w w",
“gym

mum-

A
32% n

fig"

32:”.5"
"@223;
~ mm’r .
._..
(“[2 ' >-
J'x
I

Fax)”
"1 .

 

This is to certify that the
dissertation entitled

In Vivo Analysis Of The Molecular Mechanism Of Repression
Of The Drosophila Long-Range Repressor Hairy

 

presented by

Carlos Alberto Martinez

LIBRARY
Michigan State
Unrversrty

 

 

 

has been accepted towards fulfillment
of the requirements for the

degree' In Genetics

LWM

Major Professor 5 Signature
8 / 27 [07

Date

MSU is an afiimrative—actmn, equal-opportunity employer

 

 

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/C|RC/DateDue.indd-p.1

IN VIVO ANALYSIS OF THE MOLECULAR MECHANISM OF REPRESSION OF
THE DROSOPHILA LONG-RANGE REPRESSOR HAIRY

By

Carlos Alberto Martinez

A DISSERTATION

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

DOCTOR OF PHILOSOPHY
Program in Genetics

2007

ABSTRACT

IN VIVO ANALYSIS OF THE MOLECULAR MECHANISM OF REPRESSION OF
THE DROSOPHILA LONG-RAN GE REPRESSOR HAIRY

By
Carlos Alberto Martinez
The Drosophila Hairy protein belongs to a widely conserved class of transcriptional
repressors called Hairy and Enhancer-of-Split (HES) proteins that have important roles in
embryogenesis and cell fate determination. HES family members are characterized by a
conserved basic helix-loop-helix DNA binding domain, an Orange domain, and a C-
terminal WRPW motif that binds the TLE/Groucho corepressor. While enormous
advances have been made in elucidating the biological functions of HES proteins, there is
very little molecular understanding of the mechanism by which they repress transcription.
To address this issue, Hairy repression was analyzed in the context of a highly defined
Drosophila embryo system. The system consisted of embryos carrying a lacZ reporter
transgene activated ubiquitously by the yeast Gal4 protein, the latter of which was
expressed under control of the daughterless promoter. Uniform repression of the system
was achieved by the heat shock induction of a LexA—Hairy fusion protein that can bind to
and repress the reporter. Using embryo chromatin immunoprecipitation, I have analyzed
the recruitment of activators, coactivators, repressors, corepressors, and chromatin
modifications associated with Hairy repression. The results show that during repression
activators and coactivators remain bound to the promoter, suggesting that Hairy does not
work by blocking activator and coactivator recruitment. Chromatin immunoprecipitation
analysis showed that during repression Groucho was recruited to the entire four kilobase

coding region of the Idol reporter. Groucho binding was accompanied by recruitment of

the histone deacetylase de3, a known Groucho interacting protein, and a subsequent
loss of H3 and H4 acetylation levels at the promoter and one kilobase downstream.
Concomitant with the loss of histone acetylation levels, a two-fold increase in total
histone H3 occupancy was seen at the promoter region, suggesting that Hairy repression
is associated with chromatin remodeling. Given that previous studies had characterized
Hairy as a long-range repressor capable of dominantly inhibiting distal elements, the
results suggest a model wherein Groucho spreading and subsequent histone deacetylation
can lead to long-range gene silencing. However, unlike previously characterized long-
range silencing mechanisms that involve corepressor spreading, the repression observed

in this system was transient.

ACKNOWLEDGEMENTS

I would like to acknowledge everyone who contributed to this thesis. Special
thanks to my mentor David Amosti for his patience and guidance. Also I would like to
thank the members of my committee Bill Henry, Steve Triezenberg, and Lee Kroos for
helping me to develop as a scientist. Thanks to all my lab members for their insight and
discussions. Thank you to my family in Chile and Honduras who supported me all the
way. My deepest thanks go to my wife and son who made it all worth it, gracias Chiquita

y David.

iv

TABLE OF CONTENTS

List of Figures ....................................................................................... vii
Key to Abbreviations ............................................................................... viii
CHAPTER I: Literature review .................................................................. 1
Early developmental regulation in Drosophila. . . . . . . . .............2
Transcriptional repression in Drosophila ................................................................ 4
Mechanisms of transcriptional repression .............................................................. 7
Disruption of activator firnction ................................................................... 7

Direct repression of the basal transcriptional machinery ...................... 8

Recruitment of chromatin modifying and remodeling complexes. ........... 9

Hairy ......................................................................................... 12
Hairy/E(spl) repression mechanism ............................................. 15

Hairy/E(spl) firnctions in Drosophila ........................................... l9

Hairy/E(spl) fimctions in vertebrates ........................................... 24

Gal4 .......................................................................................... 3 1

Gal4 transcriptional activation ................................................... 32

CHAPTER 11: Recruitment of Gal4 and SAGA is not affected by Hairy-mediated

repression ............................................................................................ 37
Introduction ................................................................................. 37
Materials and Methods ...................................................................... 39
Results ....................................................................................... 47

Construction and characterization of a regulated Hairy repressible

transgene ........................................................................... 47

Co-occupancy of activator and repressor during gene repression .......... 54

Transcriptional co-activators remain associated with repressed

promoter ............................................................................ 55
Discussion ................................................................................... 60

CHAPTER III: Spreading of corepressor linked to action of long-range repressor

Hairy .................................................................................................. 63
Introduction ................................................................................. 63
Materials and Methods ...................................................................... 66
Results ....................................................................................... 72

Association of Groucho and de3 co-repressors with wide tracts of
repressed gene ..................................................................... 72
Chromatin modification associated with activation and repression of a
target gene .......................................................................... 76
Discussion ................................................................................... 83

CHAPTER IV: Conclusions and Future Directions ......................................... 88

Development of a model system for the study of transcriptional regulation ......... 88
Hairy repression mechanism does not rely on activator or coactivator
displacement ................................................................................ 90
Hairy repression is associated with Groucho binding to the entire gene
locus ....................................................................................................................... 92
Hairy repression is associated with histone deacetylation and chromatin
remodeling ................................................................................... 96
Gal4 activation is associated with demethylation of K27 ............................. 97
Reactivation of gene activity after LexA-Hairy repression ........................... 97
Contrasting transcriptional states: Unactivated versus Repressed. . .. . . . . ......99
Short versus long-range repression: Two sides of the same coin .................. 102
Contextual dependency of transcriptional repression: Implications for modeling
cis regulatory elements .................................................................. 103
Concluding remarks ...................................................................... 105
APPENDIX A: Characterization analysis of the LexA-Knirps fusion protein ...... 106
APPENDIX B: Pol II accumulates in the ORF during repression ..................... 109
REFERENCES .................................................................................... 111

vi

Figure 1-1.
Figure 1-2.
Figure 1-3.
Figure 1-4.
Figure 2-1.
Figure 2-2.
Figure 2-3.
Figure 2-4.
Figure 2-5.
Figure 2-6.
Figure 2-7.

Figure 3-1.

LIST OF FIGURES

Long-range and short-range repression .............................................. 6
Mechanisms of transcriptional repression .......................................... ll
Conserved Hairy domains and expression pattern ................................ 13
Hairy/E(spl) proteins pattern PNS formation in Drosophila .................... 22
Hairy-regulated gene system in the Drosophila embryo ......................... 48
LexA-Hairy can repress in vivo ........................................................ 49
Expression of lacZ and LexA-Hairy in blastodenn embryos. . . ......51
LexA-Hairy repression is reversible ................................................. 52
Quantitation of lacZ staining after heat shock induction of LexA-Hairy. . ....53

LexA-Hairy repression does not prevent Gal4 binding .......................... 57
LexA-Hairy repression does not block SAGA recruitment by Gal4 ........... 58
LexA-Hairy repression results in recruitment of Groucho and de3 to the

entire gene locus .................................................................................... 75

Figure 3-2. LexA-Hairy repression results in histone deacetylation and higher histone
occupancy at the promoter ......................................................................... 78

Figure. 3-3. Change in acetylation levels and histone occupancy is limited to 1 kb from

promoter ............................................................................................... 81
Figure 4-1. Groucho spreading model ............................................................ 93
Figure 4-2. Proposed mechanism for Groucho-dependent reactivation .................... 101
Figure A-l. LexA-Knirps does not repress in model system ................................. 107
Figure 3-1. Chromatin immunoprecipitation of Pol II ....................................... 110

vii

AS-C
bHLH

ChIP
CK2
CtBP
CTD
Da
Dam
DSP
E(spl)
Emc
eve

Gro
GTF
HAT
HDAC
HES
Hesr
HLH
HMT
HS

kb

ORF

PcG
PCR
PIC
PNS
Pol

SAGA
Sir2
Slp 1
Su(Hw)
TAF
TBP
TLE

UAS

KEY TO ABBREVIATIONS

Achaete-Scute Complex

basic helix-loop-helix

base pair

Chromatin immunoprecipitation
Cyclin-dependent kinase 2
C-terminal binding protein
C-terminal domain
Daughterless

DNA adenine methyltransferase
Dithiobis(succinimidyl propionate )
Enhancer of split
Extramacrochaetae
even-skipped

fushi-taratzu

Groucho

General transcription factor
Histone acetyltransferase
Histone deacetylase

Hairy and Enhancer of split
Hairy and Enhancer of Split related
Helix-loop-helix

Histone methyltransferase
Heatshock

kilobase pair

Knirps

Krfippel

Open reading frame

Polycomb

Polycomb group

Polymerase chain reaction
Pre-initiation complex
Peripheral nervous system
Polymerase

rhomboid

Spt-Ada-GcnS -acetyl transferase complex
Silent information regulator 2
Sloppy-paired 1

Suppressor of Hairy wing
TBP-associated factor

TATA binding protein
Transducin-like—enhancer-of-split
twist

Upstream activating sequence

viii

Chapter I

Literature Review

The purpose of this project is to elucidate the molecular mechanism of repression
of the long-range repressor Hairy in the context of a Drosophila embryo. Hairy belongs
to a widely conserved family of transcriptional repressors called Hairy and Enhancer-of-
Split (HES) proteins that have important roles in development and in determining cell
fate decisions. While a wealth of information on the fiinction of HES proteins exists,
relatively little is known about the underlying mechanism of repression [1, 2]. A
difficulty in studying developmental regulation in vivo is that only a specific subset of
genes are expressed in different tissues. Thus, in early Drosophila development only a
fraction of the cells express Hairy [3]. In order to solve this problem, I created an in vivo
system wherein a transcriptional switch homogenously regulates the expression of a lacZ
reporter within the Drosophila embryo. Activation of the switch was achieved by the
ubiquitous expression of the yeast Gal4 activator, while repression was brought about by
the heatshock induction of Hairy fused to the DNA-binding bacterial LexA protein.
Using this system, both the repressor and the activator could be tracked on the gene, as
well as the associated cofactors and chromatin modifications. Furthermore, by
manipulating the expression of Gal4 and LexA-Hairy, the system allowed the
differentiation between activator- and repressor-specific effects.

In this chapter I will review the literature with a focus on placing Hairy repression
in a developmental and mechanistic context. To that end, I will discuss the functional

roles of Hairy and related proteins in regulating Drosophila and vertebrate development.

I will also discuss in general terms the different mechanisms of transcriptional repression
and describe what is known about the mechanism of repression by HES proteins. Given
that the repression mechanism studied here takes place in the context of a reporter system
activated by Gal4, I will include a section on the mechanism of Ga14 activation.

The results of this study show several important aspects of Hairy repression. First,
tethering Hairy to the reporter silences the gene yet it does not displace the activator or
coactivator from the promoter (Chapter 2). Second, repression is associated with the
binding of the Hairy corepressor Groucho to an extensive region encompassing several
kilobases, suggesting that Groucho can spread throughout the gene (Chapter 3). Third,
repression was accompanied by recruitment of the de3 histone deacetylase and a
subsequent loss of acetylated histones at the promoter (Chapter 3). Lastly, repression was
not permanent as it was readily reversed upon degradation of LexA-Hairy with a

corresponding increase in histone acetylation levels (Chapter 2 and 3).

Early developmental regulation in Drosophila

In recent years, great strides have been made in understanding how gene activity
is regulated in eukaryotes. The use of unicellular model systems such as yeast has
provided a wealth of information on the molecular mechanism of regulation [4].
However, much less is known about the mechanisms specifying the complex temporal
and spatial patterns of gene expression in developing organisms. One of the difficulties in
studying gene regulation in metazoans is that each different cell type requires a specific
genetic program that leads to the activation and repression of different subsets of genes

[5]. Thus, in order to elucidate the mechanistic details of transcriptional regulation in

multicellular organisms, it is necessary to take into account the tissue and developmental
time frame in which it occurs.

The fi'uitfly Drosophila melanogaster has been the organism of choice for
studying developmental regulation. Its relatively small genome and the abundance of
genetic tools provide an ideal platform for the systematic analysis of individual genes, as
well as for the study of gene networks. Another advantage is the rapid rate of Drosophila
deve10pment [6]. In only 24 hours, a single fertilized egg develops into a larva composed
of multiple different tissues and organs [6]. This rapid rate of development greatly
reduces the time needed to generate and characterize different Drosophila lines. In
addition, early development occurs in a scyncytial embryo containing only nuclei and no
cellular membranes, simplifying transcription regulatory models by allowing intercellular
signaling events to be ignored [6]. Multiple studies of Drosophila development have
shown that most regulatory mechanisms are conserved in higher organisms, including
humans [7].

Early Drosophila development is characterized by a hierarchical cascade of
transcriptional factors that progressively refine the expression patterns of genes involved
in the establishment of segment identity [8]. Initially, maternally deposited RNA at the
anterior and posterior poles of the embryo creates wide gradients of the maternal genes
(bicoid, hunchback, caudal, and nanos) that serve as a coordinate system for specifying
positional information [8]. In turn, the maternal genes regulate the expression of the
zygotically expressed gap genes (zygotic hunchback, knirps, kruppel, giant, tailless,
huckebein), so-called because null mutations cause a “gap” phenotype wherein multiple

contiguous segments of the larvae are missing [9]. Gap genes then function to specify the

seven-stripe pattern of the pair-rule genes (even-skipped, fushi-tarazu, hairy, runt, paired,
odd-skipped, odd-paired, sloppy-paired), mutations of which cause the loss of every
other segment [9]. Finally, pair-rule genes are involved in regulating the expression of the
segment polarity genes (Wingless, hedgehog, and engrailed), which are expressed in a

series of 14 stripes and function to define the metameric segment boundaries [10].

Transcriptional repression in Drosophila

A common theme of Drosophila segmentation is that repressors play a critical
role in establishing the boundaries of gene expression domains [11]. For example,
formation of the seven stripe pattern of even-skipped (eve) involves the additive action of
modular cis regulatory elements or “enhancers” that establish stripes 3/7, 2, 4/6, 1 and 5
respectively [12]. Initially, broadly expressed activators such as Bicoid, Hunchback, and
Stat bind to these elements and activate expression throughout the embryo [8]. However,
as development progresses, stripes become more defined as the gap genes bind to these
modules and repress their activity in the interstripe regions [8]. Thus, formation of the eve
stripe 2, involves activation in the anterior half of the embryo by Bicoid and Hunchback
and repression by Giant and Kruppel at the anterior and posterior borders respectively [8,
11].

Despite the importance of repression in developmental regulation of eukaryotes,
early studies focused mostly on the mechanism of transcriptional activation [13]. This
focus was due to the idea that since only 7% of genes are transcribed at any one time in
an eukaryotic cell, it is more cost effective to activate the genes needed than it is to

repress the rest [13]. Thus, it was believed that most transcription factors should function

as activators [14]. In addition, repression was viewed simply as a disruption of the
activation process [13]. However, with the discovery and characterization of many
repressor complexes, it is now clear that repression plays a big part in transcriptional
regulation and can involve multiple mechanisms [15, 16].

In Drosophila, repressors are often grouped into two firnctional classes depending
on their range of activity [17]. Short-range repressors like Snail, Knirps, Kruppel, and
Giant can disrupt the firnctions of activators or the promoter only if they are bound within
100-150 bp (Figure 1-1A) [18-21]. Such a limited range seems well suited to maintaining
the enhancer autonomy of modular cis regulatory elements such as those found on eve
[18]. In contrast, long-range repressors like Hairy and Polycomb (Pc) can silence multiple
enhancers in a dominant fashion even when bound over 1 kb away from the nearest

activator or promoter (Figure l-lB) [22-24].

 

 

Figure 1-1. Long-range and short-range repression.

(A) Short-range repressors like Knirps, Snail, Giant, and Kruppel interfere with the
function of activators that lie within 100-150 bp from the repressor. By repressing
locally short-range repressors maintain enhancer autonomy. (B) Long-range repressors
like Hairy can dominantly repress distant enhancers over distances of several kilobases.
Repressors are represented by squares and activators by ovals. Images in this dissertation
are presented in color.

Mechanisms of transcriptional repression

Transcriptional repressors can be characterized as being global or gene-specific.
Global repressors typically firnction by sequestering or modifying components of the
basal transcriptional machinery, such that it is not available for transcription, resulting in
the down-regulation of genes that require that component [16]. For example, herpes virus
infection can lead to aberrant phosphorylation of the Pol II CTD, resulting in a down-
regulation of genes transcribed by Pol II [25]. Gene-specific repressors on the other hand
regulate a much smaller subset of genes. This type of repression can be effected by
disrupting the formation of an activator/DNA complex or by blocking their stimulatory
interactions with the promoter (Figure 1-2A) [16, 26]. Alternatively, gene-specific
repression can occur by interactions with the basal machinery at a particular promoter
(Figure 1-2B) [16]. In addition, repressors can recruit chromatin modifying enzymes to
locally alter the chromatin structure so that it becomes refractive to transcription (Figure
l-2C) [16]. Many gene-specific repressors contain DNA-binding domains that enable
them to specifically bind to their targets [26]. In contrast, some repressors are recruited to

genes via interaction with other DNA binding proteins [26].

Disruption of activator function

Transcriptional repression can be achieved by processes that degrade, sequester,
covalently modify, or alter the localization of activators in the nucleus (Figure 1-2A,
insets 1-4) [26]. Additionally, disruption of activator function can occur through binding
site competition, or by masking protein-protein interactions necessary for activation

(Figure 1-2A, insets 5 and 6) [16]. For example, ubiquitination of p53 by Mdrn2 can

promote protein turnover and nuclear export [27]. Similarly, deacetylation of E2Fl by the
Rb complex can inhibit activator binding [28]. Another example of activator disruption is
the sequestration of the MyoD activator by heterodimerization to the non-DNA-binding
Id protein [26]. Other examples include masking of the Gal4 activation domain by the
Ga180 repressor, binding site competition between the Engrailed repressor and the Fushi-
tarazu activator, and blocking NFKB import into the nucleus by binding to the Ich

repressor [29-3 1].

Direct repression of the basal transcriptional machinegy

Some transcriptional repressors can disrupt transcription by direct interactions
with the basal machinery in a promoter-specific manner (Figure 1-2B) [16]. Repressors
can target the basal machinery by three known mechanisms. The first is through
modification of the locally bound Pol 11 holoenzyme (Figure 1-ZB, inset 1). For example,
the Tupl-Ssn6 repressor complex has been suggested to mediate repression by interacting
with the SrblO/ 11 cyclin/cyclin—dependent-kinase pair to misregulate Pol II CTD
phosphorylation [32-34]. The second mechanism is by blocking TBP binding to the
TATA box (Figure 1-2B, inset 2) [16]. The homeodomain-containing repressor Eve has
been shown to interact with TBP and disrupt formation of the pre-initiation complex
(PIC) [35]. Finally, repression can also involve inhibiting the protein-protein interactions
between the general transcription factors (GTF) (Figure I-ZB, inset 3) [16]. While
monomeric Kruppel can activate transcription, 3 Kruppel dimer can mediate repression

by interacting with the TFIIEB subunit and preventing PIC assembly [3 6].

Recruitment of chromatin modifying and remodeling complexes

Transcriptional regulation in eukaryotes takes place in a highly complex
environment consisting of a dense DNA/protein matrix called chromatin. Typical
eukaryotic genomes are in the order of 106-109 bp, which uncoiled would be 101-102 cm
long. Since an eukaryotic cell is approximately 10-100 um long, DNA must be tightly
packaged in order to fit inside the nucleus. The basic packaging unit is called the
nucleosome, which is a highly stable complex consisting of 147 bp of DNA wrapped 1.67
lefi-handed turns around an octamer of histone proteins (HZA, H2B, H3, and H4) [37].
The linker histone H1 loosely associates with the nucleosome and promotes chromatin
compaction [3 8]. When viewed under an electron microscope, chromatin has the
appearance of an ll-nm “beads on a string”, with the distance between each nucleosome
varying between 10-100 bp [38]. In turn, chromatin can be packaged in higher order
secondary and tertiary structures [39, 40]. Early studies on chromatin structure classified
chromatin as euchromatin or heterochromatin depending on the degree with which they
stained with DNA-specific dyes like Giemsa [41]. The lighter staining euchromatic
regions are characterized by being relatively gene rich and have a looser structure [40]. In
contrast, heterochromatic regions stain darkly, contain fewer active genes, and are highly
condensed [40]. Since then, studies have shown that euchromatin and heterochromatin
differ at a nucleosomal and DNA level, with specific covalent modifications of both
histones and DNA nucleotides associated with each region [16, 40].

Many repressors work by altering the chromatin structure in order to create a
repressive state (Figure 1-2C). A common mechanism is through the recruitment of

histone modifying enzymes such as histone deacetylases (HDACs) and histone

methyltransferases (HMTs) [26]. Active genes are generally associated with high levels
of histone acetylation, possibly because acetylation of the N-terminal histone tails might
decrease the stability of the nucleosome complex [42]. Alternatively, acetylation might
target the binding of chromatin remodeling complexes such as SWI/SNF to create a
looser chromatin structure [43]. Thus, recruitment of HDAC complexes like de3, Hdal,
and Sir2 can decrease histone acetylation and create a more compact chromatin structure
[26]. Histone methylation is often associated with gene silencing [44]. Recruitment of
HMTs such as Su(var)3-9 and SETl can result in the specific methylation of the lysine
residues H3 K9 and K27 respectively [26, 44]. Methylation of these residues can in turn
promote the binding of repressive complexes like HPl and PC to silence gene activity
[44]. Other known histone modifications include phosphorylation, ubiquitination,
sumoylation, ADP ribosylation, deimination, and proline isomerization [44].
Additionally, in many organisms DNA methylation plays an important role in gene

silencing [26].

10

0
0“ ii A

i Me Me Me
EL Me

Figure 1-2. Mechanisms of transcriptional repression.

(A) Disruption of activator function can occur by degradation (1), covalent modification
(2), translocation outside of the nucleus (3), sequestration (4), competition for binding
sites (5), or blocking of activator interactions (6). (B) Alternatively, repressors can
directly target the basal machinery by modifying the Pol II CTD (1), preventing TBP
binding to the TATA box (2), or by blocking protein-protein interactions necessary to
promote PIC assembly (3). (C) Additionally, repressors can recruit chromatin remodeling
and modifying enzymes in order to create a repressive chromatin structure. Images in this
dissertation are presented in color.

Lia—iry

The hairy (h) pair-rule gene encodes a 337-amino-acid transcriptional repressor
that plays an important role in segmentation and sensory bristle formation in Drosophila
[45—47]. Hairy is closely related to a family of proteins belonging to the Enhancer of split
complex [E(spl)], which negatively regulate neurogenesis and are regulated by Notch
signaling [48, 49]. Homologs for the Hairy/E(spl) (HES) have been found in both
vertebrates and invertebrates, where they are involved in neurogenesis, neural cell fate,
vascular development, mesoderm segmentation, and myogenesis [1, 50-53]. Hairy/E(spl)
proteins are characterized by a conserved basic helix-loop-helix (bHLH) DNA-binding
domain, an Orange repression domain, and a C-terrninal WRPW motif that binds to the
corepressor Groucho (Figure 1-3A) [54, 55]. Interestingly, studies of the mammalian
hairy/E(spl) homolog HES6 suggests that the WRPW motif also acts as a protein
degradation signal, promoting polyubiquitination and proteosomal degradation [56].
Hairy/E(spl) proteins are believed to bind to their targets as homodimers or heterodimers
through interactions in the bHLH and Orange domains, with the Orange domain
providing functional specificity [54, 57, 58]. However, yeast two-hybrid studies have
shown that Hairy does not interact with any of the E(spl) proteins and binds presumably
as a homodimer [57].

Initially Hairy is expressed in a seven stripe pattern in early Drosophila
blastoderrn embryos where it helps to establish the borders of alternate metameric
segments in the fly by repressing the expression of the pair-rule genes fiuhi-tarazu, runt,

and oddskz’pped (Figure 1-3B) [59, 60]. Later, in the larval stage, Hairy is involved in the

12

b HLH Orange

 

 

c 4"”"1.

‘1.“ at’

 

 

 

Figure 1-3. Conserved Hairy domains and expression pattern.

(A) Schematic diagram of Hairy showing the basic helix-loop-helix DNA-binding
domain (bHLH), the Orange domain, and the C-terrninal WRPW motif. Arrows point to
the regions involved in interacting with the Sir2, CtBP, and Groucho corepressors. (B) in
situ of Hairy mRNA expression pattern in early blastoderrn embryos. Images in this
dissertation are presented in color.

patterning of the sensory bristles in wing and leg imaginal discs through direct repression
of the proneural gene achaete [61].

In contrast to short-range repressors, Hairy is able to silence gene activity even
when bound 1 kb away from a heterologous enhancer [22]. However, in some cases
Hairy can allow enhancer autonomy and act in a non-dominant fashion to locally repress
enhancers [62]. The “hot chromatin model” has been proposed to explain the context
dependency in the range of Hairy repression, this model postulates that Hairy binding to
single sites might require the presence of nearby activators to “open” the chromatin.
Alternatively multiple Hairy binding sites placed close together could enable cooperative
protein-protein interactions, allowing Hairy to bind in the absence of activators [62].

Interestingly, while the long-range repression ability of Hairy has been studied
only in the context of transgenic reporter constructs, the few promoter sequences that
have been analyzed have Hairy binding sites placed very near activators or the
transcriptional start site [22]. For example a single Hairy binding site at -302, 50 bp
upstream of a cluster of 3 activator sites in the promoter region of achaete has been
shown to be critical for repression [61]. Similarly, the mouse Hairy/E(spl) homolog HES-
1 can negatively auto-regulate its own transcription by binding to 4 sites placed within
20-170 bp of the transcriptional start site [63]. Additionally, repression of sex lethal by
the Hairy-related protein Deadpan occurs by direct binding to a region between 93-125
bp upstream of the start site [64]. Therefore, it is unclear whether the long-range
repression activity of Hairy is actually required for silencing endogenous targets.
Interestingly, the Dorsal morphogenic protein, which also interacts with Groucho and has

long-range repression activity, can repress the zerknult gene by binding a silencer

14

element approximately 1 kbp upstream of the start site [65]. However, since it is not
known what activates this gene at the early blastoderm stage, it is difficult to determine if

Dorsal is behaving in a long-range fashion with respect to the activators.

Haigy/Egspl) repression mechanism

Basic HLH proteins bind as homo- or heterodimers to a consensus CANNTG
sequence known as an E-box. Both Hairy and E(spl) proteins have been shown to bind to -
E-boxes. However, Hairy homodimers prefers to bind to an alternate CACGCG sequence
[47, 61]. Interestingly, the E(spl) proteins and a heterodimer between the bHLH
activators Daughterless and Lethal-of-scute can both bind to the same consensus E-box in
vitro, suggesting that binding site competition could be a possible repression mechanism
in some contexts [66].

A conserved characteristic of Hairy/E(spl) proteins is the ability to interact with
the Groucho corepressor through a WRPW motif at the C-tenninal end [55]. Multiple
lines of evidence have shown that Groucho plays a key role in mediating Hairy repression
[67, 68]. However, Groucho binding to Hairy is not the sole repression mechanism since
Hairy can still repress scute even without the WRPW motif [54]. Indeed, the corepressors
CtBP (C-terminal binding protein) and Sir2 (silent information regulator 2) have been
shown to interact biochemically and genetically with Hairy [69-71]. CtBP, like Groucho
is also a conserved transcriptional corepressor present in both vertebrates and
invertebrates, but unlike Groucho, CtBP is thought to mediate short-range repression by
binding to PXDLS-type motifs present on short-range repressors such as Knirps, Snail

and Kruppel ([72-74]. The Hairy CtBP binding region is located only 9 amino acids from

15

the C-terrninal WRPW and corresponds to a divergent PLSLV motif with weaker binding
affinity than the consensus [69, 72]. The functional importance of CtBP-Hairy interaction
is unknown, but genetic evidence suggests that CtBP can antagonize Hairy repression, as
reduction in CtBP levels suppresses the Hairy mutant phenotype [69]. Additionally,
Hairy mutations that enhance CtBP binding inhibit Hairy-mediated repression, while
mutations that block CtBP interaction have the opposite effect [72]. Since the CtBP
binding region is very close to the C-tenninal WRPW, it has been suggested that CtBP
and Groucho compete for binding to Hairy. However, both CtBP and Groucho can
simultaneously bind a Hairy monomer in vitro [70, 72]. Intriguingly, CtBP seems to be
required for proper expression of three putative Hairy target genes identified by DamID
profiling, wherein tissue culture expression of a Hairy protein fused to the Dam
methyltransferase leads to the methylation of the DNA surrounding the sites of Hairy
recruitment [75]. However, it is unclear whether the CtBP requirement is direct or
indirect.

Unlike the effect seen with CtBP, reduction in the maternal dSir2 levels causes an
increase in the Hairy mutant phenotype, suggesting a positive contribution to Hairy
repression [71]. In vitro binding assays have shown that dSir2 can bind to Hairy, and that
the region of Hairy required corresponds to a RRAR motif present in the basic N-
terminal part of the bHLH domain [71]. Additionally, damlD profiling in Kc cells,
wherein Hairy and dSir2 were fused to the DNA adenine methyltransferase (dam),
revealed a 30% overlap between Hairy and dSir2 targets [75]. It seems likely that dSir2-
Hairy interaction is important for Hairy repression, because a similar biochemical and

functional interaction was shown for two human Hairy/E(spl) homologs hHESl and

16

hHEY2, indicating that dSir2-Hairy interaction is evolutionary conserved [76]. Sir2 was
first isolated as a gene involved in the repression of the yeast silent mating type locus;
later it was shown to be involved in the establishment of silent heterochromatic states at
yeast telomeres and rDNA [77, 78]. The molecular mechanism of Sir2 silencing seems to
involve a NAD-dependent histone deacetylation activity of Sir2 and the spreading of a
Sir2 complex over large distances [79]. This is accomplished by stepwise recruitment by
Sir2 of Sir3 and Sir4 to nearby hypoacetylated histones, which in turn recruit more Sir2,
generating a positive feedback loop [80]. This type of mechanism would be compatible
with the long-range repression activity of Hairy.

The Groucho (Gro) corepressor is a widely conserved metazoan protein that is
recruited by a large number of transcriptional repressors. Groucho does not have any
intrinsic DNA-binding ability, instead relying on direct protein interactions to bind to the
target gene. Two types of signature motifs have been found that interact with Groucho,
the C-terminal WRPW and a FxIxxIL motif called the Engrailed homology-1 (ehl) [81,
82]. In humans, Groucho homologs are called Transducin-like-enhancer-of—split (TLE)
and share with Groucho a glutamine—rich N-terminal tetrarnerization domain (Q-domain)
and seven WD repeats, each 40 amino acids long (WD-40), placed in tandem near the C-
terminus, which are thought to be involved in protein-protein interactions [83]. Gro/T LE
proteins exert their influence on gene activity by multiple mechanisms. Like Sir2,
Groucho-mediated repression appears to involve histone deacetylase activity, in this case
through direct interactions with de3 [84]. Additionally, Groucho binds hypoacetylated
histone tails, suggesting a repression mechanism analogous to that of Sir2, wherein

Groucho spreads along the chromatin by means of a positive feedback loop involving

17

de3-mediated deacetylation of surrounding histones that could allow more Groucho
binding [85]. Post-translational modifications of Groucho play a role in its repression
activity, as phosphorylation of Groucho by dHIPK2 can relieve target gene repression
[86]. Interestingly, phosphorylation of mammalian TLE by CK2 on Ser 239 can enhance
its repression activity [87]. Genome-wide studies of Hairy, CtBP, Sir2, and Groucho
binding have shown that Hairy colocalizes with different combinations of cofactors at
multiple loci, suggesting that Hairy can repress by alternate mechanisms via context-
dependent cofactor recruitment [75].

The ability of HLH proteins to form homo- or heterodimers raises the possibility
of an additional repression mechanism involving heterodimerization between
Hairy/E(spl) proteins and bHLH activators in order to titrate them off the DNA. An
example of such a mechanism is the ability of the protein Extramacrochaetae (Emc) to
block DNA binding of Daughterless (Da) and members of the Achaete-Scute Complex
(AS-C) by heterodimerization through the mutually conserved HLH domain [88]. Some
Hairy/E(spl) proteins can heterodimerize with Da/AS-C, however it is unlikely they
work in this manner in viva since target gene repression is abolished when the
Hairy/E(spl) binding sites are mutated indicating that the repressor needs to be present at
the gene for repression to occur [47, 61]. Furthermore, addition of Hairy/E(spl) proteins
does not displace Da/AS-C from E-boxes in vitro [64, 88]. Interestingly,
heterodimerization with activators does seem to play a role in repression for some
vertebrate Hairy/E(spl) homologs. For example, HESS can heterodimerize with the E47
HLH activator and prevent it from binding to DNA in vitro [89]. Similarly, the related

Heyl repressor can heterodimerize with MyoD to block myogenic terminal

18

differentiation, while Hey2 can block DNA binding by the ARNT/EPAS (HIFZ) complex
and suppress VEGF transcription [90, 91]. Likewise, HEY], HEY2, and HESl can
interact with the PTF 1-p48 activator to block exocrine cell differentiation in the

developing pancreas [92].

Haig/Eppl) functions in Drosophila
The high degree of functional conservation of Hairy/E(spl) repressors and bHLH

activators that exist among different species has led some to speculate that both form part
of a functional “cassette” involved in the establishment of cell fate decisions [93]. The
best studied of these functions involves their role in prepatteming proneural clusters and
subsequent cell fate determination within the cluster. A model system for this process is
the formation of the Drosophila sensory bristles where precursor cells gives rise to shaft,
socket, glial, and neuronal cells that make up part of the peripheral nervous system (PNS)
[94]. Establishment of the proneural fate occurs through the activation of the proneural
genes achaete, scute, lethal of scute, and asense belonging to the AS-C, as well as
daughterless and atonal [95-97]. These activators belong to a class of bHLH proteins that
typically function in opposition to the Hairy/E(spl) repressors to establish cell fate
decisions. Along with hairy/E(spl), they have coevolved to regulate cell differentiation in
multiple tissues for both vertebrates and invertebrates [93]. In the case of PNS formation,
repression of these genes by members of the Hairy/E(spl) family has been shown to
prevent the formation of ectopic sensory organs [98-100] (Figure 1-4A). Initially,
prepatteming of the PNS is brought about by direct repression of achaete by Hairy in

cells outside of the proneural cluster, resulting in a complementary pattern between hairy

l9

and achaete gene expression [47] (Figure 1-4B, C, and D). Cells inside the proneural
cluster belong to a so-called equivalence group, where any one of them has the potential
to become a precursor cell [101] (Figure 1-4D). Only one cell within the equivalence
group will then go on to become a precursor cell, with the rest assuming an epidermal
cell fate by means of lateral inhibition of proneural genes by E(spl) proteins in response
to Delta/Notch signaling [95, 102] (Figure 1-4D).

In addition to regulating the formation of sensory bristles, hairy also plays a role
in the differentiation of the eye. Like other arthropods, flies have compound eyes
characterized by an array of repeating units called omatidia, each of which acts as an
independent photoreceptor. In the Drosophila eye imaginal discs, development of each
ommatidium starts with the differentiation of the R8 photoreceptor, which then recruits
the surrounding cells to form the lens, cone, pigment cells, and visual receptor that make
up the ommatidia [103]. The differentiation of the R8 cells is temporally and spatially
controlled by a wave of differentiation called the morphogenetic furrow that sweeps from
the posterior to the anterior of the eye imaginal disc [104]. Hairy has been shown to be
part of a network of genes regulating the progression of the morphogenetic firrrow, acting
in conjunction with emc immediately ahead of the furrow to place undifferentiated cells
in a pre-proneural state [104].

Apart from their role in neurogenesis, Drosophila hairy/E(spl) genes are also
important in patterning of the somatic musculature. Analogously to PNS development,
muscle formation involves the establishment of a pluripotent field of cells expressing
high levels of the mesodermal specific twist activator [105, 106]. The regulatory

mechanism by which twist gene expression is prepattemed in the mesoderm has not been

20

Figure 1-4. Hairy/E(spl) proteins pattern the PNS in Drosophila.

(A) Hairy mutants lead to formation of ectopic bristles which form part of the Drosophila
peripheral nervous system [107]. (B) Hairy prepattems the PNS by repressing the
proneural gene achaete. (C) Schematic diagram of the achaete promoter showing binding
sites for the Hairy repressor and the HLH activators AS-C and Daughterless. (D)
Representation of the PNS patterning process by Hairy/E(spl) proteins. Hairy along with
emc initial represses achaete expression in cells outside a proneural cluster. A single cell
within the proneural cluster is selected to become the precursor cell. Selection occurs
through the process of lateral inhibition via the Notch signaling pathway. The resulting
precursor cell expresses genes that will lead to formation of the bristle sensory organ.
Images in this dissertation are presented in color.

21

a Hairy
E Achaete

 

 

Achaete

       

a é
W Lateraunhibitm W
Hairy Achaete-scute E(spl) Deadpan
emc atonal Asense

Daughterless

Figure 1-4. Hairy/E(spl) proteins pattern PNS formation in Drosophila.

22

fully elucidated, but evidence points to a role of Notch signaling both in the activation of
twist inside the field as well as repression outside it through the actions of Supressor of
Hairless [Su(H)] and members of the E(spl) class of repressors [108]. Once the
pluripotent field has been formed, lateral inhibition of myogenic genes such as lethal of
scute by E(spl) proteins in response to Notch activation, allows for a single precursor cell
to be selected, which then divides into two myogenic founder cells that can in turn recruit
surrounding cells to form the muscle [109, 110].

The Drosophila Malpighian tubes are part of the fly excretory system and form as
invaginations between the hindgut and the foregut [111]. Following the same theme as
before, development of the Malpighian tubes are characterized by the formation of an
equivalence group of 6 to 8 cells that express the AS-C genes, followed by restriction of
expression to a single non-dividing cell that migrates to the tip of the developing tube and
regulates the invagination and growth of the tube [112]. It is likely that the selection of
the tip cell involves E(sp1) proteins, because mutations in the Notch pathway have shown
misregulation of AS-C expression [112].

Deadpan is a transcriptional repressor involved in sex determination that is
structurally related to the Hairy/E(spl) class, in that it contains a bHLH DNA-binding
domain an Orange domain as well as a conserved WRPW motif [64, 100]. In flies, gender
is determined by the ratio of X to autosomal chromosomes (X:A), where X:A = 1 leads to
a female and a X:A = 0.5 leads to a male [64]. The genetic mechanism involves the
transcriptional repression of the female-specific sex-lethal gene by Deadpan in response
to the relative difference in X-linked sex determination genes or “numerator elements”

such as scute (also known as sis-B), sis-A, sis-C, and mat [113].

23

Hairy has been found to regulate the formation of the Drosophila salivary gland
secretory tubes, which are unbranched tubular structures that connect the salivary duets
with the larval mouth. Hairy works to suppress the formation of additional branches by
repressing genes required for apical membrane grth such as huckebein, in cells outside
the growth zone [114].

In summary, an important role of Hairy/E(spl) proteins in Drosophila is to
determine the outcome of binary cell fate decisions (Figure 1-4D). The process involves
establishment of an equivalence group of cells with the potential to assume a specific cell
fate. Cells outside the equivalence group are prevented fi'om assuming the same fate
through repression of tissue-specific genes by Hairy. A single cell within the equivalence
group then is selected to become the precursor cell of a particular cell type. Selection of
the precursor cell is done by lateral inhibition of precursor-specific genes by E(spl)

proteins as a response to Notch signaling.

Haigy/E1spl) functions in vertebrates
Vertebrate Hairy/E(spl) homologs (HES) are involved in many of the same

processes seen in flies such as neurogenesis, myogenesis, and mesoderm segmentation, as
well as roles in angiogenesis, and oncogenesis [1]. Thus, understanding the molecular
basis by which these proteins work in flies would shed light on the transcription
regulation of multiple developmental pathways in a wide range of organisms.

Paralleling the fly system, mammalian Hairy/E(spl) homologs HESl and HESS
work to select neuronal precursors from a population of pluripotent cells in the

developing nervous system [115]. Genetic interaction studies using mouse models

24

suggest that HES protiens function to suppress the expression of proneural genes such as
achaete-schute homolog MASH] and neurogenin] (ngnI) in response to Notch signaling
[115-117]. Similarly, the zebrafish E(spl) homolog HER4 can inhibit neurogenesis by
repressing ngnI, while the Xenopus Enhancer-of-split related 1 (ESRl) protein can
repress the proneural gene Xath5 [118, 119]. In addition to their prepatterning role in the
nervous system, HES], HES5, and the hairy/E(spl) related heer (also known as Hey2)
have been implicated in determining the cell fate of neuronal precursors by dictating
whether they will become neurons or glial cells [120-122]. In contrast to the other HES
proteins, HES6 is expressed in neuronal precursors and promotes neuronal differentiation
by forming a non-functional heterodimer with HES] [123]. Along with their role in
differentiating neurons, HES genes are also crucial for the maintenance of an
undifferentiated neuronal stem cell population [124]. Overexpression of HES] or HES5
in the developing mouse brain causes an overproliferation of undifferentiated neural stem
cells, while knockdowns of HES], HES3, and HES5 cause premature neuronal
differentiation and brain malformations [125, 126]. Recently, HES] has also been
implicated in the formation of brain boundaries such as the zona limitans intrathalamica
and the isthmus, which separate thalamus/prethalamus and midbrain/hindbrain
respectively [127].

Another example of the importance of HES proteins in cell fate decisions can be
seen in the development of the vertebrate blood circulatory system, where these proteins
dictate whether cells will adopt a venous or arterial fate [128]. Evidence in zebrafish has
shown that hypomorphic mutations or knockdowns of heer result in defects in dorsal

aorta formation, which is correlated with a decrease in the arterial marker ephrinBZa and

25

an increase in the venal marker ephB4 [129]. The functional role of HES proteins in
blood vessel formation is likely to be conserved in vertebrates because mutations of the
mouse heer and heer likewise result in a loss of arterial identity [130, 131]. Analogous
to other developmental processes, HES proteins act by repressing genes involved in vein
formation such as VEGF-A and VEGFR2 in response to Notch signaling [91, 132, 133]. It
is interesting to note that the process of blood vessel formation in vertebrates is similar to
the development of salivary ducts and Malpighian tubes in Drosophila, both of which are
tubular structures. Since flies and other arthropods have an open circulatory system, one
can speculate that evolution adapted a preexisting functional genetic cassette for
generating tubular structures in order to develop blood vessels. The roles that HES
proteins play in blood vessel formation suggest that they could be possible targets for
cancer therapies seeking to block angiogenesis in developing tumors.

Recent studies have also implicated HES proteins in the development of proper
heart morphology. Here, HES proteins are again involved in regulating cell fate decisions
by specifying a chamber myocardium identity versus an atrioventricular canal (AVC) and
inner curvature (IC) fate [134]. In chick embryos, the HES-related proteins Heyl and
Hey2 firnction to repress the gene Bmp2 in cells destined to become part of the chamber
myocardium [134]. Bmp2 is exclusively expressed in the AVC/IC where it activates the
gene be2 which in turn represses Heyl and Hey2 expression; this feedback loop
allowing a sharp demarcation between the AVC/IC and the chamber region [134].
Interestingly, only Hey] responds to Notch signaling, which makes it the direct effector

in this context [134]. The regulatory mechanism for negatively regulating AVC/IC cell

26

fate seems to be evolutionary conserved, as similar mechanisms have been reported for
both zebrafish and mouse [134, 135].

In addition to their role in patterning the heart AVC/IC regions, HES proteins
have recently been discovered to be involved in valve and septum formation. Double
mutants of mouse HES homologs Hey] and HeyL show severe malformations in the
ventricular septum as well as in the puhnonary and atrioventricular valves. This was due
to a failure to establish a mesenchymal fate by the epithelial cell precursors in the AVC.
Establishment of the mesenchymal cell fate is likely to involve Notch signaling, since
Notch], Notch2, and the notch ligand Jagged] are coexpressed with Hey] and HeyL in
the endocardium of the AVC. Additionally, a Notch] knockout abolishes the epithelial to
mesenchymal transition [136]. It is interesting to note that Heyl and Hey2 also seem to
have a role in preventing calcification of the aortic valve (an important cause of heart
disease in humans) by repressing the activity of the osteoblast specific Runx2
transcriptional activator and downstream target gene expression [137]. It is likely that a
similar mechanism operates in humans because Notch] mutations have been shown to
lead to aortic valve defects [137].

In the vertebrate embryo, patterning of the mesoderm occurs through a process of
sequential segmentation from a terminal grth zone that leads to the development of
groups of mesoderrnal cells called somites placed symmetrically along the neural tube
[138]. In contrast, long gerrnband insect segmentation like that of Drosophila occurs
simultaneously by the parallel formation of seven stripes of gene expression from the
pair-rule genes hairy, even-skipped, and fushi-tarazu in a syncytial embryo [45, 139].

Curiously, despite the difference in segmentation, Hairy functions in both insects and

27

vertebrates to prepattem the segmental body plan. Hairy regulates sonritogenesis as part
of a molecular oscillator termed the “segmentation clock”, which generates waves of
transcriptional activity that move fi'om the posterior to the anterior in the presomitic
mesoderm (PSM) [140]. For example, the chicken hairy homologs hairy], hairy2, and
Hey2 are each expressed cyclically in the PSM, with hairy] and HeyZ expressed in the
posterior or caudal half of the developing somite, while hairy2 is expressed in the
anterior or rostral half [138, 141, 142]. Similarly, mouse HES] and HES7 also show
oscillatory expression patterns as well as the zebrafish homolog her] [141, 143, 144].
The molecular mechanism by which these oscillations occur has not been completely
elucidated, but evidence points to the involvement of both extrinsic signaling via the
Notch pathway, as well as cell-autonomous oscillations consisting of a negative feedback
loop of HES proteins repressing their own transcription and subsequent proteosome-
mediated degradation and reactivation [145-147].

As in Drosophila, HES proteins have also been implicated in skeletal muscle
formation [90, 148, 149]. In the Xenopus embryo, overexpression of Hairyl leads to
repression of the myogenic inducing factor MyoD, while expression of a Hairy-VP16
chimera leads to MyoD upregulation, suggesting that Hairyl binds directly to the MyoD
promoter and represses it [148]. Furthermore, HESl was shown to inhibit myogenesis in
both rat and mouse derived cell lines [50, 150]. Also, another HES member CHF2 was
shown to block Myogenin activation by MyoD in developing myoblasts by
heterodimerization with the activator [90]. In contrast, mouse HES6 seems to be able to
stimulate myogenesis by repressing transcription of the myogenic repressor MyoR [149].

Like other developmental processes where HES proteins are involved, it is likely that

28

Notch signaling is also active in myogenesis; overexpression of the Notch ligand 0111 in
chick limb buds arrests myogenic differentiation and decreases MyoD expression [151].
Additionally, Notch signaling can inhibit myogenesis in both Xenopus embryos as well as
in frog- or mouse-derived cell lines [152, 153]. The importance of HES proteins in
myogenesis raises the possibility of a possible role in age or disease related muscle
deterioration.

Another recently discovered role of HES proteins is in the developing thymus
where they are required for proper T-cell differentiation and proliferation. In mice, HES]
mutants mostly lack a thymus and show arrested T-cell differentiation at an early stage
due to a lack of precursor cell proliferation [154]. It is not clear exactly how HESl
regulates T-cell differentiation, but it has been shown that HESl directly repress the CD4
coreceptor by binding to a silencer element in the first intron of the gene [155, 156].
Proper transcriptional regulation of both CD4 and the similar coreceptor CD8 are crucial
for T-cell differentiation and subsequent maturation [155]. Regulation of precursor
proliferation has been suggested to occur through direct repression of the cyclin-
dependent kinase inhibitor p27(Kip1) [157].

HES proteins have also been implicated in the development of the digestive
system. For example, in the developing pancreas HESl acts to maintain pancreatic stem
cells by repressing Ptf] a, which promotes exocrine differentiation, as well as by
repressing Ngn3, which promotes endocrine differentiation. HES] knockout mice show
ectopic expression of Ptf] a and Ngn3, which leads to the formation of an extra pancreas
[158, 159]. Similarly, HESl is also important for the maintenance of the stem cell

progenitors in the mouse intestinal epithelium by repressing MASHl [160-162]. Loss of

29

HES] activity causes the differentiation of intestinal stem cells into an endocrine cell
fate, suggesting an involvement in specifying enterocyte versus non-enterocyte fates
[160, 161, 163, 164]. Additionally, HES] knockout mice show lack of bile duct
formation in the liver, suggesting a role of HES] in establishing a hepatocyte versus
biliary epithelial cell fate [165].

In general HES proteins seem to have two crucial functions, the first one being to
maintain stem cells in an undifferentiated proliferating state, while the second one
involves regulating the outcome of binary cell fate decisions [124]. Disruption of HES
activity in either of these functions has been shown to have both a positive and negative
impact in oncogenesis. For example, some evidence suggests that HES] misregulation
could be involved in the development of sympathetic neuroblastomas, as neuroblastoma
cell lines show almost no HES] expression [166]. Furthermore, differentiating
neuroblastomas show an upregulation of HES] and a downregulation of the proneural
gene HASH] which is expressed in the sympathetic nervous system [166]. Similarly,
studies have shown that activation of the Notch pathway and the downstream HES]
effector can lead to inhibition of cellular proliferation in both gastrointestinal and lung
carcinoma cells and direct repression of HASH] [167-170]. In contrast, inappropriate
Notch signaling and HES activation is thought to promote the development of several T-
cell leukenrias by promoting undifferentiated precursor cell proliferation [171-174].
Likewise, inhibition of Notch signaling leads to differentiation of mice intestinal
adenomas into post-mitotic goblet cells presumably by suppressing HES] expression

[175].

30

Q11]

The yeast Ga14 protein has been used as a model for transcriptional activation for
the past 20 years. In yeast, Gal4 firnctions to activate the expression of the GAL regulon
when galactose is the sole carbon source [176]. Pioneering studies from several labs have
led to a significant level of understanding of the molecular mechanisms by which Gal4
exerts its activity. In Drosophila, the Gal4/UAS system for the spatial and temporal
control of gene activity has been a powerful tool for elucidating gene function [177].

G314 is an 881-amino-acid protein consisting of a Zn—Cys binuclear cluster-type
DNA-binding domain, a linker domain, a dirnerization domain and two acidic activation
domains [176]. Gal4 binds as a homodimer to a 17-mer palindromic sequence 5’ CGG
AGG AC(T/A) GTC CTC CG 3’ called the upstream activating sequence (UAS) [178].
DNA binding by Gal4 has been shown to be cooperative, with multiple UAS sites
showing synergistic activation [179].

The mechanism of Gal4 activation seems to involve both the creation of an
“open” chromatin state by recruitment of chromatin remodeling and histone modifying
enzymes, as well as direct targeting of the Mediatior complex and Pol 11 holoenzyme to
the promoter through interactions with its activation domain [180]. For example, in vitro
studies have shown that Gal4 can interact with both TBP and TFHD [181, 182].
Similarly, biochemical and genetic interactions with Gal4 have been shown for Cdk8,
Srb4, and Gall] [183-185]. In addition, evidence has shown that Gal4 can also interact
with the SAGA histone acetylase (HAT), as well as with the SWI/SNF chromatin
remodeling complex [186, 187]. Interestingly, recent work has also implicated the

proteosome subunits Sugl and Sug2 as direct targets for Gal4 [188, 189].

31

The yeast Gal4 protein belongs to the class of “acidic” activators which includes
the mammalian glucocorticoid receptor, c-myb, the viral VP16 protein, and the yeast
Gcn4 [190-192]. Acidic activators are characterized by having short modular domains
with a highly charged polar residues interspersed with hydrophobic groups [190]. The
ubiquitous nature of these domains and the fact that they can fimction across species
suggests an evolutionary conserved activation pathway. In Drosophila, the maternal
morphogen bicoid, which specifies the anterior half of the embryo, also contains an
acidic activation domain at the C-terrninus [193]. Interestingly, expressing a chimeric
protein containing the DNA-binding domain of Bicoid firsed to a heterologous acidic
activation domain can completely rescue the anterior pattern in a bicoid mutant embryo

[194].

Gal4 transcriptional activation

Multiple lines of evidence have implicated SAGA as one of the main players in
Gal4-mediated activation. Indeed, the importance of SAGA to gene activation has made
it into one of the most well characterized HAT complexes. SAGA, which stands for Spt-
Ada—anS-acetyl transferase complex, is a 1.8 MDa complex containing the histone
acetyltransferase Gcn5 as well as three other groups of proteins [195]. The first class
consists of the Spt proteins Spt3, Spt7, Spt8, and Spt20/Ad35 that affect TBP function
[195]. Spt3 and Spt8 have been shown to directly interact with TBP, suggesting a direct
role in PIC assembly [195, 196]. The second class belongs to the Ada transcriptional
adapter proteins Ada], Ada2, Ada3. Previous work has shown that Ada2 and Ada3 are

crucial for Gcn5 HAT activity in vivo and can also interact with TBP [197-199]. The

32

third class corresponds to a group of TBP-associated factors (TAFs) and Tral, the latter
which is shared with the NuA4 HAT complex [200, 201]. Recently, it was found that the
deubiquitylase pr8 is also a part of the SAGA complex, this protein plays an important
role in regulating histone H2B ubiquitylation [202].

Chromatin immunoprecipitation studies have shown that recruitment of SAGA is
one of the earliest steps in an ordered process leading to Mediator recruitment and
subsequent assembly of the pre-irritiation complex (PIC) [203, 204]. In vivo imaging
using fluorescence resonance energy transfer (FRET) has shown that Gal4 interacts
directly with the Tral subunit of SAGA [204]. This interaction requires the presence of
Spt20, a critical SAGA component required for complex integrity, suggesting that Tral
needs to be in a complex to interact with Gal4 [204]. Interestingly, binding of Tral to
Ga14 requires the presence of both the Gal4 DNA binding- and activation domains,
suggesting that Ga14 needs to be bound to the UAS in order to interact with Tral [204].
In support of a direct Gal4-Tral interaction is the fact that SAGA recruitment by Gal4
can occur on a minimal UAS site and in the absence of a promoter [186]. Curiously, it
seems that in some contexts Gcn5 is not required for PIC assembly, raising the possibility
that SAGA actually acts as a scaffold for recruitment of the Mediator and Pol 11
holoenzyme [186]. It should be noted however that SAGA HAT activity seems to be
important for the activation of some yeast genes such as PH05 and PH08 [205, 206]. In
addition to its catalytic activity, Gcn5 also contains a bromodomain that interacts with
acetylated lysine residues [207]. Irnportantly, functional homologs of the SAGA complex
have been characterized in Drosophila as well as in other metazoans, suggesting a

conserved mechanism of transcriptional activation [208].

33

In addition to SAGA, Gal4 activity has also been associated with recruitment of
the SWI/SNF chromatin remodeling complex [187, 209]. SWI/SNF is an ATP-dependent
chromatin remodeling complex that plays a key role in “opening” the chromatin to
facilitate transcription. The exact mechanism of function is unclear, but recent in vivo
evidence suggests that it acts by directly displacing the histone octamers from the DNA
[210]. SWI/SNF consists of 1] subunits, of which the Swi2/Snf2 contains the ATP-
dependent catalytic activity. In addition, SWi2/Snf2 also has a bromodomain, suggesting
that histone acetylation can enhance SWI/SNF activity. Indeed, SAGA and SWI/SNF
complexes seem to work in concert for the activation of many genes [43]. Furthermore, it
has been shown that for some yeast genes like H0, SAGA-dependent histone acetylation
is required prior to SWI/SNF recruitment [211]. Additional evidence of the functional
importance of SWI/SNF in Gal4 activity comes from studies looking at the GAL genes.
In the inactive state the promoter region of the GAL genes contains nucleosomes
positioned over the TATA box and the transcriptional start site. Upon galactose induction
these nucleosomes become disrupted in a Gal4-dependent manner, facilitating the
recruitment of Pol II and PIC assembly [212]. It is likely that SWI/SNF is involved in this
process, as chromatin immunoprecipitation studies have shown binding of Swi2/Snf2 to
the GAL1-10 promoter region upon Gal4 activation [213]. Like SAGA, functional
homologs of the SWI/SNF complex have been found conserved from yeast to humans
[214].

Apart from histone acetylation and nucleosome disruption, recent work has
uncovered a role for histone ubiquitylation in Gal4-dependent activation. Transient

monoubiquitylation of H2B K123 was shown to be required for maximum activation of

34

GAL] [202]. H2B ubiquitylation and deubiquitylation seems to be a tightly coordinated
process regulated by the Rad6 ubiquitin-conjugating enzyme and the deubiquitylation
activity of the SAGA subunit pr8 [202, 215]. Targeting of Rad6 to the Gal] promoter is
dependent on the presence of both Gal4 and the E3 ligase Brel [215]. The kinetics of
Rad6 binding follows very closely that of H23 ubiquitylation and seems to occur at a
step immediately prior to SAGA recruitment [215]. After SAGA binding, monoubiquityl-
HZB levels quickly go down as a consequence of pr8 activity [202]. Concomitant with
the loss of monoubiquitylated H2B, Rad6 levels also drop, while SAGA remains bound
to the gene [215]. The functional importance of HZB K123 ubiquitylation seems to lie in
a “trans-tail” process whereby the ubiquitin moiety leads to the mono-, di-, and tri-
methylation of H3 K4 and K79, which are dual purpose marks important for the
activation and silencing of genes [202, 216, 217].

Mechanistic studies on the “trans-tail” process have uncovered close links
between H2B ubiquitylation and transcriptional elongation. For example, Rad6 has been
shown to interact genetically and biochemically with several elongation factors including
Pafl [218]. Furthermore, this interaction seems to be dependent on the presence of both
Brel and the elongating form of Pol H, which is phosphorylated on Ser5 of the CTD
[218]. Moreover, mutations that affect Pol II Ser5 phosphorylation or transcriptional
elongation abolish HZB ubiquitylation and reduce Rad6 binding, with a resulting loss in
H3 K4 and K79 methylation [218]. Like Pafl, the Set] histone methyltransferase (HMT),
which catalyze H3 K4 methylation, can also associate with the elongating Pol II [219].
This association is mediated by Pafl and requires phosphorylation of Pol II Ser5 by the

Kin28 kinase, which is part of the TFIIH complex [219]. It is likely that a similar

35

association occurs with the Dot] HMT, which catalyzes H3 K79 methylation, as Pafl
knockouts result in complete loss of this modification [220]. Interestingly, it has been
shown that H2B ubiquitylation is not required for monomethylation of H3 K4 and K79,
rather it is needed for processive di- and trimethylation [221]. Lastly, recent studies
suggests that phosphorylation of Ser120 of Rad6 by the Burl/Bur2 cyclin-dependent
protein kinase is necessary for firll ubiquitylase activity [222]. Since Burl is required for
efficient transcriptional elongation and is recruited to the elongating Pol H, it suggests a
model where Rad6-Pafl association allows Burl to activate Rad6 [223]. Thus, by
associating with the elongating transcriptional complex, Rad6-mediated histone
ubiquitylation could help target Set] and Dotl activity to the promoter and open reading
frame (ORF). The resulting methylation marks could flag these regions for remodeling by
the Iswl and Chd] chromatin remodelers, which have chromodomains that recognize H3

K4 methylation [224, 225].

36

Chapter 11

Recruitment of Gal4 and SAGA is not affected by Hairy-mediated repression1

Introduction

Transcriptional repression plays central roles in developmental gene regulation,
providing the temporal and spatial specificity required for complex expression patterns.
In Drosophila, the Hairy transcriptional repressor directs patterning of segmental pair-
rule stripes in the blastoderm embryo, and in later stages directs neuronal differentiation.
[47, 61, 98, 226]. Hairy and related transcription factors belong to a conserved metazoan
family of Hairy Enhancer of Split (HES) proteins involved in cell fate decisions in
neurogenesis, vascular development, mesoderm segmentation, and myogenesis [1, 227].
Understanding the molecular basis by which these proteins exert their function will shed
light on the transcriptional regulatory mechanism of multiple developmental processes.

In general, repressors can act by three basic mechanisms disruption of activator
function, direct targeting of the basal machinery, or by chromatin remodeling [16].
Repressors typically use a combination of all three mechanisms in a wide array of
alternate strategies. Thus, no common mechanism is shared by all repressors. However,
the high degree of conservation between the different HES proteins suggests a similar
mechanism of repression for this family.

The focus of this chapter is to investigate possible mechanisms by which Hairy
could disrupt activator function. Disruption of activator function can occur by

suppressing the formation of activator/DNA complexes or by blocking the protein-protein

 

' Parts of this chapter were excerpted from the paper titled “Spreading of corepressor linked to action of
long-range repressor Hairy” by Carlos Martinez and David Amosti, submitted to Molecular and Cellular
Biology (2007)

37

interactions that activators make when bound to the DNA. Therefore, a key question to
answer regarding Hairy repression is whether it is permissive to activator and coactivator
binding. To this end, a highly defined embryo system has been set up so that a reporter
construct behaves homogenously within the embryo in response to repressor and
activator inputs. In order to achieve a tight ON/OFF response a lacZ reporter was placed
under the control of a transcriptional switch. Activation of the reporter was accomplished
by the ubiquitous expression of the yeast Gal4 protein under control of the constitutive
daughterless promoter. Transcriptional repression was brought about by heatshock-
induced expression of a LexA-Hairy fusion protein that lacks the Hairy bHLH DNA-
binding domain as well as the Sir2 interaction domain (Figure 2-1A).

This approach provides several advantages. First, by studying Hairy-mediated
repression in the context of a Drosophila embryo, tissue-specific effects can be
minimized. Additionally, the highly defined nature of the system allows all possible
genotypic combinations of activator and repressor transgenes to be analyzed. Thus, the
individual contributions of the activator and repressor can be elucidated by studying the
promoter state when just the activator or repressor is expressed. Moreover, the
availability of high-grade LexA and Gal4 antibodies allows the tracking of both activator
and repressor binding. Lastly, the activation mechanism of Gal4 has been extensively
studied, simplifying the analyis of Hairy repression by providing an initial set of testable
hypotheses [180]. Thus, the role of Hairy in counteracting each step in Gal4-mediated

activation can be investigated.

38

Materials and Methods

Construction of a transcriptional switch system in Drosophila embryos. A Hairy-
repressible transcriptional switch pC2L5U2L was constructed by placing binding sites for
a LexA-Hairy fusion and the yeast Gal4 activator (UAS) into the P-element
transformation vector pC4PLZ [228], 55 bp from the basal transposase-lacZ reporter and
350 bp from the divergently transcribed mini white reporter. Based on the architecture of
previous modules shown to be effectively repressed in the embryo [229], the 212 bp
regulatory region contains two LexA binding sites inserted 5’ and two LexA sites 3’ of
five high affinity UAS sites derived from a modified UAS-lacZ plasmid [230].
Oligonucleotides bearing 2 LexA binding sites (DA-721/722) were cloned 5’ of 5 tandem
UAS sites into NotI/HinDIII sites and 3’ into an SphI site (DA-641/642) in pBluescript
SK+, and the regulatory switch was inserted as a NotI/SphI fragment into pC4PLZ at -55
and -350 bp from the lacZ and white transcriptional start sites respectively (5’ GCG GCC
G_C_ CTG TAT ATA TAT ACA GCA TCT AGA ACC TGT ATA TAT ATA CAG
AAG CTT GCC TGC AGG T [CGG ACT ACT GTC CTC CGA G]x5 CGG AGA
CTC TAG CAT GG CTG TAT ATA TAT ACA GCA GGT ACC TGC TGT ATA
TAT ATA CAG CAT GC 3’, binding sites for LexA and Gal4 in bold, restriction sites
underlined). This construct was introduced into flies by P-element mediated gerrnline
transformation as described [231] and inserted into the third chromosome. An auto-
activating fly line was then generated by crossing and recombining the reporter line to a
line containing a daughterless enhancer expressing Gal4 ubiquitously in the embryo
(Bloomington Stock Center #5460). Finally, a heat shock-inducible LexA-Hairy fusion

(hsp70 LexA-Hairy) was introduced by recombination onto the same chromosome. This

39

gene was created by joining a BamHI/Kpnl fiagment containing a Kozak sequence,
initiator ATG, and coding sequence for the entire LexA protein (amino acid residues 1-
202) in frame to a KpnI/Xbal fragment containing the portion of Hairy C-terrninal to the
DNA-binding domain (residues 93-337), and introducing it into the Bng and XbaI sites
of pCaSper-hs [232], (5’ GGA TCC ACC AAA ATG AAA...TGG CTG GAA TTC
CCG GGC CGG GGT ACC GCA GCC...TGG TAG TCT AGA 3’; coding sequences
for LexA and Hairy in bold, restriction sites underlined, dots represent the rest of the
coding sequence). The resulting line, containing all three transgenes on the third
chromosome, behaves as a transcriptional switch in which the default state is ON before
heat shock and OFF after induction of the LexA-Hairy repressor. The triple recombinant
chromosome is carried over a TM3 Sb balancer chromosome and was found to be non-
homozygosable. The pKruppel-LexA-Hairy plasmid was constructed by exchanging the
Gal4 DNA-binding domain coding sequence from the pKreg vector [74] for LexA
(residues 1-202) via restriction digest with BamHI/KpnI and ligation of a BamHI/KpnI
LexA fragment PCR amplified from the pLexA vector (Clontech) using primer DA-645
(5’ GCG GAT CCA CCA AAA TGA AAG CGT TAA CGG CCA GG 3’) and DA-646
(5’ CGG GGT ACC CCG GCC CGG GAA TTC CAG CCA GTC GC 3’. The reporter
shown in figure 2-2 containing LexA binding sites 3’ of the rho and twi enhancers was
constructed by removing the Giant binding sites in the gt-55 vector [21] by SphI
digestion and insertion of DA-641/2 (5’ CTG TAT ATA TAT ACA GCA GGT ACC

TGC TGT ATA TAT ATA CAG CCA TG 3’) containing two LexA binding sites.

40

In situ hybridization and antibody staining of Drosophila embryos. Embryos were
fixed for in situ hybridization and stained using a digoxigenin—UTP labeled antisense
RNA probe to lacZ as described [231]. Antibody staining for LexA-Hairy expression was
done using a 1:2:1 mixture of three mouse monoclonal antibodies raised against LexA,
YN-lexA-2-12 (2 jig/ml, 1:2100), YN-lexA-6-10 (4 jig/ml, 1:2023), YN-lexA-16-7 (4
rig/ml, 1:3950) obtained as a gift from Steve Triezenberg [233]. Quantitation of lacZ
staining was done by scoring the stained embryos into unstained, lightly stained, and
darkly stained from 10-13 random fields. Total embryos scored per slide varied between

135 and 457.

Formaldehyde crosslinking of embryos for chromatin immunoprecipitation. For
chromatin immunoprecipitation assays, embryos (0.25-0.5 gm) were collected on 20%
apple juice plates supplemented with yeast paste from 25-50 plastic laying bottles
containing 300-500 flies each. Typically, fresh plates were placed in the laying bottles
and embryos were subsequently collected for 3 hours and aged for 2 hours at room
temperature. Low-level expression of the heat shock construct was observed in some
cases, but not enough to repress lacZ expression (in 20 biological replicates, some LexA-
Hairy signal was present prior to heat shock in half of the experiments, but most
experiments showed 8-10 fold increase in signal after induction). Depending on the
conditions of the experiment, embryos were then either immediately fixed, or heat
shocked (typically for 20 min) by floating plates in a 38°C water bath and allowed to
recover for a variable length of time at room temperature. Embryos were then collected in

a nylon mesh and dechorionated with 100% bleach for 2.5 min and washed with

41

deionized water for 1 min For formaldehyde crosslinking, embryos were placed in a 50
ml Corning tube and fixed for 20 min with vigorous shaking in 10 ml 3% formaldehyde
fixing buffer (50 mM HEPES pH 7.6, 1 mM EDTA, 0.5 mM EGTA, 100 mM NaCl) and
30 ml heptane. The formaldehyde was added from 37% stock (J .T. Baker # 2106-01)
immediately before use. Embryos were then centrifuged at 3000 rpm (2000 x g) using a
clinical Beckman centrifuge. Supernatant was removed and the crosslinking reaction was
stopped by addition of 25 ml of crosslinking stop buffer (0.125 M glycine, 0.01% Triton
X-100 in phosphate-buffered saline (1.37 M NaCl, 27 mM KCl, 100 mM Nazi-IP04, 18
mM KHzPO4, pH 7.4)) and shaking vigorously for 30 min After centrifugation as before
and removal of the supernatant, embryos were either immediately processed for

chromatin or flash frozen in liquid nitrogen and placed at -75°C.

Preparation of chromatin from whole embryos. 0.25-0.5 gm crosslinked embryos
were washed in 10 m1 embryo wash buffer (10 mM HEPES pH 7.6, 1 mM EDTA, 0.5
mM EGTA, 0.1% sodium deoxycholate, 0.02% sodium azide) for 10 min with vigorous
agitation on a shaker, centrifuged at 3000 rpm (2000 x g) at 4°C and resuspended in 5 ml
of sonication buffer (10 mM HEPES pH 7.6, 1 mM EDTA, 0.5 mM EGTA, 0.1% Triton
X-100) and transferred to a 15 ml Corning tube. A proteinase inhibitor tablet (Roche,
11836153001) was added and the embryos were sonicated 30 sec (100% duty cycle) with
1 nrin cooling interval for a total of 12 times using a Branson sonicator S250 fitted with a
tapered microtip 5mm in diameter. During each pulse the output was gradually increased
from setting 1 to 7 to avoid foaming. Crude chromatin samples were aliquoted into

microcentrifuge tubes, centrifuged at 14000 rpm (16000 x g) for 15 min at 4°C to remove

42

debris, and without disturbing the pellet, supernatant was transferred to a 15 ml Corning
tube. An equal volume of 2xRIPA buffer (2% Triton X-100, 280 mM NaCl, 20 mM Tris-
HCl pH 8.0, 2 mM EDTA pH 8.0, 0.2% SDS) was added. In experiments with multiple
preparations, samples were visually inspected to match color and optical density, and in
cases where they differed due to a different starting amount of embryos, total volume was
adjusted by adding lxRIPA, up to 20% of initial volume. Chromatin samples were then
precleared by adding 10 [Ll/ml of a 50% slurry containing a an equal mixture of agarose
beads coupled to protein A and protein G (Upstate, 16-125 and 16-266) that had been
previously washed 3 times with lxRIPA buffer. Chromatin samples were then aliquoted
into microcentrifuge tubes in 1 ml fractions and either flash-frozen in liquid nitrogen and
stored at —75°C or immediately immunoprecipitated. The DNA size distribution
(measured after reversing cross-links) of typical chromatin preparations was less then

500-1000 bp.

Immunoprecipitation. Immunoprecipitations were carried out in nricrocentrifuge tubes
by overnight incubation (12-16 hr) of 1 ml of precleared chromatin with the appropriate
antibody at 4°C on a rotary mixer. At the same time, protein A and protein G agarose
beads were mixed in equal proportions, washed 3 times with 1x RIPA buffer and then
incubated overnight with 0.1 mg/ml BSA and 0.2 mg/ml salmon sperm DNA in lxRIPA
buffer. Before addition of mixed protein A/G agarose beads, chromatin/antibody
reactions were centrifuged at 14000 rpm (16000 x g) for 15 min to remove insoluble
material, 900 pl of the supernatant was transferred to a new tube, and 40 ul of 50% slurry

of blocked protein A/G beads were added and incubated 4 h at 4°C on a rotary mixer.

43

Beads were centrifuged at 1000 rpm (80 x g) for 1 nrin and supernatant discarded, beads
were washed with 0°C buffer 3 times with 1 ml low salt buffer (0.1% SDS, 1% Triton X-
100, 2 mM EDTA, 20 mM Tris-Cl pH 8.0, 150 mM NaCl]), 3 times with 1 ml high salt
buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-Cl pH 8.0, 500 mM
NaCl), once with LiCl buffer (0.25 M LiCl, 1% NP-40, 1% sodium deoxycholate, 1 mM
EDTA, 10 mM Tris-Cl pH 8.0), and twice with Tris-EDTA (10 mM Tris pH 8.0, 1 mM
EDTA), 3 min per wash. Chromatin was eluted by incubating the beads at room
temperature in a rotary mixer with 250 u] freshly made elution buffer (1% SDS, 0.1 M
monobasic NaHCO3) pH 8.0, for 15 min Beads were centrifuged and the supernatant was
transferred to a screw cap microfuge tube, a second elution was performed, and
supernatants were combined giving a total of 500 pl. Twenty-five #1 of 4 M NaCl was
added to each tube and crosslinks were reversed by overnight incubation at 65°C. In
parallel, 200 it] of input chromatin (20%) was mixed with 300 [1.1 of elution buffer and
incubated overnight at 65°C for input titration controls. To each tube, 10 pl 0.5 M EDTA
pH 8.0, 20 pl 1 M Tris-Cl pH 6.5, 1 pl 10 mg/ml RNAse A, and 1 pl 20 myml
Proteinase K were added and incubated for 2 h at 42°C. DNA was extracted with 500 pl
of phenol-chloroforrn, 400 it] of the aqueous phase was placed in a new microcentrifuge
tube, and DNA precipitated by adding 1 u] (15 ug) Glycoblue pellet paint (Ambion), 44
ul 3 M NaOAc pH 5.2, and 444 u] isopropanol. Tubes were incubated (at room
temperature to prevent SDS precipitation) for 1 h, centrifirged at 14000 rpm (16000 x g)
for 15 min, and pellets were carefirlly washed with 0.5 ml 70% ethanol, dried at 65°C
under vacuum for approximately 15-30 min and resuspended in 50 ul purified water

(Millipore) and analyzed by PCR with the appropriate primers.

Antibodies for chromatin immunoprecipitation. The antibodies and amounts used for
chromatin immunoprecipitation were as follows: non-specific mouse IgG (10 pg,
Upstate), Rabbit anti-LexA (3 pg, Upstate), Rabbit anti-Gal4-TA (5 pg, Santa Cruz),
Rabbit anti-dGCNS (2 pl, gift from Jerry Workman), Rabbit anti-dAda3 (1 pl, gift from

Jerry Workman), and Mouse monoclonal anti-Pol II-CTD 8WG16 (20 pl, Covance).

PCR analysis. Irnmunoprecipitated DNA was analyzed by PCR on a Robocycler
Gradient 96 using Platinum hot-start polymerase (Invitrogen) to reduce primer-dimer
formation. PCR analysis used 2 pl of immunoprecipitated or input DNA. The primer used
to amplify the different regions of the reporter were white forward (DA-1027: 5’ATA
CAG GCG GCC GCG GAT CTG AT 3’), white reverse (DA-1028: 5’ AGA TAG CGG
ACG CAG CGG CGA A 3’), promoter forward (DA-942: 5’ ATC AGA TCC GCG GCC
GCC TGT AT 3’), promoter reverse (DA-943: 5’ CGT CCG CAC ACA ACC TTT CCT
CTC 3’), +1kb forward (DA-865: 5’ CGG GCG CTG GGT CGG TTA CG 3’), +1kb
reverse (DA-873: 5’ GGT GCC GCT GGC GAC CTG C 3’), +2kb forward (DA-948: 5’
AAC CGT CAC GAG CAT CAT CC 3’), +2kb reverse (DA-949: 5’ ATT CAT TCC
CCA GCG ACC AG 3’), +4kb forward (DA 1012: 5’ CGG TCG CTA CCA TTA CCA
GT 3’), +4kb reverse (DA 1013: 5’ ATT GTA ACA GTG GCC CGA AG 3’). Primers
used to amplify intergenic regions were X intergenic forward and reverse (DA-954: 5’
CAC AGT GGA CAC ATA CCA TAG 3’ and DA-955: 5’ CGG AAA ATA TCA GTG
CGA AAG 3’), and chromosome 3 intergenic forward and reverse (DA-960: 5’ GTT
GAG AAT GTG AGA AAG CGG 3’ and DA-96l: 5’ CGA AAA AGG AGA AGG CAC

AAAG 3’).

45

Densitometric analysis of gel images was performed using Image] software.
Briefly, images were background subtracted using the rolling ball algorithm with a ball
diameter of 50 pixels. A rectangular box was placed over the bands, and a plot profile
was generated and saved as a tab formatted file. Microsoft Excel software was then used
to analyze the plot profile, and a standard curve was generated using the area under the
peaks of the input titration. Corresponding input percentage for each band was calculated
by determining the area under the peak of each antibody, and interpolating against the
standard curve. Typical experiments showed R2 values between 0.98-0.99.

Gel images shown in figures 2-5 were background subtracted and the brightness
and contrast adjusted so that the input titrations in experiments using multiple chromatin

samples matched as closely as possible.

46

m
Construction and characterization of a regulated Hairy repressible transgene. To
characterize the activity of the Hairy repressor protein, we employed chromatin
immunoprecipitation assays, which have been used extensively in yeast and cell-culture
studies to identify protein complexes and chromatin modifications associated with genes
during activation and repression. The advantage to these single-cell systems is that a
relatively homogenous population of cells can be obtained, providing a “snapshot” of the
promoters under different physiological states. Our objective was to study the activity of
Hairy in the context of embryo development, but due to the spatially and temporally
limited action of the endogenous Hairy protein, genes in a relatively small percentage of
the total nuclei of the embryo would be repressed by Hairy at any point in time.
Therefore, we engineered a gene regulatory system that allows us to place a constitutively
expressed activator on the promoter, and induce the expression of a LexA-Hairy fusion
protein in a facultative manner, converting the embryo from an “all on” to “all off” state
(Figure 2-1B). The target gene contains five binding sites for the Gal4 activator, flanked
by two pairs of LexA binding sites to accommodate the repressors; similar configurations
had already been tested and shown to be repressed by endogenous Hairy protein [229].
To test the ability of LexA-Hairy to repress a Gal4-activated reporter in transgenic
embryos, we expressed LexA-Hairy in a central stripe in embryos with a lacZ reporter
regulated by the Gal4 activator or endogenous enhancers (Figure 2-2). Embryos
expressing the LexA-Hairy gene driven by the Kruppel enhancer showed a wide swathe
of understained nuclei in the central region, indicating that the repressor is functional

(Figure2-2). We therefore constructed an inducible form of the LexA-Hairy transgene

47

LexA

 

 

1 2025 93 337
I r l J [1:]
: Orange CtBP Gro
B
I ’ ' ‘ ’~—‘-:
’ ’ I ’ ~ ‘ \
oGal4 LexA-Hairy ‘,

00 a an N LacZ
Da [30 (r hsp70 F’n (( rat-1mm l—>

Figure 2-1. Hairy-regulated gene system in the Drosophila embryo.

(A) Schematic diagram of the LexA-Hairy fusion construct. The bacterial LexA protein
was fused in frame to a Hairy fragment lacking the bHLH DNA-binding domain. The
resulting construct contains the Orange, CtBP binding, and Groucho binding domains.
(B) Three transgenes were combined onto a single chromosome to create a regulated
on/off system; a lacZ reporter containing Ga14 and LexA sites, Gal4 activator driven by
the daughterless (da) enhancer for broad expression in the embryo, and a heat-inducible
LexA-Hairy construct. G314 and LexA-Hairy are represented by circles and squares
respectively. Images in this dissertation are presented in color.

48

Reporter Kriippel:
alone LexA-Hairy

 

 

5xUAS
2xLexA 2xLexA

 

 

 

 

 

 

 

 

 

rho twi ,
2xLexA ”r" ““- J

 

 

 

 

 

 

Figure 2-2. LexA-Hairy can repress in vivo.

In situ hybridization of lacZ mRNA expression of embryos carrying either a reporter, or a
reporter and a LexA-Hairy transgene driven by a Kruppel enhancer. Upper, the lacZ
reporter construct shown in Figure 2-1; below, a reporter activated by rho and twi
enhancers, with LexA binding sites at promoter. In both cases, a central swathe of
repression demonstrates effect of the LexA-Hairy chimera. Ga14 and LexA-Hairy are
represented by circles and squares respectively. White rectangles represent enhancers.
Embryos are shown anterior to the left, dorsal surface up. Images in this dissertation are
presented in color.

49

under control of the hsp 70 promoter, allowing variable levels of induction by titration of
heat shock conditions. As measured by in situ hybridization, lacZ transcript levels in the
entire embryo dropped markedly alter even a short (5 min) heat shock. Successively
longer heat shocks resulted in complete loss of lacZ staining (Figure 2-3, 2-4). It should
be noted however, that there was a great degree of heterogeneity in lacZ staining between
embryos within each treatment class. Therefore, in order to get a more accurate reflection
on the effect of LexA-Hairy repression, embryos within each group were scored with
respect to lacZ staining as either “dark blue” (high lacZ expression), “light blue” (low
lacZ expression), or “unstained” (no lacZ expression). Heat shock induction of LexA-
Hairy resulted in a very obvious shift in lacZ staining with most embryos being classified
as “unstained” by 30 and 60 min post-heat shock (Figure 2-5). In typical experiments, the
percentage of embryos showing strong staining dropped from almost half to less than 1%
one hour after induction of the repressor, indicating that the repression was effective in
the vast majority of nuclei and embryos (Figure 2-5). Embryos aged for two hours after
the heat shock induction showed restoration of lacZ expression, indicating that the
repression is reversible (Figure 2-4). Heat shock treatment had no effect on embryos
carrying the reporter gene and the activator in the absence of the LexA-Hairy repressor
protein, indicating that heat shock itself does not interfere with transcription of this gene

(Figure 2-3, lower panels).

50

HS:LexA-Hairy
HS (min) 0 20 4O 60

. . mm - _
. . .t
_-.. -
“$31.7 ’

 

 

 

 

 

 

 

 

 

 

 

 

Reporter alone
HS (min) 0 60

 

 

 

 

 

 

Figure 2-3. Expression of lacZ and LexA-Hairy in blastoderm embryos

Embryos containing transgenes were heat shocked for varying times and fixed for
analysis of lacZ expression by in situ hybridization against the mRNA (top) or antibody
staining of the LexA-Hairy protein (bottom). As LexA protein accumulates, lacZ mRNA
decreases. Heat shock has no effect on lacZ in embryos lacking the LexA-Hairy protein
(two lower embryos). Embryos are shown anterior to the left, dorsal surface up. Images
in this dissertation are presented in color.

51

Recovery (min)
0 30 60 1 20

 

 

 

 

 

 

 

 

 

 

 

 

 

HS (gun)

 

20
i

 

 

 

 

 

 

 

 

 

Figure 2-4. LexA-Hairy repression is reversible.

Repression and reactivation of transgenes in embryos after heat shocks of varying
duration. Shown, in situ hybridization against lacZ mRNA. A 5 min induction of the
LexA-Hairy protein is sufficient to cause significant loss of lacZ mRNA within 30 min
RNA levels remain low at 60 min, and show recovery after 120 min (at this point, most
embryos have aged to germband extended stage). Longer heat shocks (10 and 20 min)
show similar recovery kinetics. Embryos are shown anterior to the left, dorsal surface up.
Images in this dissertation are presented in color.

52

Effect of LexA-Hairy on LacZ in situ

 

 

  

 

 

 

 

 

 

 

 

100%

90% 7 7 7

80% ,7 ,— 7

70% 7 7 7
w 60% —
g iii..- , L
g o I Dark Blue
2 50/0 ' * ,, ' ' lLight Blue
‘3, E1 Unstained
°\ 40% - 7 "7’77”

30% -r 7

20% - 7

10% —7 - -

0% ‘ r r r L—‘
No heatshock 30 min 60 min 120 min

Figure 2-5. Quantitation of lacZ staining after heat shock induction of LexA-Hairy.
3-5 hour old embryos were either immediately stained or first heat shocked for 20 min at
37°C and then recovered at room temperature for 30, 60, and 120 min Embryos were then
classified according to the level of lacZ staining as dark blue, light blue, or unstained.
The percentage of dark blue embryos, corresponding to those with the highest amount of
lacZ, drops dramatically after 30 and 60 min recovery. There is an apparent shift in the
amount of lacZ staining from dark blue to light blue, and light blue to unstained,
indicating that all embryos containing the reporter are becoming repressed. By 120 min
lacZ expression has restarted with an increase in the percentage of stained embryos.
Images in this dissertation are presented in color.

53

Co—occupancy of activator and repressor during gene repression. After establishing
the efficacy of the system, large-scale collections were carried out, and 3-5 hour old
embryos were treated with formaldehyde prior to chromatin preparation for
immunoprecipitation reactions. We initially sought to detect the regulatory proteins that
bound the promoter region. As expected, Gal4 protein was detected at the promoter only
in lines carrying the dazGal4 driver and the lacZ reporter gene (Figure 2-6A, top two
panels). LexA-Hairy was readily detected at the promoter region, only in the presence of
the lexA-hairy transgene. A very strong signal is detected at the promoter after the heat
shock induction; a weak signal was also sometimes detected prior to heat shock, which is
likely due to background expression of LexA-Hairy protein, although not at levels
sufficient to inhibit transcription or to be detected by antibody staining (Figure 2-3). As a
confirmation of the specificity of this interaction, neither Hairy nor Gal4 were found to
associate with an intergenic region chosen randomly in another part of the genome
(Figure 2-6A, lower panels).

Strikingly, the appearance of the LexA-Hairy repressor at the promoter did not
preclude association of the Gal4 activator, indicating that the repressor appears not to rely
on displacement of the activator for its activity (Figure 2-6A, fourth and fifth panels). A
robust signal for the Gal4 protein was evident 30 minutes after induction of the repressor,
at a time point when most lacZ mRNA has disappeared from the embryo (Figure 2-4,
Figure 2-6A). The Gal4 signal does not represent the reassembly of active promoter
complexes at this 30 min time point, because the embryos continue to show repression
even at 60 minutes (Figure 2-4). The persisting Ga14 signal also does not represent

material bound to a separate population of unrepressed promoters, because virtually all of

54

the embryos are uniformly repressed at this stage. The signal for Gal4 and LexA was
strongly concentrated over the promoter region; considerably weaker signals were
detected further 3’ within the open reading flame (see Chapter3, Figure 3-1), possibly
because of a higher order chromatin structure, or the presence of a small flaction of larger
chromatin flagrnents >500 bp. In order to further confirm that G314 and LexA-Hairy can
co-occupy the promoter, sequential ChIPs were carried out. The results show a specific
signal for both LexA and Gal4 co-occupying the reporter during repression (Figure 2-
6B). As negative controls, a LexA or Gal4 immunoprecipitation followed by IgG did not
show any signal (Figure 2-6B). In contrast, LexA/LexA and Gal4/Gal4 sequential ChIPs

showed a strong signal (Figure 2-6B).

Transcriptional co-activators remain associated with repressed promoter. The
persistence of the activators at the promoter indicates that Hairy is not blocking their
access, raising the question of why the promoter is not activated by Gal4 in this situation.
We reasoned that repressor exclusion of co-activators might prevent Gal4 flom having a
stimulatory effect on the promoter. The Gal4 activation domain has been reported to
recruit the SAGA co-activator complex [29, 234], therefore we carried out
immunoprecipitations using antibodies against the Ada3 and Gcn5 subunits of SAGA
(Figure 2-7). A promoter-localized signal for both of these proteins was detected in lines
carrying the Gal4 activator (Figure 2-7, compare top two panels), consistent with the
recruitment of the complex to the active promoter by Gal4. Binding of these coactivators
was not detected at an intergenic locus chosen randomly in another part of the genome

(Figure 2-7, lower panels). After induction of the Hairy repressor, the co-activators

55

Figure 2-6. LexA-Hairy repression does not prevent Gal4 binding.

(A) Promoter occupancy by Gal4 activator and LexA-Hairy repressor measured by
chromatin irmnunoprecipitation. At top, no Gal4 or LexA-Hairy protein is detected with
chromatin prepared flom a line containing solely the lacZ reporter gene, as expected. The
second panel contains a strong signal for the Gal4 protein, from chromatin containing the
activated reporter. These embryos were heat shocked and recovered to parallel the
treatment used for LexA-Hairy containing strains. Below, results flom lines containing
all three transgenes prior to and directly after induction of the LexA-Hairy repressor.
Prior to induction, a signal for Gal4 is visible at the promoter, as well as a weak LexA-
Hairy signal due to low level expression. After 20 min heat shock and 30 min recovery, a
strong signal for LexA-Hairy is visible, in addition to the Gal4 signal. Both signals
remain visible after 60 min recovery. LexA-Hairy signal drops by 120 min No signals are
seen in the intergenic region on chromosome X. (B) Sequential chromatin
immunoprecipitation of embryos expressing LexA-Hairy and Gal4 show that both
proteins can co-occupy the promoter. Input titrations are shown for each chromatin
preparation (2%, 1%, 0.5%, 0.25%, 0.125%, and 0.0625%). At left, results of quantitative
PCR reactions, at right, heatshock and recovery conditions (heatshock done at 37 °C for
20 min, recovered at room temperature). Presence or absence of Gal4, and the heatshock-
inducible LexA-Hairy transgene noted by +/-. Images in this dissertation are presented in
color.

56

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.E‘

:‘c"

S 3 <i:

a w s

(35$ 0 (5 -.—!

0’3 r? 8 2

[L - -

_ L j 30 + -
a)

5]] " j + +
E

g L'- I 30 + +

l "' I 60 + +

(I "—1-...” 1+ 120 + +

(L I--"I I" ' "

9 L l-~~nm l‘l' 30 + '

éJL I’m" l' + +

93 L |........ 1+ 30 + +

c .

— [ I------- 1 + 60 + +

(L |---- - 7 120 + +

.3

cu

(V’v I

x_— 2‘ |

33388 s as

3%33; lnut g ('5' To?

3035’th i’ at ‘6’ E

Promoter L. w- ~---u---.-m |+ 30 + +

lntergenicl — — «- 1 + 30 + +

 

Figure 2-6. LexA-Hairy repression does not prevent Gal4 binding.

57

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A?

‘I"

E s. {5

to co 5 8 0’

(5 58 Input (0 8 i6 3

90“; I a: '5 5

(I I-........ I- - -

b I .. --I----- I+ 30 + -
o

‘gyl -~I---~ l- + +

09- I I----~ I 30 + +

I 1......“ I+ 60 + +

(I -- "1..--.. | 120 + +

rI Inc—u... I' ' '

o I I-nncq |+ 30 + "

.§)<l I-ut‘l. l' + +

E). I J--:::-- I 30 + +

E I Inc-u. I 60 + +

\I I-a-m- I 120 + +

 

 

Figure 2-7. LexA-Hairy repression does not block SAGA recruitment by Gal4.

Promoter occupancy by SAGA coactivator constituents Gcn5 and Ada3 measured by
chromatin immunoprecipitation. Protein occupancy of lacZ reporter promoter region in
different states. The top panel shows no signal for these coactivators on the promoter
lacking Gal4 activators, as expected. Both proteins can be detected on the promoter in a
Gal4 activator containing strain, regardless of whether the embryos had been head
shocked or not, and whether or not LexA-Hairy had been induced. No signals are seen in
the intergenic region on chromosome X. Input titrations are shown for each chromatin
preparation (2%, 1%, 0.5%, 0.25%, 0.125%, and 0.0625%). At left, results of quantitative
PCR reactions, at right, heatshock and recovery conditions (heatshock done at 37 °C for
20 min, recovered at room temperature). Presence or absence of Gal4, and the heatshock-
inducible LexA-Hairy transgene noted by +/-. Images in this dissertation are presented in
color.

58

remained detectable at the promoter, at a point in time at which lacZ gene expression in
the embryo has ceased (Figure 2-7, fourth and fifth panels). Thus, Hairy apparently does

not exclude either activator or co-activator to effect repression of the promoter.

59

Discussion

Previously, repression had been thought of as being either passive or active.
Passive repression was characterized as competition for binding sites between activators
and repressors or the formation of an inactive repressor/activator complex [16]. For
example the homeobox-containing repressor Engrailed competes for binding sites with
the Fushi-tarazu activator [31]. Similarly, the bHLH protein Extrarnachrochaetae can
bind to members of the Achaete-Scute complex and inhibit their binding to the DNA
[88]. In contrast, active repression involves direct recruitment of the repressor to the
locus where it can inhibit the basal machinery, block activator-promoter interactions, or
change the chromatin structure. Evidence that Hairy is an “active” repressor comes flom
previous studies that have shown that Hairy can repress the proneural gene achaete by
binding to a noncanonical E-box 50 bp upstream of three activator binding sites [47].
Furthermore, Hairy binding does not prevent activator binding in vitro [88]. Moreover,
expression of a Hairy protein fused to a heterologous activating domain leads to
transcriptional activation of putative Hairy targets, suggesting that it binds these targets
directly [60]. In contrast, mutation of the Hairy binding site in the achaete promoter
abolishes Hairy-mediated repression [61].

Our finding that activators and coactivators are still present under conditions
when the gene is repressed by Hairy suggests that repression does not involve
activator/coactivator displacement. The Gal4 activator might represent a particularly
stably bound protein, as it does not show the high rate of exchange noted for other
transcriptional activators [235]. Thus, it is possible that Hairy mediated repression does

interfere with binding of some activators on endogenous loci. However, repression can be

60

quite effective even in the absence of activator displacement, perhaps by targeting the
basal machinery, similar to Tupl-Mediator interactions seen in yeast [236-238]. One
might suppose that the promoter proximal location of the repressor in our system might
bias the system to such interactions, but this arrangement is physiologically relevant, as
Hairy is found in such proximal locations on endogenous genes. In addition, the LexA-
Hairy repressor is also active when bound at —2 kbp, indicating that promoter proximity
is not required for activity (L. Li, unpublished observations). Interestingly, a recent
chromatin immunoprecipitation survey of enhancers targeted by the Snail short-range
repressor suggests that this repressor can be bound to inactive enhancers simultaneously
with activators, raising the possibility that short-range repression might also involve
direct interactions with the basal machinery [23 9].

Intriguingly, the results presented here point to a possible threshold effect for
Hairy repression, where the levels of Hairy protein necessary to repress is directly
proportional to the levels of activator present. Chromatin immunoprecipitation of LexA-
Hairy and Gal4 in time course experiments showed a significant level of LexA-Hairy
remaining two hours after induction at a point when transcription is restarting (Figure 2-
4, 2-5, and 2-6). Similarly, before heat shock, there is a detectable LexA-Hairy signal
presumably due to leaky expression of the repressor, yet the embryos show robust lacZ
staining (Figure 2-4, 2-5, and 2-6). A possible explanation for these results is that it is not
the total level of repressor bound but rather the ratio of activator/repressor binding to the
gene that determines the transcriptional state. Thus, even in the presence of a significant
amount of LexA-Hairy, repression is readily reversed by relatively higher levels of Gal4

activator.

61

The importance of such an effect to endogenous HES targets can be seen in the
regulation of the sex determinant gene sex-lethal. Sex determination in Drosophila is
determined by the ratio of X to autosomal chromosomes, where a ratio of l is female and
0.5 is male [64]. In males sex-lethal is repressed by the HES family member Deadpan by
binding to a region between 93-125 bp upstream of the start site [64]. In contrast, sex-
Iethal becomes activated in females due to a higher transcription of X-linked activators
(numerator elements) to deadpan (denominator element) [113]. Since deadpan is an
autosomal gene, the concentration of the repressor will be the same in both females and
males. Assuming that the flactional occupancy of Deadpan depends solely on the Km and
repressor concentration, then it should be similar in both males and females. Therefore, in
order to maintain sex-lethal expression in females, activators must remain bound
simultaneously with Deadpan and overcome its repression mechanism.

As mentioned in the first chapter, even in the absence of the Gcn5 histone
acetyltransferase subunit, SAGA can still recruit the basal machinery through direct
protein-protein interactions [186]. Since SAGA is not displaced upon LexA-Hairy
repression, it is possible that Pol II is still being recruited to the promoter but is prevented
flom transcribing the gene. Preliminary chromatin immunoprecipitation experiments with
Pol II antibodies showed a decrease in Pol H signal at the promoter after repression
(Appendix B). Paradoxically, there was an increase in Pol II in the coding region,

suggesting a paused complex (Appendix B).

62

Chapter III

Spreading of corepressor linked to action of long-range repressor Hairy2

Introduction

An important functional distinction between different repressors is their ability to
interfere with proximally- or distally-located activators. In Drosophila, Hairy can inhibit
the activity of activators located over 1 kbp away, leading to its characterization as a
long-range repressor [22]. In contrast, short-range repressors are limited to interfering
with activators bound within ~100 bp [18]. The limited range of short-range repressors
appears to be well adapted to the architectures of the regulatory regions they control. In
the Drosophila embryo, short-range repressors such as Knirps and Giant repress the
modular enhancers controlling pair-rule genes such as even-skipped and hairy. The
independent activity of such enhancers is guaranteed by the local action of the repressors;
if enhancers are brought into artificially close proximity, or if binding sites for short
range repressors are moved close to the transcriptional start site, unwanted cross-
regulation can occur [19, 240].

In contrast, the molecular logic of cis regulatory elements controlled by long-
range repressors is less obvious. At the Drosophila achaete gene, the long-range
repressor protein Hairy binds at -300 bp, 50 bp 5’ of a cluster of activator proteins, a
position flom which short-range repressors would also presumably work well [61].
Similarly, the Hairy homolog HES] binds to its own promoter at four sites 20-170 bp

flom the transcriptional start site [63]. Dorsal protein-regulated ventral repression

 

2 Parts of this chapter were excerpted flom the paper titled “Spreading of corepressor linked to action of
long-range repressor Hairy” by Carlos Martinez and David Amosti, submitted to Molecular and Cellular
Biology (2007)

63

elements (VRE) flom Drosophila zen, tld, and dpp genes can similarly act over long
distances, but at least in the case of the zen VRE, the activators bind immediately 5’ of
the repression element [241]. Thus, it is not clear if the long range of activity in these
instances is essential to the normal regulatory function. Perhaps the strength of repression
of Hairy is the most important feature, which is only incidentally associated with long-
range effects.

Hairy/E(spl) proteins possess a conserved basic helix-loop-helix DNA-binding
domain and effector domains that include motifs important for interaction with
corepressors [1]. Hairy interacts physically and genetically with three corepressors:
Groucho, the C-terminal binding protein (CtBP), and the Sir2 histone deacetylase [55, 69,
71]. The C-terrninus of Hairy contains a WRPW motif that directly contacts the Groucho
corepressor, and removal of the motif compromises the activity of Hairy. A motif
adjacent to the Groucho interacting region binds to the CtBP corepressor. Hairy protein
has been shown to possess CtBP-mediated repression activity in certain circumstances.
However, CtBP has also been suggested to play an antagonistic role in repression by
Hairy because binding of Groucho and CtBP might be mutually exclusive, and removal
of the CtBP interacting motif has a less drastic effect on repression than removal of the
Groucho motif [72]. The histone acetylase Sir2 interacts with the Hairy DNA-binding
domain, and genetic interactions between hairy and Sir2 have been reported [71].

Whole genome mapping of binding sites for Hairy and cofactors indicates that at
many loci, Hairy is not associated with all three cofactors. In fact, Hairy was rarely found
to colocalize with regions bound by Groucho, while colocalization with CtBP was

observed in a majority of cases [75]. These studies indicate that Hairy may associate with

64

specific cofactors in a context-dependent manner, perhaps invoking different modes of
transcriptional regulation. A limitation of these studies is that the physical resolution is
limited, so that it is not known whether Hairy and the corepressor proteins are in direct
contact, or if in some cases other transcription factors might be recruiting these cofactors.
In addition, it is not known for most loci whether the observed binding event is
functional. Thus, while genetic and physical interactions hint at potential complexity, the
activity of Hairy and its set of possible corepressors is not understood at a molecular
level. To better understand molecular mechanisms of long-range repression, we have
employed a novel approach to measure the activity of the Hairy repressor on a highly
defined system in the Drosophila embryo.

Using transgenic lines containing a transcriptional switch that can be repressed
uniformly in the embryo, we have analyzed the recruitment of the corepressors and
histone modifications associated with Hairy repression. The results show that repression
is associated with binding and spreading of the Groucho corepressor and the histone
deacetylase de3 throughout the coding region of a lacZ reporter. In addition, Hairy
repression is associated with a marked decrease in histone acetylation levels and an
increase in total histone occupancy. These results strongly support a model of Hairy

repression involving the formation of a heterochromatin-like state.

65

Materials and Methods

Formaldehyde crosslinking of embryos for chromatin immunoprecipitation. For
chromatin immunoprecipitation assays, embryos (0.25-0.5 gm) were collected flom
plastic laying bottles for 3 hours and aged for 2 hours at room temperature. Embryos
were then either immediately fixed, or heat shocked (20 min.) by floating plates in a 38°C
water bath and allowed to recover for a variable length of time at room temperature.
Embryos were collected and dechorionated with bleach. For single crosslinking, embryos
were fixed for 20 min with vigorous shaking in a 50 ml Corning tube in 10 ml 3%
formaldehyde fixing buffer [50 mM HEPES pH 7.6, 1 mM EDTA, 0.5 mM EGTA, 100
mM NaCl, with formaldehyde (J .T. Baker # 2106-01) added immediately before use flom
37% stock] and 30 ml heptane. Embryos were then centrifuged at 2000 x g in a clinical
centrifuge, supernatant removed, and the crosslinking reaction stopped with 25 m1 stop
buffer [0.125 M glycine, 0.01% Triton X-100 in phosphate-buffered saline (1.37 M NaCl,
27 mM KCl, 100 mM NazHPO4, 18 mM KH2P04, pH 7.4)], shaking vigorously for 30
min. Embryos were centrifuged as before and immediately processed for chromatin or

flash flozen and stored at -75°C.

Double crosslinking of embryos for chromatin immunoprecipitation. Embryos were
collected, heatshocked, and dechorionated as above, placed in a 50 ml Corning tube with
8 ml of phosphate-buffered saline. Two ml of a fleshly prepared 25 mM
dithiobis(succinimidyl propionate (DSP) crosslinking solution in DMSO was added to

yield a final concentration of 5 mM DSP and 20% DMSO. Embryos were shaken

66

vigorously for 30 min., centrifuged at 2000 x g, supernatant removed, and crosslinked

with formaldehyde as above.

Preparation of chromatin from whole embryos. O.25-O.5 gm crosslinked embryos
were washed in 10 ml embryo wash buffer (10 mM HEPES pH 7.6, 1 mM EDTA, 0.5
mM EGTA, 0.1% sodium deoxycholate, 0.02% sodium azide) for 10 min with vigorous
agitation, centrifuged at 2000 x g at 4°C and resuspended in 5 ml of sonication buffer (10
mM HEPES pH 7 .6, 1 mM EDTA, 0.5 mM EGTA, 0.1% Triton X-100) and transferred
to a 15 ml Corning tube. A proteinase inhibitor tablet (Roche, 11836153001) was added
and the embryos were sonicated 30 sec (100% duty cycle) with l min cooling interval 12
times using a Branson sonicator. With each pulse, the output was gradually increased
flom setting 1 to 7 to avoid foaming. The embryo lysate was aliquoted into
microcentrifuge tubes, centrifuged at 16000 x g 15 min at 4°C, and without disturbing the
pellet, supernatant was transferred to a 15 ml Corning tube. An equal volume of 2xRIPA
buffer (2% Triton X-100, 280 mM NaCl, 20 mM Tris-HCl pH 8.0, 2 mM EDTA pH 8.0,
0.2% SDS) was added. For Groucho immunoprecipitation experiments flom double
crosslinked embryos the SDS was omitted. Chromatin was precleared by adding 10 pl/ml
of a 50% slurry containing an equal mixture of agarose beads coupled to protein A and
protein G (Upstate, 16-125, 16-266) previously washed 3 times with llePA buffer.
Chromatin was then aliquoted into microcentrifuge tubes in 1 ml flactions and flash-

flozen and stored at -—75°C or immediately immunoprecipitated.

67

Immunoprecipitation. Immunoprecipitations were carried out by overnight incubation
on a rotary mixer of 1 ml of precleared chromatin at 4°C with the antibody. At the same
time, protein A and protein G agarose beads were mixed in equal proportions, washed 3
times with 1x RIPA buffer and then incubated overnight with 0.1 mg/ml BSA and 0.2
mg/ml salmon sperm DNA in 1xRIPA buffer. For Groucho immunoprecipitation
experiments, 20 pg of Rabbit anti-Mouse IgG bridging antibody was added and the tube
was incubated for an additional 1-2 hours. Before addition of mixed protein A/G agarose
beads, chromatin/antibody reactions were centrifuged at 16000 x g for 15 min, 900 pl of
the supernatant was transferred to a new tube, and 40 pl of 50% slurry of blocked protein
A/G beads were added and incubated 4 h on a rotary rrrixer. Beads were centrifuged at 80
x g for l min, washed (0 °C) 3 times with 1 ml portions of low salt buffer (0.1% SDS, 1%
Triton X-100, 2 mM EDTA, 20 mM Tris-Cl pH 8.0, 150 mM NaCl]), 3 times with high
salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-Cl pH 8.0, 500 mM
NaCl), once with LiCl buffer (0.25 M LiCl, 1% NP-40, 1% sodium deoxycholate, 1 mM
EDTA, 10 mM Tris-Cl pH 8.0), and twice with Tris-EDTA (10 mM Tris pH 8.0, 1 mM
EDTA. In Groucho immunoprecipitation experiments using double-crosslinked
chromatin, 6 washes were performed with 1xRIPA lacking SDS (1% Triton X-100, 140
mM NaCl, 10 mM Tris-HCl pH 8.0, 1 mM EDTA). Chromatin was eluted at room
temperature in a rotary mixer with 250 pl elution buffer (1% SDS, 0.1 M monobasic
NaHCO3, pH 8.0), for 15 min. Beads were centrifuged and the supernatant was
transferred to a screw cap microfuge tube, a second elution was performed, supematants
combined, 25 pl of 4 M NaCl added, and crosslinks were reversed overnight at 65°C. In

parallel, 200 pl of input chromatin (20%) was mixed with 300 pl elution buffer and

68

incubated overnight at 65°C for input titration controls. Then, 10 pl 0.5 M EDTA pH 8.0,
20 pl 1 M Tris-Cl pH 6.5, 1 pl 10 mg/ml RNAse A, and 1 pl 20 mg/ml Proteinase K were
added and incubated for 2 h at 42°C. DNA was extracted with 500 pl of phenol-
chloroforrn, 400 pl of the aqueous phase was placed in a new tube and DNA precipitated
with 15 pg Glycoblue pellet paint (Ambion), 44 pl 3 M NaOAc pH 5.2, and 444 pl
isopropanol. Tubes were incubated (at room temperature to prevent SDS precipitation)
for l h, centrifuged at 16000 x g for 15 min, and pellets carefully washed with 0.5 ml
70% ethanol, dried at 65°C under vacuum for approximately 15-30 min and resuspended
in 50 pl purified water for PCR analysis.

Using the simple formaldehyde crosslinking protocol, we sometimes obtained
crosslinking of Groucho to the transcribed region but never to the promoter proximal
sequences. This signal appears to be specific, because Groucho is never detected on other
regions. Using the double crosslinking protocol, we found that Groucho, when detected,
was also associated with the promoter, in addition to downstream regions. Even using
this protocol, we were not always successfirl in detecting Groucho, consistent with the
reported difficulty of detecting transcriptional cofactors indirectly bound to the DNA

[242].

Antibodies for chromatin immunoprecipitation. We used the following antibodies:
non-specific mouse IgG (10 pg, Upstate, cat # 12-371), rabbit anti-LexA (3 pg, Upstate,
cat # 06-719), rabbit anti-Gal4-TA (5 pg, Santa Cruz, cat # sc-429), mouse monoclonal
anti-Groucho (50, 100 pl, Iowa Hybridoma Bank), rabbit anti-HDAC] (1-4 pl, Abcarn,

cat # ab1767), rabbit anti-H3 (1 pl, Abcarn, cat # ab1791), rabbit monoclonal anti-

69

mono/di/trimethyl Histone H3 K4 (1 pl, Upstate, cat # 05-791), rabbit anti-acetyl Histone
H4 (5 pl, Upstate, cat # 06-598), rabbit anti-dimethyl Histone H3 K27 (5 pl, Upstate, cat
# 07-452), rabbit anti-acetyl Histone H3 (1 pl, Upstate, cat # 06-599), and rabbit anti-

Mouse IgG (20 pg, Upstate, cat # 06-371).

PCR analysis. 2 pl samples of immunoprecipitated DNA was analyzed on a Robocycler
Gradient 96 with Platinum hot-start polymerase (Invitrogen). Primers used were white
forward (DA-1027: 5’ATA CAG GCG GCC GCG GAT CTG AT 3’), white reverse
(DA-1028: 5’ AGA TAG CGG ACG CAG CGG CGA A 3’), promoter forward (DA-
942: 5’ ATC AGA TCC GCG GCC GCC TGT AT 3’), promoter reverse (DA-943: 5’
CGT CCG CAC ACA ACC TTT CCT CTC 3’), +1kb forward (DA-865: 5’ CGG GCG
CTG GGT CGG TTA CG 3’), +1kb reverse (DA-873: 5’ GGT GCC GCT GGC GAC
CTG C 3’), +2kb forward (DA-948: 5’ AAC CGT CAC GAG CAT CAT CC 3’), +2kb
reverse (DA-949: 5’ ATT CAT TCC CCA GCG ACC AG 3’), +4kb forward (DA 1012:
5’ CGG TCG CTA CCA TTA CCA GT 3’), +4kb reverse (DA 1013: 5’ ATT GTA ACA
GTG GCC CGA AG 3’). Primers used to amplify intergenic regions were X intergenic
forward and reverse (DA-954: 5’ CAC AGT GGA CAC ATA CCA TAG 3’ and DA-
955: 5’ CGG AAA ATA TCA GTG CGA AAG 3’), and chromosome 3 intergenic
forward and reverse (DA-960: 5’ GTT GAG AAT GTG AGA AAG CGG 3’ and DA-
961: 5’ CGA AAA AGG AGA AGG CAC AAA G 3’).

Densitometric analysis of gel images was performed using Image] software. Gel

images shown in figures 2-5 were background subtracted and the brightness and contrast

70

adjusted so that the input titrations in experiments using multiple chromatin samples

matched as closely as possible.

71

m

Association of Groucho and de3 co-repressors with wide tracts of repressed gene.
Using the defined embryo system (described in Chapter 2), we tested whether the
corepressor Groucho was recruited to the lacZ reporter during repression (Figure 3-1).
Using formaldehyde crosslinking techniques, Groucho was detected after induction of the
Hairy repressor. No signal was detected at the promoter. However, strong signals were
evident at regions 3’ of the promoter at +1 kbp and +2 kbp, and less signal was detected
at +4 kbp (Figure 3-1B, lanes 6 and 7). Groucho was not found to crosslink to an
intergenic region, indicating that the signal is specific to the Hairy-repressed gene (Figure
3-1B). A similar pattern of crosslinking was obtained with the de3 histone deacetylase,
with strongest signals observed at +1 and +2 kbp (Figure 3-1B, lanes 3-5). This
deacetylase has been identified as a Groucho-interacting protein in biochemical assays
[84].

Detection of Groucho using this crosslinking protocol was variable, therefore we
tested a double crosslinking procedure recently described for crosslinking of cofactors in
yeast, which involves a two-step treatment employing the bifirnctional crosslinker DSP
first followed by formaldehyde [243]. Using this procedure we were again able to detect
Groucho on downstream regions of the lacZ reporter gene. Under these crosslinking
conditions, a strong signal was also detected over the promoter region (Figure 3-1C, lane
5). Again, no Groucho was found on intergenic regions, indicating the specificity of the
signal (Figure 3-1C). The signal seen for the LexA-Hairy protein in this experiment is
also detected more weakly at regions within the transcribed region, suggesting that there

are protein contacts with sites distal to the binding sites within the promoter (Figure 3-

72

1C, lane 3). The extended signal is not simply the result of very long chromatin
flagrnents, because the average chromatin size in this experiment is less than 500 bp, and
the Groucho signal seen in Figure 3-1B is centered over the transcribed region, not the
promoter. These widespread contacts of Groucho with the transcribed region of the gene
are consistent with a model proposed by Courey, in which multimerization of protein
through the N-terrninal domain forms an extended “spread” conformation that permits the
protein to influence multiple regions of the DNA, consistent with the long-range activity
of Hairy [244].

We tested whether Hairy recruits Groucho to upstream regions, similar to the
pattern seen on the transcribed locus. More distal regions could not be specifically
sampled because of the presence of three copies of the white gene on the different P-
element vectors in these strains. However, PCR primers specific to the promoter and the
upstream mini white transgene were used to amplify proximal regions 5’ of the promoter.
These experiments showed recruitment of Groucho in chromatin prepared using the
double crosslinking protocol (Figure 3-1C), suggesting that Groucho may be
bidirectionally disposed around the LexA-Hairy binding sites.

We sought to detect the CtBP and Sir2 corepressors at this gene, but we did not
reliably detect signals above background. Sir2 interaction with Hairy has been mapped to
the DNA-binding domain, which is absent in this LexA-Hairy chimeric protein, thus it is
not surprising that no signal was detected for this protein [71]. The negative result
regarding CtBP is not very conclusive, because this particular antibody may not be

suitable for immunoprecipitations. Nonetheless, repressors such as Hairy can utilize a

73

Figure 3-1. LexA-Hairy repression results in recruitment of Groucho and de3 to
the entire gene locus.

Occupancy of promoter and transcribed regions of a lacZ transgene by Groucho and
de3 corepressors during repression by LexA-Hairy. (A) Schematic diagram of LexA-
Hairy regulated gene, showing portions amplified. (B) PCR analysis of formaldehyde-
crosslinked chromatin immunoprecipitated flom embryos that were induced for
expression of the LexA-Hairy repressor for 20 min. and allowed to recover for 30 min.
As expected, no signal was seen for the nonspecific IgG precipitation. A strong promoter-
localized signal was detected for LexA-Hairy, with weaker signal on distal regions of the
gene. Signals for the de3 histone deacetylase (l, 2, and 4 pl of de3 antibody shown)
and the Groucho corepressor (50 and 100 pl of Groucho antibody shown) were detected
at +1 kbp and +2 kbp, with a weak signal for Groucho at +4 kbp No crosslinking was
detected for an intergenic region on chromosome 3. Input titration shown at right (2%,
1%, 0.5%, 0.25%, 0.125%). (C) PCR analysis of immunoprecipitated
DSP+formaldehyde crosslinked chromatin. LexA and Gal4 are detected at the promoter,
with weaker signals in 3’ regions. In contrast to results in (B), a strong Groucho signal is
detected at the start of the white gene (-350 bp), promoter, and throughout the transcribed
region. No crosslinking was detected for an intergenic region on chromosome X.
Groucho was not detected at the transgene locus in chromatin samples derived flom
single or double crosslinked embryos that were not heat shock induced for LexA-Hairy,
or were heat shocked but lacked the lexA-hairy gene (not shown). Input titration shown at
right (10%, 2%, 1%, 0.5%, 0.25%). Images in this dissertation are presented in color.

74

A

I Eminfl I

<—->4—> <—-> 4—-> <—>
w+ Promoter +1 kb +2kb +4kb

(D g de3 Gro Input
Promoter I u- I- — .... .. I

B

 

 

 

 

 

 

 

 

 

 

+1kb I uno- .. - In- -I--- .. I
+2kb I *0 as so .. "l~-" ..1
+4kb I |--- or; I
Intergenic L I— — a... . . I
C 2 § v
- 9 Input
2 (5?» 3 8 o ‘

 

White I urn—nu—I—u-c I '

Promoter
+1 kb

 

"-.-I--pa....-

I l
l ' J
+2kb I u ...I- - - 1 .
l l
l 1,

 

 

+4kb

M -l-umc

 

lntergenic I- — -- --,

Figure 3-1. LexA-Hairy repression results in recruitment of Groucho and de3 to
the entire gene locus.

75

subset of corepressors for regulation of individual target promoters, and perhaps CtBP is

not recruited to this gene [7 5].

Chromatin modification associated with activation and repression of a target gene.
In light of the numerous connections between chromatin modifications and gene
regulation, and the association of the deacetylase de3 with the reporter, we tested the
effects of gene repression on histones and chromatin modifications in the promoter region
(Figure 3-2). Chromatin was prepared flom embryos carrying only the reporter gene
(unactivated state), embryos with both reporter and activator (activated state), embryos
with reporter, activator and repressor, induced and aged for 30 to 60 min (repressed
state), and 120 min (reactivated state). Recruitment of Gal4 to the reporter gene was
associated with an overall reduction of levels of histones, measured by anti-H3 antibody
(Figure 3-2B, third column). The levels of histone H3 and H4 acetylation, relative to total
histone H3, were elevated, consistent with the appearance of SAGA subunits at the
promoter (Figure 2-7, Chapter 2). Also associated with activation of the gene is a relative.
decrease in histone H3 K27 methylation (Figure 3-2B, fourth column). The induction of
LexA-Hairy and repression of the gene is associated with an increase in overall H3
levels, and a relative decrease in H3 and H4 acetylation levels, consistent with the
recruitment of histone deacetylases to the gene (Figure 3-2B, first and second columns).
By these measures, the effect of the repressor is the opposite of the activator. However,
the drop in relative H3 K27 methylation levels is not reversed by recruitment of the
repressor, indicating an association of this modification with activation, but not

repression, pathways. None of these effects were noted on a distal intergenic locus,

76

Figure 3-2. LexA-Hairy repression results in histone deacetylation and higher
histone occupancy at the promoter.

(A) Chromatin remodeling, assayed by chromatin immunoprecipitation. Total histone H3
levels, acetylation, and methylation were assayed in the promoter region of the reporter
gene in unactivated, activated, and repressed states, as well as with an unactivated
reporter gene with repressor bound. The top two panels demonstrate the effect of Gal4
activators on the promoter; in the presence of Gal4 (second panel), overall histone H3
levels drop, as do H3 K27 methylation levels, while relative acetylated histone H4 and
H3 levels increase. In the presence of the LexA-Hairy repressor (compare second and
third panels), 30 min after induction, total H3 levels increase modestly and relative H3
and H4 acetylation levels drop. In this experiment, H4 acetylation levels were more
affected. Relative levels of H3 K27 methylation remained low in the repressed state. The
binding of repressor to the nonactivated gene (compare first and fourth panels) did not
affect total histone H3 occupancy or H3 K27 methylation, and had only modest effects on
relative acetylation. Below, relative levels of histone modifications at an intergenic locus
on chromosome X are unchanged by the treatments. Immunoprecipitations were carried
out with antibodies that recognize total histone H3, K9/ 14 acetylation of histone H3,
K5/8/12/16 acetylation of histone H4, or K27 dimethylation of histone H3. Input titration
shown at right (10%, 2%, 1%, 0.5%, 0.25%, 0.125%). (B) Quantitation of chromatin
modifications at the promoter during activation, repression, and reactivation. Levels of
acetylation and methylation were determined using densitometry analysis. Signals were
normalized to the unactivated state. Shown are the average values and standard deviation
of at least 5 biological replicates for the activated and repressed states. The average value
of two biological replicates is shown for the reactivated state. Signals for AcH3, AcH4,
and H3 K27 methylation are divided by the signal value of H3 to take into account
differences in total histone levels. At left, results of quantitative PCR reactions, at right,
heatshock and recovery conditions (heatshock done at 37 °C for 20 min, recovered at
room temperature). Presence or absence of Gal4, and the heatshock-inducible LexA-
Hairy transgene noted by +/-. Images in this dissertation are presented in color.

77

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.5:
E
...r ‘3 % i.
3 ° <5
giaifis-k ei a?
I I-------.. I - - -
a; I-v---I--pu----.. I + 30 +
($5 I C-IOr------— I + 30 + +
I I- I-—------- I + 30 +
o I I--—---—.. I -
a I I------~ I + 30 +
E) I I-—-----—-- I + 30 + +
E
l I

 

Chromatin Modifications at Promoter

 

I Unactivated
El Activated

I Rep ressed
Reactivated

Ratio

 

 

 

ACH3/ H3 AcH4/ H3 K27me/H3 H3

Figure 3-2. LexA-Hairy repression results in histone deacetylation and higher
histoue occupancy at the promoter.

78

indicating that the modifications were specific to the reporter gene (Figure 3-2A, lower
panels). In addition, heat shock alone did not visibly alter chromatin signals, indicating
that the heat induction regimen alone was not accountable for these changes (Figure 3-
2A, lower panels). During reactivation, the changes in acetylation levels of H3 and H4, as
well as total histone H3 were reversed back near the activated state (Figure 3-2B). No
change in H3 K27 methylation levels was observed (Figure 3-2B).

Previous studies of endogenous Hairy protein indicated that it is unable to repress
a distal enhancer if it binds in a regulatory region lacking activators, possibly because
Hairy cannot access a domain that has not been subject to remodeling induced by
activators — the so-called “hot chromatin” model [62]. Simulating this situation, we
performed chromatin immunoprecipitations in a strain lacking the Gal4 activator protein,
and found that the Hairy chimera is able to access the promoter (Figure 3-2A, fourth
panel). This result suggests that locally-acting activators were not required to recruit this
form of Hairy. However, the promoter context may be more permissive than distal
enhancer regions, or the endogenous Hairy protein may be subject to more stringent
requirements for DNA-binding. The binding of the repressor did not affect total histone
H3 levels, K27 methylation, and had only modest effects on histone acetylation levels
(most flequently a slight decrease in relative levels of histone H3 acetylation).

In light of the long-range contacts seen for the Groucho corepressor, we measured
histone levels and histone acetylation levels on the coding region of the reporter gene
(Figure 3-3). The results showed that changes in acetylation and histone levels associated
with repression could be observed up to a distance of 1 kb downstream flom the

promoter. However, little change was observed at 2 kb and 4 kb downstream. Taken

79

Figure. 3-3. Change in acetylation levels and histone occupancy is limited to 1 kb
from promoter.

Chromatin changes occurring on the transcribed region of the gene during activation,
repression, and reactivation assayed by chromatin immunoprecipitation. (A) Relative
histone H3 K9/14 acetylation levels (normalized to H3) at +1, +2, and +4 kbp. A strong
increase in relative H3 acetylation at +1 kbp is observed with Gal4 activator present, and
this acetylation decreases after induction of the repressor. No large changes at +2 and +4
kbp were observed. (B) Relative histone H4 K5/8/12/16 acetylation levels (normalized to
H3) at +1, +2, and +4 kbp. Promoter-proximal acetylation status was observed to increase
after activation, and decrease during repression. A smaller decrease in acetylation at +2
and +4 kbp was usually observed. During reactivation, a hyperacetylation of histone H4
was observed in the coding region. (C) Total relative histone H3 occupancy on reporter.
Overall levels of H3 at the promoter decreased in the activated state relative to the
unactivated gene, and increase again during repression. No change was observed at +2
and +4 kb. During reactivation H3 levels decreased throughout the coding region. Shown
are the average value and standard deviation of at least 4 biological replicates for the
activated and repressed states. Average value of 2 biological replicates is shown for the
reactivated state. Signals have been normalized to the unactivated state. Images in this
dissertation are presented in color.

80

A AcH3/H3

 

 

  
 
    

4 _ ..... _ __ 7 _
l
I IUnactivated
g I DActivated
E I I Repressed
~ I IReactivated
B
AcH4/H3
6
5
4 I Unactivated
.8 3 CIActivated
E ‘ IRepressed
2 I IReactivated
1
0
C
H3
1.2 7
1.0
0 8 I Unactivated
.3 0'6 ElActivated
§ ' i IRepressed
0-4 r IReactivated
0.2 ‘
0.0

Figure. 3-3. Change in acetylation levels and histone occupancy is limited to 1 kb
from promoter.

81

together, these results suggest that activation and repression of the reporter gene is
accompanied by changes in histone occupancy and acetylation, but that the changes
induced by activation are not completely reversed by repression.

We also analyzed the changes in chromatin modifications during reactivation of
the reporter. Intriguingly, a strong acetylation signal of histone H4 was observed
throughout the coding region during reactivation. In addition, total levels of histone H3
dropped throughout the coding region during reactivation. These results suggest that the

reactivated state is different then the activated state.

82

Discussion

Despite the intensive study of Hairy and other HES proteins in development, little
is known of the mechanism of this model long-range repressor. Thus, we lack a full
understanding of how distinct corepressors may be implicated in different developmental
circuitry, or the logic of regulatory regions controlled by this protein. Our finding that
Groucho associates with a Hairy-repressed locus over the distance of several kilobases
provides a possible mechanism that explains how Hairy can influence the activity of
activator sites over a great distance. The molecular nature of the repression mechanism
may involve “spreading”, as suggested by Courey, involving progressive deacetylation
and binding to hypoacetylated histones [17, 244].

In support of this model, Groucho binds hypoacetylated histone H3 and H4 tails,
and mutations in its N-terrninal oligomerization domain block its repression activity in
vivo [84, 244]. Recruitment of histone deacetylases may thus enhance Groucho binding to
adjacent histones in a positive feedback loop. A similar mechanism has been suggested
for Tupl, the Groucho homolog in yeast [85]. As with Groucho, histone deacetylases
have been shown to be crucial for Tupl repression [245]. Moreover, Tupl also has
affinity for hypoacetylated amino-terminal histone tails, and mutations or deletions of the
tails cause derepression of Tupl targets [246, 247]. As with the Hairy repressor here,
Tupl does not change methylation status of target genes, and deletion of histone
methyltransferases do not affect Tupl mediated repression [248]. This suggests that
methylation marks may not need to be reversed to achieve repression, but they may
facilitate ready reactivation seen upon depletion of Hairy. Regarding the extent of

association of Tupl with target genes in yeast, chromatin immunoprecipitation studies

83

have yielded conflicting pictures. Tupl has been reported to interact with the a-cell
specific ST E6 gene only at the promoter, or spread flom l-3.5 kbp, encompassing the
entire gene [249-251]. This discrepancy may be due to differences in crosslinking or
immunoprecipitation conditions, reflecting the difficulty in analyzing indirectly bound
factors. Indeed, in our study, promoter interactions by Groucho were observed only with
the use of a double crosslinking protocol.

The extensive contacts of Groucho over the repressed locus are strongly
reminiscent of the extended nucleoprotein structures deposited on regions repressed by
stable, heritably acting systems such as Polycomb group (PcG) proteins in animals and
Sir proteins in silent mating type loci and subtelomeric regions of yeast. There, chromatin
regions are modified and inhibited for formation of productive transcription complexes
[80, 252, 253]. Indeed, the association of activators and components of the transcriptional
machinery with repressed loci in these systems mirrors the continued binding of
activators and coactivators in the system we study here, suggesting that the limiting factor
for transcription occurs at a later stage [254-256]. What sort of inhibitory interaction
might be involved in this case? A number of recent reports have raised the possibility that
repressed, or non-activated promoters, feature RNA polymerase II that is blocked for
elongation, similar to the paused polymerase found at the hsp 70 locus under noninducing
conditions [257]. Gergen and colleagues find that RNA polymerase II is not displaced
flom the slp] gene upon repression with Runt, a Groucho-binding protein [258]. It is
possible that Groucho itself, through contacts with histone proteins, and/or recruitment of
deacetylases such as de3, establishes a chromatin environment that is inhibitory for

transcriptional elongation. We were unable to obtain reliable signals for RNA polymerase

84

II at this promoter, precluding a definitive statement about polymerase occupancy in
activated and repressed states, however.

A difference between the repression complex assembled by the Hairy repression
domain and by these other proteins is the transience of the effect; while PcG regulation is
linked to epigenetic modifications that allow repression to persist for an extended time
when PcG proteins are depleted [259], the regulation we see here is readily reversed upon
loss of the LexA-Hairy repressor. Similar effects are observed with elements regulated by
the endogenous Hairy protein; enhancers bearing Dorsal and Twist activator sites that are
repressed by Hairy in the blastoderm embryo are reactivated minutes later in the
gerrnband extended stage (M. Kulkarni, unpublished). Thus, Hairy appears to be
designed for highly effective, but readily reversible, repression, which may be usefirl in
particular developmental settings.

In contrast to a linear spreading model, an alternative picture of Groucho
interaction that is consistent with our observations is that the corepressor may be tethered
to the promoter region, forming larger multimeric complexes around which proximal and
distal portions of the gene are wrapped (“turban” model). This model may explain why
we often detect downstream interactions, albeit weaker ones, of the LexA-Hairy
repressor, particularly when employing the more extensive double crosslinking protocol.
In either case, Groucho itself may be important for interfering with activity of
transcription factors, or transcription of distant loci. Both of these models suggest that the
extensive spread or extensive contacts of Groucho is mechanistically linked to
transcriptional repression. However, it is possible that Groucho’s extensive contacts with

downstream regions are not the main effector of Hairy-mediated repression. Promoter-

85

proximal activities of Groucho, or of other Hairy corepressor proteins, may play the
decisive role in dictating long-range effects. However, extensive experimental evidence
indicates that Groucho plays a key role in repression mediated by Hairy, thus it seems
parsimonious to assume that Groucho activity on the repressed gene is important for
repression.

Our study demonstrates that repression by Hairy is associated with histone
deacetylation, which is certainly consistent with the nature of cofactors associating with
this protein. Interestingly, this modification appears to be restricted to regions close to the
repressor binding sites, which in our configuration places them close to the
transcriptional initiation site. How might this be related to the long-range effects
mediated by Hairy? One possibility is that Hairy, regardless of where it is bound, induces
characteristic changes on chromatin close to the transcriptional start site, which would
induce a dominant (and hence long-range) effect on target genes. Alternatively, the local
chromatin deacetylation may reflect the reversal of promoter-localized histone acetylases
(e.g. SAGA), and acetylation levels on other portions of the gene are already too low to
show robust deacetylation. A third possibility is that other Hairy-induced chromatin
modifications that are not assayed here are more extensive than the deacetylation.

Our study strongly supports a model for Hairy repression that involves contacts
between the Groucho corepressor and extended regions of the silenced gene, providing a
basis for the long-range repression observed for this protein that is independent of
activator displacement. An interesting question for future studies is how Groucho
spreading is limited, and whether specific chromatin signals modulate this activity. In

addition, such extended repression complexes might be specific to subsets of Hairy

86

 

targets. Genomic surveys indicate that Hairy is likely to associate with distinct cofactors
at different loci; future work will focus on identifying the roles of individual cofactors of

this repressor at genes that represent the diversity of Hairy targets in Drosophila.

87

Chapter IV

Conclusions and Future Directions

Development of a model system for the study of transcriptional regulation

Over the last 30 years, numerous studies have sought to understand the molecular
mechanism of repression by HES proteins [1, 2, 260, 261]. Functional characterizations
of multiple HES proteins have revealed a highly conserved role as regulators of cell fate
decisions (reviewed in Chapterl). In development, HES proteins typically function as
effectors of the Notch/Delta signaling pathway to negatively regulate genes involved in
establishing cellular identity [1, 2]. However, some HES proteins, like Drosophila Hairy,
are not under the control of Notch signaling and function instead to prepattem the initial
set of regulators specifying early pluripotent cellular precursors [2, 262, 263]. Members
of the HES family have a bHLH DNA-binding domain, a conserved Orange domain, and
a C-temrinal WRPW motif that interacts with the corepressor Groucho [l]. The high
degree of homology between HES proteins suggests an evolutionarily conserved
repression mechanism. A detailed molecular analysis of HES-mediated repression would
lead to a deeper understanding of multiple developmental pathways and would provide
opportunities for developing novel treatments for diseases found to be associated with
disruption of normal HES firnction.

In this project I developed a highly defined embryo system to study Hairy
repression in vivo. As discussed in Chapter 2, the use of exogenous activator and
repressor transgenes allowed a great degree of control over the system. The spatial and

temporal control of the activator and repressor inputs permitted the embryos to be treated

88

as a uniform collection of cells, analogous to tissue culture studies. Further analytical
power was afforded through the creation of alternate Drosophila lines carrying either the
reporter alone, reporter and activator, reporter and repressor, or all three transgenes. This
combinatorial analysis of promoter states proved extremely useful in validating the
conclusions presented here, and gave insights into both the transcriptional activation as
well as repression processes.

It should be noted that other groups have also succeeded in using whole embryo
systems to study the molecular mechanisms of transcriptional regulation [239, 258]. For
example, in a recent study, Drosophila T 0111 0" mutant embryos were used to determine
the genome wide occupancy of the Twist, Dorsal, and Snail transcription factors [239].
These embryos express a constitutively active form of the Toll receptor leading to
translocation of Dorsal into the nucleus and the uniform expression of Twist and Snail
[264]. Similarly, overexpression of th and Runt during the cellular blastoderm stage of
Drosophila development result in the uniform suppression of sloppy-paired-I (slpI),
while overexpression of th and Opa results in uniform activation [258]. However,
adapting these model systems to study different activators or repressors would require
extensive reengineering of the gene network that makes up the system. In contrast, the
embryo system developed here can readily be adapted to investigate the transcription
regulatory mechanisms of any protein containing modular repression or activation
domains.

While the repression observed in this study takes place in a short-range context,
there are two reasons why the conclusions of my study should also be valid in a long-

range context. First, the LexA-Hairy protein is able to suppress the activity of a

89

rhomboid-twist driven reporter even when bound 2 kbp upstream, indicating that it can
also firnction in a long-range context (Li Li, unpublished results). Furthermore, the terms
short-range and long-range refer to a functional characterization of the repressor and do
not imply any particular molecular mechanism [16, 22]. Thus, the conclusions derived
using the model system developed here are relevant to the general mechanism of Hairy

repression.

Haigy repression mechanism does not rely on activator or coactivator displacement

Using the highly defined embryo system discussed previously, I was able to show
that the repression mechanism of Hairy works even when the activator and coactivator
remain bound to the gene (Figure 2-6, 2-7). This result raises the possibility that Hairy
repression could involve direct targeting of the basal machinery. Indeed, repression of
slp] by th and Runt, the latter of which also binds Groucho, leads to the formation of a
paused Pol H complex downstream of the promoter [258, 265].

Preliminary experiments using the 8WG16 antibody showed a decrease in P0] H
signals at the promoter during LexA-Hairy repression when compared to the activated
state (Appendix B). Unexpectedly, the downstream Pol H signals were actually higher in
the repressed state than the activated state. A possible explanation for this result is that
LexA-Hairy repression could induce a transcriptional block near the 3’ end of the gene.
Studies in yeast have shown that both Gal4 and SAGA can directly interact with
components of the basal machinery and promote PIC assembly even in the absence of
Gcn5 HAT activity [181, 182, 186]. The continued presence of Gal4 and SAGA during

repression could lead to the observed accumulation of Pol H in the coding region.

90

 

Unfortunately, the lack of downstream Pol H signals in the activated state and the poor
reproducibility of the Pol II immunoprecipitations prevented any definite conclusions to.
Therefore, future studies should focus on determining if Hairy repression involves
blocking PIC formation or inhibiting Pol H elongation.

Since Gal4 has been shown to “lock in” and form a highly stable Gal4-promoter
complex, it is possible that other activators with weaker DNA affinities could be
displaced by Hairy repression [235]. In order to test this possibility, binding sites for
endogenous activators could be used instead of Gal4 to drive expression of the reporter
construct. Antibodies against the activator of interest (if available) could be used to track
activator binding. Alternatively, the activators could be epitope-tagged, and expressed
using a ubiquitous driver.

Another possibility is to use the embryo system I have developed to study other
transcriptional repressors such as Knirps or Giant. In fact, the original goal of this project
was to understand the molecular mechanisms differentiating short- flom long-range
repression by tethering a Knirps or Hairy repression domain to LexA and using the lacZ
reporter shown in Figure 2-1. While both LexA-Knirps and LexA-Hairy could repress the
reporter described in Chapter 2 (Figure 2-2), only LexA-Hairy could suppress gene
activation in the embryo system (Figure 2-1, Appendix A). This result was unexpected,
since it had previously been shown that the short-range repressor Giant can repress a
synthetic enhancer with a similar arrangement of activator and repressor binding sites (2
Giant, 5 UAS, 2 Giant) [229]. Since Knirps has been shown to be able to suppress the
activity of three high affrrrity UAS sites, the inability of LexA-Knirps to repress five

tandem UAS sites is not due to repressor-activator specificity [230]. Perhaps Giant is a

91

 

 

stronger repressor then Knirps, and that is why it is able to repress 5 UAS sites. Indeed, a
comparison of the relative potency ‘ of the short-range repressors Giant, Knirps, and
Kruppel, in the context of three UAS sites, showed that in general Knirps has a lower
activity [230].

As discussed in the introduction, short-range repressors allow enhancer
modularity by blocking those activators bound within 150 bp flom its target sequence
[18]. One possible mechanism by which short-range repressors could block nearby
activators is by displacing locally bound activators flom the DNA. By modifying the
embryo system to include different short-range repressor domains, this hypothesis could
be directly tested. A recent genome-wide study looking at the binding of the short-range
repressor Snail, and the Twist and Dorsal activators show that all three proteins can bind
simultaneously to inactive enhancers, suggesting that short-range repressors also do not

displace activators [239].

Haifl repression is associated with Groucho binding to the entire gene locus

A key finding in this project is that the Hairy repression mechanism is associated
with Groucho binding to a wide region encompassing the gene (Figure 3-1). This result
provides strong evidence for a model, first proposed by Al Courey, that Hairy long-range
repression is mediated by a Groucho spreading mechanism, similar to the spreading of
the Sir2 complex in yeast ( Fig 4-1) [17, 79]. A similar mechanism has also been
proposed for the yeast homolog Tupl, but evidence of Tupl spreading has been

controversial [249-251]. Like Tupl, Groucho has an affinity for hypoacetylated histones,

92

 

 

Figure 4-1. Groucho spreading model

The model for Groucho spreading involves: l) recruitment of Groucho by a long-range
repressor such as Hairy, 2) deacetylation of the surrounding histones by the associated
de3 histone deacetylase, 3) Groucho binding to adjancent nucleosomes through
interactions with the hypoacetylated H3 and H4 histone tails." Images in this dissertation
are presented in color.

93

suggesting a model wherein Groucho-mediated de3 recruitment deacetylates
surrounding histones; local histone deacetylation in turn promotes further Groucho
spreading [17, 266]. The association of Groucho/de3 to large tracts of the gene could
repress transcription by catalyzing the formation of a heterochromatin-like structure.
Interestingly, Sir2-dependent heterochromatic silencing of the yeast Hsp82 gene does not
prevent the HSF activator, or TBP, flom binding to the promoter, suggesting possible
structural similarities between Sir2- and Groucho-remodeled chromatin [254]. In support
of the importance of Groucho spreading for long-range repression, mutations in the
Groucho oligomerization domain abolish its repressive activity [244].

Since the extensive Groucho binding observed in this study is in the context of an
artificial model system, future studies should focus on ascertaining the relevance of
Groucho spreading, by looking at Groucho binding to endogenous enhancers. A possible
approach could be to chromatin immunoprecipitate Groucho flom embryos or tissue
culture cells, followed by hybridization to a tiled microarray (ChIP-chip), or by direct
sequencing of the immunoprecipitated DNA (ChIP-PET) [267].

As discussed in Chapter 3, this study raises the question of what regulates
Groucho spreading so that it does not interfere with the transcriptional activation of
nearby genes. Future studies should focus on elucidating the mechanism by which
Groucho spreading is limited and whether histone modifications have any role in
regulating this activity. Additionally, analysis of LexA-Hairy repression in a long-range
context could determine if Groucho spreading is limited to the basal promoter and ORF,
or if it can also spread through upstream cis regulatory regions. A related question

regarding Hairy repression is whether boundary elements can block Groucho spreading.

94

For example, the Su(Hw) insulator protein can partially suppress heterochromatin
spreading in Drosophila [268]. Similarly, the vertebrate insulator element cHS4 functions
to prevent heterochromatin silencing in the chicken B-globin locus [269]. This type of
mechanism might account for why the Su(Hw) insulator can block Hairy repression of a
downstream enhancer while allowing closely linked activators to stimulate transcription
[22].

In addition to Groucho, Hairy has also been found to associate with the Sir2 and
CtBP corepressors [69, 71]. Future studies should address the role of these cofactors in
mediating Hairy long-range repression. Of particular interest would be to understand the
role of Sir2, because of its critical involvement in heterochromatin silencing. A genome-
wide survey of Hairy, Groucho, CtBP, and Sir2 binding, found that Hairy associates with
different combinations of cofactors at target loci, suggesting that Hairy cofactor
recruitment is context dependent [75]. Interestingly, only a small percentage of putative
Hairy targets were associated with Groucho, with the majority associated with CtBP and
Sir2 [75].

A possible explanation for the apparent context dependency of cofactor
recruitment is that the chromatin structure regulates cofactor binding. Indeed, Groucho
was found to bind preferentially to hypoacetylated histone H3 [85]. Similarly, the Sir2
complex preferentially binds deacetylated histones through the Sir3 and Sir4 subunits
[270]. However, initial binding of Sir2 does not require HDAC activity, suggesting that it
is not dependent on histone acetylation [271]. The role of chromatin in regulating
cofactor recruitment could be investigated by analyzing cofactor binding to the lacZ

reporter in the context of the reporter alone, reporter + activator, reporter + repressor, and

95

reporter + repressor + activator, since each of these states is associated with a specific
chromatin structure. Binding of both LexA-Hairy and Gal4 to the promoter was
associated with a marked decrease in the acetylation level (Figure 3-ZB). In contrast,
binding of LexA-Hairy alone has only a slight effect on H4 acetylation levels, suggesting
that in this context LexA-Hairy does not recruit a cofactor that mediates this activity

(Figure 3-2A, compare panels 3 and 4).

Hair_'y repression is associated with histone deaceglation and chromatin remodeling

It has been shown that Hairy associates with histone deacetylases [71, 84],
consistent with this finding, this study provides direct evidence of HDAC recruitment and
subsequent histone deacetylation by Hairy. As can be seen in Fig 3-1, the de3 HDAC
was shown to track with Groucho up to 2 kbp downstream of the transcriptional start site.
Curiously, LexA-Hairy binding to the promoter resulted only in local histone H3
deacetylation up to a distance of 1 kbp (Figure 3-3A, B). However, it should be noted that
even in the unrepressed state, the histone H3 acetylation signals flom the 3’ end of the
lacZ reporter decreased dramatically relative to the promoter, making the task of
detecting changes in histone acetylation more difficult (Figure 3-3A, B). Furthermore,
there is a 5’ to 3’ decrease of Groucho and de3 signals, implying a smaller reduction in
acetylated histone levels.

In this project I have been able to show that chromatin remodeling is associated
with Gal4 activation as well as Hairy repression. Comparisons of histone H3 signals
between the unactivated and activated state revealed a marked decrease in H3 promoter

occupancy during Gal4-mediated activation (Figure 3-2B). This result is consistent with

96

previous studies that Gal4 activation is associated with the loss of promoter-proximal
histones in yeast [212]. Irnportantly, during LexA-Hairy repression there was a
reproducible increase in promoter bound H3 signals. In yeast, repression by the Groucho
homolog Tupl has also been shown to involve nucleosome positioning, suggesting a
conserved repression mechanism [272]. In order to have a better understanding of Hairy-
dependent chromatin remodeling, micrococcal nuclease digestion can be used to
determine if repression involves nucleosome positioning such as that seen for a subset of

Tupl-Ssn6 repressed genes [272].

Gal4 activation is associated with demethylation of K27

Since K27 methylation is a mark of heterochromatin, I analyzed whether Hairy
repression was associated with K27 methylation. To this end, chromatin
immunoprecipitation assays were carried out using antibodies against dimethylated H3
K27. The result showed that this modification was not associated with Hairy-mediated
repression. Instead, K27 methylation was correlated with Gal4 activation (Figure 3-2B).
Future studies should continue to investigate the possible involvement of additional
histone marks such as H2B ubiquitylation. An exhaustive analysis of histone
modification marks would most likely yield additional insights into Hairy repression

activity.

Reactivation of gene activig after LexA-Haigy repression

Using the embryo system, I was able to show that induction of the LexA-Hairy

fusion protein led to transient repression of the lacZ reporter construct (Figure 2-4). In

97

contrast, other long-range repression mechanisms such as those mediated by the
Polycomb, HP], and Sir proteins, lead to stable gene silencing [16]. Reactivation of the
lacZ reporter was accompanied by a corresponding increase in histone H3 and H4
acetylation levels (Figure 3-2, 3-3). Unexpectedly, during reactivation H4 acetylation
levels in the coding region were higher than during the activated state (Figure 3-3B). One
possible model that could help explain the high H4 acetylation levels observed is that
Groucho/de3 spreading primes the gene for reactivation. Studies in yeast have shown
that the de3S HDAC complex is targeted to the open reading flame of active genes by
binding to methylated H3 K36 histones [273]. The de3S subunit Eaf3 targets the
complex through its chromodomain which specifically recognizes methylated H3 K36
[274]. Since the EaB protein is also shared with the NuA4/1“ ip60 complex that catalyzes
H4 acetylation, it suggests a reactivation model wherein the de38 HDAC complex is
exchanged for the NuA4/Tip60 HAT complex, leading to generalized H4
hyperacetylation in the ORF [275]. Therefore, future studies could address whether
Groucho/de3 spreading promotes NuA4/Tip60 recruitment during reactivation.

An alternative reactivation model involves Groucho-mediated deposition of the
conserved H2A.Z histone variant. In yeast, the SWRl chromatin remodeling complex has
been shown to catalyze the exchange of H2A with the H2A.Z histone variant [276, 277].
H2A.Z deposition functions to prevent heterochromatin spreading and to promote
efficient reactivation at repressed promoters [278-280]. Irnportantly, the Groucho
homolog Tupl promotes the deposition of H2A.Z on the GAL] promoter nucleosome,
suggesting that it primes the promoter for reactivation [281]. Since NuA4 and SWRl

share multiple subunits and are both conserved in Drosophila [282], I propose a

98

reactivation model wherein Groucho/de3 spreading induces recruitment of SWRl and
NuA4 to the coding region (Figure 4-2). Recruitment of SWR] and NuA4 could in turn
catalyze the deposition of H2A.Z and hyperacetylate H4, respectively leading to a looser
chromatin structure and facilitating reactivation. In order to test this model, H2A.Z

recruitment during repression and reactivation can be determined.

Contrasting transcriptional states: Unactivated versus Repressed

An important observation of my study is that repression does not reset the
chromatin structure to the unactivated state. As can be seen in Figure 3-2B, in the
unactivated state there is a high level of methylated H3 K27 histones, suggesting that the
reporter gene is in a heterochromatic environment. Gal4-mediated activation leads to a
reduction of methylated K27, relative to H3, consistent with an activation mechanism that
remodels chromatin flom a heterochromatic silenced state to a euchromatic active state.
Heat shock induction of LexA-Hairy leads to the suppression of lacZ expression, yet it
does not reverse the loss of K27 methylation seen during activation. Thus, the

unactivated transcriptional state is not equivalent to the repressed state.

99

Figure 4-2. Proposed mechanism for Groucho-dependent reactivation

During repression Groucho and de3 are found associated with the entire coding region.
H3 K36 methylation stabilizes the de3 binding through direct interactions with the
chromodomain-containing Eaf3 subunit. Since the Eaf3 subunit is shared between de3
and NuA4, reactivation could occur by a subunit exchange that allows NuA4 to be
recruited to the coding region. Subsequently, NuA4 can catalyze the hyperacetylation of
H4, creating a looser chromatin structure. NuA4 and the ATP-dependent chromatin
remodeling complex SWRl share four subunits. SWR] can catalyze the exchange of
H2A with H2A.Z to prime the gene for reactivation. Images in this dissertation are
presented in color.

100

 

SWFl]

NuA4

 

Figure 4-2. Proposed mechanism for Groucho-dependent reactivation

101

Short versus long-range repression: Two sides of the same coin

Previous functional characterizations of repressors in Drosophila pointed to a role
of CtBP in mediating short-range repression and a role of Groucho in mediating long-
range repression [72-74]. However, a recent study suggests that Groucho-mediated
repression can have a short-range activity [62, 283]. Additionally, biochemical
purification of the short-range repressor Knirps indicated that Groucho is a part of the
repressor complex (Sandhya Payankaularn, unpublished results). Similarly, the CtBP
cofactor which potentiates short-range repressors such as Knirps, has been found to
associate with the Hairy long-range repressor and to be important for the repression of
putative Hairy targets [69, 7 5]. Taken together, the evidence suggests that Groucho and
CtBP form part of a common repression pathway. Indeed, some repressors like the
Drosophila Hairless and Brinker have been shown to require both CtBP and Groucho to
for their activity [284, 285].

The fact that Groucho has been shown to be important for Hairy long-range
repression raises the question of what prevents Groucho-mediated long-range repression
in a short-range context [72]. Perhaps the range of Groucho repression is regulated by
limiting the degree to which Groucho can spread in chromatin. One possibility is that the
extent of Groucho spreading is determined by the total number of repressor binding sites
present. In support of this model is the fact that the range of Hairy repression increases
with additional Hairy sites [22]. Alternatively, the distance to which Groucho can spread
could depend on the chromatin context. An intriguing idea is that Groucho spreading
could be limited by the same mechanisms that limit the spreading of silent

heterochromatin.

102

In yeast, there are several chromatin-based mechanisms that prevent the Sir2
complex flom spreading into euchromatic regions [286]. The first is through site-specific
acetylation of H4 K16 by the Esal and SasZ histone acetyltransferases [287, 288]. This
modification has been shown to be specifically enriched at heterochromatin-euchromatin
boundaries [289]. A second mechanism involves the exchange of H2A for the H2A.Z
histone variant catalyzed by the SWRl complex [282]. Deposition and subsequent NuA4-
mediated acetylation of H2A.Z has been shown to be required for the formation of
heterochromatin borders [279, 290]. A third mechanism involves the creation of
nucleosome flee regions [29]]. In yeast, heterochromatic spreading flom the silent
mating type locus HMR is blocked by the neighboring tRNAn" gene, which is partially
depleted of histones [292]. Recently, it has been reported that Set2 methylation of H3
K36 can also prevent heterochromatin spreading [293]. A combination of these
mechanisms could account for the observed differences in the range of Groucho

repression.

Contextual dependency of transcriptional repression: Implications for modeling cis

regulatogy elements

One of the most pressing problems in the post-genome era is trying to develop a
conceptual flamework that can be used to find and decode cis regulatory elements. The
Drosophila embryo has become an ideal model system because many of the transcription
factors and cis regulatory elements involved in early embryogenesis have been
characterized [294]. Currently, efforts are being made by several labs to develop

transcription regulatory models that can predict gene expression patterns directly flom

103

the DNA sequence [294-296]. An underlying assumption of these efforts is that the cis
regulatory elements under study are flexible with respect to their binding site
arrangement, with more then one “solution” leading to a particular gene expression
pattern [297]. Comparative phylogenetic analysis of the eve stripe2 enhancer supports
this idea, as the stripe 2 enhancers flom four different Drosophila species show identical
stripe placement despite varying widely in binding site arrangement [298]. However, if
the range of Groucho repression is context dependent, it would place restrictions on the
evolution of cis regulatory elements that recruit Groucho-associated repressors.

The fact that Groucho-dependent repression can act in a short- or long-range
manner, suggests that new design principles are needed to model enhancer activity.
Previous work flom the Amosti lab sought to discover “grammatical” rules governing the
function and design of modular enhancers [229, 299]. As a result, four basic parameters
were determined to be critical for modeling short-range transcriptional repression: 1) total
number of repressor binding sites, 2) repressor-activator stoichiometry, 3) binding site
affinity, and 4) binding site arrangement [229]. In order to extend the applicability of this
approach to include Groucho-associated repressors, additional grammatical rules must be
implemented. For example, if we assume that Groucho spreading, like Sir2, is blocked by
H4 K16 acetylation, then activators that recruit the NuA4/Tip60 HAT, which mediates
this modification, could result in short-range repression by Groucho-associated factors. In
addition, sequences that tend to exclude nucleosome binding could also have an impact
on the type of repression activity. Therefore, future studies should focus on discovering
novel grammatical rules that take into account those parameters that influence the range

of repression.

104

Concluding remarks

In summary, I have presented strong evidence that Hairy long-range repression is
mediated by a mechanism involving Groucho spreading and chromatin remodeling. In the
literature, there are several examples of similar long-range repression mechanisms.
However, a key feature that differentiates the repression mechanism studied here flom
previously characterized long-range mechanisms is the easy reversibility of the
repression. The transient nature of repression could be a reflection of an evolutionarily
conserved mechanism whereby the corepressor Groucho primes the gene for reactivation.
Efficient reactivation of HES target genes could be of critical importance for the proper
regulation of developmental pathways. For example, the role of HES proteins as part of
the “segmentation clock” requires repetitive repression-reactivation cycles in order to
generate waves of transcriptional activity [140, 145]. Inefficient reactivation of HES
target genes could lead to uncoordinated patterns of gene expression and disruption of the
segmentation process.

In addition, I have succeeded in developing and characterizing a highly defined
embryo system that can serve as a powerful platform for future studies of Hairy long-
range repression. Such studies could lead to a better understanding of the molecular
mechanisms by which cell fate decisions are regulated. Furthermore, the different
Drosophila lines I have created, capable of producing alternative promoter states, can be
used to provide novel insights into how the chromatin context affects transcriptional
repressor activity. A better understanding of the role of chromatin in regulating the range
of transcriptional repression is essential in order to develop predictive quantitative

models of enhancer function.

105

APPENDIX A

Characterization analysis of the LexA-Knirps fusion protein

One of the initial objectives of the project was to understand the molecular
mechanism of short-range repression. To this end, a fusion construct of LexA and Knirps
was tested for its ability to repress in the embryo system and in a rhomboid-twist driven
reporter. The repression domain used for Knirps encompassed the amino acids 75-429.
This domain had previously been shown to be sufficient to mediate short-range
repression of a heterologous stripe2 enhancer when tethered by the Gal4 DNA-binding
domain [300]. The repression domain contains the CtBP P)G)LS binding motif near the
C-terrninus but lacks the endogenous DNA-binding domain. Previously, it had been
shown that a lacZ reporter containing tandem Knirps sites placed 5’ and 3’ of three Gal4
UAS sites was repressed in the region of Knirps expression [229].

In order to test the repression activity of LexA-Knirps, a Kriippel promoter was
used to drive expression in a central band of the embryo. A transgenic line carrying the
LexA-Knirps construct was crossed to both the embryo system (carrying only the Gal4
activator and reporter), and a rhomboid-twist driven reporter with LexA sites near the
promoter. Using in situ hybridizations against lacZ, repression was observed only in the

rhomboid-twist reporter but not in the embryo system (Figure A-l).

106

55 bp
SxGal4 <——-'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2xLexA 2xLexA "* _. _J ...J
55 bp
<-——>
r— .
rho [M] - I"
2xLexA ..
LexA-Knirps
I . I

f Kruppel , I" I

 

 

 

Figure A-l. LexA-Knirps does not repress in the model system

Top panel: A line containing a Kruppel driver expressing the LexA-Knirps (75-429) was
crossed to the defined embryo system; at left is a schematic diagram of the reporter
construct. Middle panel: LexA-Knirps mediates efficient repression of a rhomboid-twist
driven reporter when bound within 55 bp of the promoter. Bottom panel: Antibody
staining against LexA shows that LexA-Knirps is being expressed in these embryos.
Images in this dissertation are presented in color.

107

Materials and Methods

Plasmids. The reporter constructs were constructed as described in Chapter 2. The LexA-
Knirps fusion construct was made by joining a BamHI/Kpnl flagrnent containing a Kozak
sequence, initiator ATG, and coding sequence for the entire LexA protein (amino acid
residues 1-202) in flame to a KpnI/Xbal flagrnent containing the Knirps repression
domain (residues 75-429). The fusion construct was placed in a modified pKreg vector
containing the Kruppel driver. The resulting plasmids were introduced into the

Drosophila gerrnline by P-element transformations as described [231].

In situ hybridization and antibody staining of Drosophila embryos. Embryos were
fixed for in situ hybridization and stained using a digoxigenin-UTP labeled antisense
RNA probe to lacZ as described [231]. Antibody staining for LexA-Hairy expression was
done using a 1:221 mixture of three mouse monoclonal antibodies raised against LexA,
YN-lexA-2-12 (2 pg/ml, 1:2100), YN-lexA-6-10 (4 pg/ml, 1:2023), YN-lexA-16-7 (4
pg/ml, 1:3950) obtained as a gift flom Steve Triezenberg [233]. Quantitation of lacZ
staining was done by scoring the stained embryos into unstained, lightly stained, and
darkly stained flom 10-13 random fields. Total embryos scored per slide varied between

135 and 457.

108

APPENDIX B

Pol II accumulates in the ORF during repression

As discussed in Chapter 4, one of the possible mechanisms of Hairy repression could be
blocking Pol II binding or elongation. Using the embryo system, I collected embryos
carrying either the reporter and activator; or the reporter, activator, and repressor
transgenes. After a 20 min heatshock and 30 min recovery a chromatin
immunoprecipitation assay against the Po] H CTD revealed a loss of Pol H flom the
promoter during repression. At the same time there was an increase of Pol H in the open
reading flame, suggesting that Hairy repression does not suppress Pol H recruitment.

Instead, repression could involve a block to elongation near the 3’ end of the gene.

109

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.E‘
E
at; 3 fi'
> (U Q)
Q fi % 5:3 0. T!
93 c: f t: 58 2
* + 30 + -
Promoter
-" U‘I'” . + 30 +
-—.~.. + 30 '1' "
+1kb
» —--«~ I+w ++
. -——u.. + 30 + -
Intergenrc
——-~- + 30 + +

 

 

 

Figure B-l. Chromatin immunoprecipitation of Pol II.

Embryos expressing only the activator or the repressor were collected and heatshocked
for 20 min at 37°C and recovered for 30 min at room temperature. Activated embryos
showed a strong Pol 11 signal at the promoter but no Pol H at + 1 kb. During repression
Pol II signals decreased slightly at the promoter but increased at +1 kb. PCR using
primers against an intergenic region of chromosome 3 did not give any signals indicating
that the signals observed are specific. Presence or absence of Gal4, and the heatshock-
inducible LexA-Hairy transgene noted by +/-. Input titration was 10%, 2%, 1%, 0.5%,
and 0.25%. Images in this dissertation are presented in color.

110

10.

ll.

12.

13.

References

Davis, KL. and D.L. Turner, Vertebrate hairy and Enhancer of split related
proteins: transcriptional repressors regulating cellular diflerentiation and
embryonic patterning. Oncogene, 2001. 20(58): p. 8342-57.

Fisher, A. and M. Caudy, The function of hairy-related bHLH repressor proteins
in cell fate decisions. Bioessays, 1998. 20(4): p. 298-306.

Carroll, S.B., A. Laughon, and BS. Thalley, Expression, function, and regulation
of the hairy segmentation protein in the Drosophila embryo. Genes Dev, 1988.
2(7): p. 883-90.

Struhl, K., Yeast transcriptional regulatory mechanisms. Annu Rev Genet, 1995.
29: p. 651-74.

Wolpert, L., et al., Principles of Development. Oxford University Press, 1998.

Campos-Ortega, J .A. and V. Hartenstein, The embryonic development of
Drosophila melanogaster. 2nd Edition. Springer, 1997.

Veraksa, A., M. Del Carnpo, and W. McGinnis, Developmental patterning genes
and their conserved functions: from model organisms to humans. Mol Genet
Metab, 2000. 69(2): p. 85-100.

Small, 8., Mechanisms of segmental pattern formation in Drosophila
melanogaster Reproductive biology of invertebrates. In Reproductive biology of
invertebrates. K.G. Adiyodi and R.G. Adiyodi, eds. John Wiley and Sons, USA,
1997. VII: p. 137-178.

Nusslein-Volhard, C. and E. Wieschaus, Mutations aflecting segment number and
polarity in Drosophila. Nature, 1980. 287(5785): p. 795-801.

Sanson, B., Generating patterns from fields of cells. Examples from Drosophila
segmentation. EMBO Rep, 2001. 2(12): p. 1083-8.

Gray, 8., et al., Transcriptional repression in the Drosophila embryo. Philos
Trans R Soc Lond B Biol Sci, 1995. 349(1329): p. 257-62.

F ujioka, M., et al., Analysis of an even-skipped rescue transgene reveals both
composite and discrete neuronal and early blastoderm enhancers, and multi-
stripe positioning by gap gene repressor gradients. Development, 1999. 126(11):
p. 2527-38.

Johnson, AD, The price of repression. Cell, 1995. 81(5): p. 655-8.

111

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Alberts, B., et al., Molecular Biology of the Cell. New York, Garland, 1983: p.
440.

Herschbach, B.M. and AD. Johnson, Transcriptional repression in eukaryotes.
Annu Rev Cell Biol, 1993. 9: p. 479-509.

Gaston, K. and PS. J ayararnan, Transcriptional repression in eukaryotes:
repressors and repression mechanisms. Cell Mol Life Sci, 2003. 60(4): p. 721-4].

Courey, A]. and S. Jia, Transcriptional repression: the long and the short of it.
Genes Dev, 2001. 15(21): p. 2786-96.

Gray, 8., P. Szymanski, and M. Levine, Short-range repression permits multiple
enhancers to function autonomously within a complex promoter. Genes Dev,
1994. 8(15): p. 1829-38.

Amosti, D.N., et al., The gap protein knirps mediates both quenching and direct
repression in the Drosophila embryo. Embo J, 1996. 15(14): p. 3659-66.

Gray, S. and M. Levine, Short-range transcriptional repressors mediate both
quenching and direct repression within complex loci in Drosophila. Genes Dev,
1996. 10(6): p. 700-10.

Hewitt, G.F., et al., Transcriptional repression by the Drosophila giant protein:
cis element positioning provides an alternative means of interpreting an eflector
gradient. Development, 1999. 126(6): p. 1201-10.

Barolo, S. and M. Levine, hairy mediates dominant repression in the Drosophila
embryo. Embo J, 1997. 16(10): p. 2883-91.

Cai, H.N., D.N. Amosti, and M. Levine, Long-range repression in the Drosophila
embryo. Proc Natl Acad Sci U S A, 1996. 93(18): p. 9309-14.

Roseman, R.R., et al., Long-range repression by multiple polycomb group (PcG)
proteins targeted by fusion to a defined DNA-binding domain in Drosophila.
Genetics, 2001. 158(1): p. 291-307.

Jenkins, H.L. and CA. Spencer, RNA polymerase 11 holoenzyme modifications
accompany transcription reprogramming in herpes simplex virus type I-infected
cells. J Virol, 2001. 75(20): p. 9872-84.

Amosti, D.N., Multiple Mechanisms of Transcriptional Repression in Eukaryotes.
In M. Gossen, J. Kauflnann, and S. J. Triezenberg (ed.), Handbook of
experimental pharmacology: transcription factors, 2004. 166: p. 33-67. Springer,
Heidelberg.

Brooks, CL. and W. Gu, Ubiquitination, phosphorylation and acetylation: the
molecular basis for p53 regulation. Curr Opin Cell Biol, 2003. 15(2): p. 164-7].

112

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

Martinez-Balbas, M.A., et al., Regulation of EZF ] activity by acetylation. Embo J,
2000. 19(4): p. 662-7].

Carrozza, M.J., et al., Ga180 confers specificity on HAT complex interactions with
activators. J Biol Chem, 2002. 277(27): p. 24648-52.

Karin, M. and Y. Ben-Neriah, Phosphorylation meets ubiquitination: the control
of NF -[kappa]B activity. Annu Rev Irnmunol, 2000. 18: p. 621-63.

Jaynes, J .B. and RH. O'Farrell, Activation and repression of transcription by
homoeodomain-containing proteins that bind a common site. Nature, 1988.

336(6201): p. 744-9.

Kuchin, S. and M. Carlson, Functional relationships of SrbIO-SrbII kinase,
carboxy-terrninal domain kinase C TDK-I, and transcriptional corepressor Ssn6-
T up]. Mol Cell Biol, 1998. 18(3): p. 1163-71.

Zaman, 2., et al., Interaction of a transcriptional repressor with the RNA
polymerase 11 holoenzyme plays a crucial role in repression. Proc Natl Acad Sci
U S A, 2001. 98(5): p. 2550-4.

Green, SR. and AD. Johnson, Promoter-dependent roles for the SrbIO cyclin-
dependent kinase and the Hda] deacetylase in T up] -mediated repression in

Saccharomyces cerevisiae. Mol Biol Cell, 2004. 15(9): p. 4191-202.

Li, C. and J .L. Manley, Even-skipped represses transcription by binding TATA
binding protein and blocking the TFIID-TATA box interaction. Mol Cell Biol,
1998. 18(7): p. 3771-81.

Sauer, F., et al., Control of transcription by Kruppel through interactions with
TFIIB and T FIIE beta. Nature, 1995. 375(6527): p. 162-4.

Luger, K., et al., Crystal structure of the nucleosome core particle at 2.8 A
resolution. Nature, 1997. 389(6648): p. 251-60.

Schalch, T., et al., X-ray structure of a tetranucleosome and its implications for
the chromatin fibre. Nature, 2005. 436(7047): p. 138-4].

Tremethick, D.J., Higher-order structures of chromatin: the elusive 30 nm fiber.
Cell, 2007. 128(4): p. 651-4.

Razin, S.V., et al., Chromatin domains and regulation of transcription. J Mol
Biol, 2007. 369(3): p. 597-607.

Comings, D.E., Mechanisms of chromosome banding. IV. Optical properties of
the Giemsa dyes. Chromosoma, 1975. 50(1): p. 89-110.

113

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

Brower-Toland, B., et al., Specific contributions of histone tails and their
acetylation to the mechanical stability of nucleosomes. J Mol Biol, 2005. 346(1):
p. 135-46.

Chandy, M., et al., S WI/SNF displaces SAGA-acetylated nucleosomes. Eukaryot
Cell, 2006. 5(10): p. 1738-47.

Kouzarides, T., Chromatin modifications and their function. Cell, 2007. 128(4): p.
693-705.

Hooper, K.L., S.M. Parkhurst, and D. Ish-Horowicz, Spatial control of hairy
protein expression during embryogenesis. Development, 1989. 107(3): p. 489-
504.

Lardelli, M. and D. Ish-Horowicz, Drosophila hairy pair-rule gene regulates
embryonic patterning outside its apparent stripe domains. Development, 1993.
118(1): p. 255-66.

Ohsako, S., et al., Hairy function as a DNA-binding helix-loop-helix repressor of
Drosophila sensory organ formation. Genes Dev, 1994. 8(22): p. 2743-55.

Knust, E., et al., The enhancer of split locus and neurogenesis in Drosophila
melanogaster. Dev Biol, 1987. 122(1): p. 262-73.

(16 Celis, J .F., et al., Functional relationships between Notch, Su(H) and the
bHLH genes of the E(spl) complex: the E(spl) genes mediate only a subset of
Notch activities during imaginal development. Development, 1996. 122(9): p.
2719-28.

Sasai, Y., et al., Two mammalian helix-loop-helix factors structurally related to
Drosophila hairy and Enhancer of split. Genes Dev, 1992. 6(12B): p. 2620-34.

Ishibashi, M., et al., Molecular characterization of HES-2, a mammalian helix-
loop-helix factor structurally related to Drosophila hairy and Enhancer of split.
Eur J Biochem, 1993. 215(3): p. 645-52.

Wrischnik, LA. and OJ. Kenyon, The role of [in-22, a hairy/enhancer of split
homolog, in patterning the peripheral nervous system of C. elegans.
Development, 1997. 124(15): p. 2875-88.

Oates, AC. and R.K. Ho, Hairy/E(spl)-related (Her) genes are central
components of the segmentation oscillator and display redundancy with the

Delta/Notch signaling pathway in the formation of anterior segmental boundaries
in the zebrafish. Development, 2002. 129(12): p. 2929-46.

Dawson, S.R., et al., Specificity for the hairy/enhancer of split basic helix-loop-
helix (bHLH) proteins maps outside the bHLH domain and suggests two

114

 

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

separable modes of transcriptional repression. Mol Cell Biol, 1995. 15(12): p.
6923-31.

Paroush, 2., et al., Groucho is required for Drosophila neurogenesis,
segmentation, and sex determination and interacts directly with hairy-related

bHLH proteins. Cell, 1994. 79(5): p. 805-15.

Kang, S.A., J .H. Seol, and J. Kim, The conserved WRPW motif of Hes6 mediates
proteosomal degradation. Biochem Biophys Res Commun, 2005 . 332(1): p. 33-6.

Aliflagis, P., et al., A network of interacting transcriptional regulators involved in
Drosophila neural fate specification revealed by the yeast two-hybrid system.
Proc Natl Acad Sci U S A, 1997. 94(24): p. 13099-104.

Taelrnan, V., et al., Sequences downstream of the bHLH domain of the Xenopus
hairy-related transcription factor-I act as an extended dimerization domain that
contributes to the selection of the partners. Dev Biol, 2004. 276(1): p. 47-63.

Ish-Horowicz, D. and SM. Pinchin, Pattern abnormalities induced by ectopic
expression of the Drosophila gene hairy are associated with repression of ftz
transcription. Cell, 1987. 51(3): p. 405-15.

Jimenez, G., S.M. Pinchin, and D. Ish-Horowicz, In vivo interactions of the
Drosophila Hairy and Runt transcriptional repressors with target promoters.

Embo J, 1996. 15(24): p. 7088-98.

Van Doren, M., et al., Negative regulation of proneural gene activity: hairy is a
direct transcriptional repressor of achaete. Genes Dev, 1994. 8(22): p. 2729-42.

Nibu, Y., H. Zhang, and M. Levine, Local action of long-range repressors in the
Drosophila embryo. Embo J, 2001. 20(9): p. 2246-53.

Takebayashi, K., et al., Structure, chromosomal locus, and promoter analysis of
the gene encoding the mouse helix-Ioop-helix factor HES-1. Negative
autoregulation through the multiple N box elements. J Biol Chem, 1994. 269(7):
p. 5150-6.

Hoshijima, K., et al., Transcriptional regulation of the Sex-lethal gene by helix-
loop-helix proteins. Nucleic Acids Res, 1995. 23(17): p. 3441-8.

Ip, Y.T., et al., The dorsal morphogen is a sequence-specific DNA-binding protein
that interacts with a long-range repression element in Drosophila. Cell, 1991.
64(2): p. 439-46.

Jennings, B.H., D.M. Tyler, and SJ. Bray, Target specificities of Drosophila

enhancer of split basic helix-loop-helix proteins. Mol Cell Biol, 1999. 19(7): p.
4600-10.

115

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

Jimenez, G., Z. Paroush, and D. Ish—Horowicz, Groucho acts as a corepressor for
a subset of negative regulators, including Hairy and Engrailed. Genes Dev, 1997.
11(22): p. 3072-82.

Fisher, A.L., S. Ohsako, and M. Caudy, The WRPW motif of the hairy-related
basic helix-loop-helix repressor proteins acts as a 4-amino-acid transcription

repression and protein-protein interaction domain. Mol Cell Biol, 1996. 16(6): p.
2670-7.

Poortinga, G., M. Watanabe, and SM. Parkhurst, Drosophila CtBP: a Hairy-
interacting protein required for embryonic segmentation and hairy-mediated
transcriptional repression. Embo J, 1998. 17(7): p. 2067-78.

Phippen, T.M., et al., Drosophila C-terminal binding protein functions as a
context-dependent transcriptional co-factor and interferes with both mad and
groucho transcriptional repression. J Biol Chem, 2000. 275(48): p. 37628-37.

Rosenberg, MI. and SM. Parkhurst, Drosophila Sir2 is required for
heterochromatic silencing and by euchromatic Hairy/E(Spl) bHLH repressors in
segmentation and sex determination. Cell, 2002. 109(4): p. 447-58.

Zhang, H. and M. Levine, Groucho and dCtBP mediate separate pathways of
transcriptional repression in the Drosophila embryo. Proc Natl Acad Sci U S A,
1999. 96(2): p. 535-40.

Nibu, Y., H. Zhang, and M. Levine, Interaction of short-range repressors with
Drosophila CtBP in the embryo. Science, 1998. 280(5360): p. 101-4.

Nibu, Y., et al., dCtBP mediates transcriptional repression by Knirps, Kruppel
and Snail in the Drosophila embryo. Embo J, 1998. 17(23): p. 7009-20.

Bianchi-Frias, D., et al., Hairy transcriptional repression targets and cofactor
recruitment in Drosophila. PLoS Biol, 2004. 2(7): p. E178.

Takata, T. and F. Ishikawa, Human Sir2-related protein SIRT] associates with the
bHLH repressors HES] and HE Y2 and is involved in HESI- and HE Y2-mediated
transcriptional repression. Biochem Biophys Res Commun, 2003. 301(1): p. 250-
7.

Shore, D., M. Squire, and K.A. Nasmyth, Characterization of two genes required
for the position-effect control of yeast mating-type genes. Embo J, 1984. 3(12): p.
2817-23.

Moazed, D., Common themes in mechanisms of gene silencing. Mol Cell, 2001.
8(3): p. 489-98.

116

 

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

Rusche, L.N., A.L. Kirchmaier, and J. Rine, Ordered nucleation and spreading of
silenced chromatin in Saccharomyces cerevisiae. Mol Biol Cell, 2002. 13(7): p.
2207-22.

Moazed, D., et al., A model for step-wise assembly of heterochromatin in yeast.
Novartis Found Symp, 2004. 259: p. 48-56; discussion 56-62, 163-9.

Jennings, B.H., et al., Molecular recognition of transcriptional repressor motifs
by the WD domain of the Groucho/TLE corepressor. Mol Cell, 2006. 22(5): p.
645-55.

Copley, R.R., The EH] motif in metazoan transcription factors. BMC Genomics,
2005.6:p.169.

Neer, E.J., et al., The ancient regulatory-protein family of WD-repeat proteins.
Nature, 1994. 371(6495): p. 297-300.

Chen, G., et al., A functional interaction between the histone deacetylase de3
and the corepressor groucho in Drosophila development. Genes Dev, 1999.

13(17): p. 2218-30.

Flores-Saaib, RD. and A.J. Courey, Analysis of Groucho-histone interactions
suggests mechanistic similarities between Groucho- and T up] -mediated
repression. Nucleic Acids Res, 2000. 28(21): p. 4189-96.

Choi, C.Y., et al., Phosphorylation by the DHIPKZ protein kinase modulates the
corepressor activity of Groucho. J Biol Chem, 2005. 280(22): p. 21427-36.

Nuthall, H.N., K. Joachim, and S. Stifani, Phosphorylation of serine 239 of
Groucho/TLE] by protein kinase CK2 is important for inhibition of neuronal
diflerentiation. Mol Cell Biol, 2004. 24(19): p. 8395-407.

Van Doren, M., H.M. Ellis, and J.W. Posakony, The Drosophila
extramacrochaetae protein antagonizes sequence-specific DNA binding by

daughterless/achaete—scute protein complexes. Development, 1991. 113(1): p.
245-55.

Akazawa, C., et al., Molecular characterization of a rat negative regulator with a
basic helix-loop-helix structure predominantly expressed in the developing
nervous system. J Biol Chem, 1992. 267(30): p. 21879-85.

Sun, J ., et al., Regulation of myogenic terminal difl'erentiation by the hairy-related
transcription factor CHF2. J Biol Chem, 2001. 276(21): p. 18591-6.

Chin, M.T., et al., Cardiovascular basic helix 100p helix factor 1, a novel

transcriptional repressor expressed preferentially in the developing and adult
cardiovascular system. J Biol Chem, 2000. 275(9): p. 6381-7.

117

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

Ghosh, B. and SD. Leach, Interactions between hairy/enhancer of split-related
proteins and the pancreatic transcription factor Pth-p48 modulate function of the
PT F I transcriptional complex. Biochem J, 2006. 393(Pt 3): p. 679-85.

Jan, Y.N. and L.Y. Jan, Functional gene cassettes in development. Proc Natl Acad
Sci U S A, 1993. 90(18): p. 8305-7.

Hartenstein, V. and J .W. Posakony, Development of adult sensilla on the wing
and notum of Drosophila melanogaster. Development, 1989. 107(2): p. 389-405.

Campos-Ortega, J .A., Genetic mechanisms of early neurogenesis in Drosophila
melanogaster. Mol Neurobiol, 1995. 10(2-3): p. 75-89.

Jarrnan, A.P., et al., Role of the proneural gene, atonal, in formation of
Drosophila chordotonal organs and photoreceptors. Development, 1995. 121(7):
p. 2019-30.

Caudy, M., et al., daughterless, a Drosophila gene essential for both neurogenesis
and sex determination, has sequence similarities to myc and the achaete-scute
complex. Cell, 1988. 55(6): p. 1061-7.

Rushlow, C.A., et al., The Drosophila hairy protein acts in both segmentation and
bristle patterning and shows homology to N-myc. Embo J, 1989. 8(10): p. 3095-
103.

Nakao, K. and J.A. Campos-Ortega, Persistent expression of genes of the
enhancer of split complex suppresses neural development in Drosophila. Neuron,
1996. 16(2): p. 275-86.

Bier, E., et al., deadpan, an essential pan—neural gene in Drosophila, encodes a
helix-loop-helix protein similar to the hairy gene product. Genes Dev, 1992.
6(11): p. 2137-51.

Cubas, P., et al., Proneural clusters of achaete-scute expression and the
generation of sensory organs in the Drosophila imaginal wing disc. Genes Dev,
1991. 5(6): p. 996-1008.

Simpson, P., Lateral inhibition and the development of the sensory bristles of the
adult peripheral nervous system of Drosophila. Development, 1990. 109(3): p.
509-19.

Tomlinson, A. and DR Ready, Neuronal Diflerentiation in the Drosophila
ommatidium. Dev. Biol., 1987. 120: p. 366-376.

Greenwood, S. and G. Struhl, Progression of the morphogenetic furrow in the
Drosophila eye: the roles of Hedgehog, Decapentaplegic and the Raf pathway.
Development, 1999. 126(24): p. 5795-808.

118

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

Baylies, MK. and M. Bate, twist: a myogenic switch in Drosophila. Science,
1996. 272(5267): p. 1481-4.

Castanon, I. and MK. Baylies, A Twist in fate: evolutionary comparison of Twist
structure and function. Gene, 2002. 287(1-2): p. 11-22.

Orenic, T.V., et al., The spatial organization of epidermal structures: hairy
establishes the geometrical pattern of Drosophila leg bristles by delimiting the
domains of achaete expression. Development, 1993. 118( 1): p. 9-20.

Tapanes-Castillo, A. and MK. Baylies, Notch signaling patterns Drosophila
mesodermal segments by regulating the bHLH transcription factor twist.
Development, 2004. 131(10): p. 2359-72.

Carmena, A., M. Bate, and F. Jimenez, Lethal of scute, a proneural gene,
participates in the specification of muscle progenitors during Drosophila
embryogenesis. Genes Dev, 1995. 9(19): p. 2373-83.

Giebel, B., The notch signaling pathway is required to specify muscle progenitor
cells in Drosophila. Mech Dev, 1999. 86(1-2): p. 137-45.

Myat, M.M., Making tubes in the Drosophila embryo. Dev Dyn, 2005. 232(3): p.
617-32.

Hoch, M., et al., Sequential fates in a single cell are established by the neurogenic
cascade in the Malpighian tubules of Drosophila. Development, 1994. 120(12): p.
3439-50.

Parkhurst, S.M., D. Bopp, and D. Ish-Horowicz, X:A ratio, the primary sex-
determining signal in Drosophila, is transduced by helix-loop-helix proteins. Cell,
1990. 63(6): p. 1179-91.

Myat, M.M. and DJ. Andrew, Epithelial tube morphology is determined by the
polarized growth and delivery of apical membrane. Cell, 2002. 111(6): p. 879-91.

Ohtsuka, T., et al., Hes] and Hes5 as notch eflectors in mammalian neuronal
difi’erentiation. Embo J, 1999. 18(8): p. 2196-207.

Ishibashi, M., et al., Targeted disruption of mammalian hairy and Enhancer of
split homolog-I (HES-1) leads to up-regulation of neural helix-loop-helix factors,
premature neurogenesis, and severe neural tube defects. Genes Dev, 1995. 9(24):
p. 3136-48.

Cau, E., et al., Hes genes regulate sequential stages of neurogenesis in the
olfactory epithelium. Development, 2000. 127(11): p. 2323-32.

Takke, C., et al., her4, a zebrafish homologue of the Drosophila neurogenic gene
E(spl), is a target of NOT CH signalling. Development, 1999. 126(9): p. 1811-21.

119

 

119.

120.

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

Schneider, M.L., D.L. Turner, and ML. Vetter, Notch signaling can inhibit Xath5
function in the neural plate and developing retina. Mol Cell Neurosci, 2001.
18(5): p. 458-72.

Hojo, M., et al., Glial cell fate specification modulated by the bHLH gene Hes5 in
mouse retina. Development, 2000. 127(12): p. 2515-22.

Furukawa, T., et al., rax, Hes], and notch] promote the formation of Muller glia
by postnatal retinal progenitor cells. Neuron, 2000. 26(2): p. 383-94.

Satow, T., et al., The basic helix-loop-helix gene hesr2 promotes gliogenesis in
mouse retina. J Neurosci, 2001. 21(4): p. 1265-73.

Bae, S., et al., The bHLH gene Hes6, an inhibitor of Hes], promotes neuronal
diflerentiation. Development, 2000. 127(13): p. 2933-43.

Kageyarna, R., T. Ohtsuka, and T. Kobayashi, The Hes gene family: repressors
and oscillators that orchestrate embryogenesis. Development, 2007. 134(7): p.
1243-51.

Ohtsuka, T., et al., Roles of the basic helix-loop-helix genes Hes] and Hes5 in
expansion of neural stem cells of the developing brain. J Biol Chem, 2001.
276(32): p. 30467-74.

Hatakeyarna, J ., et al., Hes genes regulate size, shape and histogenesis of the
nervous system by control of the timing of neural stem cell differentiation.
Development, 2004. 131(22): p. 5539-50.

Baek, J.H., et al., Persistent and high levels of Hes] expression regulate boundary
formation in the developing central nervous system. Development, 2006. 133(13):
p. 2467-76.

Lamont, RE. and S. Childs, MAPping out arteries and veins. Sci STKE, 2006.
2006(355): p. pe39.

Zhong, T.P., et al., gridlock, an HLH gene required for assembly of the aorta in
zebrafish. Science, 2000. 287(5459): p. 1820-4.

Kokubo, H., et al., Mouse hesr] and heer genes are redundantly required to
mediate Notch signaling in the developing cardiovascular system. Dev Biol,
2005. 278(2): p. 301-9.

Fischer, A., et al., The Notch target genes Hey] and Hey2 are required for
embryonic vascular development. Genes Dev, 2004. 18(8): p. 901-1].

Henderson, A.M., et al., The basic helix-loop-helix transcription factor HESR]
regulates endothelial cell tube formation. J Biol Chem, 2001. 276(9): p. 6169-76.

120

 

133.

134.

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

Lawson, N.D., et al., Notch signaling is required for arterial-venous
difi’erentiation during embryonic vascular development. Development, 2001.
128(19): p. 3675-83.

Rutenberg, J.B., et al., Developmental patterning of the cardiac atrioventricular
canal by Notch and Hairy-related transcription factors. Development, 2006.
133(21): p. 4381-90.

Kokubo, H., et al., Hesr] and Hesr2 regulate atrioventricular boundary
formation in the developing heart through the repression of Tbe. Development,
2007. 134(4): p. 747-55.

Fischer, A., et al., Combined loss of Hey] and HeyL causes congenital heart
defects because of impaired epithelial to mesenchymal transition. Circ Res, 2007.
100(6): p. 856-63.

Garg, V., et al., Mutations in NOT CH1 cause aortic valve disease. Nature, 2005.
437(7056): p. 270-4.

Palmeirim, I., et al., Avian hairy gene expression identifies a molecular clock
linked to vertebrate segmentation and somitogenesis. Cell, 1997. 91(5): p. 639-48.

Lawrence, RA. and P. Johnston, Pattern formation in the Drosophila embryo:
allocation of cells to parasegments by even-skipped and fushi tarazu.
Development, 1989. 105(4): p. 761-7.

Giudicelli, F. and J. Lewis, The vertebrate segmentation clock. Curr Opin Genet
Dev, 2004. 14(4): p. 407-14.

Jouve, C., et al., Notch signalling is required for cyclic expression of the hairy-
like gene HES] in the presomitic mesoderm. Development, 2000. 127(7): p. 1421-
9.

Leimeister, C., et al., Oscillating expression of c-Hey2 in the presomitic
mesoderm suggests that the segmentation clock may use combinatorial signaling
through multiple interacting bHLH factors. Dev Biol, 2000. 227(1): p. 91-103.

Bessho, Y., et al., Dynamic expression and essential functions of Hes7 in somite
segmentation. Genes Dev, 2001. 15(20): p. 2642-7.

Holley, S.A., R. Geisler, and C. Nusslein-Volhard, Control of her] expression
during zebrafish somitogenesis by a delta-dependent oscillator and an
independent wave-front activity. Genes Dev, 2000. 14(13): p. 1678-90.

Bessho, Y. and R. Kageyama, Oscillations, clocks and segmentation. Curr Opin
Genet Dev, 2003. 13(4): p. 379-84.

121

 

146.

147.

148.

149.

150.

151.

152.

153.

154.

155.

156.

157.

158.

159.

Hirata, H., et al., Instability of Hes7 protein is crucial for the somite segmentation
clock. Nat Genet, 2004. 36(7): p. 750-4.

Rida, P.C., N. Le Minh, and Y.J. Jiang, A Notch feeling of somite segmentation
and beyond. Dev Biol, 2004. 265(1): p. 2-22.

Umbhauer, M., J .C. Boucaut, and D.L. Shi, Repression of MyoD expression and
myogenesis by Xhairy-I in Xenopus early embryo. Mech Dev, 2001. 109(1): p.
61-8.

Gao, X., et al., HES6 acts as a transcriptional repressor in myoblasts and can
induce the myogenic dijferentiation program. J Cell Biol, 2001. 154(6): p. 1161-
71.

Kuroda, K., et al., Delta-induced Notch signaling mediated by RBP-J inhibits
MyoD expression and myogenesis. J Biol Chem, 1999. 274(11): p. 7238-44.

Hirsinger, E., et al., Notch signalling acts in postmitotic avian myogenic cells to
control MyoD activation. Development, 2001. 128(1): p. 107-16.

Kopan, R., J .S. Nye, and H. Weintraub, The intracellular domain of mouse Notch:
a constitutively activated repressor of myogenesis directed at the basic helix-loop-
helix region of MyoD. Development, 1994. 120(9): p. 2385-96.

Nye, J.S., R. Kopan, and R. Axel, An activated Notch suppresses neurogenesis
and myogenesis but not gliogenesis in mammalian cells. Development, 1994.
120(9): p. 2421-30.

Tomita, K., et al., The bHLH gene Hes] is essential for expansion of early T cell
precursors. Genes Dev, 1999. 13(9): p. 1203-10.

Kim, H.K. and G. Siu, The notch pathway intermediate HES-I silences CD4 gene
expression. Mol Cell Biol, 1998. 18(12): p. 7166-75.

Siu, G., et al., A transcriptional silencer controls the developmental expression of
the CD4 gene. Embo J, 1994. 13(15): p. 3570-9.

Murata, K., et al., Hes] directly controls cell proliferation through the
transcriptional repression of p2 7Kip] . Mol Cell Biol, 2005. 25(10): p. 4262-71.

Sumazaki, R., et al., Conversion of biliary system to pancreatic tissue in Hes] -
deficient mice. Nat Genet, 2004. 36(1): p. 83-7.

Fukuda, A., et al., Ectopic pancreas formation in Hes] -knockout mice reveals
plasticity of endodermal progenitors of the gut, bile duct, and pancreas. J Clin
Invest, 2006. 116(6): p. 1484-93.

122

 

160.

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

171.

172.

Jensen, J ., et al., Control of endodermal endocrine development by Hes-1. Nat
Genet, 2000. 24(1): p. 36-44.

Yang, Q., et al., Requirement of Math] for secretory cell lineage commitment in
the mouse intestine. Science, 2001. 294(5549): p. 2155-8.

Fre, S., et al., Notch signals control the fate of immature progenitor cells in the
intestine. Nature, 2005. 435(7044): p. 964-8.

Suzuki, K., et al., Hes] -deficient mice show precocious drfi'erentiation of Paneth
cells in the small intestine. Biochem Biophys Res Commun, 2005 . 328(1): p. 348-
52.

Stanger, B.Z., et al., Direct regulation of intestinal fate by Notch. Proc Natl Acad
Sci U S A, 2005. 102(35): p. 12443-8.

Kodama, Y., et al., The role of notch signaling in the development of intrahepatic
bile ducts. Gastroenterology, 2004. 127(6): p. 1775-86.

Axelson, H., The Notch signaling cascade in neuroblastoma: role of the basic
helix-loop-helix proteins HASH-1 and HES-I. Cancer Lett, 2004. 204(2): p. 171-
8.

Kunnimalaiyaan, M., K. Traeger, and H. Chen, Conservation of the Notch]
signaling pathway in gastrointestinal carcinoid cells. Am J Physiol Gastrointest
Liver Physiol, 2005. 289(4): p. G636-42.

Kunnimalaiyaan, M., et al., Hairy Enhancer of Split-I (HES-1), a Notch] eflector,
inhibits the growth of carcinoid tumor cells. Surgery, 2005. 138(6): p. 1137-42;
discussion 1142.

Sriuranpong, V., et al., Notch signaling induces cell cycle arrest in small cell lung
cancer cells. Cancer Res, 2001. 61(7): p. 3200-5.

Chen, H., et al., Conservation of the Drosophila lateral inhibition pathway in
human lung cancer: a hairy-related protein (HES-1) directly represses achaete-
scute homolog-1 expression. Proc Natl Acad Sci U S A, 1997. 94(10): p. 5355-60.

Ellisen, L.W., et al., TAN-I, the human homolog of the Drosophila notch gene, is
broken by chromosomal translocations in T lymphoblastic neoplasms. Cell, 1991.
66(4): p. 649-6].

Shimizu, D., et al., Detection of NOT CH] mutations in adult T -cell
leukemia/lymphoma and peripheral T -cell lymphoma. Int J Hematol, 2007. 85(3):
p. 212-8.

123

173.

174.

175.

176.

177.

178.

179.

180.

181.

182.

183.

184.

185.

186.

187.

Lee, J .S., et al., Murine leukemia provirus-mediated activation of the Notch] gene
leads to induction of HES-I in a mouse T lymphoma cell line, DL-3. FEBS Lett,
1999. 455(3): p. 276-80.

Durnortier, A., et al., Notch activation is an early and critical event during T -Cell
leukemogenesis in Ikaros-deficient mice. Mol Cell Biol, 2006. 26(1): p. 209-20.

van Es, J .H., et al., Notch/gamma-secretase inhibition turns proliferative cells in
intestinal crypts and adenomas into goblet cells. Nature, 2005. 435(7044): p. 959-
63.

Lohr, D., P. Venkov, and J. Zlatanova, Transcriptional regulation in the yeast
GAL gene family: a complex genetic network Faseb J, 1995. 9(9): p. 777-87.

Duffy, J .B., GAL4 system in Drosophila: a fly geneticist's Swiss army knife.
Genesis, 2002. 34(1-2): p. 1-15.

Marrnorstein, R., et al., DNA recognition by GAL4: structure of a protein-DNA
complex. Nature, 1992. 356(6368): p. 408-14.

Giniger, E. and M. Ptashne, Cooperative DNA binding of the yeast transcriptional
activator GAL4. Proc Natl Acad Sci U S A, 1988. 85(2): p. 382-6.

Traven, A., B. Jelicic, and M. Sopta, Yeast Gal4.: a transcriptional paradigm
revisited. EMBO Rep, 2006. 7(5): p. 496-9.

Wu, Y., R]. Reece, and M. Ptashne, Quantitation of putative activator-target

aflinities predicts transcriptional activating potentials. Embo J, 1996. 15(15): p.
3951-63.

Melcher, K. and SA. Johnston, GAL4 interacts with TA TA-binding protein and
coactivators. Mol Cell Biol, 1995. 15(5): p. 2839-48.

Ansari, A.Z., R.J. Reece, and M. Ptashne, A transcriptional activating region with
two contrasting modes of protein interaction. Proc Natl Acad Sci U S A, 1998.
95(23): p. 13543-8.

Koh, S.S., et al., An activator target in the RNA polymerase II holoenzyme. Mol
Cell, 1998. 1(6): p. 895-904.

Jeong, C.J., et al., Evidence that Gal] 1 protein is a target of the Gal4 activation
domain in the mediator. Biochemistry, 2001. 40(31): p. 9421-7.

Bhaumik, SR. and MR. Green, SAGA is an essential in vivo target of the yeast
acidic activator Gal4p. Genes Dev, 2001. 15( 15): p. 1935-45.

Ferreira, M.E., et al., Mechanism of transcription factor recruitment by acidic
activators. J Biol Chem, 2005. 280(23): p. 21779-84.

124

 

188.

189.

190.

191.

192.

193.

194.

195.

196.

197.

198.

199.

200.

201.

Chang, C., et al., The Gal4 activation domain binds Sug2 protein, a proteosome
component, in vivo and in vitro. J Biol Chem, 2001. 276(33): p. 30956-63.

Gonzalez, F., et al., Recruitment of a 19S proteosome subcomplex to an activated
promoter. Science, 2002. 296(5567): p. 548-50.

Lum, J .K. and AK. Mapp, Artificial transcriptional activation domains.
Chembiochem, 2005. 6(8): p. 131 1-5.

Hollenberg, SM. and RM. Evans, Multiple and cooperative trans-activation
domains of the human glucocorticoid receptor. Cell, 1988. 55(5): p. 899-906.

Neely, K.E., et al., Activation domain-mediated targeting of the S WI/SNF complex

to promoters stimulates transcription from nucleosome arrays. Mol Cell, 1999.
4(4): p. 649-55.

Schaeffer, V., et al., Bicoid functions without its TA T A-binding protein-associated
factor interaction domains. Proc Natl Acad Sci U S A, 1999. 96(8): p. 4461-6.

Driever, W., et al., Rescue of bicoid mutant Drosophila embryos by bicoid fusion
proteins containing heterologous activating sequences. Nature, 1989. 342(6246):
p. 149-54.

Grant, RA, et al., Yeast Gcn5 fimctions in two multisubunit complexes to
acetylate nucleosomal histones: characterization of an Ada complex and the
SAGA (Spt/Ada) complex. Genes Dev, 1997. 11(13): p. 1640-50.

Sterner, D.E., et al., Functional organization of the yeast SA GA complex: distinct
components involved in structural integrity, nucleosome acetylation, and TATA-
binding protein interaction. Mol Cell Biol, 1999. 19(1): p. 86-98.

Candau, R., et al., Histone acetyltransferase activity and interaction with ADA2
are critical for GCN5 function in vivo. Embo J, 1997. 16(3): p. 555-65.

Barlev, N.A., et al., Characterization of physical interactions of the putative
transcriptional adaptor, ADA2, with acidic activation domains and TA TA-binding
protein. J Biol Chem, 1995. 270(33): p. 19337-44.

Saleh, A., et al., Identification of native complexes containing the yeast
coactivator/repressor proteins NGGI/ADA3 and ADA2. J Biol Chem, 1997.
272(9): p. 5571-8.

Grant, RA, et al., The AT M-related cofactor T ra] is a component of the purified
SAGA complex. Mol Cell, 1998. 2(6): p. 863-7.

Allard, S., et al., NuA4, an essential transcription adaptor/histone H4

acetyltransferase complex containing Esa] p and the AT M-related cofactor T ra] p.
Embo J, 1999. 18(18): p. 5108-19.

125

 

202.

203.

204.

205.

206.

207.

208.

209.

210.

211.

212.

213.

214.

215.

Henry, K.W., et al., Transcriptional activation via sequential histone HZB
ubiquitylation and deubiquitylation, mediated by SAGA-associated pr8. Genes
Dev, 2003. 17(21): p. 2648-63.

Bryant, GO. and M. Ptashne, Independent recruitment in vivo by Gal4 of two
complexes required for transcription. Mol Cell, 2003. 11(5): p. 1301-9.

Bhaumik, S.R., et al., In vivo target of a transcriptional activator revealed by

fluorescence resonance energy transfer. Genes Dev, 2004. 18(3): p. 333-43.

Barbaric, S., H. Reinke, and W. Horz, Multiple mechanistically distinct functions
of SAGA at the PHO5 promoter. Mol Cell Biol, 2003. 23(10): p. 3468-76.

Gregory, P.D., et al., Chromatin remodelling at the PH08 promoter requires
S WI-SNF and SAGA at a step subsequent to activator binding. Embo J, 1999.
18(22): p. 6407-14.

Ornaghi, P., et al., The bromodomain of Gcn5p interacts in vitro with specific
residues in the N terminus of histone H4. J Mol Biol, 1999. 287(1): p. 1-7.

Kusch, T., et al., Two Drosophila Ada2 homologues function in different
multiprotein complexes. Mol Cell Biol, 2003. 23(9): p. 3305-19.

Peterson, CL. and I. Herskowitz,Characterization of the yeast SWI], SW12, and
S W13 genes, which encode a global activator of transcription. Cell, 1992. 68(3):
p. 573-83.

Gutierrez, J .L., et al., Activation domains drive nucleosome eviction by S WI/SNF .
Embo J, 2007. 26(3): p. 730-40.

Mitra, D., et al., S WI/SNF binding to the HO promoter requires histone
acetylation and stimulates TATA-binding protein recruitment. Mol Cell Biol,
2006. 26(11): p. 4095-110.

Lohr, D., Nucleosome transactions on the promoters of the yeast GAL and PHO
genes. J Biol Chem, 1997. 272(43): p. 26795-8.

Dhasarathy, A. and MP. Kladde, Promoter occupancy is a major determinant of
chromatin remodeling enzyme requirements. Mol Cell Biol, 2005. 25(7): p. 2698-
707.

Sudarsanam, P. and F. Winston, The Swi/Snf family nucleosome-remodeling
complexes and transcriptional control. Trends Genet, 2000. 16(8): p. 345-5].

Kao, C.F., et al., Rad6 plays a role in transcriptional activation through
ubiquitylation of histone H28. Genes Dev, 2004. 18(2): p. 184-95.

126

216.

217.

218.

219.

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

Dover, J ., et al., Methylation of histone H3 by COMPASS requires ubiquitination
of histone H2B by Rad6. J Biol Chem, 2002. 277(32): p. 28368-71.

Briggs, S.D., et al., Gene silencing: trans-histone regulatory pathway in
chromatin. Nature, 2002. 418(6897): p. 498.

Xiao, T., et al., Histone HZB ubiquitylation is associated with elongating RNA
polymerase 1]. Mol Cell Biol, 2005. 25(2): p. 637-5].

Wood, A., et al., The Pafl complex is essential for histone monoubiquitination by
the Rad6—Bre] complex, which signals for histone methylation by COMPASS and
Dot1p. J Biol Chem, 2003. 278(37): p. 34739-42.

Krogan, N.J., et al., The Paf] complex is required for histone H3 methylation by
COMPASS and Dotlp: linking transcriptional elongation to histone methylation.
Mol Cell, 2003. 11(3): p. 721-9.

Shahbazian, M.D., K. Zhang, and M. Grunstein, Histone H23 ubiquitylation
controls processive methylation but not monomethylation by Dot] and Set]. Mol
Cell, 2005. 19(2): p. 271-7.

Wood, A., et al., The BurI/Bur2 complex is required for histone H23
monoubiquitination by Rad6/Ere] and histone methylation by COMPASS. Mol
Cell, 2005. 20(4): p. 589-99.

Keogh, M.C., V. Podolny, and S. Buratowski, Bur] kinase is required for efficient
transcription elongation by RNA polymerase 11. Mol Cell Biol, 2003. 23(19): p.
7005-] 8.

Pray-Grant, M.G., et al., Chd] chromodomain links histone H3 methylation with
SA GA- and SLIK-dependent acetylation. Nature, 2005. 433(7024): p. 434-8.

Santos-Rosa, H., et al., Methylation of histone H3 K4 mediates association of the
Isw1p AT Pase with chromatin. Mol Cell, 2003. 12(5): p. 1325-32.

Ingham, P., Abdominal gene organization. Nature, 1985. 313(5998): p. 98-9.

Chen, G. and A.J. Courey, Groucho/TLE family proteins and transcriptional
repression. Gene, 2000. 249(1-2): p. 1-16.

Wharton, K.A. and ST. Crews, CNS midline enhancers of the Drosophila slit and
T 011 genes. Mech. Dev., 1993. 40(3): p. 141-154.

Kulkarni, M.M. and D.N. Amosti, cis-regulatory logic of short-range

transcriptional repression in Drosophila melanogaster. Mol Cell Biol, 2005.
25(9): p. 3411-20.

127

 

230.

231.

232.

233.

234.

235.

236.

237.

238.

239.

240.

241.

242.

Kulkarni, M.M., The function and design of cis-acting enhancer elements
regulated by short-range transcriptional repressors: grammar studies from
Drosophila melanogaster. Dissertation for the degree of PhD. Michigan State
University, 2003.

Small, 8., A. Blair, and M. Levine, Regulation of even-skipped stripe 2 in the
Drosophila embryo. Embo J, 1992. 11(11): p. 4047-57.

Pirrotta, V., Vectors for P-mediated transformation in Drosophila.
Biotechnology, 1988. 10: p. 437-56.

Stebbins, J .L. and SJ. Triezenberg, Identification, mutational analysis, and
coactivator requirements of two distinct transcriptional activation domains of the
Saccharomyces cerevisiae Hap4 protein. Eukaryot Cell, 2004. 3(2): p. 339-47.

Larschan, E. and F. Winston, The S. cerevisiae SA GA complex functions in vivo as
a coactivator for transcriptional activation by Gal4. Genes Dev, 2001. 15(15): p.
1946-56.

Nalley, K., S.A. Johnston, and T. Kodadek, Proteolytic turnover of the Gal4
transcription factor is not required for function in vivo. Nature, 2006. 442(7106):
p. 1054-7.

Lee, M., S. Chatterjee, and K. Struhl, Genetic analysis of the role of Pol 11
holoenzyme components in repression by the Cyc8-T up] corepressor in yeast.
Genetics, 2000. 155(4): p. 1535-42.

Green, SR. and AD. Johnson, Genome-wide analysis of the functions of a
conserved surface on the corepressor T up] . Mol Biol Cell, 2005. 16(6): p. 2605-
13.

Hallberg, M., et al., Functional and physical interactions within the middle
domain of the yeast mediator. Mol Genet Genomics, 2006. 276(2): p. 197-210.

Zeitlinger, J ., et al., Whole-genome ChIP-chip analysis of Dorsal, Twist, and Snail
suggests integration of diverse patterning processes in the Drosophila embryo.
Genes Dev, 2007. 21(4): p. 385-90.

Small, 8., D.N. Amosti, and M. Levine, Spacing ensures autonomous expression
of different stripe enhancers in the even-skipped promoter. Development, 1993.
119(3): p. 762-72.

Jiang, J ., et al., Conversion of a dorsal-dependent silencer into an enhancer:
evidence for dorsal corepressors. Embo J, 1993. 12(8): p. 3201-9.

Kurdistani, SK. and M. Grunstein, In vivo protein-protein and protein-DNA
crosslinking for genomewide binding microarray. Methods, 2003. 31(1): p. 90-5.

128

243.

244.

245.

246.

247.

248.

249.

250.

251.

252.

253.

254.

255.

256.

Zeng, P.Y., et al., In vivo dual cross-linking for identification of indirect DNA-

associated proteins by chromatin immunoprecipitation. Biotechniques, 2006.
41(6): p. 694, 696, 698.

Song, H., et al., Groucho oligomerization is required for repression in vivo. Mol
Cell Biol, 2004. 24(10): p. 4341-50.

Davie, J .K., et al., T up] -Ssn6 interacts with multiple class I histone deacetylases
in vivo. J Biol Chem, 2003. 278(50): p. 50158-62.

Huang, L., W. Zhang, and S.Y. Roth, Amino termini of histones H3 and H4 are
required for aI-alphaZ repression in yeast. Mol Cell Biol, 1997. 17(11): p. 6555-
62.

Watson, A.D., et al., Ssn6-T up] interacts with class I histone deacetylases
required for repression. Genes Dev, 2000. 14(21): p. 2737-44.~

Tripic, T., et al., The Set2 methyltransferase associates with Ssn6 yet T up] -Ssn6
repression is independent of histone methylation. Biochem Biophys Res
Commun, 2006. 339(3): p. 905-14.

Ducker, CE. and RT. Simpson, The organized chromatin domain of the
repressed yeast a cell-specific gene ST E6 contains two molecules of the
corepressor T up] p per nucleosome. Embo J, 2000. 19(3): p. 400-9.

Wu, J ., et al., TUPI utilizes histone H3/HZB-specific HDA] deacetylase to
repress gene activity in yeast. Mol Cell, 2001. 7(1): p. 117-26.

Davie, J .K., R.J. Trumbly, and S.Y. Dent, Histone-dependent association of T up] -
Ssn6 with repressed genes in vivo. Mol Cell Biol, 2002. 22(3): p. 693-703.

Muller, J. and J .A. Kassis, Polycomb response elements and targeting of
Polycomb group proteins in Drosophila. Curr Opin Genet Dev, 2006. 16(5): p.
476-84.

Schwartz, Y.B. and V. Pirrotta, Polycomb silencing mechanisms and the
management of genomic programmes. Nat Rev Genet, 2007. 8(1): p. 9-22.

Sekinger, EA. and D.S. Gross, SIR repression of a yeast heat shock gene: UAS
and TATA footprints persist within heterochromatin. Embo J, 1999. 18(24): p.
7041-55.

Sekinger, EA. and D.S. Gross, Silenced chromatin is permissive to activator
binding and PIC recruitment. Cell, 2001. 105(3): p. 403-14.

King, I.F., N.J. Francis, and RE. Kingston, Native and recombinant polycomb
group complexes establish a selective block to template accessibility to repress
transcription in vitro. Mol Cell Biol, 2002. 22(22): p. 7919-28.

129

 

257.

258.

259.

260.

261.

262.

263.

264.

265.

266.

267.

268.

269.

Gilmour, D.S. and J .T. Lis, RNA polymerase II interacts with the promoter region
of the noninduced hsp70 gene in Drosophila melanogaster cells. Mol Cell Biol,
1986. 6(11): p. 3984-9.

Wang, X., et al., Transcription elongation controls cell fate specification in the
Drosophila embryo. Genes Dev, 2007. 21(9): p. 1031—6.

Beuchle, D., G. Struhl, and J. Muller, Polycomb group proteins and heritable
silencing of Drosophila Hox genes. Development, 2001. 128(6): p. 993-1004.

Fischer, A. and M. Gessler, Delta-Notch--and then? Protein interactions and
proposed modes of repression by Hes and Hey bHLH factors. Nucleic Acids Res,
2007.

180, T., L. Kedes, and Y. Harnamori, HES and HERP families: multiple eflectors
of the Notch signaling pathway. J Cell Physiol, 2003. 194(3): p. 237-55.

Joshi, M., et al., Delta and Hairy establish a periodic prepattem that positions
sensory bristles in Drosophila legs. Dev Biol, 2006. 293(1): p. 64-76.

Geling, A., et al., Her5 acts as a prepattem factor that blocks neurogenin] and

cer expression upstream of Notch to inhibit neurogenesis at the midbrain-
hindbrain boundary. Development, 2004. 131(9): p. 1993-2006.

Stathopoulos, A. and M. Levine, Linear signaling in the T oll-Dorsal pathway of
Drosophila: activated Pelle kinase specifies all threshold outputs of gene
expression while the bHLH protein Twist specifies a subset. Development, 2002.
129(14): p. 3411-9.

Aronson, B.D., et al., Groucho-dependent and -independent repression activities
of Runt domain proteins. Mol Cell Biol, 1997. 17(9): p. 5581-7.

Edmondson, D.G., M.M. Smith, and S.Y. Roth, Repression domain of the yeast
global repressor T up] interacts directly with histones H3 and H4. Genes Dev,
1996. 10(10): p. 1247-59.

Euskirchen, G.M., et al., Mapping of transcription factor binding regions in
mammalian cells by ChIP: comparison of array- and sequencing-based
technologies. Genome Res, 2007. 17(6): p. 898-909.

Roseman, R.R., V. Pirrotta, and PK. Geyer, The su(Hw) protein insulates
expression of the Drosophila melanogaster white gene fi'om chromosomal
position-eflects. Embo J, 1993. 12(2): p. 435-42.

Pikaart, M.J., F. Recillas-Targa, and G. F elsenfeld, Loss of transcriptional activity

of a transgene is accompanied by DNA methylation and histone deacetylation and
is prevented by insulators. Genes Dev, 1998. 12(18): p. 2852-62.

130

 

270.

271.

272.

273.

274.

275.

276.

277.

278.

279.

280.

281.

282.

283.

Liaw, H. and A.J. Lustig, Sir3 C-terminal domain involvement in the initiation
and spreading of heterochromatin. Mol Cell Biol, 2006. 26(20): p. 7616-31.

Hoppe, G.J., et al., Steps in assembly of silent chromatin in yeast: Sir3-
independent binding of a Sir2/Sir4 complex to silencers and role for Sir2-
dependent deacetylation. Mol Cell Biol, 2002. 22(12): p. 4167-80.

Malave, T.M. and S.Y. Dent, Transcriptional repression by T up] -Ssn6. Biochem
Cell Biol, 2006. 84(4): p. 437-43.

Lee, J .S. and A. Shilatifard, A site to remember: H3K36 methylation a mark for
histone deacetylation. Mutat Res, 2007. 618(1-2): p. 130-4.

Joshi, AA. and K. Struhl, Eaf3 chromodomain interaction with methylated H3-
K36 links histone deacetylation to Pol II elongation. Mol Cell, 2005. 20(6): p.
971-8.

Li, B., et al., Combined action of PHD and chromo domains directs the de3S
HDAC to transcribed chromatin. Science, 2007. 316(5827): p. 1050-4.

Krogan, N.J., et al., A Snf2 family AT Pase complex required for recruitment of the
histone H2A variant th]. Mol Cell, 2003. 12(6): p. 1565-76.

Mizuguchi, G., et al., AT P-driven exchange of histone HZAZ variant catalyzed by
S WR] chromatin remodeling complex. Science, 2004. 303(5656): p. 343-8.

Meneghini, M.D., M. Wu, and H.D. Madharri, Conserved histone variant H2A.Z
protects euchromatin from the ectopic spread of silent heterochromatin. Cell,
2003. 112(5): p. 725-36.

Raisner, R.M., et al., Histone variant H2A.Z marks the 5 ' ends of both active and
inactive genes in euchromatin. Cell, 2005. 123(2): p. 233-48.

Zhang, H., D.N. Roberts, and B.R. Cairns, Genome-wide dynamics of thl, a
histone H2A variant that poises repressed/basal promoters for activation through
histone loss. Cell, 2005. 123(2): p. 219-31.

Gligoris, T., G. Thireos, and D. Tzamarias, The T up] corepressor directs th]
deposition at a specific promoter nucleosome marking the GAL] gene for rapid
activation. Mol Cell Biol, 2007. 27(11): p. 4198-205.

Kobor, M.S., et al., A protein complex containing the conserved Swi2/Snf2-related
AT Pase Swr] p deposits histone variant H2A.Z into euchromatin. PLoS Biol,
2004. 2(5): p. E131.

Andrioli, L.P., et al., Groucho-dependent repression by slappy-paired I
diflerentially positions anterior pair-rule stripes in the Drosophila embryo. Dev
Biol, 2004. 276(2): p. 541-5].

131

 

284.

285.

286.

287.

288.

289.

290.

291.

292.

293.

294.

295.

296.

297.

Hasson, P., et al., Brinker requires two corepressors for maximal and versatile
repression in Dpp signalling. Embo J, 2001. 20(20): p. 5725-36.

Nagel, A.C., et al., Hairless-mediated repression of notch target genes requires
the combined activity of Groucho and CtBP corepressors. Mol Cell Biol, 2005.
25(23): p. 10433-41.

Shahbazian, MD. and M. Grunstein, Functions of site-specific histone acetylation
and deacetylation. Annu Rev Biochem, 2007. 76: p. 75-100.

Suka, N., K. Luo, and M. Grunstein, Sier and SasZp opposingly regulate
acetylation of yeast histone H4 lysine] 6 and spreading of heterochromatin. Nat
Genet, 2002. 32(3): p. 378-83.

Kimura, A., T. Umehara, and M. Horikoshi, Chromosomal gradient of histone
acetylation established by Sas2p and Sier fitnctions as a shield against gene
silencing. Nat Genet, 2002. 32(3): p. 370-7.

Millar, C.B., S.K. Kurdistani, and M. Grunstein, Acetylation of yeast histone H4
lysine 16: a switch for protein interactions in heterochromatin and euchromatin.
Cold Spring Harb Symp Quant Biol, 2004. 69: p. 193-200.

Babiarz, J.E., J.B. Halley, and J. Rine, T elomeric heterochromatin boundaries
require NuA4-dependent acetylation of histone variant H2A.Z in Saccharomyces
cerevisiae. Genes Dev, 2006. 20(6): p. 700-10.

Bi, X., et al., Formation of boundaries of transcriptionally silent chromatin by
nucleosome-excluding structures. Mol Cell Biol, 2004. 24(5): p. 2118-31.

Oki, M. and RT. Kamakaka, Barrier function at HMR. Mol Cell, 2005. 19(5): p.
707-16.

Tompa, R. and H.D. Madhani, Histone H3 lysine 36 methylation antagonizes
silencing in Saccharomyces cerevisiae independently of the de3S histone
deacetylase complex. Genetics, 2007. 175(2): p. 585-93.

Zinzen, RB, et al., Computational models for neurogenic gene expression in the
Drosophila embryo. Curr Biol, 2006. 16(13): p. 1358-65.

J anssens, H., et al., Quantitative and predictive model of transcriptional control of
the Drosophila melanogaster even skipped gene. Nat Genet, 2006. 38(10): p.
1 159-65.

Zinzen, RP. and D. Papatsenko, Enhancer responses to similarly distributed
antagonistic gradients in development. PLoS Comput Biol, 2007. 3(5): p. e84.

Amosti, D.N. and M.M. Kulkarni, Transcriptional enhancers: Intelligent
enhanceosomes or flexible billboards? J Cell Biochem, 2005. 94(5): p. 890-8.

132

 

298.

299.

300.

Ludwig, M.Z., N.H. Patel, and M. Kreitrnan, Functional analysis of eve stripe 2
enhancer evolution in Drosophila: rules governing conservation and change.
Development, 1998. 125(5): p. 949-58.

Kulkarni, M.M. and D.N. Amosti, Information display by transcriptional
enhancers. Development, 2003. 130(26): p. 6569-75.

Keller, S.A., et al., dCtBP-dependent and -independent repression activities of the
Drosophila Knirps protein. Mol Cell Biol, 2000. 20(19): p. 7247-58.

133

 

              
    
   

MMMMM

lIlIlIIIIlIII/IIII

i I
03 95

     

r IIIIIIIII