Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
THE CIRCADIAN REGULATION OF FEEDING IN
ADULT DROSOPHILA MELANOGASTER
by
Shreya Shekhar
A thesis submitted in conformity with the requirements
for the degree of Master of Science
Graduate Department of Cell and Systems Biology
University of Toronto
© Copyright by Shreya Shekhar (2010)
ii
The Circadian Regulation of Feeding in Adult Drosophila melanogaster
Shreya Shekhar
Master of Science
Department of Cell and Systems Biology
University of Toronto
2010
Abstract
In nature, all organisms face the daily challenges created by a fluctuating environment.
Circadian clocks synchronize behaviour and physiology allowing an organism to adapt to and
predict daily changes to environmental conditions. In the fruit fly, Drosophila melanogaster,
circadian clocks reside in a set of ~150 neurons in the brain, collectively referred to as the
central clock, and in the cells of many peripheral tissues. The central clock regulates daily
behavioural rhythms, whereas peripheral clocks are thought to regulate the local metabolic
activities of the cells in which they reside. In this thesis, I demonstrate that a peripheral clock
resides in the abdominal fat body, a tissue analogous to the mammalian liver and adipocytes.
Moreover, I show that flies display a temporal feeding pattern that is partly regulated by a
peripheral clock. I propose that the central clock and peripheral clocks coordinate to regulate the
timing of fly feeding behaviour.
iii
Acknowledgments
I would like to start by thanking my supervisor, Joel Levine, for his support and
guidance through my graduate studies. It was your confidence in me and my capabilities that
have allowed me to grow as a researcher and person. I am grateful to Josh Krupp, for being an
excellent teacher and friend, whose valuable lessons will guide me throughout my scientific
career. I thank both Joel and Josh for editing the thesis and providing helpful comments. I
appreciate Jonathan Schneider‟s assistance with the statistics performed to analyze the data
presented here. I could not have understood it without your easy-to-follow explanations.
I am grateful to my thesis supervisors, Drs. Marla Sokolowski, Tim Westwood, and
Angela Lange whose advice and suggestions have been helpful in my research. Also, thank you
to Dr. Sokolowski for her help in preparing for my exit seminar.
I would like to thank all the members of the Levine lab for making the past few years in
lab an adventure. Research life could not have been half as fun without the birthday
celebrations, badminton tournaments and Christmas parties. I would like to thank Jean-
Christophe Billeter for being an excellent teacher in Cloning School. And of course I must thank
the girls, Farheen Mohammed, Jade Atallah and Sam Jagadeesh, from „Cloning School 2009‟.
Who knew making competent cells and doing PCRs could be such fun! Also, thanks to all
members for their feedback during general lab meetings and in preparation for the defense
exam. A big thanks to past lab members including Olga Sizova, Nancy Stepek, Hania Pavlou,
Adrienne Chu and Richard Dunbar-Yaffe.
iv
I would like to thank all the members of the Sokolowski lab, our extended fly family.
The afternoon coffee or evening chats in your lab have always the highlight of my day. Also,
your advice during thesis writing and in preparation for the exit seminar was very helpful.
Finally, the past few years would not have flown by so quickly without my family. I am
grateful for having loving parents, who have always provided encouragement and support. And
I am thankful to my brother, Mukul, who has, on more than one occasion, been forced to listen
to my ramblings about the thrills and disappointments of an experiment. Lastly, I am grateful
that “Tigger”, our new kitten, walked into my life a month ago. She has been a source of
amusement in the final leg of thesis writing.
v
Table of Contents
Abstract ...................................................................................................................................................................ii
Acknowledgments .................................................................................................................................................. iii
Table of Contents..................................................................................................................................................... v
List of Tables ......................................................................................................................................................... vii
List of Figures ...................................................................................................................................................... viii
List of Appendices .................................................................................................................................................. ix
Chapter 1 . Introduction to Circadian Clocks ......................................................................................................... 1
A Network of Neural Clocks Regulate Daily Locomotor Activity Rhythms in Drosophila ......................... 2
The Neuropeptide, Pigment Dispersing Factor, Synchronizes the Cells of the Central Clock ...................... 3
Circadian clocks Residing in Peripheral Tissues Temporally Regulate Local Metabolic Functions ............. 5
The Molecular Mechanism of the Circadian Clock ....................................................................................... 6
cryptochrome-Mediated Entrainment of Circadian Clock Cells .................................................................... 9
Thesis Objectives ......................................................................................................................................... 11
Chapter 2 . A Peripheral Clock in the Fly Fat Body ............................................................................................. 12
Introduction ...................................................................................................................................................... 12
Methods ............................................................................................................................................................ 15
Strains .......................................................................................................................................................... 15
Fly Stocks .................................................................................................................................................... 15
Fly Collections for Dissections .................................................................................................................... 16
Fat Body Dissection Procedure .................................................................................................................... 16
Genomic DNA Extraction Protocol ............................................................................................................. 17
Reverse Transcription Polymerase Chain Reactions and Agarose Gel Electrophoresis .............................. 17
Quantitative PCR ......................................................................................................................................... 18
qPCR Analysis ............................................................................................................................................. 19
Statistical Analysis ....................................................................................................................................... 19
Results .............................................................................................................................................................. 19
The Core Clock Genes are Expressed in the Abdominal Fat Body ............................................................. 19
The Abdominal Fat Body Contains a Circadian Clock ................................................................................ 21
The Fat Body Clock is Dependent on period Expression ............................................................................ 21
Discussion ......................................................................................................................................................... 23
vi
The Abdominal Fat Body is a Peripheral Clock .......................................................................................... 23
Chapter 3 . Circadian Regulation of Fly Feeding .................................................................................................. 26
Introduction ...................................................................................................................................................... 26
Methods ............................................................................................................................................................ 30
Strains .......................................................................................................................................................... 30
Fly Collections for Behavioural Assays ....................................................................................................... 30
The Capillary Feeder (CAFE) Assay ........................................................................................................... 30
Analysis of Feeding ..................................................................................................................................... 32
Statistical Analysis ....................................................................................................................................... 33
Locomotor Activity Assay ........................................................................................................................... 33
Results .............................................................................................................................................................. 34
Flies Display a Circadian Feeding Pattern ................................................................................................... 34
The Feeding Pattern is Regulated by a period-Dependent Clock ................................................................ 36
Peripheral Clocks Regulate the Temporal Pattern of Food Intake ............................................................... 38
Light-Entrainment of the Circadian Clock is Essential for Maintaining Feeding Rhythms ......................... 41
The Polymorphism in foraging Affects the Circadian Regulation of Feeding ............................................. 44
Analyzing Fly Meals .................................................................................................................................... 46
Discussion ......................................................................................................................................................... 48
Endogenous Clocks Regulate the Temporal Feeding Pattern ...................................................................... 48
Continuous Illumination Disrupts Behavioural Rhythms ............................................................................ 50
The Relationship between Feeding and Locomotor Activity ....................................................................... 51
The Role of the foraging gene in Feeding Rhythms .................................................................................... 52
Defining a Fly Meal ..................................................................................................................................... 53
Chapter 4 . Discussion .............................................................................................................................................. 55
References ........................................................................................................................................................ 58
Appendix A: Detailed Method for Preparing Fly Food .................................................................................... 65
Appendix B: Fat Body Timeseries Data ........................................................................................................... 66
Appendix C: Statistical Analysis ...................................................................................................................... 69
vii
List of Tables
Table 1: The relative RNA expression levels of period, timeless and Clock genes in the fat body
tissue of Canton-S flies in a light-dark cycle.
Table 2: The relative RNA expression levels of period, timeless and Clock genes in the fat body
tissue of Canton-S flies under constant darkness.
Table 3: The relative RNA expression levels of period, timeless and Clock genes in the fat body
tissue of y w flies under constant darkness.
Table 4: The relative RNA expression levels of period, timeless and Clock genes in the fat body
tissue of per01
flies under constant darkness.
Table 5: The parameters of cosine curves fit to relative RNA expression levels in the fat body
tissue of Canton-S flies.
Table 6: The parameters of cosine curves fit to relative RNA expression levels in the fat body
tissue of y w and per01
flies under constant darkness.
Table 7: The Statistical results of General Linear Model Repeated Measures tests performed to
compare feeding amounts between fly strains.
viii
List of Figures
Figure 1.1: The anatomical location and function of central clock neurons in the fly brain.
Figure 1.2: A schematic of the molecular clock in the central clock cells of Drosophila
melanogaster.
Figure 1.3: Daily resetting of the clock involves a CRYPTOCHROME-mediated pathway.
Figure 2.1: The location of fat body cells in the adult fruit fly.
Figure 2.2: Clock gene expression is detected in the abdominal fat body of male Canton-S flies.
Figure 2.3: A period-dependent peripheral clock resides in the abdominal fat body.
Figure 3.1: A diagram of the Capillary Feeder (CAFE) assay.
Figure 3.2: Wildtype flies show a temporal feeding pattern in a light-dark cycle and constant
darkness.
Figure 3.3: Feeding patterns appear disrupted in period mutant flies.
Figure 3.4: Neuronal clocks show some involvement in regulating the feeding rhythm.
Figure 3.5: Constant light disrupts normal feeding patterns.
Figure 3.6: Allelic variation in the foraging gene affects the temporal feeding pattern.
Figure 3.7: A pattern in the consumption of large meals appears to drive the temporal feeding
pattern.
Figure 4.1: Clocks in the brain and peripheral tissues may coordinate to regulate the temporal
feeding pattern in adult Drosophila.
ix
List of Appendices
Appendix A: Detailed Method for Preparing Fly Food
Appendix B: Fat Body Timeseries Data
Appendix C: Statistical Analysis
1
Chapter 1 . Introduction to Circadian Clocks
Daily rhythms in behaviour and physiology have been observed in a wide range of
organisms from unicellular bacteria to vertebrates (Dunlap et al., 2004). In one of the first
accounts of a documented biological rhythm, the Greek naturalist, Androsthenes, observed
that legume plants raise their leaves during the day and fold them at night. In 1729, the
French astronomer DeMairan found that mimosa plants continue to raise and fold their stems
even in constant darkness (DD) (Dunlap et al., 2004). These early observations led to the
theory that biological clocks reside within all organisms and regulate the timing of behaviour
and physiology.
Internal clocks allow an organism to anticipate and adapt to changes in the physical
environment. It is common for factors such as light intensity, humidity and temperature to
fluctuate on a daily basis. Survival depends on being able to adjust body physiology and
behaviour to anticipate and adapt to such changes (Reviewed in Dubruille and Emery, 2008).
For instance, many diurnal organisms adjust their activity in response to the reduction in
daylight during winter to be able to continue to find food before nightfall. Although an
internal clock maintains time autonomously, its rhythm is not exactly 24 hours and it must
synchronize to external time-cues or zeitgebers (a German term meaning “time giver”) such
as light intensity, temperature and food availability (Reviewed in Dubruille and Emery,
2008).
In the laboratory, circadian rhythms are measured in a 12hr:12hr light-dark (LD) cycle
resembling the solar day-night cycle. Under these conditions, clock-regulated behaviours
synchronize to the lighting conditions and exhibit approximately a 24 hour period. In the
absence of external time-cues, the clock becomes “free-running” and the period either
2
shortens or lengthens, rarely keeping to 24 hrs. Such clocks are thus called circadian (“about
a day”) (Dunlap et al., 2004).
The fruit fly, Drosophila melanogaster, has served as the preeminent model to study
circadian clocks. Before beginning a study of circadian rhythms, it is necessary to provide a
brief summary of the extensive circadian research that has been done in fruit flies. This
chapter begins with an introduction to neural clocks and the regulation of daily locomotor
activity rhythms. It will be followed by a description of clocks in peripheral tissues and their
role in the temporal regulation of local cellular functions. To understand the inner-workings
of the clock, the current model of the molecular clock mechanism and its entrainment to the
light-dark cycle will be provided. The Drosophila clock is for the most part homologous to
the mammalian clock, the similarities and differences between the two systems will be
discussed throughout the chapter.
A Network of Neural Clocks Regulate Daily Locomotor Activity Rhythms
in Drosophila
Konopka and Benzer (1971) were the first to identify the link between the molecular clock
and behaviour in fruit flies. They isolated mutants with altered pupal eclosion and locomotor
activity rhythms, two well-known circadian behaviours. Two mutants displayed rhythms
with altered periods, while one mutant showed arrhythmic behaviour. Complementation tests
determined that all three phenotypes were caused by mutations in a single locus on the X
chromosome, which they named period (per) (Konopka and Benzer, 1971). Subsequently,
other studies determined a unique characteristic of per, its mRNA and protein levels show
circadian oscillations (Hardin et al., 1990; Zerr et al., 1990). The cyclic expression of period
is important for maintaining normal behavioural rhythms.
In a light-dark cycle, wild type flies display an increase in activity at dawn and dusk each
day, corresponding to a transition in lighting (Figure 1.1) (Reviewed in Dubruille and
Emery, 2008). Flies anticipate the change in lighting by becoming more active several hours
before it occurs. In constant conditions, the bimodal activity pattern becomes unimodal, as
3
either the morning peak disappears or the two peaks become merged (Reviewed in Dubruille
and Emery, 2008). It is believed that different subsets of clock neurons in the brain regulate
different components of locomotor activity rhythms (Grima et al., 2004).
Clock neurons in the brain, collectively known as the central clock, are divided into lateral,
dorsal and lateral-posterior neurons, so named based on their relative locations in the fly
brain (Figure 1.1) (Reviewed in Nitabach and Taghert, 2008). Lateral neurons (LNs) are
divided into large-ventrolateral neurons (l-LNvs), 5 small-ventrolateral neurons (s-LNvs) and
dorsolateral neurons (LNds) whereas dorsal neurons (DNs) are categorized as DN1s, DN2s
and DN3s (Reviewed in Nitabach and Taghert, 2008). Different groups of clock neurons
appear to regulate different aspects of locomotor activity rhythms (Grima et al., 2004). 4 s-
LNvs are responsible for the morning activity peak (M peak) whereas the LNds and the 5th
s-
LNv regulate the evening peak together (E peak). These clock neurons also regulate the
anticipatory increase in activity corresponding to the M and E peaks. In constant conditions
s-LNvs alone are able to maintain rhythmic activity, indicating a more important role for
these clocks in DD (Grima et al., 2004). However, the LNds and the 5th
s-LNv are thought to
be responsible for maintaining the phase of the activity pattern. It appears that a complex
interaction between the morning and evening oscillators in the fly brain drives locomotor
activity rhythms.
The Neuropeptide, Pigment Dispersing Factor, Synchronizes the Cells of
the Central Clock
The only molecular output of the central clock that has been identified in Drosophila is the
neuropeptide, Pigment Dispersing Factor (PDF). PDF transmits temporal information
between clock neurons in different regions of the brain (Reviewed in Stanewsky, 2002).
Communication between these clocks is essential for maintaining locomotor activity
rhythms, as pdf-null flies display disrupted rhythms in a light-dark cycle and in constant
darkness (Renn et al., 1999). pdf is expressed primarily in the s-LNvs, 4 of the l-LNvs in the
fly brain and in the posterior region of the ventral ganglia (Park et al., 2000). Although pdf is
4
Figure 1.1: The anatomical location and function of central clock neurons in the fly brain.
The three types of brain clock cells are lateral neurons (LNs), dorsal neurons (DNs) and lateral-posterior
neurons (LPNs), which are named based on their location in the fly brain. These cells are implicated in
regulating daily locomotor activity rhythms in flies. Four small-LNv (s-LNv) neurons, also known as M-cells,
are required for maintaining the morning activity peak in a light-dark cycle. The evening activity peak is
regulated by the 5th s-LNv and dorsolateral neurons (LNds), collectively called the E-cells. In constant
darkness, the M-cells are able to control locomotor activity rhythms, although the E-cells regulate peak phase.
OL indicates the location of the optic lobes. Figure source: Dubruille and Emery (2008).
not expressed cyclically, PDF peptide is detected in nerve terminals in a cyclic manner; with
highest and lowest levels detected in the early morning and early night hours, respectively
(Park et al., 2000). It is thought that the rhythmic release of PDF is important for
synchronizing the timing of different clocks and maintaining behavioural rhythms.
The PDF receptor has been identified as a G-protein coupled receptor. PDF expression is
found in close proximity to cells expressing PDFR (Mertens et al., 2005). However, there
are discrepancies regarding where the ligand and receptor are expressed as some LNds are
responsive to PDF but do not express PDFR (Im and Taghert, 2010). Also, mutations in pdf
and pdfr show similar but not identical phenotypes (Mertens et al., 2005). This suggests
5
there may be additional receptors which receive PDF input or that PDFR may have a second
ligand (Im and Taghert, 2010).
Circadian clocks Residing in Peripheral Tissues Temporally Regulate
Local Metabolic Functions
Oscillators also exist in tissues outside of the central clock cells. These „peripheral‟ clocks
are found in tissues such as the antennae, proboscis, prothoracic gland, gut and oenocytes
(Chatterjee et al., 2010; Emery et al., 1997; Reviewed in Hardin, 2005; Krishnan et al.,
1999; Krupp et al., 2008). Similar to central clock cells, period mRNA levels also cycle in
peripheral clock cells (Hardin, 1994). In mammals, clocks in peripheral tissues are unable to
maintain rhythmicity and are thought to be entrained by the central clock, a group of neurons
in the Suprachiasmatic Nuclei (SCN) of the hypothalamus (Reviewed in Balsalobre, 2002).
Using a luminescence assay, it was shown that cultured rat SCN maintains rhythmicity in
constant darkness whereas peripheral clocks are less robust and damp after several days
(Yamazaki et al., 2000). Since the central clock synchronizes peripheral oscillators, it
appears that the peripheral clocks are linked to the SCN in a „master-slave‟ relationship
(Reviewed in Balsalobre, 2002). A similar relationship between central and peripheral clocks
is not observed in fruit flies. Isolated peripheral tissue clocks can entrain to light without
cues from central clock cells (Plautz et al., 1997). per-driven bioluminescence was measured
in isolated tissue cultures of the proboscis, antenna, legs, and wings. When shifted lighting
conditions were imposed, these tissues re-synchronized indicating that peripheral clocks are
able to maintain time autonomously independently of the central clock (Plautz et al., 1997).
In peripheral tissues, an internal clock is thought to regulate the rhythms of local processes.
For example, the clock in oenocytes appears to regulate the synthesis of cuticular
hydrocarbons (Krupp et al., 2008). The peripheral oenocyte clock generates these rhythms
by controlling the expression of a key enzyme, Desaturase1 (Desat1), involved in cuticular
hydrocarbon synthesis. Other peripheral clocks are found in the chemosensory bristles of the
antennae and proboscis, the smell and taste organs of the fruit fly (Chatterjee et al., 2010;
6
Krishnan et al., 1999). These clocks regulate circadian rhythms in olfactory and gustatory
sensitivity. Due to the importance of taste and smell to feeding behaviour, the oscillators in
the chemosensory organs are hypothesized to regulate the timing of foraging behaviour and
food intake.
The Molecular Mechanism of the Circadian Clock
Since the identification of period, several other genes which also show circadian regulation
have been identified, and a molecular model of the circadian clock has been assembled. In
Drosophila, this involves two transcriptional-translational feedback loops which control the
expression of several genes (Reviewed in Hardin, 2005). In the first loop (Figure 1.2), the
major clock genes involved are Clock (Clk), cycle (cyc), period (per), timeless (tim) and
doubletime (dbt). CLOCK and CYCLE are transcription factors containing a PER-ARNT-
SIM (PAS) protein-protein binding domain and a basic helix-loop-helix (bHLH) DNA
binding domain. Clk is under circadian control and is expressed cyclically, whereas cyc is
constitutively expressed and remains at a constant level. Clk mRNA and protein levels peak
in the early morning and are lowest in the early evening (Reviewed in Hardin, 2005). When
expressed, CLK heterodimerizes with CYCLE and upregulates the transcription of two other
genes which also encode transcription factors, period and timeless. In the middle of the day,
the CLK-CYC heterodimer bind both the E-box sequences, conserved sequences of six
nucleotides, in the regulatory regions of per and tim genes (Reviewed in Stanewsky, 2002).
This leads to an increase in per and tim transcription in the early evening (Reviewed in
Hardin, 2005).
The PER and TIM proteins are located in the cytoplasm, where they reach peak levels by
late night. When PER is expressed, it is phosphorylated by two kinases, DBT and CASEIN
KINASE 2 (CK2), this leads to its destabilization and degradation (Reviewed in Stanewsky,
2002). It is thought that the kinase, PROTEIN PHOSPHATASE 2a (PP2a), counteracts by
removing phosphates added to PER. Together, this process regulates the timing at which
PER reaches maximum expression levels. As TIM levels increase, it heterodimerizes with
7
phosphorylated PER, and increases its stability (Reviewed in Hardin, 2005). The interaction
between the two proteins is achieved via PER‟s PAS domain and TIM‟s ARMADILLO-like
domains (Reviewed in Stanewsky, 2002). It is thought that DBT binds to PER and the entire
TIM-PER-DBT complex proceeds to the nucleus (Reviewed in Hardin, 2005). However, the
kinases, SHAGGY (SGG) and CK2 first phosphorylate TIM and PER, respectively. Once it
enters the nucleus, the TIM-PER-DBT complex binds to the CLK-CYC heterodimer and
prevents further expression of per and tim. In the nucleus, DBT phosphorylates PER and
CLK, which causes their degradation. TIM is degraded via a light-mediated pathway, which
resets the clock mechanism (Reviewed in Hardin, 2005).
In the second transcriptional-translational feedback loop, the transcription factors encoded
by vrille (vri) and par domain protein 1ε (pdp1ε) genes regulate Clk expression (Reviewed
in Hardin, 2005). The CLK-CYC heterodimers bind to the E-box sequences in the promoter
regions of these genes and activate transcription. VRI reaches peak levels late in the day, and
begins binding to a VRI/PDP1ε box (V/P box) in the regulatory region of Clk and represses
its transcription (Reviewed in Hardin, 2005). This repression is offset by PDP1ε. After
PDP1ε protein reaches peak levels in the mid to late evening hours, it competes with VRI for
the V/P box and when bound it activates Clk transcription (Cyran et al., 2003). As a result
Clk mRNA and protein levels fluctuate rhythmically each day. The two interacting feedback
loops keep time by regulating Clk, per and tim expression.
The mammalian clock is similar to the one described in Drosophila however, it is more
complex and involves multiple isoforms of some clock proteins. It is composed of two
transcriptional-translational feedback loops which regulate gene expression (Reviewed in
Reppert and Weaver, 2002). In the positive loop, BMAL1, a homolog of the Drosophila
CYCLE, and CLOCK transcriptional factors heterodimerize and upregulate the transcription
of three mPeriod (mPer) and two mCryptochrome (mCry) genes (Reviewed in Reppert and
Weaver, 2002). Following protein expression, the mCRY proteins act as negative
transcription factors that repress transcription by binding to the BMAL1-CLOCK
8
Figure 1.2: A schematic of the molecular clock in the central clock cells of Drosophila melanogaster.
The cycle begins when a CLOCK-CYCLE heterodimer binds to regulatory elements in the period (per) and
timeless (tim) genes, and activates their transcription. Once expressed, per and tim mRNAs translocate to the
cytoplasm where these are translated into proteins. A number of kinases regulate the stability of these proteins.
After TIM and PER proteins accumulate, they form a protein complex with the kinase, DOUBLETIME (DBT).
Subsequently, the TIM-PER-DBT complex represses per and tim transcription by binding to CLK-CYC. TIM,
PER and CLK are degraded via the proteasomal pathway and the clock restarts. CK2, CASEIN KINASE 2;
Sgg, SHAGGY; PP2a, PROTEIN PHOSPHATASE 2a. Figure source: Hardin (2005).
heterodimer. The negative transcriptional loop regulates Bmal1 transcription. The BMAL1-
CLOCK heterodimer activates transcription of Rev-Erbα, which encodes a transcription
factor that binds to a Rev-Erb/ROR regulatory site and represses Bmal1 transcription. The
interaction between these loops results in a 12 hour phase difference between the expression
of Bmal1 and mPer/mCry genes (Reviewed in Reppert and Weaver, 2002). Thus, the
9
mammalian clock regulates gene expression via two transcriptional/translational feedback
loops homologous to those found in flies.
cryptochrome-Mediated Entrainment of Circadian Clock Cells
Light is thought to act as the strongest zeitgeber in entraining the circadian clock (Reviewed
in Hardin, 2005). In flies, light entrainment occurs through three possible routes: external
photoreceptors in the compound eyes and possibly the ocelli, extraocular photoreceptors, and
a blue-light photoreceptor encoded by cryptochrome (cry) (Helfrich-Forster et al., 2001). As
a result, mutant flies which do not have external and extraocular eye structures and are
deficient in cry expression are unable to entrain molecular and behaviour cycles to light
cues, and instead display free-running rhythms (Helfrich-Forster et al., 2001).
The CRY photoreceptor mediates light input into neural and peripheral clock cells, however
its mode of action is not well known. Similar to other clock genes, cry transcription is under
circadian regulation. CRY protein levels are regulated by the light-dark cycle (Reviewed in
Hardin, 2005). When light enters a cell, it either causes an inhibitor to separate from CRY or
causes the protein to undergo a conformational change (Figure 1.3). This allows CRY to
interact with the TIMELESS protein and trigger its phosphorylation by a tyrosine kinase,
and subsequent proteasomal degradation (Reviewed in Hardin, 2005). It has been proposed
that the kinase SHAGGY interferes in this process by binding to CRY and preventing TIM
degradation (Reviewed in Dubruille and Emery, 2008). An F-box protein JETLAG (JET) is
also involved in upregulating TIM degradation, although it is unclear where it fits in the
scheme of events (Reviewed in Dubruille and Emery, 2008). Thus, TIM degradation via the
CRY photoreceptor leads to the daily resynchronization of the molecular clock.
10
It has been suggested that CRY also plays a direct role in entraining some peripheral
oscillators (Reviewed in Hardin, 2005). A missense mutation in cry was found to disrupt
rhythmicity in antennal clock cells whereas the central clock was undisturbed (Krishnan et
al., 2001). In constant darkness, cry mutant flies display disrupted olfactory response
rhythms. Upon measuring per-driven bioluminescence from the antennae of control and
mutant flies, it was discovered that the molecular clock in mutant antennae is less rhythmic
than control antennae (Krishnan et al., 2001). This mutation, however, does not affect
oscillator function in the brain, indicating that cry may have an additional function in
peripheral tissues (Krishnan et al., 2001; Stanewsky et al., 1998). It has also been shown that
the cry mutation affects the rhythmic expression of period and timeless in various clocks in
the fly body, confirming that cry may be a direct participant in these clocks (Levine et al.,
2002).
Figure 1.3: Daily resetting of the clock
involves a CRYPTOCHROME-
mediated pathway.
When light enters clock cells in the
brain, it activates a blue-light
photoreceptor called
CRYPTOCHROME (CRY) by
possibly triggering a conformational
change. CRY participates in the
phosphorylation of TIM‟s tyrosine
residue. The phosphorylated TIM
enters a pathway that involves the
JETLAG (JET) protein and is degraded
in the proteasome. Figure source:
Dubruille and Emery (2008).
11
Although daily light cues are required for synchronizing the molecular clock, constant light
has a detrimental effect on circadian clocks and behaviours. Constant low intensity light
affects the period of locomotor activity rhythms in wild type flies (Konopka et al., 1989).
Alternatively, high intensity light has an even more drastic effect as some flies become
completely arrhythmic (Konopka et al., 1989). cry appears to be involved in this response as
cry-knockout flies display normal locomotor activity and eclosion rhythms in constant light
(Dolezelova et al., 2007). The disruption of the molecular clock is thought to be a
consequence of continuous TIM degradation (Reviewed in Dubruille and Emery, 2008).
Thus, the CRY photoreceptor plays a key role in entraining the molecular clock to the solar
day.
Thesis Objectives
The aim of the thesis is to study the circadian regulation of feeding in adult Drosophila
melanogaster. Circadian clocks synchronize behaviour to the fluctuating environment
(Dunlap et al., 2004). In mammals, feeding behaviour is clock regulated and shows rhythmic
changes each day (Rosenwasser et al., 1981). Due to the parallels between the circadian
mechanisms of Drosophila and mammals, I hypothesize that clocks regulate feeding in fruit
flies. Feeding is directly related to the metabolic needs of the body. Since activity depletes
nutrients in the body, it is possible for feeding events to be organized around rest/activity
cycles. This suggests that metabolic tissues may contain circadian clocks which regulate the
timing of metabolic activities. In Chapter 2, I examine whether the fly fat body, which is
homologous to the mammalian hepatocyte and adipose tissues (Reviewed in Canavoso et al.,
2001), contains a circadian clock. I find that the core clock genes are circadianly expressed
in a period-dependent fashion. Additionally, I assess the temporal pattern of feeding in fruit
flies in Chapter three. I report that flies display circadian feeding rhythms, which are
regulated in part by the central clock. I propose that the fat body clock and other
metabolically related clocks may coordinate with the central clock to regulate feeding. In the
final Chapter, I present a model of the circadian regulation of feeding. This study sheds new
light on the regulation of a fundamental metabolic behaviour.
12
Chapter 2 . A Peripheral Clock in the Fly Fat Body
In mammals, peripheral clocks are known to reside in several of the primary metabolic
tissues involved in carbohydrate and lipid metabolism, including the liver and adipocytes
(Reviewed in Challet, 2010). Given that the fat body of the fly is generally considered to be
functionally homologous to these tissues (Reviewed in Canavoso et al., 2001), I hypothesize
that it too may contain a circadian clock. In this Chapter, I demonstrate that the core clock
genes are expressed in the fat body of male wildtype flies, and that the temporal expression
patterns of these genes under various light conditions and in a clock mutant fulfill the basic
criteria used to determine the existence of a circadian clock.
Introduction
In the past sixty years, extensive research examining the feeding behaviour of rodents has
allowed humans to gain a better understanding of this innate behaviour. In 1946, Brooks et
al. were the first to document the nocturnal feeding rhythm of rats. It was observed that in a
12:12 light-dark cycle, rats consume approximately 70% of their total daily food intake
during the night (Brooks et al., 1946). It was later reported that the feeding pattern of rats
consists of two peak feeding times occurring at the beginning and at the end of the dark
period (Rosenwasser et al., 1981). Although this originally appeared to be a direct response
to light cues, experiments conducted in constant darkness demonstrated that rats continue to
exhibit a feeding rhythm even in the absence of environmental cues, suggesting that the
behaviour is clock-regulated (Rosenwasser et al., 1983).
In mammals, it has yet to be determined which tissue specific clocks, or combination
thereof, are involved in regulating feeding. Previously, it was shown that nocturnal feeding
patterns are abolished in rats with hypothalamic lesions suggesting that the central clock,
which resides in the Suprachiasmatic Nuclei (SCN) of the hypothalamus, regulates feeding
*Xu et al. (2008) published a study about the fat body clock during this thesis.
13
(Brooks et al., 1946). However, given that the central clock is thought to send timing signals
to other clocks in the body it is possible that clocks in peripheral tissues may also be
involved in regulating the time of feeding (Yamazaki et al., 2000). Feeding is thought to be
regulated through the coordination between SCN clock and peripheral clocks.
Vertebrate studies suggest that clocks in the liver and adipose tissues, two of the primary
metabolic sites, may regulate food intake in mammals. There appears to be a link between
feeding and the hepatocyte clock, as restrictive feeding has a direct effect on the peripheral
clock. When rats were restrictively fed in the daytime of a light-dark cycle, the liver clock
synchronized to the feeding time, shifting 10 hours within two days (Stokkan et al., 2001)
whereas the timing of molecular clock in the SCN was unaffected. This result indicates that
there exists a connection between feeding and the liver clock that functions independently of
the SCN.
Adipocytes regulate feeding in mammals via a circadianly regulated hormone, leptin (Froy,
2007; Zhang et al., 1994). Leptin is produced primarily in fat tissues and the level of
expression is proportional to the amount of fat in the body (Reviewed in Friedman and
Halaas, 1998). When expressed, leptin is released into the blood plasma in a diurnal pattern,
with peak levels observed following lights-off (Kalsbeek et al., 2001). Leptin travels to the
brain, and binds to receptors in the hypothalamus, where it suppresses food intake
(Reviewed in Friedman and Halaas, 1998). The central clock appears to regulate leptin levels
as SCN lesioned rats do not show a diurnal pattern in plasma leptin levels (Kalsbeek et al.,
2001). This indicates that leptin may act as an indirect route through which the SCN
regulates feeding rhythms. Although the regulators of feeding remain unknown, it is
hypothesized that the mammalian liver and adipose clocks could be involved.
Since the circadian systems of mammals and Drosophila share a high degree of similarity, it
is possible that feeding behaviour in flies is also under circadian regulation. If this is the
case, it may be possible to determine the source of the feeding rhythm in flies as the brain
clock and peripheral clocks are thought to be independent circadian oscillators (Plautz et al.,
14
1997). A possible site for the feeding clock may be the fat body, the fly homologue of the
mammalian liver and adipose tissue (Reviewed in Canavoso et al., 2001).
The fat body is a metabolic tissue that stores lipids, proteins and glycogen (Reviewed in
Canavoso et al., 2001). Although it is present throughout the body, a majority of fat body
tissue is found in the abdomen (Figure 2.1) (Miller, 1950). In females, a number of fat body
cells also reside close to the ovary and play a role in reproduction (Miller, 1950). Fat body
tissue is present in the larval, pupal and adult stages of a fly. The larval fat body provides
energy during pupation and early adult stages, when the fly is unable to eat (Aguila et al.,
2007). Afterwards, cell death occurs and these cells are replaced de novo by the adult tissues
(Butterworth et al., 1988; Miller, 1950). Although the larval and adult fat body cells arise at
different developmental stages, they are thought to have similar functions.
One of the primary functions of the larval fat body is to regulate lipid storage and
mobilization. Lipid metabolism is a homologous process in mammals and fruit flies. The
main differences arise from the organs that are involved and their functions. In mammals,
lipids are processed predominantly by the liver and adipose tissues, whereas in Drosophila
Figure 2.1: The location of fat
body cells in the adult fruit fly.
Fat body tissue in the head, thorax
and abdomen are outlined in blue.
In the abdominal segment, fat
body cells lie as sheets of tissue
above the oenocytes which are
found in the segmental overlaps
(black circles). Figure source:
Miller (1950).
15
larvae, the fat body and larval oenocytes, a tissue that has been implicated in hydrocarbon
synthesis in insects (Fan et al., 2003), coordinate in metabolizing lipids (Gutierrez et al.,
2007). In a satiated larva, lipids accumulate predominantly in the fat body and midgut
epithelial cells. Once food is ingested, lipids are transported by lipophorin molecules from
the midgut to the fat body where they are converted from diacylglycerols (DAGs) to
triacylglycerols (TAGs) and stored (Reviewed in Canavoso et al., 2001). In starvation
conditions, lipids are transported from the fat body to the larval oenocytes (Gutierrez et al.,
2007). Following the movement, energy is mobilized from the oenocytes throughout the
body. In this way, the oenocyte and fat body tissue together act as the mammalian
hepatocyte and adipocyte. Although not demonstrated, the fat body of the adult fly is thought
to perform the identical function of its larval counterpart.
Here, I examine the expression of the core clock genes in the adult abdominal fat body using
quantitative PCR. I further investigate whether expression levels are altered in the fat body
of the clockgene-mutant, [ per0]. I report that clock gene expression oscillates in the fat body
in a period-dependent manner, suggesting that a clock resides in this tissue. A peripheral fat
body clock may act to synchronize the timing of metabolic processes, and could be involved
in influencing circadian behaviours.
Methods
Strains
Canton-S was used as the wild type strain for quantitative PCR experiments. Mutant strains
are described within the results.
Fly Stocks
Flies were raised in polypropylene food bottles (Fisher Scientific catalog no. AS-355) with
agar-based fly food. A detailed method for preparing fly food is provided in Appendix A.
Stock bottles were kept in an incubator with a 12:12 light-dark cycle (LD cycle). The
16
temperature and humidity in the incubator were maintained at approximately 25˚C and 40%,
respectively. Stock bottles were changed once a week and were thrown out after 16 days of
use. Food vials (9.4 cm narrow mouthed, polystyrene vials - Fisher Scientific catalog no.
AS-515), which were used to house flies prior to an experiment, also contained standard fly
food.
Fly Collections for Dissections
Flies were collected into food vials on the first day after eclosion (from 11-15 day old
bottles) and placed into an incubator. The next day male flies were sorted on a carbon-
dioxide anaesthetizing pad (CO2 pad) and females were discarded. Male flies that were
collected for mass dissections were ushered into fresh food vials. For timeseries dissections,
pairs of males were placed into individual 10x75mm glass culture tubes (VWR catalog no.
47729-568) filled with 1mL of fly food. Flies were kept in the incubator for a minimum of 3
days prior to dissections in order to entrain them to the light-dark cycle.
Fat Body Dissection Procedure
Fat body tissue samples were collected from 5-7 day old male flies. An individual fly was
first placed on a CO2 pad for 30-50 seconds to anaesthetize it. It was then transferred to a
dissecting plate where its legs were removed with forceps and it was pinned down with
tungsten pins inserted into its neck and genital areas. A 1mL glass pipette was then used to
cover the fly with liquid Shields and Sang M3 insect media (Sigma Aldrich catalog no.
S3652). Using forceps, a cut was made on the ventral side of the fly extending from the
thorax to the genitals. Forceps were used to remove the guts and to separate the thorax from
the abdomen. The cuticle was then pinned down on either side of the fly so that the inside of
the abdomen lay flat on the plate. Using a tungsten needle, fat body tissue was detached
from other abdominal tissues and suspended in a 1.5mL Eppendorf tube containing 1% ß-
mercapto-ethanol in RLT buffer. Tissue samples were stored in a -80˚C freezer. Prior to
running PCR reactions, RNA was first extracted from fat body tissue samples using
17
MinElute Spin columns (Qiagen RNeasy Micro Kit catalog no. 74004). It was reverse
transcribed to cDNA using the QUANTA cDNA synthesis kit (Quanta Biosciences catalog
no. 95047-100).
Genomic DNA Extraction Protocol
1-2 day old male flies were anaesthetized on a CO2 pad and collected in an Eppendorf tube.
To acquire genomic DNA, standard lab protocol was followed (Hamilton and Zinn, 1994).
To purify the DNA, an additional step involved extracting with a phenol, chloroform,
isoamyl alcohol solution (25:24:1). After centrifugation, the supernatant containing genomic
DNA was transferred into another tube. Using a phenol solution, a phase separation was
created and the genomic DNA was extracted into the aqueous phase.
Reverse Transcription Polymerase Chain Reactions and Agarose Gel
Electrophoresis
The gene products of cycle, Clock, timeless, period, cryptochrome and pigment dispersing
factor receptor were PCR amplified with Taq DNA polymerase (NEB catalog no. M0320L).
The positive PCR controls were set up with genomic DNA. The sequences for the primer
sets are the following: cyc F1: 5‟-GGA GCT GGA GGA CGT ATC G-3‟ and cyc R1: 5‟-
TCA AGA TGA TTA TCC TGC AAG-3‟; Clk F3: 5‟-GGA TAA GTC CAC GGT CCT
GA-3‟ and Clk R3: 5‟-CTC CAG CAT GAG GTG AGT GT-3‟; tim F11: 5‟-CCT ATG TGG
TCA ACC CGA AT-3‟ and tim R11: 5‟-TAC ATC ACG TCC ACG GAG AA-3‟; per F12:
5‟-GGT TGC TAC GTC CTT CTG GA-3‟ and per R12: 5‟-TGT GCC TCC TCC GAT
ATC TT-3‟; cry F1: 5‟-ATG TCG GGA GCT GAA TAT CG-3‟ and cry R1: 5‟-CAG GAA
GCC CAT GTT GTC TC-3‟; pdfr F1: 5‟-GCC ACG ACT AGC GGT CAT AC-3‟ and pdfr
R1: 5‟-TGG GTG GCC AGA CTC TTT AG-3‟. PCR was conducted in a thermal cycler
with the following settings: 1 cycle (94˚C for 2 min), 25 cycles (94˚C for 15s, 55˚C for 15s
and 72˚C for 45s), 1 cycle (72˚C for 5 min) and hold (4˚C). Following PCR, products were
run on a 1.2% agarose gel in a 1% Tris Acetate-EDTA buffer at a speed of 90V/hour. The
18
gel was stained in an Ethidium Bromide solution for approximately 25 minutes. It was
destained in the presence of magnesium sulphate (MgSO4) salt for 20 minutes and visualized
under UV radiation.
Quantitative PCR
In quantitative PCR (qPCR), the amount of DNA in a sample is estimated based on the level
of fluorescence emitted by a reporter molecule in the reaction mix (Stratagene., 2004).
SYBR Green is a commonly used reporter dye which fluoresces more brightly once it binds
to double stranded DNA. Thus, the level of fluorescence emitted by the reporter is directly
proportional to the amount of DNA. Real time qPCR is a sensitive technique where the
amount of DNA at the end of several PCR cycles is used to estimate the DNA quantity in the
original sample. The output of a qPCR reaction is a threshold cycle (CT) value, which is the
1st cycle where the fluorescence signal was greater than the background noise (Stratagene.,
2004). A reference dye called ROX is also used in the reaction mix to reduce differences
caused by factors such as pipetting and plastic transparency between adjacent samples.
qPCR reactions were conducted with fat body samples from time series dissection
experiments. In this experiment, 5or 8 flies were dissected in 2 hour intervals at 8 timepoints
(CT1, 4, 7, 10, 13, 16, 19 and 22) in a 12:12 light-dark cycle or on the first day of constant
darkness. As some timepoints were conducted in the dark period, glass culture tubes with
flies were wrapped in aluminum foil to prevent light exposure. The fat body samples from
two time series dissection experiments were combined to provide sufficient RNA for
quantitative PCR. After RNA extraction and reverse transcription duplicate or triplicate
qPCR reactions were prepared with the Quanta SYBR Green qPCR kit (Quanta Biosciences
catalog no. 95056-500). The relative RNA levels of ribosomal protein 49 (rp49), tim, per
and Clk genes were determined from fat body samples. The primer set used for amplifying
rp49 is the following: F1: 5‟-ATC GGT TAC GGA TCG AAC AA-3‟ and R1: 5‟-GAC
AAT CTC CTT GCG CTT CT-3‟. The primer sequences for tim, per and Clk were provided
earlier. qPCR was conducted in an Mx 3005P Sequence Detection System with the
19
following settings: 1 cycle (95˚C for 3 min), 40 cycles (95˚C for 30s, 60˚C for 1min, 72˚C
for 1min) and 1 cycle (95˚C for 1 min, 55˚C for 30s, 95˚C for 30s).
qPCR Analysis
After conducting a SYBR Green qPCR experiment, data was acquired from the Mx3005P
v.3.20 (Stratagene) program by converting the experiment to a comparative quantitation
experiment. The REST relative expression method was used to quantify the relative RNA
levels with rp49 as the normalizing gene (Pfaffl, 2001). Microsoft Excel 2007 was used to
conduct all calculations and plots were made in Sigmaplot v. 10.0.
Statistical Analysis
A non-linear regression analysis was performed using SPSS v.16.0 statistical software to fit
a cosine curve to the expression data with the following equation, y = a + (b*Cos (2π* (CT
hours - h))/d). The expression data from one timeseries dissection experiment was used for
curve fitting. Additional timeseries data is provided in Tables 1-4 in Appendix B. a, b, and h,
and d variables which were used to estimate cosine curve properties are provided in Tables 5
and 6 in Appendix C. a estimates the y-intercept of the curve while b and h represent the
amplitude and phase, respectively. d, which estimates the period, was constrained to 24
hours. 95% confidence intervals for b, which were provided in the statistical output, test the
null hypothesis b=0, and determine whether a cosine curve is significantly different from a
straight line. The r-squared values, which are also provided in the SPSS output, estimate
how well cosine curves fit the expression data.
Results
The Core Clock Genes are Expressed in the Abdominal Fat Body
In order to examine whether the abdominal fat body has a circadian clock, standard
molecular techniques were used to confirm that the core clock genes are expressed in this
20
tissue. First, abdominal fat body tissue was dissected from male wildtype flies. The
transcripts for cycle, Clock, timeless, period, cryptochrome and pigment dispersing factor
receptor were amplified from fat body cDNA via reverse transcription PCR. Genomic DNA
served as a positive control for the PCR reactions. The expression of all six clock genes was
Figure 2.2: Clock gene expression is detected in the abdominal fat body of male Canton-S flies.
cycle, Clock, timeless, period, cryptochrome and pigment dispersing factor receptor were amplified via real
time PCR from genomic DNA (G), and fat body cDNA (C) samples. Products were run on a 1.2% agarose gel
in standard 1% TAE buffer at a speed of 90V/hour. Arrowheads indicate the location of the predicted PCR
products. cDNA was acquired from dissected fat body samples of 5-7 day old male flies (n=38). Genomic
DNA was extracted from male wildtype flies (n=30) and served as the positive control.
observed in the abdominal fat body (Figure 2.2). Genomic bands are larger than fat body
cDNA bands because genomic DNA contains introns which are spliced out once the mRNA
is produced. I detected the expression of cyc, Clk, tim and per clock genes in the fat body
indicating the presence of a molecular clock. The presence of cry expression suggests that a
peripheral clock in the fat body may be directly entrained by light. pdfr expression was also
21
observed in the fat body, an indication that the fat body clock may interact with the central
clock via the neuropeptide pigment dispersing factor.
The Abdominal Fat Body Contains a Circadian Clock
The circadian clock is characterized by the cyclic expression of the timeless, period and
Clock genes, a feature thought to be an important part of the transcriptional-translational
feedback loops of the molecular clockworks. To determine whether a clock resides in the
abdominal fat body, this tissue was dissected at set time points occurring at 4 hour intervals
over a 24 hour period, and the relative expression level of each of the core clock genes was
quantified by quantitative PCR. In a light-dark cycle, the expression of timeless, period and
Clock genes is significantly cyclic in male Canton-S fat body cells (b, P-value<0.05; Figure
2.3A). Clk mRNA levels peak in the morning (~ZT3), whereas tim and per expression reach
their highest levels approximately 12 hours later. In constant darkness (Figure 2.3B), the
amplitude of the expression profiles for all three genes are statistically significant (b, P-
value<0.05). Together these results strongly support the existence of a peripheral clock in
the cells of the abdominal fat body.
The Fat Body Clock is Dependent on period Expression
The rhythmic oscillation in period expression is an integral part of the molecular
timekeeping mechanism of the circadian clock. To further study the fat body clock, I
examined whether clock gene expression is altered in the period-null mutant, per01
(in a
yellow white (y w) genetic background), a nonsense mutation in the third exon of the period
gene (Yu et al., 1987). In male y w genotype control flies, tim and per show significant
cycling in constant darkness (b, P-value<0.05, Figure 2.3C), whereas the amplitude for the
Clk cosine curve is not statistically significant (b, P-value>0.05). The levels of tim and per
mRNA peak around CT 17, similar to the time when the same genes peak in Canton-S flies
in DD. The expression profile of Clock, however, is reduced compared to the wild type
22
Figure 2.3: A period-dependent peripheral clock resides in the abdominal fat body.
timeless, period and Clock expression patterns in the fat body of male Canton-S, y w and per01
flies. In the CS
fat body, cyclic clock gene expression is detected in a 24 hour light-dark cycle (A) and in constant darkness
(B). y w flies show similar expression levels in constant darkness for per and tim genes (C). Comparatively,
per01
mutant flies show arrhythmic expression for all three genes in DD (D). Expression levels were quantified
via quantitative PCR. Cosine curves are fit to RNA expression levels from one experiment (n=1) ± SEM (see
Tables 1-4 in Appendix B for additional timeseries data). The white, grey and black horizontal bars underneath
plots represent the day, subjective day and night periods, respectively.
23
profile. The per[0] mutation disrupted the profile of clock gene expression (Figure 2.3D);
the amplitude of expression for all three genes is reduced and not significantly different from
a flat line (b, P-value>0.05). Together these results suggest that abdominal fat body cells in
Drosophila contain a period-dependent peripheral clock.
Discussion
The Abdominal Fat Body is a Peripheral Clock
I established that the cells of the abdominal fat body of D. melanogaster contain a circadian
clock. These cells express the core clock genes cycle, Clock, timeless and period as well as
the clock-related genes cryptochrome and pigment dispersing factor receptor. Furthermore, I
have demonstrated the existence of a functional fat body clock by illustrating the cyclic
expression of per, tim and Clk in a light-dark cycle and constant darkness. In wild type flies,
Clk mRNA peaked in the morning whereas tim and per levels peaked in the early evening.
The temporal expression profile of these genes in the fat body is consistent with that
previously observed in the adult head and abdomen (Hardin, 1994; Hardin et al., 1990). The
expression of the four main clock genes suggests that both the CLK-CYC and PER-TIM
feedback loops exist in the fat body (Reviewed in Hardin, 2005). Recently, it was reported
that the abdominal fat body contains a peripheral clock (Xu et al., 2008). Similar to my
results, it was shown that tim mRNA levels in the abdominal tissue preparation are highest in
the evening. Whereas I quantified clock gene levels in isolated fat body tissue, the authors of
the aforementioned study quantified gene expression from a dissected preparation which
included a mixture of several tissues associated with the abdominal cuticle (i.e. abdominal
fat body, epithelial cells, oenocytes and cardiac tissue). Several of these tissues including
the oenocytes and epidermal cells have been previously shown to be peripheral clocks (Ito et
al., 2008; Krupp et al., 2008), making the interpretation of their data impossible.
In addition, I have shown that the cyclic expression of all three clock genes is reduced or
completely absent in period-null mutants, confirming that the fat body clock is period-
dependent. In the absence of a functional PER protein, it is likely that both transcriptional
24
feedback loops are disrupted as they are interconnected, thus preventing the rhythmic
expression of all three clock genes (Reviewed in Hardin, 2005). Similarly, the Clkjrk
mutation, which causes dysfunctional CLK protein expression (Allada et al., 1998), was
previously shown to disrupt the rhythmic expression of timeless in the abdominal cuticle
preparation containing fat body tissue (Xu et al., 2008).
The fat body is known to store glycogen and lipids, two principal nutrient sources
(Reviewed in Canavoso et al., 2001). Adipokinetic hormone (AKH), a homolog of
mammalian glucagon, regulates the mobilization of trehalose and lipids from the fly fat body
into the hemolymph to provide usable energy (Lee and Park, 2004). Recently, it was found
that the corpora cardiaca, the tissue where AKH is produced and released, is also a circadian
clock (Personal communication with Ayesha Malik, Joshua Krupp and Joel Levine). While it
remains to be determined if the production or the release of AKH is under circadian
regulation, it is possible this hormone may act as a synchronizing signal from the clock in
the CC to the fat body clock. Tissues involved in lipid metabolism may time lipid
mobilization and breakdown around the fly‟s behavioural activity. Since lipids are thought to
move to the oenocyte from the fat body (Gutierrez et al., 2007), perhaps the oenocyte clock
is also involved in coordinating the timing of lipid mobilization. pigment dispersing factor or
other neuropeptides emanating from the central clock may synchronize the timing of clocks
in these tissues.
Future experiments for this project involve disrupting the fat body clock and determining its
role in regulating feeding rhythms. I plan to do this by expressing Clk-RNAi or cyc-RNAi
using a fat body driver. It is hypothesized that if the fat body clock is disrupted, the feeding
pattern will be attenuated in flies. Thus far, the only obstacle to this experiment has been
finding a proper fat body driver, expressed solely in the abdominal fat body cells. Several
larval fat body drivers have been studied, but most are not fat body specific in adults. The
two fat body drivers that can be used are the larval serum protein 2 (lsp2)-gal4 driver and
the r4-gal4 driver, both of which are expressed in a tissue-specific manner in adult
25
Drosophila (Dauwalder et al., 2002; Lee and Park, 2004). Alternatively, this tissue can be
ablated by expressing the pro-apoptotic gene, reaper, in the fat body.
26
Chapter 3 . Circadian Regulation of Fly Feeding
Feeding behaviour is strictly regulated to meet the metabolic demands of a physically active
animal. How feeding behaviour is synchronized with an animal‟s active state to meet its
metabolic requirements is not clear. In mammals, feeding rhythms are generally considered
to be under circadian regulation (Rosenwasser et al., 1981), however, the circadian system
that regulates feeding patterns, be it the clock in the SCN or a metabolic tissue like the liver,
has not been clearly identified. As a means to gain insight into the circadian processes
regulating feeding behaviour, I utilized the model organism Drosophila melanogaster. In
this Chapter, I examine the temporal organization of fly feeding behaviour, and demonstrate
(1) that feeding is under circadian regulation, and (2) that a peripheral clock, at least in part,
is involved in modulating feeding rhythms. I hypothesize that the circadian regulation of
feeding behaviour in fruit flies involves the coordination of the central clock in the brain and
one or more peripheral clocks residing in metabolic tissues.
Introduction
Rodents have been a useful model system with which to study the circadian regulation of
feeding behaviour. Rodents display a nocturnal feeding rhythm with peaks in the beginning
and end of the night (Rosenwasser et al., 1981). A rhythm in food intake appears to be
governed by both meal size and meal frequency; both parameters also show similar circadian
fluctuations. Interestingly, the circadian patterns of meal size and frequency exhibit slight
differences in peak phase, suggesting these two components of feeding may be regulated by
separate circadian systems (Rosenwasser et al., 1981). Together, this indicates that different
circadian clocks may regulate meal size and frequency separately, which together contribute
to the overall feeding rhythm.
*Xu et al. (2008) published a study about the circadian regulation of feeding during this thesis.
27
Similarly, in insects total food intake is also determined by meal size and meal frequency.
However, it is unclear if these parameters are under circadian regulation. In fruit flies, meal
size appears to be determined in part by the nutritional content of the food source. Flies
offered a sucrose-only solution consume significantly larger meals than flies offered a mixed
sucrose-yeast solution (Ja et al., 2007). Greater consumption of sucrose-only meals may be
necessary to provide adequate nutrients to satisfy hunger in the absence of the richer sucrose
yeast food, indicating that meal size is in part dependent on the nutrient content in food.
Meal frequency has been shown to be influenced by factors such as gender and the size of
the group in which individuals are housed. Females generally consume more food than
males by feeding more frequently (Wong et al., 2009). The difference in food consumption
is likely linked to the reproductive needs of females. Interestingly, group size also produces
an effect on feeding frequency. Flies housed in larger groups consume meals more
frequently than flies housed alone (Wong et al., 2009). Since meal size and frequency are
affected by different factors, they may be independently regulated perhaps through separate
mechanisms.
In insects, smell and taste are essential for feeding behaviour and food intake. Gustatory and
olfactory cues are used to direct a fly to a food source and to determine if the food is edible
(Reviewed in Melcher et al., 2007). Insects detect odorants via olfactory sensilla located on
the antenna and maxillary palps. Odorant molecules bind to receptors on the surface of the
olfactory receptor neurons in the sensillum (Reviewed in Dahanukar et al., 2005). If the
odour is perceived to be pleasant, the fly will extend its proboscis to taste the food (Dethier,
1976). Taste is perceived through gustatory receptor neurons in gustatory sensilla located on
the mouthparts, legs and wing margins (Reviewed in Dahanukar et al., 2005). In blowflies,
the stimulation of taste receptors is a major factor that regulates the amount of food
consumed (Gelperin and Dethier, 1967). For instance, when blowflies are offered a choice
between sorbitol, which is nutritious but only weakly stimulating, and fucose, which is non-
nutritious but very stimulating, they consume a greater quantity of fucose, suggesting that
28
the stimulating power of a food source can override the nutrition it provides (Gelperin and
Dethier, 1967).
As insects feed, digestive organs including the crop and foregut play a part in regulating
food intake. In blowflies, as food is consumed, the crop fills and expands in size, which is
thought to trigger stretch receptors in the body wall that limit further food consumption
(Dethier and Gelperin, 1967). The foregut also inhibits food intake via stretch receptors that
detect the amount of food passing through the digestive system. When the recurrent nerve
that connects the foregut stretch receptor to the brain is cut, it causes overeating, suggesting
that the foregut stretch receptor is important in the regulation of food intake (Dethier and
Gelperin, 1967). Due to their anatomical similarities, it is possible that such modes of
control also exist in other dipterans including Drosophila.
Feeding-related organs appear to communicate with each other through peptides released
from the brain and some metabolic tissues. In fruit flies, the peptide TAKEOUT (TO)
regulates food intake based on the nutrient levels in the body. takeout (to) shows increased
expression in response to starvation and appears to act as a signal for increasing food intake
after starvation (Meunier et al., 2007; Sarov-Blat et al., 2000). In to1 mutant flies, the
amount of food consumed after starvation is reduced compared to that of wild type flies.
This response is partly related to sugar sensitivity in the taste neurons. Normally, starvation
causes an increase in sugar sensitivity, which leads to increased feeding. In to1 mutants,
starvation does not induce a change in sugar sensitivity, thus flies may not be stimulated to
increase food intake (Meunier et al., 2007).
The short NEUROPEPTIDE F (sNPF) protein is involved in regulating food intake,
although the mechanism by which this occurs remains unclear. The localization of sNPF to
the medulla and the mushroom body calyx, a higher brain centre linked to olfaction
(Reviewed in Dahanukar et al., 2005), suggests it may act to regulate feeding based on
olfactory cues (Lee et al., 2004). Overexpression of sNPF in the central and peripheral
nervous systems of adults promotes feeding, whereas loss-of-function mutants are less
inclined to feed, suggesting that sNPF may increase the appetite (Lee et al., 2004).
29
HUGIN (HUG) is a neuropeptide that is involved in making a decision about whether or not
to consume a meal (Melcher and Pankratz, 2005). Mutant larvae that overexpress hugin
(hug) show an abnormal feeding phenotype; they stop feeding early and move away from a
food source. However, when hugin expressing neurons are inactivated with tetanus toxin,
feeding behaviour is rescued and larvae continue to feed (Melcher and Pankratz, 2005). In
larvae, hug expression is detected in the subesophageal ganglion, a gustatory information
processing center (Reviewed in Dahanukar et al., 2005), which suggests it regulates feeding
through taste (Melcher and Pankratz, 2005). In adults, hugin appears to be involved in the
initiation of feeding. Adult flies with blocked hugin neurons begin eating faster than control
flies, even if the meal contains an aversive substance (Melcher and Pankratz, 2005).
Together, these results suggest that hugin is involved in evaluating the nutritional content of
a potential meal and the decision about whether or not to consume it.
The foraging (for) gene, which encodes for a cGMP-dependent protein kinase (PKG),
regulates food intake and nutrient absorption. There are two natural alleles in foraging,
rovers (forR) and sitters (for
S) (de Belle et al., 1989). These two populations show
differences in nutrient storage and feeding, which appears to influence foraging behaviour.
In abundant food, rovers consume less food than sitters but show greater glucose absorption
(Kaun et al., 2007). Differences in food intake are reduced when larvae are kept on less
nutritious food, although carbohydrate absorption remains higher in rovers. These effects
seem to be caused by a difference in PKG expression; rovers display higher expression than
sitters (Osborne et al., 1997). foraging is also associated with feeding related behaviours in
adult flies, and may also regulate food intake at this stage of the life cycle (Kent et al.,
2009).
Feeding is an organized process that is regulated by the chemosensory organs, metabolic
tissues in the body, and higher brain centres. I propose that feeding in fruit flies is also under
circadian control. To examine fly feeding, I measure food intake in flies that carry mutations
in clock-regulated genes and report that fly feeding shows temporal regulation. Furthermore,
30
I demonstrate that the circadian system regulating feeding rhythms interacts with the
foraging gene, a gene known to affect food intake and feeding behaviour.
Methods
Strains
Canton-S was used as the wildtype strain for all behaviour experiments. Rovers, sitters and
sitter mutant flies were kindly provided by Marla Sokolowski. The strains used for different
feeding experiments are described within the results.
Fly Collections for Behavioural Assays
Fly stock bottles were emptied 12-24 hours prior to collections and eclosed flies were
collected into food vials. Male flies were sorted on a CO2 pad and placed into fresh food
vials 24 hours later. Flies were entrained to the light-dark cycle for 3-4 days.
The Capillary Feeder (CAFE) Assay
A modified version of the capillary feeder assay (Figure 3.1) developed by Ja et al. (2007)
was used to measure fly feeding. The fly chamber was made from a 9.4cm plastic vial. The
base of the vial contained 6mL of distilled water with immersed cotton whereas the top was
enclosed by a ~2cm thick sponge. The fly was provided liquid food through a 5μL pre-
calibrated microcapillary (VWR Catalog no. 53432-706). CAFE food was made by mixing
5% sucrose and 5% autolyzed yeast in distilled water. The solution was autoclaved and
stored in a glass bottle at room temperature. Prior to use, food was filter sterilized to remove
particles that could clog the microcapillary. Each microcapillary was first filled with a small
amount of mineral oil, which acts to reduce food evaporation. The microcapillary was
31
Figure 3.1: A diagram of the Capillary Feeder (CAFE) assay.
In this assay, an individual fly is kept in a humidified compartment and liquid 5% sucrose-5% yeast food is
provided through a pre-calibrated microcapillary. As the fly consumes food from the microcapillary, the oil
meniscus descends. The oil level is then measured with reference to a white marking on the capillary as
indicated (*).
subsequently filled with CAFE food via capillary action and cleaned well to remove excess
liquid on the outer surface. The food microcapillary was then fixed into pipette tips and
inserted into a premade hole in the sponge sliver at the top of the assay. An individual fly
32
was aspirated into the chamber, with a mouth aspirator. This was performed a minimum of
24 hours prior to the start of the experiment in order to acclimatize the fly to the
environment.
CAFE assays were monitored in a room where humidity was maintained between 30-50%
and the temperature was kept between 24-26˚C. Light conditions were maintained at 12:12
light-dark, constant darkness or constant light depending on the experiment requirements. To
visualize the capillaries in the dark, a lamp emitting dim red light was kept on during the
entire experiment. 2-3 control assays without flies were set up to measure the amount of
evaporated food. Assays were set up with cardboard barriers between vials to prevent
adjacent flies from observing each other‟s behaviour. Photographs of this assay were taken
in hourly intervals with a Canon PowerShot S5 IS camera using its „Remote Capture‟
software. This program allows timed photographs to be captured through a computer.
Analysis of Feeding
The amount of food consumed was determined by calculating the change in the level of the
oil meniscus in the microcapillary over time. Image J v. 1.40g was used to measure the oil
level to a reference line (a white marking on the capillary). A number of photographs were
stacked together and using the line segment tool measurements were made in pixel units.
Using Microsoft Excel 2007, the amount of food consumed per hour was calculated for each
fly. The amount of food evaporated was subtracted from hourly food consumption. A 3 point
moving average was calculated for each fly. Food intake was recorded for 26 or 50 hour
periods, so that a moving average could be calculated for the amount of food consumed at
the beginning and end of the experiment. The individual data for flies from the same strain
was then averaged and normalized. All plots were made in Sigmaplot v.10.0.
33
Statistical Analysis
All statistical analyses were carried out with SPSS v. 16.0. The moving average of food
intake was binned in 4 hour intervals starting from Zeitgeber/Circadian time 0. As multiple
feeding measurements were taken from the same sample (fly), a General Linear Model
(GLM) Repeated Measures test was used to compare differences in feeding amounts
between any two fly strains. The null hypothesis is that there are no differences in the
amount of food consumed between two groups. The GLM repeated measures test measures
the effect of genotype (between-subjects factor) and time (within-subjects factor) on feeding
differences between two groups. A significant genotypic effect indicates differences in the
total amount of food consumed between groups. GLM employs univariate tests to determine
the effects of the within-subject variables of time and genotype by time interaction. If the
assumption of sphericity tested with the Mauchly‟s test of sphericity was violated (Mauchly,
1940), the degrees of freedom value was adjusted with Greenhousse-Geisser correction
(Greenhouse and Geisser, 1959). A significant time effect indicates that the amount of food
consumed changes with time whereas a significant genotype-by-time effect suggests that
genotype affects feeding levels that change with time. The F-statistic and P-values for GLM
repeated measures tests performed on the feeding data are provided in Table 7 in Appendix
C.
Locomotor Activity Assay
The locomotor activity of individual flies was recorded in glass activity tubes containing
food. Activity tubes were prepared with ~1 inch of a 2% bactoagar and 4% sucrose food and
sealed on one side with parafilm. A single male fly was aspirated into the tube with a mouth
aspirator and the opening was closed with a small piece of cotton. The activity tubes were
monitored under the Drosophila Activity Monitoring (DAM2) system (TriKinetics, 2007-
2008) in an incubator. Locomotor activity experiments were conducted for approximately 2
weeks: a light-dark cycle was maintained for the first week, followed by a week in either
34
constant darkness or constant light. Activity data was acquired using the DAM System
Software v.303 and analyzed using MATLAB R2006b.
Results
Flies Display a Circadian Feeding Pattern
To study daily feeding patterns in Drosophila, I modified the capillary feeder (CAFE) assay
developed by Ja et al. (2007) (Figure 3.1). This simple set up allows for the real-time
quantification of the food consumed by individual flies. In the CAFE assay, a fly has access
to liquid food provided in a microcapillary; as the fly consumes food from the microcapillary
the level of food drops. From photographs taken in hourly intervals the amount of food
consumed can be calculated by measuring the change in food level over time.
To determine whether fly feeding shows a temporal pattern, I measured hourly food
consumption in male wild type flies during a 24 hour light-dark cycle. Average food
consumption was in the range of 0.38 - 370.96nL per hour. The hourly feeding profiles of
individual flies (Figure 3.2A) clearly show that feeding events occur at most times of the
day. However, the total amount of food consumed at certain hours of the day is larger than at
others, with increased food consumption occurring at times corresponding to the transitions
in lighting conditions. The feeding profile of a typical individual wild type fly, as shown in
Figure 3.2A, illustrates that there is greater food consumption following lights-on (ZT0) and
prior to lights-off (ZT12). The 3 hour moving average plot for wild type Canton-S flies
(Figure 3.2B) shows a similar pattern. Here, three peaks appear during the day, with two
peaks that occur around lights on and lights off. The broader feeding peaks visible in the
average plot are due to some variability in the times at which flies increased their feeding.
Two short feeding troughs appear during the day, while a large trough is observed in the
dark period. Overall there is a difference in daytime and nighttime feeding; the total amount
of food consumed during the day is 574.028±78.460nL compared to 417.529±56.419nL
consumed during the night. Together these results indicate that fly feeding shows a circadian
pattern.
35
Figure 3.2: Wildtype flies show a temporal feeding pattern in a light-dark cycle and constant darkness.
The feeding pattern of Canton-S flies in 24 hours of LD (plotted twice) and 48 hours of DD as measured by
CAFE assays. In LD, the feeding profiles of an individual fly (A) and the 3-point moving average (B, n=23)
show increased feeding around lights-on and lights-off. Food intake at night appears reduced compared to
daytime feeding. In constant darkness, the greatest amount of feeding occurs during each subjective day in the
individual (C) and moving average (D, n=14) plots. Actograms (double-plotted) display average locomotor
activity rhythms of CS flies (n=27) in 4 days of LD (E) and 4 days of constant darkness (F). The moving
average of feeding values ± SEM is plotted (B. D). White, grey and black horizontal bars indicate day,
subjective day and night, respectively.
The feeding of wild type Canton-S flies was examined in constant darkness in order to
determine if this feeding pattern persists in the absence of light cues. In the feeding plot of a
typical individual fly (Figure 3.2C), there is an increase in feeding in the early hours of the
first subjective day and also around CT12 and CT36, times when lights would turn off in a
36
LD cycle. Between these peaks, there is an overall reduction in the amount of food
consumed. Although the pattern is similar to the one seen in LD, the phase of peak feeding
times appears shifted in DD. When the moving average is examined (Figure 3.2D), two
feeding peaks are visible on the first subjective day in DD, around CT5 and CT11 followed
by a large feeding trough that lasts until the middle of the second subjective day. Another
broad peak appears around CT36 followed by a second feeding trough. The continuation of
the temporal pattern suggests that feeding is under clock regulation and independent of light
cues.
The temporal pattern in feeding appears to parallel the bimodal pattern in locomotor activity
(Figure 3.2E-F). In a light-dark cycle, peaks in locomotor activity are visible at dawn and
dusk each day corresponding to the increase in feeding around these times. Similar to
locomotor activity, feeding appears to anticipate lights-on and lights-off as the level of
feeding increases prior to fluctuations in lighting. In constant darkness, locomotor activity
and feeding show similar patterns with one broad peak around the end of the subjective day.
The similarity in the two patterns suggests an interaction in mechanisms regulating feeding
and locomotor activity.
The Feeding Pattern is Regulated by a period-Dependent Clock
The period gene is vital for maintaining circadian rhythms in behaviour. To investigate
whether feeding is under circadian regulation, I quantified feeding rhythms in period-null
mutant flies. per01
flies display arrhythmic locomotor activity in constant conditions
(Konopka and Benzer, 1971). In LD, the moving average plot for y w control flies (Figure
3.3A) shows a pattern with notable peaks after the onset of light and darkness; the latter peak
appearing larger than the former one. Comparatively, wild type Canton-S flies display a third
37
Figure 3.3: Feeding patterns appear disrupted in period mutant flies.
The moving average of feeding for y w control and per01
null-mutants in a 24-hour light-dark cycle (plotted
twice) and 2 days of constant darkness. In LD, y w flies (A, n=19) display a feeding pattern, with increased
feeding around light-to-dark and dark-to-light transitions. In the per01
moving average (B, n=17), a solitary
feeding peak appears several hours after lights-off. In constant darkness, a feeding pattern is visible in y w flies
(C, n=12) but the feeding pattern of mutant flies (D, n=24) appears disrupted. Actograms (double-plotted)
display average locomotor activity patterns in 3 days of LD and 2 days of DD. y w flies (E, n=30) display
rhythmic locomotor activity patterns in both lighting conditions whereas per01
flies (F, n=25) display
arrhythmic activity in DD. The moving average of feeding values ± SEM is plotted. White, grey and black
horizontal bars indicate day, subjective day and night, respectively.
feeding peak at midday, when a trough appears in y w flies. The results show that only time
has a significant effect on feeding differences between y w flies and wild type flies (time,
F3.24/41=6.90, P<0.005; genotype, F1/41=0.27, P=0.61; genotype by time, F3.24/41=0.58,
P=0.64). period-null mutants (Figure 3.3B) display a feeding pattern that is significantly
38
different from y w controls (genotype, F1/35=5.65, P<0.05; time, F3.58/35=4.62, P<0.005).
per01
flies consume a constant amount of food during the lights-on period; however after
lights-off, feeding does increase and is followed by a trough. In y w flies the evening peak
begins prior to lights-off whereas in per01
flies, it appears following lights-off, which is an
indication that per01
flies do not anticipate lights-off. The feeding profiles of per01
and y w
flies suggests that period-dependent clocks regulate feeding rhythms.
In constant darkness, y w flies continue to show a temporal feeding pattern whereas per01
flies display disrupted feeding. The moving average profile of y w flies (Figure 3.3C) shows
an increase in feeding at subjective dawn and dusk each day. The exception to this pattern is
a feeding peak that occurs in the middle of the first night instead of appearing on the early
subjective morning of the second day in DD. An extended trough in feeding appears to
follow this peak. Peaks are also seen on the second day but appear weaker than the previous
day. The feeding plot of per01
flies (Figure 3.3D) displays a strong feeding peak at the
beginning of the first day in DD. Subsequently, a greater number of minor peaks appear until
the end of the second day. This pattern neither resembles y w flies in DD nor the feeding
pattern of per01
flies in a light-dark cycle, which suggests period is required for normal
feeding rhythms. Time has a significant effect (F7.01/35=2.12, P<0.05) on feeding differences
between y w and per01
flies in DD. per-expressing clock cells are found in the brain and
peripheral tissues, this raises the question as to which clock cells regulate feeding.
Peripheral Clocks Regulate the Temporal Pattern of Food Intake
In order to examine whether peripheral or central clock cells are responsible for regulating
food consumption, I measured food intake in the per01
7.2:2 fly strain (in a ry
506 genotypic
background). In these flies, a functional period gene is only expressed in the central nervous
system, thus central clocks are active while peripheral oscillators are disrupted (Zehring et
al., 1984). In a light-dark cycle, the feeding pattern of per01
7.2:2 flies (Figure 3.4B) is
similar to the one observed in the ry506
control flies (Figure 3.4A), although time
(F3.31/30=8.93, P<0.001) and genotype (F1/30=8.64, P<0.01) have significant effects on
39
differences in the amount of food consumed. Both control and experimental flies display a
feeding peak at dawn and dusk with a single trough before and after the evening peak. The
two feeding profiles also show a small third peak between the morning and evening peaks,
which is more distinct in per01
7.2:2 flies.
In constant darkness, a pattern similar to the LD feeding profile is still visible in the ry506
flies, whereas the feeding pattern of per01
7.2:2 flies changes after the first day in DD. ry506
flies (Figure 3.4C) continue to increase feeding around dawn and dusk each day,
corresponding to times when lights would turn on or off. per01
7.2:2 flies (Figure 3.4D) show
a pattern on the first day of DD, which is similar to the one in LD; however, the frequency of
feeding peaks increases in the beginning of the second day. The results show that time
significantly affects the amount of food consumed by ry506
and per01
7.2:2 flies (F6.47/27=4.21,
P<0.001). This suggests that input from a peripheral clock is required to maintain feeding
rhythms in DD.
The synchronization of oscillators in the fly brain occurs via PDF and is essential for
retaining locomotor activity rhythms (Reviewed in Stanewsky, 2002). Since feeding and
locomotor activity show several similarities, perhaps clocks that regulate feeding also
communicate via the PDF neuropeptide. To investigate whether pdf is required for
maintaining the temporal pattern in feeding, I measured feeding in pdf-null mutants (in a
Canton-S genotypic background); pdf01
is a null allele created by a nonsense mutation (Renn
et al., 1999). Under light-dark conditions, pdf01
mutant flies (Figure 3.4E) show a single
peak in the morning followed by a combination of three smaller peaks which occur around
the light-to-dark transition. These flies appear to feed in a similar manner to Canton-S
control flies (Figure 3.2B) during the day; however the nighttime Canton-S feeding trough is
less obvious in the pdf01
feeding plot. There is a significant effect of time on feeding
(F3.19/40=3.14, P<0.05).
40
Figure 3.4: Neuronal clocks show some involvement in regulating the feeding rhythm.
The moving average feeding profiles of ry506
, per01
7.2:2 and pdf01
flies in 1 LD day (plotted twice) and 2 days
of DD. In a light-dark cycle, the feeding pattern of ry506
control (A, n=17) and per01
7.2:2 (B, n=14) flies
appears similar. Constant conditions do not alter the feeding pattern of ry506
(C, n=17) flies but seem to affect
41
the feeding pattern of per01
7.2:2 flies (D, n=11) by the second day. A feeding pattern is visible in the pdf01
profile in LD (E, n=18), but peak times are not very consistent in DD (F, n=19). However, both LD and DD
patterns are similar to the Canton-S control flies (Figure 3.2). Average locomotor activity rhythms in 3 days of
LD and 2 days of DD are shown in double-plotted actograms. ry506
(G, n=25) and per01
7.2:2 flies (H, n=28)
display rhythmic locomotor activity patterns in LD and DD. The actogram for pdf01
flies (I, n=26) shows
rhythmic locomotor activity in LD; however in DD this pattern appears disrupted by the 2nd
day in DD. The
moving average of feeding values ± SEM is plotted. White, grey and black horizontal bars indicate day,
subjective day and night, respectively.
In constant conditions, a feeding pattern is visible in pdf01
flies (Figure 3.4F) but is different
from its own LD profile. On the first day of DD, the pdf01
flies show a morning peak several
hours after the start of the subjective day. This is followed by a broad evening peak and
another peak a few hours later, occurring prior to the morning of the subsequent day. Finally,
two peaks occur around CT36 and CT48, times when there would be a transition in lighting
in a LD cycle. Although mutant flies continue to display fluctuations in constant conditions,
it seems that peak feeding times are affected by the pdf mutation. However, the DD pattern
is similar to the one observed for wildtype control flies in DD (Figure 3.2D). The results
show that although time (F5.14/32=1.67, P>0.05) does not have a significant effect on feeding
differences, genotype affects the total amount of food consumed (F1/32=8.39, P<0.01). This
suggests that pdf plays a role in the regulation of the temporal pattern of feeding.
Light-Entrainment of the Circadian Clock is Essential for Maintaining
Feeding Rhythms
In flies, the molecular clock synchronizes to light each day to maintain behavioural rhythms.
The blue-light photoreceptor Cryptochrome (CRY) is integral to the synchronization of
clocks to light cues (Reviewed in Hardin, 2005). To determine whether a cry mutation
affects the temporal feeding pattern, I examined the feeding pattern of cry-null mutant flies
(a w1118
genotypic background) in a light-dark cycle and constant darkness. In the LD
feeding profile of w1118
control flies (Figure 3.5A), a morning peak and two afternoon peaks
appear, with two peaks corresponding to the lights-on and lights-off times. The feeding
profile of cry0 mutants (Figure 3.5B) shows a small peak at dawn and a large peak at dusk.
The results show that genotype does not have a significant difference (F1/43=0.01, P=0.94)
42
on feeding levels in LD although the effect of time is significant (F2.99/43=4.30, P<0.01). In
constant darkness, the feeding pattern of w1118
flies (Figure 3.5C) appears to be disrupted as
flies only show increased feeding on the first day. It remains to be understood why w1118
flies do not show a feeding pattern in constant darkness. cry0 flies (Figure 3.5D) show a clear
feeding pattern with a combination of a narrow and broad peak, on both days in DD. Feeding
troughs are observed between peaks on both days, similar to the trough observed in LD.
Time has a significant effect (F6.19/30=5.14, P<0.01) on the feeding differences between w1118
and cry0 flies. Thus, a mutation in cry does not seem to disrupt the fly feeding pattern.
Constant light disrupts the locomotor activity rhythms of fruit flies (Konopka et al., 1989).
This effect is thought to be caused by constantly activating CRY, which leads to TIM
degradation (Reviewed in Dubruille and Emery, 2008). I determined whether feeding
patterns are also affected by constant light. To control for the effects of constant light of
feeding, Canton-S and y w flies were examined. The Canton-S feeding profile in LL (Figure
3.5E) shows a solitary peak on the first morning of the experiment followed by a constant
level of feeding until CT32 when feeding once again increases. For an unknown reason, the
afternoon peaks which were observed in LD, are not visible on the first day of LL.
Compared to the patterns observed in LD and DD (Figure 3.2B, D), it appears that feeding is
disrupted in constant light. Moreover, y w flies display a feeding pattern (Figure 3.5F) on the
first day in LL, but feeding becomes abnormal soon after. On the first day of LL, three peaks
appear which correspond with times when there would be a transition in lighting if it were
LD. Subsequently, small feeding peaks occur randomly throughout the second day. Thus, it
appears that constant light disrupts normal fly feeding patterns.
Furthermore, the feeding profile of w1118
control flies (Figure 3.5G) shows a disrupted
pattern in constant light; only one large feeding peak occurs at CT12 on the first day after
which the level of feeding remains constant. Comparatively, the cry0 feeding pattern in LL
43
Figure 3.5: Constant light disrupts normal feeding patterns.
The temporal feeding pattern of control and mutant flies in a 24-hour light-dark cycle (plotted twice), 2 days of
constant darkness and constant light. w1118
control (A, n=24) and cry0 mutant (B, n=20) flies display a feeding
pattern in LD; with peaks at dawn and dusk. However, in constant darkness, w1118
fly feeding (C, n=12) appears
disrupted while mutants (D, n=19) appear to feed rhythmically. The feeding patterns of CS (E, n=18), y w (F,
44
n=15), and w1118
(G, n=11) flies are diminished in constant light; although cry0 flies (H, n=17) continue to
display a feeding pattern. Double-plotted actograms show average locomotor activity rhythms of w1118
and cry0
flies in 3 days of LD and 2 days of LL. w1118
flies (I, n=24) become arrhythmic in constant light whereas cry0
mutants (J, n=19) retain rhythmic activity. The moving average of feeding values ± SEM is plotted. White and
black horizontal bars indicate day and night, respectively. Grey bars and diagonally patterned bars indicate
subjective day and subjective night, respectively.
(Figure 3.5H) is similar to its LD feeding profile. On the first day of constant light, a small
peak appears at dawn followed by two feeding peaks in the evening (before and after CT12).
This is in contrast to the single large peak that occurs in the evening in LD. On the following
day, a single large feeding peak is visible around CT36; indicating the persistence of a
feeding pattern. Feeding differences between cry-null flies and w1118
flies in LL appear to be
caused by time (F6.10/27=6.04, P<0.001). Constant illumination disrupts the molecular clock
through a CRY-mediated pathway, thereby interfering with the circadian regulation of
feeding.
The Polymorphism in foraging Affects the Circadian Regulation of
Feeding
The foraging gene is thought to link feeding-related behaviours to metabolic processes in
Drosophila (Reviewed in Kaun and Sokolowski, 2009). In order to determine whether
metabolic activities within the fly body direct the time of feeding, I examined the effects of
allelic differences in foraging expression on feeding patterns by quantifying food intake in
rovers (forR) and sitters (for
s). In addition, I measured food intake in sitter mutants (for
s2), a
rover strain that expresses the sitter allele (de Belle et al., 1989; Pereira and Sokolowski,
1993). Under light-dark conditions, rovers display feeding patterns that differ from those of
natural and mutant sitter flies. forR flies (Figure 3.6A) display a temporal pattern with three
feeding peaks during the day followed by a feeding trough during the night. This pattern is
distinct from the feeding patterns of sitter flies. The feeding profile of fors flies (Figure 3.6B)
shows two large feeding peaks after dawn and around ZT12, when lights turn off. The results
indicate that a genotype-by-time interaction has a significant effect on feeding differences
45
Figure 3.6: Allelic variation in the foraging gene affects the temporal feeding pattern.
The moving average of feeding for rovers, natural sitters and sitter mutants in one day of a light-dark cycle
(plotted twice). The feeding patterns of forR (A, n=8) appear different from the feeding profiles of for
s (B, n=8)
and fors2
(C, n=6) flies. The 3-point moving average of feeding values ± SEM is plotted. White and black
horizontal bars indicate day and night, respectively.
between rovers and natural sitters (F3.04/15=4.16, P<0.05). These feeding differences are
likely to be linked to the foraging gene, as sitter mutants (Figure 3.6C) display similar
patterns to natural sitters. fors2
flies show greater feeding around dawn and dusk, although
they display a larger evening peak whereas natural sitters show a larger morning peak.
However, time does have a significant effect on the amount of food consumed by natural and
mutant sitters (F2.31/13=10.18, P<0.001). The feeding pattern of fors2
flies is significantly
different from forR flies, an effect of a genotype-by-time interaction (F5/13=4.09, P<0.005).
The foraging-allele specific differences in feeding patterns suggest that PKG may interact
with the circadian system regulating food intake.
46
Analyzing Fly Meals
Mammalian studies have shown that rhythms in feeding frequency and meal size contribute
to the temporal pattern in total feeding (Rosenwasser et al., 1981). To further analyze the
circadian regulation of feeding, I categorized hourly meals of each fly and evaluated the
number of flies that consumed each meal type. Meals were grouped into small, medium or
large meal categories based on two threshold levels. The higher threshold level was the
average amount of food consumed per hour in a 24-hour period while the lower threshold
level was half of the average amount of food consumed. Meals that were below the half
average threshold were considered small. Medium meals were any meals that were in
between the half average and average threshold levels. And large meals were meals that
were larger than the average threshold. For each strain the number of small, medium and
large hourly meals were summed and plotted as frequency against time.
In LD and DD, a temporal pattern is visible in the frequency of large meals in Canton-S
(Figure 3.7A-B) and y w flies (Figure 3.7C-D). A greater number of flies consume large
meals around the light-to-dark and dark-to-light transition times in LD. Similarly, in DD, the
frequency of flies consuming large meals increases around the times when lights would turn
on or off in a light-dark cycle. These patterns resemble the temporal feeding patterns shown
earlier (CS, Figures 3.2 B, D and y w, Figures 3.3 A, C); with peaks occurring at the same
time in both plots. In contrast, the frequency pattern for small and medium meals seems to
occur out of phase with the large meal pattern. A pattern is visible in the frequency of per01
flies that consume large meals in LD (Figure 3.7E); however, in DD (Figure 3.7F) the
pattern appears disrupted. This is consistent with the temporal feeding patterns observed for
per01
flies earlier (Figures 3.3 B, D). The statistics for this data is yet to be resolved. Greater
consumption of food at particular times of the day may result from flies either increasing the
meal size or meal frequency.
47
Figure 3.7: A pattern in the consumption of large meals appears to drive the temporal feeding pattern.
The frequency of Canton-S, y w and per01
flies that consume small, medium and large hourly meals in a light-
dark cycle (plotted twice) and constant darkness. In CS (A-B) and y w (C-D) plots, the number of flies
consuming large meals increases around the light-to-dark and dark-to-light transitions. This pattern resembles
the temporal pattern in feeding. Patterns in the small and medium meal profiles appear to be shifted by a few
hours from the large meal pattern. In LD, per01
mutant flies (E) show a pattern in the number of small, medium
and large meals consumed but this is not visible in constant darkness (F). A small meal was any meal that was
smaller than half the average hourly meal. A large meal was larger than the average hourly meal. And medium
meals were any meals that fell between the small and large meal categories. Grey bars, small meals; blue bars,
medium meals; black bars, large meals. White, grey and black horizontal bars indicate day, subjective day and
night, respectively.
48
Discussion
Endogenous Clocks Regulate the Temporal Feeding Pattern
I have determined that feeding is under circadian control in Drosophila. Male wild type flies
consume food at all times of the day; however there are large bouts in feeding around the
light-to-dark and dark-to-light transition times. It appears that the increase in food
consumption during the day may be due to an increase in the feeding frequency. Female
Dahomey flies consume more food during the lights-on period by feeding more frequently
(Wong et al., 2009). When feeding was averaged, similar observations were visible for the
entire population. In addition, average feeding reveals that wildtype flies feed more in the
day than at night. This confirms earlier observations that 1 week old Canton-S flies consume
two-thirds of their daily food intake during the light period of a 12:12 light-dark cycle (Ja et
al., 2007). Thus, fruit flies show day-night fluctuations in food consumption in a light-dark
cycle.
The temporal feeding pattern persists in constant darkness, and appears not to be driven by
the light-dark cycle, suggesting it is under circadian regulation. Since feeding was observed
for only 24 or 48-hour time periods, the period of the feeding pattern could not be calculated.
Recently, a spectrophotometric method was used to demonstrate the circadian regulation of
fly feeding (Xu et al., 2008). In comparison, the reported feeding patterns are different from
those shown here. Xu et al. found that w1118
and y w flies display a LD feeding profile with a
peak in the early morning and a trough at night. I found that both fly strains display feeding
peaks around lights-on and lights-off. The discrepancy in the results may be attributed to
different methods. While Xu et al. measured feeding in groups of flies during two hour
timepoints; I measured the food intake of individual flies monitored for the entirety of the
experiment.
Additionally, I demonstrated that a period-dependent clock regulates the feeding pattern. It
is evident from the feeding plots of period mutants that a pattern is present in LD but is not
maintained in constant darkness. Comparatively, y w flies display a feeding pattern in both
conditions, although the feeding pattern becomes somewhat more variable after the first day
49
in constant darkness. This could be attributed to the yellow and white mutations, which affect
cuticle and eye pigmentation, respectively, and were originally used to map the period gene
(Konopka and Benzer, 1971). In future experiments, feeding assays should be conducted
with per01
flies in a wildtype background, to see if the yellow and white mutations may have
contributed to a disruption in DD feeding. Perhaps the white gene affects feeding in constant
conditions, as w1118
flies also display an altered feeding pattern on the second day of DD.
The feeding patterns of per01
flies resemble their locomotor activity rhythms, which are
altered but still retained in LD but abolished in DD. Although it has been suggested that light
drives LD rhythms in per01
flies, there is some evidence that these mutants retain a short-
period oscillator that allows them to entrain to LD-cycles (Helfrich-Forster, 2001). Similarly,
it was previously shown that the Clkjrk
mutant display disrupted feeding patterns in constant
darkness (Xu et al., 2008). Together, these results suggest that the Clk and per-tim loops of
the circadian clock are involved in regulating feeding patterns.
The central clock previously had not been linked to the temporal regulation of feeding; there
is evidence that it is involved in the regulation of the starvation response and in modulating
glycogen stores. Flies with disrupted neuronal clocks have higher levels of glycogen storage
and are more resistant to starvation (Xu et al., 2008). To determine if per-dependent clocks
in either brain or peripheral tissues or both regulate the temporal feeding pattern, I examined
the feeding pattern of per01
7.2:2 flies, which retain only central clock function (Zehring et
al., 1984). I found that per01
7.2:2 flies show a temporal feeding pattern in LD similar to that
of ry506
controls. In DD, however, per01
7.2:2 flies display an altered feeding pattern after the
first day of constant darkness. This indicated that clock cells in other tissues are, at least in
part, required to regulate the circadian timing of feeding. The senses of taste and smell are
intimately involved in regulating feeding, and the sensory cells of the both gustatory and
olfactory systems contain clocks, which regulate circadian changes in sensitivity (Chatterjee
et al., 2010; Krishnan et al., 1999). The olfactory and gustatory clocks are thought to
influence feeding by regulating chemosensory sensitivity to tastants and odours. It remains
to be determined which peripheral clocks contribute to the feeding pattern.
50
Clocks in the brain and metabolic tissues need to be synchronized in order to regulate the
time of feeding. As the fly‟s activity changes rhythmically, it is likely that the metabolic
needs of the fly fluctuate in phase with activity/rest cycles. Tissues such as the fat body need
to be in sync with activity rhythms so that energy is mobilized in response to the metabolic
needs of the body. After a period of activity, nutrients must be replenished by feeding
events. The receptor for pdf (pdfR) is expressed in the abdominal fat body, indicating that the
PDF neuropeptide, which is produced in lateral neurons of the central clock, may interact
with the fat body clock to affect its timing or function. Feeding was measured in pdf01
mutant flies and the feeding pattern was found to be similar to the CS feeding pattern in LD.
However, the timing of feeding peaks of pdf01
flies in DD appears shifted compared to their
own LD profile. pdf01
flies also display altered locomotor activity rhythms in constant
darkness, many becoming completely arrhythmic after three days (Renn et al., 1999).
Although only 24% of pdf01
flies display rhythmic locomotor activity patterns in DD, I
observed that 74% of pdf01
flies show rhythmic feeding patterns, which suggests that the
cellular basis of feeding and locomotion are different. It is possible that PDF may be
involved in regulating the feeding pattern by synchronizing clocks in feeding-related tissues
such as the fat body and the oenocyte. The feeding-related neuropeptide, TAKEOUT, could
be involved in the communication between different feeding-related clocks. takeout is a
circadian gene that is expressed in feeding-related tissues including the head fat body, cardia,
crop and antennae (Dauwalder et al., 2002; Sarov-Blat et al., 2000; So et al., 2000). In
addition to circadian regulation, its expression fluctuates under starvation, which indicates it
may convey information about the feeding status to stimulate food intake.
Continuous Illumination Disrupts Behavioural Rhythms
Constant light is known to disrupt circadian rhythms (Reviewed in Dubruille and Emery,
2008). Although in nature organisms normally only encounter a day-night cycle, constant
light tests are conducted to further understand how light input affects the circadian clock. In
mammals, the light-dark differences in feeding are abolished by constant illumination but
reappear when a light-dark cycle is reinstated (Zucker, 1971). In the present study, the
51
effects of constant light on feeding rhythms were studied to determine the role of
cryptochrome in entraining the clocks that regulate feeding. Constant light disrupts the
feeding pattern of wildtype flies with a functional cry photoreceptor although the effects
seem to be strain-dependent. Whereas feeding is disrupted from the first day in wildtype
flies, y w and w1118
genotype control strains display feeding patterns for at least one day in
LL. Perhaps, the effect of light would be more apparent if feeding had been observed for a
longer time in constant light. cry0 mutants however continue to display a rhythm in feeding
on both days of LL. This suggests that the clocks which regulate feeding entrain to light
through a cry-mediated pathway. As cry expression was detected in the fat body, the fat
body clock may directly entrain to light, not requiring input from the central clock. In
blowflies, constant illumination causes an increase in sugar consumption, which is attributed
to a reduction in periods of inactivity, as flies remain continuously active in constant light
(Dethier, 1976). Perhaps the lack of inactive periods in LL is a contributing factor in the
aberrant feeding patterns of fruit flies.
The Relationship between Feeding and Locomotor Activity
The feeding pattern documented in this thesis strongly resembles the rhythm in daily
locomotor activity. Feeding peaks correlate with the light-to-dark and dark-to-light transition
times, times when there are also increases in locomotor activity. The similarity of these
patterns suggests that they result from a common control mechanism or possibly that the two
behaviours influence one another. Feeding and activity have a reciprocal relationship as the
former increases nutrient levels in the body while the latter depletes them.
The takeout gene, may be a link between feeding status and locomotor activity. In wildtype
flies, there is an increase in TO expression in response to starvation. Under starvation
conditions, flies with reduced takeout expression show abnormal locomotor activity rhythms
compared to wildtype flies, which suggests that activity may be modulated based on feeding
status information conveyed by TO (Sarov-Blat et al., 2000). The relationship between these
behaviours has also been observed in the blow fly P. regina: as time after food deprivation
52
increases, there is an increase in the number and the duration of active periods (Green,
1964a). Furthermore, as the percent of sugar in the crop decreases, the percent of time active
increases (Green, 1964b). These studies indicate that as the crop empties the fly becomes
hungry, and as the hunger level increases, the fly becomes more active as it begins to forage.
To further understand this relationship, the locomotor activity and feeding behaviours of
flies should be simultaneously observed and measured.
The Role of the foraging gene in Feeding Rhythms
The link between foraging expression and feeding-related activities is well-established in
fruit flies (Pereira and Sokolowski, 1993; Sokolowski, 1980). I have shown that for is also
involved in the circadian regulation of feeding behaviour since rovers and sitters display
different feeding rhythms. This phenotype appears to be related to the foraging gene as sitter
mutants show similar patterns to natural sitters. The for gene has not yet been related to
circadian clocks, although the foraging task shows a link to circadian clocks (Moore et al.,
1998). As the duty of an adult honeybee changes from performing hive tasks to foraging, its
activity become rhythmic. Thus, it is hypothesized that the foraging gene, which is involved
in foraging behaviour in fruit flies, may regulate circadian rhythms. Perhaps feeding
differences are related to their locomotor activity. In both the larval and adult stages, rovers
travel a greater distance than sitters on a nutritive substrate (Pereira and Sokolowski, 1993;
Sokolowski, 1980). As greater locomotion reflects higher energy demands, this could be
related to differences in nutrient storage. Rovers require more energy, thus they store and
utilize more lipids than carbohydrates while the opposite is true for sitters (Kent et al.,
2009). Since rovers show greater nutrient absorption than sitters, at least in larvae, their
feeding times may vary from sitters (Kaun et al., 2007). The foraging gene may play a role
in regulating the time of feeding, based on the nutritional demands of the fly. Surprisingly
rovers and sitters display similar locomotor activity rhythms, indicating that for expressing
clock cells affect feeding rhythms but not locomotor activity rhythms (MacPherson et al.,
2004). In adults, PKG expression is detected in several peripheral tissues such as the fat
body (Personal communication with Amsale Belay and Marla Sokolowski), which suggests
53
that clocks in peripheral tissues may be involved in regulating the temporal pattern of
feeding. In future studies, the feeding patterns of these flies should be measured in constant
darkness, to confirm that differences in rover and sitter feeding patterns are circadian.
Defining a Fly Meal
In order to understand the pattern in food intake in flies, it is essential to define what
comprises a fly meal. It is possible that similar to mammals, flies also show circadian
fluctuations in meal size and/or frequency. However, due to time constraints, I did not study
whether meal size and frequency show circadian patterns. To help us better understand if
feeding could be regulated by such parameters, the total hourly meal was categorized as
small, medium or large and the frequency of flies consuming each type of meal was
quantified. I found a pattern in the consumption of all three types of meals. However, it
appears that the pattern in the frequency of flies consuming large hourly meals drives the
feeding pattern. Similar to the temporal pattern, flies consumed large hourly meals around
the lights-on and lights-off times. This appears to be circadian as these patterns were also
visible in constant darkness. Although period mutants displayed similar patterns in a light-
dark cycle, these were not apparent in constant darkness.
In blowflies, Dethier (1976) defined a „drink‟ as a volume of food that a fly consumes
uninterrupted whereas a meal consisted of a number of drinks after which the fly did not
touch the food again. Perhaps feeding differences seen here result from the modulation of the
number and frequency of drinks consumed. In a preliminary experiment where feeding was
observed in a CAFE assay for an hour, it was noted that fruit flies also consume a number of
short drinks rather than one large meal. A fly consumes a drink for a few seconds, then
walks around, and then returns for another drink. This trend of consuming short drinks with
intervals of activity continues until the fly is satisfied. It was observed that the duration of a
drink fluctuates, perhaps indicating another feeding parameter. Further experiments are
required to elucidate the underlying mechanism that regulates food intake.
54
In future experiments, it will be determined whether the TO peptide is involved in the
regulation of feeding patterns. It is a potential candidate as its expression is regulated by a
circadian clock (So et al., 2000). There are two possible roles that TO may play in the
regulation of food intake. Firstly, it could be a peptide involved in communication between
peripheral clocks in feeding-related tissues. Or it could be a clock output factor that regulates
the amount of food consumed. Feeding will be studied in to1 mutant flies and ry
506 control
flies in LD and constant darkness, to determine its role in the regulation of the feeding
pattern.
55
Chapter 4 . Discussion
A primary aim of circadian research is to learn about the intricate network of molecular
clocks in the body that functions to synchronize behaviour and physiology to varying
environmental conditions. Clock mechanisms are conserved across all organisms, indicating
their importance for survival (Dunlap et al., 2004). In mammals, feeding is regulated by a
circadian clock. Similar to other clock-controlled behaviours, a rhythm in feeding is
observed in a light-dark cycle and can persist for days in constant darkness (Rosenwasser et
al., 1981, 1983).
Here, I report that Drosophila melanogaster exhibit circadian feeding patterns. Flies appear
to feed more around dawn and dusk, which correlates with increases in locomotor activity,
indicating that these behaviours are related (Reviewed in Dubruille and Emery, 2008). It is
likely that metabolic activities regulate the amount of food that a fly consumes. As
locomotion depletes nutrient resources and increases metabolic needs, this may influence the
time of food intake. It is also possible that feeding behaviour regulates nutrient levels which
thereby control activity.
Evidence of clocks in metabolic tissues supports the idea that metabolic activities are
circadianly regulated. I demonstrate that a peripheral clock resides in the abdominal fat body
cells, which are functionally homologous to the mammalian liver and adipose tissues
(Reviewed in Canavoso et al., 2001). The fat body is a major energy storage depot that is
involved in glycogenesis and lipid metabolism (Reviewed in Canavoso et al., 2001;
Gutierrez et al., 2007). The fat body clock could synchronize to the fly‟s rest/activity cycles,
so that lipids and glycogen are mobilized from the fat body and metabolized efficiently to
provide energy for activity. Clocks in oenocyte cells (Krupp et al., 2008), which lie
downstream to the fat body, could coordinate with the fat body clock to regulate the time of
56
lipid metabolism (Gutierrez et al., 2007). If metabolic activities are timed, this could then
control the temporal regulation of feeding.
I show that the central clock is partly involved in regulating feeding behaviour. It is possible
that the central clock could directly influence the time of feeding. Or perhaps clock cells in
the brain might indirectly control feeding time since these cells are known to regulate
locomotor activity (Reviewed in Stanewsky, 2002). As clocks in peripheral tissues also
appear to be involved in regulating feeding, clock cells in the brain and peripheral tissues
could be synchronized. Synchronization of clocks is necessary to maintain normal
behavioural rhythms (Reviewed in Stanewsky, 2002). The neuropeptide, PIGMENT
DISPERSING FACTOR (PDF), could be involved in synchronizing these clocks, as the
PDF receptor is expressed in the abdominal fat body. Another molecule that could transmit
temporal information to the fat body clock is the Adipokinetic hormone (AKH). AKH is
released from the Corpora Cardiaca (CC), and regulates nutrient mobilization from the fat
body (Lee and Park, 2004). The central clock could direct the time of AKH release from the
CC, which could act as a signal between the central clock and the peripheral fat body clock.
Peripheral clocks could regulate feeding via the TAKEOUT (TO) protein. takeout (to) is a
circadian feeding-related gene that is predominantly expressed in feeding-related tissues in
the head and thorax (Meunier et al., 2007; Sarov-Blat et al., 2000; So et al., 2000). As the
expression of to increases under starvation, it is likely that it could transmit information
about the nutrient status in the body to feeding-related tissues, which could stimulate feeding
(Sarov-Blat et al., 2000).
Finally, I have shown that rovers and sitters display different feeding patterns, which
suggests that the foraging (for) gene encoded cGMP-dependent protein kinase (PKG)
interacts with the timing system of feeding. Since PKG expression is detected in peripheral
organs (Personal communication with Amsale Belay and Marla Sokolowski) it may be a link
between clocks in peripheral tissues and the temporal regulation of feeding.
57
Thus, I present a model of the inner-controls of feeding behaviour involving clocks in the
brain and peripheral tissues (Figure 4.1). By studying feeding in D. melanogaster, we can
gain a better understanding of the interactions between the circadian system and the
metabolic processes and behavioural outputs it controls.
Figure 4.1: Clocks in the brain and peripheral tissues may coordinate to regulate the temporal feeding pattern
in adult Drosophila.
The central clock appears to time feeding events either directly or through locomotor activity, which influences
the metabolic needs of the body. Clocks in peripheral tissues such as the oenocytes and the fat body, major
metabolic tissues, may partly regulate the temporal feeding pattern. These peripheral clocks could directly
synchronize to light through a cryptochrome (cry) mediated light pathway or through a signal sent by the
central clock. Possible synchronizing factors include PIGMENT DISPERSING FACTOR (PDF) and
Adipokinetic hormone (AKH). Clocks in peripheral tissues could regulate feeding via TAKEOUT (TO). Also,
the cGMP-dependent protein kinase (PKG) might be a link between peripheral clocks and the circadian
regulation of feeding. These mechanisms could control meal size and meal frequency, which contribute to
feeding rhythms. Bold arrows indicate proven data whereas dashed arrows show the hypothesized theories.
Source of fly drawings: Miller (1950).
58
References
Aguila, J.R., Suszko, J., Gibbs, A.G., and Hoshizaki, D.K. (2007). The role of larval fat cells
in adult Drosophila melanogaster. J Exp Biol 210, 956-963.
Allada, R., White, N.E., So, W.V., Hall, J.C., and Rosbash, M. (1998). A mutant Drosophila
homolog of mammalian Clock disrupts circadian rhythms and transcription of period and
timeless. Cell 93, 791-804.
Balsalobre, A. (2002). Clock genes in mammalian peripheral tissues. Cell Tissue Res 309,
193-199.
Brooks, C.M., Lockwood, R.A., and Wiggins, M.L. (1946). A study of the effect of
hypothalamic lesions on the eating habits of the albino rat. Am J Physiol 147, 735-741.
Butterworth, F.M., Emerson, L., and Rasch, E.M. (1988). Maturation and degeneration of
the fat body in the Drosophila larva and pupa as revealed by morphometric analysis. Tissue
Cell 20, 255-268.
Canavoso, L.E., Jouni, Z.E., Karnas, K.J., Pennington, J.E., and Wells, M.A. (2001). Fat
metabolism in insects. Annu Rev Nutr 21, 23-46.
Challet, E. (2010). Interactions between light, mealtime and calorie restriction to control
daily timing in mammals. J Comp Physiol B 180, 631-644.
Chatterjee, A., Tanoue, S., Houl, J.H., and Hardin, P.E. (2010). Regulation of gustatory
physiology and appetitive behavior by the Drosophila circadian clock. Curr Biol 20, 300-
309.
Cyran, S.A., Buchsbaum, A.M., Reddy, K.L., Lin, M.C., Glossop, N.R., Hardin, P.E.,
Young, M.W., Storti, R.V., and Blau, J. (2003). vrille, Pdp1, and dClock form a second
feedback loop in the Drosophila circadian clock. Cell 112, 329-341.
Dahanukar, A., Hallem, E.A., and Carlson, J.R. (2005). Insect chemoreception. Curr Opin
Neurobiol 15, 423-430.
Dauwalder, B., Tsujimoto, S., Moss, J., and Mattox, W. (2002). The Drosophila takeout
gene is regulated by the somatic sex-determination pathway and affects male courtship
behavior. Genes Dev 16, 2879-2892.
de Belle, J.S., Hilliker, A.J., and Sokolowski, M.B. (1989). Genetic localization of foraging
(for): a major gene for larval behavior in Drosophila melanogaster. Genetics 123, 157-163.
59
Dethier, V.G. (1976). The hungry fly : a physiological study of the behavior associated with
feeding (Cambridge, Mass., Harvard University Press).
Dethier, V.G., and Gelperin, A. (1967). Hyperphagia in Blowfly. Journal of Experimental
Biology 47, 191-&.
Dolezelova, E., Dolezel, D., and Hall, J.C. (2007). Rhythm defects caused by newly
engineered null mutations in Drosophila's cryptochrome gene. Genetics 177, 329-345.
Dubruille, R., and Emery, P. (2008). A plastic clock: how circadian rhythms respond to
environmental cues in Drosophila. Mol Neurobiol 38, 129-145.
Dunlap, J.C., Loros, J.J., and DeCoursey, P.J. (2004). Chronobiology : biological
timekeeping (Sunderland, Mass., Sinauer Associates).
Emery, I.F., Noveral, J.M., Jamison, C.F., and Siwicki, K.K. (1997). Rhythms of Drosophila
period gene expression in culture. Proc Natl Acad Sci U S A 94, 4092-4096.
Fan, Y., Zurek, L., Dykstra, M.J., and Schal, C. (2003). Hydrocarbon synthesis by
enzymatically dissociated oenocytes of the abdominal integument of the German Cockroach,
Blattella germanica. Naturwissenschaften 90, 121-126.
Friedman, J.M., and Halaas, J.L. (1998). Leptin and the regulation of body weight in
mammals. Nature 395, 763-770.
Froy, O. (2007). The relationship between nutrition and circadian rhythms in mammals.
Front Neuroendocrinol 28, 61-71.
Gelperin, A., and Dethier, V.G. (1967). Long-Term Regulation of Sugar Intake by Blowfly.
Physiological Zoology 40, 218-&.
Green, G.W. (1964a). The Control of Spontaneous Locomotor Activity in Phormia-Regina
Meigen .1. Locomotor Activity Patterns of Intact Flies. Journal of Insect Physiology 10,
711-&.
Green, G.W. (1964b). The control of spontaneous locomotor activity patterns in Phormia
regina Meigen. II. Experiments to determine the mechanism involved. J Insect Physiol 10,
727-752.
Greenhouse, S.W., and Geisser, S. (1959). On methods in the analysis of profile data.
Psychometrika 24, 95-112.
Grima, B., Chelot, E., Xia, R., and Rouyer, F. (2004). Morning and evening peaks of activity
rely on different clock neurons of the Drosophila brain. Nature 431, 869-873.
60
Gutierrez, E., Wiggins, D., Fielding, B., and Gould, A.P. (2007). Specialized hepatocyte-like
cells regulate Drosophila lipid metabolism. Nature 445, 275-280.
Hamilton, B.A., and Zinn, K. (1994). From Clone to Mutant Gene. In Drosophila
melanogaster: Practical Uses in Cell and Molecular Biology, L.S.B. Goldstein, and E.A.
Fyrberg, eds. (Academic Press), p. 90.
Hardin, P.E. (1994). Analysis of period mRNA cycling in Drosophila head and body tissues
indicates that body oscillators behave differently from head oscillators. Mol Cell Biol 14,
7211-7218.
Hardin, P.E. (2005). The circadian timekeeping system of Drosophila. Curr Biol 15, R714-
722.
Hardin, P.E., Hall, J.C., and Rosbash, M. (1990). Feedback of the Drosophila period gene
product on circadian cycling of its messenger RNA levels. Nature 343, 536-540.
Helfrich-Forster, C. (2001). The locomotor activity rhythm of Drosophila melanogaster is
controlled by a dual oscillator system. J Insect Physiol 47, 877-887.
Helfrich-Forster, C., Winter, C., Hofbauer, A., Hall, J.C., and Stanewsky, R. (2001). The
circadian clock of fruit flies is blind after elimination of all known photoreceptors. Neuron
30, 249-261.
Im, S.H., and Taghert, P.H. (2010). PDF receptor expression reveals direct interactions
between circadian oscillators in drosophila. J Comp Neurol 518, 1925-1945.
Ito, C., Goto, S.G., Shiga, S., Tomioka, K., and Numata, H. (2008). Peripheral circadian
clock for the cuticle deposition rhythm in Drosophila melanogaster. Proc Natl Acad Sci U S
A 105, 8446-8451.
Ja, W.W., Carvalho, G.B., Mak, E.M., de la Rosa, N.N., Fang, A.Y., Liong, J.C., Brummel,
T., and Benzer, S. (2007). Prandiology of Drosophila and the CAFE assay. Proc Natl Acad
Sci U S A 104, 8253-8256.
Kalsbeek, A., Fliers, E., Romijn, J.A., La Fleur, S.E., Wortel, J., Bakker, O., Endert, E., and
Buijs, R.M. (2001). The suprachiasmatic nucleus generates the diurnal changes in plasma
leptin levels. Endocrinology 142, 2677-2685.
Kaun, K.R., Riedl, C.A., Chakaborty-Chatterjee, M., Belay, A.T., Douglas, S.J., Gibbs,
A.G., and Sokolowski, M.B. (2007). Natural variation in food acquisition mediated via a
Drosophila cGMP-dependent protein kinase. J Exp Biol 210, 3547-3558.
Kaun, K.R., and Sokolowski, M.B. (2009). cGMP-dependent protein kinase: linking
foraging to energy homeostasis. Genome 52, 1-7.
61
Kent, C.F., Daskalchuk, T., Cook, L., Sokolowski, M.B., and Greenspan, R.J. (2009). The
Drosophila foraging gene mediates adult plasticity and gene-environment interactions in
behaviour, metabolites, and gene expression in response to food deprivation. PLoS Genet 5,
e1000609.
Konopka, R.J., and Benzer, S. (1971). Clock mutants of Drosophila melanogaster. Proc Natl
Acad Sci U S A 68, 2112-2116.
Konopka, R.J., Pittendrigh, C., and Orr, D. (1989). Reciprocal behaviour associated with
altered homeostasis and photosensitivity of Drosophila clock mutants. J Neurogenet 6, 1-10.
Krishnan, B., Dryer, S.E., and Hardin, P.E. (1999). Circadian rhythms in olfactory responses
of Drosophila melanogaster. Nature 400, 375-378.
Krishnan, B., Levine, J.D., Lynch, M.K., Dowse, H.B., Funes, P., Hall, J.C., Hardin, P.E.,
and Dryer, S.E. (2001). A new role for cryptochrome in a Drosophila circadian oscillator.
Nature 411, 313-317.
Krupp, J.J., Kent, C., Billeter, J.C., Azanchi, R., So, A.K., Schonfeld, J.A., Smith, B.P.,
Lucas, C., and Levine, J.D. (2008). Social experience modifies pheromone expression and
mating behavior in male Drosophila melanogaster. Curr Biol 18, 1373-1383.
Lee, G., and Park, J.H. (2004). Hemolymph sugar homeostasis and starvation-induced
hyperactivity affected by genetic manipulations of the adipokinetic hormone-encoding gene
in Drosophila melanogaster. Genetics 167, 311-323.
Lee, K.S., You, K.H., Choo, J.K., Han, Y.M., and Yu, K. (2004). Drosophila short
neuropeptide F regulates food intake and body size. J Biol Chem 279, 50781-50789.
Levine, J.D., Funes, P., Dowse, H.B., and Hall, J.C. (2002). Advanced analysis of a
cryptochrome mutation's effects on the robustness and phase of molecular cycles in isolated
peripheral tissues of Drosophila. BMC Neurosci 3, 5.
MacPherson, M.R., Broderick, K.E., Graham, S., Day, J.P., Houslay, M.D., Dow, J.A., and
Davies, S.A. (2004). The dg2 (for) gene confers a renal phenotype in Drosophila by
modulation of cGMP-specific phosphodiesterase. J Exp Biol 207, 2769-2776.
Mauchly, J.W. (1940). Significance test for sphericity of a normal n-variate distribution.
Annals of Mathematical Statistics 11, 204-209.
Melcher, C., Bader, R., and Pankratz, M.J. (2007). Amino acids, taste circuits, and feeding
behavior in Drosophila: towards understanding the psychology of feeding in flies and man. J
Endocrinol 192, 467-472.
Melcher, C., and Pankratz, M.J. (2005). Candidate gustatory interneurons modulating
feeding behavior in the Drosophila brain. PLoS Biol 3, e305.
62
Mertens, I., Vandingenen, A., Johnson, E.C., Shafer, O.T., Li, W., Trigg, J.S., De Loof, A.,
Schoofs, L., and Taghert, P.H. (2005). PDF receptor signaling in Drosophila contributes to
both circadian and geotactic behaviors. Neuron 48, 213-219.
Meunier, N., Belgacem, Y.H., and Martin, J.R. (2007). Regulation of feeding behaviour and
locomotor activity by takeout in Drosophila. J Exp Biol 210, 1424-1434.
Miller, A. (1950). M. Demerec (ed): Biology of Drosophila (New York.).
Moore, D., Angel, J.E., Cheeseman, I.M., Fahrbach, S.E., and Robinson, G.E. (1998).
Timekeeping in the honey bee colony: integration of circadian rhythms and division of labor
Behavioral ecology and sociobiology 43, 147-160.
Nitabach, M.N., and Taghert, P.H. (2008). Organization of the Drosophila circadian control
circuit. Curr Biol 18, R84-93.
Osborne, K.A., Robichon, A., Burgess, E., Butland, S., Shaw, R.A., Coulthard, A., Pereira,
H.S., Greenspan, R.J., and Sokolowski, M.B. (1997). Natural behavior polymorphism due to
a cGMP-dependent protein kinase of Drosophila. Science 277, 834-836.
Park, J.H., Helfrich-Forster, C., Lee, G., Liu, L., Rosbash, M., and Hall, J.C. (2000).
Differential regulation of circadian pacemaker output by separate clock genes in Drosophila.
Proc Natl Acad Sci U S A 97, 3608-3613.
Pereira, H.S., and Sokolowski, M.B. (1993). Mutations in the larval foraging gene affect
adult locomotory behavior after feeding in Drosophila melanogaster. Proc Natl Acad Sci U S
A 90, 5044-5046.
Pfaffl, M.W. (2001). A new mathematical model for relative quantification in real-time RT-
PCR. Nucleic Acids Res 29, e45.
Plautz, J.D., Kaneko, M., Hall, J.C., and Kay, S.A. (1997). Independent photoreceptive
circadian clocks throughout Drosophila. Science 278, 1632-1635.
Renn, S.C., Park, J.H., Rosbash, M., Hall, J.C., and Taghert, P.H. (1999). A pdf
neuropeptide gene mutation and ablation of PDF neurons each cause severe abnormalities of
behavioral circadian rhythms in Drosophila. Cell 99, 791-802.
Reppert, S.M., and Weaver, D.R. (2002). Coordination of circadian timing in mammals.
Nature 418, 935-941.
Rosenwasser, A.M., Boulos, Z., and Terman, M. (1981). Circadian organization of food
intake and meal patterns in the rat. Physiol Behav 27, 33-39.
Rosenwasser, A.M., Boulos, Z., and Terman, M. (1983). Circadian feeding and drinking
rhythms in the rat under complete and skeleton photoperiods. Physiol Behav 30, 353-359.
63
Sarov-Blat, L., So, W.V., Liu, L., and Rosbash, M. (2000). The Drosophila takeout gene is a
novel molecular link between circadian rhythms and feeding behavior. Cell 101, 647-656.
So, W.V., Sarov-Blat, L., Kotarski, C.K., McDonald, M.J., Allada, R., and Rosbash, M.
(2000). takeout, a novel Drosophila gene under circadian clock transcriptional regulation.
Mol Cell Biol 20, 6935-6944.
Sokolowski, M.B. (1980). Foraging strategies of Drosophila melanogaster: a chromosomal
analysis. Behav Genet 10, 291-302.
Stanewsky, R. (2002). Clock mechanisms in Drosophila. Cell Tissue Res 309, 11-26.
Stanewsky, R., Kaneko, M., Emery, P., Beretta, B., Wager-Smith, K., Kay, S.A., Rosbash,
M., and Hall, J.C. (1998). The cryb mutation identifies cryptochrome as a circadian
photoreceptor in Drosophila. Cell 95, 681-692.
Stokkan, K.A., Yamazaki, S., Tei, H., Sakaki, Y., and Menaker, M. (2001). Entrainment of
the circadian clock in the liver by feeding. Science 291, 490-493.
Stratagene. (2004). Introduction to Quantitative PCR. Methods and applications guide.
TriKinetics, inc. (2007-2008). Drosophila Activity Monitoring System. Waltham, MA,
2007-2008.
Wong, R., Piper, M.D., Wertheim, B., and Partridge, L. (2009). Quantification of food intake
in Drosophila. PLoS One 4, e6063.
Xu, K., Zheng, X., and Sehgal, A. (2008). Regulation of feeding and metabolism by
neuronal and peripheral clocks in Drosophila. Cell Metab 8, 289-300.
Yamazaki, S., Numano, R., Abe, M., Hida, A., Takahashi, R., Ueda, M., Block, G.D.,
Sakaki, Y., Menaker, M., and Tei, H. (2000). Resetting central and peripheral circadian
oscillators in transgenic rats. Science 288, 682-685.
Yu, Q., Jacquier, A.C., Citri, Y., Hamblen, M., Hall, J.C., and Rosbash, M. (1987).
Molecular mapping of point mutations in the period gene that stop or speed up biological
clocks in Drosophila melanogaster. Proc Natl Acad Sci U S A 84, 784-788.
Zehring, W.A., Wheeler, D.A., Reddy, P., Konopka, R.J., Kyriacou, C.P., Rosbash, M., and
Hall, J.C. (1984). P-element transformation with period locus DNA restores rhythmicity to
mutant, arrhythmic Drosophila melanogaster. Cell 39, 369-376.
Zerr, D.M., Hall, J.C., Rosbash, M., and Siwicki, K.K. (1990). Circadian fluctuations of
period protein immunoreactivity in the CNS and the visual system of Drosophila. J Neurosci
10, 2749-2762.
64
Zhang, Y., Proenca, R., Maffei, M., Barone, M., Leopold, L., and Friedman, J.M. (1994).
Positional cloning of the mouse obese gene and its human homologue. Nature 372, 425-432.
Zucker, I. (1971). Light-dark rhythms in rat eating and drinking behavior. Physiol Behav 6,
115-126.
65
Appendix A: Detailed Method for Preparing Fly Food
Fly food is prepared by adding 48-60g of agar (depending on humidity), 90g of sucrose,
180g of D-glucose, 210g of yeast, 90g of cornmeal, 60g of wheat germ, 60g of soy flour and
180g of molasses to 6 L of warm tap water in a large pot. This food mixture is stirred and
heated on a heating block set at 395˚C. Once the temperature of food reaches ~90˚C, the dial
on the heating pad is lowered to 120˚C and the food is heated for another 10 minutes. The
pot is then transferred to the fumehood and food is stirred with an eggbeater until it cools to
approximately 47˚C. A previously prepared mixture of 12g of tegosept in 60mL of 95%
ethanol and 30mL of propionic acid are subsequently added to the food and stirred. Using a
plastic single-use syringe, 8mL and 40mL of food are poured into fly vials and bottles,
respectively. After the food is cooled overnight, vials and bottles are capped with cotton
plugs and stored at 4˚C until further use.
66
Appendix B: Fat Body Timeseries Data
The RNA expression levels ± SEM values are provided for replicate time series experiments that were not
plotted in Figure 2.3.
Table 1: The relative RNA expression levels of period, timeless and Clock genes in the fat body tissue of
Canton-S flies in a light-dark cycle.
Zeitgeber
time
(hours)
timeless
RNA level
Standard
Error
period RNA
level
Standard
Error
Clock RNA
level
Standard
Error
1 0.078 0.002 0.100 0.005 0.586 0.019
4 0.093 0.002 0.162 0.006 0.925 0.030
7 0.141 0.005 0.181 0.003 0.849 0.047
10 0.275 0.013 0.245 0.046 0.457 0.074
13 0.747 0.004 1.000 0.022 0.325 0.016
16 0.941 0.005 0.735 0.029 0.343 0.061
19 1.000 0.006 0.621 0.026 1.000 0.045
22 0.474 0.021 0.251 0.005 0.918 0.052
Table 2: The relative RNA expression levels of period, timeless and Clock genes in the fat body tissue of
Canton-S flies under constant darkness.
Circadian
time
(hours)
timeless
RNA level
Standard
Error
period RNA
level
Standard
Error
Clock RNA
level
Standard
Error
1 0.152 0.009 0.243 0.009 0.543 0.048
4 0.128 0.010 0.318 0.017 0.651 0.063
7 0.191 0.020 0.395 0.035 1.000 0.121
67
Circadian
time
(hours)
timeless
RNA level
Standard
Error
period RNA
level
Standard
Error
Clock RNA
level
Standard
Error
10 0.410 0.028 0.611 0.055 0.847 0.081
13 0.417 0.027 0.827 0.022 0.541 0.047
16 0.651 0.042 1.000 0.031 0.365 0.036
19 0.655 0.040 0.653 0.014 0.536 0.047
22 1.000 0.087 0.549 0.044 0.734 0.079
Table 3: The relative RNA expression levels of period, timeless and Clock genes in the fat body tissue of y w
flies under constant darkness.
Circadian
time
(hours)
timeless
RNA level
Standard
Error
period RNA
level
Standard
Error
Clock RNA
level
Standard
Error
1 0.222 0.012 0.509 0.045 1.000 0.019
4 0.240 0.003 0.334 0.024 0.822 0.046
7 0.295 0.024 0.444 0.038 0.797 0.016
10 0.574 0.053 0.825 0.101 0.816 0.079
13 0.861 0.011 1.000 0.101 0.696 0.011
16 1.000 0.013 0.770 0.058 0.588 0.071
19 0.499 0.005 0.256 0.020 0.238 0.009
22 0.174 0.005 0.024 0.003 0.156 0.010
68
Table 4: The relative RNA expression levels of period, timeless and Clock genes in the fat body tissue of per01
flies under constant darkness.
Circadian
time
(hours)
timeless
RNA level
Standard
Error
period RNA
level
Standard
Error
Clock RNA
level
Standard
Error
1 0.869 0.040 1.000 0.011 0.586 0.073
4 0.646 0.009 0.677 0.044 0.885 0.069
7 0.721 0.010 0.824 0.123 1.000 0.025
10 0.780 0.073 0.758 0.207 0.770 0.138
13 1.000 0.010 0.879 0.014 0.663 0.082
16 0.700 0.008 0.648 0.091 0.578 0.012
19 0.671 0.016 0.712 0.019 0.532 0.016
22 0.960 0.083 0.845 0.095 0.434 0.070
69
Appendix C: Statistical Analysis
Table 5: The parameters of cosine curves fit to relative RNA expression levels in the fat body tissue of Canton-
S flies.
Light
treatment
Clock
gene
a b Standard
error (b)
95% confidence
interval (b)
P-value
(b)
h r-squared
value
Lower
bound
Upper
bound
LD tim 0.591 0.361 0.061 0.192 0.529 <0.05 -5.019 0.903
DD tim 0.422 -0.366 0.081 -0.592 -0.141 <0.05 6.275 0.846
LD per 0.748 -0.289 0.033 -0.382 -0.196 <0.05 5.940 0.953
DD per 0.555 -0.348 0.082 -0.574 -0.121 <0.05 5.228 0.832
LD Clk 0.596 0.345 0.100 0.068 0.622 <0.05 4.967 0.765
DD Clk 0.649 0.302 0.068 0.113 0.491 <0.05 4.720 0.842
Table summary: A non-linear regression analysis was performed to generate a curve of best fit for the
expression data of tim, per and Clk. a represents the y-intercept of the curve. h and b respectively represent the
phase and amplitude of the cosine curve. The 95% confidence intervals indicate that the amplitude of the
curves of all genes is significantly different from 0. All P-values are statistically significant. The r-squared
values indicate that the expression data for tim and per in LD and tim, per and Clk in DD fit the idealized
cosine curve well.
70
Table 6: The parameters of cosine curves fit to relative RNA expression levels in the fat body tissue of y w and
per01
flies under constant darkness.
Genotype Clock
gene
a b Standard
error (b)
95% confidence
interval (b)
P-value
(b)
h r-squared
value
Lower
bound
Upper
bound
y w tim 0.511 -0.469 0.133 -0.838 -0.099 <0.05 5.635 0.722
per01
tim 0.583 -0.164 0.148 -0.575 0.246 >0.05 4.962 0.251
y w per 0.585 -0.352 0.071 -0.550 -0.155 <0.05 5.034 0.869
per01
per 0.625 -0.194 0.136 -0.572 0.184 >0.05 -4.513 0.344
y w Clk 0.612 0.115 0.099 -0.160 0.391 >0.05 5.670 0.270
per01
Clk 0.559 -0.172 0.144 -0.573 0.228 >0.05 1.072 0.293
Table summary: A non-linear regression analysis was performed to generate a curve of best fit for the
expression data of tim, per and Clk. a represents the y-intercept of the curve. h and b respectively represent the
phase and amplitude of the cosine curve. The 95% confidence intervals indicate that the amplitude of y w tim
and per curves are significantly different from 0 but the amplitude of y w Clk and per01
tim, per, and Clk curves
are not. The P-values for y w tim and per cosine curves are significant but the P-values for y w Clk and per01
tim, per and Clk cosine curves are not significant. Variability between biological replicates is likely to be the
cause of low r-squared values.
71
Table 7: The Statistical results of General Linear Model Repeated Measures tests performed to compare
feeding amounts between fly strains.
Light
treatment
Genotypes
tested
Effect of genotype Mauchly’s
test of
Sphericity
Effect of time Effect of genotype-by-
time interaction
Df F-stat P-value P-value Df F-stat P-value Df F-stat P-value
LD Canton-S vs.
pdf01
1 1.77 0.19 0.000 3.19 3.14 0.03 3.19 1.48 0.22
LD Canton-S vs.
y w
1 0.27 0.61 0.000 3.24 6.90 0.000 3.24 0.58 0.64
LD y w vs. per01
1 5.65 0.02 0.000 3.58 4.62 0.002 3.58 0.47 0.734
LD ry506
vs.
per01
7.2:2
1 8.64 0.006 0.000 3.31 8.93 0.000 3.31 0.48 0.72
LD w1118
vs. cry0 1 0.01 0.94 0.000 2.99 4.30 0.006 2.99 1.75 0.16
LD fors vs. for
s2 1 0.62 0.45 0.000 2.31 10.18 0.000 2.31 2.61 0.09
LD fors vs. for
R 1 0.37 0.55 0.02 3.04 8.11 0.000 3.04 4.16 0.01
LD fors2
vs. forR 1 0.00 0.98 0.05 5 4.33 0.002 5 4.09 0.003
DD Canton-S vs.
pdf01
1 8.39 0.007 0.000 5.14 1.67 0.14 5.14 1.44 0.21
DD y w vs. per01
1 2.88 0.10 0.000 7.01 2.12 0.04 7.01 1.73 0.10
DD ry506
vs.
per01
7.2:2
1 2.07 0.16 0.000 6.47 4.21 0.000 6.47 2.00 0.06
DD w1118
vs. cry0 1 0.40 0.53 0.000 6.19 5.14 0.000 6.19 1.47 0.19
LL Canton-S vs.
cry0
1 4.13 0.05 0.001 7.10 3.10 0.004 7.10 1.75 0.10
LL y w vs. cry0 1 1.20 0.28 0.000 6.70 1.50 0.17 6.70 3.31 0.003
LL w1118
vs. cry0 1 1.18 0.29 0.001 6.10 6.04 0.000 6.10 2.01 0.07
Table summary: Statistical tests were performed on moving average data that was binned in 4 hour intervals.
If the assumption of sphericity tested with the Mauchly‟s test of sphericity was violated (P-value<0.05), the
degrees of freedom values were adjusted with Greenhouse-Geisser corrections. Significant test results have
been bolded in the table above.