image analysis (Ali 2019), we extracted
intensity, volume and area for all datapoints within each NEC for 3 D. melanogaster and 3 D.
simulans brains (Figure 35). NECs were similarly color-coded in all samples.
Figure 35: Species Comparison of fru NECs Using HCR.
Comparison of digitized NECs for fru (large dots) in 3 D. melanogaster (top) & 3 D. simulans (bottom). Colors
represent 14 bilateral fru NECs and 1 non-bilateral fru NEC
98
5.4.2. Comparison of NEC Relative Intensity by Hemisphere
Figure 36 shows volumetric (top) and intensity (36 bottom) comparisons of fru NECs in 29
different brain regions between Drosophila melanogaster (left & right hemisphere shown in red
& orange respectively) and D. simulans ((left & right hemisphere shown in cyan and dark blue
respectively). We detected no significant differences in signal intensity between hemispheres
within species. Interestingly, some regions varied in signal but not volume between species,
Figure 36: Hemispheric NEC Volume and Intensity Comparison
Top: Volumetric comparison of fru NECs in different brain regions between Drosophila
melanogaster (left & right hemisphere shown in red & orange respectively) and D. simulans
(left & right hemisphere shown in cyan & dark blue respectively).
Bottom: Intensity comparison of fru NECs in different brain regions between Drosophila
melanogaster (left & right hemisphere shown in red & orange respectively) and D. simulans
(left & right hemisphere shown in cyan & dark blue respectively).
Regions are ordered alphabetically: AL, aSP1, aSP2, aSP3, Ld, Lo, Lv, mAL. mcAL, P, pL,
pSP1, pSP2, & SG.
99
suggesting that NECs can increase overall expression levels by upregulating transcription while
maintaining cluster size or vice versa. We found significant variation in several reproductively-
related NECs.
100
5.4.3. Grouped Volumetric and Intensity Comparison
NECs were grouped according to 14 bilaterally symmetric regions. Some studies show that
there is little differences between hemispheres. Grouping NECs also buffers our analysis
Figure 37: Grouped Volumetric and Intensity Comparison
Top: Bilateral volumetric comparison of fru NECs in different brain regions between
Drosophila melanogaster (red) and D. simulans ((cyan).
Bottom: Bilateral fru intensity comparison of fru NECs in different brain regions between
Drosophila melanogaster (red) and D. simulans ((cyan). Regions are ordered alphabetically:
AL, aSP1, aSP2, aSP3, Ld, Lo, Lv, mAL. mcAL, P, pL, pSP1, pSP2, & SG.
101
against outliers and potential discrepancies due to molecular methods, like probe tissue
penetration, or imaging issues, such as fluorophore photo-bleaching. Additional hemispheric
analysis of all samples showed similar overall trends for NECs. We found significant variation
in several reproductively-related NECs.
5.4.4. ANOVA Analysis
We performed an ANOVA to investigate volumetric and quantitative differences between
various brain regions. As predicted, we found significant species differences in reproductively-
relevant regions that influence courtship behavior, such as mAL, AL, mcAL, aSP1, P, & aSP3,
compared to sexually monomorphic regions, such as pSP2, Lo and SOG.
102
5.5. Discussion
The species differences in development from our Chapter 2 results drove us to ask the question,
how quickly can complex traits evolve between species despite stochasticity and developmental
robustness? Several studies have shown that dimorphic traits tend to be under stronger selection
and evolve faster than monomorphic traits, especially in males. Furthermore, male-specific genes
Table 3: ANOVA Results for NECs. Colors indicate NECs that are related.
NEC Volume Species Volume x Species
AL 0.0422195 * 0.0002229 *** 0.1531248
mAL 6.245e-06 *** 3.070e-05 *** 0.001219 **
mcAL 0.215 0.0174 * 0.2018
aSP1 0.049964 * 0.004339 ** 0.867972
aSP2 0.58025 0.78049 0.03479 *
aSP3 0.734954 0.004332 ** 0.896774
Lv 0.17073 8.634e-05 *** 0.02704 *
Ld 0.5011844 0.0005263 *** 0.9533302
P 0.008693 ** 0.010094 * 0.368451
pL 0.3488691 0.0007176 *** 0.3576024
pSP1 0.03483 * 0.94675 0.8488
pSP2 0.18913 0.12177 0.07659
Lo 0.7128 0.9751 0.7293
SOG 0.5003 0.1364 0.7
103
show more variation in expression levels within and between species compared to female-specific
and monomorphic genes (Meiklejohn 2003). For example, the size and orientation of sex combs
in males of several Drosophila sister species is highly variable, despite recent divergence (Tanaka
2008). We postulated that dimorphic brains regions evolve faster than monomorphic regions. We
used HCR to image fru NECs in 11 day old pupa. We found significant differences between D.
melanogaster and D. simulans in certain brain regions.
Interestingly, some regions varied in signal but not volume between species, suggesting that
NECs can increase overall expression levels by upregulating transcription while maintaining
cluster size or vice versa. Sex-specific tissue development and neuron patterning requires
communication between multiple biological levels. How can the patterning of dimorphic neuron
clusters change between species while conserving the SDH genes? Using HCR, we found
extensive variation in fruitless (fru) mRNA expression levels and volume between D.
melanogaster and D. simulans. These fru-Neuron Expression Clusters (NECs) were primarily
divergent in sexually-dimorphic NECs relevant to species-specific mate detection, such as the
Antenna Lobe (AL), and species-specific courtship.
We predicted high divergence in AL NECs due to their role in species-specific mate detection,
singing and courtship We found extremely significant variation in species, volume and species X
volume interactions for AL NECs. For the aSP NECs, we found a strong quantitative difference
between species and some significant volume and volume X species variation. For the AL NECs,
we found highly significant quantitative differences and significant volumetric differences
between species. We found significant divergence in Lv, Ld, P and pL as well. We found relatively
small volumetric differences in pSP1 and no significant variation for pSP2, Lo and SOG. Our
104
results suggest there is significant variation in sexually dimorphic NECs between D. melanogaster
and D. simulans.
There are several interesting future directions for this project. Despite the conservation of fru
NECs are divergent. This suggests some upstream regulator of the SDH dictates which neurons
express sex-specific genes in different species. We observed some NECs in D. simulans with more
volume but a lower amounts of fru. This might suggest a target for the upstream regulator(s). This
might also be affected by Delta/Notch lateral inhibition which can control cluster size and diverge
with time between species. Are these species differences in NECs due to the evolution of fru
regulatory sequences (cis) or upstream genes (trans)? Which upstream signals regulate which
NECs? While sex-specific alternative splicing in these sex-biased genes can resolve some of the
sexual conflict, it cannot account for the dimorphic regions that are divergent between species,
since they all express these isoforms. Furthermore, numerous studies have indicated that sex-
specific splicing cannot be the only mechanism leading to sex-specific fru neuron clustering (Lee
2000 and Samson 2014). We plan on measuring fru-NEC expression in transgenic Drosophila
lines. We can also measure distances and angles between various NECs to assess the possibility
of 3-dimensional differences in D. melanogaster and D. simulans brains within our data.
105
6. Concluding Remarks
Biological development requires the careful coordination of several processes, such as
transcription, replication, differentiation and patterning. These intrinsically stochastic processes
require developmental programs to be robust. This robustness is maintained through an intricate
web of interacting Gene Regulatory Networks (GRNs) and clean-up mechanisms. As species
diverge, stochasticity in GRNs can lead to developmental systems drift, whereby genetic variation
accumulates between populations without any observable phenotypic change. How can biological
systems use variation in gene expression, or transcriptional stochasticity, to balance developmental
robustness versus evolutionary adaptability depending on developmental and environmental
contexts? The Drosophila compound eye is an excellent system to study these interlinked topics
since individual eye units, or ommatidia, specified in the larval eye disc eventually form the highly-
organized, offset patterning seen in adult eyes.
We measured the quantitative and spatial relationship of 4 key genes involved in eye
development simultaneously, using the Hybridization Chain Reaction (HCR), in 3 natural
populations of D. melanogaster and D. simulans males and females respectively. Our novel
published findings show significant transcriptional stochasticity between species, genotypes and
sexes, as well as regulatory logic divergence despite the phenotypic similarity of Drosophila
melanogaster and D. simulans eyes (Ali 2019, Chapter 2).
Afterwards, we created a mathematical model to understand the observed phenotypic
robustness in eye GRNs. Our published results quantitatively assessed this threshold response,
whereby transcriptional stochasticity is buffered up to a certain threshold, after which the system
rapidly degrades (Courcoubetis 2019, Chapter 3).
106
We further investigated the eye GRN in different sexual contexts by genetically manipulating
the Sex Determination Hierarchy (SDH) and performing HCR on several genes in male, female
and pseudomale (masculinized females) larval eye discs, which showed no differences. (Chapter
4). To understand how GRNs evolve between closely related species, we used HCR to measure
the quantitative and spatial relationships of fruitless (fru) Neuron Expression Clusters (NECs) in
the SDH between D. melanogaster and D. simulans brains, which showed transcriptional and
volumetric variation in several reproductively-relevant brain regions (Chapter 5).
Collectively, our findings improve our understanding of the genotype-to-phenotype map,
which is instrumental in elucidating the genetic basis of complex diseases.
107
References
Albert, F., & Kruglyak, L. (2015). The role of regulatory variation in complex traits and
disease. Nature Reviews Genetics, 16, 197–212
Ali, S., Signor, S. A., Kozlov, K., Nuzhdin, S. V. (2019). Novel approach to quantitative
spatial gene expression uncovers genetic stochasticity in the developing Drosophila eye.
Evolution & Development.1–15. https://doi.org/10.1111/ede.12283
Arbeitman, M. N., Fleming, A. A., Siegal, M. L., Null, B. N., & Baker, B.S. (2007). A
genomic analysis of Drosophila somatic sexual differentiation and its regulation. Development.
131: 2007-2021
Arsenault, S. V., Hunt, B. G., & Rehan, S. M. (2018). The effect of maternal care on gene
expression and DNA methylation in a subsocial bee. Nature Communications, 9, 3468
Artavanis-Tsakonas S., Rand, M.D., & Lake, R.J. (1999). Notch signaling: cell fate control
and signal integration in development. Science. 284 (5415):770-776.
doi:10.1126/science.284.5415.770
Atkins, M., Jiang, Y., Sansores-Garcia, L., Jusiak, B., Halder, G., & Mardon, G. (2013).
Dynamic rewiring of the Drosophila retinal determination network switches its function from
selector to differentiation. PLoS Genetics, 9, e1003731
Averbukh, I., Gavish, A., Shilo, B. Z., & Barkai, N. (2017). Dealing with noise: The
challenge of buffering biological variability. Current Opinion in Systems Biology, 1:69–74.
https://doi.org/10.1016/j.coisb.2016.12.011
Baker, N. E., Yu, S., & Han, D. (1996). Evolution of proneural atonal expression during
distinct regulatory phases in the developing Drosophila eye. Current Biology, 6, 1290–1301
Baker, N. E., & Yu, S. Y. (2001). The EGF receptor defines domains of cell cycle
progression and survival to regulate cell number in the developing Drosophila eye. Cell, 104,
699–708
Baker, N. E., & Yu, S. Y. (1997). Proneural function of neurogenic genes in the developing
Drosophila eye. Current Biology, 7, 122–132
Baonza, A., & Freeman, M. (2001). Notch signaling and the initiation of neural development
in the Drosophila eye. Development, 128, 3889–3898
Baonza, A., & Freeman, M. (2005). Control of cell proliferation in the Drosophila eye by
Notch signaling. Developmental Cell, 8, 529–539
108
Barkai N. & Shilo B. Z. (2009). Robust generation and decoding of morphogen gradients.
Cold Spring Harbor perspectives in biology. 1(5):a001990. PMID: 20066119
Becker H., Renner S., Technau G.M., & Berger C. (2016). Cell-autonomous and non-cell-
autonomous function of Hox genes specify segmental neuroblast identity in the gnathal region of
the embryonic CNS in Drosophila. PLoS Genetics. 12(3): 1005961
Bhattacharya, A., & Baker, N. E. (2012). The role of the bHLH protein hairy in
morphogenetic furrow progression in the developing Drosophila eye. PLoS ONE, 7, e47503
Bhattacharya, A., & Baker, N. E. (2009). The HLH protein Extramacrochaetae is required for
R7 cell and cone cell fates in the Drosophila eye. Developmental Biology, 327:288–300
Billeter, Jean-Christophe & Rideout, Elizabeth & Dornan, Anthony & Goodwin, Stephen.
(2006). Control of Male Sexual Behavior in Drosophila by the Sex Determination Pathway.
Current biology : CB. 16. R766-76. 10.1016/j.cub.2006.08.025
Bornstein, B., Zahavi, E. E., Gelley, S., Zoosman, M., Yaniv, S. P., Fuchs, O., Porat, Z.,
Perlson, E., & Schuldiner, O. (2015). Developmental Axon Pruning Requires Destabilization of
Cell Adhesion by JNK Signaling. Neuron 88(5): 926-940
Bothma, J. P., Garcia, H. G., Esposito, E., Schlissel, G., Gregor, T., & Levine, M. (2014).
Dynamic regulation of eve stripe 2 expression reveals transcriptional bursts in living Drosophila
embryos. Proceedings of the National Academy of Sciences of the United States of America, 111,
10598–10603
Brennan, C. A., Ashburner, M., & Moses, K. (1998). Ecdysone pathway is required for furrow
progression in the developing Drosophila eye. Development, 125, 2653–2664.
Brown, K. M., Costanzo, M. S., Xu, W., Roy, S., Lozovsky, E. R., & Hartl, D. L. (2010).
Compensatory mutations restore fitness during the evolution of Dihydrofolate reductase.
Molecular Biology and Evolution, 27, 2682–2690
Brown, N. L., Sattler, C. A., Markey, D. R., & Carroll, S. B. (1991). Hairy gene function in
the Drosophila eye: Normal expression is dispensable but ectopic expression alters cell fates.
Development, 113, 1245–1256
Brown, N. L., Sattler, C. A., Paddock, S. W., & Carroll, S. B. (1995). Hairy and emc
negatively regulate morphogenetic furrow progression in the Drosophila eye. Cell, 80, 879–887
Burch, C. L., & Chao, L. (1999). Evolution by small steps and rugged landscapes in the RNA
virus phi6. Genetics, 151, 921–927
Cachero S, Ostrovsky AD, Jai YY, & Dickson BJ. (2010). Sexual dimorphism in the fly brain.
Current Biology. 20(18-4): 1589–1601
109
Cagan R.L. & Ready, D.F. (1989). Notch is required for successive cell decisions in the
developing Drosophila retina. Genes & development. 3(8):1099–1112.
https://doi.org/10.1101/gad.3.8.1099
Cande, J., Stern D. L., Morita T., Prud’homme, B., and Gompel, N. (2014) Looking Under
The Lamp Post: Neither Fruitless Nor Doublesex Have Evolved to Generate Divergent Male
Courtship in Drosophila. Cell Reports 8(2): 363–370
Casci, T. (2005). Robust arguments about canalization. Nature Reviews, 6, 89
Castellanos MC, Tang JCY, Allan DW. Female-biased dimorphism underlies a female-
specific role for post-embryonic Ilp7 neurons in Drosophila fertility. Development.
2013;140(18):3915–26. PMCID: PMC3915572
Chan, Y. F., Marks, M. E., Jones, F. C., Villarreal, G., Shapiro, M. D., Brady, S. D., Kingsley,
D. M. (2010). Adaptive evolution of pelvic reduction in sticklebacks by recurrent deletion of a
Pitx1 enhancer. Science, 327, 302–305
Chang P.L., Dunham J.P., Nuzhdin S.V. & Arbeitman, M.N. (2011). Somatic sex-specific
transcriptome differences in Drosophila revealed by whole transcriptome sequencing. BMC
Genomics. 12(1): 364
Charusanti, P., Conrad, T. M., Knight, E. M., Venkataraman, K., Fong, N. L., Xie, B.,
Palsson, B. (2010). Genetic basis of growth adaptation of Escherichia coli after deletion of pgi, a
major metabolic gene. PLoS Genetics, 6, e1001186
Chen, J., Nolte, V., & Schlotterer, C. (2015). Temperature stress mediates decanalization and
dominance of gene expression in Drosophila melanogaster. PLoS Genetics, 11, e1004883
Christiansen, A.E., Keisman, E.L., Ahmad, S.M., & Baker, B.S. (2002). Sex Comes in From
the Cold: The Integration of Sex and Pattern. Trends in Genetics. 18(10): 510–516
Choi, H. M., Beck, V. A., & Pierce, N. A. (2014). Next-Generation in Situ Hybridization
Chain Reaction: Higher Gain, Lower Cost, Greater Durability. ACS Nano. 8 (5): 4284–4294.
DOI: 10.1021/nn405717p
Chowdhury, D. & Stauffer, D. (2005). Statistical Mechanics of Interacting Systems: Theory of
Phase Transitions. Principles of Equilibrium Statistical Mechanics. p. 253–255.
https://doi.org/10.1002/3527603158
Chubb J. R., Trcek, T., Shenoy, S. M., & Singer, R. H. (2006). Transcriptional pulsing of a
developmental gene. Current Biology, 16(10):1018–1025.
https://doi.org/10.1016/j.cub.2006.03.092 PMID: 16713960
Chun, S., & Fay, J. C. (2011). Evidence for hitchhiking of deleterious mutations within the
human genome. PLoS Genetics, 7, e1002240. Cook, T., Pichaud, F., Sonneville, R., Papatsenko,
110
D., & Desplan, C. (2003). Distinction between color photoreceptor cell fates is controlled by
Prospero in Drosophila. Developmental Cell, 4, 853–864
Cline TW. (1984). Autoregulatory functioning of a Drosophila gene product that establishes
and maintains the sexually determined state. Genetics. 107:231-77
CloudCompare: 3D point cloud and mesh processing software. Open Source Project.
https://www.cloudcompare.org/
Connallon, T. & Clark, A. G. (2011). Association between sex-biased gene expression and
mutations with sex-specific phenotypic consequences in Drosophila. Genome Biology and
Evolution. 3, 151–155
Corty, M.M., Tam, J. and Grueber, W.B. (2016). Dendritic diversification through
transcription factor-mediated suppression of alternative morphologies. Development 143: 1351-
1362
Cowley, D. E. & Atchley, W. R. (1988). Quantitative Genetics of Drosophila melanogaster.
Heritabilities and Genetic Correlations between Sexes for Head and Thorax Traits. Genetics.
119: 421-433
Courcoubetis G, Ali S, Nuzhdin SV, Marjoram P, Haas S (2019) Threshold response to
stochasticity in morphogenesis. PLoS ONE 14(1): e0210088. https://doi.org/10.1371/journal.
pone.0210088
Crimmins, T. R. (1985). Geometric Filter for Speckle Reduction. Applied Optics 24 (10):
1438. https://doi.org/10.1364/AO.24.001438
Crombach, A., Wotton, K. R., Jimenez-Guri, E., & Jaeger, J. (2016). Gap gene regulatory
dynamics evolve along a genotype network. Molecular Biology and Evolution, 33, 1293–1307
Debye P. Interferenz von ro¨ntgenstrahlen und wa¨rmebewegung. Annalen der Physik. 1913;
348(1):49–92. https://doi.org/10.1002/andp.19133480105
De Graeve, F., Jagla, T., Daponte, J. P., Rickert, C., Dastugue, B., Urban, J., Jagla, K. (2004).
The ladybird homeobox genes are essential for the specification of a subpopulation of neural
cells. Developmental Biology. 270(1): 122-134
Dey, B. K., Zhao, X. L., Popo-Ola, E., Campos, A.R. (2009). Mutual regulation of the
Drosophila disconnected (disco) and Distal-less (Dll) genes contributes to proximal-distal
patterning of antenna and leg. Cell Tissue Res. 338(2): 227-240
Dokucu, M. E., Zipursky, S. L., & Cagan, R. L. (1996). Atonal, rough and the resolution of
proneural clusters in the developing Drosophila retina. Development, 122, 4139–4147
111
Dominguez M. & Casares, F. (2005). Organ specification-growth control connection: new in-
sights from the Drosophila eye-antennal disc. Developmental Dynamics; 232:673–684
Egami T. & Billinge S.J. (2012). Underneath the Bragg peaks: structural analysis of complex
materials. vol. 16. Newnes
Escudero, L. M., & Freeman, M. (2007). Mechanism of G1 arrest in the Drosophila eye
imaginal disc. BMC Developmental Biology, 7, 13
Estes, S., & Lynch, M. (2003). Rapid fitness recovery in mutationally degraded lines of
Caenorhabditis elegans. Evolution, 57, 1022–1030
Estes, S., Phillips, P. C., & Denver, D. R. (2011). Fitness recovery and compensatory
evolution in natural mutant lines of C. elegans. Evolution, 65, 2335–2344
Falconer, D.S. & Mackay, T. F. C. (1996). Introduction to Quantitative Genetics. London:
Pearson
Fear, J. M., Arbeitman, M. N., Salomon, M. P., Dalton, J. E., Tower, J., Nuzhdin, S. V., &
McIntyre, L. M. (2015). The wright stuff: reimagining path analysis reveals novel components of
the sex determination hierarchy in Drosophila melanogaster. BMC Systems Biology. 9:53
Felix, M. A. & Barkoulas, M. (2015). Pervasive robustness in biological systems. Nature
Reviews Genetics, 16, 483–496
Felsenfeld, A. L., & Kennison, J. A. (1995). Positional signaling by hedgehog in Drosophila
imaginal disc development. Development, 121, 1–10
Firth, L. C., & Baker, N. E. (2005). Extracellular signals responsible for spatially regulated
proliferation in the differentiating Drosophila eye. Developmental Cell, 8, 541–551
Firth, L. C., Bhattacharya, A., & Baker, N. E. (2010). Cell cycle arrest by a gradient of Dpp
signaling during Drosophila eye development. BMC Developmental Biology, 10, 28
Fowlkes, C. C., Eckenrode, K. B., Bragdon, M. D., Meyer, M., Wunderlich, Z., Simirenko, L.,
DePace, A. H. (2011). A Conserved developmental patterning network produces quantitatively
different output in multiple species of Drosophila. PLoS Genetics, 7, e1002346
Frankel N., Davis, G.K., Vargas, D., Wang S., Payre, F., & Stern, D. L. (2010). Phenotypic
robustness conferred by apparently redundant transcriptional enhancers. Nature
(London).466(7305):490-493. doi:10.1038/nature09158
Fu, W., & Baker, N. E. (2003). Deciphering synergistic and redundant roles of Hedgehog,
Decapentaplegic and Delta that drive the wave of differentiation in Drosophila eye development.
Development, 130, 5229–5239
112
Fujioka, M., Lear, B.C., Landgraf, M., Yusibova, G.L., Zhou, J., Riley, K.M., Patel, N.H.,
Jaynes, J.B. (2003). Even-skipped, acting as a repressor, regulates axonal projections in
Drosophila. Development 130(22): 5385-5400
Fukaya, T., Lim, B., & Levine, M. (2016). Enhancer control of transcriptional bursting. Cell,
166, 358–368
Gailey DA, Taylor BJ, Hall JC. (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: 879–890
Garfield, D. A., Runcie, D. E., Babbitt, C. C., Haygood, R., Nielsen, W. J., & Wray, G. A.
(2013). The impact of gene expression variation on the robustness and evolvability of a
developmental gene regulatory network. PLoS Genetics, 11, e1001696
Gasper, P., Arif, S., Sumner-Rooney, L., Kittelmann, M., Bodey, A.J., Stern, D.L. Nunes,
M.D.S., & McGregor, A.P. (2020). Characterization of the Genetic Architecture Underlying Eye
Size Variation Within Drosophila melanogaster and Drosophila simulans. G3: Genes, Genomes,
Genetics. 10, 3; 1005-1018. doi.org/10.1534/g3.119.400877
Gavish A & Barkai N. (2016). A two-step patterning process increases the robustness of
periodic patterning in the fly eye. Journal of biological physics. 42(3):317–338.
https://doi.org/10.1007/s10867-0169409-4
Gavish, A., Shwartz, A., Weizman, A., Schejter, E., Shilo, B. Z., & Barkai, N. (2016).
Periodic patterning of the Drosophila eye is stabilized by the diffusible activator Scabrous.
Nature Communications, 7, 1–10. https://doi.org/10.1038/ncomms10461
Génétique, C. & M., Sélection, (1983). Interspecific competition between Drosophila
melanogaster and Drosophila simulans: temperature effect on competitive ability and fitness
components. Genetics Selection Evolution, 15, 376–378
Gibson G, Dworkin I. Uncovering cryptic genetic variation. Nature Reviews Genetics. 2004;
5(9):681. https://doi.org/10.1038/nrg1426
Gibson, G., & Reed, L. K. (2008). Cryptic genetic variation. Current Biology, 18, R989–R990
Gibson, G. (2009). Decanalization and the origin of complex disease. Nature Reviews
Genetics, 10, 134–140
113
Goering, L. M., Hunt, P. K., Heighington, C., Busick, C., Pennings, P., Hermisson, J., Gibson,
G. (2009). Association of orthodenticle with natural variation for early embryonic patterning in
Drosophila melanogaster. Journal of Experimental Zoology, 312B, 841–854
Goodwin, S. F. (1999). Molecular neurogenetics of sexual differentiation and behaviour.
Current Opinions Neurobiology. 9:759–765
Goodwin, S.F., Taylor, B. J., Villella, A., Foss, M., Ryner, L. C., Baker, B. S., & Hall, J.C.
(2000). Aberrant splicing and altered spatial expression patterns in fruitless mutants of
Drosophila melanogaster. Genetics. 154:725–745
Gompel, N., Prud'homme, B., Wittkopp, P. J., Kassner, V. A., & Carroll, S. B. (2005). Chance
caught on the wing: cis-regulatory evolution and the origin of pigment patterns in Drosophila.
Nature, 433, 481–487
González‐Blas B. C, Quan X., & Duran‐Romaña, R. (2020). Identification of genomic
enhancers through spatial integration of single‐cell transcriptomics and epigenomics. Molecular
Systems Biology.16(5): e9438-n/a. doi:10.15252/msb.20209438
Gray G.E., Mann, R.S., & Mitsiadis, E. (1999). Human ligands of the Notch receptor. Am J
Pathol. 154(3):785-794. doi:10.1016/S0002-9440(10)65325-4
Green, R. M., Fish, J. L., Young, N. M., Smith, F. J., Roberts, B., Dolan, K., Hallgrímsson, B.
(2017). Developmental nonlinearity drives phenotypic robustness. Nature Communications, 8,
1970
Greenwood, S., & Struhl, G. (1999). Progression of the morphogenetic furrow in the
Drosophila eye: The roles of hedgehog, Decapentaplegic and the Raf pathway. Development,
126, 5795–5808
Griffin, R. M., Dean, R., Grace, J. L., Ryden, P. & Friberg, U. (2013). The shared genome is a
pervasive constraint on the evolution of sex-biased gene expression. Molecular Biology and
Evolution. 30, 2168–2176
Gursky, V. V., Surkova, S. Y., & Samsonova M. G. (2012). Mechanisms of developmental
robustness. Biosystems 109(3):329–335. https://doi.org/10.1016/j.biosystems.2012.05.013
PMID: 22687821
Habel, K., Grasman, R., Gramacy, R. B., Mozharovskyi, P., & Sterratt, D. C. (2019).
Geometry:
114
Mesh Generation and Surface Tessellation. R package version 0.4.5. https://CRAN.R-
project.org/package=geometry
Halperin B. & Nelson D. R. (1978). Theory of two-dimensional melting. Physical Review
Letters. 41(2):121. https://doi.org/10.1103/PhysRevLett.41.121
Haynie JL, Bryant PJ. Development of the eye-antenna imaginal disc and morphogenesis
of the adult head in Drosophila melanogaster. Journal of Experimental Zoology Part A:
Ecological Genetics and Physiology. 1986; 237(3):293–308.
https://doi.org/10.1002/jez.1402370302
Heberlein, U., Wolff, T., & Rubin, G. M. (1993). The TGF[beta] homolog dpp and the
segment polarity gene hedgehog are required for propagation of a morphogenetic wave in the
Drosophila retina. Cell, 75, 913–926
Heranz, H., & Cohen, S. M. (2010). MicroRNAs and gene regulatory networks:
Managing the impact of noise in biological systems. Genes & Development, 24, 1339–1344
Hermisson, J., & Wagner, G. (2004). The population genetic theory of hidden variation
and genetic robustness. Genetics, 168, 2271–2284
Hilbrant, M., Almudi, I., Leite, D. J., Kuncheria, L., Posnien, N., Nunes, M., &
McGregor, A. P. (2014). Sexual dimorphism and natural variation within and among species in
the Drosophila retinal mosaic. BMC Evolutionary Biology, 14, 240
Hirth, F., Kammermeier, L., Frei, E., Walldorf, U., Noll, M., & Reichert, H. (2003). An
urbilaterian origin of the tripartite brain: developmental genetic insights from Drosophila.
Development. 130(11): 2365-2373
Hitier, R., Chaminade, M., Preat, T. (2001). The Drosophila castor gene is involved in
postembryonic brain development. Mech. Dev. 103(1-2): 3-11
Hoekstra, H. E., & Coyne, J. A. (2007). The locus of evolution: Evo devo and the
genetics of adaptation. Evolution, 61, 995–1016
Horabin J. I., Walthall S., Vied C. & Moses M. (2003). A positive role for Patched in
Hedgehog signaling revealed by the intracellular trafficking of Sex-lethal, the Drosophila sex
determination master switch. Development. 130:6101-9
Horabin, J. I. (2005). Splitting The Hedgehog signal: sex patterning in Drosophila.
Development 132(21), 4801-4810
Ito, H., Fujitani, K., Usui, K., Shimizu-Nishikawa, K., Tanaka, S., & Yamamoto, D.
(1996). Sexual orientation in Drosophila is altered by the satori mutation in the sex-
determination gene fruitless that encodes a zinc finger protein with a BTB domain. Proceedings
of the National Academy of Science USA. 93:9687–9692
115
Ito, K., and Awasaki, T. (2008). Clonal Unit Architecture of the Adult Fly Brain.
Advanced Experimental Medical Biology. 628, 137–158
Ito, Y., Harigai, A., Nakata, M., Hosoya, T., Araya, K., Oba, Y., Niimi, T. (2013). The
role of doublesex in the evolution of exaggerated horns in the Japanese rhinoceros beetle. Nature
Neuroscience, 14, 561–567
Iyer, E.P., Iyer, S.C., Sullivan, L., Wang, D., Meduri, R., Graybeal, L.L., Cox, D.N.
(2013). Functional genomic analyses of two morphologically distinct classes of Drosophila
sensory neurons: post-mitotic roles of transcription factors in dendritic patterning. PLoS ONE.
8(8): e72434
Janzer B & Steinmann-Zwicky M. (2001). Cell-autonomous and somatic signals control
sex-specific gene expression in XY germ cells of Drosophila. Mechanisms of Development.
100:3-13
Jarman, A. P., Grell, E. H., Ackerman, L., Jan, L. Y., & Jan, Y. N. (1994). Atonal is the
proneural gene for Drosophila photoreceptors. Nature, 369, 398–400
Jeong, S., Rebeiz, M., Andolfatto, P., Werner, T., True, J., & Carroll, S. B. (2008). The
evolution of gene regulation underlies a morphological difference between two Drosophila sister
species. Cell, 132, 783–793
Jiao, Y., Lau T., Hatzikirou, H., Meyer-Hermann, M., Corbo, J.C., & Torquato, S.
(2014). Avian photoreceptor patterns represent a disordered hyperuniform solution to a
multiscale packing problem. Physical Review E. 89(2):022721.
https://doi.org/10.1103/PhysRevE.89.022721
Johnston, R. J., & Desplan, C. (2014). Interchromosomal communication coordinates
intrinsically stochastic expression between alleles. Science, 343, 661–665
Johnston, R. J., Jr, Otake, Y., Sood, P., Vogt, N., Behnia, R., Vasiliauskas, D., &
Desplan, C. (2011). Interlocked feedforward loops control cell-type-specific rhodopsin
expression in the Drosophila eye. Cell, 145, 956–968
Kaminker, J.S., Canon, J., Salecker, I., Banerjee, U. (2002). Control of photoreceptor
axon target choice by transcriptional repression of Runt. Nat. Neurosci. 5(8): 746-750
Keller EF. Developmental robustness. Annals of the New York Academy of Sciences.
2002;981(1):189–201. https://doi.org/10.1111/j.1749-6632.2002.tb04918.x PMID: 12547680
Keren, L., van Dijk, D., Weingarten-Gabbay, S., Davidi, D., Jona, G., Weinberger, A.,
Segal, E. (2015). Noise in gene expression is coupled to growth rate. Genome Research, 25,
1893–1902
116
Kim, S., Cassidy, J.J., Yang, B., Carthew R.W., & Hilgenfeldt, S. (2016). Hexagonal
patterning of the insect compound eye: Facet area variation, defects, and disorder. Biophysical
journal.111(12):2735–2746. https:// doi.org/10.1016/j.bpj.2016.11.004 PMID: 28002749
Kimura, K., Hachiya, T., Koganezawa, M., Tazawa, T. & Yamamoto, D. (2008).
Fruitless and Doublesex Coordinate to Generate Male-specific Neurons That Can Initiate
Courtship. Neuron. 59: 759–769
Kittel C. (2004). Introduction to Solid State Physics. Wiley; Available from:
https://books.google.com/ books?id=kym4QgAACAAJ
Kohatsu, S., Koganezawa, M. & Yamamoto, D. (2011). Female Contact Activates Male-
specific Interneurons That Trigger Stereotypic Courtship Behavior in Drosophila. Neuron. 69,
498–508
Kopp A, Duncan I, Godt, D, & Carroll S. B.. Genetic control and evolution of sexually
dimorphic characters in Drosophila. Nature. 2000; 408(6812):553–9
Kozlov, K., Kosheverova, V., Kamentseva, R., Kharchenko, M., Sokolkova, A.,
Kornilova, E., & Samsonova, M. (2017). Quantitative analysis of the heterogeneous population
of endocytic vesicles. Journal of Bioinformatics and Computational Biology, 10, 1750008
Kozlov, K., A., Pisarev, J. Kaandorp, & Samsonova, M. (2008). ProStack, an image
Analysis Software for Identification and Quantification of Objects Visualized with Microscope.
В Absract book of The 9th Int. Conf. on Systems Biology, 191. Goteborg
Kumar, J. P. (2012). Building an Ommatidium One Cell at a Time. Developmental
Dynamics: An Official Publication of the American Association of Anatomists. 241(1): 136–149.
http://doi.org/10.1002/dvdy.23707
Kumar J. P. (2013). Catching the next wave: Patterning of the Drosophila eye by the
morphogenetic furrow, in: Molecular genetics of axial patterning, growth and disease in the
Drosophila eye. In Molecular genetics of axial patterning (pp. 75–97). New York, NY: Springer
New York
Kuntz, S. G., & Eisen, M. B. (2014). Drosophila embryogenesis scales uniformly across
temperature in developmentally diverse species. PLoS Genetics, 10, e1004293
Kurusu, M., Awasaki, T., Masuda-Nakagawa, L.M., Kawauchi, H., Ito, K., & Furukubo-
Tokunaga, K. (2002). Embryonic and larval development of the Drosophila mushroom bodies:
concentric layer subdivisions and the role of fasciclin II. Development. 129(2): 409-419
Land M.F. & Eckert H. (1985). Maps of the acute zones of fly eyes. Journal of
Comparative Physiology. 5;156:525–538. doi: 10.1007/BF00613976
117
Landry, C. R., Wittkopp, P. J., Taubes, C. H., Ranz, J. M., Clark, A. G., & Hartl, D. L.
(2005). Compensatory cis-trans evolution and the dysregulation of gene expression in
interspecific hybrids of Drosophila. Genetics, 171, 1813–1822
Lawrence P. A. & Green S. M. (1979). Cell lineage in the developing retina of
Drosophila. Developmental Biology. 71(1):142–152. https://doi.org/10.1016/0012-
1606(79)90088-5 PMID: 467785
Lawrence, P. A., & Johnston, P. (1986). The muscle pattern of a segment of Drosophila
may be determined by neurons and not by contributing myoblasts. Cell. 45:505–513
Lee, G. Foss, M., Goodwin, S.F., Carlo, T., Taylor, B.J., & Hall, J.C. (2000). Spatial,
Temporal, and Sexually Dimorphic Expression Patterns of the fruitless gene in the Drosophila
Central Nervous System. Journal of Neurobiology. 15: 404–426
Lee, G., Bahn, J. H., & Park, J. H. (2006). Sex- and clock-controlled expression of the
neuropeptide F gene in Drosophila. PNAS. 103 (33) 12580-12585.
https://doi.org/10.1073/pnas.0601171103
Li, W., Ohlmeyer, J. T., Lane, M. E., & Kalderon, D. (1995). Function of protein kinase A in
hedgehog signal transduction and Drosophila imaginal disc development. Cell, 80, 553–562
Li, Y., & Baker, N. E. (2001). Proneural enhancement by Notch overcomes Suppressor-of-
Hairless repressor function in the developing Drosophila eye. Current Biology, 11, 330–338
Liu, F., Morrison, A. H., & Gregor, T. (2013). Dynamic interpretation of maternal inputs by
the Drosophila segmentation gene network. Proceedings of the National Academy of Sciences of
the United States of America, 110, 6724–6729
Lindenmeyer P, & Hosemann, R. (1963). Application of the theory of paracrystals to the
crystal structure analysis of polyacrylonitrile. Journal of Applied Physics. 34(1):42–45.
https://doi.org/10.1063/1.1729086
Lott, S. E., Kreitman, M., Palsson, A., Alekseeva, E., & Ludwig, M. (2007). Canalization of
segmentation and its evolution in Drosophila. Proceedings of the National Academy of Sciences
of the United States of America, 104, 10926–10931
Lubensky DK, Pennington MW, Shraiman BI, Baker NE. (2011). A dynamical model of
ommatidial crystal formation. Proceedings of the National Academy of Sciences. 108(27):11145–
11150. https://doi. org/10.1073/pnas.1015302108
Lucas, T., Ferraro, T., Roelens, B., De Las Heras Chanes, J., Walczak, A. M., Coppey, M., &
Dostatni, N. (2013). Live imaging of bicoiddependent transcription in Drosophila embryos.
Current Biology, 23, 2135–2139
118
Lynch, M. and Walsh, B. (1998). Genetics and analysis of quantitative traits (pp. 980).
Sunderland, MA: Sinauer Associates
Ma, C., & Moses, K. (1995). Wingless and patched are negative regulators of the
morphogenetic furrow and can affect tissue polarity in the developing Drosophila compound
eye. Development, 121, 2279–2289
Ma, C., Zhou, Y., Beachy, P. A., & Moses, K. (1993). The segment polarity gene hedgehog is
required for progression of the morphogenetic furrow in the developing Drosophila eye. Cell, 75,
927–938
Mackay, T., Stone, E. & Ayroles, J. (2009). The genetics of quantitative traits: challenges and
prospects. Nat Rev Genet 10, 565–577 https://doi.org/10.1038/nrg2612
Mackay, T. F. C., Richards, S., Stone, E. A., Barbadilla, A., Ayroles, J. F., Zhu, D., Casillas,
S., Gibbs, R. A. (2012). The Drosophila melanogaster genetic reference panel. Nature, 482,
173–178
MacNeil, L. T., & Walhout, A. J. M. (2011). Gene regulatory networks and the role of
robustness and stochasticity in the control of gene expression. Genome Research, 21, 645–657
Maisnier-Patin, S., & Andersson, D. I. (2004). Adaptation to the deleterious effects of
antimicrobial drug resistance mutations by compensatory evolution. Research in Microbiology,
155, 360–369
Manton, J.D., Ostrovsky, A.D., Goetz L., Costa, M., Rohlfing, T., & Jefferis, G.S.X.E. (2014).
Combining Genome-scale Drosophila 3D Neuroanatomical Data by Bridging Template Brains.
BioRxiv. Cold Spring Harbor Labs Journals: 1–36
Mauss, A., Tripodi, M., Evers, J.F., Landgraf, M. (2009). Midline signalling systems direct
the formation of a neural map by dendritic targeting in the Drosophila motor system. PLoS
Biology. 7(9): e1000200
Marinov, G. K., Williams, B. A., McCue, K., Schroth, G. A., Gertz, J., Myers, R. M., & Wold,
B. J. (2014). From single-cell to cell-pool transcriptomes: Stochasticity in gene expression and
RNA splicing. Genome Research, 24, 496–510
Markov A (1954). The theory of Algorithms. Academy of Sciences of the USSR. 42:3–375
McKenzie, J. A. (1993). Measuring fitness and intergenic interactions: The evolution of
resistance to diazinon in Lucilia cuprina. Genetica, 90, 227–237
119
McKenzie, J. A., & Clarke, G. M. (1988). Diazinon resistance, fluctuating asymmetry and
fitness in the Australian sheep blowfly, Lucilia cuprina. Genetics, 120, 213–220
McKenzie, J. A., Whitten, M. J., & Adena, M. A. (1982). The effect of genetic background on
the fitness of diazinon resistance genotypes of the Australian sheep blowfly, Lucilia cuprina.
Heredity, 49, 1–9
Metzger BP, Duveau F, Yuan DC, Tryban S, Yang B, Wittkopp PJ. Contrasting frequencies
and effects of cis-and trans-regulatory mutations affecting gene expression. Molecular biology
and evolution. 2016; 33(5):1131–1146. https://doi.org/10.1093/molbev/msw011 PMID:
26782996
Metzger BP, Yuan DC, Gruber JD, Duveau F, Wittkopp PJ. Selection on noise constrains
variation in a eukaryotic promoter. Nature. 2015; 521(7552):344.
https://doi.org/10.1038/nature14244 PMID: 25778704
Meiklejohn, C. D., Parsch, J., Ranz, J. M., & Daniel L. Hartl, D. L. (2003). Rapid evolution of
male-biased gene expression in Drosophila. PNAS. 100(17): 9894-9899
Mellerick, D.M., Kassis, J.A., Zhang, S.D., Odenwald, W.F. (1992). castor encodes a novel
zinc finger protein required for the development of a subset of CNS neurons in Drosophila.
Neuron. 9(8): 789-803
Meyer, F. (1994). Topographic distance and watershed lines. Signal Processing 38 (1): 113–
25 http://dx.doi.org/10.1016/0165-1684(94)90060-4
Moore, F. B., Rozen, D. E., & Lenski, R. E. (2000). Pervasive compensatory adaptation in
Escherichia coli. Proceedings of the Royal Society B, 267, 515–522
Namba, R., Pazdera, T. M., Cerrone, R. L., & Minden, J. S. (1997). Drosophila embryonic
pattern repair: How embryos respond to bicoid dosage alteration. Development, 124, 1393–1403
Neumüller, R.A., Richter, C., Fischer, A., Novatchkova, M., Neumüller, K.G., Knoblich, J.A.
(2011). Stem Cell. 8(5): 580-593
Nijhout, H. F., Sadre-Marandi, F., Best, J., & Reed, M. C. (2017). Systems biology of
phenotypic plasticity and robustness. Integrative and Comparative Biology, 57, 171–184
Nishiyama H., Knopfel T., Endo S., & Itohara S. (2002). Glial protein S100B modulates long-
term neuronal synaptic plasticity. Proc Natl Acad Sci. 99(6):4037-42
Okabe A, Boots B, Sugihara K, & Chiu SN (2009). Spatial tessellations: concepts and
applications of Voronoi diagrams. vol. 501. John Wiley & Sons
Oliver, B., Kim, Y. J., & Baker, B. S. (1993). Sex-lethal, master and slave: A hierarchy of
germline sex determination in Drosophila. Development. 119:897-908
120
Osterwalder M., Barozzi, I., & Tissières V. (2018). Enhancer redundancy provides phenotypic
robustness in mammalian development. Nature (London). 554(7691):239-243.
doi:10.1038/nature25461
Paaby, A., & Gibson, G. (2016). Cryptic genetic variation in evolutionary developmental
genetics. Biology, 5, 28–13.
Paaby, A. B., & Rockman, M. V. (2014). Cryptic genetic variation: Evolution's hidden substrate.
NatureReviews Genetics, 15, 247–258
Pai, A. A., & Gilad, Y. (2014). Comparative studies of gene regulatory mechanisms. Current
Opinion in Genetics & Development, 29, 68–74
Pai, A. A., Pritchard, J. K., & Gilad, Y. (2015). The genetic and mechanistic basis for
variation in gene regulation. PLoS Genetics, e1004857
Pan, D., & Rubin, G. M. (1995). CAMP-dependent protein kinase and hedgehog act
antagonistically in regulating decapentaplegic transcription in Drosophila imaginal discs. Cell,
80, 543–552
Parks, A. L., Turner, F. R., & Muskavitch, M. A. (1995). Relationships between complex
Delta expression and the specification of retinal cell fates during Drosophila eye development.
Mechanisms of Development, 50, 201–216
Parrish, J.Z., Kim, M.D., Jan, L.Y., Jan, Y.N. (2006). Genome-wide analyses identify
transcription factors required for proper morphogenesis of Drosophila sensory neuron dendrites.
Genes Dev. 20(7): 820-835
Penn J. K. M. and Schedl, P. (2007). The Master Switch Gene Sex-Lethal Promotes Female
Development by Negatively Regulating the N-signaling Pathway. Developmental Cell.
12(2):275-86. doi: 10.1016/j.devcel.2007.01.009
Perry, M. W., Boettiger, A. N., Bothma, J. P., & Levine, M. (2010). Shadow Enhancers Foster
Robustness of Drosophila Gastrulation. Current biology. 20(17):1562-1567.
doi:10.1016/j.cub.2010.07.043
Penalva, L. O. & Sanchez, L. (2003). RNA Binding Protein Sex-Lethal (Sxl) and Control
of Drosophila Sex Determination and Dosage Compensation. Microbiology and Molecular
Biology Reviews 67 (3) 343-359; DOI: 10.1128/MMBR.67.3.343-359.2003
Plavicki, J., Mader, S., Pueschel, E., Peebles, P., Boekhoff-Falk, G. (2012). Homeobox gene
distal-less is required for neuronal differentiation and neurite outgrowth in the Drosophila
olfactory system. Proc. Natl. Acad. Sci. 109(5): 1578-1583
Posnien, N., Hopfen, C., Hilbrant, M., Ramos-Womack, M., Murat, S., Schönauer, A.,
Herbert, S. L., Nunes, M. D., Arif, S., Breuker, C. J., Schlötterer, C., Mitteroecker, P., &
121
McGregor A. P. (2012). Evolution of eye morphology and rhodopsin expression in the
Drosophila melanogaster species subgroup. PLoS One.7(5):e37346. doi:
10.1371/journal.pone.0037346
R Core Team (2020). R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/
Rachev, S. T., Klebanov, L., Stoyanov, S. V., & Fabozzi, F. (2013). The methods of distances
in the theory of probability and statistics. Springer Science & Business Media
Raj, A., van den Bogaard, P., Rifkin, S. A., van Oudenaarden, A., & Tyagi, S. (2008).
Imaging individual mRNA molecules using multiple singly labeled probes. Nature Methods, 5,
877–879
Raj, A., & van Oudenaarden, A. (2008). Stochastic gene expression and its consequences.
Cell, 135, 216–226
Reed, R. D., Papa, R., Martin, A., Hines, H. M., Counterman, B. A., Pardo-Diaz, C.,
McMillan, W. O. (2011). optix drives the repeated convergent evolution of butterfly wing pattern
mimicry. Science, 333, 1137–1141
Restrepo, S. Zartman, J. J., & Basler, K. (2016). Cultivation and Live Imaging of Drosophila
Imaginal Discs. Nature Communications. 1478: 203-213
Rideout, E. J., Dornan, A. J., Neville, M. C., Eadie, S., & Goodwin, S. F. (2010). Control of
Sexual Differentiation and Behavior by the Doublesex Gene in Drosophila melanogaster. Nature
Neuroscience. 13: 458–466
Roignant, J.-Y., & Treisman, J. E. (2009). Pattern formation in the Drosophila eye disc. The
International Journal of Developmental Biology, 53, 795–804
Romero, I. G., Ruvinsky, I., & Gilad, Y. (2012). Comparative studies of gene expression and
the evolution of gene regulation. Nature Reviews Genetics, 13, 505–516
Rosenblum, E. B., Römpler, H., Schöneberg, T., & Hoekstra, H. E. (2010). Molecular and
functional basis of phenotypic convergence in white lizards at White Sands. Proceedings of the
National Academy of Sciences of the United States of America, 107, 2113–2117
Rutherford, S., Yoshikazu, H., & Swalla, B. J. (2007). The Hsp90 capacitor, developmental
remodeling, and evolution: The robustness of gene networks and the curious evolvability of
metamorphosis. Critical Reviews in Biochemistry and Molecular Biology, 42, 355–372
Ryner, L.C., Goodwin, S. F., Castrillon, D. H., Anand, A., Villella, A., Baker, B. S., Hall,
J.C., & Wasserman, S. A. (1996). Control of male sexual behavior and sexual orientation in
Drosophila by the fruitless gene. Cell. 87:1079 –1089
122
Samson M. L. & Rabinow, L. (2014). Transcriptomic analysis of sexual differentiation in
somatic tissues of Drosophila melanogaster: successes and caveats. Sex Development; 8(1-
3):113–26
Sanders L.E. & Arbeitman M.N. (2008). Doublesex Establishes Sexual Dimorphism in the
Drosophila Central. Developmental Biology. 320(2):378-90
Shah, S., Lubeck, E., Schwarzkopf, M., & He, T. F. (2016). Singlemolecule RNA detection at
depth via hybridization chain reaction and tissue hydrogel embedding and clearing.
Development, 143, 2862–2867
Siegel, R. W., and Hall, J. C. (1979). Conditioned responses in courtship behavior of normal
and mutant Drosophila. Proceedings of the National Academy of Science. USA 76, 3430–3434
Signor, S. A., Liu, Y., Rebeiz, M., & Kopp, A. (2016). Genetic convergence in the evolution
of male-specific color patterns in Drosophila. Current Biology, 26, 2423–2433. Signor, S. A.,
New, F. N., & Nuzhdin, S. (2017). A Large panel of Drosophila simulans reveals an abundance
of common variants. Genome Biology and Evolution, 10, 189–206
Signor, S. A., & Nuzhdin, S. (2018). Dynamic changes in gene expression and alternative
splicing mediate the response to acute alcohol exposure in Drosophila melanogaster. Heredity,
121, 342–360
Spencer, S. A., Powell, P. A., Miller, D. T., & Cagan, R. L. (1998). Regulation of EGF
receptor signaling establishes pattern across the developing Drosophila retina. Development,
125, 4777–4790
Stern, D. L. (2000). Evolutionary developmental biology and the problem of variation.
Evolution, 54, 1079–1091
Stoebel, D. M., Hokamp, K., Last, M. S., & Dorman, C. J. (2009). Compensatory evolution of
gene regulation in response to stress by Escherichia coli lacking RpoS. PLoS Genetics, 5,
e1000671
Strutt, D. I., & Mlodzik, M. (1997). Hedgehog is an indirect regulator of morphogenetic
furrow progression in the Drosophila eye disc. Development, 124, 3233–3240
Sukontason K.L., Chaiwong, T., Piangjai, S., Upakut, S., & Moophayak, K.(2008).
Ommatidia of blow fly, house fly, and flesh fly: implication of their vision efficiency. Parasitol
Res.;103:123–131
Sun, Y., Jan, L. Y., & Jan, Y. N. (1998). Transcriptional regulation of atonal during
development of the Drosophila peripheral nervous system. Development, 125, 3731–3740
Szamecz, B., Boross, G., Kalapis, D., Kovács, K., Fekete, G., Farkas, Z., Pál, C. (2014). The
genomic landscape of compensatory evolution. PLoS Biology, 12, e1001935
123
Tantale, K., Kozulic-Pirher, A., Lesne, A., Victor, J. M., Robert, M. C., Capozi, S., Bertrand,
E. (2016). A single-molecule view of transcription reveals convoys of RNA polymerases and
multi-scale bursting. Nature Communications, 7, 1–14
Tanaka K, Barmina O, & Kopp A. (2009). Distinct developmental mechanisms underlie the
evolutionary diversification of Drosophila sex combs. Proceedings of The National Academy of
Science. 106(12):4764–4769
Testa N.D., Ghosh S.M., & Shingleton A.W. (2013) Sex-Specific Weight Loss Mediates
Sexual Size Dimorphism in Drosophila melanogaster. PLoS ONE 8(3): e58936.
https://doi.org/10.1371/journal.pone.0058936
Thompson, A., Zakon, H. H., & Kirkpatrick, M. (2015). Compensatory drift and the
evolutionary dynamics of dosage-sensitive duplicate genes. Genetics, 202, 765–774
True, J.R., & Haag, E. S. (2001). Developmentalsystemdrift and flexibility in evolutionary
trajectories. Evolution & Development, 3, 1–11
Truskett TM, Torquato S, Debenedetti PG. Towards a quantification of disorder in materials:
Distinguishing equilibrium and glassy sphere packings. Physical Review E. 2000; 62(1):993.
https://doi.org/10.1103/PhysRevE.62.993
Tsachaki, M., & Sprecher, S. G. (2011). Genetic and developmental mechanisms underlying
the formation of the Drosophila compound eye. Developmental Dynamics, 241, 40–56
Tsuji, T., Hasegawa, E., Isshiki, T. (2008). Neuroblast entry into quiescence is regulated
intrinsically by the combined action of spatial Hox proteins and temporal identity factors.
Development 135(23): 3859-3869
Usui-Aoki, K., Ito, H., Ui-Tei, K., Takahashi, K., Lukacsovich, T., Awano, W., Nakata, H.,
Piao, Z. F., Nilsson, E. E., Tomida, J., & Yamamoto, D. (2000). Formation of the Male-specific
Muscle in Female Drosophila by Ectopic Fruitless Expression. Nature Cell Biology. 2:500–506
Vied C. & Horabin J. I. (2001). The sex determination master switch, Sex-lethal, responds to
Hedgehog signaling in the Drosophila germline. Development; 128:2649-60
Villella, A., Gailey, D. A., Berwald, B., Ohshima, S., Barnes, P. T., & Hall, J. C. (1997).
Extended reproductive roles of the fruitless gene in Drosophila melanogaster revealed by
behavioral analysis of new fru mutants. Genetics. 147:1107–1130
Vincent, L. (1993). Morphological grayscale reconstruction in image analysis: Applications
and efficient algorithms. IEEE Transactions on Image Processing, 2, 176–201
124
Voas, M. G. & Rebay, I. (2004). Signal integration during development: insights from the
Drosophila eye. Developmental dynamics: an official publication of the American Association of
Anatomists. 229(1):162– 175. https://doi.org/10.1002/dvdy.10449
von Hilchen, C. M., Hein, I., Technau, G. M., & Altenhein, B. (2010). Netrins guide migration
of distinct glial cells in the Drosophila embryo. Development 137(8): 1251-1262
Von Philipsborn, A. C., Liu, T., Yu, J. Y., Masser, C., Bidaye, S. S., & Dickson, B. J. (2011).
Neuronal Control of Drosophila Courtship Song. Neuron. 69: 509–522
Waddington, C. H. Canalization of development and genetic assimilation of acquired
characters. Nature. 1959; 183(4676):1654. https://doi.org/10.1038/1831654a0 PMID: 13666847
Waddington, C. H. & Perry, M. M. (1960). The ultrastructure of the developing eye of
Drosophila. Proceedings to the Royal Society of Biological Sciences. 153:155–178
Waller, I. (1923). Zur frage der einwirkung der wa¨rmebewegung auf die interferenz von
ro¨ntgenstrahlen. Zeitschrift fu¨r Physik. 17(1):398–408. https://doi.org/10.1007/BF01328696
Waterbury J. A., Horabin J. I., Bopp D., & Schedl P. (2000). Sex determination in the
Drosophila germline is dictated by the sexual identity of the surrounding soma. Genetics. 155:
1741-56
Xiao, Q., Komori, H., Lee, C.Y. (2012). klumpfuss distinguishes stem cells from progenitor
cells during asymmetric neuroblast division. Development 139(15): 2670-2680
Wernet, M. F., Mazzoni, E. O., Çelik, A., Duncan, D. M., Duncan, I., & Desplan, C. (2006).
Stochastic spineless expression creates the retinal mosaic for colour vision. Nature, 440, 174–
180
Wolff T & Ready D. F. (1991). Cell death in normal and rough eye mutants of Drosophila.
Development 113(3): 825–839
Wolff T. & Ready D. F. (1991). The beginning of pattern formation in the Drosophila
compound eye: the morphogenetic furrow and the second mitotic wave. Development
113(3):841–850
Wotton, K. R., Jimenez-Guri, E., Crombach, A., Janssens, H., AlcaineColet, A., Lemke, D.,
Jaeger, J. (2015). Quantitative system drift compensates for altered maternal inputs to the gap
gene network of the scuttle fly Megaselia abdita. Elife, 4, 1–28
Wray, G. A. (2007). The evolutionary significance of cis-regulatory mutations. Nature
Reviews Genetics, 8, 206–216
125
Wunderlich, Z., Bragdon, M. D., Eckenrode, K. B., Lydiard-Martin, T., Pearl-Waserman, S.,
& DePace, A. H. (2012). Dissecting sources of quantitative gene expression pattern divergence
between Drosophila species. Molecular Systems Biology, 8, 1–15
Yamamoto D. & Koganezawa M. (2013). Genes and Circuits of Courtship Behaviour in
Drosophila Males. Nature Reviews Neuroscience. 14(10): 681–92
Yassin, A., Delaney, E. K., Reddieux, A., Bastide, H., Seher, T., Appleton, N. C., Kopp, A.
(2016). pdm3 is responsible for recurrent evolution of female-limited color dimorphism in
Drosophila. Current Biology, 26, 2412–2422
Yu J.Y., Kanai M.I., Demir E., Jefferis G.S.X.E., & Dickson, B.J. (2010). Cellular
Organization of the Neural Circuit That Drives Drosophila Courtship Behavior. Current Biology.
20(18):1602–14
Abstract (if available)
Abstract
Biological development requires the careful coordination of several processes, such as transcription, replication, differentiation and patterning. These intrinsically stochastic processes require developmental programs to be robust. This robustness is maintained through an intricate web of interacting G̲ene R̲egulatory N̲etworks (GRNs) and clean-up mechanisms. As species diverge, stochasticity in GRNs can lead to developmental systems drift, whereby genetic variation accumulates between populations without any observable phenotypic change. How can biological systems use variation in gene expression, or transcriptional stochasticity, to balance developmental robustness versus evolutionary adaptability depending on developmental and environmental contexts? The Drosophila compound eye is an excellent system to study these interlinked topics since individual eye units, or ommatidia, specified in the larval eye disc eventually form the highly-organized, offset patterning seen in adult eyes. ❧ We measured the quantitative and spatial relationship of 4 key genes involved in eye development simultaneously, using the Hybridization Chain Reaction (HCR), in 3 natural populations of D. melanogaster and D. simulans males and females respectively. Our novel published findings show significant transcriptional stochasticity between species, genotypes and sexes, as well as regulatory logic divergence despite the phenotypic similarity of Drosophila melanogaster and D. simulans eyes (Ali 2019, Chapter 2). ❧ Afterwards, we created a mathematical model to understand the observed phenotypic robustness in eye GRNs. Our published results quantitatively assessed this threshold response, whereby transcriptional stochasticity is buffered up to a certain threshold, after which the system rapidly degrades (Courcoubetis 2019, Chapter 3). ❧ We further investigated the eye GRN in different sexual contexts by genetically manipulating the Sex Determination Hierarchy (SDH) and performing HCR on several genes in male, female and pseudomale (masculinized females) larval eye discs, which showed no differences. (Chapter 4). To understand how GRNs evolve between closely related species, we used HCR to measure the quantitative and spatial relationships of fruitless (fru) Neuron Expression Clusters (NECs) in the SDH between D. melanogaster and D. simulans brains, which showed transcriptional and volumetric variation in several reproductively-relevant brain regions (Chapter 5). ❧ Collectively, our findings improve our understanding of the genotype-to-phenotype map, which is instrumental in elucidating the genetic basis of complex diseases.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Innovative sequencing techniques elucidate gene regulatory evolution in Drosophila
PDF
The evolution of gene regulatory networks
PDF
Exploring the genetic basis of quantitative traits
PDF
Patterning and growth in biological systems: a computational exploration of biological robustness, epithelial growth dynamics and development
PDF
Complex mechanisms of cryptic genetic variation
PDF
Behavioral choice assays and alcohol preference in Drosophila melanogaster
PDF
Biological interactions on the behavioral, genomic, and ecological scale: investigating patterns in Drosophila melanogaster of the southeast United States and Caribbean islands
PDF
Genetic architectures of phenotypic capacitance
PDF
Evolutionary genomic analysis in heterogeneous populations of non-model and model organisms
PDF
Studies in bivalve aquaculture: metallotoxicity, microbiome manipulations, and genomics & breeding programs with a focus on mutation rate
PDF
Characterization of Drosophila longevity and fecundity regulating genes
PDF
Probing the genetic basis of gene expression variation through Bayesian analysis of allelic imbalance and transcriptome studies of oil palm interspecies hybrids
PDF
Investigating the potential roles of three mammalian traits in female reproductive investment
PDF
Transcriptional and morphological impacts of copper on Mytilus californianus larval development in current and future ocean conditions
PDF
Male-female conflict after mating: function and dynamics of the copulatory plug in mice (Mus domesticus)
PDF
Applications of next generation sequencing in sessile marine invertabrates
PDF
Model selection methods for genome wide association studies and statistical analysis of RNA seq data
PDF
Genetic engineering of fungi to enhance the production and elucidate the biosynthesis of bioactive secondary metabolites
PDF
Decoding the embryo: on spatial and genomic tools to characterize gene regulatory networks in development
PDF
Detecting joint interactions between sets of variables in the context of studies with a dichotomous phenotype, with applications to asthma susceptibility involving epigenetics and epistasis
Asset Metadata
Creator
Ali, Sammi
(author)
Core Title
Robustness and stochasticity in Drosophila development
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Molecular Biology
Publication Date
12/01/2020
Defense Date
10/26/2020
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Biology,Development,Drosophila,genetics,HCR,hybridization chain reaction,microscopy,molecular,OAI-PMH Harvest,quantitative,robustness,stochasticity
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Nuzhdin, Sergey (
committee chair
), Ehrenreich, Ian (
committee member
), Haas, Stephan (
committee member
), Tower, John (
committee member
)
Creator Email
SammiAli.usc@gmail.com,SammiAli@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-401455
Unique identifier
UC11666701
Identifier
etd-AliSammi-9176.pdf (filename),usctheses-c89-401455 (legacy record id)
Legacy Identifier
etd-AliSammi-9176.pdf
Dmrecord
401455
Document Type
Dissertation
Rights
Ali, Sammi
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
Drosophila
genetics
HCR
hybridization chain reaction
molecular
quantitative
robustness
stochasticity